UTS site search

Professor Louise Ryan

Biography

After completing her undergraduate degree in statistics and mathematics at Macquarie University, Louise Ryan left Australia in 1979 to pursue her PhD in statistics at Harvard University in the United States. In 1983, Louise then took up a postdoctoral fellowship in Biostatistics, jointly between Dana-Farber Cancer Institute and the Harvard School of Public Health.  She was promoted to Assistant Professor in 1985, eventually becoming the Henry Pickering Walcott Professor and Chair of the Department of Biostatistics at Harvard.  Louise returned to Australia in early 2009 to take up the role as Chief of CSIRO’s Division of Mathematics, Informatics and Statistics.  In 2012, she joined UTS as a distinguished professor of statistics in the School of Mathematical Sciences.  Louise is well known for her methodological contributions to statistical methods for cancer and environmental health research.  She is loves the challenge and satisfaction of multi-disciplinary collaboration.  She has received numerous prestigous awards, most recently her 2012 election to the Australian Academy of Science.

Professional

Editor in Chief of Statistics in Medicine

Vice President of the NSW Branch of the Statistical Society of Australia

Adjunct Professor of Biostatistics, Harvard School of Public Health

Image of Louise Ryan
Distinguished Professor, School of Mathematical and Physical Sciences
Core Member, AAI - Advanced Analytics Institute
BA (Hons), MA, Ph D
 
Phone
+61 2 9514 2275

Research Interests

Applied statistics, especially for applications in medicine and health (biostatistics).

Technical areas of expertise include

   *  Survival Analysis

   *  Correlated Data Analysis, including clustered data, multiple outcomes and longitudinal

   *  Quantitative Risk Assessment

   *  Meta-Analysis

  

Can supervise: Yes
General statistical methods, with an emphasis on applications. 

Books

Whitney, M. & Ryan, L. 2008, Response to Comments.
View/Download from: Publisher's site
Catalano, P.J. & Ryan, L.M. 1994, 14 Statistical methods in developmental toxicology.
View/Download from: Publisher's site
This chapter describes some of the statistical issues surrounding analysis of data from controlled developmental toxicity experiments in animals. After discussion of relevant background information, a typical experimental design is reviewed with discussion of the main statistical considerations involved in analyzing data from these experiments. Discussion follows to popular approaches for dose-response modeling, including both likelihood based and quasi-likelihood models. Attention is given to the general problem of handling multiple outcomes, a common occurrence in developmental toxicology. The issue of conducting risk assessment is then taken up by first reviewing the current method of defining allowable doses based on determining a no observed adverse effect level (NOAEL), or the dose preceding the lowest dose that differs significantly from controls. This is contrasted with more quantitative approaches such as determining a benchmark dose. Some recent advances in the area of risk assessment for multiple outcomes are also reviewed. © 1994 Elsevier Science B.V. All rights reserved.

Chapters

Whitney, M. & Ryan, L. 2008, 'Quantifying Dose-Response Uncertainty Using Bayesian Model Averaging' in Uncertainty Modeling in Dose Response: Bench Testing Environmental Toxicity, pp. 165-179.
View/Download from: Publisher's site
Lindsey, J.C. & Ryan, L.M. 2005, 'Survival Models: Methods for Interval-Censored Data' in Tutorials in Biostatistics, Statistical Methods in Clinical Studies, pp. 141-160.
View/Download from: Publisher's site

Conferences

Lovell, M., Luckett, T., Phillips, J., Agar, M., Ryan, L., Lam, L., McCaffrey, N., Boyle, F., Stubbs, J., Shaw, T., Currow, D., Hosie, A. & Davidson, P. 2015, 'Clinical Trial Protocol - Implementing Clinical Practice Guidelines For Cancer Pain In Adults To Ensure Equitable, Cost-Effective, Evidence-Based, Person-Centred Care: A Phase III Pragmatic Stepped Wedge Cluster Randomised Controlled Trial Of Guidelines And Screening With Implementation Strategies Versus Guidelines And Screening Alone To Improve Pain In Adults With Cancer Attending Outpatients Oncology And Palliative Care Centres', Asia-Pacific Journal of Clinical Oncology, pp. 162-162.
Hart, J.E., Ryan, L., Dockery, D.W., Smith, T.J., Garshick, E. & Laden, F. 2008, 'Long-Term Ambient Multipollutant Exposures and Mortality', EPIDEMIOLOGY, pp. S333-S333.
Hart, J.E., Yanosky, J.D., Puett, R.C., Ryan, L., Dockery, D.W., Smith, T.J., Garshick, E. & Laden, F. 2008, 'Spatial Modeling of Air Pollution 1985-2000 in the Continental US', EPIDEMIOLOGY, pp. S332-S333.
Morara, M., Ryan, L.M., Houseman, A. & Strauss, W. 2007, 'Optimal Design For Epidemiological Studies Subject To Designed Missingness', Lifetime Data Analysis, Conference on Statistical Methods for Emerging Issues in Observational Studies and Epidemiology, Springer, Seattle, WA, pp. 583-605.
View/Download from: Publisher's site
In large epidemiological studies, budgetary or logistical constraints will typically preclude study investigators from measuring all exposures, covariates and outcomes of interest on all study subjects. We develop a flexible theoretical framework that in
Clark, D., Lucas, F. & Ryan, L.M. 2007, 'Predicting Hospital Mortality, Length Of Stay, And Transfer To Long-term Care For Injured Patients', Journal Of Trauma-injury Infection And Critical Care, 19th Annual Meeting of the Eastern-Association-for-the-Surgery-of-Trauma, Lippincott Williams & Wilkins, Orlando, FL, pp. 592-600.
View/Download from: Publisher's site
Background: Using hospital length of stay (LOS) to measure trauma care efficiency is complicated by short LOS resulting from early mortality or transfer to long-term care (LTC). Methods: Records from the 1999 to 2003 National Trauma Data Bank were used t
McCracken, J., Schwartz, J., Mittleman, M., Ryan, L., Diaz Artiga, A. & Smith, K.R. 2007, 'Biomass smoke exposure and acute lower respiratory infections among Guatemalan children', EPIDEMIOLOGY, pp. S185-S185.
Missmer, S.A., Pearson, K.R., Ryan, L.M., Meeker, J.D., Cramer, D.W. & Hauser, R. 2007, 'Methodologic issues and statistical approaches to the analysis of multiple cycle data from couples undergoing in vitro fertilization (IVF).', FERTILITY AND STERILITY, pp. S123-S123.
View/Download from: Publisher's site
Davis, M., Smith, T., Laden, F., Hart, J., Ryan, L. & Garshick, E. 2006, 'Structural equation modeling in exposure assessment', EPIDEMIOLOGY, pp. S466-S466.
View/Download from: Publisher's site
Suglia, S.F., Ryan, L., Bellinger, D. & Wright, R.J. 2006, 'Violence exposure predicts adverse child behavior: Use of item response theory to characterize violence experience.', AMERICAN JOURNAL OF EPIDEMIOLOGY, pp. S233-S233.
Suglia, S.F., Ryan, L., Bellinger, D. & Wright, R. 2006, 'The influence of the social and physical environment on child behavior', EPIDEMIOLOGY, pp. S387-S387.
View/Download from: Publisher's site
McCracken, J., Diaz, A., Castro, E., Edwards, R., Ryan, L., Schwartz, J., Chowdhury, Z. & Smith, K.R. 2006, 'Biomass smoke exposure among Guatemalan infants participating in a randomized trial of chimney stoves', EPIDEMIOLOGY, pp. S35-S36.
View/Download from: Publisher's site
McCarty, K.M., Houseman, A., Su, L., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Smith, T., Ryan, L. & Christiani, D.C. 2005, 'Drinking water exposure to arsenic, polymorphisms in GSTT1, GSTM1 and GSTP1 and methylation capacity', EPIDEMIOLOGY, pp. S113-S113.
View/Download from: Publisher's site
McCarty, K.M., Houseman, A., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Smith, T., Ryan, L. & Christiani, D.C. 2005, 'The impact of age and gender on arsenic methylation capacity', AMERICAN JOURNAL OF EPIDEMIOLOGY, pp. S30-S30.
Meeker, J., Ryan, L., Barr, D., Herrick, R., Bennett, D. & Hauser, R. 2004, 'Contemporary use insecticides and human semen quality', EPIDEMIOLOGY, pp. S190-S190.
View/Download from: Publisher's site
Morales, K., Ibrahim, J., Ryan, L.M. & Chen, C. 2001, 'Bayesian Model Averaging With Applications To The Risk Assessment For Arsenic In Drinking Water', Arsenic Exposure And Health Effects Iv, 4th International Conference on Arsenic Exposure and Health Effects, Elsevier Science Bv, SAN DIEGO, CA, pp. 145-151.
Model selection often presents a challenge in the risk assessment process, especially when biologically based models are not apparent or fully developed. In a recent dose-response assessment for arsenic in drinking water, risk estimates were found to hig
Morales, K., Ryan, L.M., Brown, K., Kuo, T., Chen, C. & Wu, M. 1999, 'Model Sensitivity In An Analysis Of Arsenic Exposure And Bladder Cancer In Southwestern Taiwan', Arsenic Exposure And Health Effects, 3rd International Conference on Arsenic Exposure and Health Effects, Elsevier Science Bv, SAN DIEGO, CA, pp. 207-215.
View/Download from: Publisher's site
The Environmental Protection Agency (EPA) is under congressional mandate to revise its current standards for arsenic in drinking water. This chapter addresses issues surrounding model choice in a quantitative risk assessment. The data used were collected
Lin, A.E., Herring, A.H., Amstutz, K.S., Westgate, M.N., Lacro, R.V., Al-Jufan, M., Ryan, L. & Holmes, L.B. 1999, 'Cardiovascular malformations: Changes in prevalence and birth status, 1972-1990', AMERICAN JOURNAL OF MEDICAL GENETICS, pp. 102-110.
View/Download from: Publisher's site
Egan, K., Ryan, L.M. & Gragoudas, E. 1998, 'Survival Implications Of Enucleation After Definitive Radiotherapy For Choroidal Melanoma - An Example Of Regression On Time-dependent Covariates', Archives Of Ophthalmology, 67th Annual Meeting of the Association-for-Research-in-Vision-and-Ophthalmology, Amer Medical Assoc, FT LAUDERDALE, FLORIDA, pp. 366-370.
Objective: To evaluate whether the removal of the eye after radiotherapy alters the rates of metastatic death in patients with melanoma of the choroid. Patients and Methods: Using an extension of the Cox model, we based our analysis on a cohort of 1541 c
Andersen, J.W., Goetghebeur, E.J. & Ryan, L. 1996, 'Analysis of survival data under competing risks with missing cause of death information: Application and implications for study design', LIFETIME DATA: MODELS IN RELIABILITY AND SURVIVAL ANALYSIS, pp. 13-19.
Catalano, P., Ryan, L.M. & Senchaudhuri, P. 1994, 'Developmental Toxicity Modeling For Risk Assessment', American Statistical Association 1994 Proceedings Of The Section On Statistics And The Environment, Annual Meeting of the Section-on-Statistics-and-the-Environment of the American-Statistical-Association, Amer Statistical Assoc, TORONTO, CANADA, pp. 38-45.
NA
Lindsey, J.C. & Ryan, L.M. 1991, 'A comparison of continuous- and discrete- time three-state models for rodent tumorigenicity experiments', Environmental Health Perspectives, International Biostatistics Conference in the Study of Toxicology, Natl Inst Environ Health Sci, TOKYO, JAPAN, pp. 9-17.
View/Download from: UTS OPUS
The three-state illness-death model provides a useful way to characterize data from a rodent tumorigenicity experiment. Most parametrizations proposed recently in the literature assume discrete time for the death process and either discrete or continuous
Lindsay, J.C. & Ryan, L.M. 1994, 'A comparison of continuous-and discrete-time three-state models for rodent tumorigenicity experiments', Environmental Health Perspectives, pp. 9-17.
The three-state illness-death model provides a useful way to characterize data from a rodent tumorigenicity experiment. Most parametrizations proposed recently in the literature assume discrete time for the death process and either discrete or continuous time for the tumor onset process. We compare these approaches with a third alternative that uses a piecewise continuous model on the hazards for tu nor onset and death. All three models assume proportional hazards to characterize tumor lethality and the effect of dose on tumor onset and death rate. All of the models can easily be fitted using an Expectation Maximization (EM) algorithm. The piecewise continuous model is particularly appealing in this context because the complete data likelihood corresponds to a standard piecewise exponential model with tumor presence as a time-varying covariate. It can be shown analytically that differences between the parameter estimates given by each model are explained by varying assumptions about when tumor onsets, deaths, and sacrifices occur within intervals. The mixed-time model is seen to be an extension of the grouped data proportional hazards model. We argue that the continuous-time model is preferable to the discrete- and mixed-time models because it gives reasonable estimates with relatively few intervals while still making full use of the available imformation. Data from the ED 01 experiment illustrate the results.
BLUM, R.H., EDMONSON, J., RYAN, L. & PELLETIER, L. 1993, 'EFFICACY OF IFOSFAMIDE IN COMBINATION WITH DOXORUBICIN FOR THE TREATMENT OF METASTATIC SOFT-TISSUE SARCOMA', CANCER CHEMOTHERAPY AND PHARMACOLOGY, pp. S238-S240.
RYAN, L. 1993, 'POTENCY MEASURES FOR DEVELOPMENTAL TOXICITY', ENVIRONMETRICS, pp. 507-518.
View/Download from: Publisher's site
ELIAS, A., RYAN, L., AISNER, J. & ANTMAN, K.H. 1990, 'MESNA, DOXORUBICIN, IFOSFAMIDE, DACARBAZINE (MAID) REGIMEN FOR ADULTS WITH ADVANCED SARCOMA', SEMINARS IN ONCOLOGY, pp. 41-49.
ANTMAN, K.H., ELIAS, A. & RYAN, L. 1990, 'IFOSFAMIDE AND MESNA - RESPONSE AND TOXICITY AT STANDARD-DOSE AND HIGH-DOSE SCHEDULES', SEMINARS IN ONCOLOGY, pp. 68-73.
HERMAN, T.S., JOCHELSON, M.S., TEICHER, B.A., SCOTT, P.J., HANSEN, J., CLARK, J.R., PFEFFER, M.R., GELWAN, L.E., MOLNARGRIFFIN, B.J., FRASER, S.M., SVENNSON, G., BORNSTEIN, B.A., RYAN, L. & COLEMAN, C.N. 1989, 'A PHASE-I-II TRIAL OF CISPLATIN, HYPERTHERMIA AND RADIATION IN PATIENTS WITH LOCALLY ADVANCED MALIGNANCIES', INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS, pp. 1273-1279.

