Development and validation of a very brief questionnaire measure of physical activity in adults with coronary heart disease

Abstract

Background: The aim of this study was to validate the Total Activity Measure, a brief questionnaire, to measure physical activity in an older adult population with heart disease.

Methods: Two versions of the Total Activity Measure were administered twice, 7 days apart. The Total Activity Measure 1 asked respondents for the frequency and average duration of bouts of physical activity at three different intensity levels per week, whereas the Total Activity Measure 2 asked respondents for the total time spent in activity at each activity level per week. Questionnaire accuracy was studied in 62 men and 15 women aged 47-84 years, by repeatability and comparison of both administrations of the Total Activity Measure 1 and Total Activity Measure 2 with 7-day RT3 accelerometer data.

Results: Seventy-three adults (58 men, 15 women) were used for all statistical analyses. Intraclass correlation coefficients for the Total Activity Measure 1 and Total Activity Measure 2 total activity scores (metabolic equivalent per minute) were r=0.73 (95% confidence intervals, 0.56-0.83) and r=0.82 (95% confidence intervals, 0.71-0.88), respectively. Correlations between the Total Activity Measure 1 and RT3 accelerometer for total activity score (metabolic equivalent per minute) were significant, r=0.26 at time 1 and r=0.27 at time 2 for moderate intensity activities. Correlations between the Total Activity Measure 2 and RT3 accelerometer for total activity score (metabolic equivalent per minute) were also significant, r=0.38 at time 1 and r=0.36 at time 2, r=0.31 at time 2 for strenuous intensity activities and r=0.29 at time 1 and r=0.25 at time 2 for moderate intensity activities. Participants overestimated the amount of physical activity on both questionnaires as compared with the RT3 accelerometer.

Conclusions: The Total Activity Measure 2 was reasonably accurate in assessing total and moderate intensity activity over a 7-day period and demonstrated good test-retest reliability. The Total Activity Measure 1 was less accurate. The Total Activity Measure 2 is a suitable measure of total or moderate intensity physical activity for surveys and audits in an adult cardiac population.