Measuring change in adolescent physical activity: Responsiveness of a single item

Abstract

Self-report measures are frequently used to assess change in physical activity (PA) levels. Given the limited data from adolescent populations, the primary objective of this study was to examine the responsiveness of a single item measure (SIM) of PA for adolescents to detect change in moderate-to-vigorous physical activity (MVPA) using accelerometer data as the reference measure. A secondary objective was to provide further data on the validity of the measure at one point in time. The validity of the SIM to determine the number of days ≥60 minutes of MVPA was based on data from 200 participants (62% female; age: 14.0 ± 1.6 years) and analysis of change was based on data from 177 participants (65% female; age: 14.0 ± 1.6 years). Validity of change in days ≥60 minutes of MVPA was examined through agreement in classification of change between the SIM and accelerometry as the reference measurement and Spearman's correlation. Cohen's d and standardised response means were used to assess the responsiveness to change of the measure. The responsiveness of the SIM and accelerometer data were comparable and modest (0.27-0.38). The correlation for change in number of days ≥60 minutes MVPA between the SIM and accelerometery was low (r = 0.11) and the accuracy of the SIM for detecting change, using accelerometry as the reference, was only marginally above chance (53%). Therefore, the adolescent version of the SIM is adequate for assessing PA at a single time point but not recommended for assessing change.

Fig 1

Difference in days ≥60 minutes of moderate-to-vigorous intensity physical activity (MVPA) between the single item measure (SIM) and accelerometry: (A) Week one and (B) Week two. Median difference and inter-quartile range for the ordinal distributions are presented as dotted horizontal lines against the median of the two measures (single-item and accelerometry).