This case-control study uses Medicare-linked data from the National Health and Aging Trends Study to examine whether limitations in daily functioning existed in the 3 years before a diagnosis of Parkinson disease in a sample of Medicare beneficiaries.
Key Points
Question
Are aspects of daily functioning more likely to be impaired in individuals with prodromal Parkinson disease compared with the general population?
Findings
In this case-control study of 6674 Medicare beneficiaries aged 65 years or older, individuals with Parkinson disease were more likely to have impairment in activities involving strength and mobility up to 3 years prior to diagnosis compared with participants without Parkinson disease.
Meaning
The findings suggest that individuals who have prodromal or unrecognized Parkinson disease may have motor symptoms up to 3 years before diagnosis, which may warrant earlier intervention.
Abstract
Importance
Daily functioning in individuals with prodromal Parkinson disease requires more detailed description.
Objective
To evaluate whether functional limitations exist in individuals with Parkinson disease prior to diagnosis compared with the general population.
Design, Setting, and Participants
This case-control study used Medicare-linked data from the National Health and Aging Trends Study (NHATS), a longitudinal survey in the US, for a random subsample of Medicare beneficiaries aged 65 years or older, with Black and older individuals oversampled by design. Patients with incident Parkinson disease were defined as having 2 or more Medicare diagnoses from January 2011 to December 2017, with Medicare eligibility 2 or more consecutive years prior to the first diagnosis. Controls were defined as individuals with Medicare eligibility at a baseline year and 2 or more years prior, with no Parkinson disease diagnosis. Analyses were conducted from November 2021 to June 2022.
Exposures
Responses to survey questions addressing dexterity, eating, mobility, mood, pain, sleep, speech, strength, and vision.
Main Outcome and Measures
Associations between survey responses and Parkinson disease diagnosis in the first year of diagnosis (baseline) and up to 3 years prior to diagnosis (ie, during the prodromal phase) were examined using logistic regression.
Results
A total of 6674 participants were included. The participant numbers and case prevalence each year varied from 3492 to 5049 and from 700 to 1180 per 100 000 population, respectively. The median age groups were 75 to 79 years and 80 to 84 years, and the percentage of females varied from 48.21% (27 of 56 cases) to 59.98% (2079 of 3466 controls) across all years, with similar proportions among cases and controls. Individuals with prodromal Parkinson disease were less likely to report being able to walk 6 blocks (odds ratio [OR], 0.34; 95% CI, 0.15-0.82), stand independently from a kneeling position (OR, 0.30; 95% CI, 0.11-0.85), or lift a heavy object above one’s head (OR, 0.36; 95% CI, 0.15-0.87) and were more likely to report imbalance (OR, 2.77; 95% CI, 1.24-6.20) 3 years prior to diagnosis.
Conclusions and Relevance
The findings suggest that individuals with prodromal or unrecognized Parkinson disease may have greater impairment in activities involving mobility and strength up to 3 years prior to diagnosis compared with the general population. Identification of prodromal disease may facilitate earlier intervention to improve function.
Introduction
Parkinson disease (PD) is a neurodegenerative condition defined by motor signs collectively called parkinsonism.1 Both mild parkinsonian signs (MPSs) that do not meet PD criteria and nonmotor signs, including rapid eye movement sleep behavior disorder, precede PD.2 Prodromal PD (PPD) criteria incorporate these signs and demographic and biologic markers.3 However, while many studies have evaluated risk markers, few have characterized functional status in PPD. While MPSs are associated with reduced quality of life4 and functional limitations have been described before PD diagnosis,5 further descriptions of how PPD impacts daily living may help justify interventions at the prodromal stage. Studies of PPD are often limited to higher-risk populations (eg, persons with rapid eye movement sleep behavior disorder or gene variations), which maximizes efficiency but limits generalizability and introduces recall bias of patient-reported data.3 Alternatively, results from prospective national surveys are generalizable and likely reflect unbiased recognition of function. In this exploratory case-control study, we assessed responses regarding aspects of daily living associated with PD from a longitudinal survey of Medicare beneficiaries up to 3 years before PD diagnosis.
