Association of gait speed with risk of diabetes mellitus among older adults: findings from the China health and retirement longitudinal study

Gait speed, a simple and reliable measure of physical function, has been associated with various health outcomes. However, limited evidence exists on the relationship between gait speed and diabetes mellitus (DM) risk among older adults in China. We included 3,241 older adult participants from the 2011–2020 China Health and Retirement Longitudinal Study (CHARLS). Gait speed was assessed by a 4-meter walking test at baseline (2011–2012). Incident DM events were recorded during follow-up until 2020. Cox proportional hazards models were used to estimate hazard ratio (HR) and 95% confidence interval (95% CI) for the association between gait speed, including per 1-unit or 1-standard difference increase and tertile group, and DM risk, adjusting for potential confounders. During a median follow-up of 9.0 years, 347 incident DM events were documented. In multivariable-adjusted analyses that adjusted for sociodemographic, lifestyle variables, and health status, compared to participants in the lowest tertile of gait speed (< 1.16 m/s), participants in the middle (1.16–1.51 m/s; adjusted HR: 0.791; 95% CI: 0.609–1.026) and highest (> 1.51 m/s; adjusted HR: 0.845; 95% CI: 0.647–1.102) tertile had a decreased risk of DM. Each per standard deviation increase in gait speed was associated with a 1.1% lower risk of DM (adjusted HR: 0.989; 95% CI: 0.885–1.106). In the restricted cubic spline regression, although a potential U-shaped pattern was observed, it was not statistically significant (P for non-linearity = 0.090). Although faster gait speed showed a potential association with decreased risk of DM in older adults, the observed association did not reach statistical significance.