Tracking of Cholesterol Among Individuals With and Without Diagnosed Cardiovascular Disease

Date

2003-05-01

Authors

Fisher, Bettina L.

ORCID

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Fisher, Bettina L., Tracking of Cholesterol Among Individuals With and Without Diagnosed Cardiovascular Disease. Master of Public Health (Epidemiology), May 2003, 46 pp., 6 tables, 5 figures, bibliography, 33 titles. Cardiovascular disease is a major public health problem among the elderly in the United States, and cholesterol is the number one risk factor for coronary heart disease. Tracking is a method of analysis used to identify at-risk subjects at an early age in order to institute preventive measures before physical implications of the disease arise. The purpose of this study is to determine the stability and predictability of total serum cholesterol, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) values for people with and without diagnosed cardiovascular disease through tracking. Data was obtained from the Baltimore Longitudinal Study on Aging (BLSA) and comprised men, 45 years of age and older who had at least two measurements. The average length of time subjects were in the study was 21 years, and the average number of repeated measurements was seven. The dataset was divided into two subsets – one for subjects entering the study with diagnosed cardiovascular disease, and a second for subjects entering the study without diagnosed cardiovascular disease. Tracking coefficients were measured using the linear regression model and the linear mixed effects model. There was a high degree of tracking for HDL using the linear regression model and the linear mixed effects model (overall dataset: 0.9283, 0.8216 respectively). Comparing tracking coefficients for cholesterol among subjects with and without diagnosed cardiovascular disease (linear mixed effects: 0.6469 and 0.7668; linear regression model: 0.5408 and 0.6022) reveals that the former subset has less stable cholesterol values than that the latter subset. The linear mixed effects model was the best fit for this data, because it corrects for the variation in the BLSA in interperiod and repeated measurements between subjects.

Description

Citation