A model of dysmetabolic syndrome X in an ethnically diverse community sample of healthy men and women

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)


Clinical Psychology

First Committee Member

Barry E. Hurwitz - Committee Chair


The Dysmetabolic Syndrome X (DSX) is a constellation of risk factors integrally involved in the pathogenesis of type 2 diabetes and atherosclerotic cardiovascular disease (CVD). Psychological risk factors, such as stress, hostility, and anger, have been positively associated with both cardiovascular disease risk and DSX. Understanding the interrelationships among the elements of this syndrome within the context of risk for these diseases, along with the contribution of psychological factors, is an essential step toward improving cardiovascular disease prevention. Although previous research has investigated the relationships between various components of DSX, often studies emphasize different aspects of the syndrome, use surrogate measures for primary components such as insulin sensitivity, and furthermore, have reported some inconsistent findings. In addition, the complexity of the syndrome and the difficulty of experimentally testing such a large number of interacting factors advocate integrating these relationships into a single conceptual model as a strategic solution to further understanding of the syndrome. The purpose of this dissertation was to derive a model of DSX using data obtained from an ethnically diverse community sample of 339 healthy men and women. Structural equation modeling (SEM) was used to determine a model of DSX that had both good statistical fit and was based on theory, empirical evidence, and logical grounds.Preliminary model specification for the DSX model proposed included the following: A factor comprised of indices of psychological distress and hostility was the sole exogenous variable. Endogenous mediating variables included body mass and fat distribution, sympathetic nervous system activity, inflammation, insulin and glucose metabolic function, lipidemia, and procoagulation, and endogenous outcome variables included cardiac structure, cardiac function, and vascular function. Initially, measurement models were tested using confirmatory factor analysis. Next, structural models (the direct paths between the latent variables) were tested. Model trimming and/or building was facilitated in part by the use of the SEM modification indexes and path coefficients. (Abstract shortened by UMI.)


Health Sciences, Pathology; Psychology, Physiological

Link to Full Text


Link to Full Text