[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 34.204.52.4. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Table 1.  
Demographic and Outcome Characteristics of the Follow-up Cohort
Demographic and Outcome Characteristics of the Follow-up Cohort
Table 2.  
Estimated HRs for Clinical Outcomes in PD and PRSa
Estimated HRs for Clinical Outcomes in PD and PRSa
1.
Chaudhuri  KR, Healy  DG, Schapira  AH; National Institute for Clinical Excellence.  Non-motor symptoms of Parkinson’s disease: diagnosis and management.  Lancet Neurol. 2006;5(3):235-245. doi:10.1016/s1474-4422(06)70373-8PubMedGoogle ScholarCrossref
2.
Martinez-Martin  P, Rodriguez-Blazquez  C, Kurtis  MM, Chaudhuri  KR; NMSS Validation Group.  The impact of non-motor symptoms on health-related quality of life of patients with Parkinson’s disease.  Mov Disord. 2011;26(3):399-406. doi:10.1002/mds.23462PubMedGoogle ScholarCrossref
3.
Santos-García  D, de la Fuente-Fernández  R.  Impact of non-motor symptoms on health-related and perceived quality of life in Parkinson’s disease.  J Neurol Sci. 2013;332(1-2):136-140. doi:10.1016/j.jns.2013.07.005PubMedGoogle ScholarCrossref
4.
Do  CB, Tung  JY, Dorfman  E,  et al.  Web-based genome-wide association study identifies two novel loci and a substantial genetic component for Parkinson’s disease.  PLoS Genet. 2011;7(6):e1002141. doi:10.1371/journal.pgen.1002141PubMedGoogle ScholarCrossref
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.3043PubMedGoogle ScholarCrossref
6.
Nalls  MA, Escott-Price  V, Williams  NM,  et al; International Parkinson’s Disease Genomics Consortium (IPDGC).  Genetic risk and age in Parkinson’s disease: continuum not stratum.  Mov Disord. 2015;30(6):850-854. doi:10.1002/mds.26192PubMedGoogle ScholarCrossref
7.
Lill  CM, Hansen  J, Olsen  JH, Binder  H, Ritz  B, Bertram  L.  Impact of Parkinson’s disease risk loci on age at onset.  Mov Disord. 2015;30(6):847-850. doi:10.1002/mds.26237PubMedGoogle ScholarCrossref
8.
Davis  AA, Andruska  KM, Benitez  BA, Racette  BA, Perlmutter  JS, Cruchaga  C.  Variants in GBA, SNCA, and MAPT influence Parkinson disease risk, age at onset, and progression.  Neurobiol Aging. 2016;37:209.e1-209.e7. doi:10.1016/j.neurobiolaging.2015.09.014PubMedGoogle ScholarCrossref
9.
Pihlstrøm  L, Morset  KR, Grimstad  E, Vitelli  V, Toft  M.  A cumulative genetic risk score predicts progression in Parkinson’s disease.  Mov Disord. 2016;31(4):487-490. doi:10.1002/mds.26505PubMedGoogle ScholarCrossref
10.
Hoehn  MM, Yahr  MD.  Parkinsonism: onset, progression and mortality.  Neurology. 1967;17(5):427-442. doi:10.1212/WNL.17.5.427PubMedGoogle ScholarCrossref
11.
Costello  S, Cockburn  M, Bronstein  J, Zhang  X, Ritz  B.  Parkinson’s disease and residential exposure to maneb and paraquat from agricultural applications in the Central Valley of California.  Am J Epidemiol. 2009;169(8):919-926. doi:10.1093/aje/kwp006PubMedGoogle ScholarCrossref
12.
Gatto  NM, Rhodes  SL, Manthripragada  AD,  et al.  α-Synuclein gene may interact with environmental factors in increasing risk of Parkinson’s disease.  Neuroepidemiology. 2010;35(3):191-195. doi:10.1159/000315157PubMedGoogle ScholarCrossref
13.
Kang  GA, Bronstein  JM, Masterman  DL, Redelings  M, Crum  JA, Ritz  B.  Clinical characteristics in early Parkinson’s disease in a central California population-based study.  Mov Disord. 2005;20(9):1133-1142. doi:10.1002/mds.20513PubMedGoogle ScholarCrossref
14.
Ritz  B, Rhodes  SL, Bordelon  Y, Bronstein  J.  α-Synuclein genetic variants predict faster motor symptom progression in idiopathic Parkinson disease.  PLoS One. 2012;7(5):e36199. doi:10.1371/journal.pone.0036199PubMedGoogle ScholarCrossref
15.
Hughes  AJ, Daniel  SE, Kilford  L, Lees  AJ.  Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases.  J Neurol Neurosurg Psychiatry. 1992;55(3):181-184. doi:10.1136/jnnp.55.3.181PubMedGoogle ScholarCrossref
16.
Fahn  S, Elton  R; Members of the UPDRS Development Committee. In: Fahn S, Marsden CD, Calne DB, Goldstein M, eds. Recent Developments in Parkinson’s Disease, Vol 2. Florham Park, NJ: Macmillan Health Care Information; 1987: 153-163, 293-304.
17.
Folstein  MF, Folstein  SE, McHugh  PR.  “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician.  J Psychiatr Res. 1975;12(3):189-198.PubMedGoogle ScholarCrossref
18.
Hensel  A, Angermeyer  MC, Riedel-Heller  SG.  Measuring cognitive change in older adults: reliable change indices for the Mini-Mental State Examination.  J Neurol Neurosurg Psychiatry. 2007;78(12):1298-1303. doi:10.1136/jnnp.2006.109074PubMedGoogle ScholarCrossref
19.
Tombaugh  TN.  Test-retest reliable coefficients and 5-year change scores for the MMSE and 3MS.  Arch Clin Neuropsychol. 2005;20(4):485-503. doi:10.1016/j.acn.2004.11.004PubMedGoogle ScholarCrossref
20.
Shulman  LM, Gruber-Baldini  AL, Anderson  KE, Fishman  PS, Reich  SG, Weiner  WJ.  The clinically important difference on the Unified Parkinson’s Disease Rating Scale.  Arch Neurol. 2010;67(1):64-70.PubMedGoogle ScholarCrossref