Journal articles

Anderson, C. & Ryan, L.M. 2017, 'A Comparison of Spatio-Temporal Disease Mapping Approaches Including an Application to Ischaemic Heart Disease in New South Wales, Australia.', International Journal of Environmental Research and Public Health, vol. 14, no. 2.
View/Download from: UTS OPUS or Publisher's site
The field of spatio-temporal modelling has witnessed a recent surge as a result of developments in computational power and increased data collection. These developments allow analysts to model the evolution of health outcomes in both space and time simultaneously. This paper models the trends in ischaemic heart disease (IHD) in New South Wales, Australia over an eight-year period between 2006 and 2013. A number of spatio-temporal models are considered, and we propose a novel method for determining the goodness-of-fit for these models by outlining a spatio-temporal extension of the Moran's I statistic. We identify an overall decrease in the rates of IHD, but note that the extent of this health improvement varies across the state. In particular, we identified a number of remote areas in the north and west of the state where the risk stayed constant or even increased slightly.
Wang, J.J.J., Bartlett, M. & Ryan, L. 2017, 'On the impact of nonresponse in logistic regression: application to the 45 and Up study.', BMC Med Res Methodol, vol. 17, no. 1, p. 80.
View/Download from: Publisher's site
BACKGROUND: In longitudinal studies, nonresponse to follow-up surveys poses a major threat to validity, interpretability and generalisation of results. The problem of nonresponse is further complicated by the possibility that nonresponse may depend on the outcome of interest. We identified sociodemographic, general health and wellbeing characteristics associated with nonresponse to the follow-up questionnaire and assessed the extent and effect of nonresponse on statistical inference in a large-scale population cohort study. METHODS: We obtained the data from the baseline and first wave of the follow-up survey of the 45 and Up Study. Of those who were invited to participate in the follow-up survey, 65.2% responded. Logistic regression model was used to identify baseline characteristics associated with follow-up response. A Bayesian selection model approach with sensitivity analysis was implemented to model nonignorable nonresponse. RESULTS: Characteristics associated with a higher likelihood of responding to the follow-up survey include female gender, age categories 55-74, high educational qualification, married/de facto, worked part or partially or fully retired and higher household income. Parameter estimates and conclusions are generally consistent across different assumptions on the missing data mechanism. However, we observed some sensitivity for variables that are strong predictors for both the outcome and nonresponse. CONCLUSIONS: Results indicated in the context of the binary outcome under study, nonresponse did not result in substantial bias and did not alter the interpretation of results in general. Conclusions were still largely robust under nonignorable missing data mechanism. Use of a Bayesian selection model is recommended as a useful strategy for assessing potential sensitivity of results to missing data.
Wang, J.J.J., Bartlett, M. & Ryan, L. 2017, 'Non-ignorable missingness in logistic regression.', Stat Med, vol. 36, no. 19, pp. 3005-3021.
View/Download from: Publisher's site
Nonresponses and missing data are common in observational studies. Ignoring or inadequately handling missing data may lead to biased parameter estimation, incorrect standard errors and, as a consequence, incorrect statistical inference and conclusions. We present a strategy for modelling non-ignorable missingness where the probability of nonresponse depends on the outcome. Using a simple case of logistic regression, we quantify the bias in regression estimates and show the observed likelihood is non-identifiable under non-ignorable missing data mechanism. We then adopt a selection model factorisation of the joint distribution as the basis for a sensitivity analysis to study changes in estimated parameters and the robustness of study conclusions against different assumptions. A Bayesian framework for model estimation is used as it provides a flexible approach for incorporating different missing data assumptions and conducting sensitivity analysis. Using simulated data, we explore the performance of the Bayesian selection model in correcting for bias in a logistic regression. We then implement our strategy using survey data from the 45 and Up Study to investigate factors associated with worsening health from the baseline to follow-up survey. Our findings have practical implications for the use of the 45 and Up Study data to answer important research questions relating to health and quality-of-life. Copyright © 2017 John Wiley & Sons, Ltd.
Huque, M.H., Bondell, H.D., Carroll, R.J. & Ryan, L.M. 2016, 'Spatial regression with covariate measurement error: A semiparametric approach.', Biometrics, vol. 72, no. 3, pp. 678-686.
View/Download from: UTS OPUS or Publisher's site
Spatial data have become increasingly common in epidemiology and public health research thanks to advances in GIS (Geographic Information Systems) technology. In health research, for example, it is common for epidemiologists to incorporate geographically indexed data into their studies. In practice, however, the spatially defined covariates are often measured with error. Naive estimators of regression coefficients are attenuated if measurement error is ignored. Moreover, the classical measurement error theory is inapplicable in the context of spatial modeling because of the presence of spatial correlation among the observations. We propose a semiparametric regression approach to obtain bias-corrected estimates of regression parameters and derive their large sample properties. We evaluate the performance of the proposed method through simulation studies and illustrate using data on Ischemic Heart Disease (IHD). Both simulation and practical application demonstrate that the proposed method can be effective in practice.
Waller, D., Mondy, P., Brama, T., Fisher, J., King, A., Malkov, K., Wall-Smith, D., Ryan, L. & Irving, D.O. 2016, 'Determining the effect of vein visualization technology on donation success, vasovagal symptoms, anxiety and intention to re-donate in whole blood donors aged 18-30 years: A randomized controlled trial.', Vox sanguinis, vol. 111, no. 2, pp. 135-143.
View/Download from: Publisher's site
Vein visualization technology (VVT) devices use near-infrared light to assist location of peripheral veins. The current study investigated the impact of VVT on donor experience and collection success for young blood donors at the Australian Red Cross Blood Service.The study in donors aged 18 to 30 years used a two intervention to one control randomized trial design with 285 new and 587 returning donors recruited at two sites. Donors reported presyncopal symptoms, phlebotomy pain, anxiety and intentions to redonate along with other measures. Participating phlebotomists rated usefulness of the technology. Flow rates, collection volumes and other donation information were taken from routine data.No significant differences were found between control and intervention groups on presyncopal symptoms, phlebotomy pain, anxiety, intentions to redonate, flow rates, collection volumes or vasovagal reactions (all P's > 005). Phlebotomist ratings of VVT were significantly more positive when they had less than 5 years of experience (P < 001) or when the vein was not visible to the naked eye (P < 001).Results suggest that VVT does not improve the donation experience for younger blood donors. Staff reports indicate that VVT may have some utility for assisting with difficult phlebotomies.
Huque, M.H., Carroll, R.J., Diao, N., Christiani, D.C. & Ryan, L.M. 2016, 'Exposure Enriched Case-Control (EECC) Design for the Assessment of Gene-Environment Interaction.', Genetic epidemiology, vol. 40, no. 7, pp. 570-578.
View/Download from: Publisher's site
Genetic susceptibility and environmental exposure both play an important role in the aetiology of many diseases. Case-control studies are often the first choice to explore the joint influence of genetic and environmental factors on the risk of developing a rare disease. In practice, however, such studies may have limited power, especially when susceptibility genes are rare and exposure distributions are highly skewed. We propose a variant of the classical case-control study, the exposure enriched case-control (EECC) design, where not only cases, but also high (or low) exposed individuals are oversampled, depending on the skewness of the exposure distribution. Of course, a traditional logistic regression model is no longer valid and results in biased parameter estimation. We show that addition of a simple covariate to the regression model removes this bias and yields reliable estimates of main and interaction effects of interest. We also discuss optimal design, showing that judicious oversampling of high/low exposed individuals can boost study power considerably. We illustrate our results using data from a study involving arsenic exposure and detoxification genes in Bangladesh.
Huque, M.H., Anderson, C., Walton, R. & Ryan, L. 2016, 'Individual level covariate adjusted conditional autoregressive (indiCAR) model for disease mapping.', International journal of health geographics, vol. 15, no. 1, p. 25.
View/Download from: UTS OPUS or Publisher's site
Mapping disease rates over a region provides a visual illustration of underlying geographical variation of the disease and can be useful to generate new hypotheses on the disease aetiology. However, methods to fit the popular and widely used conditional autoregressive (CAR) models for disease mapping are not feasible in many applications due to memory constraints, particularly when the sample size is large. We propose a new algorithm to fit a CAR model that can accommodate both individual and group level covariates while adjusting for spatial correlation in the disease rates, termed indiCAR. Our method scales well and works in very large datasets where other methods fail.We evaluate the performance of the indiCAR method through simulation studies. Our simulation results indicate that the indiCAR provides reliable estimates of all the regression and random effect parameters. We also apply indiCAR to the analysis of data on neutropenia admissions in New South Wales (NSW), Australia. Our analyses reveal that lower rates of neutropenia admissions are significantly associated with individual level predictors including higher age, male gender, residence in an outer regional area and a group level predictor of social disadvantage, the socio-economic index for areas. A large value for the spatial dependence parameter is estimated after adjusting for individual and area level covariates. This suggests the presence of important variation in the management of cancer patients across NSW.Incorporating individual covariate data in disease mapping studies improves the estimation of fixed and random effect parameters by utilizing information from multiple sources. Health registries routinely collect individual and area level information and thus could benefit by using indiCAR for mapping disease rates. Moreover, the natural applicability of indiCAR in a distributed computing framework enhances its application in the Big Data domain with a large number of individual/group level c...
Cramb, S.M., Mengersen, K.L., Lambert, P.C., Ryan, L.M. & Baade, P.D. 2016, 'A flexible parametric approach to examining spatial variation in relative survival', Statistics in Medicine, vol. 35, no. 29, pp. 5448-5463.
View/Download from: Publisher's site
Ryan, L.M., Wand, M.P. & Malecki, A.A. 2016, 'Bringing coals to Newcastle', Significance, vol. 13, no. 6, pp. 32-37.
View/Download from: UTS OPUS or Publisher's site
Cramb, S.M., Baade, P.D., White, N.M., Ryan, L.M. & Mengersen, K.L. 2015, 'Inferring lung cancer risk factor patterns through joint Bayesian spatio-temporal analysis', Cancer Epidemiology, vol. 39, pp. 430-439.
View/Download from: UTS OPUS or Publisher's site
Background: Preventing risk factor exposure is vital to reduce the high burden from lung cancer. The leading risk factor for developing lung cancer is tobacco smoking. In Australia, despite apparent success in reducing smoking prevalence, there is limited information on small area patterns and small area temporal trends. We sought to estimate spatio-temporal patterns for lung cancer risk factors using routinely collected population-based cancer data. Methods: The analysis used a Bayesian shared component spatio-temporal model, with male and female lung cancer included separately. The shared component reflected lung cancer risk factors, and was modelled over 477 statistical local areas (SLAs) and 15 years in Queensland, Australia. Analyses were also run adjusting for area-level socioeconomic disadvantage, Indigenous population composition, or remoteness. Results: Strong spatial patterns were observed in the underlying risk factor estimates for both males (median Relative Risk (RR) across SLAs compared to the Queensland average ranged from 0.48 to 2.00) and females (median RR range across SLAs 0.53-1.80), with high risks observed in many remote areas. Strong temporal trends were also observed. Males showed a decrease in the underlying risk across time, while females showed an increase followed by a decrease in the final 2 years. These patterns were largely consistent across each SLA. The high underlying risk estimates observed among disadvantaged, remote and indigenous areas decreased after adjustment, particularly among females. Conclusion: The modelled underlying risks appeared to reflect previous smoking prevalence, with a lag period of around 30 years, consistent with the time taken to develop lung cancer. The consistent temporal trends in lung cancer risk factors across small areas support the hypothesis that past interventions have been equally effective across the state. However, this also means that spatial inequalities have remained unaddressed, highlighti...
Hunt, P.R., Friesen, M.C., Sama, S., Ryan, L. & Milton, D. 2015, 'Log-Linear Modeling of Agreement among Expert Exposure Assessors', ANNALS OF OCCUPATIONAL HYGIENE, vol. 59, no. 6, pp. 764-774.
View/Download from: Publisher's site
Ryan, L. 2015, 'A Conversation with Nan Laird', STATISTICAL SCIENCE, vol. 30, no. 4, pp. 582-596.
View/Download from: Publisher's site
Poon, A.H., Houseman, E.A., Ryan, L., Sparrow, D., Vokonas, P.S. & Litonjua, A.A. 2014, 'Variants of Asthma and Chronic Obstructive Pulmonary Disease Genes and Lung Function Decline in Aging', JOURNALS OF GERONTOLOGY SERIES A-BIOLOGICAL SCIENCES AND MEDICAL SCIENCES, vol. 69, no. 7, pp. 907-913.
View/Download from: Publisher's site
Welsh, P., Woodward, M., Hillis, G.S., Li, Q., Marre, M., Williams, B., Poulter, N., Ryan, L., Harrap, S., Patel, A., Chalmers, J. & Sattar, N. 2014, 'Do Cardiac Biomarkers NT-proBNP and hsTnT Predict Microvascular Events in Patients With Type 2 Diabetes? Results From the ADVANCE Trial', DIABETES CARE, vol. 37, no. 8, pp. 2202-2210.
View/Download from: Publisher's site
Huque, M.H., Bondell, H.D. & Ryan, L. 2014, 'On the impact of covariate measurement error on spatial regression modelling', Environmetrics, vol. 25, no. 8, pp. 560-570.
View/Download from: Publisher's site
Maity, A., Williams, P.L., Ryan, L., Missmer, S.A., Coull, B.A. & Hauser, R. 2014, 'Analysis of in vitro fertilization data with multiple outcomes using discrete time-to-event analysis', Statistics in Medicine, vol. 33, no. 10, pp. 1738-1749.
View/Download from: Publisher's site
D'Agostino, R., Day, S., Greenhouse, J. & Ryan, L. 2014, 'Editors' Note', Statistics in Medicine, vol. 33, no. 1, p. 1.
View/Download from: Publisher's site
Whitney, M. & Ryan, L.M. 2013, 'Uncertainty due to low-dose extrapolation: modified BMD methodology for epidemiological data', Environmetrics, vol. 24, no. 5, pp. 289-297.
View/Download from: UTS OPUS or Publisher's site
Traditional environmental risk assessment methodologies, including benchmark dose (BMD) estimation, were originally developed to be used with animal toxicology data. We discuss some problems that can occur when toxicology-based methods are applied to hum
Whitney, M., Ryan, L.M. & Walkowiak, J. 2013, 'On the Use of Bayesian Model Averaging for Covariate Selection in Epidemiological Modeling', Journal of Statistical Theory and Practice, vol. 7, pp. 233-247.
View/Download from: UTS OPUS or Publisher's site
Bayesian model averaging (BMA) is a powerful technique to address model selection uncertainty and recent computational advances have led to a proliferation of usage. BMA methods are of particular interest in environmental health risk assessment because of the high degree of uncertainty that typically arises in that context. In this article, we review a variety of approaches to conducting BMA and compare four implementations in a setting where there are a number of potential predictors. We then use these four methods to calculate risk assessment measures that account for the uncertainty involved in modeling environmental exposures. These methods are used to reexamine data from a study conducted by Walkowiak et al. (2001) to investigate the effects of maternal polychlorinated biphenyl exposure on cognitive development in early childhood. This case study reveals that different strategies for implementing BMA can yield varying risk assessment results. We conclude with some practical recommendations.
Boyle, J., Crilly, J., Keijzers, G., Wallis, M., Lind, J., Sparks, R. & Ryan, L.M. 2012, 'Impact Of Influenza Across 27 Public Emergency Departments In Australia: A 5-year Descriptive Study', Emergency Medicine Journal, vol. 29, no. 9, pp. 725-731.
View/Download from: UTS OPUS or Publisher's site
Objective To describe the incidence, characteristics and outcomes of patients with influenza-like symptoms presenting to 27 public hospital emergency departments (EDs) in Queensland, Australia. Methods A descriptive retrospective study covering 5 years (20059) of historical data from 27 hospital EDs was undertaken. State-wide hospital ED Information System data were analysed. Annual comparisons between influenza and non-influenza cases were made across the southern hemisphere influenza season (JuneSeptember) each year. Results Influenza-related presentations increased significantly over the 5 years from 3.4% in 2005 to 9.4% in 2009, reflecting a 276% relative increase. Differences over time regarding characteristics of patients with influenza-like symptoms, based on the influenza season, occurred for admission rate (decreased over time from 28% in 2005 to 18% in 2009), length of stay (decreased over time from a median of 210 min in 2005 to 164 min in 2009) and access block (increased over time from 33% to 41%). Also, every year there was a significantly (p<0.001) higher percentage of access block in the influenza cohort than in the non-influenza cohort
Shen, K.-.K., Fripp, J., Meriaudeau, F., Chetelat, G., Salvado, O., Bourgeat, P. & Neuroimaging, A.D. 2012, 'Detecting global and local hippocampal shape changes in Alzheimer's disease using statistical shape models', NEUROIMAGE, vol. 59, no. 3, pp. 2155-2166.
View/Download from: Publisher's site
Graham, P., Ryan, L.M. & Luszcz, M. 2011, 'Joint Modelling Of Survival And Cognitive Decline In The Australian Longitudinal Study Of Ageing', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 60, no. 2, pp. 221-238.
View/Download from: UTS OPUS or Publisher's site
The paper describes the use of a longitudinal tobit model to characterize cognitive decline over a 13-year period in a cohort of 2087 elderly Australians. Use of a tobit formulation allows for the so-called `ceiling effect wherein many subjects achieve perfect test scores. A Bayesian hierarchical joint model is presented that allows for random subject-specific intercepts and slopes, as well as for informative dropout. Results suggest several potential areas of intervention. For example, there is a clear doseresponse effect of exercise whereby increasing levels of exercise are associated with higher cognitive scores.
Missmer, S., Pearson, K., Ryan, L.M., Meeker, J., Cramer, D. & Hauser, R. 2011, 'Analysis Of Multiple-cycle Data From Couples Undergoing In Vitro Fertilization Methodologic Issues And Statistical Approaches', Epidemiology, vol. 22, no. 4, pp. 497-504.
View/Download from: UTS OPUS or Publisher's site
The number of in vitro fertilization (IVF) cycles in the United States increased from fewer than 46,000 in 1995 to more than 120,000 in 2005. IVF and other assisted reproductive technology (ART) data are routinely collected and used to identify outcome predictors. However, researchers do not always make full use of the data due to their complexity. Design approaches have included restriction to first-cycle attempts only, which reduces power and identifies effects only of those factors associated with initial success. Many statistical techniques have been used or proposed for analysis of IVF data, ranging from simple t tests to sophisticated models designed specifically for IVF. We applied several of these methods to data from a prospective cohort of 2687 couples undergoing ART from 1994 through 2003. Results across methods are compared and the appropriateness of the various methods is discussed with the intent to illustrate methodologic validity. We observed a remarkable similarity of coefficient estimates across models. However, each method for dealing with multiple cycle data relies on assumptions that may or may not be expected to hold in a given IVF study. The robustness and reported magnitude of effect for individual predictors of IVF success may be inflated or attenuated due to violation of statistical assumptions, and should always be critically interpreted. Given that risk factors associated with IVF success may also advance our understanding of the physiologic processes underlying conception, implantation, and gestation, the application of valid methods to these complex data is critical.
Ngo, L., Ryan, L.M., Mezzetti, M., Bois, F. & Smith, T. 2011, 'Estimating Metabolic Rate For Butadiene At Steady State Using A Bayesian Physiologically-based Pharmacokinetic Model', Environmental and Ecological Statistics, vol. 18, no. 1, pp. 131-146.
View/Download from: UTS OPUS or Publisher's site
In a study of 133 volunteer subjects, demographic, physiologic and pharmacokinetic data through exposure to 1,3-Butadiene (BD) were collected in order to estimate the percentage of BD concentration metabolized at steady state, and to determine whether this percentage varies across gender, racial, and age groups. During the 20 min of continuous exposure to 2 parts per million (ppm) of BD, five measurements of exhaled concentration were made on each subject. In the following 40 min washout period, another five measurements were collected. A Bayesian hierarchical compartmental physiologically-based pharmacokinetic model (PKPB) was used. Using prior information on the model parameters, Markov Chain Monte Carlo (MCMC) simulation was conducted to obtain posterior distributions. The overall estimate of the mean percent of BD metabolized at steady state was 12.7% (95% credible interval: 7.7-17.8%). There was no significant difference in gender with males having a mean of 13.5%, and females 12.3%. Among the racial groups, Hispanic (13.9%), White (13.0%), Asian (12.1%), and Black (10.9%), the significant difference came from the difference between Black and Hispanic with a 95% credible interval from -5.63 to -0.30%. Those older than 30 years had a mean of 12.2% versus 12.9% for the younger group; although this was not a statistically significant difference. Given a constant inhalation input of 2 ppm, at steady state, the overall mean exhaled concentration was estimated to be 1.75ppm (95% credible interval: 1.64-1.84). An equivalent parameter, first-order metabolic rate constant, was also estimated and found to be consistent with the percent of BD metabolized at steady state across gender, race, and age strata.
Suglia, S., Ryan, L.M., Bellinger, D., Enlow, M. & Wright, R. 2011, 'Children's Exposure To Violence And Distress Symptoms: Influence Of Caretakers' Psychological Functioning', INTERNATIONAL JOURNAL OF BEHAVIORAL MEDICINE, vol. 18, no. 1, pp. 35-43.
View/Download from: UTS OPUS or Publisher's site
Background Previous studies linking violence exposure to adverse child behavior have typically relied on parental report of child symptoms without accounting for the informant's mental well-being, despite evidence that parental mental health can influence children's mental health and the parent's report of distress symptoms. Purpose We assess the influence of maternal depression on the violence exposure and child distress association in a subset of the Maternal Infant Smoking Study of East Boston, a prospective birth cohort. Results Among the 162 children ages 7 to 11, 51% were boys and 43% self-identified as Hispanic. When using child self-report, increased violence exposure was significantly associated with a broader range of distress symptoms (numbness, arousal, intrusion, avoidance subscales) compared to parent reported findings, which were only significantly related to the intrusion and avoidance subscales. Moreover, a significant mediation effect of maternal depression on the violence and distress association was noted only when mother's report of exposure and outcome was used. Conclusion Considering both parent and child self-report of violence is necessary to obtain a complete picture of violence exposure because parents and children may be offering different, although equally valid information. The influence of maternal depressive symptoms on preadolescent's distress symptoms may be attributed to reporting bias as opposed to more direct effects; thus, the parent's psychological functioning should be taken into consideration when relying on parental report of the child's psychological functioning.
Hart, J., Garshick, E., Dockery, D., Smith, T., Ryan, L.M. & Laden, F. 2011, 'Long-term Ambient Multipollutant Exposures And Mortality', American Journal of Respiratory and Critical Care Medicine, vol. 183, no. 1, pp. 73-78.
View/Download from: UTS OPUS or Publisher's site
Rationale: Population-based studies have demonstrated associations between ambient air pollution exposures and mortality, but few have been able to adjust for occupational exposures. Additionally, two studies have observed higher risks in individuals with occupational dust, gas, or fume exposure. Objectives: We examined the association of ambient residential exposure to particulate matter less than 10 mm in diameter (PM10), particulate matter less than 2.5 mm in diameter (PM2.5), NO2, SO2, and mortality in 53,814 men in the U.S. trucking industry. Methods: Exposures for PM10, NO2, and SO2 at each residential address were assigned using models combining spatial smoothing and geographic covariates. PM2.5 exposures in 2000 were assigned from the nearest available monitor. Single and multipollutant Cox proportional hazard models were used to examine the association of an interquartile range (IQR) change (6 mg/m3 for PM10, 4 mg/m3 for PM2.5, 4ppb for SO2, and 8ppb for NO2) and the risk of all-cause and cause-specific mortality. Measurements and Main Results: An IQR change in ambient residential exposures to PM10 was associated with a 4.3% (95% confidence interval [CI], 1.17.7%) increased risk of all-cause mortality. The increase for an IQR change in SO2 was 6.9% (95% CI, 2.311.6%), for NO2 was8.2%(95%CI, 4.512.1%), andforPM2.5 was3.9%(95%CI, 1.06.9%). Elevated associations with cause-specific mortality (lung cancer, cardiovascular and respiratory disease) were observed for PM2.5, SO2, and NO2, but not PM10. None of the pollutants were confounded by occupational exposures. In multipollutant models, overall, the associations were attenuated, most strongly forPM10. In sensitivity analyses excluding long-haul drivers, who spend days away from home, larger hazard ratios were observed. Conclusions: In this population of men, residential ambient air pollution exposures were associated with mortality.
Boyle, J.R., Sparks, R.S., Keijzers, G.B., Crilly, J.L., Lind, J.F. & Ryan, L.M. 2011, 'Prediction and surveillance of influenza epidemics', MEDICAL JOURNAL OF AUSTRALIA, vol. 194, no. 4, pp. S28-S33.
View/Download from: UTS OPUS
Graham, P.L., Ryan, L.M. & Luszcz, M.A. 2011, 'Joint modelling of survival and cognitive decline in the Australian Longitudinal Study of Ageing', JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS, vol. 60, pp. 221-238.
View/Download from: Publisher's site
Sparks, R., Carter, C., Graham, P., Muscatello, D., Churches, T., Kaldor, J., Turner, R., Zheng, W. & Ryan, L. 2011, 'Understanding sources of variation in syndromic surveillance for early warning of natural or intentional disease outbreaks (vol 42, pg 613, 2010)', IIE TRANSACTIONS, vol. 43, no. 3, pp. 231-231.
View/Download from: Publisher's site
Choi, J.H., Ryan, L.M., Cramer, D.W., Hornstein, M.D. & Missmer, S.A. 2011, 'Effects of Caffeine Consumption by Women and Men on the Outcome of In Vitro Fertilization.', Journal of caffeine research, vol. 1, no. 1, pp. 29-34.
The objective of this study was to estimate the association between caffeine consumption and in vitro fertilization (IVF) outcomes.A total of 2474 couples were prospectively enrolled prior to undergoing their first cycle of IVF, contributing a total of 4716 IVF cycles. Discrete survival analysis adjusting for observed confounders was applied to quantify the relation between caffeine consumption and livebirth. Secondary outcomes of interest were oocyte retrieval, peak estradiol level, implantation rate, and fertilization rate.Overall, caffeine consumption by women was not significantly associated with livebirth (p trend=0.74). Compared with women who do not drink caffeine, the likelihood of livebirth was not significantly different for women who drank low (>0-800mg/week; odds ratio [OR]=1.00, 95% confidence interval [CI])=0.83-1.21), moderate (>800-1400mg/week; OR=0.89, 95% CI=0.71-1.12), or high levels of caffeine (>1400mg/week; OR=1.07, 95% CI=0.85-1.34). Greater caffeine intake by women was associated with a significantly lower peak estradiol level (p trend=0.03), but was not associated with the number of oocytes retrieved (p trend=0.75), fertilization rate (p trend=0.10), or implantation rate (p trend=0.23). There was no significant association between caffeine intake by men and livebirth (p trend=0.27), fertilization (p trend=0.72), or implantation (p trend=0.24). The individual effects of consumption of coffee, tea, or soda by women or men were not related to livebirth.Caffeine consumption by women or men was not associated with IVF outcomes.
Chen, Y., Christiani, D., Su, H., Hsueh, Y., Smith, T., Ryan, L.M., Chao, S., Lee, J. & Guo, Y. 2010, 'Early-life Or Lifetime Sun Exposure, Sun Reaction, And The Risk Of Squamous Cell Carcinoma In An Asian Population', Cancer Causes & Control, vol. 21, no. 5, pp. 771-776.
View/Download from: UTS OPUS or Publisher's site
It has been widely accepted that sun exposure is a risk factor of squamous cell carcinoma (SCC) among fair-skinned populations. However, sun exposure and sun reaction have not been explored in Asians and no gender-specific data were available. In a case-
Cai, T., Parast, L. & Ryan, L.M. 2010, 'Meta-analysis for rare events', Statistics in Medicine, vol. 29, no. 20, pp. 2078-2089.
View/Download from: UTS OPUS or Publisher's site
Meta-analysis provides a useful framework for combining information across related studies and has been widely utilized to combine data from clinical studies in order to evaluate treatment efficacy. More recently, meta-analysis has also been used to assess drug safety. However, because adverse events are typically rare, standard methods may not work well in this setting. Most popular methods use fixed or random effects models to combine effect estimates obtained separately for each individual study. In the context of very rare outcomes, effect estimates from individual studies may be unstable or even undefined. We propose alternative approaches based on Poisson random effects models to make inference about the relative risk between two treatment groups. Simulation studies show that the proposed methods perform well when the underlying event rates are low. The methods are illustrated using data from a recent meta-analysis (N. Engl. J. Med. 2007; 356(24):2457-2471) of 48 comparative trials involving rosiglitazone, a type 2 diabetes drug, with respect to its possible cardiovascular toxicity
Strauss, W., Ryan, L.M., Morara, M., Iroz-elardo, N., Davis, M., Cupp, M., Nishioka, M., Quackenboss, J., Galke, W., Ozkaynak, H. & Scheidt, P. 2010, 'Improving Cost-effectiveness Of Epidemiological Studies Via Designed Missingness Strategies', Statistics in Medicine, vol. 29, no. 13, pp. 1377-1387.
View/Download from: UTS OPUS or Publisher's site
Modern epidemiological studies face opportunities and challenges posed by an ever-expanding capacity to measure a wide range of environmental exposures, along with sophisticated biomarkers of exposure and response at the individual level. The challenge of deciding what to measure is further complicated for longitudinal studies, where logistical and cost constraints preclude the collection of all possible measurements on all participants at every follow-up time. This is true for the National Children's Study (NCS), a large-scale longitudinal study that will enroll women both prior to conception and during pregnancy and collect information on their environment, their pregnancies, and their children's development through early adulthood with a goal of assessing key exposure/outcome relationships among a cohort of approximately 100 000 children. The success of the NCS will significantly depend on the accurate, yet cost-effective, characterization of environmental exposures thought to be related to the health outcomes of interest. The purpose of this paper is to explore the use of cost saving, yet valid and adequately powered statistical approaches for gathering exposure information within epidemiological cohort studies. The proposed approach involves the collection of detailed exposure assessment information on a specially selected subset of the study population, and collection of less-costly, and presumably less-detailed and less-burdensome, surrogate measures across the entire cohort. We show that large-scale efficiency in costs and burden may be achieved without making substantive sacrifices on the ability to draw reliable inferences concerning the relationship between exposure and health outcome. Several detailed scenarios are provided that document how the targeted sub-sampling design strategy can benefit large cohort studies like the NCS, as well as other more focused environmental epidemiologic studies
Sparks, R., Carter, C., Graham, P., Muscatello, D., Churches, T., Kaldor, J., Turner, R., Zheng, W. & Ryan, L.M. 2010, 'Understanding Sources Of Variation In Syndromic Surveillance For Early Warning Of Natural Or Intentional Disease Outbreaks', IIE Transactions, vol. 42, no. 9, pp. 613-631.
View/Download from: UTS OPUS or Publisher's site
Daily counts of computer records of hospital emergency department arrivals grouped according to diagnosis (called here syndrome groupings) can be monitored by epidemiologists for changes in frequency that could provide early warning of bioterrorism events or naturally occurring disease outbreaks and epidemics. This type of public health surveillance is sometimes called syndromic surveillance. We used transitional Poisson regression models to obtain one-day-ahead arrival forecasts. Regression parameter estimates and forecasts were updated for each day using the latest 365 days of data. The resulting time series of recursive estimates of parameters such as the amplitude and location of the seasonal peaks as well as the one-day-ahead forecasts and forecast errors can be monitored to understand changes in epidemiology of each syndrome grouping. The counts for each syndrome grouping were autocorrelated and non-homogeneous Poisson. As such, the main methodological contribution of the article is the adaptation of Cumulative Sum (CUSUM) and Exponentially Weighted Moving Average (EWMA) plans for monitoring non-homogeneous counts. These plans were valid for small counts where the assumption of normally distributed one-day-ahead forecasts errors, typically used in other papers, breaks down. In addition, these adaptive plans have the advantage that control limits do not have to be trained for different syndrome groupings or aggregations of emergency departments. Conventional methods for signaling increases in syndrome grouping counts, Shewhart, CUSUM, and EWMA control charts of the standardized forecast errors were also examined. Shewhart charts were, at times, insensitive to shifts of interest. CUSUM and EWMA charts were only reasonable for large counts.We illustrate our methods with respiratory, influenza, diarrhea, and abdominal pain syndrome groupings.
Guha, S., Ryan, L.M. & Morara, M. 2009, 'Gauss-Seidel Estimation Of Generalized Linear Mixed Models With Application To Poisson Modeling Of Spatially Varying Disease Rates', Journal of Computational and Graphical Statistics, vol. 18, no. 4, pp. 818-837.
View/Download from: UTS OPUS or Publisher's site
Generalized linear mixed models (GLMMs) are often fit by computational procedures such as penalized quasi-likelihood (PQL). Special cases of GLMMs are generalized linear models (GLMs), which are often fit using algorithms like iterative weighted least sq
Hart, J., Yanosky, J., Puett, R., Ryan, L.M., Dockery, D., Smith, T., Garshick, E. & Laden, F. 2009, 'Spatial Modeling Of Pm10 And No2 In The Continental United States, 1985-2000', Environmental Health Perspectives, vol. 117, no. 11, pp. 1690-1696.
View/Download from: UTS OPUS or Publisher's site
Background: Epidemiologic studies of air pollution have demonstrated a link between long-term air pollution exposures and mortality. However, many have been limited to city-specific average pollution measures or spatial or land-use regression exposure models in small geographic areas. Objectives: Our objective was to develop nationwide models of annual exposure to particulate matter < 10 &Ecirc;m in diameter (PM10) and nitrogen dioxide during 1985.2000. Methods: We used generalized additive models (GAMs) to predict annual levels of the pollutants using smooth spatial surfaces of available monitoring data and geographic information system. derived covariates. Model performance was determined using a cross-validation (CV) procedure with 10% of the data. We also compared the results of these models with a commonly used spatial interpolation, inverse distance weighting. Results: For PM10, distance to road, elevation, proportion of low-intensity residential, highintensity residential, and industrial, commercial, or transportation land use within 1 km were all statistically significant predictors of measured PM10 (model R2 = 0.49, CV R2 = 0.55). Distance to road, population density, elevation, land use, and distance to and emissions of the nearest nitrogen oxides.emitting power plant were all statistically significant predictors of measured NO2 (model R2 = 0.88, CV R2 = 0.90). The GAMs performed better overall than the inverse distance models, with higher CV R2 and higher precision. Conclusions: These models provide reasonably accurate and unbiased estimates of annual exposures for PM10 and NO2. This approach provides the spatial and temporal variability necessary to describe exposure in studies assessing the health effects of chronic air pollution.
Meeker, J., Missmer, S., Altshul, L., Vitonis, A., Ryan, L.M., Cramer, D. & Hauser, R. 2009, 'Serum And Follicular Fluid Organochlorine Concentrations Among Women Undergoing Assisted Reproduction Technologies', Environmental health, vol. 8, no. 32, pp. 1-10.
View/Download from: UTS OPUS or Publisher's site
Sanchez, B., Budtz-jorgensen, E. & Ryan, L.M. 2009, 'An Estimating Equations Approach To Fitting Latent Exposure Models With Longitudinal Health Outcomes', Annals of Applied Statistics, vol. 3, no. 2, pp. 830-856.
View/Download from: UTS OPUS or Publisher's site
The analysis of data arising from environmental health studies which collect a large number of measures of exposure can benefit from using latent variable models to summarize exposure information. However, difficulties with estimation of model parameters may arise since existing fitting procedures for linear latent variable models require correctly specified residual variance structures for unbiased estimation of regression parameters quantifying the association between (latent) exposure and health outcomes. We propose an estimating equations approach for latent exposure models with longitudinal health outcomes which is robust to misspecification of the outcome variance. We show that compared to maximum likelihood, the loss of efficiency of the proposed method is relatively small when the model is correctly specified. The proposed equations formalize the ad-hoc regression on factor scores procedure, and generalize regression calibration. We propose two weighting schemes for the equations, and compare their efficiency. We apply this method to a study of the effects of in-utero lead exposure on child development.
Clark, C., Kawachi, I., Ryan, L.M., Ertel, K., Fay, M. & Berkman, L. 2009, 'Perceived Neighborhood Safety And Incident Mobility Disability Among Elders: The Hazards Of Poverty', BMC Public Health, vol. 9, no. 162, pp. 0-0.
View/Download from: UTS OPUS or Publisher's site
Weuve, J., Korrick, S., Weisskopf, M., Ryan, L.M., Schwartz, J., Nie, H., Grodstein, F. & Hu, H. 2009, 'Cumulative Exposure To Lead In Relation To Cognitive Function In Older Women', Environmental Health Perspectives, vol. 117, no. 4, pp. 574-580.
View/Download from: UTS OPUS or Publisher's site
BACKGROUND: Recent data indicate that chronic low-level exposure to lead is associated with accelerated declines in cognition in older age, but this has not been examined in women. OBJECTIVE: We examined biomarkers of lead exposure in relation to performance on a battery of cognitive tests among older women. METHODS: Patella and tibia bone lead-measures of cumulative exposure over many years-and blood lead, a measure of recent exposure, were assessed in 587 women 47-74 years of age. We assessed their cognitive function 5 years later using validated telephone interviews. RESULTS: Mean +/- SD lead levels in tibia, patella, and blood were 10.5 +/- 9.7 mu g/g bone, 12.6 +/- 11.6 mu g/g bone, and 2.9 +/- 1.9 mu g/dL, respectively, consistent with community-level exposures. In multivariable-adjusted analyses of all cognitive tests combined, levels of all three lead biomarkers were associated with worse cognitive performance. The association between bone lead and letter fluency score differed dramatically from the other bone lead-cognitive score associations, and exclusion of this particular score from the combined analyses strengthened the associations between bone lead and cognitive performance. Results were statistically significant only for tibia lead: one SD increase in tibia lead corresponded to a 0.051-unit lower standardized summary cognitive score (95% confidence interval: -0.099 to -0.003; P = 0.04), similar to the difference in cognitive scores we observed between women who were 3 years apart in age. CONCLUSIONS: These findings suggest that cumulative exposure to lead, even at low levels experienced in community settings, may have adverse consequences for women's cognition in older age.
Kile, M., Hoffman, E., Hsueh, Y., Afroz, S., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Ryan, L.M. & Christiani, D. 2009, 'Variability In Biomarkers Of Arsenic Exposure And Metabolism In Adults Over Time', Environmental Health Perspectives, vol. 117, no. 3, pp. 455-460.
View/Download from: UTS OPUS or Publisher's site
OBJECTIVES: To characterize inter- and intraindividual variability in IUAs in healthy individuals. METHODS: In a longitudinal study conducted in Bangladesh, we collected water and spot urine samples from 196 participants every 3 months for 2 years. Water arsenic (As) was measured by inductively coupled plasma-mass spectrometry and urinary As [arsenite, arsenate, monomethylarsonic acid (MMA), and dimethylarsinic acid (DMA)] were detected using high-performance liquid chromatography-hydride-generated atomic absorption spectrometry. We used linear mixed-effects models to compute variance components and evaluate the association between UAs and selected factors. RESULTS: The concentrations of IUAs were fairly reproducible within individuals, with intraclass correlation coefficients (ICCs) of 0.41, 0.35, 0.47, and 0.49 for inorganic As (InAs), MMA, DMA, and total urinary As (TUA). However, when expressed as a ratio, the percent InAs (%InAs), %MMA, and %DMA were poorly reproducible within individuals, with ICCs of 0.16, 0.16, and 0.17, respectively. Arsenic metabolism was significantly associated with sex, exposure, age, smoking, chewing betel nut, urinary creatinine, and season. Specificity and sensitivity analyses showed that a single urine sample adequately classified a participant's urinary As profile as high or low, but TUA had only moderate specificity for correctly classifying drinking water exposures. CONCLUSIONS: Epidemiologic studies should use both urinary As concentrations and the relative proportion of UAs to minimize measurement error and to facilitate interpretation of factors that influence As metabolism.
Owen, M., Ryan, L.M. & Holmes, L. 2009, 'Effects Of Retinoic Acid On Dominant Bemimelia Expression In Mice', Birth Defects Research Part A: Clinical and Molecular Teratology, vol. 85, no. 1, pp. 36-41.
View/Download from: UTS OPUS or Publisher's site
BACKGROUND: Dominant hemimelia (Dh) is an autosomal dominant mutation that arose spontaneously in mice. Dh animals are asplenic and they exhibit asymmetric hindlimb defects in association with reduced numbers of lumbar vertebrae. These defects suggest that Dh acts early in embryonic development to affect patterning of the anterior-posterior (A-P) and left-right axes. This study was undertaken to determine whether retinoic acid (RA), which is involved in A-P patterning and coordination of bilaterally synchronized somitogenesis, affects phenotypic expression of the Dh gene. METHODS: Thirty-four pregnant females were given, by oral intubation, a single dose of 50 or 75 mg all-trans RA per kilogram body weight at GD 9, 10, or 11. The pregnant females were then euthanized at GD 18 and fetuses removed by cesarean section. A total of 326 fetuses were identified by phenotype and linked DNA and their skeletons were analyzed. RESULTS: There was a differential effect of RA on the axial skeleton and hindlimb of Dh/+ mice as compared to their wild-type littermates. Dose- and stage-specific effects on sternebrae and vertebrae were observed. CONCLUSIONS: The effects of RA dosing on numbers of sternebrae and vertebrae suggest that Dh embryos have a primary defect in retinoid-mediated A-P patterning. Dosing with RA may produce the observed effects on phenotypic expression of Dh/+ by indirectly or directly modifying an already existing altered Hox expression pattern. As the relationship between axial patterning and the asymmetric limb is unknown, Dh is an important model for studying this relationship. Birth Defects Research (Part A) 85:36-41, 2009,
