Physical Activity, Body Composition And Their Associations With Health In Yup'ik People

Abstract

Being active and preventing excess body fat are important for maintaining good health. The ability to measure activity and body composition accurately is important to understanding the role of activity and adiposity in health. This dissertation highlights key findings regarding assessment tools for physical activity and body composition and the associations between physical activity and body composition with health in Yup'ik people. The main objectives of this dissertation were to: l) assess the accuracy of bioimpedance and multiple regression models from anthropometry for estimating body composition (fat mass, fat-free mass, and percent body fat) as compared with doubly-labeled water (DLW) body composition estimates, 2) determined the associations between body size estimates, including simple anthropometry and body composition estimates, and obesity-related health risk factors and disease outcomes, 3) assess the accuracy of a combined heart rate/movement monitor (Actiheart) for determining physical activity energy expenditure (PAEE) as compared with DLW PAEE, and 4) determine the associations between physical activity subcomponents and obesity-related health risk factors in Yup'ik people in southwestern Alaska. Body composition can accurately be estimated using only three variables -sex, waist circumference (WC), and hip circumference with a multiple R 2=0.9730 with DLW fat mass. WC and other anthropometries were more highly correlated with a number of obesity-related risk factors than were direct estimates of body composition. When determining the accuracy of the Actiheart for determining PAEE as compared with DLW PAEE, none of the software PAEE models investigated were significantly correlated with DLW PAEE (ranging from r=0.02 {95% e1 (-0.38, 0.41) to r=0.22 {-0.20, 0.56)). Limits of agreement (mean difference +/- 1.9650) for all software models were large, ranging from ---4540 to 1600 kcal/day. The best correlate of DLW PAEE from the Actiheart was the sum of accelerometer counts per day (r=O.SO (95% Cl = 0.13, 0.74)). The associations between physical activity and obesity-related health risk factors showed that total movement throughout the day was positively associated with HDL cholesterol (r<sub>s=0.13, r<sub>s =0.15 men and women respectively) and negatively associated with body weight, body mass index, we, percent body fat and triglycerides (r<sub>s range from -0.17 to -0.25 in men and -0.19 to -0.21 in women). Sedentary time was positively associated with body weight, we, and percent body fat (r<sub>s range from 0.10 to 0.18 in women) and negatively associated with HDL cholesterol (r<sub>s =-0.19 in women). Moderate-to-vigorous physical activity was only associated with fasting glucose. We conclude that 1) body composition in Yup'ik people can be accurately estimated from simple anthropometries and that simple anthropometries like WC can be used to assess obesity-related health risk, 2) the Actiheart software does not accurately estimate free-living PAEE in Yup'ik people however, total movement per day correlates with DLW PAEE and can be used as a proxy for PAEE, and 3) the accumulation of regular movement of any intensity while decreasing sedentary time may be more important for health in Yup'ik people than moderate-to-vigorous physical activity.