High dimensional regression with survey design: an application to the association between physical activity and mortality in NHANES

Document Type

Presentation Abstract

Presentation Date

8-31-2020

Abstract

Understanding the mutual interactions between physical activity and health is crucial for developing clinical and public health intervention. Unfortunately, physical activity data used in public health research is often collected using self- or proxy-reports, which provide crude summaries of physical activity and are subject to substantial recollection, cognitive, and social desirability biases. To address this problem, a growing number of studies use accelerometers to objectively quantify physical activity in the free-living environment. Currently, the National Health Examination Study (NHANES) is the largest US-population study that contains publicly available physical activity data obtained from wearable accelerometers. Analyzing the association between physical activity in NHANES and mortality raises important and practical methodological challenges: (1) physical activity data in NHANES is high dimensional with one observation per minute for up to seven days for each study participant; (2) NHANES study participants are recruited form the US population according to a survey design procedure; and (3) the structure of the association between physical activity trajectories and outcomes is a-priori unknown. In this talk I will describe the problem of predicting health outcomes (five-year all-cause mortality) using high dimensional data (minute-level accelerometry summaries) while accounting for survey design and weights. I will also provide an easy to use tutorial for starting working with the NHANES physical activity data.

Additional Details

August 31, 2020 at 3:00 p.m. via Zoom

This document is currently not available here.

Share

COinS