Background/Purpose: Knee osteoarthritis (OA) is associated with mortality, and walking difficulty is known to mediate this relationship. However, little is known about what specific aspects of walking difficulty increase mortality risk. We investigated the relationship of walking speed, as a measure of walking difficulty, to mortality and studied what thresholds that best discriminate risk of mortality in people with knee OA. Since knee OA frequently occurs in mid-life, we additionally investigated if walking speed is a relevant predictor of mortality and studied the thresholds that best discriminate the mortality risk in younger adults (aged < 65 years) with knee OA.

Methods: We used data from the Osteoarthritis Initiative (OAI). Walking speed was measured via the 20-meter (m) walk test. We quantified time to mortality over 9 years, i.e., until the 108-month OAI visit. To examine the association of walking speed with mortality over 9 years, we calculated hazard ratios (HR) and 95% confidence intervals (CI) from the Cox regression model adjusted for potential confounders. To identify an optimal threshold of walking speed predictive of mortality risk, we used the Maximal Likelihood Ratio Chi-square Approach. This approach selects the threshold that is most concordant with mortality risk over the observed follow-up times, a metric similar to maximizing Youden index when using a receiver operating curve method. Specifically, we ran adjusted Cox regression models for different thresholds of the walking speed iteratively. We then identified the optimal threshold from the model that yielded the maximal chi-square value (Table). We repeated these analyses separately for younger adults with knee OA.

Results: Of 4775 participants who completed the 20-m walk at baseline (age [mean (± sd)] 61.1 ± 9.2 years, BMI 28.6 ± 4.8 kg/m², 59% female), 6.5% (312/4755) died over 9 years.  For younger adults, 3.0% (88/2966) died. Walking 0.1 meters/second (m/s) slower during the 20-m walk was associated with 16% [Adjusted HR 1.16, 95% CI (1.09, 1.23)] and 14% [Adjusted HR 1.14, 95% CI (1.02, 1.28)] higher risk of  mortality over 9 years in the overall analytic sample and younger adults, respectively. Walking slower than 1.2 m/s was found to be an optimal threshold to best discriminate those with and without mortality for both the entire cohort, and among younger adults (Table).

Conclusion: Walking speed, an objective measure of walking difficulty is associated with mortality in knee OA. Specifically, walking slower than 1.2 m/s during a 20-m walk test may identify the risk of mortality in people with OA. These same thresholds also apply to younger adults with knee OA. Thus, walking speed may be a simple indicator of health in adults with knee OA. Health care professionals may use walking speed to assess expected health and tailor goals of care for knee OA population.

« Back to ACR/ARP Annual Meeting 2019

ACR Meeting Abstracts - https://acrabstracts.org/abstract/optimal-threshold-of-walking-speed-predictive-of-mortality-risk-over-9-years-in-knee-osteoarthritis-data-from-osteoarthritis-initiative/