21.
Purcell  SM, Wray  NR, Stone  JL,  et al; International Schizophrenia Consortium.  Common polygenic variation contributes to risk of schizophrenia and bipolar disorder.  Nature. 2009;460(7256):748-752. doi:10.1038/nature08185PubMedGoogle Scholar
22.
Hernán  M, Robins J. Causal inference. https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/. Published March 5, 2017. Accessed June 9, 2017.
23.
Aarsland  D, Kurz  MW.  The epidemiology of dementia associated with Parkinson’s disease.  Brain Pathol. 2010;20(3):633-639. doi:10.1111/j.1750-3639.2009.00369.xPubMedGoogle ScholarCrossref
24.
Crosiers  D, Verstraeten  A, Wauters  E,  et al.  Mutations in glucocerebrosidase are a major genetic risk factor for Parkinson’s disease and increase susceptibility to dementia in a Flanders-Belgian cohort.  Neurosci Lett. 2016;629:160-164.PubMedGoogle ScholarCrossref
25.
Mata  IF, Johnson  CO, Leverenz  JB,  et al.  Large-scale exploratory genetic analysis of cognitive impairment in Parkinson’s disease.  Neurobiol Aging. 2017;56:211.e1-211.e7.PubMedGoogle ScholarCrossref
26.
Winder-Rhodes  SE, Evans  JR, Ban  M,  et al.  Glucocerebrosidase mutations influence the natural history of Parkinson’s disease in a community-based incident cohort.  Brain. 2013;136(pt 2):392-399. doi:10.1093/brain/aws318PubMedGoogle ScholarCrossref
27.
Srivatsal  S, Cholerton  B, Leverenz  JB,  et al.  Cognitive profile of LRRK2-related Parkinson’s disease.  Mov Disord. 2015;30(5):728-733. doi:10.1002/mds.26161PubMedGoogle ScholarCrossref
28.
Thaler  A, Mirelman  A, Gurevich  T,  et al; LRRK2 Ashkenazi Jewish Consortium.  Lower cognitive performance in healthy G2019S LRRK2 mutation carriers.  Neurology. 2012;79(10):1027-1032. doi:10.1212/WNL.0b013e3182684646PubMedGoogle ScholarCrossref
29.
Zhu  XC, Cao  L, Tan  MS,  et al.  Association of Parkinson’s disease GWAS-linked loci with Alzheimer’s disease in Han Chinese.  Mol Neurobiol. 2017;54(1):308-318. doi:10.1007/s12035-015-9649-5PubMedGoogle ScholarCrossref
30.
Goris  A, Williams-Gray  CH, Clark  GR,  et al.  Tau and α-synuclein in susceptibility to, and dementia in, Parkinson’s disease.  Ann Neurol. 2007;62(2):145-153. doi:10.1002/ana.21192PubMedGoogle ScholarCrossref
31.
Morley  JF, Xie  SX, Hurtig  HI,  et al.  Genetic influences on cognitive decline in Parkinson’s disease.  Mov Disord. 2012;27(4):512-518. doi:10.1002/mds.24946PubMedGoogle ScholarCrossref
32.
Poulopoulos  M, Levy  OA, Alcalay  RN.  The neuropathology of genetic Parkinson’s disease.  Mov Disord. 2012;27(7):831-842. doi:10.1002/mds.24962PubMedGoogle ScholarCrossref
33.
Guella  I, Evans  DM, Szu-Tu  C,  et al; SNCA Cognition Study Group.  α-Synuclein genetic variability: a biomarker for dementia in Parkinson disease.  Ann Neurol. 2016;79(6):991-999. doi:10.1002/ana.24664PubMedGoogle ScholarCrossref
34.
O’Bryant  SE, Humphreys  JD, Smith  GE,  et al.  Detecting dementia with the Mini-Mental State Examination in highly educated individuals.  Arch Neurol. 2008;65(7):963-967. doi:10.1001/archneur.65.7.963PubMedGoogle ScholarCrossref
35.
Woodford  HJ, George  J.  Cognitive assessment in the elderly: a review of clinical methods.  QJM. 2007;100(8):469-484. doi:10.1093/qjmed/hcm051PubMedGoogle ScholarCrossref
36.
Iverson  GL.  Interpretation of Mini-Mental State Examination scores in community-dwelling elderly and geriatric neuropsychiatry patients.  Int J Geriatr Psychiatry. 1998;13(10):661-666. doi:10.1002/(SICI)1099-1166(1998100)13:10<661:AID-GPS838>3.0.CO;2-0PubMedGoogle ScholarCrossref
37.
Neumann  J, Bras  J, Deas  E,  et al.  Glucocerebrosidase mutations in clinical and pathologically proven Parkinson’s disease.  Brain. 2009;132(pt 7):1783-1794. doi:10.1093/brain/awp044PubMedGoogle ScholarCrossref
38.
Brockmann  K, Srulijes  K, Hauser  AK,  et al.  GBA-associated PD presents with nonmotor characteristics.  Neurology. 2011;77(3):276-280. doi:10.1212/WNL.0b013e318225ab77PubMedGoogle ScholarCrossref
39.
Goker-Alpan  O, Lopez  G, Vithayathil  J, Davis  J, Hallett  M, Sidransky  E.  The spectrum of parkinsonian manifestations associated with glucocerebrosidase mutations.  Arch Neurol. 2008;65(10):1353-1357. doi:10.1001/archneur.65.10.1353PubMedGoogle ScholarCrossref
40.
Oeda  T, Umemura  A, Mori  Y,  et al.  Impact of glucocerebrosidase mutations on motor and nonmotor complications in Parkinson’s disease.  Neurobiol Aging. 2015;36(12):3306-3313. doi:10.1016/j.neurobiolaging.2015.08.027PubMedGoogle ScholarCrossref
41.
Setó-Salvia  N, Pagonabarraga  J, Houlden  H,  et al.  Glucocerebrosidase mutations confer a greater risk of dementia during Parkinson’s disease course.  Mov Disord. 2012;27(3):393-399. doi:10.1002/mds.24045PubMedGoogle ScholarCrossref
42.
Mata  IF, Leverenz  JB, Weintraub  D,  et al.  GBA variants are associated with a distinct pattern of cognitive deficits in Parkinson’s disease.  Mov Disord. 2016;31(1):95-102. doi:10.1002/mds.26359PubMedGoogle ScholarCrossref
Original Investigation
March 2018

