Extracting Objective Estimates of Sedentary Behavior from Accelerometer Data: Measurement Considerations for Surveillance and Research Applications

Accelerometer-based activity monitors are widely used in research and surveillance applications for quantifying sedentary behavior (SB) and physical activity (PA). Considerable research has been done to refine methods for assessing PA, but relatively little attention has been given to operationalizing SB parameters (i.e., sedentary time and breaks) from accelerometer data - particularly in relation to health outcomes. This study investigated: (a) the accrued patterns of sedentary time and breaks; and (b) the associations of sedentary time and breaks in different bout durations with cardiovascular risk factors.Accelerometer data on 5,917 adults from the National Health Examination and Nutrition Survey (NHANES) 2003–2006 were used. Sedentary time and breaks at different bout durations (i.e., 1, 2–4, 5–9, 10–14, 15–19, 20–24, 25–29, and ≥30-min) were obtained using a threshold of <100 counts per minute. Sedentary time and breaks were regressed on cardiovascular risk factors (waist circumference, triglyceride, and high-density lipoprotein cholesterol) and body mass index across bout durations.The results revealed that the majority of sedentary time occurred within relatively short bout durations (≈70% and ≈85% for <5-min and