OBJECTIVE: To determine whether serum thyrotropin measurement performed at diagnosis of diabetes mellitus or at initial patient contact predicts subsequent development of hypothyroidism. METHODS: We retrospectively reviewed the computerized records of patients attending annual visits between January 2008 and December 2008 at a hospital diabetes mellitus clinic. Serum free thyroxine and thyrotropin at current and baseline annual visits were documented. A Cox regression model was used to analyze the relationship between development of thyroid dysfunction and patient characteristics including age, sex, type of diabetes, and baseline serum thyrotropin concentration. Kaplan-Meier survival curves were generated for predictors of hypothyroidism. RESULTS: Clinical records of 1101 patients were reviewed (595 men [54%] and 506 women [46%]). Mean age was 60.0 ± 17 years. Two hundred twenty-three patients (20.3%) had type 1 DM and 878 (79.7%) had type 2 diabetes. Thyroid dysfunction was present in 136 patients (12.4%) at baseline and developed in 71 patients (6.4%) at follow-up (median duration, 37 months). Overt and subclinical hypothyroidism developed in 28 (2.5%) and 38 (3.5%) patients, respectively. Incident hypothyroidism was associated with baseline thyrotropin concentration greater than 2.2 mIU/L (relative risk, 10.4; confidence interval, 5.6-19.6; P<.001) and female sex (relative risk, 1.8; confidence interval, 1.1-2.9; P = .007). The predictive influence of sex was abolished in patients with a thyrotropin value greater than 2.2 mIU/L. This TSH threshold yielded an optimal sensitivity and specificity of 83% and 72%, respectively, for predicting hypothyroidism. CONCLUSIONS: Baseline serum thyrotropin predicted hypothyroidism in patients with diabetes mellitus even at thyrotropin concentrations within the reference range. Selective annual thyroid screening in diabetic patients with baseline thyrotropin concentrations greater than 2.2 mIU/L may be more cost-effective than universal screening.
OBJECTIVE: To determine whether serum thyrotropin measurement performed at diagnosis of diabetes mellitus or at initial patient contact predicts subsequent development of hypothyroidism. METHODS: We retrospectively reviewed the computerized records of patients attending annual visits between January 2008 and December 2008 at a hospital diabetes mellitus clinic. Serum free thyroxine and thyrotropin at current and baseline annual visits were documented. A Cox regression model was used to analyze the relationship between development of thyroid dysfunction and patient characteristics including age, sex, type of diabetes, and baseline serum thyrotropin concentration. Kaplan-Meier survival curves were generated for predictors of hypothyroidism. RESULTS: Clinical records of 1101 patients were reviewed (595 men [54%] and 506 women [46%]). Mean age was 60.0 ± 17 years. Two hundred twenty-three patients (20.3%) had type 1 DM and 878 (79.7%) had type 2 diabetes. Thyroid dysfunction was present in 136 patients (12.4%) at baseline and developed in 71 patients (6.4%) at follow-up (median duration, 37 months). Overt and subclinical hypothyroidism developed in 28 (2.5%) and 38 (3.5%) patients, respectively. Incident hypothyroidism was associated with baseline thyrotropin concentration greater than 2.2 mIU/L (relative risk, 10.4; confidence interval, 5.6-19.6; P<.001) and female sex (relative risk, 1.8; confidence interval, 1.1-2.9; P = .007). The predictive influence of sex was abolished in patients with a thyrotropin value greater than 2.2 mIU/L. This TSH threshold yielded an optimal sensitivity and specificity of 83% and 72%, respectively, for predicting hypothyroidism. CONCLUSIONS: Baseline serum thyrotropin predicted hypothyroidism in patients with diabetes mellitus even at thyrotropin concentrations within the reference range. Selective annual thyroid screening in diabeticpatients with baseline thyrotropin concentrations greater than 2.2 mIU/L may be more cost-effective than universal screening.
Authors: A Giandalia; G T Russo; E L Romeo; A Alibrandi; P Villari; A A Mirto; G Armentano; S Benvenga; D Cucinotta Journal: Endocrine Date: 2014-01-03 Impact factor: 3.633
Authors: G Bellastella; M I Maiorino; L Scappaticcio; O Casciano; M Petrizzo; M Caputo; V A Paglionico; D Giugliano; K Esposito Journal: J Endocrinol Invest Date: 2017-08-30 Impact factor: 4.256