A Comparison Of Subjective And Objective Measures Of Physical Exertion

Better methods are needed to improve the validity and reliability of physical activity (PA) assessment instruments when measuring intensity of PA. Research suggests people tend to overestimate level of exertion when self reporting activity on Physical Activity Questionnaires (PAQs). PURPOSE: To compare a subjective method, Borg's Rate of Perceived Exertion (RPE) survey to an objective method utilizing ambulatory PA recorded by an accelerometer. METHODS: Accelerometers were fitted around the waist and positioned on the right side according to manufacturer protocols on 117 participants, (43% male, 67% female, age range 18-74, BMI 26.1 + 5.4, 53% reporting regular exercise). Participants performed three, 0.29-mile laps of increasing self-selected intensity, and then rated exertion by filling out a Borg's 15-point RPE form for each lap immediately following exercise. Exercise categories were divided into four intervals based on validated Hendelman 2 cut-points for comparing metabolic equivalents (METs) and subjective categories (i.e., light, moderate, hard, and very hard). A sub sample was selected to complete the same RPE survey 6 to 8 weeks later recalling their RPE during the three 0.29 mile laps, to assess differences over time. Chi-square tests, logistic regression, and Wilcoxon signed ranks tests measured differences between groups, develop odds ratios where appropriate, and measured discrepancies over time. RESULTS: There were significant differences between genders and overall accuracy (χ2 = 10.9, p = 0.004), as well as whether or not the individual self-reported exercising (χ2 = 15.5, p<0.001). Neither BMIs nor age had an impact on overall accuracy. Odds of underestimating RPE for a regular exerciser are 3.67 times greater than that of a nonexerciser while controlling for all other variables (95% CI = 1.48, 9.11). In assessing differences over time, the sub sample underestimated RPE to a greater extent 6 to 8 weeks after the activity (Z = -2.324, p = 0.020). CONCLUSION: Males and those regularly exercising were more likely to underestimate exertion, while sedentary women tended to overestimate. Recall of PA intensity may decline over time. Misreporting exercise intensity on PAQs is more likely to occur in a subset of the population.