Associations of Ultraprocessed Food Intake, Wearable-Measured Lifestyle Factors, and Genetic Susceptibility With Incident Type 2 Diabetes

Growing evidence links ultraprocessed food (UPF) intake to type 2 diabetes (T2D) risk; however, whether reducing UPF consumption is associated with lower risk across genetic susceptibility strata and whether protective lifestyle factors are associated with lower risk among high UPF consumers remain uncertain. We studied 56,964 UK Biobank participants without diabetes at baseline. Diet was assessed using 24-h dietary records, and foods were classified according to the Nova system. Physical activity and sleep were measured by accelerometers for 7 consecutive days, and polygenic risk was derived from 424 T2D-associated single nucleotide polymorphisms. Multivariable Cox proportional hazards models and restricted cubic splines were used to estimate hazard ratios (HRs) and dose-response relationships. Each 10% higher UPF intake was associated with 7% higher T2D risk (HR 1.07 [95% CI 1.03, 1.12]). Dose-response analyses revealed a linear association with UPF intake and an L-shaped association with moderate to vigorous physical activity (MVPA; Pnonlinearity < 0.001). Among individuals with high genetic susceptibility, lower risk was observed with lower UPF intake (HR 0.65 [95% CI 0.49, 0.85]), higher total physical activity (0.79 [0.65, 0.97]), sufficient MVPA (0.67 [0.55, 0.81]), and adequate sleep (0.57 [0.39, 0.84]). In joint analyses, sufficient MVPA (HR 0.69 [95% CI 0.50, 0.96]) and adequate sleep (0.42 [0.24, 0.75]) were associated with lower risk among participants with high UPF intake and high genetic risk. Together, our findings show that healthier modifiable behaviors were associated with lower diabetes risk across genetic susceptibility strata, and wearable-derived evidence highlights MVPA and adequate sleep as behavioral targets among individuals with high UPF intake. ARTICLE HIGHLIGHTS: Dietary intake, inactivity, poor sleep, and genetic risk may all contribute to diabetes risk, but their combined effects are unclear. Whether modifiable lifestyle factors are associated with diabetes risk across genetic risk strata and whether wearable-derived behaviors are associated with lower risk among individuals with high ultraprocessed food (UPF) intake remain uncertain. Lower UPF intake was associated with lower diabetes risk across all levels of genetic risk, and meeting moderate to vigorous physical activity targets and getting adequate sleep were associated with lower risk among individuals with high UPF intake. Wearable-derived behavioral targets may be practical solutions for reducing UPF-associated diabetes risk and supporting diabetes prevention.