Abstract
Aim
Bipolar disorder (BD) and major depression have been associated with an increased risk of developing Parkinson's disease (PD); however, few studies have directly compared the risk of PD development between patients with BD and major depression while considering relevant risk factors and psychotropic medications.
Methods
Using the Taiwan National Health Insurance Research Database, 21,186 patients with BD, 21,188 patients with major depression, and 42,374 controls were enrolled between 2001 and 2009, and followed until the end of 2011. Individuals who developed PD during the follow‐up period were identified. Cox regression models were used to analyze the hazard ratio (HR) of developing PD, adjusting for demographic factors, comorbidities, and psychotropic medication usage.
Results
Both patients with BD [HR 8.63, 95% confidence interval (CI) 6.35–11.72] and those with major depression (HR 5.68, 95% CI 4.15–7.78) had an elevated risk of subsequent PD compared to the controls. Patients with BD were associated with a 51% increased risk of subsequent PD compared with patients with major depression. Long‐term treatment with antiepileptic mood stabilizers was associated with increased PD risk among patients with late‐onset BD and high Charlson comorbidity index scores. Lithium was not associated with an increased PD risk.
Conclusions
The study highlights an elevated PD risk in patients with BD and major depression compared to the controls, with BD patients at highest risk. Further research is needed to elucidate the complex interplay between psychotropic medications and neurodegenerative processes in BD, aiming to optimize therapeutic strategies and improve patient outcomes.
Keywords: bipolar disorder, lithium, mood stabilizers, parkinsonism
Emerging evidence underscores the potential linkage between bipolar disorder (BD) and a susceptibility to neurodegenerative diseases, while no definite evidence of degenerative neuropathology has been confirmed in BD. 1 Our previous study has indicated an increased risk of developing Parkinson's disease (PD) among patients with BD compared with the general population, 2 and a recent meta‐analysis has also found a 3.35‐fold increased risk of PD among patients with BD compared to controls. 3 Compared with the recently revealed association between BD and PD, the bi‐directional relationship between major depression and PD has been proposed and investigated, involving shared neurobiological pathways and mutual impact on disease progression and symptomatology. Depression represents a prevalent symptom of PD, with a prevalence around 40% 4 ; individuals experiencing depression display an increased likelihood of subsequent PD diagnosis. 5 Previous investigations have indicated an elevated risk of PD development associated with both BD and major depression. Harrison et al. further reported that patients with BD had an odds ratio of 2.15 (1.99–2.33) for developing PD compared to patients with depression. 6 Comparative studies directly contrasting the PD development risk between patients with BD, patients with major depression, and matched controls are still lacking. Furthermore, the majority of preceding studies have not accounted for relevant PD‐related risk factors, notably cerebrovascular diseases, metabolic syndrome, and the usage of psychotropic medications, within their regression models.
An important confounder when evaluating risk of PD in psychiatric disorders is the usage of psychotropic medications. The standard treatment for BD that includes mood stabilizers (lithium and antiepileptic drugs) and antipsychotic medications can contribute to drug‐induced parkinsonism (DIP), which may be clinically indistinguishable from PD, both being characterized by bradykinesia, resting tremor, and rigidity. 7 In experimental models, lithium holds promise as a neuroprotective agent by promoting neurogenesis, enhancing mitochondrial function, and modulating inflammatory processes, through the inhibition of enzymes such as glycogen synthase kinase‐3 beta (GSK‐3β). 8 Patients with BD who had been receiving long‐term lithium treatment had reduced risk of dementia and cognitive decline; neuroimaging studies have also shown increased hippocampal volume and cortical thickness after lithium treatment. 9 While neuroprotective actions of lithium have also been demonstrated in animal models of PD, 10 a retrospective cohort study using administrative databases suggested lithium use may increase risk for PD among patients with BD. 11 Whether lithium predisposes individuals to PD is still uncertain and warrants further investigation. In addition, since PD tends to affect males more than females, 12 the risk estimates of PD in BD per sex should be examined, as this aspect has not been thoroughly explored in prior research.
Using the Taiwan National Health Insurance Research Database (NHIRD) in a large‐scale longitudinal study, we aimed to examine the differential risk of PD occurrence among patients with major depression and BD. The risk of PD per sex and the influence of psychotropic agents on PD development were also examined. Our hypothesis posited an elevated risk of PD among both patients with BD and major depression, with potentially heightened risk among those with BD.
Methods
Data source
The Taiwan National Health Insurance Program, initiated in 1995 as a universal single‐payer system covering nearly all of Taiwan's population of approximately 23 million individuals, boasts a remarkable coverage rate of approximately 99.6% by the end of 2010 (http://www.nhi.gov.tw/). Administered by the National Health Research Institute, the NHIRD offers a wealth of anonymized patient data, including demographic details and clinical records. Each patient is assigned a unique identifier to safeguard privacy, enabling continuous tracking within the database. Utilizing the International Classification of Diseases, 9th Revision, Clinical Modification (ICD‐9‐CM) for diagnostic codes, the NHIRD serves as a valuable resource for health service research, reducing selection bias due to its nationwide scope. Access to the NHIRD requires formal application, overseen by the Department of Health and the Bureau of the NHI Program. This database has been widely utilized in numerous epidemiological studies in Taiwan. 13 The Institutional Review Board of the Taipei Veterans General Hospital granted approval for this study, with informed consent waived due to the deidentified nature of the NHIRD. The study was conducted in accordance with the Declaration of Helsinki.