Mccracken, J., Schwartz, J., Bruce, N., Mittleman, M., Ryan, L.M. & Smith, K. 2009, 'Combining Individual- And Group-level Exposure Information Child Carbon Monoxide In The Guatemala Woodstove Randomized Control Trial', Epidemiology, vol. 20, no. 1, pp. 127-136.
View/Download from: UTS OPUS or Publisher's site
Background: Epidemiology frequently relies on surrogates of long-term exposures, often either individual-level short-term measurements or group-level based on long-term characteristics of subjects and their environment. Whereas individual-level measures
Ryan, L. 2009, 'Spatial Epidemiology Some Pitfalls and Opportunities', EPIDEMIOLOGY, vol. 20, no. 2, pp. 242-244.
View/Download from: Publisher's site
Sanchez, B.N., Houseman, E.A. & Ryan, L.M. 2009, 'Residual-Based Diagnostics for Structural Equation Models', BIOMETRICS, vol. 65, no. 1, pp. 104-115.
View/Download from: Publisher's site
Forno, E., Onderdonk, A.B., McCracken, J., Litonjua, A.A., Laskey, D., Delaney, M.L., DuBois, A.M., Gold, D.R., Ryan, L.M., Weiss, S.T. & Celedon, J.C. 2009, 'Diversity of the Gut Microbiota and Eczema in Infants', AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE, vol. 179.
Horton, N., Roberts, K., Ryan, L.M., Suglia, S. & Wright, R. 2008, 'A maximum likelihood latent variable regression model for multiple informants', Statistics in Medicine, vol. 27, pp. 4992-5004.
View/Download from: UTS OPUS or Publisher's site
Studies pertaining to childhood psychopathology often incorporate information front multiple sources (or informants). For example. measurement Of Some factor of particular interest might be collected from parents, teachers Lis well as the children being
Suglia, S., Ryan, L.M. & Wright, R. 2008, 'Creation Of A Community Violence Exposure Scale: Accounting For What, Who, Where, And How Often', Journal of Traumatic Stress, vol. 21, no. 5, pp. 479-486.
View/Download from: UTS OPUS or Publisher's site
Previous research has used the Rasch model, a method for obtaining a continuous scale from dichotomous survey items measuring a single latent construct, to create a scale of community violence exposure. The authors build upon previous work and describe the application of a Rasch model using the continuation ratio model to create an exposure to community violence (ETV) scale including event circumstance information previously shown to modify the impact of experienced events. They compare the Rasch ETV scale to a simpler sum ETV score, and estimate the effect of ETV on child posttraumatic stress symptoms. Incorporating detailed event circumstance information that is grounded in traumatic stress theory may reduce measurement error in the assessment of children's community violence exposure.
Ryan, L.M. 2008, 'Combining data from multiple sources, with applications to environmental risk assessment', Statistics in Medicine, vol. 27, no. 5, pp. 698-710.
View/Download from: UTS OPUS or Publisher's site
The classical statistical paradigm emphasizes the development and application of methods to account for sampling variability. Many modern day applications, however, require consideration of other sources of uncertainty that are not so easy to quantify. This paper presents a case study involving an assessment of the impact of in-utero methylmercury exposure on the Intelligence Quotient (IQ) of young children. We illustrate how familiar techniques such as hierarchical modeling, Bayesian methods and sensitivity analysis can be used to aid decision making in settings that involve substantial uncertainty
Schwartz, J., Coull, B., Laden, F. & Ryan, L.M. 2008, 'The Effect Of Dose And Timing Of Dose On The Association Between Airborne Particles And Survival', Environmental Health Perspectives, vol. 116, no. 1, pp. 64-69.
View/Download from: UTS OPUS or Publisher's site
BACKGROUND: Understanding the shape of the concentration-response curve for particles is important for public health, and lack of such understanding was recently cited by U.S. Environmental Protection Agency (EPA) as a reason for not tightening the standards. Similarly, the delay between changes in exposure and changes in health is also important in public health decision making. We addressed these issues using an extended follow-up of the Harvard Six Cities Study. METHODS: Cox proportional hazards models were fit controlling for smoking, body mass index, and other covariates. Two approaches were used. First, we used penalized splines, which fit a flexible functional form to the concentration response to examine its shape, and chose the degrees of freedom for the curve based on Akaike's information criterion. Because the uncertainties around the resultant curve do not reflect the uncertainty in model choice, we also used model averaging as an alternative approach, where multiple models are fit explicitly and averaged, weighted by their probability of being correct given the data. We examined the lag relationship by model averaging across a range of unconstrained distributed lag models. RESULTS: We found that the concentration-response curve is linear, clearly continuing below the current U.S. standard of 15 mu g/m(3), and that the effects of changes in exposure on mortality are seen within two years. CONCLUSIONS: Reduction in particle concentrations below U.S. EPA standards would increase life expectancy.
Phipatanakul, W., Celedon, J., Hoffman, E., Abdulkerim, H., Ryan, L.M. & Gold, D. 2008, 'Mouse Allergen Exposure, Wheeze And Atopy In The First Seven Years Of Life', Allergy, vol. 63, no. 11, pp. 1512-1518.
View/Download from: UTS OPUS or Publisher's site
Background: Little is known about mouse allergen exposure in home environments and the development of wheezing, asthma and atopy in childhood. Objective: To examine the relation between mouse allergen exposure and wheezing, atopy, and asthma in the first 7 years of life. Methods: Prospective study of 498 children with parental history of allergy or asthma followed from birth to age 7 years, with longitudinal questionnaire ascertainment of reported mouse exposure and dust sample mouse urinary protein allergen levels measured at age 2-3 months. Results: Parental report of mouse exposure in the first year of life was associated with increased risk of transient wheeze and wheezing in early life. Current report of mouse exposure was also significantly associated with current wheeze throughout the first 7 years of life in the longitudinal analysis (P = 0.03 for overall relation of current mouse to current wheeze). However, early life mouse exposure did not predict asthma, eczema or allergic rhinitis at age 7 years. Exposure to detectable levels of mouse urinary protein in house dust samples collected at age 2-3 months was associated with a twofold increase in the odds of atopy (sensitization to >=1 allergen) at school age (95% confidence interval for odds ratio = 1.1-3.7; P = 0.03 in a multivariate analysis. Conclusions: Among children with parental history of asthma or allergies, current mouse exposure is associated with increased risk of wheeze during the first 7 years of life. Early mouse exposure was associated with early wheeze and atopy later in life.
Boynton-jarrett, R., Ryan, L.M., Berkman, L. & Wright, R. 2008, 'Cumulative Violence Exposure And Self-rated Health: Longitudinal Study Of Adolescents In The United States', Pediatrics, vol. 122, no. 5, pp. 961-970.
View/Download from: UTS OPUS or Publisher's site
Surkan, P., Kawachi, I., Ryan, L.M., Berkman, L., Vieira, C. & Peterson, K. 2008, 'Maternal Depressive Symptoms, Parenting Self-efficacy, And Child Growth', American Journal of Public Health, vol. 98, no. 1, pp. 125-132.
View/Download from: UTS OPUS or Publisher's site
Objectives. We assessed whether maternal depressive symptoms and parenting self-efficacy were associated with child growth delay. Methods. We collected data from a random sample of 595 low-income mothers and their children aged 6 to 24 months in Teresina, Piau&iacute;, Brazil, including information on sociodemographic characteristics, mothers depressive symptoms and parenting self-efficacy, and childrens anthropometric characteristics. We used adjusted logistic regression models in our analyses. Results. Depressive symptoms among mothers were associated with 1.8 times higher odds (95% confidence interval [CI] = 1.1, 2.9) of short stature among children. Parenting self-efficacy was not associated with short stature, nor did it mediate or modify the relationship between depressive symptoms and short stature. Maternal depressive symptoms and self-efficacy were not related to child underweight. Conclusions. Our results showed that among low-income Brazilian families maternal depressive symptoms, but not self-efficacy, were associated with short stature in children aged 6 to 24 months after adjustment for known predictors of growth.
Clark, C., Ryan, L.M., Kawachi, I., Canner, M., Berkman, L. & Wright, R. 2008, 'Witnessing Community Violence In Residential Neighborhoods: A Mental Health Hazard For Urban Women', Journal Of Urban Health-bulletin Of The New York Academy Of Medicine, vol. 85, no. 1, pp. 22-38.
View/Download from: UTS OPUS or Publisher's site
We examined the prevalence and psychological correlates of witnessing community violence among women of low socioeconomic status living in urban neighborhoods in the northeastern United States. Three hundred eighty-six women receiving their health care at an urban community health center were sampled to assess their violence exposures. Women were asked to report the location and timing of their exposure to witnessing violent neighborhood events in which they were not participants. The Brief Symptoms Inventory was used to assess anxiety and depressive symptoms. Controlling for marital status, educational attainment, age, and intimate partner violence victimization, women who witnessed violent acts in their neighborhoods were twice as likely to experience depressive and anxiety symptoms compared to women who did not witness community violence. Central American-born women had particularly high exposures. We conclude that witnessing neighborhood violence is a pervasive experience in this urban cohort, and is associated with anxiety and depressive symptoms, even among women who are not direct participants in violence to which they are exposed. Community violence interventions must incorporate efforts to protect the mental health of adult women who witness events in their neighborhoods.
Suglia, S., Ryan, L.M., Laden, F., Dockery, D. & Wright, R. 2008, 'Violence Exposure, A-chronic Psychosocial Stressor, And Childhood Lung Function', Psychosomatic Medicine, vol. 70, no. 2, pp. 160-169.
View/Download from: UTS OPUS or Publisher's site
Objective: Chronic psychosocial stressors, including violence, and neuropsychological and behavioral development in children as well as physiologic alterations that may lead to broader health effects. Methods: We studied the relationship between violence and childhood lung function in a prospective birth cohort of 313 urban children (age range = 6-7 years). Mothers reported on their child's lifetime exposure to community violence (ETV) and interparental conflict in the home (Conflict Tactics Scale (CTS)) within I year of the lung function assessment. Results: In linear regression analyses, adjusting for maternal education, child's age, race, birthweight, tobacco smoke exposure, and medical history, girls in the highest CTS verbal aggression tertile had a 5.5% (95% confidence interval (CI) = -9.6, -1.5) decrease in percent predicted forced expiratory volume (FEV,) and a 5.4% (95% CI = -9.7, -1.1) decrease in forced vital capacity (FVC) compared with girls in the lowest tertile. The CTS verbal aggression subscale was associated with lung function among boys in the same direction, albeit this was not statistically significant. Boys in the highest ETV tertile had a 3.4% (95% CI = -8.0, 1.1) lower FEV, and 5.3% lower FVC (95% Cl = -10.2, -0.4) compared with boys in the lowest tertile. The ETV score was not a significant predictor of girls' lung function. Conclusions: Interparental conflict, specifically verbal aggression, and ETV were associated with decreased childhood lung function independent of socioeconomic status, tobacco smoke exposure, birthweight, and respiratory illness history. Gender differences were noted based on the type of violence exposure, which may warrant further exploration.
Beard, J., Earnest, A., Morgan, G., Chan, H., Summerhayes, R., Dunn, T., Tomaska, N. & Ryan, L.M. 2008, 'Socioeconomic Disadvantage And Acute Coronary Events - A Spatiotemporal Analysis', Epidemiology, vol. 19, no. 3, pp. 485-492.
View/Download from: UTS OPUS or Publisher's site
Background: The associations between socioeconomic disadvantage and ischemic heart disease are not well understood. We explore the relationship between socioeconomic factors and acute coronary events using spatiotemporal analysis. Methods: We studied all deaths from acute myocardial infarction and hospital admissions for acute coronary syndrome and related revascularization procedures for the state of New South Wales, Australia, from 1996 through 2002. We used conditional autoregressive models to describe how characteristics of subjects place of residence (socioeconomic disadvantage, proportion of the population of indigenous background, and metropolitan versus nonmetropolitan area) influenced admissions and mortality. Results: There were 32,534 deaths due to acute myocardial infarction and 129,045 admissions for acute coronary syndrome. We found a relationship between increasing socioeconomic disadvantage and mortality (unadjusted relative risk for highest quartile of disadvantage relative to lowest 1.40; 95% confidence interval 1.271.54) as well as admissions (1.41; 1.28 1.55). After accounting for admission rates, socioeconomic disadvantage was associated with lower rates of angiography (0.75; 0.63 0.88) and interventional angiography (0.70; 0.560.85). After adjusting for socioeconomic disadvantage, areas with higher proportions of the population identified as indigenous had higher rates of admission and mortality, while residency in the state capital was associated with higher admission rates and more interventional angiography. After accounting for admission rates, the association of socioeconomic disadvantage with mortality was reduced.
Forno, E., Onderdonk, A.B., McCracken, J., Litonjua, A.A., Laskey, D., Delaney, M.L., DuBois, A.M., Gold, D.R., Ryan, L.M., Weiss, S.T. & Celedón, J.C. 2008, 'Diversity of the gut microbiota and eczema in early life', Clinical and Molecular Allergy, vol. 6.
View/Download from: Publisher's site
Clark, D., Ryan, L.M. & Lucas, F. 2007, 'A Multi-state Piecewise Exponential Model Of Hospital Outcomes After Injury', Journal of Applied Statistics, vol. 34, no. 10, pp. 1225-1239.
View/Download from: Publisher's site
To allow more accurate prediction of hospital length of stay (LOS) after serious injury or illness, a multi-state model is proposed, in which transitions from the hospitalized state to three possible outcome states (home, long-term care, or death) are assumed to follow constant rates for each of a limited number of time periods. This results in a piecewise exponential (PWE) model for each outcome. Transition rates may be affected by time-varying covariates, which can be estimated from a reference database using standard statistical software and Poisson regression. A PWE model combining the three outcomes allows prediction of LOS. Records of 259,941 injured patients from the US Nationwide Inpatient Sample were used to create such a multi-state PWE model with four time periods. Hospital mortality and LOS for patient subgroups were calculated from this model, and time-varying covariate effects were estimated. Early mortality was increased by anatomic injury severity or penetrating mechanism, but these effects diminished with time; age and male sex remained strong predictors of mortality in all time periods. Rates of discharge home decreased steadily with time, while rates of transfer to long-term care peaked at five days. Predicted and observed LOS and mortality were similar for multiple subgroups. Conceptual background and methods of calculation are discussed and demonstrated. Multi-state PWE models may be useful to describe hospital outcomes, especially when many patients are not discharged home.
Surkan, P.J., Ryan, L.M., Vieira, L.M., Berkman, L.F. & Peterson, K.E. 2007, 'Maternal social and pyschological conditions and physical growth in low-income children in Piaui, Northeast Brazil', Social Science & Medicine, vol. 64, no. 2, pp. 375-388.
View/Download from: UTS OPUS or Publisher's site
Prevalence of child undernutrition remains high in many developing countries. In settings with scarce resources, modifiable maternal social conditions may influence feeding and parenting practices, in turn affecting child growth. This study aims to quantify the association between maternal social support and depression to children's physical growth outcomes in Teresina, Piaui, northeast Brazil. Interviews were conducted with a random sample of 595 mothers of children 6-24 months old in four low-income sections of Teresina, Piaui. We collected data on sociodemographic factors, mothers' social support, mothers' depressive symptomatology, and child's weight and recumbent length. Weight-for-height z-scores (WHZ), height-for-age z-scores (HAZ) and weight-for-age z-scores (WAZ) were calculated using the National Center for Chronic Disease Prevention and Health Promotion Center SAS program based on the 2000 Centers for Disease Control reference growth curves. Multivariable linear regression was used to model the association between maternal social support and depression to child growth, adjusting for biological and socio-demographic variables. Interviewer and neighborhood variation was accounted for through the inclusion of random effects. In adjusted models, material support, measured by number of friends or family members available to mothers when needing food or milk, was related to 0.3 higher average WHZ and 0.2 higher average WAZ in their children. Maternal positive social interaction, which reflects engagement in leisure-time activities with others, was associated with 0.3 higher average WHZ. Mothers' affectionate support was related to 0.2 higher average children's WHZ and WAZ, whereas social support for resolving a conflict was associated with 0.2 lower average HAZ. Maternal depression was not associated with child growth. It is concluded that inadequate growth in children may be sensitive to maternal social support.
Houseman, E.A., Marsit, C., Karagas, M. & Ryan, L.M. 2007, 'Penalized item response theory models: Application to epigenetic alterations in bladder cancer', Biometrics, vol. 63, no. 4, pp. 1269-1277.
View/Download from: UTS OPUS or Publisher's site
Increasingly used in health-related applications, latent variable models provide an appealing framework for handling high-dimensional exposure and response data. Item response theory (IRT) models, which have gained widespread popularity, were originally developed for use in the context of educational testing, where extremely large sample sizes permitted the estimation of a moderate-to-large number of parameters. In the context of public health applications, smaller sample sizes preclude large parameter spaces. Therefore, we propose a penalized likelihood approach to reduce mean square error and improve numerical stability. We present a continuous family of models, indexed by a tuning parameter, that range between the Rasch model and the HIT model. The tuning parameter is selected by cross validation or approximations such as Akaike Information Criterion. While our approach can be placed easily in a Bayesian context, we find that our frequentist approach is more computationally efficient. We demonstrate our methodology on a study of methylation silencing of gene expression in bladder tumors. We obtain similar results using both frequentist and Bayesian approaches, although the frequentist approach is less computationally demanding. In particular, we find high correlation of methylation silencing among 16 loci in bladder tumors, that methylation is associated with smoking and also with patient survival.
Mccarty, K., Smith, T., Zhou, W., Gonzalez, E., Quamruzzaman, Q., Mahiuddin, G., Ryan, L.M., Su, L. & Christiani, D. 2007, 'Polymorphisms In Xpd (asp312asn And Lys751gln) Genes, Sunburn And Arsenic-related Skin Lesions', Carcinogenesis, vol. 28, no. 8, pp. 1697-1702.
View/Download from: Publisher's site
Background: Single-nucleotide polymorphisms in genes related to DNA repair capacity and ultraviolet exposure have not been well investigated in relation to skin lesions associated with arsenic exposure. This population based case-control study, of 600 ca
Celedon, J., Milton, D., Ramsey, C., Litonjua, A., Ryan, L.M., Platts-mills, T. & Gold, D. 2007, 'Exposure To Dust Mite Allergen And Endotoxin In Early Life And Asthma And Atopy In Childhood', Journal Of Allergy And Clinical Immunology, vol. 120, no. 1, pp. 144-149.
View/Download from: UTS OPUS or Publisher's site
Background: There has been no longitudinal study of the relation between concurrent exposure to dust mite allergen and endotoxin in early life and asthma and atopy at school age. Objectives: To examine the relation between exposure to dust mite allergen and endotoxin at age 2 to 3 months and asthma, wheeze, and atopy in high-risk children. Methods: Birth cohort study of 440 children with parental history of atopy in the Boston metropolitan area. Results: In multivariate analyses, early exposure to high levels of dust mite allergen (>= 10 mu g/g) was associated with increased risks of asthma at age 7 years (odds ratio [OR], 3.0; 95% CI, 1.1-7.9) and late-onset wheeze (OR, 5.0; 95% Cl, 1.5-16.4). Exposure to endotoxin levels above the lowest quartile at age 2 to 3 months was associated with reduced odds of atopy at school age (OR, 0.5; 95% CI, 0.2-0.9). In contrast with its inverse association with atopy, endotoxin exposure in early life was associated with an increased risk of any wheeze between ages I and 7 years that did not change significantly with time (hazard ratio for each quartile increment in endotoxin levels, 1.23; 95% CI, 1.07-1.43). Conclusion: Among children at risk of atopy, early exposure to high levels of dust mite allergen is associated with increased risks of asthma and late-onset wheeze. In these children, endotoxin exposure is associated with a reduced risk of atopy but an increased risk of wheeze. Clinical implications: Early endotoxin exposure may be a protective factor against atopy but a risk factor for wheeze in high-risk children.
Mccarty, K., Chen, Y., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Hsueh, Y., Su, L., Smith, T., Ryan, L.M. & Christiani, D. 2007, 'Arsenic Methylation, Gstt1, Gstm1, Gstp1 Polymorphisms, And Skin Lesions', Environmental Health Perspectives, vol. 115, no. 3, pp. 341-345.
View/Download from: UTS OPUS or Publisher's site
OBJECTIVE: We investigated whether primary and secondary arsenic methylation ratios were associated with skin lesions and whether GSTT1, GSTP1, and GSTM1 polymorphisms modify these relationships. METHODS: A case-control study of 600 cases and 600 controls that were frequency matched on age and sex was conducted in Pabna, Bangladesh, in 2001-2002. Individual well water, urine, and blood samples were collected. Water arsenic concentration was determined using inductively coupled plasma mass spectrometry (ICP-MS). Urinary arsenic speciation was determined using high performance liquid chromatography hydride with generator atomic absorption spectrometry and ICP-MS. Genotyping was conducted using multiplex polymerase chain reaction and TaqMan. RESULTS: A 10-fold increase in primary methylation ratio [monomethylarsonic acid (MMA)/(arsenite + arsenate] was associated with a 1.50-fold increased risk of skin lesions (multivariate odds ratio = 1.50; 95% confidence interval, 1.00-2.26). We observed significant interaction on the multiplicative scale between GSTT1 wildtype and secondary methylation ratio [dimethylarsinic acid/MMA; likelihood ratio test (LRT), p = 0.01]. No significant interactions were observed for GSTM1 or GSTP1 or for primary methylation ratios. CONCLUSION: Our findings suggest that increasing primary methylation ratios are associated with an increase in risk of arsenic-related skin lesions. The interaction between GSTT1 wildtype and secondary methylation ratio modifies risk of skin lesions among arsenic-exposed individuals
Ryan, L.M., Scirica, C.V., Gold, D.R., Abulkerim, H., Celedon, J.C., Platts-Mills, T.A., Naccara, L.M., Weiss, S.T. & Litonjua, A.A. 2007, 'Predictors of cord blood IgE levels in children at risk for asthma and atopy', Journal of Allergy and Clinical Immunology, vol. 119, no. 1, pp. 81-88.
View/Download from: UTS OPUS or Publisher's site
Background: Increased cord blood IgE levels, in conjunction with a family history of atopy, are associated with the development of allergic diseases in children. However, little is known about predictors of cord blood IgE levels. Objective: Our objective was to identify predictors of cord blood IgE levels among infants at increased risk of atopy. Methods: Cord blood IgE levels were measured in 874 infants who were screened for participation in a birth cohort. Questionnaires were administered after birth of the infant, and maternal and cord blood was obtained for measurement of IgE levels. Logistic and tobit regression models were used to study the association between perinatal factors and cord blood IgE levels. Results: In multivariable models infant male sex, increased maternal total IgE level, maternal allergen sensitization, Hispanic ethnicity, and residence in low-income areas were associated with detectable or increased cord blood IgE levels, whereas increasing maternal age was associated with undetectable or lower cord blood IgE levels. Although maternal smoking during pregnancy was positively associated with cord blood IgE levels in univariable models, the effect did not persist after adjusting for potential confounders. Conclusion: Maternal allergen sensitization, markers of socioeconomic disadvantage and race/ethnicity, maternal age, and infant sex might influence fetal production of IgE. We found no association of maternal parity, mode of delivery, gestational age, or season of birth with cord blood IgE levels. Clinical implications: The identification of these definable familial and environmental factors that predict cord blood IgE levels might help in the early detection of infants at risk for atopic disorders.
Earnest, A., Morgan, G., Mengersen, K., Ryan, L.M., Summerhayes, R. & Beardall, J. 2007, 'Evaluating the effect of neighbourhood weight matrices on smoothing properties of Conditional Autoregressive (CAR) models', International Journal of Health Geographics, vol. 6, no. 54.
Axelrad, D.A., Bellinger, D.C., Ryan, L.M. & Woodruff, T.J. 2007, 'Dose-response relationship of prenatal mercury exposure and IQ: an integrative analysis of epidemiologic data', Environmental Health Perspectives, vol. 115, no. 4, pp. 609-615.
View/Download from: UTS OPUS or Publisher's site
BACKGROUND: Prenatal exposure to mercury has been associated with adverse childhood neurologic outcomes in epidemiologic studies. Dose-response information for this relationship is useful for estimating benefits of reduced mercury exposure. OBJECTIVES: We estimated a dose-response relationship between maternal mercury body burden and subsequent childhood decrements in intelligence quotient (IQ), using a Bayesian hierarchical model to integrate data from three epidemiologic studies. METHODS: Inputs to the model consist of dose-response coefficients from studies conducted in the Faroe Islands, New Zealand, and the Seychelles Islands. IQ coefficients were available from previous work for the latter two studies, and a coefficient for the Faroe Islands study was estimated from three IQ subtests. Other tests of cognition/achievement were included in the hierarchical model to obtain more accurate estimates of study-to-study and end point-to-end point variability. RESULTS: We find a central estimate of -0.18 IQ points (95% confidence interval, -0.378 to -0.009) for each parts per million increase of maternal hair mercury, similar to the estimates for both the Faroe Islands and Seychelles studies, and lower in magnitude than the estimate for the New Zealand study. Sensitivity analyses produce similar results, with the IQ coefficient central estimate ranging from -0.13 to -0.25. CONCLUSIONS: IQ is a useful end point for estimating neurodevelopmental effects, but may not fully represent cognitive deficits associated with mercury exposure, and does not represent deficits related to attention and motor skills. Nevertheless, the integrated IQ coefficient provides a more robust description of the dose-response relationship for prenatal mercury exposure and cognitive functioning than results of any single study.
McCarty, K.M., Ryan, L.M., Houseman, E.A., Williams, P.L., Miller, D.P., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Smith, T., Gonzalez, E., Su, L. & Christiani, D.C. 2007, 'A case-control study of GST polymorphisms and arsenic related skin lesions', Environmental health, vol. 6, no. 5, pp. 1-10.
View/Download from: UTS OPUS
Background: Polymorphisms in GSTT1, GSTM1 and GSTP1 impact detoxification of carcinogens by GSTs and have been reported to increase susceptibility to environmentally related health outcomes. Individual factors in arsenic biotransformation may influence disease susceptibility. GST activity is involved in the metabolism of endogenous and exogenous compounds, including catalyzing the formation of arsenic-GSH conjugates. Methods: We investigated whether polymorphisms in GSTT1, GSTP1 and GSTM1 were associated with risk of skin lesions and whether these polymorphisms modify the relationship between drinking water arsenic exposure and skin lesions in a case control study of 1200 subjects frequency matched on age and gender in community clinics in Pabna, Bangladesh in 2001-2002. Results and discussion: GSTT1 homozygous wildtype status was associated with increased odds of skin lesions compared to the null status (OR1.56 95% CI 1.10-2.19). The GSTP1 GG polymorphism was associated with greater odds of skin lesions compared to GSTP1 AA, (OR 1.86 ( 95% CI 1.15-3.00). No evidence of effect modification by GSTT1, GSTM1 or GSTP1 polymorphisms on the association between arsenic exposure and skin lesions was detected. Conclusion: GSTT1 wildtype and GSTP1 GG are associated with increased risk of skin lesions.
Ramsey, C.D., Gold, D.R., Litonjua, A.A., Sredl, D.L., Ryan, L.M. & Celedon, J.C. 2007, 'Respiratory illnesses in early life and asthma and atopy in childhood', Journal Of Allergy And Clinical Immunology, vol. 119, no. 1, pp. 150-156.
View/Download from: UTS OPUS or Publisher's site
Background The relation between respiratory illnesses in early life and the development of asthma and atopy in childhood is incompletely understood. Objective We sought to examine the relationship between respiratory illnesses in early life and atopic diseases at school age. Methods We performed a prospective birth cohort study of the relationship between respiratory illnesses in the first year of life and asthma, atopy (sensitization to =1 allergen), and allergic rhinitis at school age in 440 children with a parental history of atopy. Logistic regression was used to examine the relationship between respiratory illnesses and asthma, atopy, and allergic rhinitis. The relationship between respiratory illnesses in early life and repeated measures of wheezing between the ages of 1 and 7 years was investigated by using a proportional hazards models. Results Physician-diagnosed croup (adjusted odds ratio [OR], 0.30; 95% CI, 0.12-0.72) and having 2 or more physician-diagnosed ear infections (adjusted OR, 0.58; 95% CI, 0.35-0.98) in the first year of life were inversely associated with atopy at school age. Physician-diagnosed bronchiolitis before age 1 year was significantly associated with asthma at age 7 years (adjusted OR, 2.77; 95% CI, 1.23-6.22). Recurrent nasal catarrh (=3 episodes of a runny nose) in the first year of life was associated with allergic rhinitis at age 7 years (adjusted OR, 2.99; 95% CI, 1.03-8.67). Conclusion The relationship between early-life respiratory illnesses and asthma and atopy is complex and likely dependent on the type of infection and immune response it initiates.
Hauser, R., Meeker, J.D., Singh, N.P., Silva, M.J., Ryan, L.M., Duty, S. & Calafat, A.M. 2007, 'DNA damage in human sperm is related to urinary levels of phthalate monoester and oxidative metabolites', Human Reproduction, vol. 22, no. 3, pp. 688-695.
View/Download from: Publisher's site
BACKGROUND: The ubiquitous use of phthalate esters in plastics, personal care products and food packaging materials results in widespread general population exposure. In this report, we extend our preliminary study on the relationship between urinary concentrations of phthalate metabolites and sperm DNA damage among a larger sample of men and include measurements of mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHAIR) and mono-(2-ethyl-5-oxohexyl) phthalate (MEOHP), two oxidative metabolites of di-(2-ethylhexyl) phthalate (DEHP). METHODS: Among 379 men from an infertility clinic, urinary concentrations of phthalate metabolites were measured using isotope-dilution high-performance liquid chromatography-tandem mass spectrometry. Sperm DNA damage measurements, assessed with the neutral comet assay, included comet extent (CE), percentage of DNA in tail (Tail%) and tail distributed moment (TDM). RESULTS: Monoethyl phthalate (MEP), a metabolite of diethyl phthalate, was associated with increased DNA damage, confirming our previous findings. Mono-(2-ethylhexyl) phthalate (MERP), a metabolite of DEHP, was associated with DNA damage after adjustment for the oxidative DEHP metabolites. After adjustment for MEHHP, for an interquartile range increase in urinary MEHP, CE increased 17.3% [95% confidence interval (CI) = 8.7-25.7%], TDM increased 14.3% (95% CI = 6.8-21.7%) and Tail% increased 17.5% (95% CI = 3.5-31.5%). CONCLUSIONS: Sperm DNA damage was associated with MEP and with MEHP after adjusting for DEHP oxidative metabolites, which may serve as phenotypic markers of DEHP metabolism to 'less toxic' metabolites. The urinary levels of phthalate metabolites among these men were similar to those reported for the US general population, suggesting that exposure to some phthalates may affect the population distribution of sperm DNA damage.
Choi, A.L., Levy, J.I., Dockery, D.W., Ryan, L.M., Tolbert, P.E., Altshul, L.M. & Korrick, S.A. 2006, 'Does living near a Superfund site contribute to higher polychlorinated biphenyl (PCB) exposure?', Environmental Health Perspectives, vol. 114, no. 7, pp. 1092-1098.
View/Download from: UTS OPUS or Publisher's site
We assessed determinants of cord serum polychlorinated biphenyl (PCB) levels among 720 infants born between 1993 and 1998 to mothers living near a PCB-contaminated Superfund site in Massachusetts, measuring the sum of 51 PCB congeners (SPCB) and ascertaining maternal address, diet, sociodemographics, and exposure risk factors. Addresses were geocoded to obtain distance to the Superfund site and neighborhood characteristics. We modeled log10(SPCB) as a function of potential individual and neighborhood risk factors, mapping model residuals to assess spatial correlates of PCB exposure. Similar analyses were performed for light (mono-tetra) and heavy (penta-deca) PCBs to assess potential differences in exposure pathways as a function of relative volatility. PCB-118 (relatively prevalent in site sediments and cord serum) was assessed separately. The geometric mean of SPCB levels was 0.40 (range, 0.06818.14) ng/g serum. Maternal age and birthplace were the strongest predictors of SPCB levels. Maternal consumption of organ meat and local dairy products was associated with higher and smoking and previous lactation with lower SPCB levels. Infants born later in the study had lower SPCB levels, likely due to temporal declines in exposure and site remediation in 19941995. No association was found between SPCB levels and residential distance from the Superfund site. Similar results were found with light and heavy PCBs and PCB-118. Previously reported demographic (age) and other (lactation, smoking, diet) correlates of PCB exposure, as well as local factors (consumption of local dairy products and Superfund site dredging) but not residential proximity to the site, were important determinants of cord serum PCB levels in the study community.
Houseman, E.A., Coull, B.A. & Ryan, L.M. 2006, 'A functional-based distribution diagnostic for a linear model with correlated outcomes', Biometrika, vol. 93, no. 4, pp. 911-926.
View/Download from: UTS OPUS or Publisher's site
In this paper we present an easy-to-implement graphical distribution diagnostic for linear models with correlated errors. Houseman et al. (2004) constructed quantile-quantile plots for the marginal residuals of such models, suitably transformed. We extend the pointwise asymptotic theory to address the global stochastic behaviour of the corresponding empirical cumulative distribution function, and describe a simulation technique that serves as a computationally efficient parametric bootstrap for generating representatives of its stochastic limit. Thus, continuous functionals of the empirical cumulative distribution function may be used to form global tests of normality. Through the use of projection matrices, we generalised our methods to include tests that are directed at assessing the normality of particular components of the error. Thus, tests proposed by Lange & Ryan (1989) follow as a special case. Our method works well both for models having independent units of sampling and for those in which all observations are correlated.
Louis, G.B., Dukic, V., Heagerty, P.J., Louis, T.A., Lynch, C.D., Ryan, L.M., Schisterman, E.F. & Trumble, A. 2006, 'Analysis of repeated pregnancy outcomes', Statistical Methods in Medical Research, vol. 15, no. 2, pp. 103-126.
View/Download from: UTS OPUS or Publisher's site
Women tend to repeat reproductive outcomes, with past history of an adverse outcome being associated with an approximate two-fold increase in subsequent risk. These observations support the need for statistical designs and analyses that address this clus
Morales, K.H., Ibrahim, J.G., Chen, C. & Ryan, L.M. 2006, 'Bayesian model averaging with applications to benchmark dose estimation for arsenic in drinking water', Journal Of The American Statistical Association, vol. 101, no. 473, pp. 9-17.
View/Download from: UTS OPUS or Publisher's site
An important component of quantitative risk assessment involves characterizing the dose-response relationship between an environmental exposure and adverse health outcome and then computing a benchmark dose, or the exposure level that yields a suitably low risk. This task is often complicated by model choice considerations, because risk estimates depend on the model parameters. We pro pose using Bayesian methods to address the problem of model selection and derive a model-averaged version of the benchmark dose. We illustrate the methods through application to data on arsenic-induced lung cancer from Taiwan.
Davis, E., Smith, T.J., Laden, F., Hart, J.E., Ryan, L.M. & Garshick, E. 2006, 'Modeling particle exposure in U.S. trucking terminals', Environmental Science & Technology, vol. 40, no. 13, pp. 4226-4232.
View/Download from: UTS OPUS or Publisher's site
Multi-tiered sampling approaches are common in environmental and occupational exposure assessment, where exposures for a given individual are often modeled based on simultaneous measurements taken at multiple indoor and outdoor sites. The monitoring data from such studies is hierarchical by design, imposing a complex covariance structure that must be accounted for in order to obtain unbiased estimates of exposure. Statistical methods such as structural equation modeling (SEM) represent a useful alternative to simple linear regression in these cases, providing simultaneous and unbiased predictions of each level of exposure based on a set of covariates specific to the exposure setting. We test the SEM approach using data from a large exposure assessment of diesel and combustion particles in the U. S. trucking industry. The exposure assessment includes data from 36 different trucking terminals across the United States sampled between 2001 and 2005, measuring PM2.5 and its elemental carbon (EC), organic carbon (OC) components, by personal monitoring, and sampling at two indoor work locations and an outdoor "background" location. Using the SEM method, we predict the following: (1) personal exposures as a function of work-related exposure and smoking status; (2) work-related exposure as a function of terminal characteristics, indoor ventilation, job location, and background exposure conditions; and ( 3) background exposure conditions as a function of weather, nearby source pollution, and other regional differences across terminal sites. The primary advantage of SEMs in this setting is the ability to simultaneously predict exposures at each of the sampling locations, while accounting for the complex covariance structure among the measurements and descriptive variables. The statistically significant results and high R-2 values observed from the trucking industry application supports the broader use of this approach in exposure assessment modeling.
Li, Y. & Ryan, L.M. 2006, 'Inference on survival data with covariate measurement error - an imputation approach', Scandinavian Journal of Statistics, vol. 33, no. 2, pp. 169-190.
View/Download from: UTS OPUS or Publisher's site
We propose a new method for fitting proportional hazards models with error-prone covariates. Regression coefficients are estimated by solving an estimating equation that is the average of the partial likelihood scores based on imputed true covariates. For the purpose of imputation, a linear spline model is assumed on the baseline hazard. We discuss consistency and asymptotic normality of the resulting estimators, and propose a stochastic approximation scheme to obtain the estimates. The algorithm is easy to implement, and reduces to the ordinary Cox partial likelihood approach when the measurement error has a degenerate distribution. Simulations indicate high efficiency and robustness. We consider the special case where error-prone replicates are available on the unobserved true covariates. As expected, increasing the number of replicates for the unobserved covariates increases efficiency and reduces bias. We illustrate the practical utility of the proposed method with an Eastern Cooperative Oncology Group clinical trial where a genetic marker, c-myc expression level, is subject to measurement error.
McCarty, K.M., Houseman, E.A., Quamruzzaman, Q., Rahman, M., Mahiuddin, G., Smith, T., Ryan, L.M. & Christiani, D.C. 2006, 'The impact of diet and betel nut use on skin lesions associated with drinking-water arsenic in Pabna, Bangladesh', Environmental Health Perspectives, vol. 114, no. 3, pp. 334-340.
View/Download from: UTS OPUS or Publisher's site
An established exposure-response relationship exists between water arsenic levels and skin lesions. Results of previous studies with limited historical exposure data, and laboratory animal studies suggest that diet may modify arsenic metabolism and toxicity. In this study, we evaluated the effect of diet on the risk of arsenic-related skin lesions in Pabna, Bangladesh. Six hundred cases and 600 controls loosely matched on age and sex were enrolled at Dhaka Community Hospital, Bangladesh, in 2001-2002. Diet, demographic data, and water samples were collected. Water samples were analyzed for arsenic using inductively coupled plasma mass spectroscopy. Betel nut use was associated with a greater risk of skin lesions in a multivariate model [odds ratio (OR)=1.67; 95% confidence interval (CI), 1.18-2.36]. Modest decreases in risk of skin lesions were associated with fruit intake 1-3 times/month (OR=0.68; 95% CI, 0.51-0.89) and canned goods at least 1 time/month (OR=0.41; 95% CI, 0.20-0.86). Bean intake at least 1 time/day (OR=1.89; 95% CI, 1.11-3.22) was associated with increased odds of skin lesions. Betel nut use appears to be associated with increased risk of developing skin lesions in Bangladesh. Increased intake of fruit and canned goods may be associated with reduced risk of lesions. Increased intake of beans may be associated with an increased risk of skin lesions. The results of this study do not provide clear support for a protective effect of vegetable and overall protein consumption against the development of skin lesions, but a modest benefit cannot be excluded.
Morris, J.S., Arroyo, C., Coull, B.A., Ryan, L.M., Herrick, R. & Gortmaker, S.L. 2006, 'Using wavelet-based functional mixed models to characterize population heterogeneity in accelerometer profiles: A case study', Journal of the American Statistical Association, vol. 101, no. 476, pp. 1352-1364.
View/Download from: UTS OPUS or Publisher's site
We present a case study illustrating the challenges of analyzing accelerometer data taken from a sample of children participating in an intervention study designed to increase physical activity. An accelerometer is a small device worn on the hip that records the minute-by-minute activity levels throughout the day for each day it is worn. The resulting data are irregular functions characterized by many peaks representing short bursts of intense activity. We model these data using the wavelet-based functional mixed model. This approach incorporates multiple fixed-effects and random-effects functions of arbitrary form, the estimates of which are adaptively regularized using wavelet shrinkage. The method yields posterior samples for all functional quantities of the model, which can be used to perform various types of Bayesian inference and prediction. In our case study, a high proportion of the daily activity profiles are incomplete (i.e., have some portion of the profile missing), and thus cannot be modeled directly using the previously described method. We present a new method for stochastically imputing the missing data that allows us to incorporate these incomplete profiles in our analysis. Our approach borrows strength from both the observed measurements within the incomplete profiles and from other profiles, from the same child as well as from other children with similar covariate levels, while appropriately propagating the uncertainty of the imputation throughout all subsequent inference. We apply this method to our case study, revealing some interesting insights into children's activity patterns. We point out some strengths and limitations of using this approach to analyze accelerometer data.