Association of Polygenic Risk Score With Cognitive Decline and Motor Progression in Parkinson Disease

Author Affiliations
  • 1Department of Epidemiology, UCLA (University of California, Los Angeles) Fielding School of Public Health
  • 2Genetic and Molecular Epidemiology Group, Institute of Neurogenetics, University of Lübeck, Lübeck, Germany
  • 3Department of Neurology, UCLA David Geffen School of Medicine
JAMA Neurol. 2018;75(3):360-366. doi:10.1001/jamaneurol.2017.4206
Key Points

Question  Do genome-wide association study–identified susceptibility risk loci for Parkinson disease modify cognitive and motor symptom progression?

Findings  In this population-based study of 285 patients with Parkinson disease, a higher polygenic risk score based on 23 Parkinson disease genome-wide association study–identified single-nucleotide polymorphisms was associated with faster cognitive decline and progression of motor symptoms. Results did not change after removing GBA loci from the risk score.

Meaning  Susceptibility risk alleles had a cumulative association with cognitive and motor decline in Parkinson disease; these alleles may influence not only Parkinson disease susceptibility but also disease progression in multiple domains.

Abstract

Importance  Genetic factors have a well-known influence on Parkinson disease (PD) susceptibility. The largest genome-wide association study (GWAS) identified 26 independent single-nucleotide polymorphisms (SNPs) associated with PD risk. Among patients, the course and severity of symptom progression is variable, and little is known about the potential association of genetic factors with phenotypic variance.

Objective  To assess whether GWAS-identified PD risk SNPs also have a cumulative association with the progression of cognitive and motor symptoms in patients with PD.

Design, Setting, and Participants  This longitudinal population-based cohort study of 285 patients of European ancestry with incident PD genotyped 23 GWAS SNPs. One hundred ninety-nine patients were followed up for a mean (SD) of 5.3 (2.1) years for progression (baseline: June 1, 2001, through November 31, 2007; follow-up: June 1, 2007, through August 31, 2013, with mortality surveillance through December 31, 2016); 57 patients had died or were too ill for follow-up, and 29 withdrew or could not be contacted. Movement disorder specialists repeatedly assessed PD symptom progression.

Main Outcomes Measures  The combined association of PD risk loci, after creating a weighted polygenic risk score (PRS), with cognitive decline, motor progression, and survival, relying on Cox proportional hazards regression models and inverse probability weights to account for censoring.

Results  Of the 285 patients undergoing genotyping, 160 were men (56.1%) and 125 were women (43.9%); the mean (SD) age at diagnosis was 69.1 (10.4) years. The weighted PRS was associated with significantly faster cognitive decline, measured by change in the Mini-Mental State Examination (hazard ratio [HR] per 1 SD, 1.44; 95% CI, 1.00-2.07). The PRS was also associated with faster motor decline, measured by time to Hoehn & Yahr Scale stage 3 (HR, 1.34; 95% CI, 1.00-1.79) and change in Unified Parkinson’s Disease Rating Scale part III score (HR, 1.42; 95% CI, 1.00-2.01).

Conclusions and Relevance  Susceptibility SNPs for PD combined with a cumulative PRS were associated with faster motor and cognitive decline in patients. Thus, these genetic markers may be associated with not only PD susceptibility but also disease progression in multiple domains.

Introduction

Parkinson disease (PD) is the second most common neurodegenerative disorder worldwide. Although PD is typically described in terms of motor dysfunction, nonmotor features are common.1 Cognitive impairment in particular is frequent in PD and especially detrimental to patients’ quality of life and increases in caregiver burden.1,2 Furthermore, the rates of symptom development and symptom severity in PD are highly heterogeneous.3 Identifying factors that influence progression could provide insight into etiologic pathways and ultimately prevention and therapies.

Idiopathic PD likely has a considerable genetic component.4 Genome-wide association studies (GWASs) have successfully identified 26 independent single-nucleotide polymorphisms (SNPs) associated with PD at a genome-wide significance level (α = 5 × 10−8).5 The risk alleles of several of these PD loci may also lower the age of PD onset.6-8 However, the cumulative influence of these risk loci on onset age, as assessed with a constructed polygenic risk score (PRS), is rather small.6-8

On the other hand, the collective association of these PD genetic risk loci with other PD phenotypes, including disease progression and survival, is largely unexplored. A Nordic study9 generated a PRS based on 19 GWAS SNPs in 336 patients and found that the PRS estimated a faster time from diagnosis to Hoehn & Yahr Scale (H&Y)10 stage 3 but was not associated with patient survival. To our knowledge, no study has yet assessed the association of a PD GWAS PRS with cognitive decline during disease progression.

