Reliability and Validity of Walking Questions in the Behavioral Risk Factor Surveillance System (Brfss)

1106 In, 1997 a BRFSS Physical Activity Workshop identified the need for new survey questions to measure the moderate-intensity physical activity recommendations. The CDC developed new BRFSS questions to capture non-occupational walking. Test-retest reliability was performed by telephone administration of the questions at a one week interval. Sixty-nine men and women were asked to record the time spent in non-occupational walking in activities for recreation, exercise, or transportation during the last week. Spearman test-retest correlations between the days walked/wk (r=0.61), min walked/wk (r=0.44), and MET-min walked/wk (r=0.42) were significant (P<0.001). Questions were validated with a one week concurrent physical activity log, completed by the subjects at the end of each day. Correlations of non-occupational walking between the survey and the log were significant (P<0.001) for days/wk (r=0.38), min/wk (r=0.44), and MET-min/wk (r=0.44). Correlations between survey walking min/wk and types of walking from the physical activity logs were significant for walking for exercise (r=0.43, P<0.001), but not for walking for recreation (r=0.22) or transportation (r=0.18)(P>0.05). Valid and reliable survey questions about walking provide accurate data that will help guide public health surveillance efforts in the year 2000 and beyond.