Abstract presented at the Annual Meeting of the ACSM 2007, New Orleans, LA. USATo develop and validate more accurate self-report measures of physical activity it is important to describe and quantify the various sources of measurement error. Purpose: The purpose of this study is to examine the validity of a computerized (recall-based) self-report measure known as the 24 hour physical activity recall (24PAR). Methods: Twenty healthy adult males (n=10) and females (n=10) wore two different pattern recognition activity monitors (the Intelligent Device for Estimating Energy Expenditure and Activity - IDEEA and the Sensewear Pro II armband - SP2) on two different days while going about their normal daily activities. The morning after each trial, the participants returned to the lab to complete a computerized version of the 24PAR administered by a trained interviewer. Participants were randomly assigned to complete a prospective diary of their daily activities in one of the trials to directly examine the influence of recall bias. Group level estimates of EE (kcal/day) from the 24PAR were compared to the temporally matched data from the 2 monitors using t-tests and Bland Altman graphical procedures. Results: Estimates of EE from the 24PAR were highly correlated with estimates from both the IDEEA (average r = .90) and the SP2 (average r = .88). The PDPAR yielded significantly (p<.05) higher estimates of EE than the IDEEA (difference = 102.0 kcal/day) and the SP2 (difference = 164 kcal/day), but the effect sizes for these comparisons were small (< .20) indicating that the differences may not be of much practical significance. There were no differences in correlations or absolute agreement between the two trials indicating that the use of the diary did not improve overall recall accuracy. Bland Altman plots revealed no evidence of systematic bias across the range of values for either trial. Estimates of the percentage of time spent sitting were similar between the 24PAR (65%) and the IDEEA (66%) and correlations between these measures were high (r = 0.80). Correlations with estimated minutes of MVPA from the IDEEA and the SP2 were lower and more variable (r < .60). Conclusions: The results of this study support the overall utility of the 24PAR for group level estimates of daily EE. Additional research is needed to better understand the factors that influence individual agreement for estimates of EE and MVPA. Publication: Abstract presented at the Annual Meeting of the American College of Sports Medicine 2007, New Orleans, LA. USA |
|
