Hashimoto's thyroiditis might increase polycystic ovary syndrome and associated comorbidities risks in Asia.

BACKGROUND: To investigate whether increased the comorbidities such as coronary artery disease (CAD) and risks between Hashimoto's thyroiditis (HT) and polycystic ovary syndrome (PCOS) in Taiwanese women.
METHODS: Patients newly diagnosed as having HT during 2000-2012 were assigned to the case group. Cases and controls were matched for age and comorbidities at a 1:2 ratio using propensity score matching. Incidence was calculated in the unit of 1000 person-year. Univariate and multivariate Cox proportional hazard regression, multivariate Cox, logistic regression, and Kaplan-Meier analyses were performed.
RESULTS: Among 3,996 participants, 2,664 constituted the control group and 1,332 constituted the case group. The PCOS risk in patients with HT increased by 2.37 times [95% confidence interval (CI): 1.22-4.62] compared with the controls. Hypertension (HTN) [adjusted odds ratio (OR): 1.31, 95% CI: 1.03-1.66] and hyperlipidemia (adjusted OR: 1.55, 95% CI: 1.2-1.9) were more common in HT patients without PCOS than in other patients. The adjusted OR for CAD in patients with HT was 1.51 (95% CI: 1.11-2.06), whereas that in patients with HT and PCOS was 5.92 (95% CI: 1.32-26.53).
CONCLUSIONS: In our study, the PCOS risk in patients with HT increased by 2.37 times, which is lower than the increase in HT risk in Asian patients with PCOS (4.56 times). The proportion of CAD increased significantly by 5.92 times in patients with HT and PCOS compared with patients with HT only. 2020 Annals of Translational Medicine. All rights reserved.

Introduction

Hashimoto’s thyroiditis (HT) was first described by Hakaru Hashimoto in 1912, but the autoimmunity of HT was demonstrated in the 1950s. The etiology of HT is not completely known. According to epidemiological studies, HT is the most common autoimmune disease and also the most common endocrine disorder to cause hypothyroidism (1-3). HT was classified into primary and secondary forms according to its etiology (4). The clinicopathological spectrum of HT comprises six primary entities, such as the classic form and silent thyroiditis (5,6). Lymphocytic infiltration of the thyroid gland is the predominant pathological feature. Secondary form is usually iatrogenic, which is commonly induced by immunomodulatory drugs, such as interferon-alpha for hepatitis C viral infection treatment and monoclonal antibodies blocking CTLA-4 for cancer therapy (7,8). In general, HT is recognized as an organ-specific autoimmune disease with circulating antibodies, such as thyroperoxidase (TPO) and thyroglobulin (Tg), to thyroid antigens. This immune response can engender impaired conversion of estrogen precursors (4). High estrogen levels instead of androgens and progesterone are implicated in humoral immunity enhancement (9,10). Polycystic ovary syndrome (PCOS) with an increased estrogen-to-progesterone ratio is the most common endocrine disorder of women, affecting 5–10% of all reproductive-aged women worldwide (11,12). Studies have reported that among Whites and Asians, women are more likely to have HT than men (4,13). Hence, we investigated the association between HT and PCOS in Taiwanese women, including the comorbidities such as diabetes, hypertension (HTN), coronary artery disease (CAD) and risks between both diseases. Previous studies investigating the association of PCOS with HT have always focused on PCOS; by contrast, we attempted to examine the association from the perspective of HT. We present the following article in accordance with the STROBE reporting checklist (available at http://dx.doi.org/10.21037/atm-19-4763).

Methods

Data source

In March 1995, the Taiwan National Health Insurance (NHI) program was initiated. The program provides health care coverage for approximately 99% of Taiwan’s population. The National Health Insurance Administration established the National Health Insurance Research Database (NHIRD), which is managed by the National Health Research Institute. The NHIRD includes health information such as health status, medication, or treatment of insured individuals. In this study, we used the Longitudinal Health Insurance Database, a sub-database of the NHIRD with 1 million randomly selected insured individuals. Diseases were coded according to International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes.