Methods
This study was deemed exempt from approval by the University of Pennsylvania institutional review board since data were deidentified, and informed consent was not required. Analysis and reporting complied with National Health and Aging Trends Study (NHATS) and Centers for Medicare & Medicaid Services data use agreement policies. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was followed.
Data Source
We used Medicare-linked data from the NHATS, an annual survey of a representative sample of Medicare beneficiaries aged 65 years or older in the US that is funded by the National Institute on Aging.6 The sample was first interviewed in 2011 and replenished in 2015, with response rates of 71% (8245 of 11 629) and 76% (8334 of 10 937), respectively. Individuals aged 85 years or older and Black participants were oversampled by design.
Identification of Cases and Controls
We identified cases with newly diagnosed PD with 2 or more Medicare diagnoses from January 2011 to December 2017 (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] 332.0 or International Statistical Classification of Diseases, Tenth Revision, Clinical Modification [ICD-10-CM] G20). We required (1) complete Medicare Part A and B eligibility and (2) no codes for atypical parkinsonian disorders or drug-induced or vascular parkinsonism in the 2 or more consecutive years before the first qualifying diagnosis. This definition has high sensitivity and positive predictive value (89.2% and 79.4%, respectively) to identify PD in claims data.7 Controls had Medicare eligibility at baseline and 2 years prior, with no PD diagnosis.
Variable Selection
NHATS contains screening questions regarding daily activities, quality of life, and symptoms that are completed by all participants regardless of current clinical characteristics or medical history. We examined questions about dexterity, eating, mobility, mood, pain, sleep, speech, strength, and vision from the standard annual questionnaire, presented regardless of other responses, to ensure the same participants were evaluated across variables in a given year. Demographic data including age group (65-69, 70-74, 75-79, 80-84, 85-89, and ≥90 years) and gender were collected. Whether participants required a proxy was identified.
Statistical Analysis
Analyses were performed at baseline (year 0) and each year up to 3 prior (years −1, −2, and −3). For cases, year 0 was the year of the first PD diagnosis. For controls, year 0 was 2014 or 2017 if enrolled in 2011 or 2015, respectively (2011-2017 responses were available during analysis).
Multiple imputation accounted for missing responses (eMethods in the Supplement).8 Associations between case-control status and age group, gender, proxy status, and presence or absence of imputed data were evaluated using the Pearson χ2 test. Logistic regression was performed with case-control status as the dependent variable and item response as the independent variable, with age group, gender, and proxy status included in multivariable models. A response of no was the reference for yes-or-no questions, and ordinal variables were treated as continuous. Two-sided P ≤ .05 was considered significant (correction for multiple comparisons was not performed). Stata, version 17.0 (StataCorp LLC), was used. Analyses were performed from November 2021 to June 2022.
Results
Table 1 shows demographic and descriptive data for the 6674 total participants at years −3 through 0. Sample size and PPD or PD prevalence varied from 3492 to 5049 participants and from 700 to 1180 per 100 000 population, respectively. Median age groups were 75 to 79 years and 80 to 84 years, and the percentage of females varied from 48.21% (27 of 56 cases) to 59.98% (2079 of 3466 controls) across all years, with similar proportions among cases and controls. Cases more frequently had a proxy and at least 1 imputed response at baseline. A minority of individuals required imputation at all years.