Study population
Adults aged between 45 and 65 years who were diagnosed as having major depression (ICD‐9‐CM codes: 296.2× and 296.3×) or BD (ICD‐9‐CM codes: ICD‐9‐CM codes: 296.0×, 296.1×, 296.4×, 296.5×, 296.6×, 296.7×, 296.80, 296.81, 296.89) by board certificated psychiatrists at least twice between January 1, 2001 and December 31, 2009, and who had no history of PD and related diseases (ICD‐9‐CM code 332) before enrollment, were identified from the specialized dataset of mental disorders in the NHIRD and were included as the mood disorder cohort in this study. Enrollment time was defined as the time of BD or major depression diagnosis. Exact matching (1:1) based on age, sex, income, and urbanization level was performed between two groups (BD, major depression) (Fig. 1). A control cohort was then randomly chosen and exact matched to the study cohort (BD, major depression) in a 1:1 fashion in terms of age, sex, income, and urbanization level, excluding participants with previous diagnoses of major psychiatric disorders (ICD‐9‐CM codes: 295, 296, and 297) or PD. A diagnosis of PD (ICD‐9‐CM code 332.0) made by board‐certificated neurologists was recorded during the follow‐up period (from enrollment to December 31, 2011, or until death). Potential PD‐related medical comorbidities were evaluated as confounders, encompassing alcohol‐related disorders (ICD‐9 codes: 291, 303, 305.0, 357.5, 425.5, 535.3, 571.0, or 571.1–571.3), hypertension (ICD‐9‐CM codes: 401–405), dyslipidemia (ICD‐9‐CM code: 272), diabetes mellitus (ICD‐9 code: 250), cerebrovascular diseases (ICD‐9 codes: 430–438), and traumatic brain injury (ICD‐9 codes: 800–804, 850–854, 959). Each diagnosis required confirmation by corresponding physicians at least twice between January 1, 2001 and December 31, 2009 to ensure diagnostic accuracy. The Charlson Comorbidity Index (CCI) score was used to assess participants' overall systemic health and the severity of comorbid medical conditions at enrollment. 14 Income categories were defined as low (monthly income below New Taiwan Dollar [NTD] 15,840), medium (monthly income NTD 15,840–25,000), or high (monthly income >NTD 25,000), with NTD 15,840 representing tier 1 monthly salaries according to the 38‐tier insured amount designated by the NHI. The average post‐graduation salary was NTD 25,000, as per data from the Ministry of Labor in Taiwan in 2015. Additionally, the level of urbanization (ranging from level 1 to level 5, with level 1 being the most urbanized and level 5 the least) was assessed. 15 The NHIRD provides data on medication use based on the World Health Organization's Anatomical Therapeutic Chemical (ATC) classification system, enabling an analysis of the dose–response relationship between psychotropic drug use and the development of Parkinson's disease. To ensure comparability across different medications, we utilized the defined daily dose (DDD), which standardizes drug consumption. Additionally, we calculated the cumulative defined daily dose (cDDD) for each participant throughout the follow‐up period. The DDD, as recommended by the World Health Organization, represents the assumed average maintenance dose per day for a drug's primary indication. For example, the cDDD for lithium was determined by dividing the total lithium dose during the study period by the DDD for lithium, where one DDD equals 24 mmol of lithium. The cDDD for each psychotropic drug was then categorized into three groups: <30 cDDD, 30–364 cDDD, and ≥365 cDDD.
Figure 1.
Study flowchart.
Statistical analysis
For intergroup comparisons, Analysis of Variance (ANOVA) was utilized for continuous variables, and the Pearson χ 2 test was applied for nominal variables, as appropriate. Levene's test assessed variance homogeneity among groups. When variance distribution was uneven, Welch's one‐way ANOVA was used, followed by Games‐Howell post‐hoc pairwise comparisons. Three Cox regression models were used to investigate the hazard ratio (HR) with 95% CI of developing PD: model 1 for the crude HR, model 2 for the adjusted HR after controlling for demographic data (age, sex, income, level of urbanization, all cause clinical visits), and model 3 for the adjusted HR after controlling for demographic data, Charlson comorbidity index (CCI), smoking and medical comorbidities. Sensitivity analyses by excluding data from the initial 1‐ and 3‐year observations were conducted to validate our results. In addition, another Cox regression analysis was performed to compare the risk of subsequent PD between patients with major depression and those with BD by only including these two groups, adjusting for demographic factors, CCI score, smoking, and medical comorbidities. In addition, we performed subgroup analyses on the risk of PD in sex‐ and age‐stratified BD groups: age <50 years and age ≥50 years, as most research considers BD that begins at 50 years old or later to be late‐onset BD (LOBD). 16 While the medium score of CCI for patients with BD was 2 (interquartile range, 1–4) in this study, additional subgroup analyses by dividing patients to high (>2) and low CCI score (≤2) group were performed. 17 A 2‐tailed P‐value <0.05 was considered statistically significant. All data processing and statistical analyses were performed using the Statistical Package for Social Science (SPSS) Version 17 software (Chicago: SPSS Inc.) and Statistical Analysis Software (SAS) Version 9.1 (SAS Institute, Cary, NC).
Results
A total of 21,186 patients with BD, 21,188 patients with major depression, and 42,374 controls were enrolled. During the follow‐up period, patients with BD had a higher incidence of developing PD than patients with major depression and the controls; patients with BD or major depression had a shorter duration from enrollment to PD diagnosis than the control group (Table 1). Patients with BD or major depression had an elevated prevalence of medical comorbidities, including alcohol‐related disorders, hypertension, dyslipidemia, diabetes mellitus, cerebrovascular diseases, traumatic brain injury, and obesity (Table 1). A higher proportion of patients with BD received long‐term treatment of lithium, antiepileptic mood stabilizers (AED‐MSs), and second‐generation antipsychotics compared to the other groups (Table 1).
Table 1.
