A novel approach to childhood obesity: circulating chemokines and growth factors as biomarkers of insulin resistance

Abstract

Background Insulin resistance (IR) in children with obesity constitutes a risk factor that should be precisely diagnosed to prevent further comorbidities. Objective Chemokines were evaluated to identify novel predictors of IR with clinical application. Methods We analysed the levels of cytokines (tumour necrosis factor [TNF] alpha and interleukins [ILs] 1 beta, 4, 6 and 10), chemokines (stromal cell derived factor 1 alpha, monocyte chemoattract protein [MCP] 1, eotaxin and fractalkine) and growth factors (brain-derived neurotrophic factor, pro-fibrotic platelet-derived growth factor [PDGF-BB] and insulin-like growth factor 1) in serum of prepubertal children with obesity (61 girls/59 boys, 50% IR and 50% non-IR) and 32 controls. Factor analysis, correlation, binary logistic regression and receiver operating characteristic analysis of combined biomarkers were used to validate their capability for preventive interventions of IR. Results Changes in MCP1, eotaxin, IL1 beta and PDGF-BB were observed in IR children with obesity. Bivariate correlation between stromal cell derived factor 1 alpha, MCP1, eotaxin, TNF alpha, brain-derived neurotrophic factor and/or PDGF-BB explained the high variance (65.9%) defined by three components related to inflammation and growth that contribute towards IR. The combination of leptin, triglyceride/high-density lipoprotein, insulin-like growth factor 1, TNF alpha, MCP1 and PDGF-BB showed a sensitivity and specificity of 93.2% for the identification of IR. The percentage of correct predictions was 89.6. Conclusions Combined set of cytokines, adipokines and chemokines constitutes a model that predicts IR, suggesting a potential application in clinical practice as biomarkers to identify children with obesity and hyperinsulinaemia.