Women with alopecia exhibit a higher risk for thyroid cancer: A nationwide cohort study.

Women with alopecia exhibit a higher risk for thyroid cancer: A nationwide

cohort study

Li-Min Sun

, Ming-Chia Lin

, Chih-Hsin Muo

, Ji-An Liang

, Fung-Chang Sung

, Chia-Hung Kao

*

1. Introduction Androgenic alopecia (AA) is the most common cause of hair loss, followed by alopecia

areata and telogen effluvium. AA affects nearly 50% of men [1,2]. In females, AA occurs much more

frequently than is generally believed [3]. By contrast, approxi-mately 2% of the general population

develops alopecia areata [4]. AA is usually described as a genetically determined condition leading to

the permanent loss of hair in men and women [5]. Japanese researchers have suggested that the

Sox21 gene is a master regulator of hair shaft cuticle differentiation, which provides evidence on the

possible causes of human hair disorders [6]. It has been suggested that androgens play a central role

in the pathophysiology of both prostate cancer and AA. A considerable amount of evidence supports the role of androgens in prostate cancer: Eunuchs rarely develop prostate cancer, castration has a palliative effect on prostate cancer, and testosterone alone can produce prostatic adenocarcinoma in

rats [7,8]. Therefore, prostate cancer is typically considered a hormonally linked cancer. Likewise,

eunuchs do not develop baldness if castrated before the age of 25 [9]. Several studies have

investigated the relationship between AA and prostate cancer and have yielded inconsistent results

[1,10–14]. Little information is available regarding the relationship between alopecia and the risk of hormone-related cancers in women. Thyroid cancer is thought to be a kind of hormone-related cancers

[15]. Because hormones play an essential role in AA and alopecia was suggested to be related to

certain thyroid diseases [16], we hypothesize that AA has an association with thyroid cancer in women. Based on a thorough

review of relevant literature, there are no studies that outline the possible relationship between alopecia and the risk of cancer in women. This study investigates the risk of overall and thyroid cancers among women with alopecia in Taiwan. The results were generated from a retrospective cohort study of women with alopecia. The data was derived from the database of the National Health Insurance (NHI) system in Taiwan. 2. Methods This retrospective cohort study used the Longitudinal Health Insurance Database (LHID) and the Registry for Catastrophic Illness Patients. Those databases were formed by the Taiwan Bureau of National Health Insurance (TBNHI) and maintained by the National Health Research Institutes (NHRI). The TBNHI

established a single-payer NHI program on March 1, 1995 and this program covers over 99.5% of population in Taiwan [17]. The

LHID comprised one million insurants randomly selected from the original beneficiaries in 2000 and included all medical records for insurants from 1996 to 2010. The NHRI scrambled the identification of the insurants in accordance with the Personal Information Protection Act before releasing the information to researchers. This study was provided the Institutional Review

Board of the China Medical University and Hospital. Our research group has referenced the LHID in previous studies [18,19]. We

collected information on women with diagnosed alopecia [The International Classification of Diseases, 9th Revision, clinical modification (ICD-9-CM) 704.0] from 2000 to 2010. We excluded those with any history of cancer (ICD-9-CM 140–208) and the duration of follow-up was <0.5 year. Controls were selected from women without a history of alopecia before the entry date. We randomly assigned the year and month to controls and the entry date was the middle of month. Four controls were randomly frequency matched with age (5 years stratified: for example 0–4, 5–9, 10–14 and so on.), entry month, and entry year.

The excluded criteria were the same in the control and case groups. The demographic differences between the groups were ana-lyzed using a chi-square test for the categorical variables of age group and comorbidity [including hypertension (ICD-9-CM 401- 405) and diabetes (ICD-9-CM 250)], and a t-test for the continuous variable of age. We counted the person-years from the entry date to the date of cancer occurrence, or until the end of 2010, and calculated the incidence per 10,000 person-years. The hazard ratio (HR) and 95% confidence intervals (CIs) for cancer were assessed using Cox proportional hazard regression and the multivariable model was adjusted for age and comorbidity. The risks for cancer type assessed and the types were thyroid and others. According to Bonferroni correction, the significant level was set at 0.025 for multiple hypothesis testing. A Kaplan–Meier analysis was used to plot the cumulative incidence for cancer and a log-rank test was used to test the difference between the groups. In the further analysis, we estimated the risk levels for the subtypes of alopecia: unspecified, areata, and telogen effluvium. We also evaluated the effect of 2 thyroid autoimmune diseases (Graves’ disease and Hashimoto thyroiditis) on the relationship between alopecia and thyroid cancer. The NHRID encrypts the patients’ personal information for privacy protection and provides researchers with anonymous identification numbers associated with the relevant claim infor-mation, which includes the patient’s sex, date of birth, registry of medical services, and medication prescriptions. Patient consent is not required for accessing the NHIRD. This study was approved by the Institutional Review Board of China Medical University (CMU-REC-101-012). Our IRB specifically waived the requirement for consent. 3. Results The case cohort comprised 4534 women with alopecia and the control cohort comprised 18,128 women. The majority of women were 20–34 years old (41.6%), and the mean age was 32.8 (standard deviation = 14.6). The case cohort was more likely to have a history of diabetes and hypertension

than the control cohort, but only diabetes showed a significant difference (Table 1). After a nine-year follow-up, 59 and 192

events occurred in the case and control cohorts, respectively (Fig. 1A and Table 2). The incidences of overall cancer were 27.50

and 22.56 per 10,000 person-years for the alopecia and non-alopecia groups, respec-tively, and the HR was 1.22 (97.5% CI 0.87– 1.70) compared with the control group. After analyzing the risk of subtypes of cancer, only thyroid cancer showed a significantly higher risk in the case cohort, with an HR of 2.39 (97.5% CI 1.05–5.42). Compared to controls, the women with unspecific

