Association between hyperthyroidism and risk of incident in Parkinson’s disease

Introduction

Hyperthyroidism is a condition in which excess thyroid hormone is synthesized and secreted by the thyroid gland. The prevalence of symptomatic hyperthyroidism is approximately 0.5% and varies geographically (1). In Taiwan, the prevalence of symptomatic hyperthyroidism is 0.37%, and the leading causes of hyperthyroidism are Graves’ disease (95%), followed by toxic nodular goiter (2%) and other causes (3%) (2). Hyperthyroidism has been implicated in many other diseases and is widely studied for its association with cardiovascular disease (3). Hyperthyroidism has also been shown to contribute to periodic thyrotoxic paralysis, osteoporosis, and abnormalities of the reproductive system, as well as neurodegenerative diseases (1).

Parkinson’s disease (PD) is one of the most common neurodegenerative diseases and is characterized by the progressive loss of dopaminergic neurons in the substantia nigra (4). The cause of PD remains unknown; however, genetic and environmental factors have been implicated (4). Genetic association studies have identified several loci that are significantly associated with PD (4). An increased risk in individuals exposed to pesticides and those who have suffered head injuries has also been demonstrated (4).

Epidemiological evidence has suggested a link between hyperthyroidism and incident PD. A nationwide population-based cohort study in Sweden that followed 34,735 patients with Graves’ disease/hyperthyroidism has reported a 1.42-fold increased risk of developing PD (5). Another nationwide case-control study in Denmark that retrospectively followed 13,695 patients with PD has reported twice the risk of developing PD in patients with a prior diagnosis of Graves’ disease/hyperthyroidism (6). The two studies previously used data from the inpatient registry without outpatient records, which may underestimate those with less severe illness conditions. Besides, they did not have access to the drug/surgery database, limiting them from using this database to confirm the diagnosis of hyperthyroidism/PD or to study the influence of antithyroid therapy on the risk of incident PD.

To more fully determine if hyperthyroidism is linked to an increased risk of incident PD, Taiwan’s National Health Insurance Research Database (NHIRD), a nationally representative database of medical claims, was used to conduct a longitudinal follow-up study of the risk of PD in patients with hyperthyroidism.

Subjects and methods

Results

Risk of PD

The incidence rates for PD were 1.21 (case cohort) and 0.45 (control cohort) per 1000 person-years, respectively (Table 2). After adjusting for the covariates mentioned previously, the case cohort had a significantly higher risk of PD than the control cohort (HR: 2.69, 95% CI: 1.08–6.66). Kaplan–Meier analysis with a log-rank test also revealed a significant association between having hyperthyroidism and the subsequent risk of PD (P < 0.01; Fig. 1).

View Full Size

Cumulative incidence of Parkinson’s disease in patients with hyperthyroidism (case cohort) and comparative controls (control cohort). Kaplan–Meier analysis with a log-rank test revealed a significant association between the case cohort and Parkinson’s disease (P < 0.01).

Citation: Endocrine Connections 10, 1; 10.1530/EC-20-0554

• Download Figure
• Download figure as PowerPoint slide

Table 2

Cox regression analyses of demographic characteristics associated with Parkinson’s disease for the entire cohort.