Ethics statement

The NHIRD encrypts patient personal information to protect privacy and provides researchers with anonymous identification numbers associated with relevant claims information, including sex, date of birth, medical services received, and prescriptions. Therefore, patient consent is not required to access the NHIRD. This study was approved to fulfill the condition for exemption by the Institutional Review Board (IRB) of China Medical University (CMUH104-REC2-115-CR4). The IRB also specifically waived the consent requirement.

Study sample

This cohort study included a case group and control group. Patients newly diagnosed as having HT (ICD-9-CM code: 245.21) during the 2000–2012 period were assigned to the case group. The definition of HT patients was subjects with two or more outpatient visits or one hospitalization record of HT diagnosis. Individuals without HT constituted the control group. We excluded individuals with a history of PCOS before entry in the study and those aged younger than 20 years from both groups. Cases and controls were matched according to age and index year, which was the date of HT diagnosis, at a 1:2 ratio using propensity score matching.

Main event and covariates

The endpoint of the present study was diagnosis of PCOS (ICD-9-CM code 2564). We considered the following comorbidities, which developed before the index date, as potential confounders of PCOS: diabetes mellitus (DM) (ICD-9-CM code 250), hyperlipidemia (ICD-9-CM code 272), HTN (ICD-9-CM codes 401–405), stroke (ICD-9-CM codes 430–438), CAD (ICD-9-CM codes 410–414), and heart failure (ICD-9-CM codes 428, 428.0, 428.1, and 428.9).

Statistical analysis

A chi-square test was used to compare the baseline categorical variables between the case and control group. A Student t-test was used to assess mean of age between two groups. Incidence was calculated in the unit of 1,000 person-year. Univariate Cox proportional hazard models were used to estimate hazard ratios for PCOS. The multivariate Cox model included age and all comorbidities as variables to adjust the hazard ratio of PCOS. Logistic regression was then applied to investigate the effect of comorbidities on the occurrence of HT or PCOS. The cumulative incidence of PCOS in the case and control groups was evaluated using the Kaplan-Meier method. A Log-rank test was then used to verify the difference between the two curves. The statistical significance level was set as a P value less than 0.05. All statistical analyses were performed by SAS (version 9.4; SAS Institute, Inc., Cary, NC, USA).

Results

Of 3,996 participants included in this study, 2,664 constituted the control group and 1,332 constituted the case group. The follow-up time for the case group was 6.69 (±3.81) years, and that for the control group was 6.69 (±3.77) years. Most of the participants were aged older than 40 years. The mean ages in the control group was 45.1 (±16.3) years, and that in the case group was 43.7 (±14.4) years. The case group had higher proportions of HTN and hyperlipidemia. In the case and control groups 1.4% (n=19) and 0.6% (n=16) of the participants developed PCOS ().

Table 1

The baseline characters of patient with and without Hashimoto thyroiditis

Variables	Hashimoto thyroiditis					P value
	No=2,664			Yes=1,332
	n	%		n	%
Age						1.00
<30	510	19.0		255	19.0
30–39	632	24		316	24
≥40	1,522	57		761	57
Mean (SD)†	45.1	16.3		43.7	14.4	<0.001
Comorbidities
DM	275	10		175	13	0.01
HTN	594	22		278	21	0.32
Hyperlipidemia	433	16		298	22	<0.001
Stroke	162	6.1		94	7.1	0.26
CAD	328	12		192	14	0.07
Heart failure	67	2.5		37	2.8	0.70
PCOS	16	0.6		19	1.4	0.01

Chi-square test; †Student t-test. DM, diabetes mellitus; HTN, hypertension; CAD, coronary artery disease; PCOS, polycystic ovary syndrome.

lists the incidence and hazard ratios of PCOS. The PCOS risk in patients with HT increased by 2.37 times [95% confidence interval (CI): 1.22–4.62] compared with that in participants without HT. The risk of PCOS decreased by 0.88-fold (95% CI: 0.85–0.92) with an increase in age. Comorbidities were not statistically significant risk factors for PCOS.

