Abstract
Background:
Although the dopaminergic system is interconnected with the hypothalamic-pituitary-thyroid axis, few studies have explained the causal relationship between thyroid disease and Parkinson’s disease (PD).
Objective:
The goal of this study was to investigate the association between thyroid diseases and PD in Korean residents.
Methods:
The Korean National Health Insurance Service-National Sample Cohort, which includes individuals aged ≥40 years, was assessed from 2002 to 2015. A total of 5,586 PD patients were matched by age, sex, income, and the region of residence with 22,344 control participants at a ratio of 1:4. In the PD and control groups, previous histories of levothyroxine treatment, goiter, hypothyroidism, thyroiditis, and hyperthyroidism were investigated.
Results:
The rates of levothyroxine treatment for more than 3 months, hypothyroidism, and hyperthyroidism were higher in the PD group than the control group (3.2%, 3.8%, and 2.8% vs. 2.5%, 2.9%, and 1.9%, respectively, p < 0.05). The adjusted odds ratios (ORs) in model 2, which was adjusted for all potential confounders, for hypothyroidism and hyperthyroidism in the PD group were 1.25 (95% confidence interval (CI) 1.01–1.55, p = 0.044) and 1.37 (95% CI 1.13–1.67, p = 0.002), respectively. In subgroup analyses, the association between hypothyroidism and PD was maintained in men older than 70 years and the association between hyperthyroidism and PD was maintained in women younger than 70 years.
Conclusion:
Both hyperthyroidism and hypothyroidism were associated with higher risk of PD, particularly for women younger than 70 years and men older than 70 years, respectively.
Keywords
INTRODUCTION
Thyroid hormones (THs) are essential for growth, neuronal development, reproduction and the regulation of energy metabolism in virtually every vertebrate, including humans [1, 2]. TH synthesis is stimulated by thyroid stimulating hormone (TSH) from the anterior pituitary gland, and TH secretion is finely modulated by thyrotropin-releasing hormone (TRH) from the paraventricular nucleus of the hypothalamus [3]. TSH binds to a G-protein coupled receptor on thyroid follicular cells to stimulate TH (T4) synthesis and release [4]. T4 is converted in the periphery to the more active molecule triiodothyronine (T3), which elicits most thyroid cellular actions. TRH synthesis is induced by cold, thereby activating the TH response to hypothermia. TSH synthesis and production are markedly suppressed by several factors, including THs, dopamine, somatostatin, and glucocorticoids. Thyroid dysfunction is a relatively common condition, and it has a wide range of symptoms. Hypothyroidism and hyperthyroidism are common conditions with potentially devastating health sequences that affect populations worldwide. In particular, it is well known that abnormal TH signaling causes a variety of neurological and metabolic disorders.
Parkinson’s disease (PD) is a heterogeneous neurodegenerative disorder caused by the progressive degeneration of dopamine-containing neurons in the substantia nigra pars compacta within the midbrain. PD is the second most common neurodegenerative disorder and affects 1–3% of the population over 50 years of age [5]. The deficiency of striatal dopaminergic neurons due to the loss of substantia nigra neurons produces key clinical manifestations of PD, such as bradykinesia, rigidity, resting tremor, gait disturbance, and postural instability. Despite considerable progress in understanding PD pathophysiology, the exact mechanisms of dopaminergic neuron death in the substantia nigra are still not completely understood.
The regulatory interaction between the hypothala-mic-pituitary-thyroid (HPT) axis and dopaminergic system and several epidemiologic studies regarding PD and hypothyroidism suggest that these two diseases might share common pathophysiological features or that one disease may affect the development or progression of the other disease [6, 7].
Therefore, the main purpose of this study was to examine the association between thyroid diseases and PD in an age-, sex-, income-, and the region of residence-matched cohort. The second goal was to investigate whether these associations were significantly different among the groups in a subgroup analysis stratified by age and sex. Additionally, we discussed the pathophysiology of the association between thyroid diseases and PD by reviewing previous studies.
METHODS
Study population
The ethics committee of Hallym University (2019-10-023) approved this study. Written informed consent was waived by the Institutional Review Board. All analyses adhered to the guidelines and regulations of the ethics committee of Hallym University. A detailed description of The Korean National Health Insurance Service (NHIS)-Health Screening Cohort data is described elsewhere [8].
Levothyroxine users (independent variable)
Levothyroxine users were enrolled if participants had taken levothyroxine for ≥3 months.
Definition of goiter (independent variable)
Goiter was defined as a diagnosis with the International Statistical Classification of Diseases and Related Health Problems (ICD)-10 code E04 (other nontoxic goiter). Among these patients, we selected individuals who were treated ≥2 times.
