Publication Date

2012-05-04

Availability

Embargoed

Embargo Period

2014-05-04

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PHD)

Department

Psychology (Arts and Sciences)

Date of Defense

2012-04-10

First Committee Member

Maria M. Llabre

Second Committee Member

Neil Schneiderman

Third Committee Member

Frank J. Penedo

Fourth Committee Member

Ralph L. Sacco

Fifth Committee Member

David J. Lee

Abstract

Using preliminary data from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), this cross-sectional study employed latent class analysis to investigate 1) whether distinct subtypes of metabolic syndrome (MetS) could be identified among a large and diverse sample of US Hispanic/Latinos (H/Ls), 2) how identified MetS subtypes differed in demographic, socioeconomic, clinical, and behavioral characteristics, and 3) the association between identified MetS subtypes and cardiovascular disease (CVD) prevalence. Incorporating continuous measures of MetS components (waist circumference, systolic and diastolic blood pressure, HDL cholesterol, triglycerides, and fasting glucose) and data on antihypertensive, lipid-lowering, and glucose-lowering medication use into analyses, two latent clusters were identified as best representing the data among the entire sample (n = 10970), as well as among men (n = 4429) and women (n = 6541) separately. One cluster was characterized by individuals exhibiting relatively healthy mean levels across most MetS components (Non-MetS cluster), while the other cluster was characterized by individuals exhibiting clinically elevated mean levels across most MetS components (MetS cluster). The presence of additional, meaningful subtypes of MetS was not confirmed. This two-cluster model was associated with multiple covariates and prevalent CVD outcomes in a manner generally consistent with previous scientific knowledge, demonstrating adequate construct validity. For example, individuals who were older and had a positive family history of CHD exhibited greater odds of being classified into the MetS cluster, and, in turn, those classified into the MetS cluster demonstrated greater odds of having prevalent coronary heart disease. While study results largely converged with current conceptualizations of MetS as a distinct cardiometabolic state, valuable information pertaining to the presentation of MetS specifically among US H/Ls was obtained. For instance, compared to other MetS components, findings suggest that HDL cholesterol may poorly differentiate between individuals who have and do not have MetS among this ethnic group. Additionally, incipient evidence is provided suggesting that currently identified thresholds for some MetS components (i.e., the waist circumference cutoff proposed for US females by NCEP-ATP III criteria) might not be optimal for diagnosing MetS among US H/Ls. Given the exploratory nature of this methodology, and study design constraints (i.e., lack of a non-H/L comparison cohort), these results are tentative and warrant replication. The potential insights offered by adopting this analytic approach to the study of MetS are discussed, as are associated strengths, limitations, and directions for future research.

Keywords

metabolic syndrome, cardiovascular disease, hispanics, latent class analysis

Share

COinS