Gold, D.R., Willwerth, B.M., Tantisira, K.G., Finn, P.W., Schaub, B., Perkins, D.L., Tzianabos, A., Ly, N.P., Schroeter, C., Gibbons, F., Campos, H., Oken, E., Gillman, M.W., Palmer, L.J., Ryan, L.M. & Weiss, S.T. 2006, 'Associations of cord blood fatty acids with lymphocyte proliferation, IL-13, and IFN-gamma', Journal of Allergy and Clinical Immunology, vol. 117, no. 4, pp. 931-938.
View/Download from: UTS OPUS or Publisher's site
Background N-3 and n-6 polyunsaturated fatty acids (PUFAs) have been hypothesized to have opposing influences on neonatal immune responses that might influence the risk of allergy or asthma. However, both n-3 eicosapentaenoic acid (EPA) and n-6 arachidonic acid (AA) are required for normal fetal development. Objective We evaluated whether cord blood fatty acid levels were related to neonatal immune responses and whether n-3 and n-6 PUFA responses differed. Methods We examined the relation of cord blood plasma n-3 and n-6 PUFAs (n = 192) to antigen- and mitogen-stimulated cord blood lymphocyte proliferation (n = 191) and cytokine (IL-13 and IFN-?; n = 167) secretion in a US birth cohort. Results Higher levels of n-6 linoleic acid were correlated with higher IL-13 levels in response to Bla g 2 (cockroach, P = .009) and Der f 1 (dust mite, P = .02). Higher n-3 EPA and n-6 AA levels were each correlated with reduced lymphocyte proliferation and IFN-? levels in response to Bla g 2 and Der f 1 stimulation. Controlling for potential confounders, EPA and AA had similar independent effects on reduced allergen-stimulated IFN-? levels. If neonates had either EPA or AA levels in the highest quartile, their Der f 1 IFN-? levels were 90% lower (P = .0001) than those with both EPA and AA levels in the lowest 3 quartiles. Reduced AA/EPA ratio was associated with reduced allergen-stimulated IFN-? level. Conclusion Increased levels of fetal n-3 EPA and n-6 AA might have similar effects on attenuation of cord blood lymphocyte proliferation and IFN-? secretion. Clinical implications The implications of these findings for allergy or asthma development are not yet known.
Meeker, J.D., Ryan, L., Barr, D.B. & Hauser, R. 2006, 'Exposure to nonpersistent insecticides and male reproductive hormones', EPIDEMIOLOGY, vol. 17, no. 1, pp. 61-68.
View/Download from: Publisher's site
Ly, N.P., Ruiz-Pérez, B., Onderdonk, A.B., Tzianabos, A.O., Litonjua, A.A., Liang, C., Laskey, D., Delaney, M.L., DuBois, A.M., Levy, H., Gold, D.R., Ryan, L.M., Weiss, S.T. & Celedón, J.C. 2006, 'Mode of delivery and cord blood cytokines: A birth cohort study', Clinical and Molecular Allergy, vol. 4.
View/Download from: Publisher's site
Harezlak, J., Ryan, L.M., Giedd, J.N. & Lange, N. 2005, 'Individual and population penalized regression splines for accelerated longitudinal designs', Biometrics, vol. 61, no. 4, pp. 1037-1048.
View/Download from: UTS OPUS or Publisher's site
In an accelerated longitudinal design (ALD), individuals enter the study at different points of their growth trajectory and are observed over a short time span relative to the entire time span of interest. ALD data are combined across independent units to provide an estimate of an overall population curve and predictions of individual patterns of change. As a modest extension of the work of Ruppert et al. (2003, Semiparametric Regression, Cambridge University Press), we develop a computationally efficient procedure for the application of longitudinal serniparametric methods under ALD sampling schemes. We compare balanced and complete longitudinal designs to ALDs using the Berkeley Growth Study data and apply our method to longitudinal magnetic resonance imaging (MRI) brain structure size (volume) measurements from an ongoing developmental study. Potential applications extend beyond growth studies to many other fields in which cost and feasibility constraints impose restrictions on sample size and on the numbers and timings of repeated measurements across subjects.
Sánchez, B.N., Budtz-Jørgensen, E., Ryan, L.M. & Hu, H. 2005, 'Structural equation models: A review with applications to environmental epidemiology', Journal of the American Statistical Association, vol. 100, no. 472, pp. 1443-1455.
View/Download from: UTS OPUS or Publisher's site
Structural equation models (SEMs) have been discussed extensively in the psychometrics and quantitative behavioral sciences literature. However, many statisticians and researchers in other areas of application are relatively unfamiliar with their implementation. Here we review some of the SEM literature and describe basic methods, using examples from environmental epidemiology. We make connections to recent work on latent variable models for multivariate outcomes and to measurement error methods, and discuss advantages and disadvantages of SEMs compared with traditional regressions. We give a detailed example in which two models fit the same data well, yet one is physiologically implausible. This underscores the critical role of subject matter knowledge in the successful implementation of SEMs. A brief discussion on open research areas is included.
Bellamy, S.L., Li, Y., Lin, X. & Ryan, L.M. 2005, 'Quantifying PQL bias in estimating cluster-level covariate effects in genearlized linear mixed models for group-randomized trials', Statistica Sinica, vol. 15, no. 4, pp. 1015-1032.
View/Download from: UTS OPUS
We derive the asymptotic bias and variance of the penalized quasilikelihood (PQL) estimator of the cluster-level covariate effect in generalized linear mixed models for group-randomized trials where the number of clusters n is small and the cluster size m is large. We show that the asymptotic bias is of order O-p(1/m) and the asymptotic variance is of order Op(1/n) + Op{1/(nm)}. The practical implication of our results is that the PQL method works well in settings involving small numbers of large clusters which are typical in grouped randomized trials. We illustrate the results using simulation studies.
Abraham, J.H., Gold, D.R., Dockery, D.W., Ryan, L.M., Park, J. & Milton, D.K. 2005, 'Within-home versus between-home variability of house dust endotoxin in a birth cohort', Environmental Health Perspectives, vol. 113, no. 11, pp. 1516-1521.
View/Download from: UTS OPUS or Publisher's site
Endotoxin exposure has been proposed as an environmental determinant of allergen responses in children. To better understand the implications of using a single measurement of house dust endotoxin to characterize exposure in the first year of life, we evaluated room-specific within-home and between-home variability in dust endotoxin obtained from 470 households in Boston, Massachusetts. Homes were sampled up to two times over 511 months. We analyzed 1,287 dust samples from the kitchen, family room, and babys bedroom for endotoxin. We fit a mixed-effects model to estimate mean levels and the variation of endotoxin between homes, between rooms, and between sampling times. Endotoxin ranged from 2 to 1,945 units per milligram of dust. Levels were highest during summer and lowest in the winter. Mean endotoxin levels varied significantly from room to room. Cross-sectionally, endotoxin was moderately correlated between family room and bedroom floor (r = 0.30), between family room and kitchen (r = 0.32), and between kitchen and bedroom (r = 0.42). Adjusting for season, the correlation of endotoxin levels within homes over time was 0.65 for both the bedroom and kitchen and 0.54 for the family room. The temporal within-home variance of endotoxin was lowest for bedroom floor samples and highest for kitchen samples. Between-home variance was lowest in the family room and highest for kitchen samples. Adjusting for season, within-home variation was less than between-home variation for all three rooms. These results suggest that room-to-room and home-to-home differences in endotoxin influence the total variability more than factors affecting endotoxin levels within a room over time.
Litonjua, A.A., Celedon, J.C., Hausmann, J., Nikolov, M., Sredl, D., Ryan, L.M., Platts-Mills, T.A., Weiss, S.T. & Gold, D.R. 2005, 'Variation in total and specific IgE: Effects of ethnicity and socioeconomic status', Journal of Allergy and Clinical Immunology, vol. 115, no. 4, pp. 751-757.
View/Download from: UTS OPUS or Publisher's site
Chen, J., Christiani, D.C. & Ryan, L.M. 2005, 'Predicting exposure levels', Epidemiology, vol. 16, no. 1, pp. 135-135.
Burden, S., Guha, S., Morgan, G., Ryan, L.M., Sparks, R. & Young, L. 2005, 'Spatio-temporal analysis of acute admissions for ischemic heart disease in NSW, Australia', Environmental and Ecological Statistics, vol. 12, pp. 427-448.
View/Download from: UTS OPUS or Publisher's site
The recently funded Spatial Environmental Epidemiology in New South Wales (SEE NSW) project aims to use routinely collected data in NSW Australia to investigate risk factors for various chronic diseases. In this paper, we present a case study focused on the relationship between social disadvantage and ischemic heart disease to highlight some of the methodological challenges that are likely to arise.
Pfeiffer, R.M., Ryan, L.M., Litonjua, A. & Pee, D. 2005, 'A case-cohort design for assessing covariate effects in longitudinal studies', Biometrics, vol. 61, no. 4, pp. 982-991.
View/Download from: UTS OPUS or Publisher's site
The case-cohort design for longitudinal data consists of a subcohort sampled at the beginning of the study that is followed repeatedly over time, and a case sample that is ascertained through the course of the study. Although some members in the subcohort may experience events over the study period, we refer to it as the "control-cohort." The case sample is a random sample of subjects not in the control-cohort, who have experienced at least one event during the study period. Different correlations among repeated observations on the same individual are accommodated by a two-level random-effects model. This design allows consistent estimation of all parameters estimable in a cohort design and is a cost-effective way to study the effects of covariates on repeated observations of relatively rare binary outcomes when exposure assessment is expensive. It is an extension of the case-cohort design (Prentice, 1986, Biornetrika 73, 1-11) and the bidirectional case-crossover design (Navidi, 1998, Biometrics 54, 596-605). A simulation study compares the efficiency of the longitudinal case-cohort design to a full cohort analysis, and we find that in certain situations up to 90% efficiency can be obtained with half the sample size required for a full cohort analysis. A bootstrap method is presented that permits testing for intra-subject homogeneity in the presence of unidentifiable nuisance parameters in the two-level random-effects model. As an illustration we apply the design to data from an ongoing study of childhood asthma.
Morales, K.H. & Ryan, L.M. 2005, 'Benchmark dose estimation based on epidemiologic cohort data', Environmetrics, vol. 16, no. 5, pp. 435-447.
View/Download from: UTS OPUS or Publisher's site
Risk assessments based on epidemiologic studies are becoming increasingly common in evaluating environmental health risks and setting health standards. This article will discuss and compare some of the available methods for exposure-response modeling and risk estimation based on environmental epidemiologic studies with age-specific incidence and mortality data. Recommendations will be made regarding approaches that can be used in practice
Stark, P.C., Celedon, J.C., Chew, G.L., Ryan, L.M., Burge, H.A., Muilenberg, M.L. & Gold, D.R. 2005, 'Fungal levels in the home and allergic rhinitis by 5 years of age', Environmental Health Perspectives, vol. 113, no. 10, pp. 1405-1409.
View/Download from: UTS OPUS or Publisher's site
Studies have repeatedly demonstrated that sensitization to fungi, such as Alternaria, is strongly associated with allergic rhinitis and asthma in children. However, the role of exposure to fungi in the development of childhood allergic rhinitis is poorly understood. In a prospective birth cohort of 405 children of asthmatic/allergic parents from metropolitan Boston, Massachusetts, we examined in-home high fungal concentrations (> 90th percentile) measured once within the first 3 months of life as predictors of doctor-diagnosed allergic rhinitis in the first 5 years of life. In multivariate Cox regression analyses, predictors of allergic rhinitis included high levels of dust-borne Aspergillus [hazard ratio (HR) = 3.27; 95% confidence interval (Cl), 1.50-7.14], Aureobasidium (HR = 3.04; 95% Cl, 1.33-6.93), and yeasts (HR = 2.67; 95% Cl, 1.26-5.66). The factors controlled for in these analyses included water damage or mild or mildew in the building during the first year of the child's life, any lower respiratory tract infection in the first year, male sex, African-American race, fall date of birth, and maternal IgE to Alternaria > 0.35 U/mL. Dust-borne Alternaria and non-sporulating and total fungi were also predictors of allergic rhinitis in models excluding other fungi but adjusting for all of the potential confounders listed above. High measured fungal concentrations and reports of water damage, mold, or mildew in homes may predispose children with a family history of asthma or allergy to the development of allergic rhinitis. Key words: allergic rhinitis, fungi, mold, respiratory health effects, water damage.
Meeker, J.D., Barr, D.B., Ryan, L.M., Herrick, R.F., Bennett, D.H., Bravo, R. & Hauser, R. 2005, 'Temporal variability of urinary levels of nonpersistent insecticides in adult men', Journal of Exposure Analysis and Environmental Epidemiology, vol. 15, no. 3, pp. 271-281.
View/Download from: UTS OPUS
Widespread application of contemporary-use insecticides results in low-level exposure for a majority of the population through a variety of pathways. Urinary insecticide biomarkers account for all exposure pathways, but failure to account for temporal within-subject variability of urinary levels can lead to exposure misclassification. To examine temporal variability in urinary markers of contemporary-use insecticides, nine repeated urine samples were collected over 3 months from 10 men participating in an ongoing study of male reproductive health. These 90 samples were analyzed for urinary metabolites of chlorpyrifos (3,5,6-trichloro-2-pyridinol (TCPY)) and carbaryl (1-naphthol (1N)). Volume- based (unadjusted), as well as creatinine (CRE)- and specific gravity (SG)-adjusted concentrations were measured. TCPY had low reliability with an intraclass correlation coefficient between 0.15 and 0.21, while 1N was moderately reliable with an intraclass correlation coefficient between 0.55 and 0.61. When the 10 men were divided into tertiles based on 3-month geometric mean TCPY and 1N levels, a single urine sample performed adequately in classifying a subject into the highest or lowest exposure tertiles. Sensitivity and specificity ranged from 0.44 to 0.84 for TCPY and from 0.56 to 0.89 for 1N. Some differences in the results between unadjusted metabolite concentrations and concentrations adjusted for CRE and SG were observed. Questionnaires were used to assess diet in the 24 h preceding the collection of each urine sample. In mixed-effects models, TCPY was significantly associated with season as well as with consuming grapes and cheese, while 1N levels were associated with consuming strawberries. In conclusion, although a single sample adequately predicted longer-term average exposure, a second sample collected at least 1 month following the first sample would reduce exposure measurement error.
Chen, C., Wang, X., Wang, L., Yang, F., Tang, G., Xing, H., Ryan, L.M., Lasley, B., Overstreet, J.W., Stanford, J.B. & Xu, X. 2005, 'Effect of environmental tobacco smoke on levels of urinary hormone markers', Environmental Health Perspectives, vol. 113, no. 4, pp. 412-417.
View/Download from: UTS OPUS or Publisher's site
Our recent study showed a doseresponse relationship between environmental tobacco smoke (ETS) and the risk of early pregnancy loss. Smoking is known to affect female reproductive hormones. We explored whether ETS affects reproductive hormone profiles as characterized by urinary pregnanediol-3-glucuronide (PdG) and estrone conjugate (E1C) levels. We prospectively studied 371 healthy newly married nonsmoking women in China who intended to conceive and had stopped contraception. Daily records of vaginal bleeding, active and passive cigarette smoking, and daily first-morning urine specimens were collected for up to 1 year or until a clinical pregnancy was achieved. We determined the day of ovulation for each menstrual cycle. The effects of ETS exposure on daily urinary PdG and E1C levels in a &plusmn;10 day window around the day of ovulation were analyzed for conception and nonconception cycles, respectively. Our analysis included 344 nonconception cycles and 329 conception cycles. In nonconception cycles, cycles with ETS exposure had significantly lower urinary E1C levels (= 0.43, SE = 0.08, p < 0.001 in log scale) compared with the cycles without ETS exposure. There was no significant difference in urinary PdG levels in cycles having ETS exposure (= 0.07, SE = 0.15, p = 0.637 in log scale) compared with no ETS exposure. Among conception cycles, there were no significant differences in E1C and PdG levels between ETS exposure and nonexposure. In conclusion, ETS exposure was associated with significantly lower urinary E1C levels among nonconception cycles, suggesting that the adverse reproductive effect of ETS may act partly through its antiestrogen effects.
Abraham, J.H., Finn, P.W., Milton, D.K., Ryan, L.M., Perkins, D.L. & Gold, D.R. 2005, 'Infant home endotoxin is associated with reduced allergen-stimulated lymphocyte proliferation and IL-13 production in childhood', Journal Of Allergy And Clinical Immunology, vol. 116, no. 2, pp. 431-437.
View/Download from: UTS OPUS or Publisher's site
Background: Infant endotoxin exposure has been proposed as a factor that might protect against allergy and the early childhood immune responses that increase the risk of IgE production to allergens. Objective: Using a prospective study design, we tested the hypothesis that early-life endotoxin exposure is associated with allergen- and mitogen-induced cytokine production and proliferative responses of PBMCs isolated from infants with a parental history of physician-diagnosed asthma or allergy. Methods: We assessed household dust endotoxin at age 2 to 3 months and PBMC proliferative and cytokine responses to cockroach allergen (Bla g 2), dust mite allergen (Der f 1), cat allergen (Fel d 1), and the nonspecific mitogen PHA at age 2 to 3 years. Results: We found that increased endotoxin levels were associated with decreased IL-13 levels in response to cockroach, dust mite, and cat allergens, but not mitogen stimulation. Endotoxin levels were not correlated with allergen- or mitogen-induced IFN-gamma, TNF-alpha, or IL-10. Increased endotoxin levels were associated with decreased lymphocyte proliferation after cockroach allergen stimulation. An inverse, although nonsignificant, association was also found between endotoxin and proliferation to the other tested stimuli. Conclusion: Increased early-life exposure to household endotoxin was associated with reduced allergen-induced production of the T(H)2 cytokine IL-13 and reduced lymphoproliferative responses at age 2 to 3 years in children at risk for allergy and asthma. Early-life endotoxin-related reduction of IL-13 production might represent one pathway through which increased endotoxin decreases the risk of allergic disease and allergy in later childhood.
Duty, S.M., Calafat, A.M., Silva, M.J., Ryan, L.M. & Hauser, R. 2005, 'Phthalate exposure and reproductive hormones in adult men', Human Reproduction, vol. 20, no. 3, pp. 604-610.
View/Download from: UTS OPUS or Publisher's site
BACKGROUND: Phthalates are used in personal and consumer products, food packaging materials, and polyvinyl chloride plastics and have been measured in the majority of the general population of the USA. Consistent experimental evidence shows that some phthalates are developmental and reproductive toxicants in animals. This study explored the association between environmental levels of phthalates and altered reproductive hormone levels in adult men. METHODS: Between 1999 and 2003, 295 men were recruited from Massachusetts General Hospital. Selected phthalate metabolites were measured in urine. Linear regression models explored the relationship between specific gravity-adjusted urinary phthalate monoester concentrations and serum levels of reproductive hormones, including FSH, LH, sex hormone-binding globulin, testosterone, and inhibin B. RESULTS: An interquartile range (IQR) change in monobenzyl phthalate (MBzP) exposure was significantly associated with a 10% [95% confidence interval (CI): -16, -4.0] decrease in FSH concentration. Additionally, an IQR change in monobutyl phthalate (MBP) exposure was associated with a 4.8% (95% CI: 0, 10) increase in inhibin B but this was of borderline significance. CONCLUSIONS: Although we found associations between MBP and MBzP urinary concentrations and altered levels of inhibin B and FSH, the hormone concentrations did not change in the expected patterns. Therefore, it is unclear whether these associations represent physiologically relevant alterations in these hormones, or whether they represent associations found as a result of conducting multiple comparisons.
Hoffmann, K. 2005, 'Predicting exposure levels.', Epidemiology (Cambridge, Mass.), vol. 16, no. 1, pp. 134-135.
Chen, Y.C., Xu, L., Guo, Y.L., Su, H.J., Smith, T.J., Ryan, L.M., Lee, M. & Christiani, D.C. 2004, 'Polymorphisms in GSTT1 and p53 and urinary transitional cell carcinoma in south-western Taiwan: A preliminary study', Biomarkers, vol. 9, no. 4-5, pp. 386-394.
View/Download from: UTS OPUS or Publisher's site
Little is known about the relevance of genetic polymorphisms to arsenic-related bladder cancer. A preliminary case-control study was conducted to explore the association between genetic polymorphisms of GSTT1, p53 codon 72 and bladder cancer in southern Taiwan, a former high arsenic exposure area. Fifty-nine urinary transitional cell carcinoma (TCC) patients from a referral centre in south-western Taiwan and 81 community controls matched on residence were recruited from 1996 to 1999. A questionnaire was administered to obtain arsenic exposure and general health information. Genotypes of p53 codon 72 and GSTT1 were analysed by polymerase chain reaction-restriction fragment length polymerase. The combined variant genotypes (heterozygous or homozygous variant) of p53 codon 72 and GSTT1 null were observed in 29% of cases and in 44% of controls, respectively. In this preliminary study, bladder cancer risk was slightly elevated for subjects carrying the variant genotype of p53 codon 72 or in subjects carrying the GSTT1 null genotype. Variants in p53 codon 72 increased the risk of bladder cancer among smokers. However, the results were not statistically significant and larger confirmatory studies are needed to clarify the role of candidate gene polymorphisms and bladder cancer risk in arsenic exposed populations.
Houseman, E.A., Ryan, L.M. & Coull, B.A. 2004, 'Cholesky residuals for assessing normal errors in a linear model with correlated outcomes', Journal Of The American Statistical Association, vol. 99, no. 466, pp. 383-394.
View/Download from: UTS OPUS or Publisher's site
Despite the widespread popularity of linear models for correlated outcomes (e.g., linear mixed models and time series models), distribution diagnostic methodology remains relatively underdeveloped in this context. In this article we present an easy-to-implement approach that lends itself to graphical displays of model fit. Our approach involves multiplying the estimated marginal residual vector by the Cholesky decomposition of the inverse of the estimated marginal variance matrix. The resulting "rotated" residuals are used to construct an empirical cumulative distribution function and pointwise standard errors. The theoretical framework, including conditions and asymptotic properties, involves technical details that are motivated by Lange and Ryan. Pierce, and Randles. Our method appears to work well in a variety of circumstances. including models having independent units of sampling (clustered data) and models for which all observations are correlated (e. g.. a single time series). Our methods can produce satisfactory results even for models that do not satisfy all of the technical conditions stated in our theory.
Chen, J., Chang, W., Shih, T., Chen, C., Chang, W., Dennerlein, J., Ryan, L.M. & Christiani, D. 2004, 'Using 'exposure Prediction Rules' For Exposure Assessment - An Example On Whole-body Vibration In Taxi Drivers', Epidemiology, vol. 15, no. 3, pp. 293-299.
View/Download from: Publisher's site
Background: It is often difficult and expensive to make direct measurements of an individual's occupational or environmental exposures in large epidemiologic studies. Methods: In this study, we used information collected in validation studies to develop
Chen, J., Dennerlein, J.T., Shih, T., Chen, C., Cheng, Y., Chang, W.P., Ryan, L.M. & Christiani, D.C. 2004, 'Knee pain and driving duration: A secondary analysis of the taxi drivers' health study', American Journal of Public Health, vol. 94, no. 4, pp. 575-581.
View/Download from: UTS OPUS or Publisher's site
Objectives. We explored a postulated association between daily driving time and knee pain. Methods. We used data from the Taxi Drivers' Health Study to estimate 1-year prevalence of knee pain as assessed by the Nordic musculoskeletal questionnaire. Results. Among 1242 drivers, the prevalence of knee pain, stratified by duration of daily driving (less than or equal to6, > 6 through 8, > 8 through 10, and >10 hours), was 11%, 17%, 19%, and 22%, respectively. Compared with driving 6 or fewer hours per day, the odds ratio of knee pain prevalence for driving more than 6 hours per day was 2.52 (95% confidence interval= 1.36, 4.65) after we adjusted for socioeconomic, work-related, and personal factors in the multiple logistic regression. Conclusions. The dose-related association between driving duration and knee pain raises concerns about work-related knee joint disorders among professional drivers.
Stern, A.H., Jacobson, J.L., Ryan, L.M. & Burke, T.A. 2004, 'Do recent data from the Seychelles Islands alter the conclusions of the NRC Report on the toxicological effects of methylmercury?', Environmental Health: A Global Access Science Source, vol. 3, no. 2.
View/Download from: UTS OPUS
Li, Y. & Ryan, L.M. 2004, 'Survival analysis with heterogeneous measurement error', Journal of the American Statistical Association, vol. 99, no. 467, pp. 724-735.
View/Download from: UTS OPUS or Publisher's site
This article is motivated by a time-to-event analysis where the covariate of interest was measured at the wrong time. We show that the problem can be formulated as a special case of survival analysis with heterogeneous covariate measurement error and develop a general analytic framework. We study the asymptotic behavior of the naive partial likelihood estimates and analytically demonstrate that under the heterogeneous measurement error structure and the assumption that all components of the covariate vector and the measurement error vector combined are mutually independent, these naive estimates will shrink toward 0, and that the degree of attenuation increases as the measurement error increases. We also give counterexamples for reverse attenuation when the independence conditions are violated. We use our analytical results to derive a simple bias-correcting estimator that performs well in simulations for small and moderate amounts of measurement error. Our framework can be used to provide insight into the behavior of the commonly used partial likelihood score test for testing no association between a failure outcome and an exposure, for example, in the presence of measurement error or mistiming error. In particular, we derive the asymptotic distribution of the naive partial likelihood score test under a series of local alternatives and discuss the asymptotic relative efficiency. As a result, a simple sample size formula to account for the contamination of covariates is obtained.
Elashoff, M. & Ryan, L.M. 2004, 'An EM algorithm for estimating equations', Journal of Computational and Graphical Statistics, vol. 13, no. 1, pp. 48-65.
View/Download from: UTS OPUS or Publisher's site
This article presents an algorithm for accommodating missing data in situations where a natural set of estimating equations exists for the complete data setting. The complete data estimating equations can correspond to the score functions from a standard, partial, or quasi-likelihood, or they can be generalized estimating equations (GEEs). In analogy to the EM, which is a special case, the method is called the ES algorithm, because it iterates between an E-Step wherein functions of the complete data are replaced by their expected values, and an S-Step where these expected values are substituted into the complete-data estimating equation, which is then solved. Convergence properties of the algorithm are established by appealing to general theory for iterative solutions to nonlinear equations. In particular, the ES algorithm (and indeed the EM) are shown to correspond to examples of nonlinear Gauss-Seidel algorithms. An added advantage of the approach is that it yields a computationally simple method for estimating the variance of the resulting parameter estimates.
Bellamy, S.L., Yi, L., Ryan, L.M., Lipsitz, S., Canner, M.J. & Wright, R. 2004, 'Analysis of clustered and interval censored data from a community-based study in asthma', Statistics In Medicine, vol. 23, no. 23, pp. 3607-3621.
View/Download from: UTS OPUS or Publisher's site
Many authors in recent years have proposed extensions of familiar survival analysis methodologies to apply in dependent data settings, for example, when data are clustered or subject to repeated measures. However, these extensions have been considered largely in the context of right censored data. In this paper, we discuss a parametric frailty model for the analysis of clustered and interval censored failure time data. Details are presented for the specific case where the underlying time to event data follow a Weibull distribution. Maximum likelihood estimates will be obtained using commercially available software and the empirical efficiency of these estimators will be explored via a simulation study. We also discuss a score test to make inferences about the magnitude and significance of over-dispersion in clustered data settings. These methods will be illustrated using data from the East Boston Asthma Study.
Arroyo, C., Hu, F.B., Ryan, L.M., Kawachi, I., Colditz, G.A., Speizer, F.E. & Manson, J. 2004, 'Depressive symptoms and risk of type 2 diabetes in women', Diabetes Care, vol. 27, no. 1, pp. 129-133.
View/Download from: UTS OPUS or Publisher's site
OBJECTIVE - To explore the relationship between depressive symptoms and incidence of type 2 diabetes in women. RESEARCH DESIGN AND METHODS - We conducted an analysis of 72,178 female nurses aged 45-72 years who did not have diagnosed diabetes and who answered the Medical Outcomes Study 36-Item Short-Form Health Status Survey (SF-36) at baseline in 1992. We calculated relative risks (RR) of type 2 diabetes for women with presence of depressive symptoms (i.e., Five-Item Mental Health Index [MHI-5] score >52). RESULTS - During 4 years of follow-up (282,317 person-years), 973 incident cases of type 2 diabetes were documented. Age-adjusted RR of developing type 2 diabetes for women with presence of depressive symptoms was 1.55 (95% CI 1.27-1.90). Additional adjustment for BMI resulted in a RR of developing type 2 diabetes of 1.36 (1.11-1.67). The muitivariate RR of developing type 2 diabetes was 1.22 (1.00-1.50). After ex eluding women diagnosed with diabetes between 1992 and 1994, 472 incident cases of type 2 diabetes were documented for the follow-up period from 1994 to 1996 (148,889 person-years). The multivariate RR of developing type 2 diabetes for women with depressive symptoms was 1.29 (0.96-1.72). CONCLUSIONS- Our data suggest that depressive symptoms are associated with a modest increase in the risk of type 2 diabetes.
Meeker, J.D., Ryan, L.M., Barr, D.B., Herrick, R.F., Bennett, D.H., Bravo, R. & Hauser, R. 2004, 'The relationship of urinary metabolites of carbaryl/naphthalene and chlorpyrifos with human semen quality', Environmental Health Perspectives, vol. 112, no. 17, pp. 1665-1670.
View/Download from: UTS OPUS or Publisher's site
Most of the general population is exposed to carbaryl and other contemporary-use insecticides at low levels. Studies of laboratory animals, in addition to limited human data, show an association between carbaryl exposure and decreased semen quality. In the present study we explored whether environmental exposures to 1-naphthol (1N), a metabolite of carbaryl and naphthalene, and 3,5,6-trichloro-2-pyridinol (TCPY), a metabolite of chlorpyrifos and chlorpyrifos-methyl, are associated with decreased semen quality in humans. Subjects (n = 272) were recruited through a Massachusetts infertility clinic. Individual exposures were measured as spot urinary concentrations of 1N and TCPY adjusted using specific gravity. Semen quality was assessed as sperm concentration, percent motile sperm, and percent sperm with normal morphology, along with sperm motion parameters (straight-line velocity, curvilinear velocity, and linearity). Median TCPY and 1N concentrations were 3.22 and 3.19 mug/L, respectively. For increasing 1N tertiles, adjusted odds ratios (ORs) were significantly elevated for below-reference sperm concentration (OR for low, medium, and high tertiles = 1.0, 4.2, 4.2, respectively; p-value for trend = 0.01) and percent motile sperm (1.0, 2.5, 2.4; p-value for trend = 0.01). The sperm motion parameter most strongly associated with 1N was straight-line velocity. There were suggestive, borderline-significant associations for TCPY with sperm concentration and motility, whereas sperm morphology was weakly and nonsignificantly associated with both TCPY and 1N. The observed associations between altered semen quality and 1N are consistent with previous studies of carbaryl exposure, although suggestive associations with TCPY are difficult to interpret because human and animal data are currently limited.
Ryan, L.M., Huang, W., Thurston, S.W., Kelsey, K.T., Wiencke, J.K. & Christiani, D.C. 2004, 'On the use of biomarkers for environmental health research', Statistical Methods in Medical Research, vol. 13, no. 3, pp. 207-255.
View/Download from: UTS OPUS
This article discusses the growing interest in the use of biomarker data in environmental health research and considers some of the challenging statistical issues that arise. We specify a modeling framework that links environmental exposure, biomarkers and outcome, and discuss in Conceptual terms how such a formulation could be used to inform dose response modeling for the purpose of quantitative risk assessment. We then analyse some biomarker data from a case-control study designed to elucidate the mechanisms of smoking induced lung cancer. Because of sample size limitations, we use a likelihood-based analysis which subsumes both cohort and case-control designs as special cases. Our analysis allows us to 1) investigate the extent to which the markers explain the pathway from exposure to outcome; 2) quantify the degree to which biomarker data can improve on predicting outcome over and above exposure; and 3) estimate the association among multiple markers.
Venners, S.A., Wang, X., Chen, C., Wang, L., Chen, D., Guang, W., Huang, A., Ryan, L.M., O'Connor, J., Lasley, B., Overstreet, J., Wilcox, A. & Xu, X. 2004, 'Paternal smoking and pregnancy loss: A prospective study using a biomarker of pregnancy', American Journal of Epidemiology, vol. 159, no. 10, pp. 993-1001.
View/Download from: Publisher's site
Results of studies on paternal smoking and spontaneous abortions have been inconsistent. The authors examined the effect of paternal smoking on the risk of pregnancy loss in a prospective cohort of 526 newly married, nonsmoking, female textile workers in China between 1996 and 1998. Upon stopping contraception, subjects provided daily urine specimens and records of vaginal bleeding for up to 1 year or until clinical pregnancy. Daily urinary human chorionic gonadotropin was assayed to detect conception and early pregnancy losses, and pregnancies were followed to detect clinical spontaneous abortions. Subjects were grouped by the number of cigarettes that husbands reported smoking daily: nonsmokers (group 1, n = 216), fewer than 20 cigarettes (group 2, n = 239), and 20 or more cigarettes (group 3, n = 71). Compared with that for group 1, the adjusted odds ratio of early pregnancy loss of any conception for group 2 was 1.04 (95% confidence interval (CI): 0.67, 1.63) and for group 3 was 1.81 (95% CI: 1.00, 3.29). The adjusted hazard ratio of conception for group 2 was 0.90 (95% CI: 0.70, 1.18) and for group 3 was 0.96 (95% CI: 0.66, 1.39), while the adjusted hazard ratio of clinical pregnancy for group 2 was 0.93 (95% CI: 0.72, 1.20) and for group 3 was 0.78 (95% CI: 0.55, 1.12). The authors conclude that heavy paternal smoking increased the risk of early pregnancy loss through maternal and/or paternal exposure.
Meeker, J.D., Singh, N.P., Ryan, L.M., Duty, S.M., Barr, D.B., Herrick, R.F., Bennett, D.H. & Hauser, R. 2004, 'Urinary levels of insecticide metabolites and DNA damage in human sperm', Human Reproduction, vol. 19, no. 11, pp. 2573-2580.
View/Download from: UTS OPUS or Publisher's site
Members of the general population are exposed to non-persistent insecticides at low levels. The present study explored whether environmental exposures to carbaryl and chlorpyrifos are associated with DNA damage in human sperm. METHODS: Subjects (n 5 260) were recruited through a Massachusetts infertility clinic. Individual exposures were measured as spot urinary metabolite concentrations of chlorpyrifos [3,5,6-trichloro-2- pyridinol (TCPY)] and carbaryl [1-naphthol (1N)], adjusted using specific gravity. Sperm DNA integrity was assessed by neutral comet assay and reported as comet extent, percentage DNA in comet tail (Tail%) and tail distributed moment (TDM). RESULTS: A statistically significant increase in Tail% was found for an interquartile range (IQR) increase in both 1N [coefficient 5 4.1; 95% confidence interval (CI) 1.96.3] and TCPY (2.8; 0.9 4.6), while a decrease in TDM was associated with IQR changes in 1N (22.2; 24.9 to 0.5) and TCPY (22.5; 24.7 to 20.2). A negative correlation between Tail% and TDM was present only when stratified by comet extent, suggesting that Tail% and TDM may measure different types of DNA damage within comet extent strata. CONCLUSIONS: Environmental exposure to carbaryl and chlorpyrifos may be associated with increased DNA damage in human sperm, as indicated by a change in comet assay parameters.
Raby, B.A., Celedon, J.C., Litonjua, A.A., Phipatanakul, W., Sredl, D., Oken, E., Ryan, L., Weiss, S.T. & Gold, D.R. 2004, 'Low-normal gestational age as a predictor of asthma at 6 years of age', PEDIATRICS, vol. 114, no. 3, pp. E327-E332.
View/Download from: Publisher's site
Serrano-Trespalacios, P.I., Ryan, L. & Spengler, J.D. 2004, 'Ambient, indoor and personal exposure relationships of volatile organic compounds in Mexico City Metropolitan Area', JOURNAL OF EXPOSURE ANALYSIS AND ENVIRONMENTAL EPIDEMIOLOGY, vol. 14, pp. S118-S132.
View/Download from: Publisher's site
Duty, S.M., Calafat, A.M., Silva, M.J., Brock, J.W., Ryan, L., Chen, Z.Y., Overstreet, J. & Hauser, R. 2004, 'The relationship between environmental exposure to phthalates and computer-aided sperm analysis motion parameters', JOURNAL OF ANDROLOGY, vol. 25, no. 2, pp. 293-302.
Li, Y. & Ryan, L. 2004, 'Survival analysis with heterogeneous covariate measurement error', JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, vol. 99, no. 467, pp. 724-735.
View/Download from: Publisher's site
Mezzetti, M., Ibrahim, J.G., Bois, F.Y., Ryan, L.M., Ngo, L. & Smith, T.J. 2003, 'A Bayesian compartmental model for the evaluation of 1,3-butadiene metabolism', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 52, pp. 291-305.
View/Download from: UTS OPUS or Publisher's site
We propose a Bayesian model for physiologically based pharmacokinetics of 1,3-butadiene (BD). BD is classified as a suspected human carcinogen and exposure to it is common, especially through cigarette smoke as well as in urban settings. The main aim of the methodology and analysis that are presented here is to quantify variability in the rates of BD metabolism by human subjects. A three-compartmental model is described, together with informative prior distributions for the population parameters, all of which represent real physiological variables. The model is described in detail along with the meanings and interpretations of the associated parameters. A four-compartment model is also given for comparison. Markov chain Monte Carlo methods are described for fitting the model proposed. The model is fitted to toxicokinetic data obtained from 133 healthy subjects (males and females) from the four major racial groups in the USA, with ages ranging from 19 to 62 years. Subjects were exposed to 2 parts per million of BD for 20 min through a face mask by using a computer-controlled exposure and respiratory monitoring system. Stratification by ethnic group results in major changes in the physiological parameters. Sex and age were also tested but not found to have a significant effect.
Yeap, B.Y., Catalano, P.J., Ryan, L.M. & Davidian, M. 2003, 'Robust two-stage approach to repeated measurements analysis of chronic ozone exposure in rats', Journal of Agricultural, Biological, and Environmental Statistics, vol. 8, no. 4, pp. 438-454.
View/Download from: UTS OPUS or Publisher's site
A robust two-stage approach is used to reanalyze the repeated measurements from an experiment of airway responsiveness in rats randomized to long-term exposure at four ozone doses. The concentration-response data generated for each rat may be represented as a hierarchical nonlinear model encompasing the sources of variation within and between individual profile for each rat, the conditional modeling approach can assess the adequacy of an assumed mean model, a fundamental advantage not intrinsic to marginal techniques. The two-stage population inference is based on the estimated individual parameters, thus maintaining an intuitive appeal to the toxicologists who traditionally have fitted a separate curve for each animal and then applied ANOVA to the summary statistics. However, we formally adjust the standard errors for the extra variability due to the initial estimation of the individual parameters and also allow for their within-rat correlation. The robust two-stage method appropriately down weights the a berrant responses arising sporadically within individualsand, more importantly, the rats which may be outlying relative to the usual population variation. The true effect of chronic ozone exposure, including a significant gender interaction, may be masked by a few rats which exert undue influence on the population estimates in a nonrobust analysis.
Stark, P.C., Burge, H.A., Ryan, L.M., Milton, D.K. & Gold, D.R. 2003, 'Fungal levels in the home and lower respiratory tract illnesses in the first year of life', American Journal of Respiratory and Critical Care Medicine, vol. 168, no. 2, pp. 232-237.
View/Download from: UTS OPUS or Publisher's site
The association between home dampness and lower respiratory symptoms in children has been well documented. Whether fungal exposures contribute to this association is uncertain. In a prospective birth cohort of 499 children of parents with asthma/allergies, we examined in-home fungal concentrations as predictors of lower respiratory illnesses (LRI) (croup, pneumonia, bronchitis, and bronchiolitis) in the first year. In multivariate analyses, we found a significant increased relative risk (RR) between LRI and high levels (more than the 90th percentile) of airborne Penicillium (RR = 1.73, 95% confidence interval [CI], 1.23, 2.43), dust-borne Clodosporium (RR = 1.52; Cl, 1.02, 2.25), Zygornycetes (RR = 1.96; CI, 1.35, 2.83), and Alternaria (RR = 1.51; CI, 1.00, 2.28), after controlling for sex, presence of water damage or visible mold/mildew, born in winter, breastfeeding, and being exposed to other children through siblings. In a multivariate analysis, the RR of LRI was elevated in households with any fungal level at more than the 90th percentile (RR = 1.86; CI, 1.21, 2.88). Exposure to high fungal levels increased the risk of LRI in infancy, even for infants with nonwheezing LRI. Actual mechanisms remain unknown, but fungi and their components (glucans, mycotoxins, and proteins) may increase the risk of LRI by acting as irritants or through increasing susceptibility to infection.
Roy, J., Lin, X. & Ryan, L.M. 2003, 'Scaled marginal models for multiple continuous outcomes', Biostatistics, vol. 4, no. 3, pp. 371-383.
View/Download from: UTS OPUS or Publisher's site
In studies that involve multivariate outcomes it is often of interest to test for a common exposure effect. For example, our research is motivated by a study of neurocognitive performance in a cohort of HIV-infected women. The goal is to determine whethe
Ryan, L.M., Chen, Y., Su, H.J., Guo, Y.L., Hsueh, Y., Smith, T.J., Lee, M. & Christiani, D.C. 2003, 'Arsenic methylation and bladder cancer risk in Taiwan', Cancer Causes & Control, vol. 14, no. 4, pp. 303-310.
View/Download from: Publisher's site
Objective: The mechanism of arsenic detoxification in humans remains unclear. Data are especially lacking for low-level arsenic exposure. We hypothesize that arsenic methylation ability, defined as the ratios of monomethylarsonic acid (MMA(V))/inorganic
Okuno, S., Ryan, L.M., Edmonson, J., Priebat, D. & Blum, R. 2003, 'Phase Ii Trial Of Gemcitabine In Patients With Advanced Sarcomas (e1797) - A Trial Of The Eastern Cooperative Oncology Group', Cancer, vol. 97, no. 8, pp. 1969-1973.
View/Download from: Publisher's site
BACKGROUND. The current study was conducted to evaluate the antitumor activity 2 and toxicity of gemcitabine in patients with advanced sarcoma. METHODS. Twenty-five patients with advanced sarcomas, who previously were untreated for metastatic disease, we
Lohstroh, P.N., Chen, J., Ba, J., Ryan, L.M., Xu, X., Overstreet, J.W. & Lasley, B.L. 2003, 'Bone resorption is affected by follicular phase length in female rotating shift workers.', Environmental Health Perspectives, vol. 111, no. 4, pp. 618-622.
View/Download from: UTS OPUS or Publisher's site
