Moderating effect of individual characteristics and neighborhood greenness on the association between ambient temperature and habitual physical activity levels

Background Ambient temperature could impact physical activity behavior, as extreme temperatures may be a barrier for physical activity. Adopting a within-subject design, we examined the association between ambient temperature and habitual physical activity levels and explored how individual characteristics and neighborhood greenness influenced this relation. Methods Physical activity data was obtained for 8 consecutive days using a thigh-worn triaxial accelerometer during three measurement periods with distinct ambient temperatures: COLD (5.9 ± 0.4 °C), MODERATE (13.9 ± 3.4 °C) and WARM (26.0 ± 0.7 °C). Physical activity parameters were compared between conditions using linear mixed model analysis, with MODERATE as the reference. The moderating effects of age, sex, cardiovascular health status, baseline physical activity levels, and neighborhood greenness on the association between ambient temperature and physical activity was subsequently explored. Results Participants ( n = 151, 62 ± 11 years) had lower daily step counts in WARM (10,576 ± 3395 steps/day, p = 0.003) and COLD (10,641 ± 3483 steps/day, p < 0.001) compared to MODERATE (11,602 ± 3636 steps/day). LIPA and MVPA were lower in COLD (244 ± 62 min/day, p = 0.004, and 91 ± 30 min/day, p < 0.001, respectively) compared to MODERATE (255 ± 66 min/day and 104 ± 36 min/day, respectively). Sitting time (545 ± 89 min/day) did not differ across conditions ( p = 0.68). Baseline physical activity was an effect modifier for step count, sitting time, and MVPA, whereas neighborhood greenness was an effect modifier for sitting time. Conclusion Colder ambient temperatures were associated with reduced physical activity levels ( i.e. , lower step count, LIPA, and MVPA), while warmer ambient conditions were only associated with reduced step count. Effects were more prominent in the most active versus least active individuals, with little impact of neighborhood greenness.