Accuracy of Direct Observation to Assess Physical Activity in Older Adults.

The purpose of the study was to evaluate the accuracy of direct observation (DO) to estimate MET level and intensity category during laboratory-based and free-living activity in older adults. Older adults engaged in unstructured laboratory and free-living activity. Participants wore a portable metabolic system to measure energy expenditure and were directly observed. DO recorded MET-level point estimates. 32,401 in-laboratory and 87,715 free-living data points (9 participants, 67% male, 71.0 ± 6.9 years, 27.1 ± 4.3 kg·m) were included in final analysis. Results revealed 45.4% of in-laboratory and 61.1% of free-living mean DO activities fell within 0.5 METs of the measured MET values. DO accurately classified intensity category 45.0% of the time in-laboratory and 50.9% of free-living observations. DO-estimated activity cost resulted in low point estimate accuracy however there was low variability between the mean measured and estimated METs. This suggests, dependent on the desired outcome, DO could provide a viable option for activity assessment, however, the low point estimate accuracy presents a need for further research to continue to refine the approach to increase accuracy.