Definition of hypothyroidism (independent variable)
Hypothyroidism was defined as a diagnosis with the ICD-10 codes E02 (subclinical iodine-deficiency hypothyroidism) and E03 (other hypothyroidism). Among these patients, we selected individuals who were treated ≥2 times.
Definition of thyroiditis (independent variable)
Thyroiditis was defined as a diagnosis with the ICD-10 code E06 (thyroiditis). Among these patients, we selected individuals who were treated ≥2 times.
Definition of hyperthyroidism (independent variable)
Hyperthyroidism was defined as a diagnosis with the ICD-10 code E05 (hyperthyroidism). Among these patients, we selected individuals who were treated ≥2 times.
Definition of PD (dependent variable)
PD was defined as a diagnosis with the ICD-10 code G20 (Parkinson’s disease). To ensure the accuracy of the diagnosis, we selected only participants who visited hospitals or clinics ≥2 times with the diagnosis of PD.
Participant selection
PD patients were selected from 514,866 participants with 615,488,428 medical claim codes from 2002 through 2015 (n = 6,483). The control group included participants who were not defined as having PD from 2002 through 2015 (n = 508,383). To select PD patients who were diagnosed for the first time, PD patients diagnosed in 2002 were excluded (washout period, n = 402). Participants who were treated for head trauma histories (ICD-10 codes: S00 to S09, diagnosed by neurologists, neurosurgeons, or emergency medicine doctors) ≥2 times and underwent head and neck computed tomography (CT) evaluations (claim codes: HA401–HA416, HA441–HA443, HA451–HA453, HA461–HA463, and HA471–HA473) were excluded (n = 481 for PD patients, n = 12,989 for control participants). Participants who were treated for brain tumors (ICD-10 codes: C70 to C72) ≥2 times were excluded (n = 14 for PD patients, n = 855 for control participants). PD patients were matched 1:4 for age, sex, income, and the region of residence with control participants. To minimize selection bias, control participants were selected with a random number order. The index date of each PD patient was set as the time of treatment for PD. The index date of control participants was set as the index date of their matched PD patients. Therefore, each PD patient and the matched control participant had the same index date. During the matching procedure, 472,195 control participants were excluded. Ultimately, 5,586 PD patients were matched 1:4 with 22,344 control participants (Fig. 1).
A schematic illustration of the participant selection process used in this study. Among a total of 514,866 participants, 5,586 PD patients were matched 1:4 for age, sex, income, and the region of residence with 22,344 control participants.
Covariates
Individuals were divided into age groups based on 5-year intervals: 40–44, 45–49, 50–54 ... , and 85+ years old. A total of 10 age groups were specified. Individuals were placed into income groups with 5 classes (class 1 [lowest income]-5 [highest income]). The region of residence was characterized as urban (Seoul, Busan, Daegu, Incheon, Gwangju, Daejeon, and Ulsan) or rural (Gyeonggi, Gangwon, Chungcheongbuk, Chungcheongnam, Jeollabuk, Jeollanam, Gyeongsangbuk, Gyeongsangnam, and Jeju) areas.
Tobacco smoking was categorized based on the participant’s current smoking status (nonsmoker, past smoker, and current smoker). Alcohol consumption was categorized on the basis of the frequency of alcohol consumption (<1 time a week and ≥1 time a week). Obesity was measured using body mass index (BMI, kg/m2). BMI was categorized as <18.5 (underweight), ≥18.5 to <23 (normal), ≥23 to <25 (overweight), ≥25 to <30 (obese I), and ≥30 (obese II) based on the Asia-Pacific criteria based on the Western Pacific Regional Office (WPRO) 2000 [9]. Missing BMI values (28/27,930 [0.100%]) were replaced by mean values for the variable from final selected participants. The Charlson Comorbidity Index (CCI) is widely used to measure disease burden using 17 comorbidities. In our study, we excluded cancer and metastatic cancer from the CCI score. A score was given to each participant depending on the severity and number of diseases. The CCI was measured as a continuous variable (0 [no comorbidities] through 29 [multiple comorbidities]) [10].
For PD patients, patients were also considered to have thyroid cancer (ICD-10 code: C73) if they were treated ≥2 times.
Statistical analyses
General characteristics were compared between the PD and control groups using the Chi-square test.
