All questionnaires and medical procedures were approved by the Institutional Review Board of Fudan University (Shanghai, China). Written informed consent was prospectively obtained from all study participants.

A total of 59 patients (age range, 18–60 years), diagnosed with type I HT at the Fifth People's Hospital of Shanghai, Fudan University between March and November 2013, were enrolled in the present study and assigned to the HT experimental group. An additional 38 healthy age- and gender-matched volunteers were recruited to the control group. The diagnosis criteria for type I HT were as follows: Normal thyroid function and characteristic clinical manifestations associated with elevated serum anti-thyroid peroxidase antibodies (anti-TPO) or anti-thyroglobulin antibodies (anti-Tg) at ≥60 U/ml (23). The clinical manifestations were detected twice for each patient, with a 1-week interval. The exclusion criteria were as follows: Patients with a previous and/or new diagnosis of DM, determined by oral glucose tolerance test (OGTT), patients with a confirmed diagnosis of hyperthyroidism, and those who received anti-thyroid drugs or thyroid hormone replacement therapy. In addition, patients with cardiovascular and cerebrovascular diseases, severe liver or kidney dysfunction, malignant tumors, severe mental disorders, were currently taking glucocorticoids, or were pregnant or breastfeeding were excluded. The diagnoses of DM, impaired glucose regulation and hypertension were made according to the guidelines approved by the World Health Organization (24).

Questionnaires containing demographic information, medical history of hypertension, weight, height, body mass index (BMI), and waist and hip circumferences were administered to all study participants. An OGTT was performed using fasting blood samples and blood samples collected 30 and 120 min following drinking 75 g glucose dissolved in 250–300 ml water. The levels of plasma glucose, serum lipid, C-peptide and insulin were measured, and thyroid function was assessed. The insulin level in the blood at the time points of 0, 30, and 120 min were designated as Ins0, Ins30 and Ins120, respectively. In addition, the plasma glucose levels at those time points were designated as Glu0, Glu30 and Glu120, with the average glucose and insulin levels in the OGTT designated as GluAve and InsAve, respectively. The area under the curve (AUC) for insulin and glucose (InsAuc and GluAuc, respectively) were determined by the insulin secretion curve adjusted for glucose.

The plasma glucose levels were measured using a glucose oxidation assay kit (Shanghai Institute of Biological Products Co., Ltd. Shanghai, China), using a Beckman Glucose Lab Analyzer 2 (Model 6517; Beckman Coulter, Inc., Danvers, MA, USA). The serum lipid contents, including total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C), were measured using an enzymatic colorimetric analyzer (Hitachi 7600 Clinical Analyzer; Hitachi, Ltd., Tokyo, Japan). The serum levels of C-peptide and insulin, and the thyroid function indicators total triiodothyronine (TT₃), total thyroxine (TT₄), free triiodothyronine (FT₃), free thyroxine (FT₄), thyroid stimulating hormone (TSH), anti-Tg and anti-TPO were measured using specific radioimmunoassay kits purchased from Linco Research, Inc. (St. Charles, MO, USA) in a Beckman immunoassay analyzer (Beckman Coulter, Inc.).

To assess β-cell function and insulin sensitivity, the following parameters were measured: i) homeostasis model assessment of β-cell function (HOMA-β) [20 × Ins0 (µU/ml)]/[Glu0 (mmol/l)-3.5]; ii) change in insulin and glucose ratio at 30 min OGTT (ΔI30/ΔG30) [Ins30-Ins0 (µU/ml)]/[Glu30-Glu0 (mmol/l)]; iii) early-phase insulin secretion (InsAuc30/GluAuc30) [Ins0+Ins30 (pmol/L)]/[Glu0+Glu30 (mmol/l)]; iv) ratio of insulin to glucose AUC values at 120 min (InsAuc120/GluAuc120) [Ins0+4 × Ins30+3 × Ins120 (pmol/l)]/[Glu0+4′Glu30+3 × Glu120 (mmol/l)]; v) HOMA of insulin resistance (HOMA-IR) calculated as [Ins0 (µU/ml)x Glu0 (mmol/l)/22.5]; vi) disposition index (DI) calculated as [HOMA-β/HOMA-IR]; vi) Matsuda index (ISI_M) of insulin sensitivity [10,000/((Glu0 (mg/dL) × Ins0 (µU/ml) × GluAve (mg/dl) × InsAve (µU/ml))^1/2]; vii) DI at 30 min OGTT (DI30) [InsAuc30/GluAuc30 × ISI_M]; and viii) DI at 120 min OGTT (DI120) [InsAuc120/GluAuc120 × ISI_M].

