Patients aged between 6–18 years, who had been admitted to the Pediatric Endocrinology Outpatient Clinic of Bursa Yuksek Ihtisas Hospital with the complaint of being overweight between 1 October 2018 and 30 March 2019 and whose BMI >95th percentile for age and sex were included in the study. The study protocol was approved by the ethics committee of the Bursa Yuksek Ihtisas Training and Research Hospital (2011KAEK-25 2019/03-05) in accordance with the principles of the Helsinki Declaration. Data regarding age, sex, puberty, anthropometric measurements, and laboratory and radiology results were retrospectively collected from patient files. Patients who had missing data, whose thyroid auto-antibodies (antithyroid peroxidase, antithyroglobulin) were positive, whose TSH was >10 mIU/L and fT4 was low according to age, whose thyroid ultrasonography results were not normal, who had chronic diseases or medication use and secondary obesity, were excluded from the study. A total of 215 obese patients, the complete data of whom could be reached, whose thyroid auto-antibodies were negative and thyroid ultrasonography results were normal, were included in the study. Weight (kg), height (cm), BMI (kg/m2), BMI standard deviation score (SDS), puberty status and waist circumference of the patients were recorded. Laboratory analyses included TSH, fT4, fasting plasma glucose, homeostasis model assessment for insulin resistance (HOMA-IR), triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), aspartate aminotransferase, and alanine aminotransferase values. The patients were allocated to 2 groups according to thyroid function test results. Subjects whose fT4 level was normal and TSH level was 5–10 mIU/L were enrolled in the SH group. The subjects whose fT4 level was normal and TSH level was <5 mIU/L were enrolled in the control group. Both groups were compared with regard to fasting plasma glucose, insulin, HOMA-IR, and lipid profile.

The weight measurements of all patients were carried out using the Tess BM 300 device (Comak tarti, Istanbul, Turkey), at 0.1-kg sensitivity and on an empty stomach, without shoes and with daily clothes. The height measurements were made with a stadiometer with an error of 1 mm with bare feet, standing upright, and the shoulder and gluteal area touching the wall. All measurements were made between 9:00 AM and 16:00 PM by the same person. BMI was calculated by using the formula: body weight/height² (kg/m²).

The serum fasting blood glucose, TC, TG, HDL-C, LDL-C, and insulin levels were tested after 12 hours of fasting. The serum blood glucose and lipids were tested using the standard methods with the ‘Cobas 8000’ (Roche Diagnostics Turkey, Istanbul, Turkey) equipment. Insulin was measured using the standard method with the ‘Cobas 8000’ equipment. Insulin resistance was calculated using the HOMA-IR formula (fasting blood glucose [mg/dL]×fasting insulin [µIU/mL]/405). The serum TSH (reference range, 0.5–5 mIU/L) and fT4 (reference range, 0.8–1.9 ng/mL) were analyzed using chemiluminescence immune assays method.

Statistical analyses were performed using the IBM SPSS Statistics ver. 21.0 (IBM Co., Armonk, NY, USA). The categorical variables were analyzed using the chi-square test. The Student t test and the Mann-Whitney U test were used for comparison of groups according to the parametric values. For correlation analyses, the Pearson test was used for the parametric variables and the Spearman test was used for the nonparametric variables. A P value <0.05 was accepted as statistically significant.

A total of 215 patients were included in the study. There were 141 females (66%) and 74 males (34%). Mean ages of subjects were 12.33±2.86 years. The anthropometric and laboratory data of the patients have been summarized in Table 1.

SH was determined in a total of 77 patients (36%). There were 138 patients (64%) in the control group whose thyroid function test results were normal. No difference was found between the groups with regard to sex distribution (P=0.65). The puberty status was similar in both groups (P=0.92). While the mean BMI SDS was 2.47±0.67 in the SH group, it was 2.55±0.57 in the control group. Both groups were similar with regard to BMI SDS (P=0.16). When groups were compared with regard to glucose metabolism, the fasting insulin and the HOMA-IR were found to be higher in the SH group (P=0.007 and P=0.004, respectively). When the groups were compared with regard to lipid metabolism, while the HDL-C level was found to be lower, the TG level was seen to be higher in the SH group than control group (P=0.02 and P=0.01, respectively). The anthropometric and laboratory measurements of both groups have been summarized in Table 2.

There was a positive correlation between TSH level and some metabolic parameters such as insulin (r=0.20, P=0.003), HOMA-IR (r=0.20, P=0.003), and TG (r=0.21, P=0.001). There was no correlation between TSH level and BMI and waist circumference. The correlation between TSH and metabolic parameters has been shown in Figs. 1–3.