1461-P: HbA1c-Independent Association between DNA Methylation (DNAm) and 28-Year Risk of Microvascular Complications in Type 1 Diabetes (T1D)

Abstract

DNAm has been associated with microvascular disease but the few epigenome-wide association studies (EWAS) of complications in T1D were largely cross-sectional with limited data on glycemic exposure. We thus performed separate EWAS of 28-yr overt nephropathy (ON) and proliferative retinopathy (PR) risk to determine whether DNAm associations are independent of concurrent HbA1c or differ by long-term HbA1c trajectory in the Pittsburgh Epidemiology of Diabetes Complications study of childhood onset (<17 yrs) T1D (baseline mean age 29, T1D duration 21 yrs). After quality control, baseline whole blood DNAm (Illumina EPIC array) at 683,597 CpGs was analyzed in Cox models for time-to-ON (albumin excretion rate >200µg/min in ≥2 of 3 timed urines, n=306 free of ON at baseline) and PR (ETDRS grade≥60 or laser photocoagulation, n=265 free of PR), adjusting for T1D duration, sex, smoking pack-years, est. blood cell composition, and technical covariates. Additional EWAS further adjusted for baseline HbA1c. CpGs with false discovery rate (FDR) <0.20 were examined by previously derived latent HbA1c trajectory (Low: stable 7-8%, Improved: 10% decreasing to 7%, Intermediate-High: stable 9%-11%). Over 28 yrs, 65 (21.2%) developed ON; 139 (52.5%) PR. For ON, DNAm at cg19693031 (TXNIP) was associated (β per 5% DNAm=−0.43±0.09, p=2.84E-07, FDR=0.19) but attenuated in a baseline HbA1c-adjusted EWAS (β=−0.32±0.08, p=7.76E-05, FDR=0.98). For PR, 3 CpGs (cg27512687 [KIF16B]; cg04202206 [ISCA1]; cg14678509 [intergenic]) had FDR<0.20; only cg27512687 remained so after HbA1c adjustment (β=−2.6±0.49, p=8.46E-08, FDR=0.06). There was no interaction with baseline HbA1c (p=0.49); the magnitude of the cg27512687-PR association was similar regardless of HbA1c trajectory. Thus, further study of cg27512687 in KIF16B, a locus encoding a kinesin involved in receptor recycling/degradation, is warranted as it may make an HbA1c-independent contribution to PR risk T1D. Disclosure R.G.Miller: None. J.Mychaleckyj: None. S.Onengut-gumuscu: None. T.J.Orchard: None. T.Costacou: None. Funding American Diabetes Association (1-19-JDF-109 to R.G.M.); National Institute of Diabetes and Digestive and Kidney Diseases (R01DK034818)