Stressors as subtle as night work or shift work can lead to irregular menstrual cycles, and changes in reproductive hormone profiles can adversely affect bone health. This study was conducted to determine if stresses associated with the disruption of regular work schedule can induce alterations in ovarian function which, in turn, are associated with transient bone resorption. Urine samples from 12 rotating shift workers from a textile mill in Anqing, China, were collected in 1996-1998 during pairs of sequential menstrual cycles, of which one was longer than the other (28.4 vs. 37.4 days). Longer cycles were characterized by a prolonged follicular phase. Work schedules during the luteal-follicular phase transition (LFPT) preceding each of the two cycles were evaluated. All but one of the shorter cycles were associated with regular, forward phase work shift progression during the preceding LFPT. In contrast, five longer cycles were preceded by a work shift interrupted either by an irregular shift or a number of "off days." Urinary follicle-stimulating hormone levels were reduced in the LFPT preceding longer cycles compared with those in the LFPT preceding shorter cycles. There was greater bone resorption in the follicular phase of longer cycles than in that of shorter cycles, as measured by urinary deoxypyridinoline. These data confirm reports that changes in work shift can lead to irregularity in menstrual cycle length. In addition, these data indicate that there may be an association between accelerated bone resorption in menstrual cycles and changes of regularity in work schedule during the preceding LFPT.
Chen, Y., Xu, L., Guo, Y.L., Su, H.J., Hsueh, Y., Smith, T.J., Ryan, L.M., Lee, M., Chaor, S., Lee, J.Y. & Christiani, D.C. 2003, 'Genetic polymorphism in p53 codon 72 and skin cancer in Southwestern Taiwan', Journal Of Environmental Science And Health Part A-toxic/hazardous Substances & Environmental Engineering, vol. 38, no. 1, pp. 201-211.
View/Download from: UTS OPUS or Publisher's site
The Pro/Pro polymorphism of p53 codon 72 has been reported to be related to bladder and lung cancer, but its relationship with skin cancer is unclear. We assessed the hypothesis that there is a relationship between the p53 codon 72, Pro/Pro polymorphism, cumulative arsenic exposure, and the risk of skin cancer in a hospital-based case-control study in southwestern Taiwan. From 1996 to 1999, 93 newly-diagnosed skin cancer patients at the National Cheng-Kung University (NCKU) Hospital and 71 community controls matched on residence were recruited in southwestern Taiwan. The genotype of p53 codon 72 (Arg/Arg, Arg/Pro, or Pro/Pro) was determined for all subjects by polymerase chain reaction-restricted fragment length polymorphism (PCR-RFLP). A questionnaire was administered to each subject for collection of demographic information, personal habits, disease history, diet information, and other relevant questions. The Pro/Pro (homozygous) genotype was more frequent in skin cancer patients (cases, 20%; controls, 12%; P = 0.37). Subjects with the susceptible genotype Pro/Pro and heterozygous (intermediate) genotype Pro/Arg had 2.18 and 0.99 times risk of skin cancer than the wild type Arg/Arg (95% confidence interval, 0.74-4.38; 95% confidence interval, 0.44-2.21), respectively. Compared with subjects with 18.5 < BMI < 23, subjects with BMI > 18.5 had 5.78 times risk of skin cancer (95% confidence interval, 1.06 to 31.36) after adjusting for other risk factors. There was no interaction between BMI and genotype, but the sample size was small. The risk of skin cancer did not significantly vary by tumor cell-type. The risk of skin cancer is increased in individuals with the Pro/Pro genotype. Larger, confirmatory studies are needed to clarify the role of constitutional polymorphisms in p53 and skin cancer risk.
Coull, B.A., Mezzetti, M. & Ryan, L.M. 2003, 'A Bayesian hierarchical model for risk assessment of methylmercury', Journal of Agricultural, Biological, and Environmental Statistics, vol. 8, no. 3, pp. 253-270.
View/Download from: UTS OPUS or Publisher's site
This article uses a Bayesian hierarchical model to quantify the adverse health effects associated with in-utero exposure to methylmercury. By allowing for study-to-study as well as outcome-to-outcome variability, the approach provides a useful meta-analytic tool for multi-outcome, multi-study environmental risk assessments. The analysis presented here expands on the findings of a National Academy of Sciences (NAS) committee, charged with advising the United States Environmental Protection Agency (EPA) on an appropriate approach to conducting a risk assessment for methylmercury. The NAS committee, for which the senior author (Ryan) was a committee member, reviewed the findings from several conflicting studies and reported the results from a Bayesian hierarchical model that synthesized information across several studies and for several outcomes. Although the NAS committee did not suggest that the hierarchical model be used as the actual basis for a methylmercury risk assessment, the results from the model were used to justify and support the final recommendation that the risk analysis be based on data from a study conducted in the Faroe Islands, which had found an association between in-utero exposure to methylmercury and impaired neurological development. We consider a variety of statistical issues, but particularly sensitivity to model specification.
Chen, J.C., Chang, W.R., Shih, T.S., Chen, C.J., Chang, W.P., Dennerlein, J.T., Ryan, L.M. & Christiani, D.C. 2003, 'Predictors of whole-body vibration levels among urban taxi drivers', Ergonomics, vol. 46, no. 11, pp. 1075-1090.
View/Download from: UTS OPUS or Publisher's site
To identify a set of important WBV predictors that could be used to develop a statistical instrument for exposure assessment in a large epidemiologic study, a total of 432 WBV measures were taken from a sample of 247 male drivers in Taipei City, Taiwan.
Duty, S.M., Singh, N.P., Silva, M.J., Barr, D.B., Brock, J.W., Ryan, L.M., Herrick, R.F., Christiani, D.C. & Hauser, R. 2003, 'The relationship between environmental exposures to phthalates and DNA damage in human sperm using the neutral comet assay', Environmental Health Perspectives, vol. 111, no. 9, pp. 1164-1169.
View/Download from: UTS OPUS or Publisher's site
Phthalates are industrial chemicals widely used in many commercial applications. The general population is exposed to phthalates through consumer products as well as through diet and medical treatments. To determine whether environmental levels of phthalates are associated with altered DNA integrity in human sperm, we selected a population without identified sources of exposure to phthalates. One hundred sixty-eight subjects recruited from the Massachusetts General Hospital Andrology Laboratory provided a semen and a urine sample. Eight phthalate metabolites were measured in urine by using high-performance liquid chromatography and tandem mass spectrometry; data were corrected for urine dilution by adjusting for specific gravity. The neutral single-cell microgel electrophoresis assay (comet assay) was used to measure DNA integrity in sperm. VisComet image analysis software was used to measure comet extent, a measure of total comet length (micrometers); percent DNA in tail (tail%), a measure of the proportion of total DNA present in the comet tail; and tail distributed moment (TDM), an integrated measure of length and intensity (micrometers). For an interquartile range increase in specific gravityadjusted monoethyl phthalate (MEP) level, the comet extent increased significantly by 3.6 &igrave;m [95% confidence interval (95% CI), 0.746.47]; the TDM also increased 1.2 &igrave;m (95% CI, 0.05 to 2.38) but was of borderline significance. Monobutyl, monobenzyl, monomethyl, and mono-2-ethylhexyl phthalates were not significantly associated with comet assay parameters. In conclusion, this study represents the first human data to demonstrate that urinary MEP, at environmental levels, is associated with increased DNA damage in sperm.
Edmonson, J.H., Ryan, L.M., Falkson, C.I., Hicks, D.G. & Blum, R.H. 2003, 'Phase II study of Ifosfamide+Doxorubicin in patients with advanced synovial sarcomas (E1793): a trial of the Eastern Cooperative Oncology Group', Sarcoma, vol. 7, no. 1, pp. 9-11.
View/Download from: UTS OPUS
Purpose Because we had observed in the synovial sarcoma subgroup of a broad phase III advanced soft tissue sarcoma study a significantly greater objective regression rate from ifosfamide&thorn;doxorubicin (88%) than from doxorubicin alone (20%) ( P&frac14;0.02), the Eastern Cooperative Oncology Group (ECOG) decided to further assess this two drug combination in a subsequent Phase II study. Patients Between 1994 and 1999, twelve adult patients with advanced synovial sarcomas were enrolled to receive, as their initial chemotherapy, ifosfamide 7.5 gm/m2 plus doxorubicin 60 mg/m2, given intravenously over two consecutive days every 3 weeks. Methods Each day for 2 days doxorubicin 30 mg/m2 was infused over 5 min through a running i.v., followed by ifosfamide 3750 mg/m2 over 4 h. Continuous i.v. fluid was infused at 300mL/h for 3 h on day 1, before chemotherapy was begun; then the infusion was continued at 100mL/h for a total of 3 days. Mesna 750 mg/m2 was given 15 min before ifosfamide and at 4 and 8 h after ifosfamide on days 1 and 2 of each treatment cycle. Filgrastim (G-CSF) 5 mg/kg was given subcutaneously each day for 14 days beginning on day 3 of each treatment cycle to limit the severity of neutropenia. Results Five of our 12 patients (42%) experienced partial regression of their advanced synovial sarcomas; however, this first stage result was borderline for proceeding to the second planned stage of accrual and our case accrual was quite poor. Thus, the study was closed after stage one accrual. Our patients received a median of four cycles of chemotherapy (range: 1 to 6). All patients experienced at least grade 3 neutropenia (grade 4 in nine of them), and one patient died of treatment-related sepsis following the initial cycle of chemotherapy. Median survival was 11 months.
Chen, Y., Guo, Y.L., Su, H.J., Hsueh, Y., Smith, T.J., Ryan, L.M., Lee, M., Chaor, S., Lee, J.Y. & Christiani, D.C. 2003, 'Arsenic methylation and skin cancer risk in Southwestern Taiwan', Journal of Occupational & Environmental Medicine, vol. 45, no. 3, pp. 241-248.
View/Download from: UTS OPUS or Publisher's site
Arsenic is a known carcinogen, but data are especially lacking on the health effects of low-level exposure, and on the health significance of methylation ability. We conducted a case-control study (76 cases and 224 controls from 1996 to 1999) in southwestern Taiwan to explore the association among primary and secondary arsenic methylation index (PMI and SMI, respectively), cumulative arsenic exposure (CAE), and the risk of skin cancer. As compared with the controls, the skin cancer group reported more sun exposure (P = 0.02) and had a lower BMI (P = 0.03), as well as lower education level (P = 0.01). Skin cancer patients and controls were similar with regard to age, gender, smoking and alcohol consumption. Given a low SMI (less than or equal to5), CAE > 15 mg/L-year was associated with an increased. risk of skin cancer (OR, 7.48; 95% CI, 1.65-33.99) compared to a CAE less than or equal to2 mg/L-year. Given the same level of PMI, SMI, and CAE, men had a higher risk of skin cancer (OR, 4.04; 95% CI, 1.46-11.22) when compared to women. Subjects with low SMI and high CAE have a substantially increased risk of skin cancer. Males in all strata of arsenic exposure and methylation ability had a higher risk of skin cancer than women.
Li, Y., Ryan, L.M., Bellamy, S. & Satten, G.A. 2003, 'Inference on clustered survival data using imputed frailties', Journal of Computational and Graphical Statistics, vol. 12, no. 3, pp. 1-24.
This article proposes a new method for fitting frailty models to clustered survival data that is intermediate between the fully parametric and nonparametric maximum likelihood estimation approaches. A parametric form is assumed for the baseline hazard, but only for the purpose of imputing the unobserved frailties. The regression coefficients are then estimated by solving an estimating equation that is the average of the partial likelihood score with respect to the conditional distribution of frailties given the observed data. We prove consistency and asymptotic normality of the resulting estimators and give associated closed-form estimators of their variance. The algorithm is easy to implement and reduces to the ordinary Cox partial likelihood approach when the frailties have a degenerate distribution. Simulations indicate high efficiency and robustness of the resulting estimates. We apply our new approach to a study with clustered survival data on asthma in children in east Boston.
Ronnenberg, A.G., Wang, X., Xing, H., Chen, C., Chen, D., Guang, W., Guang, A., Wang, L., Ryan, L.M. & Xu, X. 2003, 'Low preconception body mass index is associated with birth outcome in a prospective cohort of Chinese women', The Journal of Nutrition, vol. 133, no. 11, pp. 3449-3455.
View/Download from: UTS OPUS
Low maternal prepregnancy BMI is associated with adverse birth outcomes, but the BMI at which risk increases is not well defined. We assessed whether the relationship between prepregnancy BMI and birth outcomes is influenced by the extent to which mothers were underweight in a prospective study in Anhui, China. The women (n = 575) were 20-34 y old, married, nulliparous and nonsmokers. All measures of infant growth increased with increasing maternal BMI until a plateau was reached at a BMI of 22-23 kg/m(2). Infants born to the 27% of women who were severely underweight before pregnancy (BMI less than or equal to 18.5 kg/m(2)) were at increased risk for fetal growth deficits associated with infant morbidity. Compared with a normal BMI, being severely underweight was associated with mean (+/- SEM) reductions of 219 +/- 40 g in infant birthweight and 6.7 +/- 1.3% in the birthweight ratio and an 80% increase in risk of intrauterine growth restriction [odds ratio (OR) 1.8; 95% Cl: 1.0, 3.3; P = 0.05]. Being severely underweight was also associated with smaller infant head circumference and lower ponderal index. Being moderately underweight (18.5 < BMI < 19.8 kg/m(2)) was not significantly associated with adverse pregnancy outcomes. Gestational age and risk of preterm birth were not associated with maternal BMI. More than half of the women in this study were underweight before pregnancy. Although being moderately underweight was not associated with increased risk of adverse pregnancy outcomes, being severely underweight was an important risk factor for reduced fetal growth.
Golden, C.M., Ryan, L.M. & Holmes, L.B. 2003, 'Chorionic villus sampling: A distinctive teratogenic effect on fingers?', Birth Defects Research Part A: Clinical and Molecular Teratology, vol. 67, pp. 557-562.
View/Download from: UTS OPUS
BACKGROUND: An increased frequency of major limb malformations, especially terminal transverse limb defects, have been described in several studies of birth defects in children who had been exposed to the prenatal diagnosis procedure known as chorionic villus sampling (CVS). Vascular disruption has been proposed as the mechanism behind the fetal effect. We postulate that this mechanism is more likely to affect one or two middle fingers, rather than all five fingers. A recent report of the frequency of defects in any or all fingers in an unexposed control population enabled us to assess whether CVS is associated with an increased frequency of defects involving one or two fingers, as well as terminal transverse limb defects. METHODS: The frequency of limb-reduction defects affecting one or more fingers or toes, including those with constriction rings and tissue loss, in published studies of 20,236 children who had been exposed to CVS was compared with the frequency in 161,252 newborn infants who had not been exposed to CVS. Children with recognized genetic disorders were excluded. RESULTS: Several aspects of the limb deficiencies were more common in the CVS-exposed infants than in unexposed controls. The former were more likely to have: 1) any type of limb deficiency involving one or more fingers (p < .001); 2) absence/hypoplasia of two fingers (1) < .001); and 3) absence/hypoplasia of all five fingers (p = .015). The absence of the distal portion of the third finger was a distinctive type of limb-reduction defect in CVS-exposed infants. CONCLUSIONS: The occurrence of deficiencies in one or two fingers, including those designated as "amniotic band deformities," are as common as terminal transverse limb defects in CVS-exposed infants, and both are much more common than in unexposed infants. The absence of the distal portion of the third finger, with tapering and stiff joints, appears to be a distinctive effect of exposure to CVS
Hamlett, A.C., Ryan, L.M., Serrano, P. & Wolfinger, R. 2003, 'Mixed models for assessing correlation in the presence of replication', Journal of the Air & Waste Management Association, vol. 53, no. 4, pp. 442-450.
View/Download from: UTS OPUS
The need to assess correlation in settings where multiple measurements are available on each of the variables of interest often arises in environmental science. However, this topic is not covered in introductory statistics texts. Although several ad hoc approaches can be used, they can easily lead to invalid conclusions and to a difficult choice of an appropriate measure of the correlation. Lam et al. approached this problem by using maximum likelihood estimation in cases where the replicate measurements are linked over time, but the method requires specialized software. We reanalyze the data of Lam et al. using PROC MIXED in SAS and show how to obtain the parameter estimates of interest with just a few lines of code. We then extend Lam et al.'s method to settings where the replicate measurements are not linked. Analysis of the unlinked case is illustrated with data from a study designed to assess correlations between indoor and outdoor measurements of benzene concentration in the air.
Ryan, L.M. 2003, 'Epidemiologically based environmental risk assessment', Statistical Science, vol. 18, no. 4, pp. 466-480.
View/Download from: UTS OPUS or Publisher's site
Environmental health research aims to discover and understand the links between environmental exposure and disease and to inform the regulatory community so that society can be protected against cancer, birth defects and other adverse health effects associated with chemical, industrial and other exposures. Statistical science has a critical role to play in terms of providing the appropriate tools to design and analyze the studies needed to address the questions of interest, as well as quantifying risks and characterizing uncertainty. Recent years have seen some dramatic changes in the way that environmental risk assessment is accomplished. One such change is a move away from a traditional reliance on toxicological studies in animals to incorporate more epidemiological data. This shift has been facilitated by scientific advances that now allow researchers to accurately characterize human exposures in a variety of settings, as well as to measure genetic and other biomarkers that reflect subtle health effects and variations in susceptibility. This article will use a high profile case study to highlight some of the challenging statistical issues arising from this shifting emphasis from animal based toxicology to environmental epidemiology in the risk assessment world. Among the topics to be discussed are the uses of biologically based models and biomarkers, as well as the role of Bayesian methods to characterize uncertainty due to population heterogeneity, unmeasured confounders, exposure measurement error and model uncertainty.
Celedon, J.C., Wright, R.J., Litonjua, A.A., Sredl, D., Ryan, L.M., Weiss, S.T. & Gold, D.R. 2003, 'Day care attendance in early life, maternal history of asthma, and asthma at 6 years', American Journal of Respiratory and Critical Care Medicine, vol. 167, pp. 1239-1243.
View/Download from: Publisher's site
Among children not selected on the basis of a parental history of atopy, day care attendance in early life is inversely associated with asthma at school age. We examined the relation between day care in the first year of life and asthma, recurrent wheezing, and eczema at the age of 6 years and wheezing in the first 6 years of life among 453 children with parental history of atopy followed from birth. Among all study participants, day care in the first year of life was inversely associated with eczema (odds ratio [OR] = 0.3, 95% confidence interval [CI] = 0.1-0.8). Day care attendance in early life was associated with a decreased risk of asthma (OR = 0.3, 95% CI = 0.1-0.7) and recurrent wheezing (OR = 0.3, 95% CI = 0.1-0.9) at the age of 6 years and with a decreased risk of any wheezing after the age of 4 years only among children without maternal history of asthma. Among children with maternal history of asthma, day care in early life had no protective effect on asthma or recurrent wheezing at the age of 6 years but was instead associated with an increased risk of wheezing in the first 6 years of life. Our findings suggest that maternal history of asthma influences the relation between day care-related exposures and childhood asthma.
Li, Y., Ryan, L., Bellamy, S. & Satten, G.A. 2003, 'Inference on clustered survival data using imputed frailties', JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, vol. 12, no. 3, pp. 640-662.
View/Download from: Publisher's site
Hauser, R., Chen, Z.Y., Pothier, L., Ryan, L. & Altshul, L. 2003, 'The relationship between human semen parameters and environmental exposure to polychlorinated biphenyls and p,p '-DDE', ENVIRONMENTAL HEALTH PERSPECTIVES, vol. 111, no. 12, pp. 1505-1511.
View/Download from: Publisher's site
Duty, S.M., Silva, M.J., Barr, D.B., Brock, J.W., Ryan, L., Chen, Z.Y., Herrick, R.F., Christiani, D.C. & Hauser, R. 2003, 'Phthalate exposure and human semen parameters', EPIDEMIOLOGY, vol. 14, no. 3, pp. 269-277.
View/Download from: Publisher's site
Betensky, R., Lindsey, J., Wand, M. & Ryan, L.M. 2002, 'A local likelihood proportional hazards model for interval censored data', Statistics in Medicine, vol. 21, pp. 263-275.
View/Download from: Publisher's site
We discuss the use of local likelihood methods to fit proportional hazards regression models to right and interval censored data. The assumed model allows for an arbitrary, smoothed baseline hazard on which a vector of covariates operates in a proportional manner, and thus produces an interpretable baseline hazard function along with estimates of global covariate effects. For estimation, we extend the modified EM algorithm suggested by Betensky, Lindsey, Ryan and Wand. We illustrate the method with data on times to deterioration of breast cosmeses and HIV-1 infection rates among haemophiliacs
Duty, S.M., Singh, N.P., Ryan, L.M., Chen, Z., Lewis, C., Huang, T. & Hauser, R. 2002, 'Reliability of the comet assay in cryopreserved human sperm', Human Reproduction, vol. 17, no. 5, pp. 1274-1280.
View/Download from: UTS OPUS or Publisher's site
Reliability of the comet assay in cryopreserved human sperm
Hauser, R., Altshul, L., Chen, Z., Ryan, L.M., Overstreet, J., Schiff, I. & Christiani, D.C. 2002, 'Environmental organochlorines and semen quality: Results of pilot study', Environmental Health Perspectives, vol. 110, no. 3, pp. 229-233.
View/Download from: UTS OPUS or Publisher's site
There have been numerous studies that suggest that sperm concentrations (sperm counts) are declining in men. However, other studies suggest that sperm counts are not declining or may be increasing in some areas. Although there is disagreement on whether there is a downward temporal trend in sperm counts, the studies provide evidence that sperm counts vary by geographic location. It has been hypothesized that the geographic variation in sperm concentrations may be due to environmental exposures, lifestyle factors, or some unknown causes. To determine whether contemporary ambient levels of polychlorinated biphenyls (PCBs) and p,p'-DDE are associated with altered semen quantity and quality, we selected a study population without specific exposure to PCBs or p,p'-DDE. The present study presents the results from a pilot study on the relationship between serum PCBs and p,p'-DDE and semen quality in 29 subjects recruited from the Massachusetts General Hospital Andrology Laboratory. Of the 29 subjects, 3 had sperm concentrations < 20 million/mL, 7 had < 50% motile sperm, 9 had < 4% normal morphology, and 6 were below normal in more than one semen parameter. The 18 subjects with normal spermatozoa concentration, motility, and morphology were used as comparison subjects. The mean (SE) concentration of the sum of PCBs and p,p'-DDE was 242 ng/g lipids (34.0) and 354 ng/g lipids (120), respectively, for men with below normal motility as compared to 202 ng/g lipids (16.6) and 240 ng/g lipids (31-1), respectively, for the comparison subjects. The data showed general trends that were suggestive of an association between PCBs and p,p'-DDE and abnormal motility, as well as with sperm concentration and morphology. A full-scale study is currently in progress.
Celedon, J.C., Litonjua, A.A., Ryan, L.M., Weiss, S.T. & Gold, D.R. 2002, 'Lack of association between antibiotic use in the first year of life and asthma, allergic rhinitis or eczema at age 5 years', American Journal of Respiratory and Critical Care Medicine, vol. 166, pp. 72-75.
View/Download from: UTS OPUS or Publisher's site
Five retrospective studies have reported an association between antibiotic use in early life and asthma in childhood. We studied the relationship between the use of oral antibiotics in the first year of life and asthma, allergic rhinitis, and eczema at age 5 years among 448 children with a parental history of atopy monitored from birth. After adjustment for potential confounders, we found no significant association between antibiotic use in the first year of life and asthma (odds ratio [OR] for one versus no courses of antibiotics, 0.5; 95% confidence interval [CI] for OR, 0.2 to 1.5; OR for two or more versus no courses of antibiotics, 1.0; 95% CI for OR, 0.5 to 2.2), recurrent wheezing, allergic rhinitis, or eczema at age 5 years. There was no significant association between antibiotic use in the first year of life and having at least one of three atopic diseases (asthma, allergic rhinitis, or eczema) at age 5 years (OR for one versus no courses of antibiotics, 0.7, 95% CI, 0.4 to 1.4; OR for two or more versus no courses of antibiotics, 0.9; 95% CI, 0.5 to 1.4). Our findings do not support the hypothesis that antibiotic use in early life is associated with the subsequent development of asthma and atopy in childhood.
Litonjua, A.A., Milton, D.K., Celedon, J.C., Ryan, L.M., Weiss, S.T. & Gold, D.R. 2002, 'A longitudinal analysis of wheezing in young children:The independent effects of early life exposure to house dust endotoxin, allergens, and pets', Journal of Allergy and Clinical Immunology, vol. 110, no. 5, pp. 736-742.
View/Download from: UTS OPUS or Publisher's site
Background: It has been postulated that exposure to bacterial endotoxins and animals early in life might confer protection against the development of asthma and allergies. Objective: We investigated the longitudinal effects of exposure to house dust endotoxin (HDE), allergen levels, and the presence of a dog in the home on wheezing in young children over a 4-year period. Methods: Two hundred twenty-six children younger than 5 years were followed for 4 years. Endotoxin and allergen levels were measured from house dust collected at baseline. Longitudinal associations were investigated by using a proportional hazards technique that allowed for multiple outcomes per subject. Results: Exposure to high concentrations of HDE of greater than the median level was associated with an increased risk for wheezing over the period of observation (multivariate relative risk, 1.52; 95 % CI, 1.07-2.14), but this risk rapidly decreased over time (P for trend = .005). Exposure to cockroach allergen was associated with increased risk for wheezing, whereas exposure to cat allergen and the presence of a dog in the home were both associated with decreased risk for wheezing. The risks associated with cockroach allergen, cat allergen, and dog did not change over the period of observation. Conclusion: The negative associations between exposures to dogs and cat allergen and wheeze appear to be independent of the effects of endotoxin and suggest that separate mechanisms might mediate the effects of HDE exposure and pet exposure on the developing immune system.
Celedon, J.C., Litonjua, A.A., Ryan, L.M., Platts-Mills, T., Weiss, S.T. & Gold, D.R. 2002, 'Exposure to cat allergen, maternal history of asthma, and wheezing in first 5 years of life', The Lancet, vol. 360, no. 9335, pp. 781-782.
View/Download from: UTS OPUS
We looked for an association between early exposure to pets and asthma and wheezing in children whose mothers or fathers did or did not have a history of asthma. We followed up 448 children, who had at least one parent with a history of atopy, from birth to 5 years. Among children whose mothers had no history of asthma, exposure to a cat or a Fel d 1 concentration of at least 8 mug/g at the age of 2-3 months was associated with a reduced risk of wheezing between the ages of 1 and 5 years. However, among children whose mothers did have a history of asthma, such exposures were associated with an increased risk of wheezing at or after the age of 3 years. There was no association between wheezing and exposure to dog or dog allergen, and the father's allergy status had no effect on the relation between childhood wheezing and cat exposure.
Sammel, M.D. & Ryan, L.M. 2002, 'Effects of covariance misspecification in a latent variable model for multiple outcomes', Statistica Sinica, vol. 12, no. 4, pp. 1207-1222.
View/Download from: UTS OPUS
Sammel and Ryan (1996) developed a latent variable model that allows for covariate effects on multiple continuous outcomes. While the approach provides an effective tool for data reduction and global test for covariate effects, it makes strong assumptions about the covariance among the outcomes. In addition, some parameters axe common to both the mean and variance suggesting that robustness could be a problem. This manuscript evaluates model misspecification on tests of exposure effects derived from the latent variable model. We develop a robust score test which is valid under misspecified variance assumptions and compare it to one based on Generalized Estimating Equations (GEE) (Liang and Zeger (1986)), under varying assumptions on the true model. Both models have similar loss in power under variance misspecification while the estimated global effect of the covariate is more biased towards the null for the GEE model than the LV model. As the variance/scale of the outcomes increases, the performance of the LV model improves. As for asymptotic comparisons, test performance depends upon the amount of variability and correlation among the outcomes. The LV model test is superior when the data are highly correlated, p > 0.3, and with large variance. When uncorrelated outcomes are incorporated, the GEE model is superior, except when only the correlated outcomes axe impacted by the exposure.
Cho, S., Goldman, B., Ryan, L.M., Chen, C., Damokosh, A.I., Christiani, D.C., Lasley, B.L., O'Connor, J.F., Wilcox, A.J. & Xu, X. 2002, 'Reliability of serial urine HCG as a biomarker to detect early pregnancy loss', Human Reproduction, vol. 17, no. 4, pp. 1060-1066.
View/Download from: Publisher's site
Background: To examine the reliability of HCG as a biomarker for early pregnancy loss, five experienced researchers independently assessed data from 153 menstrual cycles, determining whether each cycle represented,no conception,' a 'continuing conception' or a 'conception lost.' Methods: Urine samples were analysed by immunoradiometric assay using a combination of capture antibodies for the intact heterodimer (B109) and for an epitope common to the beta subunit and the beta core fragment (B204). For each cycle, HCG data were presented as graphs of daily assay results. Summary statistics for HCG assays from 46 women who had undergone bilateral tubal ligation represented baseline values. Results: Pairwise agreement among the assessors for any of the three options ranged from 78-89%. At least three experts agreed for 147 cycles (96%), accounting for 28 conception losses and 19 continuing conceptions. The multi-rater kappa was 0.62 for the conception lost category and 0.68 for continuing conceptions, indicating substantial agreement. Conclusion: The main sources of disagreement involved deciding whether there was sufficient information for assessment, interpreting cycle parameters such as cycle length or bleeding event, and interpreting a distinct HCG rise pattern that does not exceed the baseline value obtained from the sterilized women.
Li, Y. & Ryan, L.M. 2002, 'Modeling Spatial Survival Data Using Semiparametric Frailty Models', Biometrics, vol. 58, no. 2, pp. 287-297.
View/Download from: UTS OPUS or Publisher's site
We propose a new class of semiparametric frailty models for spatially correlated survival data. Specifically, we extend the ordinary frailty models by allowing random effects accommodating spatial correlations to enter into the baseline hazard function multiplicatively. We prove identifiability of the models and give sufficient regularity conditions. We propose drawing inference based on a marginal rank likelihood. No parametric forms of the baseline hazard need to be assumed in this semiparametric approach. Monte Carlo simulations and the Laplace approach are used to tackle the intractable integral in the likelihood function. Different spatial covariance structures are explored in simulations and the proposed methods are applied to the East Boston Asthma Study to detect prognostic factors leading to childhood asthma.
Stoler, J.M., Ryan, L.M. & Holmes, L.B. 2002, 'Alcohol dehydrogenase 2 genotypes, maternal alcohol use, and infant outcome', Journal Of Pediatrics, vol. 141, no. 6, pp. 780-785.
View/Download from: UTS OPUS or Publisher's site
Objective: To determine whether different alleles of the ADH2 gene (ADH2-1, ADH2-2 and ADH2-3) with differing levels of enzymatic activity can alter the risk of fetal alcohol effects. Study design: ADH2 genotypes were performed on 404 pregnant high-risk women and 139 infants as part of a larger study of alcohol use in pregnancy. Mothers were interviewed about alcohol use during pregnancy, and their infants were examined for alcohol-related features without knowledge of the exposure status. Results: The ADH2-1/3 genotype was more prevalent among black women (46%) than expected (33%); the rate among white women was low as expected (2%). More black women who reported high alcohol use during the pregnancy had the ADH2-1/3 genotype compared with those who reported no alcohol use (70% vs 44%). Sixty percent of the affected black infants had the ADH2-1/3 genotype compared with 29% of the unaffected infants (P < .045). The maternal genotype correlated with her chance of having an infant with alcohol-related physical features (odds ratio = 2.49). This association remained significant after accounting for confounders, such as smoking and maternal weight gain. Alcohol exposure was not significantly associated with infant outcome in black infants after accounting for genotype, smoking, and maternal weight gain, but this association could only be tested in 10 infants of mothers with high exposure. Conclusion: Women with the ADH2-1/3 genotype may be at greater risk for having an affected infant, which may be the result of greater ingestion of alcohol.
Clark, D.E. & Ryan, L.M. 2002, 'Concurrent prediction of hospital mortality and length of stay from risk factors on admission', Health Services Research, vol. 37, no. 3, pp. 631-645.
View/Download from: UTS OPUS or Publisher's site
Objective. To develop a method for predicting concurrently both hospital survival and length of stay (LOS) for seriously ill or injured patients, with particular attention to the competing risks of death or discharge alive as determinants of LOS. Data Sources. Previously collected 1995-1996 registry data on 2,646 cases of injured patients from three trauma centers in Maine. Study Design. Time intervals were determined for which the rates of discharge or death were relatively constant. Poisson regression was used to develop a model for each type of terminal event, with risk factors on admission contributing proportionately to the subsequent rates for each outcome in each interval. Mean LOS and cumulative survival were calculated from a combination of the resulting piecewise exponential models. Principal Findings. Age, Glasgow Coma Scale, Abbreviated Injury Scores, and specific mechanisms of injury were significant predictors of the rates of death and discharge, with effects that were variable in different time intervals. Predicted probability of survival and mean LOS from the model were similar to actual values for categorized patient groups. Conclusions. Piecewise exponential models may be useful in predicting LOS, especially if determinants of mortality are separated from determinants of discharge alive.
Ha, E., Cho, S., Chen, D., Chen, C., Ryan, L.M., Smith, T.J., Xu, X. & Christiani, D.C. 2002, 'Parental exposure to organic solvents and reduced birth weight', Archives of Environmental Health, vol. 57, no. 3, pp. 207-214.
View/Download from: UTS OPUS or Publisher's site
The authors investigated the association of birth weight with maternal and paternal exposure to organic solvents in 1,222 couples employed in a large petrochemical corporation in Beijing, China, during the period between 1994 and 1998. A trained interviewer assessed parental exposures to organic solvents. The authors used generalized additive models to examine the association between birth weight and parental exposure to organic solvents. After the authors adjusted for potential confounders, maternal exposure to solvents was significantly associated with reduced birth weight (-81.7 gm, 95% confidence interval = -106.3, -3.1), and reduced birth weights of female babies and of younger mothers' babies were statistically significant. Maternal exposure to organic solvents was associated with reduced birth weight in this population, but paternal exposure to organic solvents was not similarly associated.
Celedon, J.C., Litonjua, A.A., Ryan, L.M., Weiss, S.T. & Gold, D.R. 2002, 'Day care attendance, respiratory tract illnesses, wheezing, asthma, and total serum IgE level in early childhood', Archives of Pediatrics & Adolescent Medicine, vol. 156, pp. 241-245.
View/Download from: Publisher's site
Background: It has been hypothesized that day care-related infections may explain the inverse relation between day care attendance in early life and asthma in childhood. Objective: To examine the relation between day care attendance or respiratory tract illnesses in the first year of life and wheezing and asthma in the first 4 years of life among children with a parental history of atopy who were followed up from birth. Results: Day care attendance in the first year of life was inversely associated with geometric mean total serum IgE level (12.9 [+/- 1 SD = 3.3, 51.4] IU/mL for day care vs 18.5 [+/- 1 SD = 5.3, 64.7] IU/mL for no day cared P=.03) at 2 years of age but not significantly associated with wheezing at or after 2 years of age. Having at least 1 physician-diagnosed lower respiratory tract illness in the first year of life was significantly associated with recurrent wheezing (odds ratio [OR], 2.0; 95% confidence interval [CI], 1.0-4.1) and asthma (OR, 2.5; 95% CI, 1.1-5.5) at 4 years of age, but not with any wheezing (infrequent and frequent) at 3 years or older. Illnesses of the upper respiratory tract (greater than or equal to1 physician-diagnosed upper respiratory tract illness or greater than or equal to3 episodes of nasal catarrh) in the first year of life were associated with any wheezing (frequent and infrequent) between the ages of 1 and 4 years, but not with recurrent wheezing or asthma at 4 years of age. Conclusions: Our results suggest that among children with a parental history of atopy the protective effect of day care attendance in early life against the development of atopy has begun by 2 years of age, and that a protective effect of day care attendance in early life against wheezing may not be observed until after 4 years of age.
Celedon, J.C., Litonjua, A.A., Ryan, L.M., Weiss, S.T. & Gold, D.R. 2002, 'Bottle feeding in the bed or crib before sleep time and wheezing in early childhood', Pediatrics, vol. 110, no. 6, p. e77.
View/Download from: UTS OPUS
Objective. Bottle feeding in the bed or crib before sleep time has been associated with an increased risk of wheezing in the first year of life. We examined whether bottle feeding in the bed or crib before sleep time in the first year of life is associated with wheezing in the first 5 years of life. Methods. In a prospective cohort study of 448 children with parental history of atopy followed from birth, we examined the relation between the number of bimonthly parental reports of bottle feeding in the bed or crib before sleep time in the first year of life (range: 0-6 reports) and parental report of wheezing in the first 5 years of life. Additional outcome measures included recurrent wheezing (greater than or equal to2 episodes of wheezing in the previous year) and asthma (physician-diagnosed asthma and greater than or equal to1 episode of wheezing in the previous year) at the age of 5 years. Results. The risk of recurrent wheezing and asthma at 5 years of age increased significantly with each additional report of bottle feeding in the bed or crib before sleep time in the first year of life. The risk of wheezing between the ages of I and 5 years increased with each additional report of bottle feeding in the bed or crib before sleep time in the first year of life. As an example, a child whose parents reported bottle feeding in the bed or crib before sleep time on 3 occasions in the first year of life had 1.5 times higher risk of wheezing between the ages of 1 and 5 years than a child whose parents did not report bottle feeding in the bed or crib before sleep time in the first year of life (95% confidence interval for relative risk: 1.12-2.12). Conclusions. Among children with parental history of atopy, bottle feeding in the bed or crib before sleep time in the first year of life is a risk factor for recurrent wheezing and asthma at 5 years of age and a risk factor for wheezing between the ages of 1 and 5 years.
Thomson, C.C., Roberts, K., Curran, A., Ryan, L.M. & Wright, R.J. 2002, 'Caretaker-child concordance for child's exposure to violence in a preadolescent inner-city population', Archives of Pediatrics & Adolescent Medicine, vol. 156, no. 8, pp. 818-823.
View/Download from: UTS OPUS or Publisher's site
Background: Effective screening for exposure to violence (ETV) in the pediatric setting depends on informant reliability and recognition of patients at increased risk. Pediatricians screening for children's ETV often rely on parent reporting. Hypothesis: That there would be poor caretaker-child concordance given that children would report events occurring outside the home not witnessed by the caretaker and that ETV would be higher among immigrant families. Objectives: To examine concordance between caretaker and child self-report of the child's ETV in a preadolescent population and to explore factors related to increased risk. Design: Community-based survey. Setting: Urban community health center. Participants: One hundred sixty-five caretaker-child pairs. Methods: The ETV was assessed by means of a standardized interview questionnaire on location and frequency of ETV. A Rasch model was used to develop summary scores of ETV (frequency and severity). Results: Caretaker-child concordance on reports of child's ETV was poor. The K statistics ranged from-0.04 for seeing someone knifed to 0.39 for witnessing a shooting. Children reported ETV more often in their neighborhood or at school, whereas caretakers reported more events near or at home. Univariate predictors of child's self-reported ETV were female sex (beta +/- SE, -10.1 +/- 4.6; P=.03) and caretaker being divorced (beta +/- SE, 12.6 +/- 6.0; P=.04). In multivariate analyses, country of origin predicted child's ETV, adjusting for child's age and sex, and caretaker educational status and marital status. Conclusions: Caretakers and their children have poor agreement on reports of the child's ETV. Intervention strategies around ETV should include assessment of the child independent of caretaker report for preadolescents. Screening may be more effective if pediatricians are aware of factors related to increased risk, including immigration status and caretaker marital status.
Houseman, E.A., Ryan, L.M., Levy, J.I. & Spengler, J.D. 2002, 'Autocorrelation in real-time continuous monitoring of microenvironments', Journal of Applied Statistics, vol. 29, no. 6, pp. 855-872.
View/Download from: UTS OPUS or Publisher's site
Interpretation of continuous measurements in microenvironmental studies and exposure assessments can be complicated by autocorrelation, the implications of which are often not fully addressed. We discuss some statistical issues that arose in the analysis of microenvironmental particulate matter concentration data collected in 1998 by the Harvard School of Public Health. We present a simulation study that suggests that Generalized Estimating Equations, a technique often used to adjust for autocorrelation, may produce inflated Type I errors when applied to microenvironmental studies of small or moderate sample size, and that Linear Mixed Effects models may be more appropriate in small-sample settings. Environmental scientists often appeal to longer averaging times to reduce autocorrelation. We explore the functional relationship between averaging time, autocorrelation, and standard errors of both mean and variance, showing that longer averaging times impair statistical inferences about main effects. We conclude that, given widely available techniques that adjust for autocorrelation, longer averaging times may be inappropriate in microenvironmental studies.
Claeskens, G., Aerts, M., Molenberghs, G. & Ryan, L. 2002, 'Robust benchmark dose determination based on profile score methods', ENVIRONMENTAL AND ECOLOGICAL STATISTICS, vol. 9, no. 4, pp. 357-377.
View/Download from: Publisher's site
Chen, Y.C., Su, H.J.J., Guo, Y.L.L., Hsueh, Y.M., Smith, T.J., Ryan, L.M., Lee, M.S. & Christani, D.C. 2002, 'Arsenic methylation and bladder cancer risk', EPIDEMIOLOGY, vol. 13, no. 4, pp. S164-S164.
Clark, C., Berkman, L., Kawachi, I. & Ryan, L. 2002, 'Perceptions of neighborhood safety and consequences for mobility in elderly populations.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 155, no. 11, pp. s74-s74.
