Seven-day activity and self-report compared to a direct measure of physical activity.

To determine how well a seven-day interview-administered activity recall used in a large epidemiologic study at Stanford University reflected seven days of self-reported activity and directly measured physical activity, 30 white males, mean age 52 years, recorded daily physical activity for a week, and half of these wore an ambulatory solid-state minicomputer (Vitalog) which measures continuous heart rate and motion. Total hours of moderate, hard, and hard plus very hard activity were not significantly different for weekdays and weekends for self-report and recall and were significantly correlated. Total energy expenditure for subjects wearing the Vitalog averaged 38.5 +/- 6.7 kcal/kg/day compared to an average of 37.7 +/- 4.5 kcal/kg/day for recall or 39.6 +/- 7.2 kcal/kg/day for self-report. Conditioning activities are best remembered followed by home or leisure and job activities. Mean hours of sleep per week night were significantly greater reported by self-report than reported by recall, but the two were significantly correlated. It is concluded that a seven-day activity recall accurately reflects mean kcal/day expenditure, with conditioning activities being the best recalled. A self-report log used in conjunction with an interview-based seven-day recall might maximize accuracy of recall.