Association between decreased thyroid stimulating hormone and hyperuricemia in type 2 diabetic patients with early-stage diabetic kidney disease

The study protocol was approved by the Medicine and Pharmacy Ethics Committee of Beijing Chao-Yang Hospital, Capital Medical University. Written informed consent was collected from each subject.

The participants aged from 18 to 80 years were recruited consecutively from May 2017 to March 2019 in a group of outpatients at the Department of Endocrinology, Beijing Chao- Yang Hospital, Capital Medical University, Beijing, China.

Patients were diagnosed with T2DM according to the World Health Organization (WHO) criteria, and were diagnosed with early-stage DKD as defined by the consensus on DKD by Chinese Medical Association as urinary albumin-to-creatinine ratio (UACR) in 30 ~ 300 mg/g [17]. Subjects were selected in line with both results of UACR from two urine samples collected on different days in 1 month were in 30 ~ 300 mg/g. Hyperuricemia was defined as the SUA level > 420 μmol/L in males or > 360 μmol/L in females [18].

People with primary glomerulonephritis or kidney diseases caused by secondary conditions other than diabetes, with gout or medicines that influence UA metabolism (including diuretics, sodium-dependent glucose transporter 2 inhibitors, UA lowering agents, etc.), with history of thyroid diseases or medicines that influence thyroid function, with infection, malignancies, autoimmune disease, hypertension, cardiovascular disease or pregnancy, and drinking alcohol in last 2 weeks were excluded from all groups.

People with type 1 diabetes or other specific types of diabetes as defined by the WHO classification of diabetes mellitus were excluded from the study. Moreover, people with recent acute diabetic complications, including ketoacidosis, hyperosmolar non-ketotic diabetic coma and lactic acidosis, were also excluded from the study. In addition, none of the control participants had a history of glycometabolism abnormality.

Two hundred fifty-four type 2 diabetic patients with early-stage DKD were enrolled for the study, and were further divided into high SUA group (n = 101) and normal SUA group (n = 153). Eighty-five control subjects were recruited as control group.

The subject’ health status and medical history were obtained using a standard questionnaire (see Additional file 1) and medical records via the face-to-face surveys. Subjects wore only underwear for weight and height measurements. Body mass index (BMI) was determined using weight (kg) / [height (m)]². Blood pressures were collected from the non-dominant arm after 5 min of subjects sitting using a same calibrated standard mercury sphygmomanometer.

Three tubes of serum [3 ml for each tube, for total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), fasting blood glucose (FBG), creatinine (CR) and SUA, for fasting insulin (FINS), and for FT3, FT4 and TSH, respectively] and one tube of plasma [2 ml, for glycated hemoglobin (HbA1c)] from each subject were collected in the morning after an 8-h overnight fast. Random spot urine samples were obtained after 24 h without physical exercise. UACR was determined as the mean of measurements of two urine samples obtained on different days in 1 month.

FBG level was determined by the hexokinase method, TG level was determined by the glycerol phosphate oxidase-peroxidase method, HDL-C and LDL-C levels were determined by the direct method, TC and serum CR levels were determined by the enzymatic method, and SUA level was determined by the uricase method, using a Dade-Behring Dimension RXL Autoanalyzer (Dade Behring Diagnostics, Marburg, Germany). FINS was measured by the electrochemiluminescence method using a Beckman Access 2 (Fullerton, CA, USA). HbA1c was measured by high-performance liquid chromatography on a HLC-723G7 analyzer (Tosoh Corporation, Japan). Insulin resistance was calculated with the equation: homeostasis model assessment of insulin resistance (HOMA-IR) = FINS (mIU/L) × FBG (mmol/L) / 22.5. The estimated glomerular filtration rate (eGFR) was calculated using the Chinese modification of diet in renal disease (MDRD) equation: eGFR (ml/min/1.73 m²) = 175 × (serum creatinine) ^-1.234 × (age)^-0.179 × (0.79 for female) [14]. Urinary creatinine was detected by enzymatic method (BNProSpec, SIEMENS, Germany); urinary albumin was measured by scattering turbidimetry (Abbot c16000, Abbott Diagnostics, IL, USA). FT3, FT4 and TSH were measured by electrochemiluminescence immunoassay using an Abbott Architect i2000 (Abbott Diagnostics, IL, USA). The reference intervals for FT3, FT4 and TSH were 2.30 ~ 4.20 pg/ml, 0.89 ~ 1.76 ng/dl and 0.55 ~ 4.78 μIU/ml, respectively.

All operations were performed by the same well-trained group, and all obtained data were inspected by the principal investigator. Blood and urine analyses were conducted twice in the Central Laboratory of Beijing Chao-Yang Hospital, Capital Medical University, and finally the average values were taken. The factors that might interfere with the measurement results were governed by the laboratory technologists. The laboratory reagents and instruments were inspected using the quality control samples before the samples of participants were run. Moreover, if the findings were outside the reference intervals, the measurements would be checked and repeated by laboratory technologists.

