Abstract
This genetic association study assesses genetic etiologies and associations between Parkinson disease and bipolar disorder.
A recent systematic review and meta-analysis1 suggests that a previous diagnosis of bipolar disorder (BD) increases the likelihood of a subsequent diagnosis of idiopathic Parkinson disease (PD). Using data from the International Parkinson Disease Genomics Consortium (IPDGC) and the UK Biobank (UKBB), we assessed the possibility of a shared genetic etiology or association between PD and BD.
Methods
To explore overlapping genetics, linkage disequilibrium score regression (LDSC)2 was performed using recent genome-wide association study (GWAS) meta-analyses in PD3 and BD.4 We applied the genetic risk score (GRS) as previously described5 using 86 of the 90 independent risk loci associated with PD (imputation quality R2, >0.3)3 to assess BD status in 1106 patients with BD and 9980 controls (who were of European ancestry without BD, PD, or a first-degree relative affected by PD) with available imputed data from the UKBB. We applied the GRS using 27 of 30 risk loci associated with BD4 to determine the association of PD status in 19 249 patients with PD and 22 875 controls from the clinically ascertained subset of IPDGC. We applied BD GRS to explore the association with PD in the UKBB data. All data included from published IPDGC GWAS studies complied with the relevant institutional review boards. Software used was plink 1.9 and threshold for significance = .05. We also explored the direction of the association between PD and BD in a mendelian randomization (MR) analysis in an attempt to draw potential causal inferences.
Results
In the LDSC analysis, no evidence for a genetic correlation between both diseases was found (rg = 0.068; SE = 0.04;P = .10). This null association persisted when restricting the analysis to only clinically confirmed PD cases. Although bias might be a concern, LDSC intercepts were close to 1 in PD and BD summary statistics (PD: h2 = 0.221; SE = 0.021; intercept = 0.97; BD: h2 = 0.346; SE = 0.016; intercept = 1.01).
The GRS for PD was not associated with BD status (β = 0.036; SE = 0.04; P = .36). No differences were found when stratifying BD cases with or without a diagnosis of PD (Figure A). The GRS for PD was significantly associated with PD status in the UKBB data (β = 0.041; SE = 0.016; P = .01). The GRS for BD was not associated with PD status in the IPDGC cohort (β = 0.015; SE = 0.011; P =.15) (Figure B).
Figure. Genetic Risk Scores (GRSs) for Parkinson Disease (PD) and Bipolar Disorder (BD).

A, Parkinson disease GRS violin plots divided by BD cases and controls in the UK Biobank (UKB). B, Bipolar disorder GRS violin plots divided by PD cases and controls from the International Parkinson Disease Genomics Consortium (IPDGC). C, Bipolar disorder GRS violin plots divided by PD cases, bipolar disorder cases with a PD diagnosis (BD-PD), and controls from the UKB.
When we stratified PD cases with and without a BD diagnosis, BD GRS was not associated with PD status in the PD group without a diagnosis of BD (β = 0.012; SE = 0.026; P = .63; sampling 1696 cases and 9980 controls) (Figure B). In 30 samples with PD and BD, there was a higher mean GRS than controls (t test, P = .002) that was perhaps driven by high variance in the very small sample size of this subset. The GRS for BD was significantly associated with BD status in the UKBB data (β = 0.12; SE = 0.03; P = .001). In MR analysis, we were not able to infer a potential a causal association between PD and BD in either direction (Table) despite having sufficient statistical power to prove the null hypothesis (F statistic = 89.73) (Table).
Table. Association, Reverse Association, and Sensitivity Analyses for Parkinson Disease and Bipolar Disorder Using 2-Sample Mendelian Randomization.
| Association type | Exposure | Outcome | No. of SNVs | IVW | MR Egger | Sensitivity analyses | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β (SE) | P value | β (SE) | P value | Egger intercept (SE) | P value | IVW Q | Q df | P value | ||||
| Association | Bipolar disorder | Parkinson disease | 24 | 0.07 (0.11) | .55 | −0.34 (0.59) | .56 | 0.01 (0.01) | .48 | 59.11 | 23 | 5.13E-05 |
| Reverse association | Parkinson disease | Bipolar disorder | 83 | −0.008 (0.02) | .75 | 0.05 (0.05) | .35 | −0.006 (0.005) | .23 | 166.01 | 82 | 1.20E-07 |
Abbreviations: IVW, inverse variance weighted; MR, mendelian randomization; SNVs, single-nucleotide variants.
Discussion
Parkinson disease affects approximately 1% of the population older than 60 years.6 The mean (SD) age at recruitment for the BD UKBB cohort (code F31) was 61.65 (5.94) years and the prevalence for PD was 2.7% (30/1106). This indicates a higher prevalence of PD in concordance with the published meta-analysis.1 Based on what to our knowledge are the largest PD and BD genetic data sets to date, this analysis of common variability suggests no genetic association between both diseases. However, we believe that the original study was well powered and supported by higher estimates of PD among BD in the UKBB cohort. Limitations of this study include that important pleiotropic associations may have been overlooked when using LDSC, and that additional external factors, rare or structural genetic variation, and indirect unmeasured confounding may have contributed to this co-occurrence. Notably, a portion of the diagnosed PD cases in the original BD cohort may represent misattribution of drug-induced parkinsonism, driven by the use of commonly applied medications in BD (a potential limitation that Faustino and colleagues1 acknowledge).
References
- 1.Faustino PR, Duarte GS, Chendo I, et al. Risk of developing Parkinson disease in bipolar disorder: a systematic review and meta-analysis. JAMA Neurol. Published online October 14, 2019. doi: 10.1001/jamaneurol.2019.3446 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bulik-Sullivan BK, Loh P-R, Finucane HK, et al. ; Schizophrenia Working Group of the Psychiatric Genomics Consortium . LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291-295. doi: 10.1038/ng.3211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Nalls MA, Blauwendraat C, Vallerga CL, et al. Expanding Parkinson’s disease genetics: novel risk loci, genomic context, causal insights and heritable risk. bioRxi. Preprint. Posted online March 4, 2019. doi: 10.1101/388165 [DOI] [Google Scholar]
- 4.Stahl EA, Breen G, Forstner AJ, et al. ; eQTLGen Consortium; BIOS Consortium; Bipolar Disorder Working Group of the Psychiatric Genomics Consortium . Genome-wide association study identifies 30 loci associated with bipolar disorder. Nat Genet. 2019;51(5):793-803. doi: 10.1038/s41588-019-0397-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nalls MA, Pankratz N, Lill CM, et al. ; International Parkinson’s Disease Genomics Consortium (IPDGC); Parkinson’s Study Group (PSG) Parkinson’s Research: The Organized GENetics Initiative (PROGENI); 23andMe; GenePD; NeuroGenetics Research Consortium (NGRC); Hussman Institute of Human Genomics (HIHG); Ashkenazi Jewish Dataset Investigator; Cohorts for Health and Aging Research in Genetic Epidemiology (CHARGE); North American Brain Expression Consortium (NABEC); United Kingdom Brain Expression Consortium (UKBEC); Greek Parkinson’s Disease Consortium; Alzheimer Genetic Analysis Group . Large-scale meta-analysis of genome-wide association data identifies six new risk loci for Parkinson’s disease. Nat Genet. 2014;46(9):989-993. doi: 10.1038/ng.3043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.de Lau LML, Breteler MMB. Epidemiology of Parkinson’s disease. Lancet Neurol. 2006;5(6):525-535. doi: 10.1016/S1474-4422(06)70471-9 [DOI] [PubMed] [Google Scholar]
