Abstract 9113: Metabolic Syndrome and Principal Component Analysis: Multi-Ethnic Study of Atherosclerosis (MESA)
Introduction: The NCEP definition of the Metabolic Syndrome (MetS) is based on an ad hoc and non-linear combination of six metabolic risk factors from predominantly Caucasian populations. We propose a new definition for the MetS based on principal component analysis (PCA) of the same metabolic risk factors in a large multi-ethnic cohort and compare prediction of incident CVD events with the NCEP MetS definition.
Methods: We performed PCA of the elements of the MetS (waist, HDL, TG, fasting blood glucose, SBP, and DBP) in 2610 Caucasian- Americans, 801 Chinese-Americans, 1875 African-Americans, and 1494 Hispanic-Americans in the MESA cohort. We selected the first principal component as a continuous metabolic syndrome score (MetS - PC). Multivariable Cox proportional hazards models were used to examine the association between MetS-PC and 5.5 years of adjudicated all CVD events (n=384) with and without adjustment for age, gender, race, smoking and LDL. Ethnicity specific analyses were performed. Kaplan-Meier survival analyses were performed to test for a trend between increasing quartiles of MetS-PC and CVD events. To compare the MetS-PC with the NCEP definition, a MetS-PC cut-point was chosen to yield the same prevalence of MetS as the NCEP definition (37%) in the MESA cohort. Hazard ratios for CVD events were estimated using the NCEP and Mets-PC derived binary definition.
Results: The HR (95% CI) for CVD events using the MetS-PC was, 1.46 (1.36–1.58) (p<0.0001). Stratified by ethnicity HR's were: Caucasian, 1.42 (1.26–1.60), Chinese, 1.39 (1.06–1.83), African, 1.44 (1.25–1.65), Hispanic, 1.69 (1.43–1.99), all p<0.0001 except Chinese (p<0.02) after adjustment. Log rank tests for trend for CVD events with increasing quartiles of MetS-PC were highly significant overall and in each ethnicity (p < 0.0001) except Chinese (p<0.02). The HR for CVD events (fully adjusted) increased from 1.54 to 3.66 with increasing quartiles of MetS-PC. HR for CVD events (fully adjusted) using the NCEP MetS was, 1.79 (1.46–2.20) vs. 2.34 (1.91–2.87) when using the binary definition of MetS-PC.
Conclusions: MetS-PC is a continuous measure of cardiovascular risk and is an excellent predictor of CVD events overall and in individual ethnicities and outperforms NCEP MetS.
- Metabolic syndrome
- Epidemiologic methods
- Risk factors
- Cardiovascular disease
- Cardiovascular disease prevention
- © 2010 by American Heart Association, Inc.