Greater variability in daily physical activity is associated with poorer mental health profiles among obese adults

Abstract Research is inconclusive about whether physical activity (PA) should be performed every day or performed less frequently but in longer bouts to obtain mental health benefits. The current study examined the extent to which day-to-day variability in PA is associated with adults' mental health, and if this association differed by Body Mass Index (BMI). Adults (N = 116) completed three waves of data collection (each lasting 4 days) during which participants completed a questionnaire assessing mental health (life satisfaction, depressive symptoms, perceived stress), wore a waist accelerometer, and had height and weight measured. This study employed a novel two-stage data analysis approach using the standalone program MIXWILD. The first-stage model partitioned mean level as well as between-subject and within-subject variances in daily PA by estimating a random location (subject-level mean) and a random scale (subject-level variability) for daily PA. In the second-stage, these random subject effects for daily PA along with their interactions with BMI were used as predictors for subject-level mental health outcomes. Associations between subject-level variability in daily PA and mental health outcomes significantly differed depending on adults’ BMI (life satisfaction: β = −0.05, p