When analysing physical activity (PA) levels using accelerometry, the epoch setting is critical to capture intensity-specific PA correctly. The aim of the present study was to investigate the PA intensity signatures related to metabolic health in children using different epoch settings. A sample of 841 Norwegian children (age 10.2 ± 0.3 years; BMI 18.0 ± 3.0; 50% boys) provided data on accelerometry (ActiGraph GT3X+) and several indices of metabolic health (aerobic fitness, abdominal fatness, insulin sensitivity, lipid metabolism, blood pressure) that were used to create a composite metabolic health score. We created intensity spectra from 0-99 to ≥ 10000 counts per minute (cpm) for files aggregated using 1, 10, and 60-second epoch periods and used multivariate pattern analysis to analyse the data. The association patterns with metabolic health differed substantially between epoch settings. The intensity intervals most strongly associated with metabolic health were 7000-8000 cpm for data analysed using 1-second epoch, 5500-6500 cpm for data analysed using 10-second epoch, and 4000-5000 cpm analysed using 60-second epoch. Aggregation of data over different epoch periods has a clear impact on how PA intensities in the moderate and vigorous range are associated with childhood metabolic health.
Keywords: Multivariate analysis; accelerometry; child; intensity; risk factors.