Venous blood was obtained from the study participants by EDTA-anticoagulation. For examination of different lymphocyte populations, two 100-µL blood aliquots were incubated with CD4-FITC/CD8-PE-Cy5/CD28-PE and CD19-PE-Cy5/CD24-PE/CD27-APC/CD38-FITC (BD Biosciences, San Jose, CA, USA) for 20 min at room temperature in the absence of light. The stained samples were treated with Simultest™ (Beckman Coulter, Inc.) to lyse erythrocytes. The samples were washed with PBS twice prior to flow cytometric analysis. The following isotype controls were used: IgG1-FITC, IgG1-PE-Cy5, IgG2a-PE and IgG1-APC. For examination of intracellular IL-10, 100 µl whole blood was cultured in RPMI 1640 supplemented with 10% heat inactivated fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA), 25 µg/l phorbol 12-myristate 13-acetate, 1 mg/l ionomycin and 20 µg/l monensin (Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) at a density of 2×10⁶ cells/ml. Following incubating for 4–6 h at 37°C and in 5% CO₂, CD19-Cy5 (BD Biosciences) was added to the cells for 30 min at room temperature. Subsequently, the cells were permeabilized according to the manufacturer's protocol (Beckman Coulter, Inc.), followed by incubation with the monoclonal antibodies IL-10-PE (20 µl, cat. no. JES3-9D7; BD Biosciences) or isotype control IgG2a-PE (BD Biosciences) for 30 min at room temperature. The stained cells were then analyzed on a FACSCalibur flow cytometer (BD Biosciences). Lymphocytes were collected to measure the percentages of CD8⁺, CD8⁺CD28⁻, CD19⁺CD24⁺CD27⁺, CD19⁺CD24^hiCD38^hi and CD19⁺IL-10⁺ (B10) cell subpopulations.

Data are expressed as the mean ± standard deviation or percentage. Comparison of continuous data between groups was performed using either a parametric test, such as Student's t-test and one-way analysis of variance), or a nonparametric test (Mann-Whitney U or Kruskal-Wallis H tests). The comparison of categorical data between groups was performed using a χ² test. Pearson's correlation analysis was used to investigate the association between different parameters. P<0.05 (two-tailed) was considered to indicate a statistically significant difference. SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) was used to analyze data.

Laboratory measurements were recorded in patients with type I HT (HT group; n=59) and healthy volunteers (normal group; n=38). The results are summarized in Table I. No significant differences were found in the mean age (47.2±14.7, vs. 40.6±16.9 years) or female-to-male ratio between the HT and control groups. Average hip circumference was similar between the two groups, however, a marginal but significant increase in waist circumference was observed in the HT group (P<0.05). Abdominal circumference has previously been associated with diabetes and cardiovascular disease, suggesting that those in the HT group may be predisposed to such risk factors. Biochemical parameters in the blood, including those for pancreatic and thyroid functions, were also measured. No significant differences were observed in blood pressure, BMI, lipid profiles (TC, TG, HLD-C and LDL-C), thyroid indicators (TT₃, TT₄, FT₃, FT₄ and TSH) or diabetes-associated autoantibodies (GADA, IAA and ICA) between the two groups. The incidence of impaired glucose regulation was also similar between the two groups (Table I). Consistent with the diagnosis, the levels of anti-Tg and anti-TPO, two specific indicators for HT, were significantly higher in the HT group, compared with those in the control group (P<0.05). The level of C-peptide at 30 min was significantly different between the groups (Fig. 1B). In addition, OGTT was performed to determine insulin resistance, and the results showed markedly elevated insulin levels at 30 and 120 min following glucose intake in the HT group (Fig. 1C).

