Recent evidence suggests that HDL particle concentration (HDL-P) and cholesterol efflux capacity (CEC) are better predictors of coronary artery disease than HDL-C in type 1 diabetes (T1D) . We hypothesized that these novel HDL markers also predict all-cause mortality better than HDL-C in T1D. HDL-P (calibrated ion mobility analysis) and CEC (validated cell-based assay) were quantified in the first available stored sample from 550 Epidemiology of Diabetes Complications study participants with childhood onset (<17 years) T1D (baseline mean age 27.8 years and T1D duration, 19.6 years) . Vital status was assessed as of September 30, 2019. During a median follow-up of 26 years, there were 167 deaths. At analytic baseline, deceased were older, with longer T1D duration and a worse cardiovascular risk factor profile compared with survivors. In Cox proportional hazards models, after multivariable adjustments including for HDL-C, the concentration of total HDL-P (T-HDL-P, HR=0.90, 95% CI: 0.84-0.96) and that of three of the four major HDL subpopulations - extra small (XS-HDL-P, HR=0.44, 0.21-0.90) , small (S-HDL-P, HR=0.74, 0.60-0.93) and medium (M-HDL-P, HR=0.80, 0.70-0.91) particles - were associated with a lower mortality risk. Marginal protective associations were also observed for the concentration of large-sized particles (L-HDL-P, HR=0.84, 0.68-1.06) , as well as for CEC (HRs per SD was 0.86 (0.74-1.00) and 0.86 (0.73-1.01) for overall J774 and ABCA1-specific CEC, respectively) . These results remained largely unchanged by further adjustment for HDL-P (in the case of HDL-C and CEC) or APOA1 and triglyceride concentrations (for HDL-P) and were similar by sex. HDL-C was not a significant predictor of mortality in either univariate or multivariable analyses. Our findings show that higher HDL-P concentrations are negatively associated with mortality in T1D independent of other risk factors.


T.Costacou: None. R.G.Miller: Stock/Shareholder; Becton, Dickinson and Company. T.J.Orchard: None. T.Vaisar: Consultant; AstraZeneca.


National Institutes of Health (R01HL130153 and R01DK034818) and the Rossi Memorial Fund.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at