Table 2

The incidence rate and hazard ratio of polycystic ovary syndrome

Variables	PCOS				Crude			Adjusted
	Events	PY	Rate#		HR	95% CI		HR	95% CI
HT
No	16	17,831	0.90		1.00	Reference		1.00	Reference
Yes	19	8,917	2.13		2.37	1.22–4.62*		2.37	1.21–4.64*
Age					0.88	0.85–0.92***		0.87	0.83–0.91***
<30	25	5,489	4.55		1.00	Reference
30–39	9	6,805	1.32		0.29	0.14–0.63***
≥40	1	14,454	0.07		0.01	0.00–0.11***
Comorbidities
No	29	18,320	1.58		1.00	Reference		1.00	Reference
Yes	6	8,428	0.71		0.43	0.18–1.05		2.22	0.87–5.65

Adjusted HR: adjusted with age and comorbidities of DM, HTN, hyperlipidemia, stroke, CAD and heart failure. Rate#, incidence rate per 10,000 person-years. *, P<0.05; ***, P<0.001. CI, confidence interval; HR, hazard ratio; PY, person-years; DM, diabetes mellitus; HTN, hypertension; CAD, coronary artery disease; PCOS, polycystic ovary syndrome; HT, Hashimoto’s thyroiditis.

The results of comorbidity-specific analysis are listed in . HTN [adjusted odds ratio (OR): 1.31, 95% CI: 1.03–1.66] and hyperlipidemia (adjusted OR: 1.55, 95% CI: 1.2–1.9) were more common in HT patients without PCOS than in the controls. The adjusted OR for CAD in patients with HT was 1.51 (95% CI: 1.11–2.06), and that in patients with HT and PCOS was 5.92 (95% CI: 1.32–26.53).

Table 3

The odd ratios of comorbidities in patients with and without Hashimoto thyroiditis and polycystic ovary syndrome

Variables	Event, N (HT/PCOS)	Adjusted
		OR	95% CI
DM
No/no	2,648/187	1.00	Reference
Yes/no	1,313/83	0.95	0.73–1.25
No/yes	16/1	1.81	0.23–13.92
Yes/yes	19/2	3.03	0.69–13.41
HTN
No/no	2,648/196	1.00	Reference
Yes/no	1,313/120	1.31	1.03–1.66*
No/yes	16/2	2.76	0.62–12.38
Yes/yes	19/1	1.04	0.14–7.85
Hyperlipidemia
No/no	2,648/263	1.00	Reference
Yes/no	1,313/185	1.55	1.27–1.9***
No/yes	16/0	–	–
Yes/yes	19/1	0.74	0.1–5.62
Stroke
No/no	2,648/103	1.00	Reference
Yes/no	1,313/57	1.32	0.94–1.85
No/yes	16/0	–	–
Yes/yes	19/1	4.86	0.63–37.73
CAD
No/no	2,648/108	1.00	Reference
Yes/no	1,313/73	1.51	1.11–2.06**
No/yes	16/0	–	–
Yes/yes	19/2	5.92	1.32–26.53*
Heart failure
No/no	2,648/77	1.00	Reference
Yes/no	1,313/28	0.91	0.58–1.43
No/yes	16/0	–	–
Yes/yes	19/0	–	–

Adjusted OR: adjusted with age. *, P<0.05; **, P<0.01; ***, P<0.001. DM, diabetes mellitus; HTN, hypertension; CAD, coronary artery disease; PCOS, polycystic ovary syndrome; HT, Hashimoto’s thyroiditis; OR, odds ratio.

The cumulative incidence of PCOS in patients with HT was higher than that in the controls, and the difference was statistically significant (P=0.009; ).

Figure 1

The cumulative incidence of PCOS (polycystic ovary syndrome) in patients with HT was higher than that in the controls, and the difference was statistically significant (P=0.009). PCOS, polycystic ovary syndrome; HT, Hashimoto’s thyroiditis.

