Biopsychosocial factors that classify meeting/exceeding minimally-protective physical activity dose for depression among older adults: Results from The Irish Longitudinal Study on Ageing

The authors have previously shown that physical activity doses as low as 100 min of moderate-intensity activity per week were associated with 43 % lower odds of depression across 10 years among older adults from The Irish Longitudinal Study on Ageing (TILDA). Identifying biopsychosocial profiles associated with these protective doses is critical to optimizing activity prescription for mental health. Herein, we identified biopsychosocial elements that classify individuals as meeting or not meeting minimally-protective activity dose for depression among 6747 older adults from TILDA. Eighty-four biopsychosocial variables of logical, practical and/or prior empirical relation to physical activity, depression, and/or their association were considered. Feature selection was performed using Chi-square and planned contrasts to identify differences between groups that did not meet and those that met or exceeded the optimal dose. Thirty-nine variables were retained for random forest modelling, which correctly identified 63.5 % of cases. The model performed worse in classifying participants who failed to meet the optimal dose (41 % misclassification; n = 1348) compared to those who met/exceeded the optimal dose (32 % misclassification; n = 1116). Grip strength and number of physical limitations displayed the strongest influence in model classification, with mean decreased accuracy of 26.3 % and 24.1 %, respectively. Collective expression of mostly physical function-related characteristics appears a hallmark of those meeting/exceeding the minimally-protective activity dose for depression. However, heterogeneity in these features among those not meeting optimal dose made classification prone to error, reinforcing the challenges associated with precision health initiatives to promote physical activity for mental health among older adults.