Medicine & Science in Sports & Exercise:
D-32 Free Communication/Poster - Methods: Self-report: JUNE 2, 2011 1:00 PM - 6:00 PM: ROOM: Hall B
1Carolinas HealthCare System, Charlotte, NC. 2University of South Carolina, Columbia, SC.
(No relationships reported)
PURPOSE: Reviewing the scientific evidence relating physical activity (PA) and cardiorespiratory fitness (CRF) to health outcomes is critical in establishing national PA recommendations. The purpose of this study was to determine the degree of misclassification between different PA/CRF assessment methods in a large, heterogeneous population.
METHODS: A total of 814 participants (18-49 yr) who participated in the 2003-2004 NHANES study cycle were included. Self-report PA assessed with a questionnaire, objective PA and objective sedentary behavior assessed by accelerometry, and CRF assessed with a submaximal treadmill test were each categorized into three groups: 1) Least desirable 20% (i.e. least active/fit or most sedentary); 2) Middle 40%; and 3) Most desirable 40% (i.e. most active/ fit or least sedentary). Information on demographics and adiposity measures (BMI, waist circumference, and % body fat) were also collected. Survey adjusted observed agreements, expected agreements, Kappas, and observed:expected agreements ratios were computed. In addition, significant observed agreement differences by demographics or adiposity measures were investigated.
RESULTS: There was poor agreement across the 20/40/40 categories between all four PA/CRF assessment methods and between any two methods (κ=-0.003-0.115). Objective sedentary behavior was most likely to disagree compared with the other methods. Higher observed agreements were generally found in: 1) the combination of the most desirable 40% and the normal BMI/% body fat group; 2) the combination of the least desirable 20% and the obese BMI/% body fat group; 3) females; and 4) young adults (18-29 years) between all four assessment methods and between any two assessment methods.
CONCLUSION: Results suggest that different assessment methods capture different constructs of PA/CRF with little overlap, classifying the same individuals in different categories. In addition, results support the idea that PA and sedentary behavior are two distinct types of behaviors meaning individuals can be physically active, but also could lead very sedentary lifestyles outside of their PA regimen. Researchers need to carefully consider the PA assessment used when stating their conclusions.