To analyze the odds ratios (ORs) with 95% confidence intervals (CIs), a conditional logistic regression model for PD and thyroid diseases was generated. An unadjusted model, model 1 (adjusted for obesity, smoking, alcohol consumption, CCI scores and thyroid cancer), and model 2 (adjustments from model one and additionally adjustments for levothyroxine treatment, goiter, hypothyroidism, thyroiditis, and hyperthyroidism) were generated. Because levothyroxine treatment, goiter, hypothyroidism, thyroiditis, and hyperthyroidism histories were closely related (Supplementary Table 1), they were adjusted for in model 2. Analyses were stratified by age, sex, income, and the region of residence.
For subgroup analyses, we divided participants by age and sex (<70 years old and ≥70 years old; men and women) and by income and the region of residence (low and high; urban and rural) using the unadjusted model, model 1, and model 2.
Two-tailed analyses were performed, and significance was defined as p values less than 0.05. SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses.
RESULTS
The time distances between the independent variables and the index date were 53.59 months (standard deviation [SD] = 37.67) vs. 56.11 (SD = 39.04) for goiter, 51.72 (SD = 42.17) vs. 61.59 (SD = 41.38) for hypothyroidism, 59.64 (SD = 43.74) vs. 60.96 (SD = 39.64) for thyroiditis, and 53.64 (SD = 43.28) vs. 70.28 (SD = 42.27) for hyperthyroidism in the PD and control groups, respectively.
The rates of levothyroxine treatment for more than 3 months, hypothyroidism, and hyperthyroidism for the PD group (3.2%, 3.8%, and 2.8%) were significantly higher than those for the control group (2.5%, 2.9%, and 1.9%, p < 0.05, Table 1). The general characteristics (age, sex, income, and the region of residence) of the participants were the same between the 2 groups due to the matching procedures (p = 1.000, Table 1). The rate of smoking and alcohol consumption and CCI scores significantly differed between the PD and control groups (p < 0.001, Table 1).
General characteristics of participants
CCI, Charlson comorbidity index; PD, Parkinson’s disease. *Chi-square test. Significance at p < 0.05. †Obesity (BMI, body mass index, kg/m2) was categorized as <18.5 (underweight), ≥18.5 to <23 (normal), ≥23 to <25 (overweight), ≥25 to <30 (obese I), and ≥30 (obese II).
The adjusted ORs in model 1, which was adjusted for obesity, smoking status, alcohol consumption, CCI scores and thyroid cancer, were 1.32 (95% CI 1.09–1.60, p = 0.005), 1.23 (95% CI 1.03–1.46, p = 0.021), 1.35 (95% CI 1.15–1.60, P < 0.001), and 1.47 (95% CI 1.21–1.78, p < 0.001, Table 2) in the PD group for levothyroxine treatment, goiter, hypothyroidism and hyperthyroidism, respectively. However, only hypothyroidism and hyperthyroidism were significantly associated with PD after adjusting for all thyroid conditions. The adjusted ORs for hypothyroidism and hyperthyroidism in model 2, which was adjusted for levothyroxine treatment, goiter, hypothyroidism, thyroiditis and hyperthyroidism, were 1.25 (95% CI 1.01–1.55, p = 0.044) and 1.37 (95% CI 1.13–1.67, p = 0.002, Table 2), respectively, for the PD group.
Unadjusted and adjusted odds ratios (95% confidence intervals) for levothyroxine treatment, goiter, hypothyroidism, thyroiditis, and hyperthyroidism for PD
CCI, Charlson comorbidity index; PD, Parkinson’s disease. *Conditional logistic regression model. Significance at p < 0.05. †PD patients were matched 1:4 for age, sex, income, and the region of residence with control participants. ‡Model 1 was adjusted for obesity, smoking, alcohol consumption, CCI scores, and thyroid cancer. §Model 2 was adjusted from model 1 and additionally adjusted for levothyroxine treatment, goiter, hypothyroidism, thyroiditis, and hyperthyroidism.
In subgroup analyses stratified for age and sex (Supplementary Table 2, Fig. 2), the adjusted OR in model 2 for levothyroxine treatment was higher in the PD group among women ≥70 years (p < 0.05). The adjusted ORs in model 2 for goiter and hyperthyroidism were higher in the PD group among women aged <70 years, and the adjusted OR for hypothyroidism was higher in the PD group among men aged ≥70 years (each p < 0.05). In subgroup analyses stratified for income and the region of residence (Supplementary Table 3, Fig. 3), the adjusted OR in model 2 for hypothyroidism was higher in the PD group among urban participants with a high income (p < 0.05). The adjusted OR in model 2 for hyperthyroidism was higher in the PD group for both urban and rural participants with high incomes (each p < 0.05).
Adjusted ORs and 95% CIs were calculated for each thyroid condition while adjusting for all covariates according to age and sex.