Table 1. Descriptive Data.
| Characteristic | Participants, No. (%) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Year 0 | Year −1 | Year −2 | Year −3 | |||||||||
| Cases (n = 56) | Controls (n = 4764) | P valuea | Cases (n = 41) | Controls (n = 4835) | P valuea | Cases (n = 35) | Controls (n = 5014) | P valuea | Cases (n = 26) | Controls (n = 3466) | P valuea | |
| Age group, y | ||||||||||||
| 65-69 | 5 (8.93) | 353 (7.41) | .27 | 4 (9.76) | 620 (12.82) | .47 | 2 (5.71) | 911 (18.17) | .40 | 3 (11.54) | 667 (19.24) | .49 |
| 70-74 | 7 (12.50) | 1038 (21.79) | 9 (21.95) | 1030 (21.30) | 8 (22.86) | 1046 (20.86) | 4 (15.38) | 752 (21.70) | ||||
| 75-79 | 14 (25.00) | 944 (19.82) | 7 (17.07) | 982 (20.31) | 8 (22.86) | 1066 (21.26) | 5 (19.23) | 737 (21.26) | ||||
| 80-84 | 14 (25.00) | 954 (20.03) | 9 (21.95) | 954 (19.73) | 8 (22.86) | 980 (19.55) | 9 (34.62) | 712 (20.54) | ||||
| 85-89 | 12 (21.43) | 800 (16.79) | 10 (24.39) | 715 (14.79) | 7 (20.00) | 620 (12.37) | 4 (15.38) | 386 (11.14) | ||||
| ≥90 | 4 (7.14) | 675 (14.17) | 2 (4.88) | 534 (11.04) | 2 (5.71) | 391 (7.80) | 1 (3.85) | 212 (6.12) | ||||
| Gender | ||||||||||||
| Female | 27 (48.21) | 2785 (58.46) | .12 | 21 (51.22) | 2835 (58.63) | .34 | 17 (48.57) | 2925 (58.34) | .24 | 14 (53.85) | 2079 (59.98) | .53 |
| Male | 29 (51.79) | 1979 (41.54) | 20 (48.78) | 2000 (41.37) | 18 (51.43) | 2089 (41.66) | 12 (46.15) | 1387 (40.02) | ||||
| Proxy required | 11 (19.64) | 428 (8.98) | .006 | 4 (9.76) | 332 (6.87) | .47 | 3 (8.57) | 250 (4.99) | .33 | 3 (11.54) | 167 (4.82) | .11 |
| ≥1 Imputed item | 14 (25.00) | 648 (13.60) | .01 | 9 (21.95) | 621 (12.84) | .08 | 5 (14.29) | 537 (10.71) | .50 | 3 (11.54) | 285 (8.22) | .54 |
Pearson χ2 test.
Table 2 describes adjusted associations, eTable 1 in the Supplement describes univariable associations, and eTable 2 in the Supplement reports verbatim NHATS questions and responses. At baseline, PD was associated with fear of falling, assistive device use, speech difficulty, and upper extremity weakness after adjustment. People who could lift and carry 20 pounds, stand independently from a kneeling position, and lift a heavy object overhead were less likely to have PD.
Table 2. Multivariable Associations Between Positive or Higher Survey Item Response and Parkinson Disease Diagnosis From Year –3 to Year 0.
