Socioeconomic status, race/ethnicity, and physical activity as measured by accelerometry: Which direction does the social gradient go and for whom?

Higher socioeconomic status (SES) is associated with better health, and physical activity is one potential contributing factor. We examined SES gradients in relation to objectively measured physical activity from accelerometry.

Data come from the 2003-2006 National Health and Nutrition Examination Survey, including activity measured using Actigraph model 7164 uniaxial accelerometers. We analyzed weekday data from participants ages 20-64 who wore accelerometers for >=4 days. The cutpoint for moderate to vigorous physical activity (MVPA) was 2020 counts per minute (CPM). Regression-adjusted means were calculated for active and sedentary time using SES predictors (education and income), stratifying by race/ethnicity within gender.

Higher SES was related to greater sedentary time for every group (P<0.001 for all). Less educated Hispanic and black non-Hispanic men had higher average CPM/day than more educated men (P<0.001 for both), as did less educated black non-Hispanic women (P<0.001). Lower income Hispanic men had greater average CPM than higher income Hispanic men (P<0.001). Only white women showed a positive association between activity – as measured by MVPA - and SES (P<0.001 for education and income); this trend was reversed in Hispanic men by income (P=0.002).

Higher SES is associated with greater sedentary time. Only higher SES non-Hispanic white women showed a positive association between SES and MVPA. For non-Hispanic black men and women and for Hispanic men, the classic social gradient works in the opposite direction: those with lower SES are more active, suggesting research should explore how SES gradients in health may be determined by factors besides physical activity.