Intraindividual variability and reliability in a 7-day exercise record

Purpose: High levels of day-to-day or intraindividual variability implies u nreliability of a measure of physical activity. Unreliability in a measure leads to attenuation of correlations with other variables. As intraindividu al variability increases, the number of days necessary to assess physical a ctivity to achieve the desired level of reliability increases. The use of a n intraclass correlation to assess day-to-day reliability in a measure assu mes compound symmetry. Methods: This study reports on these issues in a sam ple of 165 elementary school teachers who maintained a 7-d record of physic al activity each year for 3 yr. Analyses were conduced with physical activi ty measured as minutes, MET minutes, and kcal. Analyses were conducted with PROC MIXED in SAS controlling for the clustering effect by school. Results : Compound symmetry could not be supported across 7 d of the record. The we ekdays tended to intercorrelate, Saturday correlated at very low levers, an d Sunday correlated with Monday only. Compound symmetry was supported acros s the three weeks. Conclusions: To achieve a reliability of 0.8 using a 7-d activity record requires 2 wk of assessment. The reliability of measures o f physical activity require more careful attention, and likely require more points of assessment to achieve desired levels.