Should thyroid-stimulating hormone goals be reviewed in patients with Type 1 diabetes mellitus? Results from The Brazilian Type 1 Diabetes Study Group

Aims To investigate if thyroid-stimulating hormone (TSH) levels are associated with any differences in glycaemic control or diabetes-related complications in individuals with Type 1 diabetes. Methods This observational, cross-sectional and multicentre study included patients with Type 1 diabetes for >= 5 years, with a recent TSH measurement and without a known previous thyroid disease. Patients were divided into three groups according to TSH levels: 0.4-2.5 mU/l; 2.5-4.4 mU/l; and >= 4.5 mU/l. Results We included 1205 individuals with a mean +/- SD age of 23.8 +/- 11.3 years. Seven patients had TSH levels <0.4 mU/l and were excluded from the comparison between groups. HbA(1c) levels, systolic and diastolic blood pressure, LDL cholesterol and disease duration were similar in all groups (P = 0.893, P = 0.548, P = 0.461, P = 0.575 and P = 0.764, respectively). The rates of diabetic retinopathy and GFR < 60/mL/min/1.73 m(2) differed between groups (P = 0.006 and P < 0.001, respectively) and were lower in those with lower TSH levels. Multivariate analysis confirmed these associations. The frequencies of retinopathy and GFR < 60 mL/min/1.73 m(2) were higher not only in patients with TSH = 4.5 mU/l (odds ratio 1.878 and 2.271, respectively) but also in those with TSH levels of 2.5-4.4 mU/l (odds ratio 1.493 and 2.286, respectively), when compared with patients with TSH levels of 0.4-2.5 mU/l. Conclusions TSH levels of 0.4-2.5 mU/l are associated with a lower risk of diabetic retinopathy and renal failure in individuals with Type 1 diabetes, independently of glycaemic control and duration of the disease.