Thyroid Function in 35 Patients with Gitelman Syndrome

Objective . In this study, we aimed to analyze thyroid function and related risk factors for thyroid dysfunction in 35 patients with Gitelman syndrome (GS). Methods . This study included 35 patients with GS who were referred to West China Hospital of Sichuan University from Aug 2013 to Jan 2018. General patient characteristics were collected, and thyroid function was assessed. To evaluate the potential contribution of hypokalemia to thyroid dysfunction, 636 patients who were clinically diagnosed with primary aldosteronism (PA) during the same period were included as the control group; these patients were divided into a hypokalemia group ( N = 528 ) and a normokalemia group ( N = 108 ). Logistic regression was used to screen for signi ﬁ cant determinants of thyroid dysfunction in the GS patients. Results . Patients with GS had a signi ﬁ cantly di ﬀ erent prevalence of subclinical hypothyroidism, hypothyroidism, and hyperthyroidism than patients with hypokalemic PA and normokalemic PA (28.6%, 2.9%, and 11.4% vs. 15.5%, 6.1%, and 0.7% vs. 8.3%, 4.6%, and 2.8%, P < 0 : 001 ). No signi ﬁ cant di ﬀ erence was observed in the distribution of thyroid function between the hypokalemic PA group and the normokalemic PA group ( P > 0 : 05 ). No signi ﬁ cant di ﬀ erences were seen in the positive rates of thyrotropin receptor antibody (TRAb), thyroglobulin antibody (TGAb), and thyroid peroxidase antibody (TPOAb) among the three groups ( P > 0 : 05 ). In the logistic regression, only sex (OR, 7.4; 95% CI, 1.555-35.479; P = 0 : 012 ) was signi ﬁ cantly correlated with thyroid dysfunction in GS patients. Conclusion . GS is complicated with a greater rate of thyroid dysfunction than primary aldosteronism. The risk of thyroid dysfunction in female patients with GS is higher than that in male patients.


Introduction
Gitelman syndrome (GS) is an autosomal recessive hereditary disorder that is characterized by hypokalemic metabolic alkalosis, hypomagnesemia, and hypocalciuria [1]. Case reports have demonstrated that GS is associated with thyroid diseases and thyroid dysfunction [2][3][4], as described in a pedigree of patients with GS complicated by thyroid disease or dysfunction, reported by Liu et al. [2]. A 16-year-old teenager (the proband) was diagnosed with GS and Graves' disease [2], and his family members who carried SLC12A3 gene mutations without clinical manifestations of GS were also affected with subclinical hypothyroidism or thyroid autoantibodies [2]. Zhou et al. [3] reported a male patient with a 12year history of Graves' disease (GD) who was finally diagnosed with GS. Aoi et al. [4] reported two GS patients who were diagnosed using genetics and whose diseases were complicated by GD and autoimmune thyroid disease (AITD). Since according to multiple case reports, GS patients experienced thyroid dysfunction; it is necessary to verify this finding in additional GS patients.
It is generally recognized that thyrotoxicosis is the most common cause of hypokalemic periodic paralysis (HPP) in Graves' disease (GD), especially in Asian men [5]. Other studies have reported an association between HPP and other thyroid disorders, such as painless thyroiditis, toxic multinodular goiter (MNG), toxic thyroid adenoma, thyroidstimulating hormone-(TSH-) secreting pituitary adenoma, and primary hypothyroidism [6][7][8]. Therefore, whether the thyroid dysfunction observed in GS is due to hypokalemia requires further investigation. In this study, we aimed to analyze thyroid function in 35 GS cases to determine the potential influence of hypokalemia on thyroid dysfunction in GS compared with primary aldosteronism (PA) and to explore the risk factors of thyroid dysfunction in GS patients.