Demographic data and incidence of Parkinson's disease among patients with major depression or bipolar disorder and the control group
| Control (n = 42,374) | Major depression (n = 21,188) | Bipolar disorder (n = 21,186) | P | Post‐hoc | |
|---|---|---|---|---|---|
| Age at enrollment (years, SD) | 53.09 (5.50) | 53.11 (5.47) | 53.09 (5.48) | 0.842 | |
| Sex (female, %) | 25,033 (59.1) | 12,517 (59.1) | 12,516 (59.1) | 1 | |
| Parkinson's disease (n, %) | 50 (0.12) | 210 (0.99) | 300 (1.42) | <0.001 | BD > MDD > HC |
| Age at diagnosis (years, SD) | 66.40 (5.04) | 61.99 (6.24) | 62.06 (6.00) | <0.001 | MDD, BD < HC |
| Duration between enrollment and PD (years, SD) | 6.50 (2.28) | 4.17 (3.22) | 4.71 (3.08) | <0.001 | MDD, BD < HC |
| Comorbidities (n, %) | |||||
| Alcohol use disorder | 1221 (2.9) | 1952 (9.2) | 2664 (12.6) | <0.001 | BD > MDD > HC |
| Hypertension | 13,471 (31.8) | 8716 (41.1) | 8339 (39.4) | <0.001 | MDD > BD > HC |
| Dyslipidemia | 8793 (20.8) | 6446 (30.4) | 6062 (28.6) | <0.001 | MDD > BD > HC |
| Diabetes mellitus | 6592 (15.6) | 4565 (21.5) | 5217 (24.6) | <0.001 | BD > MDD > HC |
| Cerebrovascular diseases | 2944 (6.9) | 3030 (14.3) | 2670 (12.6) | <0.001 | MDD > BD > HC |
| Traumatic brain injury | 948 (2.2) | 1117 (5.3) | 1505 (7.1) | <0.001 | BD > MDD > HC |
| Obesity | 775 (1.8) | 605 (2.9) | 930 (4.4) | <0.001 | BD > MDD > HC |
| CCI score (SD) | 1.48 (1.72) | 2.52 (2.11) | 2.42 (2.09) | <0.001 | MDD > BD > HC |
| Smoking (n, %) | 816 (1.9) | 929 (4.4) | 1176 (5.6) | <0.001 | BD > MDD > HC |
| All‐cause clinical visits (times per year, SD) | 9.89 (9.73) | 20.99 (21.16) | 19.73 (22.34) | <0.001 | MDD > BD > HC |
| All‐cause mortality, n (%) | 507 (1.2) | 909 (4.3) | 1214 (5.7) | <0.001 | BD > MDD > HC |
| Antidepressants treatment (n, %) | <0.001 | ||||
| <30 cDDD | 41,052 (96.9) | 3021 (14.3) | 6379 (30.1) | ||
| 30–364 cDDD | 1005 (2.4) | 7943 (37.5) | 4935 (23.3) | ||
| ≧365 cDDD | 317 (0.7) | 10,224 (48.3) | 9872 (46.6) | ||
| Lithium treatment (n, %) | <0.001 | ||||
| <30 cDDD | 42,365 (99.9) | 20,865 (98.5) | 15,079 (71.2) | ||
| 30–364 cDDD | 8 (0.0) | 247 (1.2) | 2582 (12.2) | ||
| ≧365 cDDD | 1 (0.0) | 76 (0.4) | 3525 (16.6) | ||
| Antiepileptic mood stabilizers (n, %) | <0.001 | ||||
| <30 cDDD | 41,615 (98.2) | 19,151 (90.4) | 10,587 (50.0) | ||
| 30–364 cDDD | 608 (1.4) | 1607 (7.6) | 5815 (27.4) | ||
| ≧365 cDDD | 151 (0.6) | 430 (2.0) | 4784 (22.6) | ||
| Antipsychotics treatment (n, %) | <0.001 | ||||
| <30 cDDD | 42,278 (99.8) | 18,520 (87.4) | 11,253 (53.1) | ||
| 30–364 cDDD | 86 (0.2) | 1840 (8.7) | 5303 (25.0) | ||
| ≧365 cDDD | 10 (0.02) | 828 (3.9) | 4630 (21.9) | ||
| Urbanization level (n, %) | 1 | ||||
| 1 (most urbanized) | 9768 (23.1) | 4884 (23.1) | 4884 (23.1) | ||
| 2 | 14,112 (33.3) | 7056 (33.3) | 7056 (33.3) | ||
| 3 | 4012 (9.5) | 2006 (9.5) | 2006 (9.5) | ||
| 4 | 5460 (12.9) | 2730 (12.9) | 2730 (12.9) | ||
| 5 (most rural) | 9022 (21.3) | 4512 (21.3) | 4510 (21.3) | ||
| Income‐related insured amount (n, %) † | 1 | ||||
| ≤15,840 NTD/month | 22,708 (53.6) | 11,354 (53.6) | 11,354 (53.6) | ||
| 15,841–25,000 NTD/month | 10,907 (25.7) | 5454 (25.7) | 5453 (25.7) | ||
| ≥25,001 NTD/month | 8759 (20.7) | 4380 (20.7) | 4379 (20.7) | ||
CCI, Charlson Comorbidity Index; cDDD, cumulative defined daily dose; NTD, new Taiwan dollar; SD, standard deviation.
1.00 Taiwanese new dollar = US$0.035.
Kaplan–Meier survival analysis demonstrated a significantly elevated risk of PD development among individuals with BD or major depression compared to controls (Fig. 2). Cox regression analysis, adjusting for demographic variables and medical comorbidities, revealed that both BD (HR 8.63, 95% CI 6.35–11.72) and major depression (HR 5.68, 95% CI 4.15–7.78) were associated with an increased risk of subsequent PD during follow‐up, relative to the control cohort (Table 2). Furthermore, we compared the risk of subsequent PD development among patients with BD with that among major depressive patients. The Cox regression model indicated that patients with BD had a higher risk (HR 1.51, 95% CI 1.26–1.80, P < 0.001) of developing PD compared with those with major depression after adjusting for demographic factors, CCI score, smoking, and medical comorbidities.
Figure 2.
Survival curve of developing Parkinson's disease among patients with bipolar disorder or major depression and control group.
Table 2.
Risk of developing Parkinson's disease among patients with major depression or bipolar disorder
| Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|
| Crude HR | 95% CI | Adjusted HR | 95% CI | Adjusted HR | 95% CI | |
| Control | 1 | 1 | 1 | |||
| Major depression | 8.70 | 6.39–11.85 | 6.64 | 4.87–9.06 | 5.68 | 4.15–7.78 |
| Bipolar disorder | 12.56 | 9.31–16.94 | 9.83 | 7.27–13.30 | 8.63 | 6.35–11.72 |
| Lithium‐free cohort | ||||||
| Control | 1 | 1 | 1 | |||
| Major depression | 8.72 | 6.40–11.88 | 6.64 | 4.86–9.06 | 5.51 | 4.01–7.56 |
| Bipolar disorder | 11.75 | 8.60–16.06 | 8.68 | 6.32–11.91 | 7.45 | 5.40–10.28 |
| Mood stabilizers‐free cohort | ||||||
| Control | 1 | 1 | 1 | |||
| Major depression | 9.27 | 6.74–12.76 | 7.10 | 5.14–9.80 | 5.65 | 4.06–7.86 |
| Bipolar disorder | 11.30 | 7.98–16.01 | 7.58 | 5.30–10.83 | 6.56 | 4.57–9.40 |
Model 2 adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits). Model 3 adjusted for demographic data, Charlson comorbidity index, smoking, and medical comorbidities.