Thus, the aim of the present study was to investigate for the first time, to our knowledge, the cumulative association of established PD genetic risk factors with cognitive impairment. Furthermore, we reexamined the association of the PRS with PD motor progression and survival.9

Methods
Participants

The Parkinson Environment and Gene (PEG) study is a population-based study in central California. More details on recruitment methods,11,12 case definition criteria,13 and the longitudinal patient cohort14 have been published previously. Patients with idiopathic PD diagnosed less than 3 years previously were recruited from June 1, 2001, through November 31, 2007. Patients were confirmed as having clinically probable or possible PD by a team of movement disorder specialists, led by one of us (J.M.B.), and were generally cognitively unimpaired at baseline.15 All procedures described were approved by the UCLA human subjects committee, and written informed consent was obtained from all participants.

Initially, 360 patients with PD were enrolled. We limited our study sample to patients of European ancestry (n = 285), all of whom are included in the survival analyses. We assessed symptom progression during 1 to 3 follow-up examinations by our movement disorder team (June 1, 2007, to August 31, 2013; mean [SD] follow-up, 5.3 [2.1] years; mean [SD] time from disease onset, 7.3 [2.8] years). We reexamined 199 patients (69.8%) during follow-up (57 had died or were too ill to participate in follow-up, and 29 withdrew or could not be contacted). A total of 192 participants provided the biosamples and data necessary for analyses of cognitive and motor symptom progression.

PD Symptom Assessment

Trained interviewers collected detailed information on demographic and risk factors. UCLA movement disorder specialists conducted in-person physical examinations to record progression for each participant. Motor symptoms were assessed with the Unified Parkinson’s Disease Rating Scale part III (UPDRS-III; range, 0-108, with higher scores indicating worse motor symtoms).16 If possible, patients were examined functionally while not receiving PD medications (off-score; 229 [80.4%] at the baseline examination and 146 [79.8%] at follow-up). For patients only examined while receiving medication (on-score), we estimated an off-score by adding the difference of the whole study population’s mean off- and on-scores at the time of examination to the patient’s on-score.14 During the examination, the neurologists also assigned patients to one of the modified H&Y stages (1-5, with higher stages indicating worse motor symptoms) based on the clinical descriptions of each stage.10 Cognitive function was assessed with the Mini-Mental State Examination (MMSE; range, 0-30, with lower scores indicating worse cognitive function)17 at the same neurologic examination, although after the off-score UPDRS-III examination, patients were asked to take their PD medications as needed.

Outcomes
Survival

Continued mortality surveillance has been performed throughout the PEG follow-up, primarily through vital statistics data, review of public obituaries, and continued active follow-up with patients and their families. Although our patients were examined last by study neurologists in 2013, mortality surveillance extended to December 31, 2016, which was 16 years after our study began (2001), with as many as 19 years of passive follow-up for mortality for patients first diagnosed in 1998.

Cognitive Decline

Cognitive decline was determined with the MMSE. Multiple studies investigating reliable change indices, which provide estimates of the probability that an individual’s change in test scores is not due to chance, have suggested that a reliable change in MMSE score during long intervals is 3 to 4 points.18,19 Because our mean (SD) follow-up was 5.3 (2.1) years (Table 1), we defined cognitive decline as a 4-point decrease from baseline MMSE score and time to event as the time from the baseline to follow-up examinations in which a 4-point decrease was first measured.

Motor Decline

Changes of 2.5 to 5.2 points on the UPDRS-III motor score represent clinically meaningful differences.20 We thus defined fast motor progression, based on the mean 5.3 years of follow-up, as a 20-point increase in UPDRS-III score (mean of 4 points per year) and time to event as the time from the baseline to follow-up examinations in which a 20-point increase was first measured.

H&Y Stage

Previous research using a PRS investigated conversion to H&Y stage 3.9 Time to conversion to H&Y stage 3 was defined as the time from the baseline to first follow-up examinations in which the patient scored at least stage 3. The 23 patients with this score at baseline (Table 1) were excluded from these analyses.

Genotyping and Calculation of the Weighted PRS

Participants provided biospecimens (blood or saliva) from which we extracted DNA. The SNPs were selected based on published GWASs (eTable 1 in the Supplement). Genotyping was performed using a high-throughput system (BioMark HD; Fluidigm Corporation). Of all 26 SNPs showing genome-wide significant evidence for association with PD risk in the latest GWAS meta-analysis,5 23 SNPs were successfully genotyped. The genotyping assays of SNPs rs8118008 and rs7681154 did not work properly, and rs13201101 failed our genotyping efficiency criterion (ie, <97%). The genotyping efficiency for the remaining 23 SNPs was greater than 99.6%. Furthermore, all SNPs were in Hardy-Weinberg equilibrium in PEG controls (significance cutoff of α = .002 based on Bonferroni correction for 23 SNPs) as determined based on Pearson χ2 test implemented in the PLINK tool set (version 1.07; http://zzz.bwh.harvard.edu/plink/).

The PRS was calculated using R software (version 3.3.1; https://www.r-project.org). This score is defined as the sum of the number of risk alleles per individual weighted by their β coefficient (ie, their effect size estimate corresponding to the logarithm of the odds ratio, as described by Nalls et al5 and Purcell et al21). We extracted the β coefficients from the combined discovery and replication stage analysis of Nalls et al5 so that the weights were ascertained independently from the data set used for the present study. We standardized the PRS by z transformation.

Statistical Analysis

To assess associations between the PRS and the survival and progression outcomes, we used Cox proportional hazards regression models to estimate hazard ratios (HRs) with 95% CI and a Wald χ2 test for significance (α = .05). Each outcome was modeled individually. In each model, we controlled for sex and age at diagnosis. For the MMSE and UPDRS-III, we chose not to analyze the rate of change as a continuous variable. Symptom progression is likely to be nonlinear; thus, we selected analyzing time to meaningful differences based on changes in UPDRS-III motor and MMSE scores (as described in the Outcomes subsection of the Methods section).