As shown in Table II, the parameters for HOMA-β, insulin response (ΔI30/ΔG30), and estimated insulin resistance in the fasting state (HOMA-IR) and postprandial state (ISI_M) showed no significant differences between the HT and control groups. Furthermore, the early-phase DI (DI30) and the total DI (DI120) were similar between the two groups. However, the early-phase and total insulin secretions (InsAuc30/GluAuc30 and InsAuc120/GluAuc120, respectively) were markedly higher in the HT group, compared with those in the control group. In addition, the basal insulin DI, which reflected the capacity of β-cell compensation function for IR in the fasting state, was significantly lower in the HT group (P<0.05). These results indicated that the patients with type I HT had increased insulin secretion in the postprandial state. No significant differences were found in the fasting glucose or insulin secretion, indicated by plasma C-peptide and insulin differences between the HT and control groups (Fig. 1), however, the fasting β-cell compensation function (indicated by DI) was markedly reduced in the HT group (Table I).

To analyze lymphocyte populations, peripheral blood was obtained from patients in the HT group and the healthy controls. The percentages of different lymphocyte populations were measured using flow cytometry. No significant differences in the overall percentage of CD4⁺ T cells or the CD8⁺CD28⁻ TReg cell subset were observed between the HT and control groups (Fig. 2A, upper left and lower panels). However, patients in the HT group had an increased percentage of CD8⁺ T cells, compared to the control group (P<0.05; Fig. 2A, upper right panel).

To assess the percentages of different B cell subsets, cells were stained with specific lineage markers. As shown in Fig. 2B, the percentage of the specific Breg cell subset, CD19⁺CD24⁺CD27⁺, was significantly higher in the HT group (P<0.05) (upper right panel). However, no differences in the other regulatory Breg cell subset (CD19⁺CD24^hiCD38^hi) or the B10 cells were found between the HT and control groups (lower panels).

To determine the correlation between the enhancement of Bregs with insulin sensitivity, the study subjects were divided into CD19⁺CD24⁺CD27⁺ Breg enhancer and Breg control groups. A reference range for different subsets of peripheral lymphocytes was set as the interquartile range (25–75th percentile) of the percentages of lymphocyte subsets in the control group. According to the reference range, all study subjects were divided into two groups: Those with a higher percentage of CD19⁺CD24⁺CD27⁺ Breg cells, compared with the reference range (>7.4325%; elevated group), and those with a percentage of CD19⁺CD24⁺CD27⁺ Breg cells within the reference range (≤7.4325%; normal group). Pancreatic β-cell function and insulin sensitivity were compared between the two groups. As shown in Fig. 3, HOMA-β and HOMA-IR were significantly increased in the elevated group, indicating that the insulin secretory capacity of β-cells and the insulin resistance in the fasting state were enhanced in individuals with an elevated level of CD19⁺CD24⁺CD27⁺ Breg cells. However, no differences were observed in the insulin secretion index in the postprandial state (InsAuc30/GluAuc30 and InsAuc120/GluAuc120) between groups.

To investigate whether the elevated level of CD19⁺CD24⁺CD27⁺ Breg cells was associated with alterations in insulin secretion and insulin sensitivity, correlation analysis was performed. The results demonstrated that the CD19⁺CD24⁺CD27⁺ Bregs were positively correlated with HOMA-IR (r=0.338; P=0.029) and HOMA-β (r=0.254; P=0.005; Fig. 4).

Insulin resistance is commonly observed in patients with HT accompanied with clinical or subclinical hypothyroidism (1). Previous investigations to elucidate the mechanism of insulin resistance in patients with HT have yielded inconclusive results. The present study demonstrated for the first time, to the best of our knowledge, the action of insulin in patients with type I HT with preserved normal thyroid function. The results indicated that these patients had insufficient insulin secretion following adjusting insulin resistance (DI) and increased insulin secretion in the postprandial state. An elevated level of CD19⁺CD24⁺CD27⁺ Bregs was observed in these patients, which was positively correlated with insulin secretion and insulin resistance in the fasting state. These results revealed a close association between immune dysregulation and insulin resistance in type I HT.

