Gait Speed as a Predictor for Diabetes Incidence in People with or at Risk of Knee Osteoarthritis: A Longitudinal Analysis from the Osteoarthritis Initiative

Abstract

Background: This study examined the association between baseline gait speed with incident diabetes mellitus (DM) among people with or at elevated risk for knee OA. Materials and Methods: Participants from the Osteoarthritis Initiative, aged 45 to 79 years, where included. Participants with or at risk of knee OA from baseline to the 96-month visit were included. Participants with self-reported DM at baseline were excluded. DM incidence was followed over the 4-time points. Gait speed was measured at baseline using a 20-m walk test. Generalized estimating equations with logistic regression were utilized for analyses. Receiver operator characteristic curves and area under the curve were used to determine the cutoff score for baseline speed. Results: Of the 4313 participants included in the analyses (58.7% females), 301 participants had a cumulative incidence of DM of 7.0% during follow-up. Decreased gait speed was a significant predictor of incident DM (RR 0.44, p = 0.018). The threshold for baseline gait speed that predicted incident DM was 1.32 m/s with an area under the curve of 0.59 (p < 0.001). Conclusions: Baseline gait speed could be an important screening tool for identifying people at risk of incident diabetes, and the determined cutoff value for gait speed should be examined in future research.