alopecia had a significant 1.87-fold higher risk for thyroid cancer, but not for other cancers (Table 3). The cumulative incidence

of thyroid cancer in the case cohort was significantly 0.3% higher than in the control cohort (log-rank P = .024, Fig. 1B). Women

with alopecia had a significantly higher risk for Graves’ disease compared with women without alopecia (1.94% vs. 1.09%, P < . 05), but the risk for Hashimoto thyroiditis was less and non-significant (0.26% vs. 0.33%). Because of the relatively small number of thyroid cancer cases in both groups of autoimmune diseases, the incidence rate does not reflect the statistical significance (Table 4). Table 5 illustrates that the positive relationship between alopecia and thyroid cancer was limited to women younger than 50 years. 4. Discussion This nationwide cohort study discovered that patients with alopecia show a significantly higher risk of developing thyroid cancer. Further analyses showed that the significantly increased risk is limited to women younger than 50 years. The alopecia group showed a higher incidence of Graves’ disease but not Hashimoto thyroiditis. Stratified analyses of these thyroid diseases did not reveal any significant difference in thyroid cancer between the alopecia and non-alopecia groups because of the relatively small number of cases in the subgroups. Alopecia is not uncommon in women. An earlier survey conducted on women in the United Kingdom found frontal and frontoparietal recessions in 13% of premenopausal and 37% of

postmenopausal women [20]. A recent study revealed that the prevalence of alopecia in Taiwanese women 30 years and older

was 11.8% (95% CI 11.5%–12.2%) and increased with advancing age [21]. Alopecia was suggested to be related to certain

thyroid diseases [16]. Lo Sicco et al. performed a retrospective study to investigate thyroid structural abnormalities in alopecia

ultrasound finding [22]. The exact etiology of alopecia areata remains unclear, but evidence suggests that autoimmunity and

endocrine dysfunction are involved [23,24]. A significant association between alopecia areata and thyroid autoimmunity has

been reported in several studies [25,26]. Kurtev et al. diagnosed autoimmune thyroiditis in 22 of 46 children with alopecia

areata [25]. Puavilai et al. determined that the prevalence of thyroid disease among patients with alopecia areata was 7.2%

[27]. Our data revealed a significantly higher risk of thyroid cancer among women with alopecia. Fig. 1B illustrates that the difference in cumulative incidences of thyroid cancer between groups is consistent over time after a 2-year follow-up. We explored possible explanations for this result. Some autoimmune thyroid diseases have been suggested to be related to the

development of thyroid cancer, although the precise mechanism of that relationship remains unclear [28–33]. Graves’ disease is

the most prevalent autoimmune disorder in the United States [34] and is the underlying cause of 50%–80% of cases of

hyperthyroidism [35]. Hyperthyroidism does not appear to protect patients from thyroid cancer [36], and a higher risk of thyroid

cancer among patients with Graves’ disease has been reported in several studies [28–31]. A Taiwanese cohort study showed

that patients with Graves’ disease had a significantly higher risk (adjusted HR = 10.4) of developing thyroid cancer [33]. By

contrast, Hashimoto thyroiditis is also an autoimmune disease that causes the immune system to attack and destroy the thyroid

gland [37]. An epidemiological association has been identified between Hashimoto thyroiditis and thyroid cancer [28,32,38–40],

and the increased incidence of thyroid cancer in patients with Hashimoto thyroiditis suggests that it is a precancerous condition. Therefore, we hypothesized that the higher risk of thyroid cancer observed in our group of women with alopecia is partially attributable to the autoimmune thyroid diseases linked to both alopecia and thyroid cancer. However, our data did not support this assumption. The relatively small case numbers of Graves’ disease and Hashimoto thyroiditis in the study groups prevented us from detecting significant findings. The same concern applies for alopecia areata; because of the relatively small number of cases, statistically significant results were not expected. Our data showed that only women with alopecia who were younger than 50 years had a significantly higher risk for thyroid cancer. Again, the small number of cases of with alopecia who were 50 years and older hindered the possibly of significant association. Nevertheless, the prognosis of thyroid cancer is highly

dependent on the age of the patient [41], and the significantly higher risk for younger women with alopecia should prompt

public health authorities to perform extensive screenings of these patients to detect lesions as soon as possible. The strength of this study was its population-based design, which increased the generalizability of the results. However, some limitations must be addressed. First, potential misclassification of alopecia is a large concern with insurance-database-based research. Taiwan launched the NHI program in 1995, and it is operated by a single-payer, the government. All insurance claims should be scrutinized by medical reimbursement specialists and peer review. Therefore, the final diagnoses and coding of alopecia must be checked by at least 2 experienced dermatologists in the NHRID. In addition, several Taiwan studies had demonstrated the

high accuracy and validity of diseases diagnoses in NHIRD [42,43]. The diagnoses of alopecia and cancer are based on highly

reliable database and the concern of misclassification can be minimized. Second, the diagnosis code in the NHIRD did not specify AA; thus, we cannot precisely evaluate its effect on the risk of cancer. Instead, we used unspecified alopecia to assess patients. The number of cases of unspecified alopecia was not sufficiently large to cover all the cases of AA. Third, the NHIRD does not provide information regarding cancer histology, and certain types of histological thyroid cancer are more commonly

seen in thyroid autoimmune diseases [30,31,39,40]. However, we could not evaluate this because of the intrinsic limitations of

the database. In conclusion, this population-based retrospective cohort study found a significantly increased risk of thyroid cancer in Taiwanese women with alopecia compared with the general population, especially in younger women. Although the underlying mechan-isms remain unclear and our results must be verified in further investigations, the findings from this study