Falls in people with diabetes and and peripheral neuropathy and their association with physical activity and gait quality

Introduction: Falls are a major clinical problem, and frequently associated with sensory and motor disturbances in the foot and ankle, such as peripheral neuropathy and altered gait biomechanics. In people with diabetes mellitus, peripheral neuropathy is a predictor of falls. However, within this population, it is unknown which characteristics are associated with falls. Our aim was to determine fall incidence and its association with physical activity and gait quality characteristics in people with diabetes and peripheral neuropathy.

Methods: In a prospective longitudinal observational study, we included 43 participants with diabetes and neuropathy (9 females,  mean age 64years (SD:8.9), BMI 30kg/m2 (SD:5.7)) and followed them for 6 months. Falls were reported in a fall diary. Physical activity (e.g. steps, daily standing time) and gait quality (e.g. walking speed, stride length, gait pattern) were measured using a tri-axial accelerometer, worn for seven consecutive days. We tested associations between falls and independent variables using forward multivariate logistic regression analysis.

Results: During 6 months follow-up, 17 participants (40%) reported a total 23 falls. Fallers had a lower number of lying episodes per day than non-fallers (fallers: 8.8 (SD:5.1), non-fallers: 12.8 (SD:5.6)), significantly associated with falls in multivariate analysis (p=0.02). Compared to non-fallers, fallers were more often female (fallers: 35%, non-fallers: 21%), had lower walking speed (fallers: 0.8m/s (SD:0.2), non-fallers: 0.9m/s (SD:0.2)) and more consistent gait pattern (power spectral density: fallers: 0.6 (SD:0.2) vs. non-fallers: 0.5 (SD:0.1)). However, none of these characteristics were significantly associated with falls in multivariate analysis.

Conclusions: 40% of people with diabetes-related peripheral neuropathy had a fall incident during 6 months follow-up. Fewer daily lying episodes was the only characteristic statistically significantly associated with falls, suggesting limited rest or greater exposure to weight-bearing activity as underlying reasons. Gait quality characteristics did not discriminate within this high-risk population.