No Association Between the MTHFR Gene Polymorphism and Diabetic Retinopathy in Type 2 Diabetic Patients Without Overt Nephropathy: Response to Maeda et al.

Maeda et al. (1) have recently demonstrated that the presence of the C677T mutation in the methylenetetrahydrofolate reductase (MTHFR) gene in diabetic patients can be a predictive for diabetic retinopathy (DR), especially nonproliferative DR (NPDR). We also genotyped the MTHFR polymorphism (C677T) in 366 type 2 diabetic patients without overt nephropathy, and no such association was found. The subjects had a mean age of 60.0 years, duration of diabetes of 11.7 years, HbA_1c of 7.3%, and serum creatinine of 0.71 mg/dl. The patients with urinary albumin excretion >300 mg/g creatinine were excluded. The allelic frequency of the C677T mutation was 0.39, and the genotypes were in Hardy-Weinberg equilibrium (677C/677C, 36.3%, n = 133; 677C/677T, 49.7%, n = 182; and 677T/677T, 14.0%, n = 51), similar to those in their study. Of our 366 diabetic patients, 14.2% (n = 52) had NPDR and 12.6% (n = 46) had proliferative DR (PDR). The remaining 73.2% (n = 268) had no DR. There was no association between the genotypes and clinical parameters such as age, duration of diabetes, HbA_1c, serum lipids, and serum creatinine. The frequency of the MTHFR (C677T) polymorphism in the patients with DR did not significantly differ from that in patients without DR (DR: 677C/677C, 33.7%; 677C/677T, 51.0%; and 677T/677T, 15.3% vs. without DR: 677C/677C, 37.3%; 677C/677T, 49.3%; and 677T/677T, 13.4%; χ² test, P = 0.78). In addition, there was no difference in the allele frequency between the NPDR and no DR group (χ² test, P = 0.33). After adjustment for duration of diabetes, HbA_1c level, and blood pressure, multiple regression analysis also showed no significant correlation between the MTHFR gene polymorphism and diabetic retinopathy (P = 0.98).

The discrepancy between our results and those reported by Maeda et al. is unclear, but fewer subjects and those without DR in their study may explain their conclusions. Other risk factors for DR such as overt nephropathy might be included in their study because the authors did not determine the presence of overt proteinuria; instead, they selected subjects according to their serum creatinine level, although they excluded renal failure. We selected our subjects not only by serum creatinine levels but also for the presence of overt proteinuria, and therefore excluded overt nephropathy. Although associations between hyperhomocysteininemia (2) or defective homocysteine metabolism (3) and risk of diabetic retinopathy have been reported, we conclude that the MTHFR gene polymorphism cannot be a predictive marker for diabetic retinopathy.