Physical activity assessment for public health: efficacious use of the single-item measure

Abstract

The accurate mass assessment of physical activity is essential for effective Public Health policy and practice. Combined with a desire to minimize participant burden, the self-reported single-item physical activity screening measure has become increasingly attractive and widespread. To help reduce any potential misclassification, refining this instrumentation in line with any changes in prescribed activity levels is essential to optimize accuracy. This study compares the levels of agreement, sensitivity and specificity for the single-item measure versus International Physical Activity Questionnaire (IPAQ) using current physical activity recommendations. Agreement was assessed in a non-probability sample of 7650 adults. The κ statistic, sensitivity and specificity were used to assess agreement between the tools for classifying participants as sufficiently active for health (≥150 min of physical activity per week) or not, and being classified as inactive (<30 of minutes of physical activity per week) or not. The single-item measure showed weak agreement with the IPAQ for identifying participants who met the current physical activity guidelines (κ = 0.13, 95% CI 0.12 to 0.14), sensitivity was 18.7% and specificity was 97.2%. For the classification of inactive participants it showed a moderate agreement with IPAQ (κ = 0.45, 95% CI 0.43 to 0.47), sensitivity was 74.2% and specificity was 79.7%. The single-item measure had a low diagnostic capacity compared to IPAQ. Further research is needed if it is to be used in large scale surveys and interventions where screening for sufficiently active or inactive individuals is the goal.