When estimating the associations for progression outcomes, we were forced to restrict our analysis to an uncensored subset of the cohort for whom we had ascertained outcome information (n = 192). To account for the possibility of selection bias from right censoring, we used inverse probability of censoring weights (IPCW) to create a pseudopopulation mimicking the total population before censoring in distribution of measured covariates.22 We then estimated the associations between the PRS and progression outcomes in the pseudopopulation. The outcomes of participants lost to or unavailable for follow-up are represented in the pseudopopulation by increasing the weight given to the outcomes of participants with similar measured covariates who were included (ie, had outcome information). We used logistic regression to calculate the probability of being censored, including covariates in the model that may be related to loss to follow-up and PD progression (age at PD diagnosis, sex, years of schooling, baseline MMSE score, baseline UPDRS-III score, and H&Y stage 3) and the PRS.

We conducted sensitivity analyses excluding the 2 GBA (OMIM 606463) SNPs (rs35749011 and rs114138760) from the PRS to assess whether results were driven by GBA. We used SAS software (version 9.4; SAS Institute) for analyses. Unless otherwise indicated, data are expressed as mean (SD).

Results

Of the 285 patients undergoing genotyping, 160 were men (56.1%) and 125 were women (43.9%); the mean age at diagnosis was 69.1 (10.4) years. Demographic and outcome frequencies and characteristics of the European ancestry–only cohort can be found in Table 1. The mean duration of PD at the baseline examination was 2.0 (1.4) years. The cohort was followed up for a mean of 5.3 (2.1) years for symptom progression (ie, 7.3 [2.8] years after onset of disease).

Of the 285 European ancestry–only participants enrolled at baseline, 166 (58.2%) died, with a mean time to death or censoring from baseline of 8.1 (3.7) years. Among the 192 participants with follow-up data, the mean change in the MMSE was −0.9 (2.5) points per 5 years. However, among the 32 of 191 patients with a decrease of at least 4 points (16.8%), the mean change was −5.1 (3.7) points per 5 years. Among the 183 patients with data available, the mean change in the UPDRS-III score was 8.2 (12.5) points per 5 years and 24.4 (18.7) points per 5 years among the 41 patients with an increase in score of 20 points or more (22.4%) (Table 1). Correlations among our outcome results in the follow-up cohort can be found in eTable 2 in the Supplement. Although motor outcomes (change in the UPDRS-III score and H&Y stage 3) were correlated (ρ = 0.43; P < .001), the cognitive outcome was only weakly correlated with other outcomes (range, ρ = 0.17 to ρ = 0.26).

Table 2 provides the results for the PRS and each outcome analyzed. Estimates for progression events were obtained using IPCW; eTable 3 in the Supplement gives the estimates from analyses that did not apply IPCW. We did not observe an association between the PRS and survival. However, we found that a higher PRS was associated with faster time to cognitive and motor events. We found that the PRS was associated with faster (4-point) decrease on the MMSE, such that a 1-SD increase in PRS corresponds to an HR of 1.44 (95% CI, 1.00-2.07). The effect size remained similar when we removed GBA from the PRS (HR, 1.42; 95% CI, 0.98-2.07). A higher PRS also tended to be associated with motor symptom progression, including a 20-point UPDRS-III increase (HR, 1.42; 95% CI, 1.00-2.01) and progression to H&Y stage 3 (HR, 1.34; 95% CI, 1.00-1.79). Again, results were similar and formally statistically significant when we removed GBA SNPs from the score for a UPDRS-III 20-point decrease (HR, 1.50; 95% CI, 1.05-2.14) and progression to H&Y stage 3 (HR, 1.36; 95% CI, 1.01-1.83; Table 2).

Discussion

We assessed the association of a PRS with disease progression in a longitudinal population-based cohort of patients with PD. We newly suggest that the PRS, based on the established PD GWAS risk loci, is associated with cognitive decline and validates previous findings of a collective association of these risk variants with motor symptom progression. Furthermore, we support the previous observation9 that PD risk variants are not cumulatively associated with survival among patients.

Cognitive impairment and dementia are well established disorders in PD.23 As many as 80% of patients with PD who are alive 10 years after diagnosis are expected to develop dementia.23 Despite the high prevalence, our understanding of pathogenic mechanisms is limited, and treatments are largely unsuccessful. To our knowledge, all patient cohorts investigating cognitive impairment in PD have used candidate gene approaches. For instance, GBA, one of the most important genetic risk factors for PD, has also been implicated in cognitive impairment in PD.24-26 Other studies have assessed the role of LRRK2,27-29MAPT,30,31 and SNCA32,33 for dementia in PD, but results have often been inconclusive or not been replicated independently. For the first time to our knowledge, we linked the cumulative burden of PD genetic risk factors with patients’ cognitive decline.

Although the MMSE is among the most widely used screening instruments for cognitive impairment, not all changes in score reflect true clinical change.18 Furthermore, multiple cut points have been suggested as thresholds for cognitive impairment and dementia, mostly 26 or less to 24 or less.34,35 Multiple independent studies have shown that changes of 3 to 4 points in MMSE score during longer intervals are necessary to conclude with 90% confidence that an individual has experienced a reliable functional change.18,19,36 Thus, we selected a 4-point decrease in MMSE score to define a meaningful cognitive decline (by definition, all patients with a 4-point decrease will have scores ≤26).

The only study to explore progression of motor symptoms with a PRS (based on 19 PD SNPs)9 followed up 336 Nordic patients in tertiary care and reported faster progression to H&Y stage 3 with increasing score (HR per 1 SD, 1.29; 95% CI, 1.06-1.56). Because PD is a gradually progressive disorder, moving from one H&Y stage to another generally takes several years, and H&Y stage 3 signifies the point when disability advances from mild to moderate and the appearance of postural instability.10 We have replicated this finding in our own community-based study (HR per 1 SD, 1.34; 95% CI, 1.00-1.79), providing additional evidence for a multiloci genetic component in PD motor progression.

