Moderators Of The Relationship Between Worksite Walkability And Physical Activity

Abstract

PURPOSE: This study examined traditional and novel moderators (i.e., age, sex, race, income, # of children <18 yrs, ratio of cars to drivers, home walkability) of the relationship between worksite neighborhood walkability (WNW) and physical activity (PA). METHODS: Participants (n = 512, mean 44 yrs, 59% female, 70% White) were employed (not at home) during the baseline phase of a larger trial. Measures included self-reported PA (total min/week, min/week of transport-PA inside (TPAin) and outside (TPAout) the home neighborhood; NPAQ) and Actigraph-measured PA (min/week moderate-to-vigorous PA in bouts [MVPA] and sedentary-light PA [SLPA]). GIS-assessed WNW (500m network buffer of residential, intersection, and transit densities; land use mix). Negative binomial regression estimated associations of WNW with PA, each moderator, and WNW x moderator (i.e., age, sex, race, income, # of children <18, car:driver ratio, home walkability) interactions. Alpha level set at 0.10 to probe conditional effects. Main effects of covariates (residence tenure, distance to work, reason moved to residence, total household members; and in accelerometer models, wear time) were included. RESULTS: Sex and # of children <18 yrs moderated relationships of WNW with MVPA and SLPA. Race moderated the relationship of WNW with TPAout. For women, the conditional effect of WNW on MVPA was positive (p = .04) while the conditional effect of WNW on SLPA was negative (p = .04); effects not different from zero for men. For adults with no children <18, the conditional effect of WNW on MVPA was positive (p = .01) while the conditional effect of WNW on SLPA was negative (p = .01); not different from zero for those with at least 1 child. For White participants, the conditional effect of WNW on TPAout was negative (p = .07); not different from zero for non-Whites. No other moderators (age, income, car:driver ratio, or home walkability) interacted with WNW. CONCLUSIONS: For women, adults without young children, and Whites (but not men, those with young children, and non-Whites), PA is influenced by walkability of worksite environments in mostly expected directions. An underexplored aspect of behavioral ecological models is identifying who is sensitive to the environmental conditions that can bolster health promotion efforts.