Participant characteristics and quality of recall of physical activity in the distant past.

Assessment of physical activity in the distant past, usually by recall, is necessary to evaluate its etiologic effects on chronic disease. Few studies have investigated influences on such long-term recall. The authors investigated the association of gender, age, education, marital status, current physical activity, body weight, body mass index, blood pressure, and pulse with the quality of recall of physical activity in a Buffalo Health Study cohort followed since the early 1960s. Comparisons of original, recalled, and current reports of physical activity were made in 137 survivors of the cohort. The quality of recall (the difference between original and recalled reports) values near zero indicated the best recall; positive values, overestimation; and negative values, underestimation. Overestimators had the highest levels, and good recallers lower levels, of current physical activity. Although the authors found differences by gender, age, and education, the evidence did not support better recall by one group compared with the others. Moreover, no association of marital status, body weight, body mass index, blood pressure, or pulse was found with the quality of recall. The results suggest that individual respondent characteristics have little association with recall of past physical activity; however, current physical activity may be a factor to consider in studies of past physical activity and chronic disease.