Variables	PD	Person years	IRa	Crudeb HR (95% CI)	Adjustedc HR (95% CI)
Hyperthyroidism
No	32	70,370	0.45	1 (Reference)	1 (Reference)
Yes	85	70,486	1.21	2.65 (1.77–3.99)*	2.69 (1.08–6.66)*
Age
<50	18	108,752	0.17	1 (Reference)	1 (Reference)
50–65	47	24,556	1.91	11.7 (6.81–20.2)*	3.74 (2.85–4.92)*
>65	52	7548	6.89	43.5 (25.4–74.6)*	8.53 (3.58–12.7)*
Gender
Female	91	109,490	0.83	1 (Reference)	1 (Reference)
Male	26	31,366	0.83	0.99 (0.64–1.54)	0.86 (0.54–1.37)
CCI score
0	74	131,551	0.56	1 (Reference)	1 (Reference)
1–2	36	8123	4.43	7.95 (5.33–11.8)*	1.58 (1.13–2.21)*
≥3	7	1182	5.92	11.0 (5.05–23.9)*	1.63 (1.25–3.34)*
Comorbidities
Brain injury
No	100	134,505	0.74	1 (Reference)	1 (Reference)
Yes	17	6351	2.67	3.58 (2.14–6.00)*	1.57 (1.03–2.53)*
Cerebrovascular disease
No	71	133,578	0.53	1 (Reference)	1 (Reference)
Yes	46	7278	6.32	11.9 (8.20–17.2)*	3.44 (2.27–5.21)*
Hypertension
No	86	119,236	0.72	1 (Reference)	1 (Reference)
Yes	31	21,620	1.43	1.98 (1.31–2.99)*	1.08 (0.69–1.69)
Dyslipidemia
No	85	118,408	0.72	1 (Reference)	1 (Reference)
Yes	32	22,448	1.42	1.97 (1.31–2.97)*	1.29 (0.83–2.01)
Diabetes mellitus
No	88	123,500	0.71	1 (Reference)	1 (Reference)
Yes	29	17,356	1.67	2.33 (1.53–3.55)*	1.26 (0.81–1.97)

^aPer 1000 person-years; ^brelative hazard ratio; ^cadjusted for hyperthyroidism, age, gender, CCI score, comorbidities and antithyroid therapy in Cox regression analyzes (antithyroid therapy, including antithyroid drugs, radioactive iodine ablation, and surgery, were treated as time-dependent covariates); *P-value <0.05.

CCI, Charlson Comorbidity Index; HR, hazard ratio; IR, incidence rates; PD, Parkinson’s disease.

For the entire cohort, after adjusting for the covariates mentioned previously, the risk of PD was higher for those aged 50–65 (HR: 3.74, 95% CI: 2.85–4.92) and >65 years (HR: 8.53, 95% CI: 3.58–12.7) than for those <50 years. There was no significant difference in the risk estimate for PD between the two genders. Also, compared to patients whose baseline CCI score was 0, those with a CCI score of 1–2 (HR: 1.58, 95% CI: 1.13–2.21) and ≥3 (HR: 1.63, 95% CI: 1.25–3.34) were associated with a higher risk of PD after adjusting for the covariates mentioned previously. Regarding baseline comorbidities, brain injury (HR: 1.57, 95% CI: 1.03–2.53), cerebrovascular disease (HR: 3.44, 95% CI: 2.27–5.21) were associated with an increased risk of incident PD after adjusting for the covariates mentioned previously. There was no significant difference in the estimation of the PD risk in the presence of hypertension, dyslipidemia, or diabetes mellitus.

Discussion

This population-based cohort study examined the risk of PD in patients with hyperthyroidism using age-, gender-, index year-, and CCI score-matched cohort with a maximum follow-up period of 13 years in a nationally representative sample. The study demonstrated that patients with hyperthyroidism had an increased risk of PD compared with unaffected controls. This finding was consistent with reports of population-based studies in Sweden and Denmark (5, 6). Unique to the two studies, this study contained outpatient records and took into account antithyroid therapy’s effect on the risk of incidence of PD. Hyperthyroidism should still be considered an independent risk factor for PD because, after adjustment for the impact of antithyroid therapy, a significant risk association has been observed.

There are possible explanations that may account for the mechanisms underlying this observation. First, hyperthyroidism has been linked to increased oxidative stress and a deregulated inflammatory response in adult patients (9). Since persistent oxidative stress and neuroinflammation has been linked to both the initiation and progression of PD (10), hyperthyroidism may serve as a possible contributing factor to the development of PD through these mechanisms (11). Second, Graves’ disease is considered an autoimmune disease that affects the thyroid and is the leading cause of hyperthyroidism in Taiwan (12). A previous study found that Graves’ disease may share genetic pathways with PD and is closely linked to PD’s specific clinical manifestations (13). Amplified autoimmune reactions of Graves’ disease via shared genetic pathways may also contribute to PD development (14). Third, hyperthyroidism triggers symptoms in patients with subclinical vulnerability to PD by acting on PD’s biological substrate. For example, the thyroid hormone can increase dopamine catabolism or alter dopamine receptors’ sensitivity (15), which may cause symptoms in patients vulnerable to PD. It should be mentioned that these explanations should be considered preliminary and that the mechanisms of detail are worth studying in the future.

