Showing papers by "Ross L. Prentice published in 2004"

Estrogen plus progestin and risk of venous thrombosis.

Abstract: Context Postmenopausal hormone therapy increases the risk of venous thrombosis. It is not known whether other factors influencing thrombosis add to this risk. Objective To report final data on incidence of venous thrombosis in the Women’s Health Initiative Estrogen Plus Progestin clinical trial and the association of hormone therapy with venous thrombosis in the setting of other thrombosis risk factors. Design, Setting, and Participants Double-blind randomized controlled trial of 16 608 postmenopausal women between the ages of 50 and 79 years, who were enrolled in 1993 through 1998 at 40 US clinical centers with 5.6 years of follow up; and a nested case-control study. Baseline gene variants related to thrombosis risk were measured in the first 147 women who developed thrombosis and in 513 controls. Intervention Random assignment to 0.625 mg/d of conjugated equine estrogen plus 2.5 mg/d of medroxyprogesterone acetate, or placebo. Main Outcome MeasuresCentrally validated deep vein thrombosis and pulmonary embolus.Results Venous thrombosis occurred in 167 women taking estrogen plus progestin (3.5 per 1000 person-years) and in 76 taking placebo (1.7 per 1000 person-years); hazard ratio (HR), 2.06 (95% confidence interval [CI], 1.57-2.70). Compared with women between the ages of 50 and 59 years who were taking placebo, the risk associated with hormone therapy was higher with age: HR of 4.28 (95% CI, 2.38-7.72) for women aged 60 to 69 years and 7.46 (95% CI, 4.32-14.38) for women aged 70 to 79 years. Compared with women who were of normal weight and taking placebo, the risk associated with taking estrogen plus progestin was increased among overweight and obese women: HR of 3.80 (95% CI, 2.08-6.94) and 5.61 (95% CI, 3.12-10.11), respectively. Factor V Leiden enhanced the hormone-associated risk of thrombosis with a 6.69-fold increased risk compared with women in the placebo group without the mutation (95% CI, 3.09-14.49). Other genetic variants (prothrombin 20210A, methylenetetrahydrofolate reductase C677T, factor XIII Val34Leu, PAI-1 4G/5G, and factor V HR2) did not modify the association of hormone therapy with venous thrombosis. Conclusions Estrogen plus progestin was associated with doubling the risk of venous thrombosis. Estrogen plus progestin therapy increased the risks associated with age, overweight or obesity, and factor V Leiden.

Research issues and strategies for genomic and proteomic biomarker discovery and validation: a statistical perspective

Abstract: The development and validation of clinically useful biomarkers from high-dimensional genomic and proteomic information pose great research challenges. Present bottlenecks include: that few of the biomarkers showing promise in initial discovery were found to warrant subsequent validation; and biomarker validation is expensive and time consuming. Biomarker evaluation should proceed in an orderly fashion to enhance rigor and efficiency. A molecular profiling approach, although promising, has a high chance of yielding biased results and overfitted models. Specimens from cohorts or intervention trials are essential to eliminate biases. The high cost for biomarker validation motivates some novel study design features, including sequential filtering and DNA pooling. For data analysis, logistic regression (in particular, boosting logistic regression) has features of robustness against model misspecification, and has resistance to model overfitting. Model assessment and cross-validation are critical components of data analysis. Having an independent test set is a vital feature of study design.

Nutrition and Physical Activity and Chronic Disease Prevention: Research Strategies and Recommendations

Abstract: A shortage of credible information exists on practical dietary and physical activity patterns that have potential to reverse the national obesity epidemic and reduce the risk of major cancers and other chronic diseases. Securing such information is a challenging task, and there is considerable diversity of opinion concerning related research designs and priorities. Here, we put forward some perspectives on useful methodology and infrastructure developments for progress in this important area, and we list high-priority research topics in the areas of 1) assessment of nutrient intake and energy expenditure; 2) development of intermediate outcome biomarkers; 3) enhancement of cohort and cross-cultural studies; and 4) criteria for and development of full-scale nutrition and physical activity intervention trials.

Chronic disease prevention: public health potential and research needs.

Abstract: This paper, arising out of an event to honour the statistical and scientific contributions of Professor Peter Armitage, is concerned with research strategies and needs for chronic disease prevention. A few highlights from recent intervention trials for the prevention of cancer, cardiovascular disease, fractures and diabetes is provided, along with a discussion of some settings where intervention trial results seem discrepant with a body of preceding observational data. This background is used to identify research strategies and infrastructure needs for moving this vitally important research area forward, for both chemoprevention and lifestyle modification interventions. Copyright © 2004 John Wiley & Sons, Ltd.

Hazard‐based nonparametric survivor function estimation

Abstract: Summary. A representation is developed that expresses the bivariate survivor function as a function of the hazard function for truncated failure time variables. This leads to a class of nonparametric survivor function estimators that avoid negative mass. The transformation from hazard function to survivor function is weakly continuous and compact differentiable, so that such properties as strong consistency, weak convergence to a Gaussian process and bootstrap applicability for a hazard function estimator are inherited by the corresponding survivor function estimator. The set of point mass assignments for a survivor function estimator is readily obtained by using a simple matrix calculation on the set of hazard rate estimators. Special cases arise from a simple empirical hazard rate estimator, and from an empirical hazard rate estimator following the redistribution of singly censored observations within strips. The latter is shown to equal van der Laan's repaired nonparametric maximum likelihood estimator, for which a Greenwood-like variance estimator is given. Simulation studies are presented to compare the moderate sample performance of various nonparametric survivor function estimators.

Nonparametric correction for covariate measurement error in a stratified Cox model.

Abstract: SUMMARY Stratified Cox regression models with large number of strata and small stratum size are useful in many settings, including matched case-control family studies. In the presence of measurement error in covariates and a large number of strata, we show that extensions of existing methods fail either to reduce the bias or to correct the bias under nonsymmetric distributions of the true covariate or the error term. We propose a nonparametric correction method for the estimation of regression coefficients, and show that the estimators are asymptotically consistent for the true parameters. Small sample properties are evaluated in a simulation study. The method is illustrated with an analysis of Framingham data.

Ethnicity and breast cancer in the Women's Health Initiative: A unifying concept for unfavorable outcome in African American women

Abstract: 1008 Background: The lower breast cancer incidence in minorities and higher breast cancer mortality in African Americans compared to Whites are largely unexplained. Therefore, relationships among e...

Hazard-based nonparametric survivor function

Abstract: A representation is developed that expresses the bivariate survivor function as a function of the hazard function for truncated failure time variables. This leads to a class of nonparametric survivor function estimators that avoid negative mass. The transformation from hazard function to survivor function is weakly continuous and compact differentiable, so that such properties as strong consistency, weak convergence to a Gaussian process and bootstrap applicability for a hazard function estimator are inherited by the corresponding survivor function estimator. The set of point mass assignments for a survivor function estimator is readily obtained by using a simple matrix calculation on the set of hazard rate estimators. Special cases arise from a simple empirical hazard rate estimator, and from an empirical hazard rate estimator following the redistribution of singly censored observations within strips. The latter is shown to equal van der Laan's repaired nonparametric maximum likelihood estimator, for which a Greenwood-like variance estimator is given. Simulation studies are presented to compare the moderate sample performance of various nonparametric survivor function estimators.