Development of a Multivariable Equation for Predicting Healthy Step Length

Gait reflects multisystem integrity and predicts morbidity and functional decline, underscoring its growing recognition as a “sixth vital sign.” This study developed and validated a transparent multivariable regression equation to estimate healthy step length in adults, incorporating age, sex, height, weight, and habitual walking speed. Data from 252 healthy adults (ages 18–90 years) were analyzed using multiple linear regression with 5-fold cross-validation to assess predictive accuracy and model stability. The final model demonstrated strong performance (adjusted R2 = 0.83; RMSE = 2.88 cm). The resulting equation was: step length (cm) = –16.14 – 0.06·Age +0.31·Height − 0.04·Weight + 0.02·Sex +0.30·Walking speed. Individualized 95% prediction intervals defined normative ranges for clinical and research applications. This empirically derived model provides a reproducible and interpretable reference for assessing gait health, enabling early detection of deviation and supporting fall prevention, rehabilitation, and digital health monitoring across diverse adult populations.