Poster Topical Area: Energy and Macronutrient Metabolism
Location: Hall D
Poster Board Number: 528
Objectives. We hypothesized that the shape of visceral adipose tissue (VAT) is associated with the metabolic syndrome (MetSx) in adults from different ethnic groups and aimed to determine this association. Methods. Data from 1,096 participants in a cross-sectional study were analyzed to associate the abdominal shape, texture and appearance variation with the presence of the MetSx. Magnetic resonance images (MRI) of the abdomen at the L4-L5 intervertebral space were annotated using 55 landmark points to define the abdominal tissues (VAT, subcutaneous fat, muscle) using statistical appearance modeling (SAM) as described by Cootes. The SAM generated principal components (PCs) that reflect dominant modes of variation in the abdominal structures. In this analysis, only VAT PCs were included as independent variables in multiple logistic regression models to predict the presence of the MetSx. Demographic, body size, and adiposity variables (including VAT cross-sectional area [cm2] as a measure of VAT quantity) were included in the models to test if the PCs provide independent metabolic risk prediction beyond VAT amount. Results. A preliminary analysis of VAT shape, texture and appearance in a sub-sample of 580 participants (mean age=68.6 years), from 3 different ethnic groups, showed that 30 shape modes explained 95% of the VAT shape variance, 40 modes for texture (grayscale variance) and 30 modes for appearance (combination of the shape and texture variances). Shape, texture and appearance PCs remained in the models (p<0.05) even after inclusion of the covariates. The model including VAT-texture PCs had a better R2 than the models for shape or appearance (0.41 vs. 0.38 and 0.38, respectively). These R2s represent a slight improvement over the prediction of the MetSx using only demographic, body size and body adiposity variables, including VAT amount (R2=0.33). Conclusions. The shape and distribution of visceral fat in the abdomen may improve the prediction of cardiometabolic risk, even after considering total VAT amount. Further work in this project will include more participants, additional race/ethnic groups, and additional landmark points to refine the models.
Funding Source: NCI-P01CA168530 "Obesity, body fat distribution, and cancer risk in the Multiethnic Cohort"
NIH/NIDDK-R01DK109008 "Optical Body Composition and Health Assessment"
University of Sonora, Graduate Program on Chemical Biological and Health Sciences
Hermosillo, Sonora, Mexico