To explore the correlation structure of individual components of the Metabolic Syndrome and identify combinations of correlated variables we used principal component analysis (PCA) in a multiethnic group of women with a history of GDM (pGDM) (185 European, 103 South Asian, 80 AfricanCaribbean), recruited at 20.0 (18.222.1) months (geometric mean (95% CI)) after pregnancy and 482 control women of similar ethnicities. BMI, W:H ratio, systolic and diastolic BP, f. glucose, f. triglycerides, f. insulin and HDL-Cholesterol we included in the analysis. Regression analysis was used to assess the impact of GDM and ethnicity on principal components. Local ethical committee approval has been obtained.
The PCA analyses identified two components accounting for 52% of the total variability. The first PC explained 36% of the variability, was determined primarily by BMI, W:H ratio, BP and serum triglyceride and denotes individuals with the Metabolic Syndrome. pGDM women and those with abnormal glucose regulation, had significantly higher values of the first PC than the controls (P<0.001). South Asian and African Caribbean women also had higher levels of the first PC than Europeans. The second PC, explained 16% of the variability and was characterized by high systolic and diastolic BP and high HDL-Cholesterol, with the other features of the Metabolic Syndrome displaying negative loadings. High values of this second PC denote individuals of AfricanCaribbean ethnicity. Control Women had higher values of the 2nd PC than pGDM women (P<0.001).
PCA identified Metabolic Syndrome features among pGDM women, especially among non-European women. PCA also indicated that among women of African-Caribbean ethnicity, high blood pressure is not necessarily accompanied by the other features of the Metabolic Syndrome. PCA is a powerful instrument for multivariate analysis in a population at risk for the metabolic syndrome.