Revisiting The ACSM Metabolic Equation For Walking

Abstract

The ACSM has long published a metabolic equation using walking speed and grade to predict oxygen consumption. The small homogeneous sample (n=3, trained men) used to derive the speed component of this equation calls into question its generalizability. Further, the equation’s free-living application is limited by the difficulty of measuring speed. Conversely, walking cadence (steps/min) is a practical and measurable metric that has been shown to be a reasonable proxy for walking intensity. PURPOSE: To develop a metabolic equation using cadence to predict oxygen consumption (VO2; mL/kg/min) during level walking (and compare its predictive accuracy to that of the ACSM metabolic equation) in a large sample of men and women aged 21-40 years. METHODS: Sixty-nine adults (52% women, mean±SD age=30.0±5.6 years, BMI=24.6±3.3 kg/m2) completed 5-min treadmill bouts separated by 2-min rest at four speeds: 53.6, 67.1, 80.5, and 93.9 m/min (2.0, 2.5, 3.0, 3.5 mph). The cadence-VO2 relationship was quantified with a quadratic model of best fit, producing the cadence metabolic equation. For an unbiased evaluation of this equation, leave one out cross-validation (LOOCV) was then performed and the root mean square error (RMSE) was calculated. The ACSM metabolic equation for walking was then applied to these data for comparison, and its predictive accuracy was evaluated by determining its RMSE. The bias of both metabolic equations was also calculated. RESULTS: The cadence metabolic equation was [VO (mL/kg/min) = 0.0021*C2 – 0.24*C + 15.4, where C2 = cadence]. The RMSE [±95% CI] from the LOOCV of the cadence metabolic equation was 2.5 [±1.0] mL/kg/min and its bias [±95% CI] was 0.6 [±0.3] mL/kg/min. The RMSE from applying the ACSM walking metabolic equation to this data was 3.1 [±1.2] mL/kg/min, with a bias of -2.5 [±0.2] mL/kg/min. CONCLUSION: In the same way that speed is used in the ACSM metabolic equation, cadence may also be used in a walking metabolic equation with similar error and reduced bias. The greater sample size and sex distribution used herein to develop this cadence-based metabolic equation suggests greater potential to produce accurate and generalizable estimations. Future research should test this equation during overground walking and incorporate grade as an additional variable.