Comparing physical activity assessment methods in the Seasonal Variation of Blood Cholesterol Study

PURPOSE: This paper evaluated three measures of physical activity employed in the Seasonal Variation of Blood Cholesterol Study (Seasons), and it had two objectives: 1) To examine the laboratory validity of the Actillume activity monitor, and 2) To examine the relative validity of three 24-h physical activity recalls (24HR) in quantifying short-term physical activity behaviors. METHODS: Nineteen healthy middle-age adults completed seven activity trials (reading, typing, box moving, stepping, and walking (3.5, 4.25, 5.0 km x h(-1))) while oxygen consumption and Actillume measures were obtained. ANOVA, linear regression, and a scatter plot were employed to examine the validity of the Actillume. In relative validity analyses of the 24HR in the Seasons study, participants (N = 481) completed two or three 24HR (MET-h x d(-1)) and a modified Baecke Questionnaire. A subset of the cohort (N = 41) wore the Actillume for 3-8 d (counts x min(-1) x d(-1)). The relative validity of the 24HR method was examined by comparison to these criterion measures. RESULTS: In laboratory validation analyses, the monitor was found to discriminate between sedentary and moderate intensity activities, changes in walking speed, and to account for 79% of the variance in oxygen consumption across sedentary and walking trials. In relative validity analyses, correlations between the 24HR and the modified Baecke ranged from 0.29 to 0.52 (P < 0.01) across total, household, occupational, and leisure-time activities. CONCLUSIONS: In laboratory testing, the Actillume monitor discriminated between sedentary and moderate intensity activities and was highly correlated with oxygen consumption. Three 24HR of physical activity were observed to have a relative validity that was comparable to published data from other short-term activity assessments that also employed the Baecke Questionnaire and activity monitors as criterion measures.