In:
Diabetes, American Diabetes Association, Vol. 70, No. Supplement_1 ( 2021-06-01)
Abstract:
The mechanisms linking overall/central obesity with type 2 diabetes are not fully understood. In a nested case-control study of 1039 incident cases and 1039 matched controls from two prospective cohorts of Chinese adults, we measured 142 plasma metabolites using LCMS platforms. We identified and validated 28 metabolites associated with body mass index (BMI) and 34 metabolites with waist circumference (WC) in multivariable-adjusted linear models at false discovery rate & lt;0.05. Using LASSO regression, we selected 15 and 21 metabolites, respectively, to create BMI- and WC-related metabolite composite scores (MCSs), including glutamate, branched-chain amino acids, tyrosine, uric acid, dimethylguanidino valeric acid, C5-carnitine, cis-aconitic acid, and homogentisic acid (positive associations) and 1,5-anhydrosorbitol, pyruvic acid, serine, glycine, and asparagine (inverse associations). Both MCSs showed strong linear associations with diabetes risk: OR (95% CI) in the highest vs. lowest quartiles was 3.92 (1.88, 8.15) for BMI-related MCS and 7.87 (3.53, 17.53) for WC-related MCS after adjusting for diabetes risk factors including plasma glucose (both P-trend & lt;0.001). Strikingly, the MCSs showed even stronger associations with diabetes among participants with normal BMI or WC: OR (95% CI) across quartiles was 6.99 (3.97, 12.3) and 9.98 (6.11, 16.3), respectively. The MCSs mediated 50-70% of the BMI/WC-diabetes association. MCSs alone achieved similar prediction performance as traditional risk factors, including age, sex, lifestyles, diet, family history, and BMI/WC (all C-statistics~0.75). Adding MCSs to traditional risk factors and plasma glucose improved the C-statistics from 0.79 to 0.83 (P & lt;0.001). In conclusion, we identified multiple obesity-related circulating metabolites strongly associated with diabetes risk, even among individuals with normal BMI or WC. Those metabolites could explain up to 70% of obesity-related diabetes risk and further improve diabetes risk prediction. Disclosure X. Pan: None. Z. Chen: None. T. Wang: None. W. Zheng: None. R. E. Gerszten: None. X. Shu: None. D. Yu: None. Funding National Institute of Diabetes and Digestive and Kidney Diseases (R01DK108159 to T.W., X-O.S., R.E.G.), (R01DK126721 to D.Y.); National Cancer Institute (UM1CA173640 to X-O.S.), (UM1CA182910 to W.Z.)
Type of Medium:
Online Resource
ISSN:
0012-1797
,
1939-327X
Language:
English
Publisher:
American Diabetes Association
Publication Date:
2021
detail.hit.zdb_id:
1501252-9
Permalink