Subgroup analyses of men, women, age <50 years, and age ≥50 years exhibited the same trend of a higher PD risk among patients with BD (Table 3). There was no sex difference in the risk of subsequent PD among patients with BD. Among medical comorbidities, history of cerebrovascular diseases and traumatic brain injury were risk factors of subsequent PD among patients with BD. Long‐term use (≥ 365 cDDD) of lithium and AED‐MSs were associated with higher risks of PD among patients with LOBD (Table 3). We further stratified patients with LOBD according to their CCI scores, and observed that the positive association between long‐term mood stabilizer use and PD risk was only significant in those with high CCI scores (>2) (Table 4). Lithium treatment alone did not appear to influence PD risk in the subgroup of patients who were free from AED‐MSs. To determine whether the increased PD risk was related to individual mood stabilizer, we performed sub‐analyses to compare PD risk among BD patients who only received long‐term lithium treatment, those who only received long‐term AED‐MSs treatment, and those receiving both. The long‐term use of lithium alone was not associated with PD risk. The long‐term use of AED‐MSs and the use of both lithium and AED‐MSs were associated with an increased risk of PD in patients with LOBD and high CCI scores (>2) (Table S1). Second‐generation antipsychotics treatment was not related to the risk of PD development. By comparing patients with BD over 50‐year‐old of different severities of medical comorbidities, we found that a lower proportion of patients with high CCI scores received long‐term treatment of lithium, and a higher proportion of them received long‐term AED‐MSs treatment than patients with low CCI scores (Table S2). The associations between psychotropic medication use and risk of subsequent PD were not found among patients with major depression (Table S3).
Table 3.
Cox regression analyses of the risk of Parkinson's disease among patients with bipolar disorder and controls, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, smoking, medical comorbidities, and psychotropic agents
| Age <50 y HR (95% CI) | Age ≥50 y HR (95% CI) | Males HR (95% CI) | Females HR (95% CI) | Total HR (95% CI) | |
|---|---|---|---|---|---|
| Bipolar Disorder (presence vs. absence) | 10.48 (3.00–36.61) | 8.80 (5.97–12.96) | 10.55 (5.75–19.39) | 7.58 (4.75–12.10) | 9.34 (6.47–13.48) |
| Age at enrollment | 1.16 (0.91–1.46) | 1.15 (1.12–1.18) | 1.15 (1.11–1.20) | 1.14 (1.10–1.17) | 1.15 (1.12–1.17) |
| All‐cause clinical visits | 1.02 (1.02–1.02) | 1.02 (1.02–1.02) | 1.02 (1.02–1.02) | 1.02 (1.02–1.02) | 1.02 (1.02–1.02) |
| Antidepressants treatment | |||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 0.45 (0.16–1.30) | 0.85 (0.62–1.16) | 0.67 (0.41–1.10) | 1.02 (0.70–1.49) | 0.80 (0.59–1.07) |
| ≧365 cDDD | 0.63 (0.28–1.43) | 0.62 (0.46–0.83) | 0.52 (0.33–0.80) | 0.70 (0.49–1.00) | 0.61 (0.46–0.80) |
| Lithium treatment | |||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 0.74 (0.22–2.55) | 0.88 (0.58–1.35) | 0.93 (0.49–1.75) | 0.79 (0.46–1.35) | 0.88 (0.60–1.31) |
| ≧365 cDDD | 1.40 (0.61–3.22) | 1.60 (1.17–2.18) | 1.75 (1.11–2.77) | 1.48 (1.01–2.17) | 1.57 (1.17–2.10) |
| Antiepileptic mood stabilizers | |||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 2.19 (0.91–5.31) | 1.15 (0.82–1.60) | 1.09 (0.65–1.80) | 1.55 (1.05–2.30) | 1.22 (0.90–1.66) |
| ≧365 cDDD | 1.90 (0.74–4.90) | 1.53 (1.11–2.11) | 1.45 (0.91–2.30) | 1.84 (1.23–2.75) | 1.55 (1.15–2.10) |
| Antipsychotics treatment | |||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 0.43 (0.15–1.23) | 0.74 (0.53–1.03) | 0.88 (0.54–1.41) | 0.44 (0.28–0.70) | 0.65 (0.47–0.89) |
| ≧365 cDDD | 0.45 (0.18–1.13) | 0.90 (0.64–1.27) | 0.70 (0.41–1.20) | 0.86 (0.57–1.28) | 0.81 (0.59–1.12) |
| Comorbidities | |||||
| Alcohol use | 0.91 (0.33–2.46) | 1.04 (0.67–1.64) | 1.44 (0.90–2.31) | 0.46 (0.17–1.26) | 1.03 (0.69–1.56) |
| Hypertension | 0.61 (0.29–1.31) | 0.81 (0.63–1.05) | 0.60 (0.40–0.89) | 0.85 (0.63–1.16) | 0.79 (0.63–1.01) |
| Dyslipidemia | 1.56 (0.75–3.25) | 0.83 (0.64–1.07) | 0.91 (0.60–1.37) | 0.91 (0.67–1.24) | 0.88 (0.69–1.12) |
| Diabetes mellitus | 0.98 (0.43–2.24) | 1.20 (0.92–1.57) | 1.42 (0.94–2.14) | 1.14 (0.75–1.45) | 1.17 (0.91–1.52) |
| Cerebrovascular diseases | 2.08 (0.89–4.91) | 2.76 (2.14–3.57) | 3.92 (2.65–5.81) | 2.21 (1.61–3.04) | 2.69 (2.10–3.43) |
| Traumatic brain injury | 1.98 (0.80–4.91) | 1.43 (0.99–2.07) | 1.45 (0.86–2.44) | 1.56 (1.00–2.44) | 1.55 (1.11–2.17) |
| Obesity | 1.90 (0.64–5.66) | 1.01 (0.55–1.86) | 2.42 (1.05–5.60) | 0.84 (0.43–1.65) | 1.15 (0.68–1.95) |
Bold type indicates statistical significance.
cDDD, cumulative defined daily dose; HR, hazard ratio.
