Ballman, Karla V.

Objectives To analyze self‐reported changes in physical function in older women with breast cancer receiving adjuvant chemotherapy. Design Secondary analysis of the Cancer and Leukemia Group B (CALGB) 49907 prospective randomized clinical trial. Setting CALGB institutions in the United States. Participants Women aged 65 and older with Stage I to III breast cancer enrolled in CALGB 49907 who had physical function data from before and after receipt of adjuvant chemotherapy (N=256; mean age 71.5, range 65–85). Measurements Participants were administered the physical function subscale of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire before chemotherapy, at the end of chemotherapy, and 12 months after chemotherapy initiation. Functional decline was defined as a more than 10‐point decrease from baseline at each time point. Resilience was defined as return to within 10 points of baseline. Multivariable regression was used to examine pretreatment characteristics associated with physical function changes. Results Of 42% of participants who had physical function decline from before to the end of chemotherapy, 47% recovered by 12 months (were resilient). Almost one‐third experienced functional decline from before chemotherapy to 12 months later. Pretreatment fatigue was a risk factor for functional decline from before to the end of chemotherapy ( P =.02). Risk factors for functional decline at 12 months included pretreatment dyspnea ( P =.007) and being unmarried ( P= .01). Conclusion Functional decline was common in older women receiving adjuvant chemotherapy for breast cancer in a clinical trial. Although half recovered their physical function, one‐third had a clinically meaningful decline at 12 months. Strategies are needed to prevent functional decline in older adults receiving chemotherapy. J Am Geriatr Soc 67:920–927, 2019.

The clinical utility of microbiome biomarkers depends on the reliable and reproducible nature of comparative results. Underappreciation of the variation associated with common demographic, health, and behavioral factors may confound associations of interest and generate false positives. Here, we present the Midwestern Reference Panel (MWRP), a resource for comparative gut microbiome studies conducted in the Midwestern United States. We analyzed the relationships between demographic and health behavior–related factors and the microbiota in this cohort, and estimated their effect sizes. Most variables investigated were associated with the gut microbiota. Specifically, body mass index (BMI), race, sex, and alcohol use were significantly associated with microbial β-diversity (P < 0.05, unweighted UniFrac). BMI, race and alcohol use were also significantly associated with microbial α-diversity (P < 0.05, species richness). Tobacco use showed a trend toward association with the microbiota (P < 0.1, unweighted UniFrac). The effect sizes of the associations, as quantified by adjusted R 2 values based on unweighted UniFrac distances, were small (< 1% for all variables), indicating that these factors explain only a small percentage of overall microbiota variability. Nevertheless, the significant associations between these variables and the gut microbiota suggest that they could still be potential confounders in comparative studies and that controlling for these variables in study design, which is the main objective of the MWRP, is important for increasing reproducibility in comparative microbiome studies.

AbstractMotivationOne objective of human microbiome studies is to identify differentially abundant microbes across biological conditions. Previous statistical methods focus on detecting the shift in the abundance and/or prevalence of the microbes and treat the dispersion (spread of the data) as a nuisance. These methods also assume that the dispersion is the same across conditions, an assumption which may not hold in presence of sample heterogeneity. Moreover, the widespread outliers in the microbiome sequencing data make existing parametric models not overly robust. Therefore, a robust and powerful method that allows covariate-dependent dispersion and addresses outliers is still needed for differential abundance analysis.ResultsWe introduce a novel test for differential distribution analysis of microbiome sequencing data by jointly testing the abundance, prevalence and dispersion. The test is built on a zero-inflated negative binomial regression model and winsorized count data to account for zero-inflation and outliers. Using simulated data and real microbiome sequencing datasets, we show that our test is robust across various biological conditions and overall more powerful than previous methods.Availability and implementationR package is available at https://github.com/jchen1981/MicrobiomeDDA.Supplementary informationSupplementary data are available at Bioinformatics online.

Motivation Our goal was to understand why the PLIER algorithm performs so well given its derivation is based on a biologically implausible assumption. Results In spite of a non-intuitive assumption regarding the PM and MM errors made as part of the derivation for PLIER, the resulting probe level error function does capture the key characteristics of the ideal error function, assuming MM probes only measure non-specific binding and no signal.