Social epidemiology of Fitbit daily steps in early adolescence

Sociodemographic disparities in adolescent physical activity have been documented but mostly rely on self-reported data. Our objective was to examine differences in device-based step metrics, including daily step count (steps d^-1), by sociodemographic factors among a diverse sample of 10-to-14-year-old adolescents in the US.

We analyzed prospective cohort data from Year 2 (2018-2020) of the Adolescent Brain Cognitive Development (ABCD) Study (N = 6460). Mixed-effects models were conducted to estimate associations of sociodemographic factors (sex, sexual orientation, race/ethnicity, household income, parental education, and parental marital status) with repeated measures of steps d^-1 over the course of 21 days.

Participants (49.6% female, 39.0% racial/ethnic minority) accumulated an average of 9095.8 steps d^-1. In mixed-effects models, 1543.6 more steps d^-1 were recorded for male versus female sex, Black versus White race (328.8 more steps d^-1), heterosexual versus sexual minority sexual orientation (676.4 more steps d^-1), >$200,000 versus <$25,000 household income (1003.3 more steps d^-1), and having married/partnered parents versus unmarried/unpartnered parents (326.3 more steps d^-1). We found effect modification by household income for Black adolescents and by sex for Asian adolescents.

Given sociodemographic differences in adolescent steps d^-1, physical activity guidelines should focus on key populations and adopt strategies optimized for adolescents from diverse backgrounds.

Sociodemographic disparities in physical activity have been documented but mostly rely on self-reported data, which can be limited by reporting and prevarication bias. In this demographically diverse sample of 10-14-year-old early adolescents in the U.S., we found notable and nuanced sociodemographic disparities in Fitbit steps per day. More daily steps were recorded for male versus female sex, Black versus White race, heterosexual versus sexual minority, >$100,000 versus <$25,000 household income, and having married/partnered versus unmarried/unpartnered parents. We found effect modification by household income for Black adolescents and by sex for Asian adolescents.