The observation of a normal blood glucose level and elevated insulin level, described as hyperinsulinemia, indicates the presence of insulin resistance. Insulin resistance contributes to the development of various medical conditions, including type 2 DM, metabolic syndromes, obesity, hypertension, dyslipidemia, elevated free fatty acid and a deregulated stress responses (25). In the present study, the levels of plasma glucose remained normal during OGTT in the HT group. However, the early-phase and total insulin secretion increased significantly, compared with the levels in the control group, indicating the possibility of postprandial insulin resistance. The fasting insulin secretion and insulin resistance levels in the HT group were normal, however, fasting β-cell compensation function was impaired, as indicated by a significant reduction in DI. DI reflects the ability of β-cells to compensate for insulin resistance; a reduced DI indicates impaired compensation. When insulin resistance increases, insulin secretion consequently increases. However, if the insulin secreted from β-cells is insufficient to compensate for insulin resistance, then the DI decreases, indicating impaired β-cell compensation function (26). In the present study, the fasting DI was significantly decreased in the HT group, compared with that in the control group, suggesting that the patients with HT had impaired β-cell compensation function and that insulin levels were not sufficient to compensate for the insulin resistance.

Previously, immune components, including CD8⁺ cytotoxic T lymphocytes, proinflammatory T cells, and Tregs, have been shown to be important in the pathophysiology of autoimmune HT. Notably, the present study found no significant differences in CD4⁺ or CD8⁺CD28⁻ T cells, but found a decrease in the overall percentage of CD8⁺ T cells. However, previous studies have shown that the infiltration of proinflammatory T cells and overproduction of inflammatory mediators are largely present in the liver and muscle tissues of diabetic patients, which have been implicated in the development of insulin resistance (27,28). Under normal circumstances, the activation and proliferation of these proinflammatory T cells are strictly regulated by their interplay with Tregs, thus the immune balance is carefully maintained.

The present study is the first, to the best of our knowledge, to report on the involvement of CD19⁺CD24⁺CD27⁺ Bregs in HT and associated insulin resistance. The patients with HT had reduced DI, and the elevated CD19⁺CD24⁺CD27⁺ Breg population was positively correlated with fasting insulin secretion and insulin resistance. Although the regulatory mechanism of Bregs remains to be elucidated, studies in diabetic patients may provide insight to facilitate understanding of HT. Rather than producing the anti-inflammatory cytokine IL-10, Bregs in diabetic patients upregulate the expression of proinflammatory cytokines, including IL-6 and IL-8, and promote inflammation (29,30). In the present study of type I HT, the percentage of CD19⁺CD24⁺CD27⁺ Bregs was elevated significantly, compared with that in the healthy controls. This increase in CD19⁺CD24⁺CD27⁺ Bregs may also increase the expression of IL-6 and IL-8 in HT, promoting increased autoimmune inflammation, and thereby increasing risk of insulin resistance and decreased β-cell function. There may also be a feedback mechanism by which the Bregs are activated and proliferate in response to the tissue damage caused by infiltrating inflammatory cells. The causal association between Breg dysregulation and insulin resistance remains to be elucidated. Further investigation is required to define the exact role of Bregs in the development of HT.

In the patients with type I HT examined in the present study, insulin resistance, indicated by increased early-phase and total insulin secretion, but a normal glucose response was observed during OGTT. The increase in the percentage of CD19⁺CD24⁺CD27⁺ Breg cells was correlated with insulin resistance in the fasting state and homeostatic β-cell function. However, no significant association was found between the postprandial secretion of insulin and alterations in the CD19⁺CD24⁺CD27⁺ Breg population. These results indicated potential contributions from other cellular components or immune factors acting during the later phases of insulin sensitivity, which functioned with Bregs to regulate insulin signaling in the postprandial state.

The patients with HT examined in the present study showed increased postprandial insulin levels, but normal fasting insulin levels. A possible explanation for these differences may be due to the primary tissues involved in the action of insulin. Fasting insulin resistance is associated with decreased glucose uptake and utilization under the regulation of insulin in the liver, whereas postprandial insulin resistance is associated with reduced glucose uptake and utilization in muscle and adipose tissues. The required insulin level for glucose uptake in muscle or adipose tissues is substantially higher, compared with the level required in the liver. Therefore, fasting and postprandial insulin levels are increased in compensation in individuals with insulin resistance, although the postprandial insulin level is increased to a higher degree.

In conclusion, the present study demonstrated an increase in postprandial insulin secretion and impaired fasting β-cell compensation function, indicated by reduced DI, in patients with type I HT. An increased percentage of CD19⁺CD24⁺CD27⁺ Breg cells was found in these patients, which was closely associated with β-cell function and insulin resistance in the fasting state. These results may assist in the development of novel therapeutic strategies, which target the fundamental immune components involved in the pathogenesis of HT.