Materials and Methods
2.1. Patients. The data of patients with GS and PA were retrospectively collected. Patients with clinical symptoms, such as weakness, numbness, muscle pain, nocturia, palpitation, and chest pain, were referred to West China Hospital of Sichuan University from Aug 2013 to Jan 2018. Thirty-seven patients were clinically diagnosed with GS, and of these, nine were genetically diagnosed with GS. Two patients who underwent partial and total thyroidectomy due to thyroid adenoma and papillary thyroid cancer, respectively, were excluded. Thirtyfive patients with GS were finally included in this study. 636 patients who were clinically diagnosed with PA from Dec 2008 to Dec 2018 were included as the control group. To evaluate the potential contribution of hypokalemia to thyroid dysfunction, PA patients with serum potassium levels lower than 3.5 mmol/L and between 3.5 mmol/L and 5.5 mmol/L were classified into the hypokalemic group (N = 528) and the normokalemic group (N = 108), respectively. Clinical data were retrospectively obtained from the medical records.

Ethical
Approval. The study protocol was approved by the Ethical Committee of West China Hospital of Sichuan University (2019-556) and was performed according to the principles in the Declaration of Helsinki.
2.3. Diagnostic Criteria. GS was diagnosed in accordance with the consensus and guidance established in 2017 [9]. Briefly, patients with chronic hypokalemia in the absence of potassium-lowering drugs, inappropriate renal potassium wasting, metabolic alkalosis, hypomagnesemia, hypocalciuria, high plasma renin activity, low or normal-low blood pressure, and normal renal ultrasound findings were clinically diagnosed with GS. Patients who were found to harbor biallelic-inactivating mutations in SLC12A3 were genetically diagnosed with GS. Patients treated with laxatives or diuretics and those who abused alcohol or drugs were excluded. Patients with other extrarenal and renal causes of hypokalemia and transcellular shift of potassium, such as bronchodilator use and familial periodic paralysis, were excluded. Finally, patients with autoimmune diseases, such as Sjögren's syndrome, were also excluded.
PA was diagnosed in accordance with published guidelines [10]. Briefly, patients with a high plasma aldosterone (PAC, ng/dL)/renin (PRA, ng/mL/h) ratio (ARR) required one or more confirmatory tests to definitively confirm or exclude the diagnosis. PA was diagnosed when at least one of the confirmatory tests (captopril challenge test or saline infusion test) was positive.
Clinical hyperthyroidism was defined as the use of antithyroid medications, a history of hyperthyroidism or a TSH level < 0:27 mU/L, and an FT4 level > 22:0 pmol/L. Clinical hypothyroidism was defined as the use of levothyroxine treatment or a TSH level > 4:2 mU/L and an FT4 level < 12:0 pmol/L. Subclinical hypothyroidism was defined as a TSH level > 4:2 mU/L and a normal FT4 level between 12.0 and 22.0 pmol/L without administration of drugs for thyroid disease [11]. A TPOAb level > 34 IU/mL, a TGAb level > 115 IU/mL, and a TRAb level > 3 IU/mL were considered positive [11]. Patients with nonthyroidal illness, temporary changes in thyroid function due to subacute or painless thyroiditis, and thyroidectomy were not included. Patients with a clear history of hyperthyroidism due to Graves' disease were also included in the hyperthyroidism group for analysis, even though they had normal thyroid function or were treated with antithyroid drugs at the time of the study.

