Abstract
Introduction: Genetic or pharmacological inhibition of sphingolipid synthases prevented diabetes in animal studies. Scanty of cohort studies that examined circulating sphingolipids and incident diabetes yielded disparate findings. Hypothesis: Specific sphingolipids are associated with elevated risk of diabetes beyond traditional risk factors. Methods: Untargeted metabolomics profiling was performed using fasting serum samples collected from 2010 adult participants of the Hispanic Community Health Study/Study of Latinos, who were free of diabetes and other major chronic diseases at baseline (2008 to 2011). A total of 43 sphingolipids were quantified and 6 sphingolipid scores were derived considering subclasses and chemical structures. Survey Poisson regressions were applied to estimate relative risk (RR) and 95% confidence interval (CI) of incident diabetes associated with individual sphingolipids and the scores. Results: Over ~6 years of follow-up, 224 incident diabetes cases were identified using the ADA criteria. After adjustment for socioeconomic and lifestyle factors, 8 sphingolipids were associated with higher risk of diabetes significantly at the FDR-adjusted level ( Figure ). After further adjusting for general and central adiposity, blood pressure, CRP, HDL cholesterol, and triglycerides, only associations for three saturated sphingomyelins (SMs),36:0 (RR =1.41 [95% CI 1.17-1.69]), 38:0 (RR =1.32 [1.11-1.56]), and d18:0/22:0 (RR =1.45 [1.24-1.71]) remained. Of the 6 scores, only a score comprising 5 saturated SMs was associated with higher risk of diabetes (RR =1.31 [1.12-1.53]) after the full adjustment. An addition of the saturated SM score to the full model including traditional risk factor lead to a modest improvement in diabetes risk classification (net reclassification improvement = 22.0%; 95% CI: 7.8%-36.1%). Conclusions: Our findings suggest that a cluster of saturated SMs may be associated with higher risk of diabetes beyond traditional risk factors in US Hispanics/Latinos.
