Objective: Examine the association of adiponectin and metabolic syndrome components with measures of global and lobar cortical thickness.
Background: Metabolic syndrome has been associated with structural brain changes, but the relationship of adiponectin and cortical thickness is understudied.
Methods: The Northern Manhattan Study MRI Sub-Study is a mostly Hispanic, stroke-free, prospective cohort study of older adults. Cortical thickness (mm) was obtained from T1-weighted brain MRIs using the publically-available Freesurfer software. Regional cortical thickness metrics were averaged to obtain mean lobar cortical thickness. Adiponectin (μg/mL) was measured at baseline (1993-2001). Metabolic syndrome components were measured at MRI Sub-Study baseline (2003-2008). We estimated the cross-sectional associations of adiponectin (per 1 SD) and metabolic syndrome components with global and lobar cortical thickness (per 1 SD) using multivariable linear regression models adjusted for sociodemographic factors, APOE ε4 allele presence, and health-related behaviors. All hypothesis testing was two-sided with an alpha level of 5%.
Results: Freesurfer data were available in 947 participants (mean±SD age=70±9 years, 63% women, 66% Hispanics, 16% black, and 15% white). Global cortical thickness was normally distributed (mean±SD = 2.3±0.1mm). In fully adjusted models, 1 SD (4.9μg/mL) increase in adiponectin was associated with smaller overall (β [95%CI] = -0.07 [-0.14, -0.0002]) and parietal cortical thickness (β [95%CI] = -0.08 [-0.03, -0.0002]). Greater blood glucose levels significantly associated with smaller occipital cortical thickness (β [95%CI] = -0.003 [-0.006, -0.0007]). Greater waist circumference was significantly associated with smaller frontal cortical thickness (β [95%CI] = -0.02 [-0.04, -0.0007]). Neither blood pressure (systolic and diastolic) nor cholesterol (total, HDL-C, and LDL-C) were associated with global or regional cortical thickness.
Conclusions: There was heterogeneity in the cross-sectional associations between adiponectin, metabolic syndrome components, and regional cortical thickness. Further studies are needed to explore the temporal relationship between risk factors and cortical thinning.