The association of serum sulfur amino acids and related metabolites with incident diabetes: a prospective cohort study

Abstract

Aim: Plasma total cysteine (tCys) is associated with fat mass and insulin resistance, whereas taurine is inversely related to diabetes risk. We investigated the association of serum sulfur amino acids (SAAs) and related amino acids (AAs) with incident diabetes.

Methods: Serum AAs were measured at baseline in 2997 subjects aged ≥ 65 years. Diabetes was recorded at baseline and after 4 years. Logistic regression evaluated the association of SAAs [methionine, total homocysteine (tHcy), cystathionine, tCys, and taurine] and related metabolites [serine, total glutathione (tGSH), glutamine, and glutamic acid] with diabetes risk.

Results: Among 2564 subjects without diabetes at baseline, 4.6% developed diabetes. Each SD increment in serum tCys was associated with a 68% higher risk (95% CI 1.27, 2.23) of diabetes [OR for upper vs. lower quartile 2.87 (1.39, 5.91)], after full adjustments (age, sex, other AAs, adiposity, eGFR, physical activity, blood pressure, diet and medication); equivalent ORs for cystathionine were 1.33 (1.08, 1.64) and 1.68 (0.85, 3.29). Subjects who were simultaneously in the upper tertiles of both cystathionine and tCys had a fivefold risk [OR = 5.04 (1.55, 16.32)] of diabetes compared with those in the lowest tertiles. Higher serine was independently associated with a lower risk of developing diabetes [fully adjusted OR per SD = 0.68 (0.54, 0.86)]. Glutamic acid and glutamine showed positive and negative associations, respectively, with incident diabetes in age- and sex-adjusted analysis, but only the glutamic acid association was independent of other confounders [fully adjusted OR per SD = 1.95 (1.19, 3.21); for upper quartile = 7.94 (3.04, 20.75)]. tGSH was inversely related to diabetes after adjusting for age and sex, but not other confounders. No consistent associations were observed for methionine, tHcy or taurine.

Conclusion: Specific SAAs and related metabolites show strong and independent associations with incident diabetes. This suggests that perturbations in the SAA metabolic pathway may be an early marker for diabetes risk.

Keywords: Adiposity; Android/gynoid ratio; Protein intake; Renal function; Sulfur amino acids.