Change in the prevalence of leisure activity with the number of activities recalled.

Purpose: The purpose of this study was to examine how the estimated prevalence of leisure activity changed when respondents self-reported up to four activities, compared with using information only from the respondent's one, two, and three most commonly performed leisure activities.

Methods: Information on leisure activities, sociodemographics, and body mass index (BMI) categories was collected in the year 2000 during a telephone interview of 1813 randomly selected adults living in six North Carolina counties. Recommended activity was defined as moderate leisure activity > or =5x wk(-1) for > or =30 min per session or vigorous leisure activity > or =3x wk(-1) for > or =20 min per session.

Results: Among respondents, 32.1% reported participation in no leisure activity, 41.7% reported one activity, 21.5% reported two activities, 3.8% reported three activities, and 0.9% reported four activities during the past month. The largest change in prevalence occurred when using two activities compared with one activity for recommended activity, overall and by sociodemographic and BMI categories. The prevalence of recommended activity determined by using 1, 2, 3, or 4 activities was 21.0%, 26.0%, 26.5%, and 26.7%, respectively.

Conclusion: In this study, collecting information on two leisure activities was the most efficient balance for population specific accuracy of recommended activity.