We also examined changes in UPDRS-III motor scores. Shulman et al20 reported that changes of 2.5 to 5.2 points on the UPDRS-III represent clinically meaningful differences, with 10.8 points corresponding to a 1-stage change on the H&Y. We selected a 20-point increase in UPDRS-III score, corresponding to a clinically meaningful 4-point annual change in score during follow-up. We found a higher PRS to be associated with a faster time to a 20-point increase. As expected, an increase of 20 points on the UPDRS-III and conversion to H&Y stage 3 were correlated in our population (Spearman coefficient, r = 0.43; P < .001) (eTable 2 in the Supplement), indicating that both measures likely capture some of the same features of motor progression. The UPDRS-III and H&Y were assessed throughout the study by the same team of movement disorder specialists, led by one of us (J.M.B.), trained to evaluate patients in a standardized fashion. Thus, we did not expect much misclassification owing to differences in how clinicians conducted examinations. However, 89 of our participants (31.2%) could not perform at least 1 neurologic examination while in a functional off-medication state. We estimated UPDRS-III off-scores based on the study population means. However, this procedure is still likely to cause some misclassification of motor severity for patients with on-scores. We expect this misclassification to be nondifferential with respect to the PRS because the PRS does not predict medication status at any of the examinations. In addition, not enough patients had only on-score examinations to determine whether our off-score findings are consistent across medication status. This determination would be interesting when considering whether attributed genetic effects may also be related to levodopa responsiveness. Furthermore, although the MMSE was done after the UPDRS-III motor examinations, patients were allowed to take PD medications as needed after motor examinations.

The influence of GBA mutations on PD risk is well established. The weight of evidence supports that carriers of GBA mutations are not only at a higher risk of PD37-40 but also have an earlier age at onset and are more likely to develop cognitive dysfunction.26,41,42 Thus, we conducted sensitivity analyses after excluding variants in the GBA locus to determine whether our results for the PRS were driven by GBA. Estimated effect sizes did not change meaningfully (Table 2), suggesting that the remaining multiple genetic variants together contribute to faster progression cumulatively.

Strengths and Limitations

Our study is one of less than a handful of population-based prospective PD cohorts worldwide and the only study, to our knowledge, to investigate PD risk loci with a PRS and cognitive decline. Although we were unable to follow up all patients with PD enrolled at baseline because loss to follow-up occurred primarily owing to death and illness, we conducted IPCW statistical analyses to account for selective survival. This form of internal adjustment corrects for censoring, which is modeled as a function of measured baseline risk factors and PD symptom levels that affect censoring and the PD progression end points under study. However, this method can only account for factors included in the censoring model. Other unmeasured factors likely influence loss to follow-up, and thus some remaining selection bias is possible.

A notable strength in our study is the well characterized PD population. All our patients were personally examined by the same UCLA movement disorder specialists throughout the study, which minimizes outcome misclassification. Furthermore, follow-up began early in the disease course (within 3 years of diagnosis), allowing us to track the natural history of progression. Because of our community-based design, our results are more generalizable to average PD populations than patient cohorts assembled at tertiary care centers.

Conclusions

Ultimately, identifying genetic markers associated with faster progression may help elucidate pathogenesis and inform further research. Although replication is needed, our results support the collective involvement of PD susceptibility risk alleles in cognitive and motor decline, suggesting that these alleles may be associated with not only PD susceptibility but also disease progression in multiple domains. Our findings support a polygenic architecture contributing to PD progression, as has been suggested for PD susceptibility, and suggest that progression in PD may, at least in part, be driven by an accumulation of many common genetic variants, each individually having a relatively small effect size.

Back to top
Article Information

Accepted for Publication: October 18, 2017.

Corresponding Author: Beate R. Ritz, MD, PhD, Department of Epidemiology, UCLA Fielding School of Public Health, Box 951772, Los Angeles, CA 90095 (britz@ucla.edu).

Published Online: January 16, 2018. doi:10.1001/jamaneurol.2017.4206

Author Contributions: Dr Paul and Ms Schulz were co–first authors, and Drs Lill and Ritz were co–last authors. Dr Ritz had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Paul, Bronstein, Lill, Ritz.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Paul.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Paul, Schulz, Lill.

Obtained funding: Bronstein, Lill, Ritz.

Administrative, technical, or material support: Bronstein, Lill, Ritz.

Study supervision: Bronstein, Lill, Ritz.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by grants 2R01-ES010544, U54ES012078, and F32-ES028087 from the National Institute of Environmental Health Science; grant 5P30ES007048 from the Southern California Environmental Health Sciences Center (Drs Paul, Bronstein, and Ritz); a population- and laboratory-based sciences award from the Burroughs Wellcome Fund (Dr Paul); funds from the Veterans Administration Healthcare System (Southwest Parkinson’s Disease Research, Education and Clinical Centers), the Levine Foundation, and Parkinson Alliance (Dr Bronstein); an MD thesis research scholarship Exzellenzmedizin from the University of Lübeck (Ms Schulz); grants FOR2488/1 and GZ LI 2654/2-1 from the German Research Foundation; and grant J21-2016 from the University of Lübeck Section of Medicine (Dr Lill).

Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Chaudhuri  KR, Healy  DG, Schapira  AH; National Institute for Clinical Excellence.  Non-motor symptoms of Parkinson’s disease: diagnosis and management.  Lancet Neurol. 2006;5(3):235-245. doi:10.1016/s1474-4422(06)70373-8PubMedGoogle ScholarCrossref
2.
Martinez-Martin  P, Rodriguez-Blazquez  C, Kurtis  MM, Chaudhuri  KR; NMSS Validation Group.  The impact of non-motor symptoms on health-related quality of life of patients with Parkinson’s disease.  Mov Disord. 2011;26(3):399-406. doi:10.1002/mds.23462PubMedGoogle ScholarCrossref
3.
Santos-García  D, de la Fuente-Fernández  R.  Impact of non-motor symptoms on health-related and perceived quality of life in Parkinson’s disease.  J Neurol Sci. 2013;332(1-2):136-140. doi:10.1016/j.jns.2013.07.005PubMedGoogle ScholarCrossref
4.
Do  CB, Tung  JY, Dorfman  E,  et al.  Web-based genome-wide association study identifies two novel loci and a substantial genetic component for Parkinson’s disease.  PLoS Genet. 2011;7(6):e1002141. doi:10.1371/journal.pgen.1002141PubMedGoogle ScholarCrossref
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.3043PubMedGoogle ScholarCrossref
6.
Nalls  MA, Escott-Price  V, Williams  NM,  et al; International Parkinson’s Disease Genomics Consortium (IPDGC).  Genetic risk and age in Parkinson’s disease: continuum not stratum.  Mov Disord. 2015;30(6):850-854. doi:10.1002/mds.26192PubMedGoogle ScholarCrossref
7.
Lill  CM, Hansen  J, Olsen  JH, Binder  H, Ritz  B, Bertram  L.  Impact of Parkinson’s disease risk loci on age at onset.  Mov Disord. 2015;30(6):847-850. doi:10.1002/mds.26237PubMedGoogle ScholarCrossref
8.
Davis  AA, Andruska  KM, Benitez  BA, Racette  BA, Perlmutter  JS, Cruchaga  C.  Variants in GBA, SNCA, and MAPT influence Parkinson disease risk, age at onset, and progression.  Neurobiol Aging. 2016;37:209.e1-209.e7. doi:10.1016/j.neurobiolaging.2015.09.014PubMedGoogle ScholarCrossref
9.
Pihlstrøm  L, Morset  KR, Grimstad  E, Vitelli  V, Toft  M.  A cumulative genetic risk score predicts progression in Parkinson’s disease.  Mov Disord. 2016;31(4):487-490. doi:10.1002/mds.26505PubMedGoogle ScholarCrossref
10.
Hoehn  MM, Yahr  MD.  Parkinsonism: onset, progression and mortality.  Neurology. 1967;17(5):427-442. doi:10.1212/WNL.17.5.427PubMedGoogle ScholarCrossref
11.
Costello  S, Cockburn  M, Bronstein  J, Zhang  X, Ritz  B.  Parkinson’s disease and residential exposure to maneb and paraquat from agricultural applications in the Central Valley of California.  Am J Epidemiol. 2009;169(8):919-926. doi:10.1093/aje/kwp006PubMedGoogle ScholarCrossref
12.
Gatto  NM, Rhodes  SL, Manthripragada  AD,  et al.  α-Synuclein gene may interact with environmental factors in increasing risk of Parkinson’s disease.  Neuroepidemiology. 2010;35(3):191-195. doi:10.1159/000315157PubMedGoogle ScholarCrossref
13.
Kang  GA, Bronstein  JM, Masterman  DL, Redelings  M, Crum  JA, Ritz  B.  Clinical characteristics in early Parkinson’s disease in a central California population-based study.  Mov Disord. 2005;20(9):1133-1142. doi:10.1002/mds.20513PubMedGoogle ScholarCrossref
14.
Ritz  B, Rhodes  SL, Bordelon  Y, Bronstein  J.  α-Synuclein genetic variants predict faster motor symptom progression in idiopathic Parkinson disease.  PLoS One. 2012;7(5):e36199. doi:10.1371/journal.pone.0036199PubMedGoogle ScholarCrossref
15.
Hughes  AJ, Daniel  SE, Kilford  L, Lees  AJ.  Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases.  J Neurol Neurosurg Psychiatry. 1992;55(3):181-184. doi:10.1136/jnnp.55.3.181PubMedGoogle ScholarCrossref
16.
Fahn  S, Elton  R; Members of the UPDRS Development Committee. In: Fahn S, Marsden CD, Calne DB, Goldstein M, eds. Recent Developments in Parkinson’s Disease, Vol 2. Florham Park, NJ: Macmillan Health Care Information; 1987: 153-163, 293-304.
17.
Folstein  MF, Folstein  SE, McHugh  PR.  “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician.  J Psychiatr Res. 1975;12(3):189-198.PubMedGoogle ScholarCrossref
18.
Hensel  A, Angermeyer  MC, Riedel-Heller  SG.  Measuring cognitive change in older adults: reliable change indices for the Mini-Mental State Examination.  J Neurol Neurosurg Psychiatry. 2007;78(12):1298-1303. doi:10.1136/jnnp.2006.109074PubMedGoogle ScholarCrossref
19.
Tombaugh  TN.  Test-retest reliable coefficients and 5-year change scores for the MMSE and 3MS.  Arch Clin Neuropsychol. 2005;20(4):485-503. doi:10.1016/j.acn.2004.11.004PubMedGoogle ScholarCrossref
20.
Shulman  LM, Gruber-Baldini  AL, Anderson  KE, Fishman  PS, Reich  SG, Weiner  WJ.  The clinically important difference on the Unified Parkinson’s Disease Rating Scale.  Arch Neurol. 2010;67(1):64-70.PubMedGoogle ScholarCrossref
21.
