Accelerometer Cut Points For Adults With Down Syndrome

Abstract

Past research has indicated that the relationship between energy expenditure and accelerometer output is different between adults with and without Down syndrome (DS). This suggests a need for DS-specific cut points for determining levels of sedentary behavior and physical activity from accelerometer output for adults with DS. PURPOSE: To develop accelerometer output cut points for sedentary behavior and moderate and vigorous intensity physical activity for adults with DS. METHODS: Sixteen adults with DS (10 men & 6 women; age 31 ± 15 years) performed 12 tasks each lasting 6 min: sitting; playing app on tablet; drawing; folding clothes; sweeping; fitness circuit; moving a box; basketball; standing; and walking at the preferred speed and at 0.8 and 1.4 m.s-1. We measured the rate of oxygen uptake with portable indirect calorimetry (K4b2, Cosmed) and expressed it in Metabolic Equivalents (METs). Output from a triaxial accelerometer (wGT3X-BT, Actigraph) worn on the non-dominant hip was determined as Vector Magnitude. Receiver Operating Characteristic (ROC) curves were used to identify Vector Magnitude cut points for sedentary behavior and moderate (3.0 - 5.99 METs) and vigorous (≥6 METs) intensity physical activity. Overall performance of classification models was assessed with the area under the ROC curve. Optimal cut points maximizing sensitivity and specificity were selected based on Youden’s index. RESULTS: Area under the ROC curve was high for all models: (a) sedentary behavior (0.96; 95% CI: 0.93 - 0.98); (b) moderate intensity physical activity (0.92; 95% CI: 0.88 - 0.96); and (c) vigorous intensity physical activity (0.92; 95% CI: 0.85 - 0.99). The optimal Vector Magnitude cut points were: (a) sedentary behavior ≤236 counts·min-1 (sensitivity 0.98; specificity 0.90; Youden’s index 0.88); (b) moderate-intensity physical activity ≤2167 counts·min-1 (sensitivity 0.99; specificity 0.82; Youden’s index 0.81); and (c) vigorous-intensity physical activity ≥4200 counts·min-1 (sensitivity 1.00; specificity 0.84; Youden’s index 0.84). CONCLUSION: This study offers the first DS-specific accelerometer output cut-points for classifying sedentary behavior and intensity of physical activity in adults with DS. Overall classification accuracy was excellent. Supported by NIH Grant R15HD098660