Search

To facilitate comparative treatment selection when there is substantial heterogeneity of treatment effectiveness, it is important to identify subgroups that exhibit differential treatment effects. Existing approaches model outcomes directly and then define subgroups according to interactions between treatment and covariates. Because outcomes are affected by both the covariate–treatment interactions and covariate main effects, direct modeling outcomes can be hard due to model misspecification, especially in presence of many covariates. Alternatively one can directly work with differential treatment effect estimation. We propose such a method that approximates a target function whose value directly reflects correct treatment assignment for patients. The function uses patient outcomes as weights rather than modeling targets. Consequently, our method can deal with binary, continuous, time-to-event, and possibly contaminated outcomes in the same fashion. We first focus on identifying only directional estimates from linear rules that characterize important subgroups. We further consider estimation of comparative treatment effects for identified subgroups. We demonstrate the advantages of our method in simulation studies and in analyses of two real data sets.

Twenty-two sixth-grade girls who participated in a 7-month school-based resistance-training program were compared to 22 controls. In a subanalysis of Tanner breast II (T2) and III (T3) subjects (n = 21 controls subjects (CON), n = 17 subjects in the high-intervention (INT)-dose group (HI)), T2 HI had greater narrow neck (NN) width gains than T2 CON (p 0.05) and T3 HI had greater L3 bone mineral density (BMD) gains than T3 CON (p 0.05).Physical activity modulates bone growth during adolescence, but an effective activity has not been identified for general use. The purpose of this study was to examine the effect of a school-based resistance-training program on skeletal growth in pre-menarcheal females.Sixth-grade girls participated in a 7-month, resistance-training program (INT) embedded in physical education (PE) classes. Age- and maturity-matched CON from a neighboring school participated in the standard PE classes. INT dose defined high (HI) and low (LO) groups. At baseline (BL) and follow-up (FU), non-INT organized physical activity (PA, hours per week) and maturity status were recorded; DXA scans assessed total body, distal radius, proximal femur, and lumbar spine. Regression models analyzed growth in bone outcomes for HI versus CON, accounting for age, Tanner stage, height, and PA.Forty-four girls (22 HI, 22 CON) were 11.7 ± 0.3 years of age at BL; all were ≤6 months postmenarche and did not differ in bone growth over the course of the intervention (p 0.05). However, in a subanalysis limited to subjects who were T2 or T3 at BL (n = 21 CON, n = 17 HI), T2 HI had greater gains in NN width (p = 0.01) compared to T2 CON, while T3 HI had greater gains in L3 BMD (p = 0.03) compared to T3 CON.In a group of T2 and T3 sixth-grade girls, a school-based resistance-training intervention produced maturity-specific differential gains for HI versus CON at the hip and spine.

Recent advance in biotechnology and its wide applications have led to the generation of many high-dimensional gene expression data sets that can be used to address similar biological questions. Meta-analysis plays an important role in summarizing and synthesizing scientific evidence from multiple studies. When the dimensions of datasets are high, it is desirable to incorporate variable selection into meta-analysis to improve model interpretation and prediction. According to our knowledge, all existing methods conduct variable selection with meta-analyzed data in an "all-in-or-all-out" fashion, that is, a gene is either selected in all of studies or not selected in any study. However, due to data heterogeneity commonly exist in meta-analyzed data, including choices of biospecimens, study population, and measurement sensitivity, it is possible that a gene is important in some studies while unimportant in others. In this article, we propose a novel method called meta-lasso for variable selection with high-dimensional meta-analyzed data. Through a hierarchical decomposition on regression coefficients, our method not only borrows strength across multiple data sets to boost the power to identify important genes, but also keeps the selection flexibility among data sets to take into account data heterogeneity. We show that our method possesses the gene selection consistency, that is, when sample size of each data set is large, with high probability, our method can identify all important genes and remove all unimportant genes. Simulation studies demonstrate a good performance of our method. We applied our meta-lasso method to a meta-analysis of five cardiovascular studies. The analysis results are clinically meaningful.

Background: Racial disparities in blood pressure (BP) control persist, but whether differences by race in antihypertensive medication intensification (AMI) contribute is unknown. Objective: To compare AMI by race for patients with elevated home BP readings. Methods: This prospective cohort study followed adult patients from 6 rural primary care practices who used home BP monitoring (HBPM) and recorded/reported values. For providers, AMI was encouraged when mean HBPM systolic blood pressure (SBP) values were ⩾135 mm Hg; patients received phone-based coaching on HBPM technique and sharing HBPM findings. AMI was assessed between baseline and 12 months using defined daily dose (DDD) and summed to create a total antihypertensive DDD value. Results: A total of 217 patients (mean age = 61.4 ± 10.2 years; 66% female; 57% black) provided usable HBPM data. Among 90 (41%) intensification-eligible hypertensive patients (ie, mean HBPM SBP values for 6-months ⩾135 mm Hg), mean total antihypertensive DDD was increased in 61% at 12 months. Blacks had significantly higher mean DDD at baseline and 12 months, but intensification (+0.72 vs +0.65; P = 0.83) was similar by race. However, intensification was greater in males than females (+1.1 vs +0.39; P = 0.031). Reduction in mean SBP following intensification was greater in white versus black patients (−8.2 vs −3.9 mm Hg; P = 0.14). Conclusion/Relevance: Treatment intensification in HBPM users was similar by race, differed significantly by gender, and may produce a greater response in white patients. Differential AMI in HBPM users does not appear to contribute to persistent racial disparities in BP control.