Date of Award

2010

Document Type

Campus Access Dissertation

Department

Epidemiology and Biostatistics

Sub-Department

Epidemiology

First Advisor

Natalie Colabianchi

Abstract

The purpose of this dissertation was to elucidate the degree of misclassification between self-report physical activity, objective physical activity and objective sedentary behavior measured by accelerometry, and cardiorespiratory fitness and to help clarify the independent and joint associations between these four different physical activity/cardiorespiratory fitness assessment methods and adiposity related to higher blood pressure.

Participants were 814 men and women (18-49 years) in the 2003-2004 National Health and Examination Survey study cycle for whom physical activity/cardiorespiratory fitness measures, adiposity measures, and blood pressure measures were assessed. The four assessment methods were each categorized into three groups: 1) Least desirable 20% (i.e. least active, least fit, or most sedentary); 2) Middle 40%; and 3) Most desirable 40% (i.e. most active, most fit, or least sedentary). Survey adjusted observed agreements, expected agreements, Kappas, and observed:expected agreements ratios were used to evaluate the agreement between all four methods and between any two methods (Paper 1). Survey adjusted linear and quantile regression models were used to evaluate the association between each assessment method and higher blood pressure (Paper 2) and the independent and joint associations of physical activity/cardiorespiratory fitness and adiposity with higher blood pressure (Paper 3).

There was poor agreement across the 20/40/40 categories between all four physical activity/cardiorespiratory fitness assessment methods (κ=0.031) and between any two methods (κ=-0.003-0.115). The least desirable 20% was generally associated with the greatest observed:expected agreement ratios and the middle 40% was generally associated with the lowest agreement ratios.

Objective physical activity and objective sedentary behavior were most strongly associated with higher systolic blood pressure and higher diastolic blood pressure, respectively, compared with the other assessment methods. After adjusting for potential confounders, the significant systolic blood pressure differences at the 90th percentile in objective physical activity for the middle 40% and most desirable 40% compared with the least desirable 20% were -3.98 mmHg and -2.27 mmHg, respectively. In addition, after adjusting for potential confounders, the significant diastolic blood pressure differences at the 90th percentile in objective sedentary behavior for the middle 40% and the most desirable 40% compared with the least desirable 20% were 1.74 mmHg and -0.82 mmHg, respectively.

BMI was most strongly associated with both higher systolic blood pressure and higher diastolic blood pressure compared with waist circumference and percent body fat. BMI but not objective physical activity was independently associated with systolic blood pressure after adjusting for potential confounders. Further, both objective sedentary behavior and BMI were independently associated with diastolic blood pressure at the 90th percentile (P=.005, P<.000, respectively). Both these associations remained significant after adjusting for potential confounders; however, the trend for the middle 40% for objective sedentary behavior and higher diastolic blood pressure was not in the opposite direction.

The joint combination of objective physical activity and BMI was significantly associated with systolic blood pressure at the 90th percentile after adjusting for potential confounders (P<0.000) with the most active 40% and normal BMI group having the greatest negative difference in systolic blood pressure from the least active 20% and overweight/obese group (-9.25 mmHg). In addition, the joint combination of objective sedentary behavior and BMI was significantly associated with diastolic blood pressure at the 90th percentile (P<0.000); however, most combination groups were attenuated to no longer significantly different from the most sedentary 20% and overweight/obese group after adjusting for potential confounders.

Given the poor agreement values across different physical activity/cardiorespiratory fitness assessment methods in this study, the results of this dissertation suggest these different assessment methods are capturing different constructs of physical activity/fitness with little overlap, classifying the same individuals in different categories. This supports the idea that issues pertaining to the specific physical activity assessment method may underlie differences in findings between studies of physical activity, cardiorespiratory fitness, and health outcomes. It is these studies that are ultimately considered when modifying national physical activity recommendations. In addition, the results of this dissertation support that being both physically active and having a normal body weight is associated with lower blood pressure values in early and middle adulthood.

Share

COinS