Key points are not available for this paper at this time.
Patients with Type 2 diabetes (T2D) demonstrate a remarkable increased risk of mortality. Yet, underlying metabolomic profiling and relevant biochemical alterations are not fully elucidated. We conducted a prospective metabolomic study of T2D mortality risk in 152 case-control pairs within the National Newly Onset Diabetes Cohort (2015-2017), which recruited nearly 18,000 newly diagnosed T2D patients from 46 centers in China. Using incidence-density sampling method, controls were selected from the cohort who were alive at the time of case death and individually matched to cases by age, sex, assessment center, and date of blood draw. Median time from baseline fasting serum collection to death for cases was 4.8 years. We identified 736 known metabolites through the Ultrahigh-performance LC-MS/GC-MS platform. Conditional logistic regression models were used, adjusting for age, sex, body mass index, smoking status, systolic and diastolic blood pressure, HbA1c, and triglyceride. Based on a false discovery rate (FDR) of 0.05, 86 metabolites were found to be associated with risk of mortality among T2D. Integrating approaches of LASSO, forward and stepwise logistic regression, eight metabolites were identified to be independently related to risk of mortality among T2D, including perfluorooctanoate, 5α-pregnan-3β,20α-diol monosulfate, maltose, N-palmitoyl-sphingosine (d18:1/16:0), eicosenedioate (C20:1-DC), behenoyl dihydrosphingomyelin (d18:0/22:0), deoxycholic acid glucuronide, and 16-hydroxypalmitate. Multivariate ROC regression estimated an AUC of 0.63 (referent model: above-mentioned adjusted covariates) versus 0.80 (extended model: additionally adjusted for the eight selected metabolites) (Wald test for comparison: P0.0001). Our data demonstrated that several serum metabolites are independently associated with mortality risk in T2D, which may provide novel insight on the molecular basis of progression and development among individuals with T2D. Disclosure L. Gan: None. S. Yang: None. B. Zhao: None. Z. Zhou: None. J. Huang: None. Funding National Natural Science Foundation of China (82100949)
GAN et al. (Fri,) studied this question.