Statistical Analysis.
Statistical analyses were performed with SPSS 22.0. For normal distributions, numerical variables were described as the means ± standard deviations (SDs). For nonnormal distributions, numerical variables were presented as medians and quartiles. Categorical variables were described as counts. The Kruskal-Wallis test and one-way ANOVA were used to compare differences in laboratory tests among the three groups. Comparisons were performed using the chi-square test for categorical data. P < 0:05 was considered statistically significant. Univariate logistic regression was used to screen for significant determinants of thyroid dysfunction in GS patients. Factors such as sex, age, and the levels of serum magnesium, serum potassium, serum calcium, serum chlorine, TPOAb, and TGAb were included. Factors with an associated P value less than 0.05 were entered into a stepwise backwards conditional multivariable logistic regression analysis.  Table 1. Their serum potassium (K+) concentration ranged from 1.3 to 2.9 mmol/L with inappropriate renal potassium wasting (15.9-180.5 mmol/day); the serum magnesium (Mg2+) concentration in these patients ranged from 0.2 to 0.9 mmol/L, and 5 patients had a normal serum Mg2+ concentration (normal range 0.7-1.0 mmol/L). Their calcium excretion varied from 0.1 to 9.6 mmol/day (normal range 2.5-7.5 mmol/day). All patients demonstrated high plasma renin activity that ranged from 4.5 to 12.0 ng/mL/h (normal range 0.1-0.8 ng/mL/h). The pH ranged from 7.37 to 7.52.

Results
As shown in Table 2, the mean age of patients in the GS group was 36:6 ± 15:4 years, which was lower than the mean ages of patients in the normokalemic PA group (51:9 ± 11:7 years) and the hypokalemic PA group (48:4 ± 12:1 years) (P < 0:05). As expected, all patients in the GS group had   Table 2. No significant dif-ferences were found in the positive rates of TRAb, TGAb, and TPOAb among the three groups (P > 0:05).

Hypomagnesemia Is Not Associated with Thyroid
Dysfunction in GS Patients. GS patients with serum magnesium levels lower than 0.67 mmol/L and serum magnesium levels that ranged from 0.67 to 1.04 mmol/L were classified into the hypomagnesemia group (N = 30) and the normomagnesemia group (N = 5), respectively. Thirteen patients in the hypomagnesemia group and two patients in the normomagnesemia group exhibited thyroid dysfunction. No significant difference was observed between the two groups in the occurrence of thyroid dysfunction by chi-square test (P = 0:889).

Angiotensin II Is Not Associated with Thyroid
Dysfunction in GS Patients. GS patients were divided into the high angiotensin II (N = 17) and low angiotensin II groups (N = 18) according to the median value. Ten patients in the high angiotensin II group and five patients in the low angiotensin II group exhibited thyroid dysfunction. No significant difference was found between the two groups in the occurrence of thyroid dysfunction by chi-square test (P = 0:064).