Adjusted ORs and 95% CIs were calculated for each thyroid condition while adjusting for all covariates according to income and the region of residence.
DISCUSSION
This nested case-control study examined the association with various thyroid diseases and PD using an age-, sex-, income- and region of residence-matched cohort. In addition, we reported ORs for both groups that were adjusted for potential confounders, such as smoking status, alcohol consumption, CCI scores and thyroid conditions. The main findings of this study demonstrated that both hyperthyroidism and hypothyroidism were associated with higher risk of PD after adjusting for age, sex, income, region of residence, and comorbidities. The adjusted ORs for hypothyroidism and hyperthyroidism in model 2 for the PD group were 1.25 (95% CI = 1.01–1.55) and 1.37 (95% CI = 1.13–1.67), respectively. These results were consistent with those of the subgroup of men aged ≥70 years and the subgroup of women aged <70 years, respectively. Although goiter did not differ between the PD group and the control group in the overall analysis, the adjusted OR in model 2 was higher for the PD group in the subgroup of women aged <70 years (OR = 1.55, 95% CI = 1.16–2.08). Given that these associations between thyroid diseases and PD are age- and sex-dependent, it can be assumed that the susceptibility to PD development or progression may differ depending on age or sex. It seems that hypothyroidism makes older men more susceptible to PD, whereas hyperthyroidism makes younger women more susceptible to PD.
In fact, there are some conflicts regarding whether hypothyroidism or hyperthyroidism increase the susceptibility to PD. Chen et al. recently performed a large-population cohort study including a total of 4725 patients and reported that patients with hypothyroidism had increased risk of developing PD (Hazzard ratio [HR] = 1.77, 95% CI = 1.13–2.76) [6]. Additionally, Li et al. reported that follow-up of patients with Grave’s disease/hyperthyroidism or Hashimoto’s disease/hypothyroidism showed that these patients had increased risk of subsequent PD (166 patients, standardized incidence ratios [SIR] = 1.63, 95% CI = 1.39–1.90 and 66 patients, SIR = 2.40, 95% CI = 1.86–3.06, respectively) [7]. However, other studies did not find a significant difference in the presence of hypothyroidism or hyperthyroidism between a population of PD patients and a control group. Munhoz et al. showed that hypothyroidism was not more prevalent in 95 PD patients compared with 102 age-matched controls [11]. Tandeter et al. reported that hypothyroidism was not more common in patients with PD in a retrospective review of 92 PD patients and 226 randomly selected controls [12].
The relationship between THs and dopaminergic neurons is relatively well established. First, hypothyroidism often leads to the loss of dopaminergic neurons, and it has been demonstrated in mice that the Girk2 mutation that causes hypothyroidism [13]. In addition, reduced levels of serum TH in these mutant mice are associated with a decrease in striatal transforming growth factor α (TGF-α) expression, which is important for the survival of mesencephalic dopaminergic neurons [14]. A recent study demonstrated that TH induces the differentiation of mouse dopaminergic neurons from the embryonic ventral midbrain by upregulating orthodenticle homeobox 2 (OTX2) [15].
Meanwhile, oxidative stress is one of the chief contributing factors to dopaminergic neuron loss and PD progression [16, 17]. Previous studies have shown that both hyperthyroidism and hypothyroidism are related to oxidative stress and cellular damage [18–20]. One study showed that hypothyroidism may enhance the susceptibility for increased low-density lipid protein-C (LDL-C) oxidation and that elevated LDL-C can be oxidatively attacked by free radicals, while hyperthyroidism accelerates mitochondrial oxidative metabolism, resulting in increased free radical production and lipid peroxidase [21]. Moreover, neuroinflammation also plays a crucial role in the pathogenesis of PD because it leads to the loss of neurons and altered neurotransmitter levels [16]. According to Wu et al., both inflammatory processes and oxidative stress may contribute to neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-related neurodegeneration [22]. Inflammation, which results from hormone and cytokine changes, leads to oxidative stress and affects thyroid function, causing pituitary-thyroid axis depression. Inflammation and oxidative stress can affect the initiation or progression of PD, since they are closely related processes and are also related to hormonal dysregulation in a reciprocal way. A recent comprehensive review explained that the HPT axis regulates dopamine release, while PD might disturb the HPT axis; therefore, hypothyroidism and hyperthyroidism lead to disturbances in dopamine signaling and cause dopaminergic neuron loss by upregulating oxidative stress [23].