Purcell  SM, Wray  NR, Stone  JL,  et al; International Schizophrenia Consortium.  Common polygenic variation contributes to risk of schizophrenia and bipolar disorder.  Nature. 2009;460(7256):748-752. doi:10.1038/nature08185PubMedGoogle Scholar
22.
Hernán  M, Robins J. Causal inference. https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/. Published March 5, 2017. Accessed June 9, 2017.
23.
Aarsland  D, Kurz  MW.  The epidemiology of dementia associated with Parkinson’s disease.  Brain Pathol. 2010;20(3):633-639. doi:10.1111/j.1750-3639.2009.00369.xPubMedGoogle ScholarCrossref
24.
Crosiers  D, Verstraeten  A, Wauters  E,  et al.  Mutations in glucocerebrosidase are a major genetic risk factor for Parkinson’s disease and increase susceptibility to dementia in a Flanders-Belgian cohort.  Neurosci Lett. 2016;629:160-164.PubMedGoogle ScholarCrossref
25.
Mata  IF, Johnson  CO, Leverenz  JB,  et al.  Large-scale exploratory genetic analysis of cognitive impairment in Parkinson’s disease.  Neurobiol Aging. 2017;56:211.e1-211.e7.PubMedGoogle ScholarCrossref
26.
Winder-Rhodes  SE, Evans  JR, Ban  M,  et al.  Glucocerebrosidase mutations influence the natural history of Parkinson’s disease in a community-based incident cohort.  Brain. 2013;136(pt 2):392-399. doi:10.1093/brain/aws318PubMedGoogle ScholarCrossref
27.
Srivatsal  S, Cholerton  B, Leverenz  JB,  et al.  Cognitive profile of LRRK2-related Parkinson’s disease.  Mov Disord. 2015;30(5):728-733. doi:10.1002/mds.26161PubMedGoogle ScholarCrossref
28.
Thaler  A, Mirelman  A, Gurevich  T,  et al; LRRK2 Ashkenazi Jewish Consortium.  Lower cognitive performance in healthy G2019S LRRK2 mutation carriers.  Neurology. 2012;79(10):1027-1032. doi:10.1212/WNL.0b013e3182684646PubMedGoogle ScholarCrossref
29.
Zhu  XC, Cao  L, Tan  MS,  et al.  Association of Parkinson’s disease GWAS-linked loci with Alzheimer’s disease in Han Chinese.  Mol Neurobiol. 2017;54(1):308-318. doi:10.1007/s12035-015-9649-5PubMedGoogle ScholarCrossref
30.
Goris  A, Williams-Gray  CH, Clark  GR,  et al.  Tau and α-synuclein in susceptibility to, and dementia in, Parkinson’s disease.  Ann Neurol. 2007;62(2):145-153. doi:10.1002/ana.21192PubMedGoogle ScholarCrossref
31.
Morley  JF, Xie  SX, Hurtig  HI,  et al.  Genetic influences on cognitive decline in Parkinson’s disease.  Mov Disord. 2012;27(4):512-518. doi:10.1002/mds.24946PubMedGoogle ScholarCrossref
32.
Poulopoulos  M, Levy  OA, Alcalay  RN.  The neuropathology of genetic Parkinson’s disease.  Mov Disord. 2012;27(7):831-842. doi:10.1002/mds.24962PubMedGoogle ScholarCrossref
33.
Guella  I, Evans  DM, Szu-Tu  C,  et al; SNCA Cognition Study Group.  α-Synuclein genetic variability: a biomarker for dementia in Parkinson disease.  Ann Neurol. 2016;79(6):991-999. doi:10.1002/ana.24664PubMedGoogle ScholarCrossref
34.
O’Bryant  SE, Humphreys  JD, Smith  GE,  et al.  Detecting dementia with the Mini-Mental State Examination in highly educated individuals.  Arch Neurol. 2008;65(7):963-967. doi:10.1001/archneur.65.7.963PubMedGoogle ScholarCrossref
35.
Woodford  HJ, George  J.  Cognitive assessment in the elderly: a review of clinical methods.  QJM. 2007;100(8):469-484. doi:10.1093/qjmed/hcm051PubMedGoogle ScholarCrossref
36.
Iverson  GL.  Interpretation of Mini-Mental State Examination scores in community-dwelling elderly and geriatric neuropsychiatry patients.  Int J Geriatr Psychiatry. 1998;13(10):661-666. doi:10.1002/(SICI)1099-1166(1998100)13:10<661:AID-GPS838>3.0.CO;2-0PubMedGoogle ScholarCrossref
37.
Neumann  J, Bras  J, Deas  E,  et al.  Glucocerebrosidase mutations in clinical and pathologically proven Parkinson’s disease.  Brain. 2009;132(pt 7):1783-1794. doi:10.1093/brain/awp044PubMedGoogle ScholarCrossref
38.
Brockmann  K, Srulijes  K, Hauser  AK,  et al.  GBA-associated PD presents with nonmotor characteristics.  Neurology. 2011;77(3):276-280. doi:10.1212/WNL.0b013e318225ab77PubMedGoogle ScholarCrossref
39.
Goker-Alpan  O, Lopez  G, Vithayathil  J, Davis  J, Hallett  M, Sidransky  E.  The spectrum of parkinsonian manifestations associated with glucocerebrosidase mutations.  Arch Neurol. 2008;65(10):1353-1357. doi:10.1001/archneur.65.10.1353PubMedGoogle ScholarCrossref
40.
Oeda  T, Umemura  A, Mori  Y,  et al.  Impact of glucocerebrosidase mutations on motor and nonmotor complications in Parkinson’s disease.  Neurobiol Aging. 2015;36(12):3306-3313. doi:10.1016/j.neurobiolaging.2015.08.027PubMedGoogle ScholarCrossref
41.
Setó-Salvia  N, Pagonabarraga  J, Houlden  H,  et al.  Glucocerebrosidase mutations confer a greater risk of dementia during Parkinson’s disease course.  Mov Disord. 2012;27(3):393-399. doi:10.1002/mds.24045PubMedGoogle ScholarCrossref
42.
Mata  IF, Leverenz  JB, Weintraub  D,  et al.  GBA variants are associated with a distinct pattern of cognitive deficits in Parkinson’s disease.  Mov Disord. 2016;31(1):95-102. doi:10.1002/mds.26359PubMedGoogle ScholarCrossref
×