Table 4.
Hazard ratios of Parkinson's disease among patients with bipolar disorder receiving psychotropic agents, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, smoking, and medical comorbidities
| AED‐MSs free | Lithium free | |||
|---|---|---|---|---|
| Age ≥50 y & CCI ≤2 n = 7454 HR (95% CI) | Age ≥50 y & CCI >2 n = 6163 HR (95% CI) | Age ≥50 y & CCI >2 n = 3045 HR (95% CI) | Age ≥50 y & CCI >2 n = 4586 HR (95% CI) | |
| Antidepressants treatment | ||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 0.97 (0.57–1.66) | 0.80 (0.53–1.21) | 1.02 (0.53–1.97) | 0.66 (0.37–1.16) |
| ≧365 cDDD | 0.51 (0.30–0.88) | 0.65 (0.45–0.95) | 1.02 (0.55–1.91) | 0.69 (0.43–1.12) |
| Lithium treatment | ||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | |
| 30–364 cDDD | 1.02 (0.51–2.01) | 0.87 (0.50–1.52) | 1.62 (0.71–3.66) | |
| ≧365 cDDD | 1.21 (0.69–2.12) | 1.79 (1.22–2.62) | 1.65 (0.78–3.51) | |
| AED‐MSs | ||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | |
| 30–364 cDDD | 1.18 (0.68–2.05) | 1.13 (0.73–1.75) | 1.11 (0.64–1.94) | |
| ≧365 cDDD | 1.27 (0.71–2.27) | 1.74 (1.16–2.61) | 1.98 (1.18–3.31) | |
| Antipsychotics treatment | ||||
| <30 cDDD | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 30–364 cDDD | 0.89 (0.52–1.54) | 0.67 (0.43–1.05) | 0.77 (0.37–1.62) | 0.75 (0.42–1.31) |
| ≧365 cDDD | 0.48 (0.23–1.00) | 1.10 (0.73–1.66) | 0.70 (0.27–1.85) | 0.95 (0.52–1.71) |
Bold type indicates statistical significance.
AED‐MSs, antiepileptic mood stabilizers; CCI, Charlson Comorbidity Index; cDDD, cumulative defined daily dose; HR, hazard ratio.
Sensitivity analyses after excluding the first year or the first 3 years of observation consistently showed that patients with BD were associated with an elevated risk of subsequent PD in later life than the controls (Table S4). Sensitivity analyses after the exclusion of lithium or any mood stabilizers exposure before PD diagnosis also supported an elevated risk of subsequent PD among patients with BD (Table 2).
Discussion
The results of this longitudinal cohort study supported the hypothesis that patients with both major depression and BD exhibited an increased likelihood of developing PD compared with the controls, and patients with BD had a 51% increased risk of developing PD compared to those with major depression. There was no sex difference in the risk of subsequent PD among patients with BD. Furthermore, long‐term treatment of AED‐MSs may increase the risk of PD among patients with LOBD and high CCI scores, while lithium was not associated with PD risk.
Our findings further endorse and broaden the link between BD and PD from our prior report and a recent meta‐analysis. 2 , 3 Although the etiology of BD has not been definitively established, evidence suggests involvement of the dopaminergic system, as a transition from depressive to manic states appears linked to upregulation of dopamine receptors. 18 The dopamine dysregulation hypothesis proposes a connection between BD and PD. Levodopa, used to increase dopamine in PD, can induce (hypo)mania, while dopamine receptor‐blocking antipsychotics improve manic symptoms but may worsen motor symptoms in PD. 19 , 20 Clinical evidence supports this, with depressed patients with BD often experiencing more parkinsonism, while patients with PD may manifest mood changes associated with the on–off phenomenon. Independent investigations of BD and PD have also identified shared pathophysiological characteristics, such as gamma‐aminobutyric acid decline 21 , 22 and chronic inflammation. 23 , 24 Additionally, shared differentially expressed genes, such as those involved in ATPase activity and primary active transmembrane transporter activity, are observed in both BD and PD, suggesting a shared pathophysiological mechanism that may explain overlapping features in certain patients. 25 Consistent with a previous study by Harrison et al., we confirmed the elevated risk of PD in BD relative to major depression 6 ; our study further considered socioeconomic factors and additional medical comorbidities such as alcohol use and smoking in the analyses. This finding is noteworthy, particularly considering the well‐established association between depression and PD. 5 BD is often considered a more severe mood disorder compared to major depression. BD is associated with more deficits in neurocognitive function, more severe pro‐inflammatory profile, and more pathology in neuroimaging findings with more gray matter volumetric or white matter connectivity changes. 26 , 27 Consistent with these observations, our investigation revealed that both BD and major depression were related to an increased risk of developing PD, with patients with BD demonstrating a significantly higher risk of subsequent PD compared to individuals with major depression and controls.