All data were analyzed using Statistical Package for Social Sciences version 20.0 (SPSS, Inc., Chicago, IL, USA). The normality of the data distribution was checked by the Kolmogorov-Smirnov test. Normally distributed data were given as the means ± standard deviations, and non-normally distributed data were expressed as medians with 25th and 75th percentiles. Comparisons of the clinical parameters in three groups were analyzed by one-way ANOVA (normally distributed data) or Kruskal-Wallis H test (non-normally distributed data). Proportions were analyzed by the chi-squared test. Since TSH was non-normally distributed, the association between TSH and SUA in type 2 diabetic patients with early-stage DKD was examined using Spearman’s correlation coefficient analysis. Hierarchical multivariate linear and logistic regressions were adopted, adjusting for established and potential confounding factors, to explore the relationship between TSH and SUA, and the correlation between TSH and hyperuricemia in type 2 diabetic patients with early-stage DKD. The basic adjusted model adjusted for age and gender, and the fully adjusted model additionally adjusted for BMI, SBP, TG, FBG, FINS, eGFR, and UACR. All statistical tests were two-tailed, and a P-value < 0.05 was considered as statistical significance for the findings. However, P < 0.017 (0.05 divided by the times of comparison) was used to indicate statistical significance for the multiple comparison.

The study cohort included 254 type 2 diabetic participants with early-stage DKD consisting of high SUA group and normal SUA group, and 85 control participants as control group. The clinical characteristics of all study participants were listed in Table 1. The participants in three groups were similar in gender, age, systolic blood pressure (SBP), diastolic blood pressure (DBP) and FT3 (P > 0.05 for all). A significant trend was observed for BMI, TC, HDL-C, LDL-C, TG, FBG, HbA1c, FINS, HOMA-IR, CR, eGFR, UACR, FT4, TSH and SUA in three groups (P < 0.01 for all). High SUA group had the significantly increased levels of BMI, TG, FBG, HOMA-IR, CR and FT4, and the significantly decreased level of eGFR compared with normal SUA group and control group (P < 0.017 for all). The levels of BMI, TG, FBG and HOMA-IR were significantly elevated in normal SUA group than those in control group (P < 0.017 for all), and no significant difference was detected in CR, eGFR and FT4 between normal SUA group and control group (P > 0.017 for all). Both high SUA group and normal SUA group exhibited the significantly higher levels of TC, LDL-C, HbA1c, FINS and UACR, and the significantly lower level of HDL-C than control group (P < 0.017 for all), and there was no significant difference in these parameters between high SUA group and normal SUA group (P > 0.017 for all). Furthermore, compared with normal SUA group and control group, high SUA group was with a higher SUA level (Fig. 1) and a lower TSH level (Fig. 2) (P < 0.017 for both). However, no significant difference was observed in SUA and TSH between normal SUA group and control group (P > 0.017 for both). The decreased TSH in high SUA group might indicate the possible association between thyroid function and uric acid metabolism in type 2 diabetic participants with early-stage DKD.

Then, the Spearman’s analysis was performed between SUA and TSH in 254 type 2 diabetic participants with early-stage DKD including individuals in high SUA group and normal SUA group. As exhibited in Fig. 3, SUA was negatively correlated with TSH [r = − 0.35, 95% confidence interval (CI): − 0.43 to − 0.21, P < 0.001] in type 2 diabetic participants with early-stage DKD, suggesting that decreased TSH was associated with elevated SUA in type 2 diabetic participants with early-stage DKD.

Furthermore, to explore the role of TSH in uric acid metabolism, the hierarchical multiple regression analysis was performed to determine whether TSH was independently correlated with SUA in type 2 diabetic participants with early-stage DKD. As presented in Table 2, the results showed that TSH was negatively associated with SUA in unadjusted model (β = − 39.52, 95% CI: − 47.35 to − 27.69, P < 0.001), and this association remained significant after adjustment for age and gender (β = − 34.91, 95% CI: − 44.86 to − 24.95, P < 0.001), as well as additionally adjusting for BMI, SBP, TG, FBG, FINS, eGFR, and UACR (β = − 25.69, 95% CI: − 33.38 to − 18.00, P < 0.001) in type 2 diabetic participants with early-stage DKD.

Subsequently, to further clarify the role of TSH in uric acid metabolism, the hierarchical logistic regression of the association between TSH and hyperuricemia in type 2 diabetic patients with early-stage DKD was analyzed, and the results were shown in Table 3. The unadjusted odds ratio (OR) for the association between TSH and hyperuricemia was 1.57 (95% CI: 1.21 to 2.04, P = 0.001), and the association retained significant after adjustment for age and gender (OR = 1.66, 95% CI: 1.27 to 2.17, P < 0.001) in type 2 diabetic patients with early-stage DKD. Notably, decreased TSH with an OR of 1.73 (95% CI: 1.22 to 2.47, P = 0.002) was associated with hyperuricemia in type 2 diabetic patients with early-stage DKD after adjustment for age, gender, BMI, SBP, TG, FBG, FINS, eGFR, and UACR.