Ambulatory activity in men with diabetes: relationship between self-reported and real-world performance-based measures.

The measurement of physical activity, especially walking activity, is important for many outcome studies. In many investigations, the Physical Activity scale of the short-form-36 (SF-36) health assessment questionnaire is used in lieu of an actual physical measurement of walking. This study determined the relationship between the SF-36 questionnaire and the Step Activity Monitor (SAM), a real-world performance-based tool that counts the actual number of steps taken during daily activities. We studied the physical activity of 57 men with diabetes using step count monitoring and the SF-36 questionnaire. The subjects averaged 3,293 steps/day, but had a very wide range (111-11,654) and a large standard deviation (SD = 2,037). The correlations between total daily steps and the SF-36 Physical Component Summary score, and the Physical Function, Bodily Pain, and Vitality scales of the SF-36 were only fair (Pearson's r = 0.376, 0.488, 0.332, 0.380, respectively). The corresponding coefficients of determination range from only 7.7% to 23.8%. Physical activity is a complex concept not completely represented by either the SF-36 or the step counts. The correlation between actual walking activity and the SF-36 is not as strong as many researchers believe. Caution should be exercised with the use of the SF-36 to specifically measure walking activity.