Additionally, there are several studies suggesting that thyroid disease and PD symptoms are linked. Low levels of THs are associated with motor symptom severity [24], and hyperthyroidism exacerbates PD symptoms such as tremor and dyskinesia [25]. Umehara et al. showed that free T3 levels were lower in patients with the akinetic-rigid motor subtype than in those with the tremor-dominant-type or mixed-type subtype, and there was a significant negative correlation between free T3 levels and disease severity [24]. Another study showed that restriction of maternal THs downregulates the dopaminergic enzyme tyrosine hydroxylase in the striatum and increases sensitivity to Parkinson-like movement in rat offspring, suggesting a possible mechanism for the above observations in patients [26]. Thyrotoxicosis has been reported in PD patients with severe tremor [27, 28], and anti-thyroid treatment is effective in controlling the ‘on-off’ phenomenon and dyskinesia in thyrotoxic patients with PD [29, 30]. The exact mechanism that drives the relationship between hyperthyroidism and PD symptom aggravation is also unclear; the increased sensitivity of adrenergic receptors to TH and increased sensitivity of postsynaptic dopaminergic receptors could be plausible explanations [25, 27].
A strength of this study is the utilization of NIHS data, which includes all Korean citizens without exceptions; therefore, this cohort is a large, highly representative, nationwide population sample, and there were no missing participants. Due to the large number of participants, the control group was randomly selected and subsequently matched to minimize the impacts of confounders, including age, sex, income, and the region of residence. Furthermore, we adjusted our analyses for all potential confounding factors, not only risk factors but also thyroid conditions that might be related to each other.
Nevertheless, several shortcomings should be considered when interpreting our findings. First, we examined the association between hypothyroidism or hyperthyroidism and PD, but we did not account for a causal relationship between these diseases. Further studies, especially prospective and randomized controlled trials, are required to definitively show that there is a causal association between each thyroid disease and PD. Second, we did not assess the severity of PD or the level of TH among participants. PD is a complex neurodegenerative disorder with a broad spectrum of motor and nonmotor features that require a close assessment using a clinical rating scale. A variety of clinical rating scales, particularly the Unified Parkinson’s Disease Rating Scale (UPDRS; the most commonly used scale to assess disease severity) and reliable diagnostic, presymptomatic and progression biomarkers are being developed to aid in diagnosing and tracking the disease course [31]. As mentioned above, the motor type and disease severity (UPDRS) of PD were correlated with thyroid function (free T3) [24]. Consistently, a recent study revealed that TH levels are correlated with motor subtypes and disease severity in euthyroid patients with PD [32]. However, UPDRS and serum TH levels are not available in the Korean NIHS database; thus, we could not perform further analysis to determine whether the extent of thyroid function is correlated with the subtype or severity of PD. Likewise, we could not assess thyroid gland function using the thyroid function test (TFT) in this study. Because levothyroxine-treated patients can achieve normal serum TSH levels, even patients with hypothyroidism may have normal TFT results. Hence, it is necessary to accurately evaluate thyroid function with a TFT at the time of diagnosis to investigate whether there is an obvious association between hypothyroidism and PD. Although TFT results were not available in the Korean NIHS database, we observed a significant association after examining the history of levothyroxine treatment for more than 3 months and adjusting for this factor. Third, hypothyroidism or hyperthyroidism and PD often show common clinical features. Hypothyroidism patients usually have PD symptoms such as rigidity, hypokinesia, facial hypomimia and voice abnormalities [11]. In addition, hyperthyroidism exhibits clinical signs such as tremor, sweating, and weight loss, which are experienced by many PD patients, and it even aggravates symptoms such as tremor and dyskinesia [25]. Due to these clinical similarities, the diagnosis of each disease is ambiguous, or one disease may mask the appearance of the other. Finally, clinical challenges of the disease include the inability to make a definite diagnosis at the early stages and difficulties in predicting disease progression. Therefore, we only included patients who had been diagnosed with PD at least twice prior to the index date in this study in order to increase the validity of the PD diagnosis by minimizing misdiagnosis (Supplementary Figure 1). Patients diagnosed with PD who visit a clinic once may be misdiagnosed, which would increase false-positive diagnoses.
In conclusion, we found that both hyperthyroidism and hypothyroidism were significantly associated with PD. In subgroup analyses, the association between hypothyroidism and PD was maintained in the group of men older than 70 years, and the association between hyperthyroidism and PD was maintained in the group of women younger than 70 years old. For hypothyroidism and hyperthyroidism, some of the observed associations have been previously described, and our findings provide further evidence for these associations. Further investigations should be conducted to provide clear evidence and to identify the underlying mechanism responsible for the association between thyroid diseases and PD, which may provide additional insights into the pathophysiology of PD.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
