Estimating Walking Intensity In Adults With Down Syndrome Using A Portable Accelerometer

PURPOSE: Gait and physiological characteristics of adults with Down syndrome (DS) suggest that accelerometer cut-points for moderate-to-vigorous physical activity [MVPA ≥3 metabolic equivalents (METs)] for typically developed adults may be inappropriate for adults with DS. Thus, we developed MVPA cut-points for the ActiGraph wGT3X-BT worn on the non-dominant hip and both wrists during treadmill (TM) walking in adults with DS. METHODS: Energy expenditure (METs) during the final 4 mins. of each 7 min. stage of a continuous progressive TM walking protocol (0% grade), which began at 1.5 mph and increased in 0.5 mph increments until reaching vigorous intensity (≥ 6 METs) was assessed by indirect calorimetry. METs for each participant were calculated using resting metabolic rate (RMR) assessed by indirect calorimetry prior to beginning the walking protocol. Optimal MVPA cut-points for vertical axis and vector magnitude counts/5-second (5-sec) epochs were determined as the value that provided sensitivity and specificity closest to the area under the curve (AUC) and that minimized the absolute difference between sensitivity and specificity. Confusion matrices were used to assess the sensitivity (correct positive predictions/total positive predictions), specificity (correct negative predictions/total negative predictions), accuracy (correct predictions/total predictions), and AUC. RESULTS: Seventeen adults with DS (24.6 ± 5.1 yrs., 65% female, RMR: 2.3 ± 0.4 ml/kg/min) completed this study. Vertical axis cut-points for MVPA were 115, 217 and 196 counts/5-sec for the hip, right and left wrist, respectively. Vector magnitude cut-points were 233, 395 and 340 counts/5-sec for the hip, right and left wrists, respectively. Accuracy for classification of MVPA (65.9%) was highest for the vector magnitude cut-point for the ActiGraph worn on the non-dominant hip (sensitivity: 66.1%, specificity: 65.9%, AUC: 0.66) and lowest (60.3%) for the vertical axis cut-point for the ActiGraph worn on the right wrist (sensitivity: 60.2%, specificity: 60.3%, AUC: 0.60). CONCLUSIONS: The ActiGraph vector magnitude cut-point (233 counts/5-sec) provided the highest accuracy for classifying MVPA during TM walking. Confirmation of our results in a larger sample of adults with DS during both TM and overground walking is warranted.