Prevalence and heritability of clusters for diagnostic components of metabolic syndrome: the Oman family study.

BACKGROUND

Prevalence and heritability of metabolic syndrome (MetS) vary between populations according to the currently used criteria. We examined combinations for joint probabilities and heritabilities of MetS criteria from the National Cholesterol Education Program Adult Treatment Panel III (NCEP), World Health Organization (WHO), and International Diabetes Federation (IDF) in a sample of Omani families.

METHODS

We included 1277 subjects from 5 pedigrees. The likelihood ratio of diagnostic cluster dependence over clustering by chance was LDep = P(dependent)/P(independent). Heritabilities were adjusted by sex and age.

RESULTS

The highest LDep were central obesity (CO) + high glucose level (HGl) + triglycerides (IDF, 3.08; NCEP, 4.38; WHO, 3.17; P < 0.001). Triglycerides combined with any other component were the most common cluster. The lowest LDep for IDF were high blood pressure (HBP) + CO + low HDL-C (1.21, P < 0.025); for NCEP were HBP + HGl + low HDL-C (1.21, P < 0.07). These components were gathered almost by chance alone. In contrast, the lowest LDep for WHO were HGl + CO + low HDL-C (2.01, P < 0.001). The WHO criteria yielded the highest heritability for a MetS diagnosis (h(2) = 0.9), followed by NCEP (0.48) and IDF (0.38). The rationale of the MetS diagnostics is based on insulin resistance. This base would be lost if we continue lowering cut-off points for diagnosis for increasing the sensitivity. The WHO showed the highest values for LDep for all components because they used the highest cut-off points.