The roles of psychotropic medications in the development of PD among patients with BD were examined, and we found that long‐term AED‐MSs treatment was associated with an increased risk of subsequent PD among patients with LOBD and more severe medical comorbidities. We did not find an association between lithium treatment and risk of developing PD. BD treatment includes lithium, antipsychotics, and AED‐MSs, all of which have been implicated in causing DIP. 28 Consequently, there exists a potential for misdiagnosing DIP as PD and an overestimation of PD prevalence in BD. To ensure the diagnostic validity of PD in our study, we restricted our definition of PD diagnosis to include only the ICD‐9‐CM code 332.0 (primary PD), excluding code 332.1 (secondary parkinsonism). The accuracy of PD diagnosis has been proven to be around 95% in the Taiwan NHIRD. 29 We conducted sensitivity analyses to establish primary PD diagnosis, excluding exposure to any mood stabilizers. These analyses consistently revealed a significant temporal association between BD and PD. Questions regarding the nature of psychotropic agents and neurodegenerative processes in patients with BD remain open, with hypotheses suggesting a role for antipsychotic drugs, lithium, or AED‐MSs in triggering or exacerbating degenerative processes. 30 While lithium has been proposed to lower risk of dementia in patients with BD and have theoretically neuroprotective effects over PD through the inhibition of GSK‐3β and modulation of the WNT/β‐catenin pathway, 10 conflicting evidence suggests potential neurotoxicity of lithium mediated by lowered tau protein levels and nigral/cortical iron dysregulation that is closely associated with neurodegeneration and parkinsonian motor abnormalities. 31 Long‐term lithium use has been linked to a higher incidence of dopaminergic drug use in contrast to antidepressants, indicating a prescribing cascade particularly prevalent among older individuals. 32 Moreover, certain AED‐MSs, especially valproic acid, have been hypothesized to exert a neurotoxic effect on the dopaminergic pathway. 33 The associations between psychotropic medication use and the risk of developing PD were not found among patients with major depression. It may be that the number of patients taking long‐term lithium or AED‐MSs is too small to produce conclusive results (0.4% and 2.0%, respectively). Although antidepressants use has not been shown to be associated with higher risk of PD from a population‐based study using the South Korean health insurance database, 34 we demonstrated a lower risk of subsequent PD among BD patients receiving long‐term antidepressant treatment. It is possible that such associations reflected a lower risk of PD during depressive episodes than during manic episodes. While our previous study showed that manic/mixed and depressive episodes were both associated with an elevated likelihood of developing PD, comparative studies of PD risk across different mood states in BD are needed. 2 Furthermore, patients with BD have demonstrated gut microbiota dysbiosis, which may result in decreased production of short‐chain fatty acids—key molecules involved in maintaining blood–brain barrier integrity, improving neuronal function, and reducing neuroinflammation in the brain. 35 Similar gut microbiota expression has been identified in BD and PD, and such dysbiosis is likely involved in the pathogenesis and the development of PD. 36 How long‐term psychotropics treatment drive alterations in gut microbiota populations and its potential contribution to subsequent onset of PD warrants further investigation. Our study results suggested that for patients with LOBD who have more severe comorbidities, long‐term use of AED‐MSs requires attention to the risk of PD and other movement abnormalities. If parkinsonism occurs in older or comorbid patients with BD, more evaluation beyond clinical symptoms is needed, such as dopamine transporter imaging. However, the precise mechanisms underlying nigrostriatal degeneration from a concomitant use of these drugs in this subset of BD patients are not fully elucidated. These findings underscore the complex interplay between psychotropic medications and neurodegenerative processes in BD, highlighting the need for further research to elucidate their mechanisms and optimize therapeutic strategies.
While clinical symptoms alone are insufficient for distinguishing DIP from idiopathic PD, presynaptic dopamine transporter imaging studies using single photon emission computed tomography (DaT‐SPECT) have shown high sensitivity and specificity in identifying DIP. 37 As most DIP cases are categorized as a type of postsynaptic parkinsonism resulting from D2‐receptor blockade, normal findings on SPECT are typically anticipated in these individuals. 38 However, some patients with DIP exhibit reduced striatal uptake, suggesting certain psychotropics might potentially exert direct neurotoxic damage to nigrostriatal dopamine neurons, especially in individuals with genetic variants associated with familial PD susceptibility. 39 , 40 Notably, Roberto Erro et al. demonstrated that up to 20% of BD patients with clinical parkinsonism display underlying dopaminergic deficits, as evidenced by DaT‐SPECT, highlighting BD as a significant risk factor for future neurodegenerative parkinsonism. 41 Furthermore, while early PD typically exhibits greater loss of dopaminergic activity in the putamen compared to the caudate nucleus, BD patients with dopaminergic deficits show less clinical asymmetry and higher putamen‐to‐caudate ratios, suggesting a complex interplay of dopaminergic dysfunction in BD‐associated parkinsonism, potentially involving different patterns of striatal involvement compared to PD. 41 , 42 These findings underscore the crucial role of DaT‐SPECT imaging in aiding the differential diagnosis of PD or DIP in BD. Detailed movement disorder evaluations and dopamine transporter imaging examinations are needed in future studies.
The main strengths of this study are the large cohort size and the wealth of data available about demographics and medications. By matching participants according to age, sex, income, and level of urbanization, confounding was significantly reduced. Collectively, these elements imply that the findings of the present study are relatively robust. Our identification of which subgroup of patients with BD are at PD risk is of significant clinical relevance. Longitudinal motor assessments in patients with BD are advisable to detect prodromal signs of PD. When parkinsonian symptoms arise in BD, they shouldn't simply be attributed to DIP and warrant investigations, particularly in older patients or patients with late‐onset BD, with dopaminergic imaging being a recommended diagnostic tool.
Utilization of an insurance claims database restricts access to certain potentially confounding variables, including family history, disease severity of BD, and environmental influences. Further research is warranted to establish a definitive link between different mood episodes, severity of symptoms and neuroprogression, alongside exploring the impact of genetic predisposition and environmental factors on the development of neurobiological susceptibility to PD. Additionally, the DDD extracted from the NHIRD relies on filled prescriptions, serving as an approximation of the presumed average maintenance dose during the follow‐up period. This may not accurately depict the actual dose taken by the patients.
Our large‐scale longitudinal cohort study provides evidence of an increased risk of PD among patients with BD and major depression compared to the general population, with patients with BD having the highest risk. Long‐term treatment with AED‐MSs may elevate PD risk, particularly among patients with late‐onset BD and high comorbidity scores. These results emphasize the importance of vigilant monitoring and appropriate diagnostic evaluations when parkinsonian symptoms arise in BD patients, especially in older individuals or those with severe comorbidities. Further research elucidating the complex interplay between psychotropic medications and neurodegenerative processes in BD is warranted to optimize therapeutic strategies and improve patient outcomes.
Disclosure statement
All authors have no conflict of interest.
Author contributions
M.H.: conceptualization, formal analysis, investigation, methodology, writing – original draft preparation. C.C.: formal analysis, methodology, J.H.: supervision, validation. Y.B.: funding acquisition. T.S.: supervision, validation. C.L.: funding acquisition. S.T.: funding acquisition. Y.C.: validation. M.C.: data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, writing – review and editing.
