Objective: To determine the levels and patterns of daily physical activity in groups of normal-weight, overweight and obese adults using uniaxial minute-by-minute accelerometry.
Design: Cross-sectional study of physical activity levels over a 7 day period using accelerometers programmed to collect minute-by-minute data.
Setting: Participants were recruited from large companies in Bristol and Cardiff. All meetings took place on company premises.
Participants: One-hundred and eight participants volunteered for the study. Eighty-four (36 males, 48 females; 38.6+/-9.3 y) (mean+/-s.d. ) met the inclusion criteria for accelerometer measurements and were included in analyses.
Results: No significant differences in physical activity were identified between normal-weight (BMI<25) and overweight (BMI 25-29.9) participants. Obese participants (BMI>30) were significantly less active than non-obese participants (BMI< or =30) during weekdays (279.1+/-77.5 vs. 391.3+/-139.4 counts min(-1); P<0.001), weekends (222.3+/-93.9 vs. 386.2+/-177.5 counts min(-1); P<0.001) and evenings (221.1+/-126.3 vs. 380.8+/-220.7 counts min(-1); P=0.002), but not at work (307.5+/-87.2 vs 398.7+/-163.3 counts min(-1); P=0.06). Differences in activity levels between obese and non-obese participants were greater in males than females. Hour-by-hour physical activity patterns demonstrated that obese participants were less active than the non-obese for almost every hour of the weekdays and the weekend.
Conclusions: Although no differences in activity emerged between overweight and normal-weight individuals, obese participants were substantially less active than the non-obese, particularly when there was a free choice of activity or no activity. The difference in activity levels was particularly pronounced in males, and unless obese individuals are compensating by lower energy intake, these activity patterns will contribute to the maintenance of or increase in the degree of obesity. Minute-by-minute accelerometry is a valuable tool for the assessment of activity patterns and may be useful to highlight periods of inactivity for intervention design.