Stride Length Predicts Adverse Clinical Events in Older Adults: A Systematic Review and Meta-Analysis.

Abstract

Background: This meta-analysis aims to estimate the power of walking stride length as a predictor of adverse clinical events in older adults. Methods: We searched all electronic databases until April 2021 for studies reporting stride length and other spatial gait parameters, including stride velocity, stride width, step width and stride variability, and compared them with clinical outcomes in the elderly. Meta-analyses of odds ratios (ORs) of effects of stride length on clinical outcomes used the generic inverse variance method and random model effects. Clinical outcomes were major adverse events (MAEs), physical disability and mortality. Results: Eleven cohort studies with 14,167 patients (mean age 75.4 ± 5.6 years, 55.8% female) were included in the analysis. At 33.05 months follow up, 3839 (27%) patients had clinical adverse events. Baseline stride length was shorter, WMD −0.15 (−0.19 to −0.11, p < 0.001), and stride length variability was higher, WMD 0.67 (0.33 to 1.01, p < 0.001), in fallers compared to non-fallers. Other gait parameters were not different between the two groups (p > 0.05 for all). Short stride length predicted MAE OR 1.36 (95% CI; 1.19 to 1.55, p < 0.001), physical disability OR 1.26 (95% CI; 1.11 to 1.44, p = 0.004) and mortality OR 1.69 (95% CI; 1.41 to 2.02, p < 0.001). A baseline normalized stride length ≤ 0.64 m was more accurate in predicting adverse clinical events, with summary sensitivity 65% (58–71%), specificity 72% (69–75%) and accuracy 75.5% (74.2–76.7%) compared to stride length variability 5.7%, with summary sensitivity 66% (61–70%), specificity 56% (54–58%) and accuracy 57.1% (55.5–58.6%). Conclusion: The results of this meta-analyses support the significant value of stride length for predicting life-threatening clinical events in older adults. A short stride length of ≤0.64 m accurately predicted clinical events, over and above other gait measures.