Identification of risk factors underlying a decline in gait speed in older individuals

Abstract Introduction Declining gait speed is strongly associated with future long-term care (LTC) dependency in older adults; however, its risk factors remain unclear. This study aimed to identify predictors of gait speed decline and develop a simple risk scoring system. Materials and methods A total of 1,577 community-dwelling adults aged ≥65 years were followed for 6 years. Of the 725 participants followed, 37 with baseline gait speed <1 m/s were excluded, leaving 688 for cohort analysis. Among these, 98 baseline factors were initially examined, and five predictors were selected. Complete data were available for 681 participants (91 with and 590 without gait speed decline). The five predictors were entered simultaneously into a multivariable logistic regression model, and a scoring system was developed based on the regression coefficients. Results The five predictors were age, pain/discomfort, mobility problems, exercise <3 days/week, and not feeling full of energy. The apparent AUC of the logistic model was 0.735 (95% CI 0.678–0.792), and the optimism-corrected AUC after 1,000 bootstrap resamples was 0.714. A revised integer score based on the mutually adjusted coefficients ranged from 0 to 18 points. A cutoff of ≥6 points yielded a sensitivity of 0.505, specificity of 0.847, positive predictive value of 0.338, negative predictive value of 0.918, and score AUC of 0.736 (95% CI 0.676–0.795). Conclusions We developed a simple scoring system to predict gait speed decline in older adults. Currently, it should be considered an exploratory risk stratification tool; external validation may support its future clinical use.