Move More Beneficially: Physical Activity Variability as a Novel Metric of Physical Activity Pattern and an Independent Predictor of Mortality and Chronic Disease Incidence

Current physical activity guidelines emphasize the total activity volume of moderate-to-vigorous physical activity (MVPA) for health benefits, but the implications of physical activity variability (PAV) on health outcomes remain unclear. This study aimed to investigate the association of accelerometer-measured PAV, defined as daily fluctuations in physical activity levels, with mortality and chronic disease incidence. Two cohorts were analyzed: the UK Biobank (2013-2015) with follow-up through November 2022, and NHANES (2003-2006) with follow-up through December 2019. PAV, calculated as the coefficient of variation of hourly physical activity, represents high variability (peaks and troughs) versus low variability (stability). Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals for all-cause mortality, cardiovascular disease (CVD) and cancer mortality, as well as incidence of CVD, type 2 diabetes mellitus (T2DM), and chronic obstructive pulmonary disease (COPD), adjusting for total activity volume (i.e., average acceleration) and important confounders. In the UK Biobank, individuals in the highest PAV quartile had lower risks of all-cause mortality (28%, 21%-35%), CVD mortality (42%, 24%-56%), and incidences of CVD (19%, 12%-25%), T2DM (37%, 26%-46%), and COPD (45%, 35%-54%). In NHANES, the highest PAV quartile was associated with 23% (7%-36%) and 50% (18%-69%) lower risks of all-cause and CVD mortality. PAV was inversely associated with risks of mortality and chronic disease, independent of total volume of physical activity. The health implications of PAV warrant further investigation.