Compliance with objective physical activity monitoring: Considerations for public health surveillance and program evaluation

Background: Pedometers and accelerometers are increasingly used to objectively measure physical activity (PA) participation in public health related research. The utility of objective PA measures depends upon participant compliance with established wearing protocols. Purpose: To describe wearing compliance of accelerometers in the 2003-2004 National Health and Nutrition Examination Survey. Methods: All participants aged ≥6 years were asked to wear an accelerometer for 7 days from the time they woke up until going to bed. A minimum wear time of 10 hrs/day was required to constitute a valid wear day. Compliance data (for number of days worn) is summarized for 6329 participants whose accelerometers were in calibration after being returned via mail. Results: Cross tabulations of demographic variables and days worn for age, race-ethnicity and sex categories display the influence of these demographics on compliance. Overall, wearing compliance for all 7 days and any 4 days were 28.4% and 76.8%, respectively. Compliance increased with age; compliance on any 4 days ranged from 68.6% for adolescents aged 12-19 to 87.7% among participant ≥60 years. Compliance by sex was similar but slightly higher among males (~1-5%). Compliance on any 4 days by race-ethnicity was similar between non-Hispanic whites and Mexican Americans (79.7% vs. 78.9%), and lowest among non-Hispanic blacks (70.4%). Conclusions: Results inform the design of research projects that may include objective measurement of PA. Compliance differences across population demographics, namely age and race-ethnicity, warrant additional study design consideration to obtain adequate samples for specific subgroups.