Background The role of
adipokine dysregulation in determining the metabolic fate of
obesity is not well studied. We aimed to examine whether the matricellular
protein osteonectin and the profiles of certain
adipokines could differentiate metabolically healthy obese ( MHO ) versus metabolically unhealthy obese phenotypes in childhood. Methods and Results This study included 1137 obese children and 982 normal-weight healthy ( NWH ) controls recruited from the BCAMS (Beijing Child and Adolescent
Metabolic Syndrome) study. MHO was defined by the absence of
insulin resistance and/or any
metabolic syndrome components. Six
adipokines-
osteonectin,
leptin,
adiponectin,
resistin,
FGF21 (
fibroblast growth factor 21), and RBP-4 (
retinol binding protein 4)-were assessed. Approximately 20% of obese children displayed the MHO phenotype. MHO children had a more favorable
adipokine profile than metabolically unhealthy obese children, with lower
osteonectin,
leptin, and RBP -4 and higher
adiponectin (all P<0.05). Compared with normal-weight healthy controls, MHO children displayed increased
leptin,
resistin, and RBP -4 levels and reduced
adiponectin concentrations (all P<0.05) but similar
osteonectin and FGF 21 levels. Among obese subjects, decreased
osteonectin (odds ratio [OR]: 0.82; 95% confidence interval [CI] per standard deviation, 0.70-0.97), RBP -4 (OR: 0.77; 95% CI per standard deviation, 0.64-0.93), and
leptin/
adiponectin ratio (OR: 0.58; 95% CI per standard deviation, 0.43-0.77) were independent predictors of MHO . In addition, compared with children without abnormalities, those with any 3
adipokine abnormalities were 80% less likely to exhibit the MHO phenotype ( OR : 0.20; 95% CI , 0.10-0.43) and 3 times more likely to have
metabolic syndrome ( OR : 2.77; 95% CI , 1.52-5.03). Conclusions These findings suggest that dysregulation of
adipokines might govern the metabolic consequences of
obesity in children. Low
osteonectin levels, along with a healthy
adipokine profile, might be used as an early marker of the MHO phenotype.