Discussion

Associated risks between HT and PCOS

In our study, the risk of PCOS in patients with HT increased by 2.37 times (95% CI: 1.22–4.62) compared with participants without HT. Compared with other studies on HT risk in PCOS, the general risk was higher by 4.56 times in Asians, 3.27 times in Europeans, and 1.86 times in South Americans (14). One prospective study demonstrated a 3 times higher prevalence of autoimmune thyroiditis in the PCOS group compared with the control group (15). Another recent study on Danish women revealed that the event rate of thyroid disease increased significantly by 2.5 times in the PCOS group (16). Compared with other studies, although our analysis was based on clinical diagnoses using only ICD-9-CM codes and without any laboratory information, the accuracy of HT and PCOS diagnoses was acceptable. One of the reasons is that in the Taiwan NHI system, most patients, including those with HT and PCOS, are offered quick treatment by endocrinology and gynecology specialists without being transferred from primary physicians. Younger women aged younger than 30 years were the predominant group in our analysis, indicating that our government should pay more attention to these reproductive-aged women to improve aging in Taiwan.

Associated comorbidities between HT and PCOS

In general, overall comorbidities were not statistically significant. In our subgroup analysis, patients with HT only appeared to have more comorbidities, namely HTN, hyperlipidemia, and CAD, than did those with PCOS only. HT caused hypothyroidism or subclinical hypothyroidism and it resulted in hyperlipidemia (2.3) Notably, CAD was 5.92 times significantly higher in the HT patients with PCOS compared with other patients, strongly suggesting for early CAD screening in patients with HT and PCOS. The PCOS standard treatment guidelines indicate the importance of early diagnosis along with early intervention to improve not only infertility but also other risks of diabetes, HTN, and dyslipidemia for cardiovascular disease (CVD) (17,18). To our knowledge, PCOS was associated with metabolic syndrome, insulin resistance, diabetes, obesity, dyslipidemia, HTN, and CVD (14,19,20), and the common mechanisms were higher insulin level or low grade chronic inflammation due to high estrogen secretion influenced by the hypothalamic-pituitary-gonadal axis (21). The association between inflammation and autoimmunity in PCOS was extensively discussed in a recent study (22). Whether stimulating ovarian antibodies in granulosa cells play a major role in PCOS pathophysiology (18) or the concept of nonorgan-specific autoantibodies in PCOS (20) constitute the autoimmune etiology is still controversial (23,24). These discussions partially explain the observation of the recurrent association between PCOS and autoimmune diseases such as Cushing syndrome and autoimmune thyroid diseases, the most common autoimmune disorder in the world (25,26). HT, involving positive antibodies for TPO or Tg, is one of the most common causes of hypothyroidism in young women through the mechanism of T cell-mediated organ-specific autoimmnity (27-29). One study revealed the relationship between subclinical hypothyroidism and PCOS with metabolic abnormalities, such as the level of high-density lipoprotein and homeostatic model assessment for insulin resistance (30). Regardless of study design, such as autoimmunity or epidemiology designs, the association between HT and PCOS still remains unclear.

Limitations

First, our national database lacks laboratory information such as TSH, FT4, FT3, anti-TPO, anti-TG levels for HT, insulin and fasting glucose level for insulin resistance calculation, lipid profile with cholesterol or triglyceride and estrogen, progesterone, and androgen levels for PCOS; therefore, we could not include such information in our analysis. In addition, we did not conduct an imaging study, such as thyroid or ovary echograms. Second, lifestyle or other environmental factors such as nutritional state, other pollution, smoking, alcohol consumption, and exercise habits were not included in the study.

Conclusions

In our study, the risk of PCOS in patients with HT increased by 2.37 times, which is lower than the increase in the risk of HT in Asian patients with PCOS (4.56 times). The proportion of CAD increased significantly by 5.92 times in patients with HT and PCOS compared with patients with HT only. Considering the public health issue, infertility as well as CAD screening should be indicated for patients with both HT and PCOS.

The article’s supplementary files as