Validation of Energy Expenditure Prediction Algorithms in Adolescents and Teens Using the Actical Activity Monitor

PURPOSE This study validated algorithms for predicting activity energy expenditure (AEE = task EE – RMR, kcals/kg/min) from the output of the Actical (Mini Mitter Co., Inc., Bend, OR) electronic activity monitor in adolescents and teens. METHODS Fourteen boys (Mean±SD[Range]: 12.1±2.8[8–16] yrs) and 10 girls (12.9±2.3[10–17] yrs) visited the lab once to perform 9 activities: Typing, hand writing, video game playing, floor sweeping, carpet vacuuming, table surface dusting, and treadmill walking (67 and 80.4 m/min) and jogging (120.6 m/min). Subjects wore Actical monitors on the non-dominant wrist, on the ankle of the same side of the body, and on the right hip just posterior to the iliac crest during all activities. Actual AEE was determined using the VmaxST portable metabolic system (SensorMedics Corp., Yorba Linda, CA) which was carried in a small pack and worn on the back during all activities (equipment mass = 1.14 kg). The last two 60 sec sample averages from both the metabolic system and the Actical monitors (units of counts/min) were extracted from the raw data and used to generate separate AEE prediction algorithms (standard linear regression procedures) for each monitor location. Each algorithm included: 1) Lower Actical cutoff value (< 250 counts/min; no prediction equation applied); 2) Equation predicting AEE between the lower cutoff and an upper cutoff (1600–1800 counts/min); Equation predicting AEE higher than the upper cutoff. Predicted AEE from each algorithm was compared with actual AEE using PPM correlations, SEE, and standard residual analyses. RESULTS Algorithms for the ankle (R = 0.80, SEE = ±0.0771, P < 0.001), hip (R = 0.89, SEE = ±0.0587, P < 0.001), and wrist (R = 0.89, SEE = ±0.0592, P < 0.001) monitors all provided accurate predictions of AEE. CONCLUSIONS Though the hip and wrist locations appear to be most promising, further research should verify this in field settings. This study was funded by a grant from the Mini Mitter Co., Bend, OR.