The aim of this study was to examine the relation between serum total homocysteine concentrations and microvascular complications in Pima Indians with Type 2 diabetes.

Homocysteine concentrations were measured in frozen sera of 396 diabetic participants in a longitudinal study who were 40 years of age or older and who had attended one or more examinations between 1982 and1985. Retinopathy was assessed by fundoscopy and nephropathy by an albumin:creatinine ratio greater than 300 mg/g. The incidence rate ratio for a 5 µmol/l difference in homocysteine was calculated using proportional hazard regression.

The incidence of each complication was assessed in subjects without that complication at baseline and with more than one follow-up examination: 229 for nephropathy, 212 for retinopathy and 266 for proliferative retinopathy. There were 101 incident cases of nephropathy, 113 of retinopathy and 40 of proliferative retinopathy during a mean follow-up of 8.6, 7.5 and 8.9 years, respectively. Incidence of nephropathy was associated with homocysteine concentrations: IRR=1.42 (95% CI, 1.09–1.84, p=0.01); this remained statistically significant controlled for age, sex and duration of diabetes (p=0.03), but not when controlled for baseline renal function (p=0.4). Homocysteine concentrations were not associated with the incidence of any retinopathy IRR=1.14 (95%CI 0.89–1.46, p=0.3) but were associated with the incidence of proliferative retinopathy IRR=1.62 (95% CI 1.16–2.28, p=0.005); this association remained statistically significant when controlled for baseline renal function and diabetes duration (p=0.02).

Increased homocysteine concentrations are associated with an increased risk for incidence of nephropathy and proliferative retinopathy; the relation with incidence of nephropathy seems to be explained by an association with baseline albuminuria status concentrations, whereas the relation with incidence of proliferative retinopathy does not.