Discussion
In this study, we discovered that patients with GS had a high incidence of thyroid dysfunction, including 28.6% (10/35) with subclinical hypothyroidism, 2.9% (1/35) with hypothyroidism, and 11.4% (4/35) with hyperthyroidism. This prevalence of thyroid dysfunction was significantly higher than that in the general population of China (the prevalence of subclinical hyperthyroidism and subclinical hypothyroidism was 0.46% and 5.46%, respectively) [12]. This prevalence was also higher than the reported prevalence in the US and Korea [13,14]. Published studies have shown that the prevalence of subclinical hypothyroidism, hypothyroidism, and hyperthyroidism was 8.5%, 0.4%, and 0.1%, respectively, according to the US Colorado thyroid disease prevalence survey [13]. Moreover, in Korea, the prevalence of subclinical hypothyroidism and subclinical hyperthyroidism was 3.3% and 2.6%, respectively, [14]. Fujimura et al. [15] investigated the clinical and genetic characteristics of 185 Japanese patients with GS diagnosed by genetic testing. They found that 4 (4.3%) of 94 cases exhibited thyroid dysfunction: 2 (2.1%) had hyperthyroidism, while 2 (2.1%) had hypothyroidism [15]. Unfortunately, they did not present data on subclinical thyroid dysfunction, which was high in the present study, and did not mention the issue of previously diagnosed thyroid disease and those who were under treatment for thyroid disease. To date, no other cohort study on GS and thyroid dysfunction could be found.
The most prominent characteristic of GS is hypokalemia. To study the impact of hypokalemia on thyroid dysfunction in GS patients, PA patients with and without hypokalemia were used as control groups. The results showed that the distribution of thyroid function in the GS group was significantly different than that in the normokalemic PA group and the hypokalemic PA group. No significant difference was observed in the distribution of thyroid function between the hypokalemic PA group and the normokalemic PA group. The probable explanations for this finding might be that (1) the hypokalemia in the GS group was congenital, while in the PA group, it was acquired. Whether sustained inherited hypokalemia affects thyroid function in GS patients requires further study; or (2) other factors instead of hypokalemia affected thyroid function in these GS patients.
The GS patients also exhibited hypomagnesemia. In animal experiments, thyroid-deficient animals conserved magnesium much more efficiently than rats with either euthyroidism or hyperthyroidism, although the serum magnesium levels in thyroid-deficient rats were similar to those in euthyroid rats [16]. Studies have shown that patients with hyperthyroidism had lower levels of serum Mg 2+ than patients with euthyroid function [17,18]. The use of methimazole for the treatment of hyperthyroidism due to GD leads to an increase in serum Mg 2+ concentration [18]. In our study, 30 (85.7%) GS patients exhibited hypomagnesemia. We compared the occurrence of thyroid dysfunction between the hypomagnesemia group and the normomagnesemia group. Nevertheless, no significant difference was found. This negative result still requires further confirmation by studies with larger sample sizes.
It has also been shown that the rat thyroid expresses the angiotensin II receptor subtype 1, AT1, but not the other subtype, AT2 [19]. Angiotensin II acts on AT1 to produce aldosterone. GS patients have high levels of angiotensin II, which might affect thyroid function by acting on AT1 receptors in thyroid cells. We compared the rate of thyroid dysfunction in GS patients in the high angiotensin II and low angiotensin II groups (divided by the median value) and found no difference in thyroid dysfunction between these two groups.
Patients with GS and Sjögren's syndrome (SS) have been reported in the literature [20][21][22][23]. In these patients, both acquired [20] and inherited GS [21][22][23] were reported. Kim et al. reported a case of acquired GS associated with primary SS and suggested the presence of circulating autoantibodies to sodium-chloride cotransporter (NCCT), which led to GS [23]. SS and AITDs may frequently coexist in clinical practice [24]. In a recent study of the characteristics of SS in a Chinese population, 332 (11.1%) patients had thyroid disease, including hyperthyroidism (1.2%), hypothyroidism (6.0%), and subacute thyroiditis (3.9%) [25]. In our study, we did not find manifestations of SS in any of the 35 GS patients. Since a relationship among GS, SS, and thyroid disease has been  One strength of this study was that we analyzed thyroid function in a cohort of GS patients, which is different from previous case reports. Another strength was that we included a large sample size of hypokalemic PA and normokalemic PA patients as control groups, which helped us determine the impact of hypokalemia on thyroid dysfunction in GS patients. Through a comparison with PA, which is also a hypokalemic disease, we did not find that hypokalemia affected thyroid function in GS patients.
The present study also had some limitations. TRAb was detected in only 19 (54.3%) GS patients, TGAb was detected in 20 (57.1%) GS patients, and TPOAb was detected in 23 (65.7%) GS patients. Therefore, we could not identify whether high thyroid dysfunction in GS patients was due to AITD. Zhou et al. summarized the data of 18 patients with GS complicated by AITD from nine papers and found that 13 had GD, 3 had Hashimoto's thyroiditis (HT), and two had antibody-positive AITD [3]. However, case reports cannot reflect real-world situations. In our study, we did not find that the positive rates of TPOAb and TGAb were related to thyroid dysfunction in GS patients. We speculated that thyroid dysfunction in GS patients was caused by factors other than AITD. However, additional studies are needed to clarify the detailed mechanisms of thyroid dysfunction in GS patients.
In conclusion, the situation of GS patients may be complicated by thyroid dysfunction. Thyroid dysfunction is more common in GS patients than in PA patients, and the risk of thyroid dysfunction is higher in female GS patients than in male patients.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.