Accelerometer-derived physical activity patterns and incident type 2 diabetes: a prospective cohort study

Background Emerging evidence suggests a significant relationship between the duration of physical activity (PA) and the incidence of type 2 diabetes (T2D). However, the association between the "weekend warrior" (WW) pattern-characterized by concentrated moderate-to-vigorous PA (MVPA) over one to two days-and T2D remains unclear. Methods This prospective cohort study aims to utilize device-measured PA data to investigate the associations between PA patterns and T2D. Individuals were divided into three MVPA patterns on the basis of WHO guidelines: inactive (< 150 min), active WW (>= 150 min with >= 50% of total MVPA achieved in one to two days), and active regular (>= 150 min but not active WW). These patterns were also evaluated using sample percentile thresholds. The relationships between PA patterns and the risk of T2D were analysed employing Cox proportional hazards models. Results A total of 1972 participants developed T2D over a 7.9-year median follow-up period. In the fully adjusted model, both active patterns demonstrated comparable reductions in the risk of developing T2D (active WW: hazard ratio [HR] 0.64, 95% confidence interval [CI] 0.58-0.71; active regular: 0.56, 0.49-0.64). Moreover, the risk of T2D exhibited a progressive decline as the duration of MVPA increased across both active patterns. Conclusions Engaging in MVPA for one or two days per week provides comparable protective benefits against the incidence of T2D as more evenly distributed PA. Additionally, exceeding the current guidelines may confer even greater advantages.