Key points are not available for this paper at this time.
Objective: We used metabolomics data to identify metabolite profiles predictive of nephropathy, retinopathy, and/or neuropathy over 11-15 y of follow-up in DPPOS, and to assess the extent to which the profiles represent shared v. distinct pathways. Methods: We applied LASSO to 350 annotated compounds assayed from fasting blood of 1947 participants at baseline (age: 53.1±10.1 y). The models accounted for age, sex, and race. We retained compounds with nonzero estimates from ≥7 of 10 bootstrapped datasets for each outcome. Results: Five hundred seventy-two participants developed ≥1 outcome during follow-up (277 cases of nephropathy, 194 of retinopathy, and 212 of neuropathy). One hundred thirty-seven metabolites predicted at least one outcome. Of them, 6 metabolites were predictive of all three outcomes, including bilirubin, a byproduct of red blood cell breakdown suppressed in the context of hyperglycemia; serine, for which lower levels are linked to impaired nerve function; and 1-methylhistamine, a biomarker of neurological disorders. Fourteen metabolites were predictive of nephropathy and retinopathy, including some involved in branched chain amino acid (BCAA) metabolism - e.g., 2-AAA, phenylalanine, and kynurenic acid. Thirteen metabolites were predictive of nephropathy and neuropathy, including DMGV, which is implicated in fatty liver progression; and 1,5-AG which is linked to poor glycemic control. Seven metabolites were predictive of neuropathy and retinopathy, including quinolinic acid, a tryptophan metabolite; and UDP-GlcNAc, a nucleotide end-product of glucose metabolism. Conclusions: Microvascular complications of diabetes share pathophysiology related to hyperglycemia and impaired nerve function. We observed the largest overlap in metabolite predictors of nephropathy and retinopathy, for which several compounds are part of a BCAA metabolic signature implicated in worsening glycemia. Disclosure W. Perng: None. R. Shu: None. D.M. Nathan: None. J.A. Luchsinger: Consultant; Merck KGaA. Other Relationship; EMD Serono, Wolters Kluwer Health. R.E. Gerszten: None. R.J.W. Middelbeek: Research Support; Novo Nordisk. S.E. Kahn: Advisory Panel; AltPep, Boehringer-Ingelheim, Eli Lilly and Company, Intarcia Therapeutics, Inc., Merck National Institute on Aging (NIA) of the National Institutes of Health (NIH) (U19AG078558)
PERNG et al. (Fri,) studied this question.