Supporting information
Table S1. Hazard ratios of Parkinson's disease among patients with bipolar disorder receiving psychotropic agents, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, smoking, and medical comorbidities.
Table S2. Comparisons of patients over 50 years old with bipolar disorder and different severities of medical comorbidities.
Table S3. Cox regression analyses of the risk of Parkinson's disease among patients with major depression and controls, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits) and medical comorbidities.
Table S4. Sensitivity test for the risk of developing Parkinson's disease among patients with major depression or bipolar disorder, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, and medical comorbidities.
Acknowledgments
The study was supported by grant from Taipei Veterans General Hospital (V113C‐039, V113C‐011, V113C‐010), Yen Tjing Ling Medical Foundation (CI‐113‐32, CI‐113‐30), Ministry of Science and Technology, Taiwan (MOST110‐2314‐B‐075‐026, MOST110‐2314‐B‐075‐024‐MY3, MOST 109‐2314‐B‐010‐050‐MY3, MOST111‐2314‐B‐075‐014‐MY2, MOST 111‐2314‐B‐075‐013, NSTC111‐2314‐B‐A49‐089‐MY2, NSTC 113‐2314‐B‐075‐042), Taipei, Taichung, Kaohsiung Veterans General Hospital, Tri‐Service General Hospital, Academia Sinica Joint Research Program (VTA112‐V1‐6‐1, VTA113‐V1‐5‐1) and Veterans General Hospitals and University System of Taiwan Joint Research Program (VGHUST112‐G1‐8‐1, VGHUST113‐G1‐8‐1). The funding source had no role in any process of our study.
References
- 1. Harrison PJ, Colbourne L, Harrison CH. The neuropathology of bipolar disorder: Systematic review and meta‐analysis. Mol. Psychiatry 2020; 25: 1787–1808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Huang MH, Cheng CM, Huang KL et al. Bipolar disorder and risk of Parkinson disease: A nationwide longitudinal study. Neurology 2019; 92: e2735–e2742. [DOI] [PubMed] [Google Scholar]
- 3. Faustino PR, Duarte GS, Chendo I et al. Risk of developing Parkinson disease in bipolar disorder: A systematic review and meta‐analysis. JAMA Neurol. 2020; 77: 192–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. van der Hoek TC, Bus BA, Matui P, van der Marck MA, Esselink RA, Tendolkar I. Prevalence of depression in Parkinson's disease: Effects of disease stage, motor subtype and gender. J. Neurol. Sci. 2011; 310: 220–224. [DOI] [PubMed] [Google Scholar]
- 5. Gustafsson H, Nordstrom A, Nordstrom P. Depression and subsequent risk of Parkinson disease: A nationwide cohort study. Neurology 2015; 84: 2422–2429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Harrison PJ, Luciano S. Incidence of Parkinson's disease, dementia, cerebrovascular disease and stroke in bipolar disorder compared to other psychiatric disorders: An electronic health records network study of 66 million people. Bipolar Disord. 2021; 23: 454–462. [DOI] [PubMed] [Google Scholar]
- 7. Noyce AJ, Bestwick JP, Silveira‐Moriyama L et al. Meta‐analysis of early nonmotor features and risk factors for Parkinson disease. Ann. Neurol. 2012; 72: 893–901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Forlenza OV, De‐Paula VJ, Diniz BS. Neuroprotective effects of lithium: Implications for the treatment of Alzheimer's disease and related neurodegenerative disorders. ACS Chem. Nerosci. 2014; 5: 443–450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gerhard T, Devanand DP, Huang C, Crystal S, Olfson M. Lithium treatment and risk for dementia in adults with bipolar disorder: Population‐based cohort study. Br. J. Psychiatry 2015; 207: 46–51. [DOI] [PubMed] [Google Scholar]
- 10. Vallee A, Vallee JN, Lecarpentier Y. Parkinson's disease: Potential actions of lithium by targeting the WNT/beta‐catenin pathway, oxidative stress, inflammation and glutamatergic pathway. Cells 2021; 10: 230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Marras C, Herrmann N, Fischer HD et al. Lithium use in older adults is associated with increased prescribing of Parkinson medications. Am. J. Geriatr. Psychiatry 2016; 24: 301–309. [DOI] [PubMed] [Google Scholar]
- 12. Balestrino R, Schapira AHV. Parkinson disease. Eur. J. Neurol. 2020; 27: 27–42. [DOI] [PubMed] [Google Scholar]
- 13. Cheng CM, Chen MH, Tsai SJ et al. Susceptibility to treatment‐resistant depression within families. JAMA Psychiatry 2024; 81: 663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J. Chronic Dis. 1987; 40: 373–383. [DOI] [PubMed] [Google Scholar]
- 15. Liu C‐Y, Hung Y‐T, Chuang Y‐L et al. Incorporating development stratification of Taiwan townships into sampling design of large scale health interview survey. ScienceOpen, Inc 2006; 4: 1–22. [Google Scholar]
- 16. Yassa R, Nair NP, Iskandar H. Late‐onset bipolar disorder. Psychiatr. Clin. North Am. 1988; 11: 117–131. [PubMed] [Google Scholar]
- 17. Tuty Kuswardhani RA, Henrina J, Pranata R, Anthonius Lim M, Lawrensia S, Suastika K. Charlson comorbidity index and a composite of poor outcomes in COVID‐19 patients: A systematic review and meta‐analysis. Diabetes Metab. Syndr. 2020; 14: 2103–2109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Berk M, Dodd S, Kauer‐Sant'anna M et al. Dopamine dysregulation syndrome: Implications for a dopamine hypothesis of bipolar disorder. Acta Psychiatr. Scand. Suppl. 2007; 434: 41–49. [DOI] [PubMed] [Google Scholar]
- 19. Murphy DL, Brodie HK, Goodwin FK, Bunney WE Jr. Regular induction of hypomania by L‐dopa in “bipolar” manic‐depressive patients. Nature 1971; 229: 135–136. [DOI] [PubMed] [Google Scholar]
- 20. Hilty DM, Brady KT, Hales RE. A review of bipolar disorder among adults. Psychiatr. Serv. 1999; 50: 201–213. [DOI] [PubMed] [Google Scholar]
