Gait characteristics in older adults with diabetes and impaired fasting glucose: The Rotterdam Study

Aims To investigate the association of diabetes mellitus and impaired fasting glucose with gait in the general middle-aged and elderly population.
Methods We performed a cross-sectional study on 3019 participants from the population-based Rotterdam Study (aged > 45 years, 54% women). The presence of diabetes mellitus and impaired fasting glucose was evaluated by measuring serum glucose levels and by documenting anti-diabetic treatment. Participants underwent gait analysis using an electronic walkway. Thirty gait variables were summarized into five independent gait domains for normal walking (Rhythm, Variability, Phases, Pace and Base of Support), one for turning (Turning) and one for walking heel to toe (Tandem), which were averaged into Global Gait. Linear regression analyses were performed to determine the association of diabetes, impaired fasting glucose and continuous glucose levels within the normal range with gait.
Results Diabetes mellitus was associated with worse Global Gait (Z-score difference - 0.19, 95% confidence interval (CI) - 0.30; - 0.07), worse Pace (- 0.20, 95% CI - 0.30; - 0.10) and worse Tandem (- 0.21, 95% CI - 0.33; - 0.09), after adjusting for age, sex, height and weight. The association with Tandem remained significant after additional adjustment for cardiovascular risk factors. Impaired fasting glucose and continuous glucose levels within the normal range were not associated with any of the gait domains.
Conclusion In our population-based study diabetes mellitus was associated with worse Global Gait, which was mostly reflected in Pace and Tandem. These associations were partly driven by other cardiovascular risk factors, emphasizing the importance of optimal control of cardiovascular risk factor profiles in patients with diabetes.