Duty, S.M., Silva, M.J., Barr, D.B., Brock, J.W., Ryan, L., Chen, Z., Herrick, R.F., Christiani, D. & Hauser, R. 2002, 'Urinary phthalate monoesters at general population exposure levels are associated with altered semen quality.', EPIDEMIOLOGY, vol. 13, no. 4, pp. S197-S197.
Parise, H., Wand, M., Ruppert, D. & Ryan, L.M. 2001, 'Incorporation of historical controls using semiparametric mixed models', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 50, no. 1, pp. 31-42.
Cho, S., Damokosh, A., Ryan, L.M., Chen, D., Hu, Y., Smith, T., Christiani, D. & Xu, X. 2001, 'Effects Of Exposure To Organic Solvents An Menstrual Cycle Length', Journal Of Occupational And Environmental Medicine, vol. 43, no. 6, pp. 567-575.
View/Download from: Publisher's site
To investigate the association between organic solvent exposure and menstrual disturbance, we conducted a cross-sectional study among 1408 petrochemical workers in China. Based on an industrial hygiene evaluation, we classified the workshops according to
O'dwyer, P., Manola, J., Valone, F., Ryan, L.M., Hines, J., Wadler, S., Haller, D., Arbuck, S., Weiner, L., Mayer, R. & Benson, A. 2001, 'Fluorouracil Modulation In Colorectal Cancer: Lack Of Improvement With N-phosphonoacetyl-l-aspartic Acid Or Oral Leucovorin Or Interferon, But Enhanced Therapeutic Index With Weekly 24-hour Infusion Schedule - An Eastern Cooperative Oncology Group/ca', Journal Of Clinical Oncology, vol. 19, no. 9, pp. 2413-2421.
Purpose: To investigate mechanism-directed regimens in maximizing the efficacy of fluorouracil (5-FU) in advanced colorected cancer. Patients and Methods: Based on promising phase II data, a randomized comparison of various methods for the biochemical mo
Holmes, L.B., Harvey, E.A., Coull, B.A., Huntington, K.B., Khoshbin, S., Hayes, A.M. & Ryan, L.M. 2001, 'The teratogenicity of anticonvulsant drugs', New England Journal Of Medicine, vol. 344, no. 15, pp. 1132-1138.
View/Download from: UTS OPUS or Publisher's site
Background: The frequency of major malformations, growth retardation, and hypoplasia of the midface and fingers, known as anticonvulsant embryopathy, is increased in infants exposed to anticonvulsant drugs in utero. However, whether the abnormalities are caused by the maternal epilepsy itself or by exposure to anticonvulsant drugs is not known. Methods: We screened 128,049 pregnant women at delivery to identify three groups of infants: those exposed to anticonvulsant drugs, those unexposed to anticonvulsant drugs but with a maternal history of seizures, and those unexposed to anticonvulsant drugs with no maternal history of seizures (control group). The infants were examined systematically for the presence of major malformations, signs of hypoplasia of the midface and fingers, microcephaly, and small body size. Results: The combined frequency of anticonvulsant embryopathy was higher in 223 infants exposed to one anticonvulsant drug than in 508 control infants (20.6 percent vs. 8.5 percent; odds ratio, 2.8; 95 percent confidence interval, 1.1 to 9.7). The frequency was also higher in 93 infants exposed to two or more anticonvulsant drugs than in the controls (28.0 percent vs. 8.5 percent; odds ratio, 4.2; 95 percent confidence interval, 1.1 to 5.1). The 98 infants whose mothers had a history of epilepsy but took no anticonvulsant drugs during the pregnancy did not have a higher frequency of those abnormalities than the control infants. Conclusions: A distinctive pattern of physical abnormalities in infants of mothers with epilepsy is associated with the use of anticonvulsant drugs during pregnancy, rather than with epilepsy itself.
Parise, H., Dinse, G. & Ryan, L.M. 2001, 'Flexible Estimates Of Tumour Incidence For Intermediately Lethal Tumours In A Typical Long-term Animal Bioassay', Journal Of The Royal Statistical Society Series C-applied Statistics, vol. 50, no. NA, pp. 171-185.
The estimation of the incidence of tumours in an animal carcinogenicity study is complicated by the occult nature of the tumours involved (i.e. tumours are not observable before an animal's death). Also, the lethality of tumours is generally unknown, making the tumour incidence function non-identifiable without interim sacrifices, cause-of-death data or modelling assumptions. Although Kaplan-Meier curves for overall survival are typically displayed, obtaining analogous plots for tumour incidence generally requires fairly elaborate model fitting. We present a case-study of tetrafluoroethylene to illustrate a simple method for estimating the incidence of tumours as a function of more easily estimable components. One of the components, tumour prevalence, is modelled by using a generalized additive model, which leads to estimates that are more flexible than those derived under the usual parametric models. A multiplicative assumption for tumour lethality allows for the incorporation of concomitant information, such as the size of tumours. Our approach requires only terminal sacrifice data although additional sacrifice data are easily accommodated. Simulations are used to illustrate the estimator proposed and to evaluate its properties. The method also yields a simple summary measure of tumour lethality, which can be helpful in interpreting the results of a study.
Coull, B.A., Hobert, J.P., Ryan, L.M. & Holmes, L.B. 2001, 'Crossed random effect models for multiple outcomes in a study of teratogenesis', Journal of the American Statistical Association, vol. 96, no. 456, pp. 1194-1204.
View/Download from: UTS OPUS or Publisher's site
Human teratogens often manifest themselves through a broad spectrum of adverse effects. Although often not serious when considered individually, such outcomes taken together may represent a syndrome that can lead to serious developmental problems. Accordingly, studies that investigate the effect of human teratogens on fetal development typically record the presence or absence of a multitude of abnormalities, resulting in the data of multivariate binary form for each infant. Such studies typically have three objectives: (1) estimate an overall effect of exposure across outcomes, (2) identify subjects having the syndrome, and (3) identify those outcomes that constitute the syndrome so that doctors know what to look for when diagnosing the syndrome in other exposed newborns. This article proposes the use of a logistic regression model with crossed random effect structure to address all three questions simultaneously. We use the proposed models to analyze data from a study investigating the effects of in utero antiepileptic drug exposure on fetal development.
Cook, R.J., Brumback, B.B., Wigg, M.B. & Ryan, L.M. 2001, 'Synthesis of evidence from epidemiological studies with interval censored exposure due to grouping', Biometrics, vol. 57, no. 3, pp. 671-680.
View/Download from: UTS OPUS or Publisher's site
We describe a method for assessing dose-response effects from a series of case-control and cohort studies in which the exposure information is interval censored. The interval censoring of the exposure variable is dealt with through the use of retrospective models in which the exposure is treated as a multinomial response and disease status as a binary covariate. Polychotomous logistic regression models are adopted in which the dose-response relationship between exposure and disease may be modeled in a discrete or continuous fashion. Partial conditioning is possible to eliminate some of the nuisance parameters. The methods are applied to the motivating study of the relationship between chorionic villus sampling and the occurrence of terminal transverse limb reduction.
Corcoran, C., Ryan, L.M., Senchaudhuri, P., Mehta, C., Patel, N. & Molenberghs, G. 2001, 'An exact trend test for correlated binary data', Biometrics, vol. 57, no. 3, pp. 941-948.
View/Download from: UTS OPUS or Publisher's site
The problem of testing a dose-response relationship in the presence of exchangeably correlated binary data has been addressed using a variety of models. Most commonly used approaches are derived from likelihood or generalized estimating equations and rely on large-sample theory to justify their inferences. However, while earlier work has determined that these methods may perform poorly for small or sparse samples, there are few alternatives available to those faced with such data. We propose an exact trend test for exchangeably correlated binary data when groups of correlated observations are ordered. This exact approach is based on an exponential model derived by Molenberghs and Ryan (1999) and Ryan and Molenberghs (1999) and provides natural analogues to Fisher's exact test and the binomial trend test when the data are correlated. We use a graphical method with which one can efficiently compute the exact tail distribution and apply the test to two examples.
Chen, C., Wang, X., Chen, D., Li, G., Ronnenberg, A., Watanabe, H., Wang, X., Ryan, L.M., Christiani, D.C. & Xu, X. 2001, 'Tofu consumption and blood lead levels in young Chinese adults', American Journal of Epidemiology, vol. 153, no. 12, pp. 1206-1212.
View/Download from: UTS OPUS or Publisher's site
Tofu is a commonly consumed food in China. Tofu may interfere with lead absorption and retention because of its high calcium content. In this observational study, the authors examined whether dietary tofu intake was associated with blood lead levels among young adults in Shenyang, China. The analyses included 605 men and 550 women who completed baseline questionnaires and had blood lead measurements taken in 1996-1998 as part of a prospective cohort study on reproductive health. Mean blood lead levels were 13.2 mug/dl in men and 10.1 mug/dl in women. Blood lead levels were negatively associated with tofu intake in both genders. A linear trend test showed a 3.7% (0.5-mug/dl) decrease in blood lead level with each higher category of tofu intake (p = 0.003). The highest tofu intake group (greater than or equal to 750 g/week) had blood lead levels 11.3% lower (95% confidence interval: 4.1, 18.0) than those of the lowest tofu intake group (< 250 g/week). In all regression models, data were adjusted for gender, age, height, body mass index, district, cigarette smoking, alcohol drinking, education, occupation, use of vitamin supplements, season, and dietary intake of meat, fish, vegetables, eggs, and milk. In conclusion, the authors found a significant inverse dose-response relation between tofu consumption and blood lead levels in this Chinese population.
Korrick, S.A., Chen, C., Damokosh, A.I., Ni, J., Liu, X., Cho, S., Altshul, L., Xu, X. & Ryan, L.M. 2001, 'Association of DDT with spontaneous abortion: A case-control study', Annals of Epidemiology, vol. 11, no. 7, pp. 491-496.
View/Download from: UTS OPUS or Publisher's site
PURPOSE: Spontaneous abortion (SAB), the most common adverse pregnancy outcome, affects similar to 15% of clinically recognized pregnancies. Except for advanced maternal age and smoking, there are not well-established risk factors for SAB. Animal models associate increased fetal resorption or abortion with exposure to the pesticide dichlorodiphenyl trichloroethane (DDT), but epidemiologic investigations of DDT and SAB are inconsistent. We undertook a pilot investigation of the hypothesized association of DDT with SAB. METHODS: Participants in this case-control study were selected from a longitudinal study of reproductive effects of rotating shifts among female Chinese textile workers who were married, ages 22-34, nulliparous without history of SAB or infertility, and planning pregnancy. From 412 pregnancies, 42 of which ended in SAB, 15 SAB cases and 15 full-term controls were randomly selected and phlebotomized. Serum was analyzed for p,p ' -DDT, o,p ' -DDT, their metabolites (DDE and DDD), and other organochlorines including polychlorinated biphenyls. RESULTS: Cases and controls were nonsmokers and did not differ in age (mean 25 years), body mass index (BMI), passive smoke exposure, or workplace exposures. Cases had significantly (p < 0.05) higher serum levels of p,p ' -DDE (22 vs. 12 ng/g) and o,p ' -DDE (0.09 vs. 0.05 ng/g) than controls. After adjustment for age and BMI, each ng/g serum increase in p,p ' -DDE was associated with a 1.13 (Cl, 1.02-1.26) increased odds of SAB. With adjustment of serum DDE levels for excretion via breastfeeding, DDE-associated increased odds of SAB remained significant with up to 7% declines in maternal serum DDE levels for each month of breastfeeding. CONCLUSIONS: A potential increased risk of SAB is associated with maternal serum DDE levels.
Levy, J.I., Houseman, E.A., Spengler, J.D., Loh, P. & Ryan, L.M. 2001, 'Fine particulate matter and polycyclic aromatic hydrocarbon concentration patterns in Roxbury, Massachusetts: a community-based GIS analysis', Environmental Health Perspectives, vol. 109, no. 4, pp. 341-347.
View/Download from: UTS OPUS or Publisher's site
Given an elevated prevalence of respiratory disease and density of pollution sources, residents of Roxbury, Massachusetts, have been interested in better understanding their exposures to air pollution. To determine whether local transportation sources contribute significantly to exposures, we conducted a community-based pilot investigation to measure concentrations of fine particulate matter (particulate matter < 2.5 <mu>m; PM2.5) and particle-bound polycyclic aromatic hydrocarbons (PAHs) in Roxbury in the summer of 1999. Community members carried portable monitors on the streets in a 1-mile radius around a large bus terminal to create a geographic information system (GIS) map of concentrations and gathered data on site characteristics that could predict ambient concentrations. Both PM2.5, and PAH concentrations were greater during morning rush hours and on weekdays. In linear mixed-effects regressions controlling for temporal autocorrelation, PAH concentrations were significantly higher with closer proximity to the hus terminal (p < 0.05), and both pollutants were elevated, but not statistically significantly so, on bus routes. Regressions on a subset of measurements for which detailed site characteristics were gathered showed higher concentrations of both pollutants on roads reported to have heavy bus traffic. Although a more comprehensive monitoring protocol would be needed to develop robust predictive functions for air pollution, our study demonstrates that pollution patterns in an urban area can be characterized with limited monitoring equipment and that university-community partnerships can yield relevant exposure information.
Scott, L., Warram, J.H., Hanna, L.S., Laffel, L.M., Ryan, L.M. & Krolewski, A.S. 2001, 'A nonlinear effect of hyperglycemia and current cigarette smoking are major determinants of the onset of microalbuminuria in type 1 diabetes', Diabetes, vol. 50, no. 12, pp. 2842-2849.
View/Download from: UTS OPUS or Publisher's site
Cigarette smoking and poor glycemic control are risk factors for diabetic nephropathy in type 1 diabetes. However, the specifics of the relation of these risk factors to the onset of this complication have not been elucidated. To investigate these issues, we followed for 4 years 943 Joslin Clinic patients aged 15-44 years with type I diabetes who had normoalbuminuria during the 2-year baseline period. Microalbuminuria developed in 109 of the 943 individuals, giving an incidence rate of 3.3/100 person-years. The risk of onset of microalbuminuria was predicted somewhat more precisely by the measurements during the 1st and 2nd years preceding onset than by all the measurements during the longer (4-year) interval, suggesting attenuation of the impact of past hyperglycemia over time. Point estimates of the incidence rate (per 100 person-years) according to quartiles of HbA(1c) during the 1st and 2nd years preceding the outcome were 1.3 in the 1st, 1.5 in the 2nd, 3.1 in the 3rd, and 6.9 in the 4th (P=1.3 x 10(-9)). Point estimates of the incidence rate (per 100 person-years) according to smoking status were 7.9 for current smokers, 1.8 for past smokers, and 2.2 for those who had never smoked (P=2.0 x 10(-7)). In a multiple logistic model, the independent effects of HbA(1c) level and cigarette smoking remained highly significant, but their magnitudes were reduced. Using the same covariates in a generalized additive model, we examined the shape of the relationship between HbA(1c) and onset of microalbuminuria and found significant nonlinearity in the logarithm of odds scale (P=0.04). The slope was steeper with HbA(1c) >8% than <8%. Furthermore, the change in the slope was magnified among current smokers. In conclusion, patients with type I diabetes who smoke and have an HbA(1c) > 8% have the highest risk of onset of microalbuminuria.
Ku, C.H., Ryan, L. & Christiani, D. 2001, 'Skin test reactivity to fungi and dust mites and serum IgE concentrations in an occupational cohort', EPIDEMIOLOGY, vol. 12, no. 4, pp. S58-S58.
Cho, S.I., Goldman, M.B., Ryan, L.M., Chen, C., Damokosh, A.I., Christiani, D.C., Lasley, B.L., O'Connor, J.F., Wilcox, A.J. & Xu, X. 2001, 'Use of a urine human chorionic gonadotropin (hCG) biomarker to detect early pregnancy loss.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 153, no. 11, pp. S160-S160.
Wang, X., Chen, C., French, J., Cho, S.I., Liu, X., Ryan, L., Christiani, D., O'Connor, J., Lasley, B., Wilcox, A. & Xu, X. 2001, 'Time to pregnancy, subclinical conception, and early fetal loss: A population-based prospective study.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 153, no. 11, pp. S159-S159.
French, J., Chen, C., Ryan, L., Wang, X. & Xu, X. 2001, 'Bayesian method for handling missing hCG data in the detection of early fetal loss.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 153, no. 11, pp. S143-S143.
Wang, X.B., Zuckerman, B., Kaufman, G., Wise, P., Hill, M., Niu, T.H., Ryan, L., Wu, D. & Xu, X.P. 2001, 'Molecular epidemiology of preterm delivery: methodology and challenges', PAEDIATRIC AND PERINATAL EPIDEMIOLOGY, vol. 15, pp. 63-77.
View/Download from: Publisher's site
Ku, C.H., Ryan, L. & Christiani, D. 2001, 'An application of a Poisson model with GEE adjustments for assessment of occupational exposure to fungi', EPIDEMIOLOGY, vol. 12, no. 4, pp. S59-S59.
Parise, H., Dinse, G.E. & Ryan, L.M. 2001, 'Flexible estimates of tumour incidence for intermediately lethal tumours in a typical long-term animal bioassay', Journal of the Royal Statistical Society. Series C: Applied Statistics, vol. 50, no. 2, pp. 171-185.
Zanobetti, A., Wand, M., Schwartz, J. & Ryan, L.M. 2000, 'Generalized additive distributed lag models: quantifying mortality displacement', Biostatistics, vol. 1, no. 3, pp. 279-292.
View/Download from: UTS OPUS or Publisher's site
Brown, K., Kuo, T., Guo, H., Ryan, L.M. & Abernathy, C. 2000, 'Sensitivity Analysis Of Us Epa's Estimates Of Skin Cancer Risk From Inorganic Arsenic In Drinking Water', Human And Ecological Risk Assessment, vol. 6, no. 6, pp. 1055-1074.
View/Download from: Publisher's site
The current U.S. Environmental Protection Agency's (USEPA's) risk analysis on the Integrated Risk Information System (IRIS) for arsenic in drinking water is based on an epidemiological study of skill cancer in Taiwan. Assumptions used in the USEPA applic
Carlin, J.B., Ryan, L.M., Harvey, E.A. & Holmes, L.B. 2000, 'Anticonvulsant teratogenesis 4: Inter-rater agreement in assessing minor physical features related to anticonvulsant therapy', Teratology, vol. 62, no. 6, pp. 406-412.
View/Download from: UTS OPUS or Publisher's site
Background: We report on inter-rater agreement in the assessment of newborn infants with respect to a range of minor physical features in a cohort study of the fetal effects of maternal anticonvulsant use during pregnancy. Methods: Infants from three groups (exposed to anticonvulsants, seizure history but no medication exposure, and unexposed controls) were examined by both a pediatrician/teratologist, who was blinded with respect to the mother's exposure status, and an unblinded research assistant. Agreement on assessments for selected anomalies associated with anticonvulsant therapy was measured by kappa-statistics, as well as by more sensitive log-linear modeling techniques, which allow examination of possible covariate effects on the strength of agreement. Although the physician and research assistant agreed on a high proportion of cases (80-90%), kappa values were modest (0.2-0.5), partly because of the low prevalence of the anomalies considered. To explore how agreement varies within subgroups, we used recently developed methods for studying agreement based on log-linear models. Results: Log-linear modeling indicated that there was substantial variation in pattern of agreement between different individual research assistants but that other factors (e.g., exposure category, sex, and birthweight) did not appear to be related to agreement. Our results suggest that research assistants with more experience showed the highest degree of agreement with the physicians. Conclusions: Our results have implications for both clinical practice and epidemiologic research and underline the importance of thorough training of staff in the definitions to be used and also the need for multiple independent assessments of these subtle anomalies.
Padungtod, C., Savitz, D., Overstreet, J., Christiani, D., Ryan, L.M. & Xu, X. 2000, 'Occupational Pesticide Exposure And Semen Quality Among Chinese Workers', Journal Of Occupational And Environmental Medicine, vol. 42, no. 10, pp. 982-992.
View/Download from: Publisher's site
This study investigated the association between occupational pesticide exposure and semen quality among Chinese workers. Male workers, 32 who were exposed to organophosphate pesticides and 43 who were not exposed were recruited from two nearby factories
Ibrahim, J., Chen, M. & Ryan, L.M. 2000, 'Bayesian Variable Selection For Time Series Count Data', Statistica Sinica, vol. 10, no. 3, pp. 971-987.
We consider a parametric model for time series of counts by constructing a likelihood-based generalization of a model considered by Zeger (1988). We consider a Bayesian approach and propose a class of informative prior distributions for the model paramet
Lu, M.C., Sammel, M.D., Cleveland, R.H., Ryan, L.M. & Holmes, L.B. 2000, 'Digit effects produced by prenatal exposure to antiepileptic drugs', Teratology, vol. 61, no. 4, pp. 277-283.
View/Download from: UTS OPUS or Publisher's site
The hypothesis tested was that digit anomalies among individuals exposed in utero to antiepileptic drugs (AED) are best identified by a systematic search, including radiographs and dermatoglyphics, rather than relying only on visual inspection, A systematic search was made for five types of digit abnormalities in 46 AED-exposed individuals ages 5-29 years in comparison with controls: visible anomalies, size of fingernails, dermal ridge patterns, length of metacarpals and phalanges, and qualitative changes in the distal phalanges. Among the AED-exposed, nail size was not decreased. However, there was a 10.8% frequency of digit anomalies, a 12% frequency of three or more arch patterns, and significant shortening and qualitative changes in the distal phalanges, all of which are consistent with the fetal effects of AED. Among the 42 individuals who underwent all evaluations, 14.3% had two or more of these abnormalities, most of which would not be identified by clinical inspection. This frequency is much higher in these AED-exposed individuals than in the general population. Radiographs in 13 individuals over a period of several years showed that the changes in the phalanges and metacarpals persisted.
Ryan, L.M. 2000, 'Statistical issues in toxicology', Journal of the American Statistical Association, vol. 95, no. 449, pp. 304-308.
View/Download from: UTS OPUS or Publisher's site
NA
Goetghebeur, E. & Ryan, L.M. 2000, 'Semiparametric Regression Analysis of Interval-Censored Data', Biometrics, vol. 56, no. 4, pp. 1139-1144.
View/Download from: UTS OPUS or Publisher's site
We propose a semiparametric approach to the proportional hazards regression analysis of interval-censored data. An EM algorithm based on an approximate likelihood leads to an M-step that involves maximizing a standard Cox partial likelihood to estimate regression coefficients and then using the Breslow estimator for the unknown baseline hazards. The E-step takes a particularly simple form because all incomplete data appear as linear terms in the complete-data log likelihood. The algorithm of Turnbull (1976, Journal of the Royal Statistical Society, Series B 38, 290-295) is used to determine times at which the hazard can take positive mass. We found multiple imputation to yield an easily computed variance estimate that appears to be more reliable than asymptotic methods with small to moderately sized data sets. In the right-censored survival setting, the approach reduces to the standard Cox proportional hazards analysis, while the algorithm reduces to the one suggested by Clayton and Cuzick (1985, Applied Statistics 34, 148-156). The method is illustrated on data from the breast cancer cosmetics trial, previously analyzed by Finkelstein (1986, Biometrics 42, 845-854) and several subsequent authors.
Contreras, M. & Ryan, L.M. 2000, 'Fitting nonlinear and constrained generalized estimating equations with optimization software', Biometrics, vol. 56, no. 4, pp. 1268-1271.
View/Download from: UTS OPUS or Publisher's site
In this article, we present an estimation approach for solving nonlinear constrained generalized estimating equations that can be implemented using object-oriented software for nonlinear programming, such as nlminb in Splus or fmincon and Isqnonlin in Matlab. We show how standard estimating equation theory includes this method as a special case so that our estimates, when unconstrained, will remain consistent: and asymptotically normal. To illustrate this method, we fit a nonlinear dose-response model with nonnegative mixed bound constraints to clustered binary data from a developmental toxicity study. Satisfactory confidence intervals are found using a nonparametric bootstrap method when a common correlation coefficient is assumed for all the dose groups and for some of the dose-specific groups.
Ryan, L.M., Brumback, B.A., Schwartz, J.D., Neas, L.M., Stark, P.C. & Burge, H.A. 2000, 'Transitional regression models, with application to environmental time series', Journal Of The American Statistical Association, vol. 95, no. 449, pp. 16-27.
View/Download from: UTS OPUS
Environmental epidemiologists often encounter time series data in the form of discrete or other nonnormal outcomes: for example, in modeling the relationship between air pollution and hospital admissions or mortality rates. We present a case study examining the association between pollen counts and meteorologic covariates. Although such time series data are inadequately described by standard methods for Gaussian time series, they are often autocorrelated, and warrant an analysis beyond those provided by ordinary generalized linear models (GLMs). Transitional regression models (TRMs), signifying nonlinear regression models expressed in terms of conditional means and variances given past observations, provide a unifying framework for two mainstream approaches to extending the GLM for autocorrelated data. The first approach models current outcomes with a GLM that incorporates past outcomes as covariates, whereas the second models individual outcomes with marginal GLMs and then couples the error terms With an autoregressive covariance matrix. Although the two approaches coincide for the Gaussian GLM, which serves as a helpful introductory example, in general they yield fundamentally different models. We analyze the pollen study using TRM's of both types and present parameter estimates together with asymptotic and bootstrap standard errors. In several cases we find evidence of residual autocorrelation; however, when we relax the TRM to allow for a nonparametric smooth trend, the autocorrelation disappears. This kind of trade-off between autocorrelation and flexibility is to be expected, and has a natural interpretation in terms of the covariance function for a nonparametric smoother. We provide an algorithm for fitting these flexible TRM's that is relatively easy to program with the generalized additive model software in S-PLUS.
Lin, X., Ryan, L.M., Sammel, M., Zhang, D., Padungtod, C. & Xu, X. 2000, 'A scaled linear mixed model for multiple outcomes', Biometrics, vol. 56, no. 2, pp. 593-601.
View/Download from: UTS OPUS or Publisher's site
We propose a scaled linear mixed model to assess the effects of exposure and other covariates on multiple continuous outcomes. The most general form of the model allows a different exposure effect for each outcome. An important special case is a model that represents the exposure effects using a common global measure that can be characterized in terms of effect sizes. Correlations among different outcomes within the same subject are accommodated using random effects. We develop two approaches to model fitting, including the maximum likelihood method and the working parameter method. A key feature of both methods is that they can be easily implemented by repeatedly calling software for fitting standard linear mixed models, e.g., SAS PROC MIXED. Compared to the maximum likelihood method, the working parameter method is easier to implement and yields fully efficient estimators of the parameters of interest. We illustrate the proposed methods by analyzing data from a study of the effects of occupational pesticide exposure on semen quality in a cohort of Chinese men.
Bellamy, S.L., Gibberd, R., Hancock, L., Howley, P., Kennedy, B., Klar, N., Lipsitz, S. & Ryan, L.M. 2000, 'Analysis of dichotomous outcome data for community intervention studies', Statistical Methods in Medical Research, vol. 9, no. 2, pp. 135-159.
View/Download from: UTS OPUS or Publisher's site
Community intervention trials are becoming increasingly popular as a tool for evaluating the effectiveness of health education and intervention strategies. Typically, units such as households, schools, towns, counties, are randomized to receive either intervention or control, then outcomes are measured on individuals within each of the units of randomization. It is well recognized that the design and analysis of such studies must account for the clustering of subjects within the units of randomization. Furthermore, there are usually both subject level and cluster level covariates that must be considered in the modelling process. While suitable methods are available for continuous outcomes, data analysis is more complicated when dichotomous outcomes are measured on each subject. This paper will compare and contrast several of the available methods that can be applied in such settings, including random effects models, generalized estimating equations and methods based on the calculation of `design effects', as implemented in the computer package SUDAAN. For completeness, the paper will also compare these methods of analysis with more simplistic approaches based on the summary statistics. All the methods will be applied to a case study based on an adolescent anti-smoking intervention in Australia. The paper concludes with some general discussion and recommendations for routine design and analysis.
Brumback, B.A., Cook, R.J. & Ryan, L.M. 2000, 'A meta-analysis of case-control and cohort studies with interval-censored exposure data: application to chorionic villus sampling', Biostatistics, vol. 1, no. 2, pp. 203-217.
Thurston, S.W., Ryan, L.M., Christiani, D.C., Snow, R., Carlson, J., You, L., Cui, S., Ma, G., Wang, L., Huang, Y. & Xu, X. 2000, 'Petrochemical exposure and menstrual disturbances', American Journal of Industrial Medicine, vol. 38, no. 5, pp. 555-564.
View/Download from: UTS OPUS
Background Art exploratory, cross-sectional retrospective study was conducted to examine the effects of benzene exposure on menstrual problems. Methods The study was based on a survey administered to over 3,000 women who worked in a large petrochemical company in Beijing, China. An abnormal menstrual cycle length (AMCL), defined as an average menstrual cycle length of greater than 35 clays or less than 21 days, is the major outcome of interest. Results After 7 years of benzene exposure, the adjusted odds ratio of having AMCL for each additional 5 years of exposure was 1.71 (95% CI 1.27-2.31). Feeling stressed at work was also all important predictor Conclusions This study suggests a significant association of benzene exposure and perceived stress with menstrual disturbance. A prospective study is needed to confirm this finding.
Holmes, L.B., Rosenberger, P.B., Harvey, E.A., Khoshbin, S. & Ryan, L.M. 2000, 'Intelligence and physical features of children of women with epilepsy', Teratology, vol. 61, pp. 196-202.
View/Download from: UTS OPUS
The teratogenicity of maternal epilepsy has been attributed to several factors, including the antiepileptic drugs taken to prevent seizures during pregnancy, the occurrence of seizures during pregnancy, and the factors in the mother that caused her to have epilepsy. We have addressed the hypothesis that the children of women who have a history of epilepsy (seizure history), but who took no antiepileptic drugs (AED) and had no tonic-clonic seizures in pregnancy, have an increased risk of malformations and diminished intelligence. The frequency of cognitive dysfunction was determined in 57 seizure history and 57 matched control children aged 6-16 years. The masked evaluation of the children included a physical and neurologic examination and testing with the Wechsler Intelligence Scale for Children-Revised (WISC-R) and a systematic physical examination for the features of the fetal AED syndrome. The evaluation of both parents of each child included a test of reasoning (Ravens Progressive Matrix) and a physical examination. There were no differences between the two groups of children in either IQ scores or physical features; none of the seizure history children was judged to have the "anticonvulsant face" or digit hypoplasia. This study had 80% power to rule out a difference of seven or move IQ points between the two groups, based on a two-sided test at a 5% level of significance. Our confidence in concluding that there was no difference between seizure history and control infants was strengthened by the fact that no statistically significant differences were observed with respect to multiple outcomes, including eight related measures of intelligence. Thirty (53%) of the seizure history mothers resumed taking AED after the birth of the child we evaluated. Additional studies are needed to address the teratogenicity of the antiepileptic drugs as monotherapy.
Morales, K.H., Ryan, L.M., Kuo, T.L., Wu, M. & Chen, C.J. 2000, 'Risk of internal cancers from arsenic in drinking water', Environmental Health Perspectives, vol. 108, no. 7, pp. 655-661.
View/Download from: UTS OPUS or Publisher's site
The U.S. Environmental Protection Agency is under a congressional mandate to revise its current standard for arsenic in drinking water. We present a risk assessment for cancers of the bladder, liver, and lung from exposure to arsenic in water, based on data from 42 villages in an arseniasisendemic region of Taiwan. We calculate excess lifetime risk estimates for several variations of the generalized linear model and for the multistage-Weibull model. Risk estimates are sensitve to the model choice, to whether or not a comparison population is used to define the unexposed diease mortality rates, and to whether the comparison population is all of Taiwan or just the southwestem region. Some factors that may affect risk could not be evaluated quantitativel. the ecologic nature of the data, the nutritional status of the study population, and the dietary intake of arsenic. Despite all of these sources of uncertainty, however, our analysis suggests that the current standard of 50 pg/L is asoatted with a substantial increased risk of cancer and is not sufficiently protective of public health.
Levy, J.I., Houseman, E.A., Ryan, L.M., Richardson, D. & Spengler, J.D. 2000, 'Particle concentrations in urban microenvironments', Environmental Health Perspectives, vol. 108, no. 11, pp. 1051-1057.
View/Download from: UTS OPUS or Publisher's site
Although ambient particulate matter has been associated with a range of health outcomes, the health risks for individuals depend in part on their daily activities. Information about particle mass concentrations and size distributions in indoor and outdoor microenvironments can help identify high-risk individuals and the significant contributors to personal exposure. To address these issues in an urban setting, we measured particle count concentrations in four size ranges and particulate matter (3/4) 10 microm (PM(10)) concentrations outdoors and in seven indoor microenvironments in Boston, Massachusetts. Particle counts and PM(10) concentrations were continuously measured with two light-scattering devices. Because of the autocorrelation between sequential measurements, we used linear mixed effects models with an AR-1 autoregressive correlation structure to evaluate whether differences between microenvironments were statistically significant. In general, larger particles were elevated in the vicinity of significant human activity, and smaller particles were elevated in the vicinity of combustion sources, with indoor PM(10) concentrations significantly higher than the outdoors on buses and trolleys. Statistical models demonstrated significant variability among some indoor microenvironments, with greater variability for smaller particles. These findings imply that personal exposures can depend on activity patterns and that microenvironmental concentration information can improve the accuracy of personal exposure estimation.
Wang, X., Chen, D., Niu, T., Wang, Z., Wang, L., Ryan, L.M., Smith, T., Christiani, D.C., Zuckerman, B. & Xu, X. 2000, 'Genetic susceptibility to benzene and shortened gestation: Evidence of gene-environment interaction', American Journal of Epidemiology, vol. 152, no. 8, pp. 693-700.
View/Download from: Publisher's site
This study investigated whether the association between low level benzene exposure and shortened gestation is modified by two susceptibility genes, CYP1A1 and GSTT1. This report includes 542 (302 nonexposed, 240 benzene-exposed) nonsmoking and nondrinking mothers of singleton live births at Beijing Yanshan Petrochemical Corporation between June 1995 and June 1997. Epidemiologic and clinical data and blood samples were obtained from mothers. Multiple linear regression models were used to estimate the associations of benzene exposure and genetic susceptibility with gestational age, adjusting for maternal age, education, parity, stress, passive smoking, prepregnancy weight and height, and infant's sex. Without consideration of genotype, benzene exposure was associated with a decrease in mean gestational age of 0.29 (standard error (SE), 0.12) week. When stratified by the maternal CYP1A1 genotype, the estimated decrease was 0.54 (SE, 0.12) week for the AA group, which was significantly greater (p = 0.003) than that for the Aa/aa group, which showed no decrease in gestational age. When both CYP1A1 and GSTT1 were considered, the greatest decrease was found among exposed mothers with the CYP1A1 AA-GSTT1 absent group (0.79 (SE, 0.25) week) and the CYP1A1 AA-GSTT1 present group (0.50 (SE, 0.22) week). Among the nonexposed, genetic susceptibility alone did not confer a significant adverse effect. This study provides evidence of gene-environment interaction and supports further assessment of the role of genetic susceptibility in the evaluation of reproductive toxins.
Geys, H., Molenberghs, G., Declerck, L. & Ryan, L.M. 2000, 'Flexible quantitative risk assessment for developmental toxicity based on fractional polynomial predictors', Biometrical Journal, vol. 42, no. 3, pp. 279-302.
View/Download from: UTS OPUS
Risk assessment for developmental toxicity studies in rodents is faced with the fairly involved data structure of clustered multivariate binary outcomes. While likelihood methods for this setting do not abound, we show that a conditional model, combined with pseudo-likelihood inference and fractional polynomial predictor functions, as proposed by ROYSTON and ALTMAN (1994), are a promising way forward. The methods are illustrated using teratology data collected under the National Toxicology Program.
Chen, D., Cho, S., Chen, C., Wang, X., Damokosh, A.I., Ryan, L.M., Smith, T.J., Christiani, D.C. & Xu, X. 2000, 'Exposure to benzene, occupational stress, and reduced birth weight', Occupational and Environmental Medicine, vol. 57, pp. 661-667.
View/Download from: UTS OPUS or Publisher's site
Objectives-The association between birth weight and exposure to benzene, work stress, and other occupational and environmental hazards was investigated. Methods-in a large petrochemical industry, 792 pregnant workers were enrolled and followed up through delivery between May 1996 and December 1998. Exposure to benzene and other solvents was assessed by an industrial hygienist based on each woman's job title and workplace information. Other occupational and environmental exposures and personal information, including perceived work stress, exposure to noise, physical exertion at work, and passive smoking, were obtained by an interview questionnaire. Univariate and multivariate regression models were used to examine the individual and combined associations of occupational and environmental exposures with birth weight, with adjustment for major confounders including gestational age. Results-In the univariate model, birth weight was negatively associated with exposure to benzene (-58 g (95% confidence interval (95% CI), -115 to -2)) and with work stress (-84 g (95% CI, -158 to -10)). In the multivariate model, there was a significant interaction between exposure to benzene and work stress relative to reduced birth weight, after adjustment for other environmental and occupational exposures and personal variables. Adjusted mean birth weight was 3445 g (95% CI 3401 to 3489) among those with neither exposure, 3430 g for those with exposure to benzene only, 3426 g for those with work stress only, and 3262 g (95% CI 3156 to 3369) for those with both exposures. In other words, there was 183 g (95% CI 65 to 301) reduction in birth weight among those with both exposure to benzene and work stress compared with those with neither exposure. Other work or environmental factors could not explain these findings. Conclusions-Low level exposure to benzene and work stress interact to reduce birth weight in this population.
Declerck, L., Molenberghs, G., Aerts, M. & Ryan, L.M. 2000, 'Litter-based methods in developmental toxicity risk assessment', Environmental and Ecological Statistics, vol. 7, pp. 57-76.
View/Download from: UTS OPUS or Publisher's site
Developmental toxicity experiments are designed to assess potential adverse effects of drugs and other exposures on developing fetuses from pregnant dams. Extrapolation to humans is a very difficult problem. An important issue here is whether risk assessment should be based on the fetus or the litter level. In this paper, fetus and litter-based risks that properly account for cluster size are defined and compared for the beta-binomial model and a conditional model for clustered binary data. It is shown how the hierarchical structure of non-viable implants and viable but malformed offspring can be incorporated. Risks based on a joint model for death/resorption and malformation are contrasted with risks based on an adverse event defined as either death/resorption or malformation. The estimation of safe exposure levels for all risk types is discussed and it is shown how estimation of the cluster size distribution affects variance estimation. The methods are applied to data collected under the National Toxicology Program and in large sample simulations.
Cho, S.I., Goldman, M.B., Ryan, L., Chen, C., Damokosh, A.I., Christiani, D.C., Lasley, B.L., O'Connor, J.F., Wilcox, A.J. & Xu, X. 2000, 'Reliability of HCG data interpretation.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 151, no. 11, pp. S48-S48.
Warram, J.H., Scott, L.J., Hanna, L.S., Wantman, M., Cohen, S.E., Laffel, L.M.B., Ryan, L. & Krolewski, A.S. 2000, 'Progression of microalbuminuria to proteinuria in type 1 diabetes - Nonlinear relationship with hyperglycemia', DIABETES, vol. 49, no. 1, pp. 94-100.
View/Download from: Publisher's site
Chen, K.L.B., Gertig, D., Ryan, L., Willett, W., Christiani, D. & Hunter, D. 2000, 'Toenail arsenic levels and squamous cell carcinoma of the skin', EPIDEMIOLOGY, vol. 11, no. 4, pp. S58-S58.
Zanobetti, A., Wand, M., Schwartz, J. & Ryan, L. 2000, 'Mortality displacement in Milan, Italy.', EPIDEMIOLOGY, vol. 11, no. 4, pp. S77-S77.
Stoler, J.M., McGuirk, C., Holmes, L.B., Ryan, L. & Lieberman, E. 2000, 'Untitled - Response', GENETICS IN MEDICINE, vol. 2, no. 5, pp. 300-301.
View/Download from: Publisher's site
Cho, S.I., Chen, C., Chen, D., Damokosh, A.I., Yiu, L., Ryan, L., Smith, T.J., Lasley, B.L., O'Connor, J.F., Wang, L., Wilcox, A.J., Christiani, D.C. & Xu, X. 2000, 'Benzene exposure and time to conception.', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 151, no. 11, pp. S13-S13.
Clayton, D., Campbell, M.J., Longford, N.T., Armstrong, B., Richardson, S., Rushton, L., Raab, G.M., Quinn, M., Nelder, J.A., Bland, M., Macfarlane, A., Dryden, I.L., Smith, R.I., Berhane, K., Thomas, D.C., Cotton, J.W., Coull, B.A., Mezzetti, M., Ryan, L.M., Cowles, K., Cox, C., Cox, L.H., Gupta, S., Hosmane, B.S., Gurrin, L.C., Pettitt, A.N., Saez, M., Sheppard, L. & Lumley, T. 2000, 'Combining evidence on air pollution and daily mortality from the 20 largest US cities: a hierarchical modelling strategy - Discussion', JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES A-STATISTICS IN SOCIETY, vol. 163, pp. 284-302.
Harris, J.E., Ryan, L., Hoover, H.C., Stuart, R.K., Oken, M.M., Benson, A.B., Mansour, E., Haller, D.G., Manola, J. & Hanna, M.G. 2000, 'Adjuvant active specific immunotherapy for stage II and III colon cancer with an autologous tumor cell vaccine: Eastern Cooperative Oncology Group Study E5283.', Journal of clinical oncology : official journal of the American Society of Clinical Oncology, vol. 18, no. 1, pp. 148-157.
PURPOSE: A randomized phase III clinical trial of adjuvant active specific immunotherapy (ASI) with an autologous tumor cell-bacillus Calmette-Gu&eacute;rin (BCG) vaccine was conducted to determine whether surgical resection plus ASI was more beneficial than resection alone in stage II and III colon cancer patients. PATIENTS AND METHODS: Patients (n = 412) with colon cancer (297 with stage II disease, 115 with stage III disease) were randomly allocated to an observation arm or to a treatment arm in which they received three weekly intradermal vaccine injections of 10(7) irradiated autologous tumor cells beginning approximately 4 weeks after surgery. The first two weekly injections also contained 10(7) BCG organisms. Patients were observed for determination of time to recurrence and disease-free and overall survival. RESULTS: This was a negative study in that after a 7.6-year median follow-up period, there were no statistically significant differences in clinical outcomes between the treatment arms. However, there were disease-free survival (P =.078) and overall survival (P =.12) trends in favor of ASI when treatment compliance was evaluated, ie, patients who received the intended treatment had a delayed cutaneous hypersensitivity (DCH) response to the third vaccination (induration >/=5 mm). Also, the magnitude of the DCH response correlated with improved prognosis. The 5-year survival proportion was 84.6% for those with indurations greater than 10 mm, compared with 45.0% for those with indurations less than 5 mm. CONCLUSIONS: When all randomized patients were evaluated, no significant clinical benefit was seen with ASI in surgically resected colon cancer patients with stage II or III colon cancer. However, there was an indication that treatment compliance with effective immunization results in disease-free and overall survival benefits.
