Protocol for Objective Measurement of Infants' Physical Activity using Accelerometry

Purpose This study aimed to evaluate methods for infants' physical activity measurement based on accelerometry, including the minimum number of measurement days and placement of a wrist or ankle device. We also evaluated the acceptability of the device among infants and mothers. Methods A cross-sectional mixed-methods study was conducted on a convenience sample of 90 infants. Physical activity was measured using the Actigraph GT3X+ accelerometer placed on the wrist and/or ankle for 7 consecutive days (worn for 24 h), and a qualitative interview was performed to verify acceptability. The intraclass correlation coefficient (ICC) method and the Bland and Altman's dispersion diagram were used to verify the minimum number of measurement days. All analyses were stratified by walking status. Results The mean (SD) age was 12.9 (1.70) months; the mean acceleration varied between 25.8 mg (95% confidence interval (CI), 14.3-52.7) and 27.3 mg (95% CI, 17.9-44.5) using the wrist placement, and between 24.9 mg (95% CI, 10.6-48.4) and 26.2 mg (95% CI, 11.7-65.6) using the ankle placement. The ICC results showed a lower acceleration variability between days among infants incapable of walking; they achieved an ICC of 0.80 with 1 d of measurement in both placements. Among those capable of walking, the minimum number of days to achieve an ICC of 0.80 was 2 d measured at the wrist (0.85; 95% CI, 0.71-0.93) and 3 d measured at the ankle (0.92; 95% CI, 0.84-0.96). The qualitative results pointed to the wrist placement as the preferred placement among the overall sample. Conclusions Two and three measurement days with the accelerometer placed on the wrist and ankle, respectively, seemed to adequately represent a week of measurement. The accelerometer placed on the wrist had better acceptance by the infants and mothers.