Besides, this study found that older age, more comorbidities, and specific comorbidities (brain injury and cerebrovascular disease) were associated with an increased risk of developing PD. Sufficient evidence shows that age is the most critical risk factor for PD, where the incidence of PD begins to rise sharply after age 60 (16); this study confirms this finding. More medical comorbidities and specific comorbidities (brain injury and cerebrovascular disease) may increase the risk of developing PD are consistent with the previous results, which could be the possibility of a direct cause-effect relationship pathophysiological vulnerabilities shared (4). However, we still considered hyperthyroidism as an independent risk factor for PD because a significant risk association was found after adjusting for these comorbidities.

In the stratification analysis, similar risk associations linking hyperthyroidism to subsequent PD were found in all stratified age and gender subgroups (adjusted HR all >1). However, a relatively small number of PD incidents could result in HR estimates having a wider 95% CI in some age and gender subgroups and therefore become statistically insignificant. Further studies with a larger sample that could adequately adjust all of the model’s confounders are warranted.

This study’s strengths include the large population-based cohort, adjustment for known risk factors, and long period of observation. This study also contains outpatient records and considers the effect of antithyroid therapy on the risk of incident PD, which is unique to the previous two population-based studies (5, 6). However, several limitations should also be noted. First, only patients who sought medical help for the diagnosis and treatment of hyperthyroidism or PD were included; therefore, patient identification may have been biased and may have weakened the observed association. However, the patients identified in the current study were confirmed by their treatment, which improved diagnostic validity. Second, the presentation of symptoms for hyperthyroidism and PD and detailed laboratory data such as thyroid hormone levels are not available in the Taiwan NHIRD database. Therefore, additional analyses could not be applied, such as the relationship between thyroid hormone levels and PD symptoms. Third, hyperthyroidism and PD share common manifestations, making a differential diagnosis more difficult (8). Although a sensitivity analysis was performed to minimize potential bias by excluding the first year of observation, the inclusion of false positives is a common concern. Finally, Taiwan NHIRD does not have information on some known risk or protective factors associated with PD, such as family history, pesticide exposure, smoking habits, and coffee drinking (4). Therefore, the contributions of these factors could not be further evaluated.

This study shows that hyperthyroidism is associated with an increased risk of PD, corroborating previous studies in Sweden and Denmark. Further prospective studies on the pathogenesis between hyperthyroidism and PD and determine if an intervention could reduce the risk of PD are warranted.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This study is supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial Center (MOHW109-TDU-B-212-114004); MOST Clinical Trial Consortium for Stroke, Taiwan (MOST 108-2321-B-039-003); Tseng-Lien Lin Foundation, Taichung, Taiwan; Buddhist Tzu Chi Medical Foundation, Taiwan (TCMF-MP 108-01-03). The funding source has no further role in study design; in the collection, analysis, and interpretation of data; in the report’s writing; and in the decision to submit the paper for publication.

Data availability

The data are available, on request, from the corresponding author.

Statement of ethics

This research was monitored by the Institutional Review Board of China Medical University (CMUH104-REC2-115). All study approaches followed the appropriate standards and policies. The need for informed consent from individual subjects was waived because the LHID2000 contains only anonymized secondary data.

Author contribution statement

S R Lin and S F Chen managed the literature searches and wrote the ‘Introduction’ of the manuscript. Y C Yang and C Y Hsu did all the analyses and wrote the ‘Method and results’ of the manuscript. Y C Shen conceived the study and wrote the ‘Discussion’ of the manuscript. All authors have approved the final manuscript.