Abstract
IMPORTANCE: Cerebral white matter (WM) damage has been reported in childhood obesity and in metabolic syndrome (MetS) but mechanisms remain unclear. OBJECTIVES: To ascertain whether adolescents with MetS have retinal vessel alterations and if the anticipated reductions in retinal arteriolar diameter are associated with diminished cerebral WM microstructural integrity and to test a model for vascular etiology of the WM abnormalities. DESIGN, SETTING, AND PARTICIPANTS: Cross-sectional study of the brain correlates of obesity and related metabolic disease in youths. This study was conducted at the Brain, Obesity, and Diabetes Laboratory, New York University School of Medicine, New York. Thirty-nine obese adolescents with MetS and 51 matched adolescents without MetS received comprehensive endocrine, neuropsychological, retinal vessel, and diffusion tensor imaging-based cerebral WM evaluations. MAIN OUTCOMES AND MEASURES: Retinal arteriolar diameter, cerebral WM microstructural integrity, waist circumference, and insulin resistance. RESULTS: Obese adolescents with MetS had significant reductions in retinal arteriolar diameter relative to adolescents without MetS (mean [SD] central retinal arteriolar equivalent, 182.35 [16.10] vs. 198.62 [19.03] μm, respectively; P < .001). The greater the number of MetS criteria present, the greater the reduction was in retinal arteriolar diameter (β = -8.61; ∆r2 = 0.335; ∆F1,83 = 70.79; P < .001). We found that abdominal obesity (waist circumference) was the strongest MetS component related to reductions in retinal arteriolar diameter (rp[85] = -0.661; P < .001), and importantly, for the first time to our knowledge, we demonstrated that its effect was partially mediated by comorbid insulin resistance (indirect effect = -0.1355 [95% CI, -0.2471 to -0.0593]; Z = -2.56; P = .01). Consistent with our prior report of nondiabetic adolescents with MetS, we also uncovered cerebral WM microstructural damage. These subtle WM changes were associated with reductions in retinal arteriolar diameter, a proxy for cerebral microvascular health (3150 voxels or 3.15 cm3; P < .001). Importantly, some of the WM regions showing lower microstructural integrity also demonstrated associations with retinal arteriolar diameter, suggesting that the observed WM pathology is likely vascular in nature. CONCLUSIONS AND RELEVANCE: We document, for the first time to our knowledge, the associations between retinal vessel alterations and subclinical WM pathology among obese adolescents with MetS. This suggests that the subtle WM pathology in adolescents with MetS may have a vascular origin. Future work should include direct assessments of cerebral microvascular health.