Associations between minute-level smartphone GPS-derived exposure to greenness and consumer wearable-derived physical activity in the Nurses’ Health Study 3

Abstract

BACKGROUND AND AIM: Increased exposure to greenness has been linked to increased physical activity and improved health outcomes, including better mental health and lower cardiovascular disease and mortality rates. However, most studies measure greenness exposure surrounding residential addresses, and do not account for each study participants’ movements. We examine minute-level smartphone GPS-derived exposure to greenness and consumer wearable-derived physical activity. METHODS: We assessed daily GPS-derived greenness exposure using wearable accelerometer data from participants in the US-based prospective Nurses’ Health Study 3 cohort mHealth Substudy (2018-2020). The mHealth substudy consisted of 348 female participants (mean age 35.7) who undertook 7-day sampling periods, four times across a year, to capture seasonal variability in greenness and physical activity. Normalized Difference Vegetation Index data were derived from 30m resolution Landsat imagery, which was spatially joined to the GPS points that were recorded every 10 minutes during sampling periods. Fitbit proprietary algorithms provided physical activity summarized as mean number of steps per minute for each 10-minute period. Generalized linear and additive mixed models were utilized to examine near momentary associations between greenness and physical activity adjusting for socioeconomic status, season, age, and smoking. RESULTS:Across study periods mean 10-minute step count ranged from 0-111.57 steps and 10-minute NDVI ranged from 0 to 0.71. Summer months had higher values and large ranges due to seasonal and individual-level mobility changes. Contrary to our hypotheses, higher non-linear greenness exposure was associated with lower mean steps per minute. Participants in the top greenness quintiles took 2.25 (95% CI 2.42, 2.08) fewer steps per 10-minute average compared to those with the lowest greenness quintile, after adjusting for confounders. CONCLUSIONS:We utilized objective data at fine temporal and spatial scales to present novel estimates on the real time association between greenness and physical activity. Increases in greenness were associated with small decreases in steps per minute. KEYWORDS: Activity Space, Green Space, Physical Activity, Mixed Models