study the effect of decreasing the upper limit of reference range of TSH levels in diabetic patients without known dysthyroïdies on the prevalence of diabetic complications.Design and method:
Data were collected in diabetic out-patients without known dysthyroïdies in whome TSH levels were greater than 0.4mIU/L. TSH levels at the upper reference range (URR) were defined between 2.5 and 4 mIU/L. All statistical analyses were conducted with SPSS software. P < 0.05 was considered significant.Results:
Normal TSH values were observed in 76 diabetic patients, in whom 25 have TSH at URR. Subclinical hypothyroidism was noted in 18 patients. When TSH levels were at URR, we observed intermediate frequencies in hyperlipidemia, retinopathy, and nephropathy. Binary logistic regression analysis showed significant association between TSH levels compared to 2.5 mIU/L and hyperlipidemia*retinopathy interaction.Conclusions:
An independent association seemed to be present between TSH levels greater than 2.5mIU/L and diabetic complications. A systematic thyroid function tests, with a lowered upper limit of reference range, would be discussed in the management of diabetic patients.