Abstract
Sedentary (SED) and physical activity (PA) behaviors influence health and prevalence of chronic diseases. Activity measurements help to understand the influence of SED and PA on health and disease outcomes, but these assessments can be complex. Time spent in SED and PA intensities have different health outcomes. Accelerometers permit objective measurements under free-living conditions to capture time spent in SED, light, moderate, and vigorous PA. Knowing the reliability of PA and SED measurements within a specific research study population is important when interpreting impact of SED and PA behaviors on health and disease outcomes.
PURPOSE: The purpose of this study was to evaluate test-retest reliability of average time per day in SED and light, moderate, and vigorous PA parameters in accelerometer use on 2 different 7 consecutive day periods, 12 weeks apart, in adults with elevated risk for metabolic disease.
METHODS: SED and PA were measured in overweight and obese research subjects in a dietary intervention study using waist worn accelerometers (ActiGraph GT3X+) for 7 days on two occasions 12 weeks apart. Test-retest reliability coefficients of 0.6 < 0.7 were considered questionable; 0.7 < 0.8 acceptable; 0.8 < 0.9 good; and ≥ 0.9 excellent.
RESULTS: A total of 22 subjects (mean ± SD; age = 45.4 ± 14.1 y; BMI = 33.9 ± 6.3 kg.m-2) completed the study. Except for average minutes in vigorous PA, test-retest correlations (p < 0.001 for all) for SED and light, moderate, and vigorous PA of varying bout durations were in the acceptable to excellent range (r = 0.744 to r = 0.924). Test-retest means and correlations for average minutes in each category were SED (493.5 ± 119.9 min, 474.9 ± 100.0 min; r = 0.88, good), light (309.7 ± 87.2 min, 31.6 ± 87.7 min; r = 0.843, good), moderate (42.8 ± 18.5 min, 41.9 ± 20.6 min; r = 0.773, acceptable), and vigorous (2.3 ± 7.6 m, 1.5 ± 3.0 m; r = -.656, questionable).
CONCLUSIONS: In adults at increased risk for metabolic disease, test-retest reliability measured using accelerometry for 7 days varied across SED and PA intensity levels. Reliability was highest for SED and decreased as intensity of PA increased. Time spend in vigorous PA was the most variable between tests and reliability of this measure is questionable in this population.