- 21. Blaszczyk JW. Parkinson's disease and neurodegeneration: GABA‐collapse hypothesis. Front. Neurosci. 2016; 10: 269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Romeo B, Choucha W, Fossati P, Rotge JY. Meta‐analysis of central and peripheral gamma‐aminobutyric acid levels in patients with unipolar and bipolar depression. J. Psychiatry Neurosci. 2018; 43: 58–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Pajares M, Rojo AI, Manda G, Bosca L, Cuadrado A. Inflammation in Parkinson's disease: Mechanisms and therapeutic implications. Cells 2020; 9: 1687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Huang MH, Chan YE, Chen MH et al. A longitudinal study of the association between pro‐inflammatory cytokines and mood symptoms in bipolar disorder. Acta Psychiatr. Scand. 2023; 147: 81–91. [DOI] [PubMed] [Google Scholar]
- 25. Hossain MB, Islam MK, Adhikary A, Rahaman A, Islam MZ. Bioinformatics approach to identify significant biomarkers, drug targets shared between Parkinson's disease and bipolar disorder: A pilot study. Bioinform Biol Insig 2022; 16: 11779322221079232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Huang MH, Chan YE, Chen MH et al. Pro‐inflammatory cytokines and cognitive dysfunction among patients with bipolar disorder and major depression. Psychiatry Clin. Neurosci. 2022; 76: 450–458. [DOI] [PubMed] [Google Scholar]
- 27. Siegel‐Ramsay JE, Bertocci MA, Wu B, Phillips ML, Strakowski SM, Almeida JRC. Distinguishing between depression in bipolar disorder and unipolar depression using magnetic resonance imaging: A systematic review. Bipolar Disord. 2022; 24: 474–498. [DOI] [PubMed] [Google Scholar]
- 28. Factor SA, Burkhard PR, Caroff S et al. Recent developments in drug‐induced movement disorders: A mixed picture. Lancet Neurol. 2019; 18: 880–890. [DOI] [PubMed] [Google Scholar]
- 29. Lee YC, Lin CH, Wu RM et al. Discontinuation of statin therapy associates with Parkinson disease: A population‐based study. Neurology 2013; 81: 410–416. [DOI] [PubMed] [Google Scholar]
- 30. Erro R, Bhatia KP, Tinazzi M. Parkinsonism following neuroleptic exposure: A double‐hit hypothesis? Mov. Disord. 2015; 30: 780–785. [DOI] [PubMed] [Google Scholar]
- 31. Lei P, Ayton S, Appukuttan AT et al. Lithium suppression of tau induces brain iron accumulation and neurodegeneration. Mol. Psychiatry 2017; 22: 396–406. [DOI] [PubMed] [Google Scholar]
- 32. Rochon PA, Gurwitz JH. The prescribing cascade revisited. Lancet 2017; 389: 1778–1780. [DOI] [PubMed] [Google Scholar]
- 33. Brugger F, Bhatia KP, Besag FM. Valproate‐associated parkinsonism: A critical review of the literature. CNS Drugs 2016; 30: 527–540. [DOI] [PubMed] [Google Scholar]
- 34. Shin JY, Jeon HL, Jeong HE, Ma HI, Jang S. Association between use of antiemetics, antipsychotics, or antidepressants and the risk of Parkinson's disease. Int. J. Clin. Pharmacol. Ther. 2019; 57: 73–81. [DOI] [PubMed] [Google Scholar]
- 35. Ortega MA, Alvarez‐Mon MA, Garcia‐Montero C et al. Microbiota‐gut‐brain axis mechanisms in the complex network of bipolar disorders: Potential clinical implications and translational opportunities. Mol. Psychiatry 2023; 28: 2645–2673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Huang B, Chau SWH, Liu Y et al. Gut microbiome dysbiosis across early Parkinson's disease, REM sleep behavior disorder and their first‐degree relatives. Nat. Commun. 2023; 14: 2501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Brigo F, Matinella A, Erro R, Tinazzi M. [(1)(2)(3)I]FP‐CIT SPECT (DaTSCAN) may be a useful tool to differentiate between Parkinson's disease and vascular or drug‐induced parkinsonisms: A meta‐analysis. Eur. J. Neurol. 2014; 21: 1369. [DOI] [PubMed] [Google Scholar]
- 38. Kagi G, Bhatia KP, Tolosa E. The role of DAT‐SPECT in movement disorders. J. Neurol. Neurosurg. Psychiatry 2010; 81: 5–12. [DOI] [PubMed] [Google Scholar]
- 39. Hoenicka J, Vidal L, Morales B et al. Molecular findings in familial Parkinson disease in Spain. Arch. Neurol. 2002; 59: 966–970. [DOI] [PubMed] [Google Scholar]
- 40. Rollema H, Skolnik M, D'Engelbronner J, Igarashi K, Usuki E, Castagnoli N Jr. MPP(+)‐like neurotoxicity of a pyridinium metabolite derived from haloperidol: In vivo microdialysis and in vitro mitochondrial studies. J. Pharmacol. Exp. Ther. 1994; 268: 380–387. [PubMed] [Google Scholar]
- 41. Erro R, Landolfi A, D'Agostino G et al. Bipolar disorder and Parkinson's disease: A (123)I‐Ioflupane dopamine transporter SPECT study. Front. Neurol. 2021; 12: 652375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Sasannezhad P, Juibary AG, Sadri K et al. (99m)Tc‐TRODAT‐1 SPECT imaging in early and late onset Parkinson's disease. Asia Ocean J Nucl Med Biol 2017; 5: 114–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Hazard ratios of Parkinson's disease among patients with bipolar disorder receiving psychotropic agents, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, smoking, and medical comorbidities.
Table S2. Comparisons of patients over 50 years old with bipolar disorder and different severities of medical comorbidities.
Table S3. Cox regression analyses of the risk of Parkinson's disease among patients with major depression and controls, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits) and medical comorbidities.
Table S4. Sensitivity test for the risk of developing Parkinson's disease among patients with major depression or bipolar disorder, adjusted for demographic data (age, sex, income, level of urbanization, all cause clinical visits), Charlson comorbidity index, and medical comorbidities.