Sanchez, J.M., Collia, F., Stoler, J.M., McGuirk, C., Holmes, L.B., Ryan, L. & Lieberman, E. 2000, 'Letter to the editor (multiple letters)', Genetics in Medicine, vol. 2, no. 5, pp. 300-301.
Molenberghs, G. & Ryan, L.M. 1999, 'An exponential family model for clustered multivariate binary data', Environmetrics, vol. 10, no. 3, pp. 279-300.
View/Download from: UTS OPUS or Publisher's site
This paper focuses on the analysis of clustered multivariate binary data that arise from developmental toxicity studies. In these studies, pregnant mice are exposed to chemicals to assess possible adverse effects on developing fetuses. Multivariate binary outcomes arise when each fetus in a litter is assessed for the presence of malformations and/or low birth weight. We analyse the data using a multivariate exponential family model which is flexible in terms of allowing response rates to depend on duster size. Maximum likelihood estimation of model parameters and the construction of score tests for dose effect are discussed. Copyright (C) 1999 John Wiley & Sons, Ltd.
Geys, H., Molenberghs, G. & Ryan, L.M. 1999, 'Pseudolikelihood modeling of multivariate outcomes in developmental toxicology', Journal of the American Statistical Association, vol. 94, no. 447, pp. 734-745.
View/Download from: UTS OPUS or Publisher's site
The primary goal of this article is to determine benchmark doses based on the ethylene glycol study, which comprises data from a developmental toxicity study in mice. Because the data involve a vector of malformation indicators, a flexible model for multivariate clustered data is required. An exponential family model is considered and pseudolikelihood-based inferential tools are proposed, hence avoiding excessive computational requirements.
Betensky, R.A., Lindsey, J.C., Ryan, L.M. & Wand, M. 1999, 'Local EM estimation of the hazard function for interval-censored data', Biometrics, vol. 55, pp. 238-245.
View/Download from: UTS OPUS or Publisher's site
We propose a smooth hazard estimator for interval-censored survival data using the method of local likelihood. The model is fit using a local EM algorithm. The estimator is more descriptive than traditional empirical estimates in regions of concentrated information and takes on a parametric flavor in regions of sparse information. We derive two different standard error estimates for the smooth curve, one based on asymptotic theory and the other on the bootstrap. We illustrate the local EM method for times to breast cosmesis deterioration (Finkelstein, 1986, Biometrics 42, 845-854) and for times to HIV-1 infection for individuals with hemophilia (Kroner et al., 1994, Journal of AIDS 7, 279-286). Our hazard estimates for each of these data sets show interesting structures that would not be found using a standard parametric hazard model or empirical survivorship estimates.
Padungtod, C., Niu, T., Wang, Z., Savitz, D.A., Christiani, D.C., Ryan, L.M. & Xu, X. 1999, 'Paraoxonase polymorphism and its effect on male reproductive outcomes among Chinese pesticide factory workers', American Journal of Industrial Medicine, vol. 36, no. 3, pp. 379-387.
View/Download from: UTS OPUS or Publisher's site
Background Sei um paraoxonase has been associated with the metabolism of organophosphate pesticides in humans. Molecular analysis of the human paraoxonase gene (PON1) has revealed that Arg(192) homozygotes have a greater detoxifying capability than Gln(192) homozygotes. We examined the effects of PON1 genotypes on male reproductive outcomes and its interaction with exposure to organophosphate pesticides. Methods We studied 60 Chinese pesticide-factory workers and 89 textile-factory workers who were unexposed to pesticides. The respective allele frequencies of Arg(192) and Gln(192) were 0.62 and 0.38. Pesticide exposure among 36 exposed subjects and 12 unexposed subjects, regardless of gender was assessed by personal measurement of pesticide residues over an entire 8-hr shift and measurement of urinary p-nitrophenol level over a 24-hr period. We analyzed semen and hormone data collected from male subjects. Results When the three PON1 genotypes were analyzed separately, a gene dose effect was not detected, We used rite unexposed Arg(192) homo/heterozygotes as the reference group, and re-analyzed the data. Exposed Arg(192) homo/heterozygotes had significantly lower sperm count (chi(2) = 9.01, P < 0.01) and lower percentage of sperm with normal morphology (chi(2) = 4.18, P < 0.05) than the reference group. Both unexposed Gln(192) homozygotes (chi(2) = 4.90 P < 0.05) and exposed Arg(192) homo/heterozygotes (chi(2) = 10.00, P < 0.01) showed significantly lower sperm concentrations than the reference group. In addition, exposed Arg(192) homo/heterozygotes had significantly higher serum LH levels (chi(2) = 7.94, P < 0.01) than the reference group. Conclusions Because of a small sample size, our findings are highly preliminary: Nevertheless, it calls for further investigation of the interaction between the PON1 genotype and organophosphate pesticide exposure on male reproductive outcomes.
Witte, R.S., Ryan, L.M., Schutt, A.J., Carbone, P.P. & Engstrom, P.F. 1999, 'PALA versus streptozotocin, doxorubicin, and MeCCNU in the treatment of patients with advanced pancreatic carcinoma', Investigational New Drugs Investigational New Drugs Investigational New Drugs, vol. 16, no. 4, pp. 315-318.
View/Download from: UTS OPUS or Publisher's site
Seventy-three eligible, chemotherapy-naive, ambulatory patients with advanced pancreatic carcinoma were allocated to one of two treatment regimens: 35 received PALA (1250 mg/m2 daily 5 every 4 weeks) and 38 were given SAM (streptozotocin 400 mg/m2 IV daily 5, doxorubicin 45 mg/m2 IV on day 1 and 22, and methyl CCNU 60 mg/m2 orally on days 1 and 22 every 6 weeks). Doses were modified for myelo-, gi-, or cardiotoxicity. Adequate organ, bone marrow and cardiac function; a measurable lesion; adequate caloric intake; and a life expectancy of 2 months were required for treatment on this trial. One patient on each regimen had a partial response for response rates of 3% (95% confidence intervals, 0.08 to 17%). Median survival on the PALA arm was 5 months and median time to treatment failure was 2.6 months. SAM patients experienced median overall and progression free survivals of 3.4 and 1.9 months, respectively. The severe toxicity observed was almost exclusively myelosuppression on both regimens. One patient receiving SAM had lethal leukopenic sepsis during the first cycle as the only treatment-related death. Neither PALA nor SAM offer any therapeutic utility to patients with advanced pancreatic cancer.
Brumback, B.A., Holmes, L.B. & Ryan, L.M. 1999, 'Adverse effects of chorionic villus sampling: A meta-analysis', Statistics In Medicine, vol. 18, no. 16, pp. 2163-2175.
View/Download from: Publisher's site
Meta-analysis is a popular tool for combining evidence from several related studies. The technique is usually used to combine randomized clinical trials, case-control studies or prospective studies where each study has its own exposed and unexposed group
Padungtod, C., Hassold, T., Millie, E., Ryan, L.M., Savitz, D.A., Christiani, D.C. & Xu, X. 1999, 'Sperm aneuploidy among Chinese pesticide factory workers: Scoring by the FISH method', American Journal of Industrial Medicine, vol. 36, no. 2, pp. 230-238.
View/Download from: UTS OPUS
Chedid, A., Ryan, L.M., Dayal, Y., Wolf, B.C. & Falkson, G. 1999, 'Morphology and other prognostic factors of hepatocellular carcinoma', Archives of Pathology & Laboratory Medicine, vol. 123, no. 6, pp. 524-528.
View/Download from: UTS OPUS
Christiani, D.C., Ye, T., Zhang, S., Wegman, D.H., Eisen, E.A., Ryan, L.M., Olenchock, S.A., Pothier, L. & Dai, H. 1999, 'Cotton dust and endotoxin exposure and long-term decline in lung function: results of a longitudinal study', American Journal of Industrial Medicine, vol. 35, no. 4, pp. 321-331.
View/Download from: UTS OPUS
Sammel, M., Lin, X. & Ryan, L.M. 1999, 'Multivariate linear mixed models for multiple outcomes', Statistics in Medicine, vol. 18, pp. 2479-2492.
View/Download from: UTS OPUS or Publisher's site
We propose a multivariate linear mixed (MLMM) for the analysis of multiple outcomes, which generalizes the latent variable model of Sammel and Ryan, The proposed model assumes a flexible correlation structure among the multiple outcomes, and allows a global test of the impact of exposure across outcomes. In contrast to the Sammel-Ryan model, the MLMM separates the mean and correlation parameters so that the mean estimation will remain reasonably robust even if the correlation is misspecified. The model is applied to birth defects data, where continuous data on the size of infants who were exposed to anticonvulsant medications in utero are compared to controls.
Lin, A.E., Herring, A.H., Amstutz, K.S., Westgate, M., Lacro, R.V., Al-Jufan, M., Ryan, L.M. & Holmes, L.B. 1999, 'Cardiovascular malformations: Changes in prevalence and birth status, 1972-1990', American Journal of Medical Genetics, vol. 87, no. 2, pp. 102-110.
View/Download from: UTS OPUS
Through an ongoing hospital-based active malformation surveillance program, we identified cardiovascular malformations (CVMs) in 3.3 per 1,000 liveborn and stillborn infants, and fetuses from pregnancies terminated electively during a 15-year period. We excluded the children of mothers who had planned delivery elsewhere, but were transferred for care of anomalies that had been detected in prenatal screening, Birth status changed markedly during the study with a significant increase in elective terminations of fetuses with a CVM from 0 to 22% (P < 0.01 based on a test for trend). The proportion of liveborn infants with CVMs decreased from 90% to 73% (P < 0.01); the frequency of stillbirths did not change, During the study period, there was a significant increase in the prevalence of CVMs in all births (P < 0.01) and elective terminations (P < 0.01), The increase in liveborn prevalence was not statistically significant (P = 0.08), Stillborn prevalence was unchanged. The number of mothers having prenatal ultrasonography (P < 0.01 for trend) and amniocentesis (P < 0.01 for trend) increased steadily. There were significant increases in the pro-portion of mothers having any ultrasound examination (P < 0.01 for trend), the number of initial ultrasound examinations occurring in the second trimester (P < 0.01 for trend), and the proportion of mothers having amniocentesis (P < 0.01 for trend), There was a significant increasing trend in the proportion of mothers who were 35 years and older (10% in 1972-1974, 26% in 1988-1990, P < 0.01), This hospital-based active surveillance program suggests that more frequent elective terminations had a significant effect on overall birth prevalence of CVMs, This trend would not have been detected by most other surveillance systems which determine prevalence of common birth defects from birth certificates and other forms of administrative reporting, and exclude elective terminations of pregnancy
Ryan, L.M. & Molenberghs, G. 1999, 'Statistical methods for developmental toxicity - Analysis of clustered multivariate binary data', Annals of the New York Academy of Sciences, vol. 895, pp. 196-211.
View/Download from: UTS OPUS or Publisher's site
This paper discusses some of the statistical issues that arise from developmental toxicity studies, wherein pregnant mice are exposed to chemicals in order to assess possible adverse effects on developing Fetuses. We begin with a review of some current approaches to risk assessment, based on NOAELs, and provide justification for the use of methods based on dose-response models, Due to the hierarchical nature of the data, such models are more complicated in the present context than, say, in cancer studies. For example, multivariate binary outcomes arise when each fetus in a litter is assessed for the presence of malformations and/or tow birth weight. We describe a multivariate exponential family model that works well for these data and that is flexible in terms of allowing response rates to depend on cluster size. Maximum Likelihood estimation of model parameters and the construction of score tests for dose effect are briefly discussed. Results are illustrated with data from several NTP studies.
Weller, E., Long, N., Smith, A., Williams, P., Ravi, S., Gill, J., Henessey, R., Skornik, W., Brain, J., Kimmel, C., Kimmel, G., Holmes, L. & Ryan, L.M. 1999, 'Dose-rate effects of ethylene oxide exposure on developmental toxicity', Toxicological Sciences, vol. 50, pp. 259-270.
View/Download from: UTS OPUS or Publisher's site
In risk assessment, evaluating a health effect at a duration of exposure that is untested involves assuming that equivalent multiples of concentration (C) and duration (T) of exposure have: the same effect. The limitations of this approach (attributed to F. Haber, Zur Geschichte des Gaskrieges [On the history of gas warfare], in Funf Vortrage aus den Jahren 1920-1923 [Five lectures from the years 1920-1923], 1924, Springer, Berlin, pp. 76-92), have been noted in several studies. The study presented in this paper was designed to specifically look at dose-rate (C x T) effects, and it forms an ideal case study to implement statistical models and to examine the statistical issues in risk assessment. Pregnant female C57BL/6J mice were exposed, on gestational day 7, to ethylene oxide (EtO) via inhalation for 1.5, 3;, or 6 h at exposures that result in C x T multiples of 2100 or 2700 ppm-hi EtO was selected because of its short half-life, documented developmental toxicity, and relevance to exposures that occur in occupational settings. Concurrent experiments were run with animals exposed to air for similar periods. Statistical analysis using models developed to assess dose-rate effects revealed significant effects with respect to fetal death and resorptions, malformations, crown-to-rump length, and fetal weight. Animals exposed to short, high exposures of EtO on day 7 of gestation were found to have more adverse effects than animals exposed to the same C x T multiple but at longer, lower exposures. The implication for risk assessment is that applying Haber's Law could potentially lead to an underestimation of risk at a shorter duration of exposure and an overestimation of risk at a longer duration of exposure. Further research, toxicological and statistical, are required to understand the mechanism of the dose-rate effects, and how to incorporate the mechanistic information into the risk assessment decision process.
Stoler, J.M., McGuirk, C.K., Lieberman, E., Ryan, L. & Holmes, L.B. 1999, 'Malformations reported in chorionic villus sampling exposed children: A review and analytic synthesis of the literature', GENETICS IN MEDICINE, vol. 1, no. 7, pp. 315-322.
Lindsey, J.C. & Ryan, L. 1999, 'Tutorial in biostatistics: methods for interval-censored data (vol 17, pg 219, 1998)', STATISTICS IN MEDICINE, vol. 18, no. 7, pp. 890-890.
View/Download from: Publisher's site
Ku, C.H., Burge, H.A., Ryan, L.M., Pai, L., Liou, S.H. & Christiani, D.C. 1999, 'Rice field fungal exposures in young male workers: Short-term exposure assessment for fungi', AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE, vol. 159, no. 3, pp. A298-A298.
Cho, S.I., Li, Q., Yang, J., Chen, C., Padungtod, C., Ryan, L., Christiani, D.C. & XU, X. 1999, 'Drinking water source and spontaneous abortion: A cross-sectional study in a rural Chinese population.', International journal of occupational and environmental health, vol. 5, no. 3, pp. 164-169.
The authors examined the association between the risk of spontaneous abortion and the type of drinking water source in a rural Chinese population. Information about pregnancy outcomes and various exposures was collected by means of a detailed interview questionnaire. 2,876 pregnancies occurring in 1989-1993 among 2,201 nonsmoking and non-alcohol-drinking women were analyzed by logistic regression. Generalized estimating equations were used to adjust for correlations between multiple pregnancies in the same woman. The use of pond water as a drinking water source was associated with a higher risk of spontaneous abortion than the use of well or river water (odds ratio, 1.63; 95% confidence interval, 1.11 to 2.39), adjusting for potential confounders. More studies are needed to identify the agents responsible for the observed association. The results of this study emphasize the importance of monitoring potential drinking-water contamination in communities.
Novik, Y., Ryan, L.M., Haller, D.G., Asbury, R., Dutcher, J.P. & Schutt, A. 1999, 'Phase II protocol for the evaluation of new treatments in patients with advanced gastric carcinoma: results of ECOG 5282.', Medical oncology (Northwood, London, England), vol. 16, no. 4, pp. 261-266.
View/Download from: Publisher's site
The study was a Phase II randomized study to evaluate the efficacy of new agents for the treatment of advanced gastric carcinoma. Patients were randomized to receive single agent chemotherapy with mitoxantrone, etoposide, aclacinomycin-A or spirogermanium. The patients were stratified by prior use of chemotherapy, prior doxorubicin use and ECOG performance status. Patients with a history of cardiac disease or prior doxorubicin exceeding a dose of 400 mg/m2 were restrictively randomized to sopirogermanium or etoposide only. One hundred and fourteen patients were registered for the study. Among 98 evaluable patients there were only two partial responses (both in the etoposide arm), and one complete response in the mitoxantrone arm. The median survival on the study was 3.3 months. One hundred and six patients were analyzable for toxicity. There were four treatment-related deaths and four life-threatening toxicities. Because of low response rates and relatively high toxicities the studied compounds were not deemed worth further investigation for advanced gastric cancer.
Keller, S.M., Ryan, L.M., Coia, L.R., Dang, P., Vaught, D.J., Diggs, C., Weiner, L.M. & Benson, A.B. 1998, 'High dose chemoradiotherapy followed by esophagectomy for adenocarcinoma of the esophagus and gastroesophageal junction : Results of a Phase II study of the Eastern Cooperative Oncology Group', Cancer, vol. 83, pp. 1908-1916.
View/Download from: Publisher's site
Padungtod, C., Lasley, B., Christiani, D., Ryan, L.M. & Xu, X. 1998, 'Reproductive Hormone Profile Among Pesticide Factory Workers', Journal Of Occupational And Environmental Medicine, vol. 40, no. 12, pp. 1038-1047.
Serum follicle-stimulating hormone (FSH), luteinizing hormone (LH), and testosterone levels, as well as urinary levels of FSH, LH, and E1C, a metabolite of testosterone, were measured to investigate the adverse reproductive effects of organophosphate pes
Ibrahim, J.G., Ryan, L.M. & Chen, M. 1998, 'Using historical controls to adjust for covariates in trend tests for binary data', Journal of the American Statistical Association, vol. 93, no. 444, pp. 1282-1293.
View/Download from: UTS OPUS or Publisher's site
Historical data often play an important role in helping interpret the results of a current study. This article is motivated primarily by one specific application: the analysis of data from rodent carcinogenicity studies. By proposing a suitable informative prior distribution on the relationship between control outcome data and covariates, we derive modified trend test statistics that incorporate historical control information to adjust for covariate effects. Frequentist and fully Bayesian methods are presented, and novel computational techniques are developed to compute the test statistics. Several attractive theoretical and computational properties of the proposed priors are derived. In addition, a semiautomatic elicitation scheme for the priors is developed. Our approach is used to modify a widely used prevalence test for carcinogenicity studies. The proposed methodology is applied to data from a National Toxicology Program carcinogenicity experiment and is shown to provide helpful insight on the results of the analysis.
Smith, T., Ryan, L.M., Douglass, H., Haller, D., Dayal, Y., Kirkwood, J., Tormey, D., Schutt, A., Hinson, J. & Sischy, B. 1998, 'Combined Chemoradiotherapy Vs. Radiotherapy Alone For Early Stage Squamous Cell Carcinoma Of The Esophagus: A Study Of The Eastern Cooperative Oncology Group', International Journal Of Radiation Oncology Biology Physics, vol. 42, no. 2, pp. 269-276.
View/Download from: Publisher's site
Squamous carcinoma of the thoracic esophagus has an extremely poor prognosis. This study, EST-1282, was undertaken by the Eastern Cooperative Oncology Group (ECOG) to determine whether the combined use of 5-fluorouracil (5-FU), mitomycin C, and radiation
Bosch, R.J. & Ryan, L.M. 1998, 'Generalized Poisson models arising from Markov processes', Statistics & Probability Letters, vol. 39, no. 3, pp. 205-212.
View/Download from: UTS OPUS or Publisher's site
We develop a family of distributions which allow for over- and underdispersion relative to the Poisson. This latter feature is particularly appealing since many existing methods only allow for overdispersion. These distributions arise from underlying continuous-time Markov processes in which event rates depend on how many events have already occurred. The results are illustrated with underdispersed count data from a polyspermy study and overdispersed data from the Canadian Sickness Survey. (C) 1998 Elsevier Science B.V. All rights reserved.
Weller, E.A. & Ryan, L.M. 1998, 'Testing for trend with count data', Biometrics, vol. 54, no. 2, pp. 762-773.
View/Download from: UTS OPUS or Publisher's site
Among the tests that can be used to detect dose-related trends in count data from toxicological studies are nonparametric tests such as the Jonckheere-Terpstra and likelihood-based tests, for example, based on a Poisson model. This paper was motivated by a data set of tumor counts in which conflicting conclusions were obtained using these two tests. To define situations where one test may be preferable, we compared the small and large sample performance of these two tests as well as a robust and conditional version of the likelihood-based test in the absence and presence of a dose-related trend for both Poisson and overdispersed Poisson data. Based on our results, we suggest using the Poisson test when little overdispersion is present in the data. For more overdispersed data, we recommend using the robust Poisson test for highly discrete data (response rate lower than 2-3) and the robust Poisson test or the Jonckheere-Terpstra test for moderately discrete or continuous data (average responses larger than 2 or 3). We also studied the effects of dose metameter misspecification. A clear effect on efficiency was seen when the 'wrong' dose metameter was used to compute the test statistic. In general, unless there is strong reason to do otherwise, we recommend the use of equally spaced dose levels when applying the Poisson or robust Poisson test for trend.
Lindsay, J.C. & Ryan, L.M. 1998, 'Tutorial in biostatistics - Methods for interval-censored data', Statistics in Medicine, vol. 17, pp. 219-238.
View/Download from: Publisher's site
In standard time-to-event or survival analysis, occurrence times of the event of interest are observed exactly or are right-censored, meaning that it is only known that the event occurred after the last observation time, There are numerous methods available for estimating the survival curve and for testing and estimation of the effects of covariates in this context. In some situations, however, the times of the events of interest may only be known to have occurred within an interval of time. In clinical trials, for example, patients are often seen at pre-scheduled visits but the event of interest may occur in between visits. These data are interval-censored. Owing to the lack of well-known statistical methodology and available software, a common ad hoc approach is to assume that the event occurred at the end (or beginning or midpoint) of each interval, and then apply methods for standard time-to-event data. However, this approach can lead to invalid inferences, and in particular will tend to underestimate the standard errors of the estimated parameters. The purpose of this tutorial is to illustrate and compare available methods which correctly treat the data as being interval-censored. It is not meant to be a full review of all existing methods, but only those which are available in standard statistical software, or which can be easily programmed. All approaches will be illustrated on two data sets and compared with methods which ignore the interval-censored nature of the data. We hope this tutorial will allow those familiar with the application of standard survival analysis techniques the option of applying appropriate methods when presented with interval-censored data. (C) 1998 John Wiley & Sons, Ltd.
Stoler, J.M., Huntington, K.S., Peterson, C.M., Peterson, K.P., Daniel, P., Aboagye, K.K., Lieberman, E., Ryan, L.M. & Holmes, L.B. 1998, 'The prenatal detection of significant alcohol exposure with maternal blood markers', Journal of Pediatrics, vol. 133, no. 3, pp. 346-352.
View/Download from: UTS OPUS or Publisher's site
Objective: To examine the efficacy of a combination of 4 blood markers of alcohol use in detecting alcohol-abusing pregnant women. Study design: Two new markers of alcohol use, whole blood-associated acetaldehyde and carbohydrate-deficient transferrin, and 2 traditional markers of alcohol use, gamma-glutamyl transpeptidase and mean red blood cell volume, were measured in the blood of pregnant women. Each woman was interviewed about alcohol and drug use, medical and obstetric histories, and nutrition. Each infant was examined by a clinician who was blinded to exposure status. Results: All of the women who reported drinking an average of 1 or more ounces of absolute alcohol per day had at least, 1 positive blood marker. The infants of mothers with 2 or more positive markers had significantly smaller birth weights, lengths, and head circumferences than the infants with negative maternal screens. The presence of 2 or more positive markers was more predictive of infant outcome than any self-reporting measure. Conclusions: These markers, which detect more at-risk pregnant women than self-reporting methods, could lead to better efforts at detection and prevention of alcohol-induced fetal damage.
Gaylor, D., Ryan, L.M., Krewski, D. & Zhu, Y. 1998, 'Procedures for calculating benchmark doses for health risk assessment', Regulatory Toxicology and Pharmacology, vol. 28, no. 2, pp. 150-164.
View/Download from: UTS OPUS or Publisher's site
Safety assessment for noncancer health effects generally has been based upon dividing a no observed adverse effect (NOAEL) by uncertainty (safety) factors to provide an acceptable daily intake (ADI) or reference dose (RfD). Since the NOAEL does not utilize all of the available dose-response data, allows higher ADI from poorer experiments, and may have an unknown, unacceptable level of risk, the benchmark dose (BD) with a specified, controlled low level of risk has become popular as an adjunct to the NOAEL or the low observed adverse effect level (LOAEL) in the safety assessment process. The purpose of this paper is to summarize statistical procedures available for calculating BDs and their confidence limits for noncancer endpoints. Procedures are presented and illustrated for quantal (binary), quasicontinuous (proportion), and continuous data. Quasicontinuous data arise in developmental studies where the measure of an effect for a fetus is quantal (normal or abnormal) but the experimental unit is the mother (litter) so that results can be expressed as the proportion of abnormal fetuses per litter. However, the correlation of effects among fetuses within a litter poses some additional statistical problems. Also, developmental studies usually include some continuous measures, such as fetal body weight or length. With continuous data there generally is not a clear demarcation between normal and adverse measurements. In such cases, extremely high and/or low measurements at some designated percentile(s) can be considered abnormal. Then the probability (risk) of abnormal individuals can be estimated as a function of dose. The procedure for estimating a ED with continuous data is illustrated using neurotoxicity data. When multiple measures of adverse effects are available, a BD can be estimated based on a selected endpoint or the appearance of any combination of endpoints. Multivariate procedures are illustrated using developmental and reproductive toxicity data.
Xu, X.P., Cho, S.I., Sammel, M., You, L.Y., Cui, S.C., Huang, Y.M., Ma, G.H., Padungtod, C., Pothier, L., Niu, T.H., Christiani, D., Smith, T., Ryan, L. & Wang, L.H. 1998, 'Association of petrochemical exposure with spontaneous abortion', OCCUPATIONAL AND ENVIRONMENTAL MEDICINE, vol. 55, no. 1, pp. 31-36.
Weller, E.A., Ryan, L.M., Spiegelman, D. & Smith, T. 1997, 'Statistical issues in assessing human population exposures', Chemometrics and Intelligent Laboratory Systems, vol. 37, pp. 189-195.
View/Download from: UTS OPUS or Publisher's site
Relating an exposure level to a particular outcome in occupational and environmental epidemiology studies may be challenging. In these studies, researchers are interested in the health effects from exposure. Two factors that complicate this process are the measurement error that is usually present in exposure assessment and the complex nature of the exposure-dose relationship. Statistical aspects of the measurement error problem has been the focus of many researchers in the past few years. However, less attention has been given to the second issue of incorporating the relationship between exposure and dose and directly modeling the dose-response relationship, when this is of primary scientific interest. For complex exposure-dose relationships, a direct application of statistical methods to relate exposure to the health outcome may lead to bias and loss of efficiency in estimates of the dose-response relationship. In this paper, we present some of the statistical issues that arise using a study of the health effects of arc welding fumes as an illustration.
Wang, X., Ding, H., Ryan, L.M. & Xu, X. 1997, 'Association between air pollution and low birth weight: a community-based study', Environmental Health Perspectives, vol. 105, no. 5, pp. 514-520.
View/Download from: Publisher's site
The relationship between maternal exposure to air pollution during periods of pregnancy (entire and specific periods) and birth weight was investigated in a well-defined cohort. Between 1988 and 1991, all pregnant women living in four residential areas of Beijing were registered and followed from early pregnancy until delivery. Information on individual mothers and infants was collected. Daily air pollution data were obtained independently. The sample for analysis included 74,671 first-parity live births were gestational age 37-44 weeks. Multiple linear regression and logistic regression were used to estimate the effects of air pollution on birth weight and low birth weight (< 2,500 g), adjusting for gestational age, residence, year of birth, maternal age, and infant gender. There was a significant exposure-response relationship between maternal exposures to sulfur dioxide (SO2) and total suspended particles (TSP) during the third trimester of pregnancy and infant birth weight. The adjusted odds ratio for low birth weight was 1.11 (95% CI, 1.06-1.16) for each 100 micrograms/m3 increase in SO2 and 1.10 (95% CI, 1.05-1.14) for each 100 micrograms/m3 increase in TSP. The estimated reduction in birth weight was 7.3 g and 6.9 g for each 100 micrograms/m3 increase in SO2 and in TSP, respectively. The birth weight distribution of the high-exposure group was more skewed toward the left tail (i.e., with higher proportion of births < 2,500 g) than that of the low-exposure group. Although the effects of other unmeasured risk factors cannot be excluded with certainty, our data suggests that TSP and SO2, or a more complex pollution mixture associated with these pollutants, contribute to an excess risk of low birth weight in the Beijing population.
Legler, J.M. & Ryan, L.M. 1997, 'Latent variable models for teratogenesis using multiple binary outcomes', Journal of the American Statistical Association, vol. 92, no. 437, pp. 13-20.
View/Download from: UTS OPUS or Publisher's site
Multiple outcomes are commonly measured in:the study of birth defects. The reason is that most teratogens do not cause a single, uniquely defined defect, but rather result in a range of effects, including major malformations, minor anomalies, and deficiencies in birth weight, length and head circumference. The spectrum of effects associated with a particular teratogen is sometimes described as a ''syndrome.'' In this article we develop a latent variable model to characterize exposure effects on multiple binary outcomes. Not only does the method allow comparisons of control and exposed infants with respect to multiple outcomes, but it also provides a measure of the ''severity'' of each child's condition. Data from a study of the teratogenic effects of anticonvulsants illustrate our results.
Baird, S., Catalano, P., Ryan, L.M. & Evans, J. 1997, 'Evaluation Of Effect Profiles: Functional Observational Battery Outcomes', Fundamental And Applied Toxicology, vol. 40, no. 1, pp. 37-51.
View/Download from: Publisher's site
The Functional Observational Battery (FOB) is a neurotoxicity screening assay composed of 25-30 descriptive, scalar, binary, and continuous endpoints. These outcomes have been grouped into six biologically logical domains as a means to interpret the neur
Geys, H., Molenberghs, G. & Ryan, L.M. 1997, 'Pseudo-likelihood inference for clustered binary data', Communications In Statistics-Theory And Methods, vol. 26, no. 11, pp. 2743-2767.
View/Download from: UTS OPUS or Publisher's site
Molenberghs and Ryan (1996) proposed a likelihood-based model for clustered. binary data, based on a multivariate exponential family model (Cox, 1972). The model benefits from the elegance and simplicity of exponential family theory and is flexible in terms of allowing response rates to depend on cluster size. A main problem however, particularly with large clusters is the evaluation of the normalizing constant. In this paper, pseudo-likelihood is explored as an alternative mode of inference. The pseudo-likelihood equations are derived, the model is applied to data from a developmental toxicity study, and an asymptotic and small sample relative efficiency study is performed.
Sammel, M.D., Ryan, L.M. & Legler, J.M. 1997, 'Latent variable models for mixed discrete and continuous outcomes', Journal of the Royal Statistical Society Series B: Statistical Methodology, vol. 59, no. 3, pp. 667-678.
View/Download from: UTS OPUS
We propose a latent variable model for mixed discrete and continuous outcomes. The model accommodates any mixture of outcomes from an exponential family and allows for arbitrary covariate effects, as well as direct modelling of covariates on the latent variable. An EM algorithm is proposed for parameter estimation and estimates of the latent variables are produced as a by-product of the analysis. A generalized likelihood ratio test can be used to test the significance of covariates affecting the latent outcomes. This method is applied to birth defects data, where the outcomes of interest are continuous measures of size and binary indicators of minor physical anomalies. Infants who were exposed in utero to anticonvulsant medications are compared with controls.
Augenlicht, L.H., Wadler, S., Corner, G., Richards, C., Ryan, L.M., Multani, A.S., Pathak, S., Benson, A., Haller, D. & Heerdt, B.G. 1997, 'Low-level c-myc amplification in human colonic carcinoma cell lines and tumors: a frequent, p53-independent mutation associated with improved outcome in a randomized multi-institutional trial', Cancer Research, vol. 57, no. 9, pp. 1769-1775.
View/Download from: UTS OPUS
Human colonic cancer is associated with multiple genetic deletions, mutations, and alterations in gene expression; in contrast, gene amplification has not been recognized as a prominent characteristic of human colonic tumors. Although the c-myc gene is overexpressed in approximately 70% of human colonic cancers, previous studies have not detected frequent gene amplification or rearrangement of c-myc in these tumors, although such amplification has been reported in chemically induced rodent colon cancer and quantitative analysis of gene copy number has shown the gene to be amplified at a low level in mucinous and poorly differentiated human colon carcinomas. Using rigorously controlled blot methodology, we have established that the c-myc gene, Located at 8q21, exhibited amplification of 87% to 35-fold in 7 of 10 human colonic carcinoma cell lines. This was highly significant even at a low level of amplification in HT29 cells (P < 0.0001). Cytogenetic analysis by G-banding did not detect aneuploidy involving chromsome 8q, suggesting that the amplification for the c-myc gene on 8q was relatively specific, and this was consistent with a lack of amplification detected for the c-mos gene on 8q24, which was assayed similarly. The same methodology then revealed amplification of c-myc from 1.5-fold to 5-fold in 32% of tumors from 149 patients entered into a multi-institutional Phase III study of adjuvant therapy for colon cancer. c-myc status was not related to time to recurrence or death, but low levels of c-myc amplification identified a subset of patients who showed a statistically significant increase in disease-free survival, and a corresponding trend to longer overall survival, in response to adjuvant therapy with 5-fluorouracil plus levamisole. Presence of c-myc amplification was not related to incidence of p53 mutations.
Clark, D.E. & Ryan, L.M. 1997, 'Modeling injury outcomes using time-to-event methods', Journal of Trauma: Injury, Infection, and Critical Care, vol. 42, pp. 1129-1134.
View/Download from: UTS OPUS
Background: Mortality is an important measurement of injury outcomes, but measurements reflecting disability or cost also important, Hospital length of stay (LOS) has been used as an outcome variable, but reduced LOS could be achieved either by improved care or by increased mortality, A solution to this statistical problem of ''competing risks'' would enable injury outcomes based on LOS to be modeled using time-to-event methods, Methods: Time-to-event methodology was applied to 2,106 cases with complete data from the 1991-1994 registry of a regional trauma center. LOS was used as the outcome variable, modified by assigning an arbitrarily long LOS to any fatal case, A combination of proportional hazards and logistic regression models was used to explore the effects of potential predictive variables, including Trauma Score (TS), Injury Severity Score (ISS), components of TS or ISS, age, sex, alcohol use, and whether a patient was transferred. Results: The ''TRISS'' combination of TS, ISS, and age previously shown to predict mortality also predicted ''modified LOS'' (Wald p value less than 0.001 for each variable). Models using only age and certain components of ISS or TS fit our data even better, with fewer parameters, Other variables were not predictive, Modified Kapian-Meier plots provided easily interpreted graphical results, combining both mortality and LOS information. Conclusions: With a simple modification to allow for competing risks, time-to-event methods enable more informative modeling of injury outcomes than binary (lived/died) methods alone, Such models may be useful for describing and comparing groups of hospitalized trauma patients.
Ryan, L.M. 1997, 'Adjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: meta-analysis of individual data', The Lancet, vol. 350, no. 9092, pp. 1647-1654.
View/Download from: UTS OPUS
Background Individually, randomised trials have not shown conclusively whether adjuvant chemotherapy benefits adult patients with localised resectable soft-tissue sarcoma. Methods A quantitative meta-analysis of updated data from individual patients from all available randomised trials was carried out to assess whether adjuvant chemotherapy improves overall survival, recurrence-free survival, and local and distant recurrence-free intervals (RFI) and whether chemotherapy is differentially effective in patients defined by age, sex, disease status at randomisation, disease site, histology, grade, tumour size, extent of resection, and use of radiotherapy. Findings 1568 patients from 14 trials of doxorubicin-based adjuvant chemotherapy were included (median follow-up 9.4 years). Hazard ratios of 0.73 (95% CI 0.56-0.94, p=0.016) for local RFI, 0.70 (0.57-0.85, p=0.0003) for distant RFI, and 0.75 (0.64-0.87, p=0.0001) for overall recurrence-free survival, correspond to absolute benefits from adjuvant chemotherapy of 6% (95% CI 1-10), 10% (5-15), and 10% (5-15), respectively, at 10 years. For overall survival, the hazard ratio of 0.89 (0.76-1.03) was not significant (p=0.12), but represents an absolute benefit of 4% (1-9) at 10 years. These results were not affected by prespecified changes in the groups of patients analysed. There was no consistent evidence that the relative effect of adjuvant chemotherapy differed for any subgroup of patients for any endpoint. However, the best evidence of an effect of adjuvant chemotherapy for survival was seen in patients with sarcomas of the extremities. Interpretation The meta-analysis provides evidence that adjuvant doxorubicin-based chemotherapy significantly improves the time to local and distant recurrence and overall recurrence-free survival. There is a trend towards improved overall survival.
Stark, P.C., Ryan, L.M., McDonald, J.L. & Burge, H.A. 1997, 'Using meteorologic data to predict daily ragweed pollen levels', Aerobiologia, vol. 13, no. 3, pp. 177-184.
View/Download from: UTS OPUS or Publisher's site
Pollen-related allergy is a common disease resulting in symptoms of hay fever and asthma. Control of symptoms depends (generally) on avoidance and pharmacological treatment. Both of these approaches could benefit from accurate predictions of pollen levels for future days. We have constructed a model that uses meteorological data to predict ragweed pollen levels based on air samples collected daily in Kalamazoo, MI from 1991 to 1994. Ragweed pollen counts were converted to pollen grains/m3 of air (24-h average). We used Poisson regression, which appropriately handles the heterogenous variance associated with pollen data. Using standard statistical model selection procedures, combined with biological considerations, we selected rainfall, wind speed, temperature, and the time measured from the start of the season as the most significant variables. Using our model, we propose a method that uses the weather forecast for the following day to predict the ragweed pollen level. This approach differs from most previous attempts because it uses Poisson regression and because this model needs to be fit iteratively each day. By updating the coefficients of the model based on the information to date, this method allows the fundamental shape of the pollen distribution curve to change from year to year. Application to the Kalamazoo data suggests that the method has good sensitivity and specificity for predicting high pollen days.
Baird, S.J.S., Catalano, P.J., Ryan, L.M. & Evans, J.S. 1997, 'Evaluation of effect profiles: Functional observational battery outcomes', Toxicological Sciences, vol. 40, no. 1, pp. 37-51.
Ibrahim, J.G. & Ryan, L.M. 1996, 'Use of historical controls in time-adjusted trend tests for carcinogenicity', Biometrics, vol. 52, no. 4, pp. 1478-1485.
View/Download from: UTS OPUS or Publisher's site
We develop a method for incorporating historical control information into time-adjusted tests for dose effects in carcinogenicity studies. After discretizing the time scale, we use a multinomial distribution to model the number of animals dying with tumor in each interval. Data from past studies are used to estimate the parameters characterizing the prior. A score test derived from the resulting Dirichlet-multinomial generalizes the test of Tarone (1982, Biometrics 38, 215-220) and reduces, in the limit, to the log-rank test in the case of a diffuse prior. The methodology is illustrated with data from a study of the fire retardant 2,2-Bis(bromomethyl)-1,3-propanediol.
Sammel, M.D. & Ryan, L.M. 1996, 'Latent variable models with fixed effects', Biometrics, vol. 52, no. 2, pp. 650-663.
View/Download from: UTS OPUS or Publisher's site
We discuss latent variable models that allow for fixed effect covariates, as well as covariates affecting the latent variable directly. Restricted maximum likelihood and maximum likelihood are used to estimate model parameters. A generalized likelihood ratio test can be used to test significance of the covariates effecting the latent outcomes. Special cases of the proposed model correspond to factor analysis, mixed models, random effects models, and simultaneous equations. The model is applied to birth defects data, where continuous data on the size of infants who were exposed to anticonvulsant medications in utero are compared to controls.
Hansen, R.M., Ryan, L.M., Anderson, T., Krzywda, B., Quebbeman, E., Benson, I.I.I.A., Haller, D.G. & Tormey, D.C. 1996, 'Phase III study of bolus versus infusion fluorouracil with or without cisplatin in advanced colorectal cancer', Journal of the National Cancer Institute, vol. 88, no. 10, pp. 668-674.
View/Download from: UTS OPUS
Williams, P. & Ryan, L.M. 1996, 'Design of multiple binary outcome studies with intentionally missing data', Biometrics, vol. 52, no. 4, pp. 1498-1514.
View/Download from: UTS OPUS or Publisher's site
We discuss the design and analysis of studies involving multiple binary outcomes in which only a subset of these outcomes can be measured on each individual. Such studies with "intentionally missing data" may arise due to practical or economic constraints; several examples from toxicology serve as illustrations. A global test statistic based on generalized estimating equations is presented and evaluated under a variety of missing patterns and correlation structures. Extensions of the global test statistic to allow for clustered data are also described. The relative efficiency of the global test statistic with missing data relative to that for complete data is investigated, both under a common dose effect alternative and when exposure has differential effects on the multiple endpoints. The implications of these efficiency calculations on study design are explored, and several recommendations are provided.
