Physical Activity And Its Association With Other Wellness And Risk Behaviors From A College Sample

Abstract

PURPOSE: The University of Vermont Wellness Environment App Study is an app-based longitudinal research study focused on promoting health and reducing risky behaviors in a college student population. The mobile ecological momentary assessment over an academic year provides a unique opportunity to accomplish two aims in this study: 1) to assess physical activity (PA) variation across days of the week and throughout the academic year and 2) to explore the correlates that were associated with PA. METHODS: Students who enrolled in the study were asked to report their wellness and risk behaviors on a 14-item survey through a smartphone app every day. Each student was also provided an Apple Watch to track their real-time PA. Data were collected from 805 college students, with an average of 97 days of daily surveys and steps data from Sept 2017 to early May 2018. Daily survey and step data were merged by day for each participant. General estimating equations were implemented in SAS PROC MIXED, with an autoregressive covariance structure to estimate the daily steps by demographic variables and other wellness behaviors. RESULTS: Based on average daily step counts, females were significantly more active than male college students (Female=8904 and Male=8488, p=.0082). Age-related PA decline was found from freshmen to seniors, but only freshmen was significantly more active than seniors (freshmen=8714 and senior=8023, p=.0051). Students were significantly more active (p<.0001) during the weekday (Monday to Friday ranged from 8800 to 9384 steps) than weekend (Sat=8356 and Sun=7145). Temporal patterns were also revealed that students were less active during Thankgiving, Winter, and Spring breaks. Strong correlations were found between daily steps and self-reported mood (p<.0001), sleep (p<.0001), fruit and vegetable consumption (p<.0001), water intake (p<.0001), and screen time (p<.0001). No significant associations were found for marijuana use (p=.997), cigarette use (p=.2518), drug use (p=.1546) but significant associated were observed for illicit pill use (p=.0083), alcoholic drink consumption (p<.0001), and liquor shot consumption (p<.0001). CONCLUSION: The study provides a comphrensive surveillance on longitudinal PA pattern and its association with a variety of wellness and risky behaviors in college students.