Association of obesity phenotypes with left ventricular mass index and left ventricular hypertrophy in children and adolescents.

It has been argued that metabolically healthy obesity (MHO) does not increase the risk of cardiovascular disease. The aim of this study is to evaluate whether, in a population of obese children/adolescents, the metabolically unhealthy obesity (MUO) phenotype is associated with higher left ventricular mass index and/or higher prevalence of left ventricular hypertrophy than the MHO phenotype. We also tested whether the addition of an insulin resistance index (HOMA-index >90th percentile by sex and age) and the presence of hyperuricemia (serum uric acid >90th percentile by sex and age) to the definition of MUO better identified obese children with early cardiac damage. Left ventricular hypertrophy was defined as the presence of left ventricular mass index greater than or equal to the age- and sex-specific 95th percentile. The study population included 459 obese children (males 53.2%, mean age 10.6 [standard deviation, 2.6] years), of whom 268 (58.4%) were MUO. The left ventricular mass index was higher in MUO children than in MHO children (37.8 vs 36.3 g/m, p=0.015), whereas the percentage of MUO children presenting left ventricular hypertrophy was only slightly higher in MUO children (31.1 vs 40%, p=0.06). Multiple linear regression analyses showed that the variables significantly associated with higher left ventricular mass index were male gender (p<0.01), Body Mass Index z-score (p<0.001) and Waist-to-Height-ratio (p<0.001). Multiple logistic regression analyses showed that the presence of left ventricular hypertrophy was only significantly associated with higher Body Mass Index z-score (p<0.05) and Waist-to-Height-ratio (p<0.05). In spite of the higher left ventricular mass index of MUO as compared to MHO children, the MUO phenotype was not a significant predictor of either higher left ventricular mass index or higher left ventricular hypertrophy prevalence. The MUO phenotype had a low predictive ability on the presence of left ventricular hypertrophy. The area under the receiver operating characteristic curve was 0.57 (sensitivity 0.64, 1-specificity 0.55). The addition of insulin resistance and hyperuricemia to the definition of MUO did not change the results observed with the standard definition of MUO at multivariable analysis. The MUO phenotype appears to be of little usefulness in identifying the early presence of cardiac damage in a large population of obese children and adolescents. Excess weight and abdominal obesity are confirmed as an important determinant of early organ damage in obese children.