Andersen, J., Goetghebeur, E. & Ryan, L.M. 1996, 'Missing cause of death information in the analysis of survival data', Statistics in Medicine, vol. 15, pp. 2191-2201.
View/Download from: UTS OPUS or Publisher's site
Goetghebeur and Ryan proposed a method for proportional hazards analyses of competing risks failure-time data when the failure type is missing for some cases. This paper evaluates the properties of the method using data from a clinical trial in Hodgkin's disease. We generated several patterns of missingness in the cause of death in 'pseudo-studies' derived from the study database. We found that the proposed method provided regression coefficients and inferences that were less biased than those from other methods over an increasing percentage of missingness in the failure type when missingness is random, when it depends on an important covariate, when it depends on failure type, and when it depends on follow-up time. We present suggestions for study design with planned missingness in the failure type.
Bosch, R.J., Wypij, D. & Ryan, L.M. 1996, 'A semiparametric approach to risk assessment for quantitative outcomes', Risk Analysis, vol. 16, no. 5, pp. 657-665.
View/Download from: UTS OPUS or Publisher's site
Characterizing the dose-effect relationship and estimating acceptable exposure levels are the primary goals of quantitative risk assessment. A semiparametric approach is proposed for risk assessment with continuously measured or quantitative outcomes which has advantages over existing methods by requiring fewer assumptions. The approach is based on pairwise ranking between the response values in the control group and those in the exposed groups. The work generalizes the rank-based Wilcoxon-Mann-Whitney test, which for the two-group comparison is effectively a test of whether a response from the control group is different from (larger than) a response in an exposed group. We develop a regression framework that naturally extends this metric to model the dose effect in terms of a risk function. Parameters of the regression model can be estimated with standard software. However, inference requires an additional step to estimate the variance structure of the estimated parameters. An effective dose (ED) and associated lower confidence limit (LED) are easily calculated. The method is supported by a simulation study and is illustrated with a study on the effects of aconiazide. The method offers flexible modeling of the dose effect, and since it is rank-based, it is more resistant to outliers, nonconstant variance, and other departures from normality than previously described approaches.
Catalano, P.J., Ryan, L.M. & Kaden, D.A. 1996, 'Statistical design aspects of the NTP/HEI collaborative study on the health effects of chronic ozone inhalation', Inhalation Toxicology, vol. 8, no. 3, pp. 229-249.
View/Download from: UTS OPUS or Publisher's site
Kaden, D.A., Warren, J., Ryan, L.M., Boorman, G. & Mellick, P. 1996, 'The NTP/HEI collaborative ozone project on the health effects of chronic ozone inhalation', Inhalation Toxicology, vol. 8, no. 3, pp. 213-227.
View/Download from: UTS OPUS or Publisher's site
Although many people are exposed to ozone, the effects of chronic exposure to this ubiquitous pollutant, especially low-level chronic exposure, are not well understood. The U.S. Environmental Protection Agency (EPA) current national ambient air quality standard for ozone is exceeded in many communities, especially during the summer. The standard is attained when the number of days per calendar year with maximum hourly average concentrations above 0.12 ppm is equal to or less than 1. The U.S. EPA estimates that 67 million people in the United States, or slightly more than a quarter of the residents, live in areas that were out of compliance with the current National Ambient Air Quality Standard (NAAQS) for ozone in 1989. Although there have been some studies of long-term exposure to ozone, many important questions remain about the health effects of chronic ozone exposure. The Health Effects Institute (HEI), in conjunction with the National Toxicology Program (NTP) carcinogenesis studies, has completed a major effort to help answer these questions. NTP included additional animals in its study for HEI investigators. Included in this effort is a set of studies examining histopathological, biochemical, morphological, and functional alterations in rats exposed to 0, 0.12, 0.5 or 1.0 ppm ozone for 20 mo. This article describes several aspects of this effort. This project can serve as a model for other large toxicological studies for which cancer may not be the only endpoint of concern. The additional animals required for the NTP/HEI Collaborative Ozone Project only represented a modest incremental cost, yet provided information on a much broader range of potential effects of ozone than the basic NTP carcinogenesis studies.
Legler, J.M., Lefkopoulou, M. & Ryan, L.M. 1995, 'Efficiency and power of tests for multiple binary outcomes', Journal of the American Statistical Association, vol. 90, no. 430, pp. 680-693.
View/Download from: UTS OPUS
Global tests provide a useful tool for comparing two or more groups with respect to multiple correlated outcomes. We adapt and compare the performance of tests that have been suggested for use with multiple continuous outcomes to the case of multiple binary outcomes. Comparisons and guidelines are based on asymptotic relative efficiencies (ARE's) and simulations. These results are illustrated using an application from teratology. We extend the work of Lefkopoulou and Ryan to include general M-group comparisons alternatives where group effects may differ for each outcome. A concise form for this general class of score tests is derived. To compute the ARE's for this class of tests, we devise a useful characterization of the alternative space based on multivariate polar coordinates. Our findings indicate that the common outcome effect tests are efficient for a remarkably large range of circumstances. A simple formula applies to compute the maximum number of unaffected outcomes that can be included in a set of outcomes for which the common outcome effect tests remain more efficient than those derived under multidimensional alternatives. For comparison, other global tests are also considered in the simulations: two tests based on resampling( maximal and minimal z tests), a rank-sum test, a generalized least squares test, and a test based on collapsing multiple endpoints to a single binary outcome. Besides the common outcome effect tests, the resampling tests and the rank-sum test are found to perform very well for the cases under consideration.
Goetghebeur, E. & Ryan, L.M. 1995, 'Analysis of competing risks survival data when some failure types are missing', Biometrika, vol. 82, no. 4, pp. 821-833.
View/Download from: UTS OPUS or Publisher's site
We propose a method to analyse competing risks survival data when failure types are missing for some individuals. Our approach is based on a standard proportional hazards structure for each of the failure types, and involves the solution to estimating equations. We present consistent and asymptotically normal estimators of the regression coefficients and related score tests. An appealing feature is that individuals with known failure types make the same contributions as they would to a standard proportional hazards analysis. Contributions of individuals with unknown failure types are weighted according to the probability that they failed from the cause of interest. Efficiency and robustness are discussed. Results are illustrated with data from a breast cancer trial.
Ryan, L.M. 1995, 'Inference based on estimating functions in the presence of nuisance parameters: Comment', Statistical Science, vol. 10, no. 2, pp. 158-173.
View/Download from: Publisher's site
Whyte, M., Landt, M., Mulivor, R., Henthorn, P., Fedde, K., Mahuren, J., Coburn, S. & Ryan, L.M. 1995, 'Alkaline-Phosphatase - Placental And Tissue-Nonspecific Isoenzymes Hydrolyze Phosphoethanolamine, Inorganic Pyrophosphate, And Pyridoxal 5'-Phosphate - Substrate Accumulation In Carriers Of Hypophosphatasia Corrects During Pregnancy', Journal Of Clinical Investigation, vol. 95, no. 4, pp. 1440-1445.
View/Download from: UTS OPUS or Publisher's site
Hypophosphatasia features selective deficiency of activity of the tissue-non-specific (liver/bone/kidney) alkaline phosphatase (ALP) isoenzyme (TNSALP); placental and intestinal ALP isoenzyme (PALP and IALP, respectively) activity is not reduced, Three p
Catalano, P.J., Rogus, J. & Ryan, L.M. 1995, 'Consequences of prolonged inhalation of ozone on F344/N rats: collaborative studies. Part X: Robust composite scores based on median polish analysis.', Research report (Health Effects Institute), no. 65 Pt 10, pp. 1-57.
This report describes some of the statistical methods used to analyze data from the National Toxicology Program/Health Effects Institute Collaborative Ozone Project. The purpose of the collaborative study was to assess the health effects of chronic ozone inhalation. Data were obtained from a subset of 164 F344/N rats dedicated to use by the Health Effects Institute from a standard ozone inhalation study conducted by Battelle Pacific Northwest Laboratories for the National Toxicology Program. The study involved eight groups of investigators, each assessing different types of ozone-related health effects. These included studies of respiratory function and of structural, cellular, and biochemical changes in the lungs and airways. Designing and analyzing a study with several groups of investigators raises many statistical challenges. The highest design priority for this study was that each investigation be individually interpretable as an independent study. This meant that each investigator had to receive an adequate number of animals, balanced with respect to level of ozone exposure and other factors such as the gender of the rats and the time they were killed. Another feature of the collaborative study was the opportunity it provided to assess and quantify the effect of ozone exposure on a broad spectrum of endpoints, and to explore the relations between the different types of effect. Maximizing the potential to assess these correlations required that the individual animals studied by the different groups of investigators overlap as much as possible. This aspect of the statistical design required careful consideration of the compatibility between various investigations. Fortunately, the degree of compatibility was substantial. In many cases, for example, it was possible to assess respiratory function in the animals before they were killed, and then to divide the tissue among several different investigators. This report concentrates on the methods that were speciall...
Ryan, L.M. 1995, 'Comment', Statistical Science, vol. 10, no. 2, pp. 189-193.
View/Download from: Publisher's site
Catalano, P.J., Chang, L.Y., Harkema, J.R., Kaden, D.A., Last, J.A., Mellick, P.W., Parks, W.C., Pinkerton, K.E., Radhakrishnamurthy, B. & Ryan, L.M. 1995, 'Consequences of prolonged inhalation of ozone on F344/N rats: collaborative studies. Part XI: Integrative Summary.', Research report (Health Effects Institute), no. 65 Pt 11, pp. 1-54.
Legler, J.M., Ryan, L.M., Harvey, E.A. & Holmes, L.B. 1994, 'Anticonvulsant teratogenesis: 2. Statistical methods for multiple birth outcomes', Teratology, vol. 50, no. 1, pp. 74-79.
View/Download from: UTS OPUS or Publisher's site
The purpose of this paper is to demonstrate the application of generalized estimating equations to assess an exposure effect using multiple birth outcomes. This multivariate approach provides the flexibility of regression modeling and improved power, as compared to series of univariate analyses or collapsing the multiple endpoints to a single indicator of affectedness. Motivating the discussion will be a large cohort study designed to assess the effects of anticonvulsant medications on a variety of birth outcomes, including major malformations, and growth and weight parameters, as well as a broad spectrum of minor physical anomalies. Because the study is still in progress, the aim here is not to present a definitive analysis, but to present and describe the application of these recently developed statistical methods to analyze studies with multiple outcomes. For simplicity, we will focus on the control and drug-exposed groups only from that study (ignoring the seizure history group), and we will concentrate on an analysis of minor physical anomalies. (C) 1994 Wiiey-Liss, Inc.
CATALANO, P., RYAN, L. & SCHARFSTEIN, D. 1994, 'MODELING FETAL DEATH AND MALFORMATION IN DEVELOPMENTAL TOXICITY STUDIES', RISK ANALYSIS, vol. 14, no. 4, pp. 629-637.
View/Download from: Publisher's site
Rosenthal, A., Mccarty, B., Cheung, H. & Ryan, L.M. 1993, 'A Comparison Of The Effect Of Transforming Growth-Factor-Beta-1 On Pyrophosphate Elaboration From Various Articular Tissues', Arthritis And Rheumatism, vol. 36, no. 4, pp. 539-542.
View/Download from: Publisher's site
Objective. The purpose of this study was to determine the effect of transforming growth factor beta1 (TGFbeta1) on inorganic pyrophosphate (PPi) elaboration from articular tissues to better understand the pathophysiology of calcium pyrophosphate dihydrat
Catalano, P.J., Scharfstein, D.O., Ryan, L.M., Kimmel, C.A. & Kimmel, G.L. 1993, 'Statistical model for fetal death, fetal weight, and malformation in developmental toxicity studies', Teratology, vol. 47, no. 4, pp. 281-290.
View/Download from: Publisher's site
The purpose of this paper is to present a statistical model for analyzing the joint effects of exposure on fetal death, fetal weight, and malformation in a developmental toxicity study. In addition to allowing for the usual litter effect, the model allow
Lindsey, J.C. & Ryan, L.M. 1993, 'A three-state multiplicative model for rodent tumorigenicity experiments', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 42, no. 2, pp. 283-300.
View/Download from: Publisher's site
A three-state illness-death model provides a useful way to represent data from rodent tumorigenicity experiments. Some of the earliest proposals use fully parametric models based on, for example, Weibull distributional assumptions. Recently, nonparametri
Ryan, L.M. 1993, 'Using historical controls in the analysis of developmental toxicity data', Biometrics, vol. 49, no. 4, pp. 1126-1135.
View/Download from: UTS OPUS or Publisher's site
Historical control data can often aid in the interpretation of the results from laboratory dose-response experiments. For example, formal statistical methods for using historical data are well established for carcinogenicity studies. In that case, a score test derived from a beta-binomial model yields a simple modification of the standard Cochran-Armitage test for trend. However, this test cannot be used for developmental toxicity studies since it does not allow for the additional correlation structure induced by the presence of litter effects. In this paper, quasi-likelihood methods are used to incorporate historical control information into a trend test for the types of correlated binary outcomes that typically arise in developmental toxicity studies. The proposed test is asymptotically equivalent to the beta-binomial score test in the special case when all the litter sizes are equal to 1. Malformation data from a series of developmental toxicity studies illustrate the results.
Lefkopoulou, M. & Ryan, L.M. 1993, 'Global tests for multiple binary outcomes', Biometrics, vol. 49, pp. 975-988.
View/Download from: UTS OPUS or Publisher's site
The applied statistician often encounters the need to compare two or more groups with respect to more than one outcome or response. Several options are generally available, including reducing the dimension of the problem by averaging or summarizing the outcomes, using Bonferroni or other adjustments for multiple comparisons, or applying a global test based on a suitable multivariate model. For normally distributed data, it is well established that global tests tend to be significantly more sensitive than other procedures. While global tests have also been proposed for multiple binary outcomes, their properties have not been well studied nor have they been widely discussed in the context of clustered data. In this paper, we derive a class of quasi-likelihood score tests for multiple binary outcomes, and show that special cases of this class correspond to other tests that have been proposed. We discuss extensions to allow for clustered data, and compare the results to the simple approach of collapsing the data to a single binary outcome, indicating the presence or absence of at least one response. The asymptotic relative efficiencies of the tests are shown to depend not only on the correlation between the outcomes, but also on the response probabilities. Although global tests based on a multivariate model are generally recommended, our findings suggest that a test based on the collapsed data can maintain surprisingly high efficiency, especially when the outcomes of interest are rare. Data from several developmental toxicity studies illustrate our results.
Rabinowitz, D. & Ryan, L.M. 1993, 'Lower confidence bounds for time to cure', Biometrika, vol. 80, no. 3, pp. 681-687.
View/Download from: UTS OPUS or Publisher's site
Clinicians often wish to know if and when patients with cancer or other life-threatening diseases can be considered cured. Although some parametric approaches are available, no nonparametric methods exist for treating this question. Here, a nonparametric procedure based on the definition that cure occurs when the hazard for death among patients is-no longer greater than the hazard among subjects in the general, population with similar age and demographic characteristics is developed. It is argued that the use of a fully nonparametric approach cannot lead to a nontrivial upper confidence bound for time to cure, and an approach to computing lower confidence bounds for time to cure is developed and applied to data from a clinical trial in colorectal cancer.
GRODSTEIN, F., GOLDMAN, M.B., RYAN, L. & CRAMER, D.W. 1993, 'SELF-REPORTED USE OF PHARMACEUTICALS AND PRIMARY OVULATORY INFERTILITY', EPIDEMIOLOGY, vol. 4, no. 2, pp. 151-156.
View/Download from: Publisher's site
GRODSTEIN, F., GOLDMAN, M.B., RYAN, L. & CRAMER, D.W. 1993, 'RELATION OF FEMALE INFERTILITY TO CONSUMPTION OF CAFFEINATED BEVERAGES', AMERICAN JOURNAL OF EPIDEMIOLOGY, vol. 137, no. 12, pp. 1353-1360.
RYAN, L., KRAMAR, A. & BORDEN, E. 1993, 'PROGNOSTIC FACTORS IN METASTATIC MELANOMA', CANCER, vol. 71, no. 10, pp. 2995-3005.
View/Download from: Publisher's site
Edmonson, J.H., Ryan, L.M., Blum, R.H., Brooks, J.S., Shiraki, M., Frytak, S. & Parkinson, D.R. 1993, 'Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas.', Journal of clinical oncology : official journal of the American Society of Clinical Oncology, vol. 11, no. 7, pp. 1269-1275.
PURPOSE: This three-armed phase III study in adults with advanced soft tissue sarcomas was planned as a comparison of objective regression rates, toxicity, and survival of patients receiving doxorubicin alone, ifosfamide plus doxorubicin, and mitomycin plus doxorubicin plus cisplatin. PATIENTS AND METHODS: Between December 1987 and July 1990, 279 patients with histologically confirmed sarcomas were enrolled to receive treatment A (doxorubicin 80 mg/m2), treatment B (ifosfamide 7.5 g/m2 plus doxorubicin 60 mg/m2), or treatment C (mitomycin 8 mg/m2 plus doxorubicin 40 mg/m2 plus cisplatin 60 mg/m2). RESULTS: Of 262 assessable patients, 74 (29%) achieved objective tumor regression. Objective regression occurred in 20% of the 90 patients who received doxorubicin alone (complete remission [CR] rate, 2%), in 34% of the 88 who received ifosfamide plus doxorubicin (CR rate, 3%), and in 32% of the 84 who received mitomycin plus doxorubicin plus cisplatin (CR rate, 7%). With grade 3 or greater myelosuppression in 53% of group A, 80% of group B, and 55% of group C, regimen B was significantly more myelosuppressive than either regimen A or C (P = .01) with two, three, and one treatment-related deaths, respectively. Synovial sarcomas were responsive to ifosfamide plus doxorubicin, especially among patients younger than 40 years of age. CONCLUSION: Ifosfamide plus doxorubicin produced a significantly higher regression rate (P = .03) than did doxorubicin alone; however, this was achieved at a level of myelosuppression significantly more intense than that produced by the single agent or by the three-drug combination. Mitomycin, doxorubicin, and cisplatin also appeared to be more active than the single agent; however, at a myelosuppression level similar to that of doxorubicin alone, this trend (P = .07) did not attain the usual level for significance. No significant survival differences were observed.
Leisenring, W. & Ryan, L.M. 1992, 'Statistical properties of the NOAEL', Regulatory Toxicology and Pharmacology, vol. 15, pp. 161-171.
View/Download from: UTS OPUS or Publisher's site
Borden, E.C., Kim, K., Ryan, L.M., Blum, R.H., Shiraki, M., Tormey, D.C., Comis, R.H., Hahn, R.G. & Parkinson, D.R. 1992, 'Phase II trials of interferons-alpha and -beta in advanced sarcomas', Journal of Interferon Research, vol. 12, no. 6, pp. 455-458.
View/Download from: UTS OPUS or Publisher's site
Falkson, G., Ryan, L.M. & Haller, D. 1992, 'Phase-ii Trial For The Evaluation Of Trimetrexate In Patients With Inoperable Squamous Carcinoma Of The Esophagus', American Journal Of Clinical Oncology-cancer Clinical Trials, vol. 15, no. 5, pp. 433-435.
View/Download from: Publisher's site
EST 2287 was a Phase II clinical trial conducted by the Eastern Cooperative Oncology Group (ECOG) designed to evaluate trimetrexate in patients with advanced, measurable, inoperable squamous cell carcinoma of the esophagus. The drug was given at a dose o
Catalano, P.J. & Ryan, L.M. 1992, 'Bivariate latent variable models for clustered discrete and continuous outcomes', Journal of the American Statistical Association, vol. 87, no. 419, pp. 651-658.
View/Download from: UTS OPUS or Publisher's site
We use the concept of a latent variable to derive the joint distribution of a continuous and a discrete outcome, and then extend the model to allow for clustered data. The model Can be parameterized in a way that allows one to write the joint distribution as a product of a standard random effects model for the continuous variable and a correlated probit model for the discrete variable. This factorization suggests a convenient approach to parameter estimation using quasi-likelihood techniques. Our approach is motivated by the analysis of developmental toxicity experiments for which a number of discrete and continuous outcomes are measured on offspring clustered within litters. Fetal weight and malformation data illustrate the results.
Ryan, L.M. 1992, 'Quantitative risk assessment for developmental toxicity', Biometrics, vol. 48, pp. 163-174.
View/Download from: UTS OPUS or Publisher's site
Pharmaceutical companies and governmental regulatory agencies are becoming increasingly aware of the need for improved statistical methods for developmental toxicity experiments. Although a number of statisticians have become interested in this area, activity has centered mostly on the development of methods to analyze binary outcomes, such as malformations among live pups, while accounting appropriately for the correlation induced by the litter effect. In contrast, the topic of quantitative risk assessment has received relatively little attention. This paper addresses the specific question of how to assess risk appropriately when exposure causes a variety of adverse effects, including resorption and fetal death, in addition to malformations. It will be seen that risk assessments based on a single developmental outcome, such as malformation, may be conservative. A method is proposed for estimating an exposure level at which the overall risk of any adverse effect is acceptably low. The method is based on a continuation ratio formulation of a multinomial distribution, with an additional scale parameter to account for overdispersion. Comparisons are made with binary models on prenatal death and malformation, as well as a binary model that makes no distinction between death and malformation, but simply classifies each fetus as normal or abnormal. Data from several developmental toxicity studies illustrate the results and findings.
D'agostino, R.B., Lange, N. & Ryan, L.M. 1992, 'Overview', Statistics in Medicine, vol. 11, pp. 1801-1805.
RYAN, L. 1992, 'THE USE OF GENERALIZED ESTIMATING EQUATIONS FOR RISK ASSESSMENT IN DEVELOPMENTAL TOXICITY', RISK ANALYSIS, vol. 12, no. 3, pp. 439-447.
View/Download from: Publisher's site
DAGOSTINO, R.B., LANGE, N. & RYAN, L. 1992, 'PAPERS FROM THE SYMPOSIUM ON LONGITUDINAL DATA-ANALYSIS, 19-21 JUNE 1991 - OVERVIEW', STATISTICS IN MEDICINE, vol. 11, no. 14-15, pp. 1801-1805.
View/Download from: Publisher's site
Catalano, P.J. & Ryan, L.M. 1992, 'Bivariate latent variable models for clustered discrete and continuous outcomes', Journal of the American Statistical Association, vol. 87, no. 419, pp. 651-658.
View/Download from: Publisher's site
Engstrom, P., Ryan, L.M., Falkson, G. & Haller, D. 1991, 'Phase-ii Study Of Aminothiadiazole In Advanced Squamous-cell Carcinoma Of The Esophagus', American Journal Of Clinical Oncology-cancer Clinical Trials, vol. 14, no. 1, pp. 33-35.
View/Download from: Publisher's site
Twenty-three patients with advanced inoperable squamous cell carcinoma of the esophagus were treated with aminothiadiazole (A-TD) 125 mg/m2 weekly plus allopurinol daily in a phase II cooperative group trial. No patients responded to treatment; 17 patien
Cheuvart, B. & Ryan, L.M. 1991, 'Adjusting for age-related competing mortality in long-term cancer clinical trials', Statistics in Medicine, vol. 10, pp. 65-77.
View/Download from: UTS OPUS or Publisher's site
Mortality related to causes other than the treated disease may have a significant impact on overall survival in long-term clinical trials. We present a model that adjusts for age-related competing mortality when cause of death is missing or only partially available. Through use of a piecewise exponential survival model, we extend relative survival methods to continuous follow-up data, allowing the competing mortality to differ from that of the general population by a scale parameter. An EM algorithm provides a simple way to compute the maximum likelihood estimators (MLEs) and to test hypotheses using widely available software. We compare the bias and relative efficiency of this model to a piecewise exponential Cox model for overall survival. Theoretical results are confirmed by simulations and illustrated with data from a clinical trial in colorectal cancer. This example also shows how age-related and disease-related mortality can be confounded in an analysis of overall survival. We conclude with a discussion of the advantages and disadvantages of the model.
Ryan, L.M., Catalano, P.J., Kimmel, C.A. & Kimmel, G.L. 1991, 'Relationship between fetal weight and malformation in developmental toxicity studies', Teratology, vol. 44, no. 2, pp. 215-223.
View/Download from: UTS OPUS or Publisher's site
Exposure to developmental toxicants may cause fetal malformations, increase prenatal death rates and reduce fetal weight at term. However, there has been little formal study of the relationship among these effects. Certainly, no statistical methods are currently available to jointly analyze these effects of exposure. As a preliminary step in developing such methods, simple exploratory analyses were conducted using a series of ten studies conducted for the National Toxicology Program. Because fetal weight and malformation status were both reported for all live fetuses, the data permitted an exploration of the correlation between these two outcomes. The data show a clear pattern wherein malformed fetuses tended to be lighter at term than nonmalformed fetuses. While these patterns cannot be used to draw inferences regarding the biological relationship between fetal weight and malformation, they do suggest the potential value in developing statistical models for the joint effect of exposure on fetal weight and malformations.
Goetghebeur, E. & Ryan, L.M. 1990, 'A modified log rank test for competing risks with missing failure type', Biometrika, vol. 77, no. 1, pp. 207-211.
View/Download from: UTS OPUS or Publisher's site
We propose a modified log rank test for the analysis of competing risks survival data, when failure type is missing for some individuals. The proposed test reduces to a standard log rank test when all failure types are known. The test arises from a partial likelihood, constructed under semiparametric assumptions on the relationship between cause-specific hazards.
Cnaan, A. & Ryan, L. 1990, 'Survival analysis in natural history studies of disease (I: Reply)', Statistics in Medicine, vol. 9, no. 10, p. 1222.
Antman, K.H., Elias, A. & Ryan, L. 1990, 'Ifosfamide and mesna: Response and toxicity at standard- and high-dose schedules', Seminars in Oncology, vol. 17, no. 2 SUPPL. 4, pp. 68-73.
In two sequential trials, 154 patients were treated with dosages of ifosfamide, ranging between 8 and 18 g/m 2 divided over 4 days, with mesna uroprotection. The first was a phase II efficacy trial in 125 advanced sarcoma patients (Antman et al: J Clin Oncol 7:126-131, 1989), while the second was a dose escalation trial involving 29 patients (Elias et al: J Clin Oncol 8:170-178, 1990). In the first trial, patients received 8 to 10 g/m 2 ifosfamide either by bolus or continuous infusion. The response rate for the 64 patients receiving bolus administration was 23% compared with 12% for the 60 patients receiving a continuous infusion schedule (P = .09). Of the 154 patients, 144 had sarcoma and had failed at least one previous regimen. Of these 144, 4% responded completely and 23% had a complete or partial response. The maximum tolerated dose of ifosfamide was 16 g/m 2 in the second trial. Dose-limiting renal toxicity was observed at 18 g/m 2 ifosfamide (Elias et al: J Clin Oncol 8:170-178, 1990). The duration of myelosuppression and the frequency and severity of mucositis and renal tubular acidosis were dosedependent. A median of 11 days (range, 8 to 18) of granulocytopenia ( < 500/L) were observed. Thus, autologous bone marrow reinfusion was not required. Severe central nervous system toxicity (transient confusion, hallucinations, and somnolence) was observed sporadically at both low- and high-dose levels. The first four patients on the standard-dose study did not receive mesna because it was unavailable; three developed gross hematuria. In patients who received mesna, hematuria was uncommon. Hematuria in the group as a whole was significantly associated with a lack of uroprotection, but was not associated with prior cyclophosphamide, pelvic radiotherapy, age, or bolus versus a continuous infusion schedule. Patients receiving ifosfamide with mesna uroprotection can tolerate considerable dose escalation over the usual prescribed doses before nonhematologic toxi...
Elias, A., Ryan, L., Aisner, J. & Antman, K.H. 1990, 'Mesna, doxorubicin, ifosfamide, dacarbazine (MAID) regimen for adults with advanced sarcoma', Seminars in Oncology, vol. 17, no. 2 SUPPL. 4, pp. 41-49.
The mesna, doxorubicin, ifosfamide, dacarbazine regimen produced a 47% response rate (including 10% complete responses) in 105 eligible adults with advanced sarcoma. The major dose-limiting toxicity was granulocytopenia. There was one toxic death from sepsis. Central nervous system and renal toxicity occurred infrequently, perhaps as a result of the continuous-infusion schedule. This regimen is being evaluated further in advanced disease, the adjuvant setting, and in combination with bone marrow colony-stimulating factors. &copy; 1990.
Keiding, N. & Knuiman, M.W. 1990, 'Survival analysis in natural history studies of disease.', Statistics in medicine, vol. 9, no. 10, pp. 1221-1222.
Lange, N. & Ryan, L.M. 1989, 'Assessing normality in random effects models', Annals of Statistics, vol. 17, no. 2, pp. 624-642.
View/Download from: UTS OPUS or Publisher's site
Archer, L.E. & Ryan, L.M. 1989, 'Accounting for misclassification in the cause of death test for carcinogenicity', Journal Of The American Statistical Association, vol. 84, no. 407, pp. 787-791.
View/Download from: UTS OPUS or Publisher's site
NA
Siegel, R.D., Ryan, L.M. & Antman, K.H. 1989, 'Osteosarcoma in adults. One institution's experience', Clinical Orthopaedics And Related Research, vol. 240, pp. 261-269.
NA
Archer, L.E. & Ryan, L.M. 1989, 'On the role of cause of death in the analysis of rodent tumorigenicity experiments', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 38, pp. 81-93.
View/Download from: Publisher's site
Lefkopoulou, M., Moore, D. & Ryan, L.M. 1989, 'The analysis of multiple correlated binary outcomes: Application to rodent teratology experiments', Journal of the American Statistical Association, vol. 84, no. 407, pp. 810-815.
View/Download from: UTS OPUS or Publisher's site
Cnaan, A. & Ryan, L.M. 1989, 'Survival analysis in natural history studies of disease', Statistics in Medicine, vol. 8, pp. 1255-1268.
View/Download from: UTS OPUS or Publisher's site
Clinicians often wish to use data from clinical trials or hospital databases to study disease natural history. Of particular interest are estimated survival and prognostic factors. In this context, it may be appropriate to measure survival from diagnosis or some other time origin, possibly prior to study entry. We describe the application of methods for truncated survival data, and compare these with the standard product limit estimator and proportional hazards models in the measurement of survival from entry. Theoretical considerations suggest that analysis of survival from entry may under- or overestimate the survival distribution of interest, depending on the shape of the true underlying hazard. Analogous results hold for the coefficients from a proportional hazards model. We illustrate our findings with data from a multicenter clinical trial and a hospital database.
ELIAS, A., RYAN, L., SULKES, A., COLLINS, J., AISNER, J. & ANTMAN, K.H. 1989, 'RESPONSE TO MESNA, DOXORUBICIN, IFOSFAMIDE, AND DACARBAZINE IN 108 PATIENTS WITH METASTATIC OR UNRESECTABLE SARCOMA AND NO PRIOR CHEMOTHERAPY', JOURNAL OF CLINICAL ONCOLOGY, vol. 7, no. 9, pp. 1208-1216.
ANTMAN, K.H., RYAN, L., ELIAS, A., SHERMAN, D. & GRIER, H.E. 1989, 'RESPONSE TO IFOSFAMIDE AND MESNA - 124 PREVIOUSLY TREATED PATIENTS WITH METASTATIC OR UNRESECTABLE SARCOMA', JOURNAL OF CLINICAL ONCOLOGY, vol. 7, no. 1, pp. 126-131.
Archer, L.E. & Ryan, L.M. 1989, 'Accounting for misclassification in the cause-of-death test for carcinogenicity', Journal of the American Statistical Association, vol. 84, no. 407, pp. 787-791.
View/Download from: Publisher's site
The cause-of-death test (Peto 1974; Peto et al. 1980) provides a test for carcinogenicity without requiring extreme lethality assumptions. The approach has been criticized, however, because cause-of-death determinations require subjective decisions by pathologists and may be unreliable. Using a missing data formulation, we derive a simple, intuitive modification of the test that allows for misclassification of cause of death. The modified statistic is a function of the misclassification probabilities. A sensitivity analysis may be performed using a plausible range of values for these probabilities or, in certain situations, the misclassification probabilities can be estimated from the data. Both approaches are applied to data from the ED 01 experiment. &copy; 1989 Taylor & Francis Group, LLC.
Lefkopoulou, M., Moore, D. & Ryan, L. 1989, 'The analysis of multiple correlated binary outcomes: Application to rodent teratology experiments', Journal of the American Statistical Association, vol. 84, no. 407, pp. 810-815.
View/Download from: Publisher's site
In a developmental toxicity study, pregnant animals are exposed to the test substance and their offspring are assessed for defects. Often, multiple observations are made on each fetus, in which case the data are doubly nested. In this article we adapt the approach of Liang and Zeger (1986) and Zeger and Liang (1986) to yield an analysis, which appropriately allows for the correlation structure. &copy; 1989 Taylor & Francis Group, LLC.
Ryan, L.M. & Orav, J.E. 1989, ''On the use of covariates for rodent bioassay and screening experiments'', Biometrika, vol. 76, no. 2, p. 407.
View/Download from: Publisher's site
Siegel, R.D., Ryan, L.M. & Antman, K.H. 1988, 'Adults with Ewing's sarcoma. An analysis of 16 patients at the Dana-Farber Cancer Institute', American Journal Of Clinical Oncology-cancer Clinical Trials, vol. 11, no. 6, pp. 614-617.
View/Download from: Publisher's site
NA
Ryan, L.M. & Orav, E.J. 1988, 'On the use of covariates for rodent bioassay and screening experiments', Biometrika, vol. 75, no. 4, pp. 631-637.
View/Download from: UTS OPUS
NA
Falkson, G., Cnaan, A., Schutt, A., Schutt, A.J., Ryan, L.M. & Falkson, H.C. 1988, 'Prognostic factors for survival in hepatocellular carcinoma', Cancer Research, vol. 48, no. 24, pp. 7314-7318.
View/Download from: UTS OPUS
Associations between patient characteristics and survival were inves tigated in 432 patients with hepatocellular carcinoma. Those patients were prospective!}' studied by the Eastern Cooperative Oncology Group, and each had his or her diagnosis reconfirmed by a pathology review panel. There were 301 North American and 131 South African patients. Sixty-nine % of the North American patients and 82% of the South African patients were male. There were 187 Black patients, 62 of whom were from North America. The study population is unique among hepatocellular carcinoma pa tients in that eligibility, evaluability, and endpoint definitions were stand ardized, and patients from both North America and South Africa received similar treatments at a similar time. Factors with the most significant adverse effect on survival are im paired performance status, male sex, older age, and disease symptoms (jaundice and reduced appetite). There is no apparent difference in survival between White and Black patients within North America, but North American patients survived longer than South African patients. Among the different therapies, p.o. 5-fluorouracil was associated with the poorest median survival time (6 wk), and i.v. 5-fluorouracil plus semustine with the best median survival time (24 wk).
ANTMAN, K.S., GRIFFIN, J.D., ELIAS, A., SOCINSKI, M.A., RYAN, L., CANNISTRA, S.A., OETTE, D., WHITLEY, M., FREI, E. & SCHNIPPER, L.E. 1988, 'EFFECT OF RECOMBINANT HUMAN GRANULOCYTE MACROPHAGE COLONY-STIMULATING FACTOR ON CHEMOTHERAPY-INDUCED MYELOSUPPRESSION', NEW ENGLAND JOURNAL OF MEDICINE, vol. 319, no. 10, pp. 593-598.
View/Download from: Publisher's site
ANTMAN, K., SHEMIN, R., RYAN, L., KLEGAR, K., OSTEEN, R., HERMAN, T., LEDERMAN, G. & CORSON, J. 1988, 'MALIGNANT MESOTHELIOMA - PROGNOSTIC VARIABLES IN A REGISTRY OF 180 PATIENTS, THE DANA-FARBER-CANCER-INSTITUTE AND BRIGHAM-AND-WOMENS-HOSPITAL EXPERIENCE OVER 2 DECADES, 1965-1985', JOURNAL OF CLINICAL ONCOLOGY, vol. 6, no. 1, pp. 147-153.
Falkson, G., Ryan, L.M., Johnson, L.A., Simson, I.W., Coetzer, B.J., Carbone, P.P., Creech, R.H. & Schutt, A.J. 1987, 'A random phase II study of mitoxantrone and cisplatin in patients with hepatocellular carcinoma: An ECOG study', Cancer, vol. 60, no. 9, pp. 2141-2145.
View/Download from: UTS OPUS or Publisher's site
Of 86 patients entered in an Eastern Cooperative Oncology Group (ECOG) random Phase II study of mitoxantrone (DHAD) and cisplatin (DDP) in primary liver cancer, 69 were eligible. Nine of the 13 ineligible patients were excluded after a pathology review. Sixty-one percent of the patients were North American, and 39% were South African. The most common severe or the worst toxicity on DHAD was hematologic; and to DDP, hematologic and vomiting. Of the 69 eligible patients, 21 experienced severe, life-threatening or fatal toxic reactions. Two patients treated with DDP had partial responses. With a 95% confidence interval, the true response rate to DHAD was less than 8%, and to DDP, less than 17%. The median survival time was 14 weeks on both drugs. Assuming a proportional hazards model, factors that are significantly associated with survival are patient performance status, the presence of the symptoms, raised bilirubin and hepatomegaly, and clinical evidence of cirrhosis. Any differences between survival rates for South African and North American patients were largely explainable by these factors.
Finkelstein, D.M. & Ryan, L.M. 1987, 'Estimating carcinogenic potency from a rodent tumorigenicity experiment', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 36, no. 2, pp. 121-133.
View/Download from: UTS OPUS or Publisher's site
Rodent tumorigenicity experiments are conducted to determine whether a particular substance accelerates tumour development. The association between exposure to this substance and the risk of tumour development can be characterized by a measure of carcinogenic potency. The most commonly used potency measure, the dose effect on the lifetime risk of tumour, may be seriously biased, especially when control and exposed groups differ with respect to longevity. However, more appropriate measures, which account for age at death, are largely unavailable except in the special cases of instantly lethal or nonlethal tumours or tumours with observable onset. In this paper, we propose an estimate of carcinogenic potency based on a proportonal prevalence odds model which applies regardless of tumour lethality and can be calculated using standard statistical methodologies. Furthermore, we show how our estimator can be used to generalize available potency estimators to tumours of any lethalit
Buyse, M. & Ryan, L.M. 1987, 'Issues of efficiency in combining proportions of deaths from several clinical trials', Statistics in Medicine, vol. 6, no. 5, pp. 565-576.
View/Download from: UTS OPUS or Publisher's site
The Mantel-Haenszel test provides a straightforwdd method to combine results from several clinical trials when only summary information, such as the proportion of deaths, is available for each trial. More efficient tests, such as the stratified logrank test, should be used if the survival and censoring times are known for all individuals, but in practice, the cost and effort of obtaining this information may be prohibitive. The purpose of this paper is to derive a general expression for the asymptotic relative efficiency (ARE) of the Mantel-Haenszel test with respect to the stratified logrank test, and to compute the ARE in situations which rue likely to be of practical interest. The results show that under realistic assumptions about the survival distribution, losses to follow-up and duration of accrual, the ARE frequently exceeds 80 per cent. An example is given to show the usefulness of the approach when combining proportions of deaths from several cancer dinical trials.
CNAAN, A. & RYAN, L. 1987, 'ANALYZING SURVIVAL FROM INITIAL DIAGNOSIS', CONTROLLED CLINICAL TRIALS, vol. 8, no. 3, pp. 283-283.
RYAN, L.M. 1987, 'ON THE EFFICIENCY OF AGE-ADJUSTED TESTS IN ANIMAL CARCINOGENICITY EXPERIMENTS - RESPONSE', BIOMETRICS, vol. 43, no. 1, pp. 243-244.
Ryan, L.M. 1985, 'Efficiency of age adjusted tests in animal carcinogenicity experiments', Biometrics, vol. 41, no. 2, pp. 525-531.
View/Download from: UTS OPUS or Publisher's site
Lagakos, S. & Ryan, L.M. 1985, 'Statistical-analysis Of Disease Onset And Lifetime Data From Tumorigenicity Experiments', Environmental Health Perspectives, vol. 63, no. NA, pp. 211-216.
View/Download from: UTS OPUS
NA
Dempster, A.P. & Ryan, L.M. 1985, 'Weighted normal plots', Journal of the American Statistical Association, vol. 80, no. 392, pp. 845-850.
View/Download from: UTS OPUS or Publisher's site
NA
Lagakos, S.W. & Ryan, L.M. 1985, 'On the representativeness assumption in prevalence tests of carcinogenicity', Journal of the Royal Statistical Society Series C: Applied Statistics, vol. 34, no. 1, pp. 54-62.
View/Download from: UTS OPUS or Publisher's site
Conditions for the validityo f the Hoel-Walburga nd Peto tests, which compare dose groups with respect to tumour prevalence,a re replaced by a more generalc ondition, representativenesAs. large carcinogenicityex perimentp rovideda unique opportunityt o assess empiricallyif t his conditionh olds. Though representativenesws as generallyv iolated, neithert he Hoel-Walburgn or the Peto tests were seriouslyd istorted.A nalyticc onsiderationss uggestt hat such robustnessc an occur in manys ituations
RYAN, L.M. 1985, 'THE ANALYSIS OF TUMORIGENICITY EXPERIMENTS, USING INFORMATION ON TUMOR STAGE', BIOMETRICS, vol. 41, no. 2, pp. 567-567.
RYAN, L.M. & LAGAKOS, S.W. 1985, 'REPRESENTATIVENESS IN TUMOR PREVALENCE TESTS AND ITS IMPLICATIONS FOR ASSIGNING CAUSE OF DEATH', BIOMETRICS, vol. 41, no. 1, pp. 319-319.
Dempster, A.P. & Ryan, L.M. 1985, 'Weighted normal plots', Journal of the American Statistical Association, vol. 80, no. 392, pp. 845-850.
View/Download from: Publisher's site
Weighted normal plots are proposed as graphical checks on the normality of random effects in Gaussian linear models. The technique is illustrated using the one-way comparisons model Y i = i + i , where the ( i , i ), are independent pairs with i and i , independent N(0, 2 ) and N(0, 2 i ), respectively, for i = 1, &#8230;, n. When the variance components 2 and 2 i are known, an unweighted normal plot of the standardized Z i = Y i ( 2 + 2 i ) 1/2 provides a check of the overall adequacy of the model. Weighted normal plots involve a modification that gives the ith observation a sample weight of W i = ( 2 + 2 i ) 1 . Under the null hypothesis, the sample size must be larger by a factor of (1 + v/m 2 ), where m and v are the mean and variance of the weights, to produce a weighted plot with approximately the same sampling variance as an unweighted normal plot. Despite this higher variability, we show that weighted plots are more sensitive than unweighted plots to several departures from the assumed distribution on the random effects, i . Several numerical examples are included and the effects of substituting maximum likelihood estimates for the parameters 2 and 2 i are considered briefly. &copy; 1985 Taylor & Francis Group, LLC.
Cheung, H. & Ryan, L.M. 1981, 'A Method Of Determining Dna And Chondrocyte Content Of Articular-Cartilage', Analytical Biochemistry, vol. 116, no. 1, pp. 93-97.
View/Download from: Publisher's site
NA

Reports

Rai, T. & Ryan, L. 2014, Vein Visualization Trial - 2nd Interim Report, pp. 1-17, Sydney.
View/Download from: UTS OPUS
Rai, T. & Ryan, L. 2014, Vein Visualization Trial - Interim Analysis, pp. 1-20, Sydney.
View/Download from: UTS OPUS