How Many Valid Days Are Necessary To Assess Physical Activity Data From Accelerometry During Pregnancy?

Background: The authors examined whether or not <= 3 days wearing Actical (R) accelerometers provided acceptable results in comparison with the recommendation of >= 4 days in women across gestation. Methods: A total of 26, 76, and 57 participants at early, mid, and late pregnancy, respectively, were assessed. Participants were instructed to wear the device for 7 days and women who wore it for >= 4 days were included. For each participant, 3, 2, and 1 day(s) were randomly selected. Paired comparisons, intraclass correlations coefficients, and kappa statistics were performed for >= 4 days (criterion) versus 3, 2, and 1 day(s). Averages (in minutes per day) of sedentary time, light, moderate, vigorous, moderate to vigorous physical activity (PA) and steps per day were examined. Results: When 3 valid days were compared with the criterion, no significant differences were found for any gestational period. The intraclass correlations coefficients were "high" for all PA-related variables. The k values varied from.819 to.838 across pregnancy ("strong"). Two and 1 valid day(s) versus the criterion showed significant differences in some PA intensities, reduced intraclass correlations coefficients, "moderate" k values for 2 valid days (.638-.788) and "minimal-to-moderate" k values for 1 valid day (.367-.755). Conclusion: In pregnant women during early, mid, and late pregnancy, PA data obtained from 3 valid days of wear was equivalent and agreed with >= 4 valid days.