| Item | Topic | Year 0 | Year −1 | Year −2 | Year −3 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| OR (95% CI)a | P value | OR (95% CI)a | P value | OR (95% CI)a | P value | OR (95% CI)a | P value | |||
| HC14 | Falls | 2.66 (0.78-1.12) | .46 | 1.05 (0.84-1.27) | .77 | 1.16 (0.93-1.45) | .18 | 1.10 (0.32-3.73) | .88 | |
| HC15 | Fear of falls | 2.11 (1.23-3.61) | .006 | 2.15 (1.14-4.05) | .02 | 2.32 (1.17-4.62) | .02 | 1.32 (0.57-3.04) | .51 | |
| HC19A | Anhedonia | 1.26 (0.98-1.62) | .07 | 1.45 (1.09-1.92) | .01 | 1.24 (0.89-1.73) | .21 | 0.98 (0.63-1.52) | .93 | |
| HC19B | Depression | 1.25 (0.92-1.69) | .15 | 1.33 (0.94-1.90) | .11 | 1.41 (0.96-2.06) | .08 | 1.04 (0.62-1.74) | .87 | |
| HC19C | Anxiety | 1.14 (0.84-1.55) | .41 | 1.11 (0.75-1.64) | .59 | 1.39 (0.97-1.99) | .07 | 1.46 (0.97-2.19) | .07 | |
| HC19D | Worry | 1.20 (0.91-1.58) | .19 | 1.02 (0.70-1.50) | .91 | 1.15 (0.78-1.69) | .48 | 1.00 (0.62-1.61) | .99 | |
| HC20 | Sleep-onset insomnia | 1.06 (0.86-1.30) | .52 | 1.10 (0.86-1.42) | .45 | 1.06 (0.81-1.38) | .68 | 1.23 (0.89-1.71) | .21 | |
| HC21 | Sleep-maintenance insomnia | 0.97 (0.77-1.23) | .81 | 0.97 (0.73-1.28) | .83 | 1.04 (0.77-1.40) | .80 | 0.94 (0.67-1.31) | .70 | |
| HC22 | Medication for insomnia | 0.98 (0.82-1.17) | .79 | 0.96 (0.78-1.19) | .71 | 1.23 (0.90-1.67) | .20 | 1.07 (0.79-1.45) | .66 | |
| MD1 | Assistive device | 3.28 (1.80-5.98) | <.001 | 1.88 (0.96-3.69) | .07 | 1.39 (0.89-1.42) | .32 | 2.04 (0.87-4.78) | .10 | |
| PC1 | Walking 6 blocks | 0.37 (0.72-1.04) | .13 | 0.45 (0.22-0.89) | .02 | 0.39 (0.19-0.83) | .01 | 0.34 (0.15-0.82) | .02 | |
| PC3 | Walking up 20 stairs | 0.60 (0.75-1.09) | .30 | 0.54 (0.27-1.08) | .08 | 0.42 (0.20-0.86) | .02 | 0.50 (0.22-1.17) | .11 | |
| PC5 | Carrying 20 pounds | 0.45 (0.74-0.83) | .01 | 0.64 (0.30-1.37) | .25 | 0.84 (0.36-1.93) | .67 | 0.72 (0.25-2.05) | .54 | |
| PC7 | Standing independently from a kneeling position | 0.51 (0.26-0.99) | .05 | 0.34 (0.15-0.76) | .008 | 0.64 (0.30-1.37) | .25 | 0.30 (0.11-0.85) | .02 | |
| PC9 | Lifting a heavy object overhead | 0.45 (0.24-0.82) | .01 | 0.51 (0.25-1.03) | .06 | 0.37 (0.18-0.79) | .01 | 0.36 (0.15-0.87) | .02 | |
| PC11 | Opening a jar | 0.72 (0.39-1.36) | .31 | 0.70 (0.33-1.46) | .34 | 0.66 (0.30-1.47) | .31 | 0.62 (0.26-1.51) | .30 | |
| SC17 | Toileting | 1.65 (0.65-4.20) | .29 | 0.97 (0.20-4.80) | .97 | 0.69 (0.08-5.66) | .73 | 1.04 (0.12-9.07) | .98 | |
| SS10 | Eye lenses | 0.97 (0.48-1.97) | .94 | 1.14 (0.45-2.91) | .78 | 1.83 (0.55-6.02) | .32 | 2.01 (0.47-8.67) | .35 | |
| SS11 | Other vision aids | 0.62 (0.31-1.24) | .17 | 0.62 (0.28-1.36) | .23 | 0.41 (0.16-1.08) | .07 | 0.64 (0.24-1.72) | .38 | |
| SS12 | Reading newspaper print | 1.45 (0.42-5.00) | .55 | 0.47 (0.15-1.43) | .18 | 1.76 (0.23-13.40) | .58 | 0.46 (0.13-1.60) | .22 | |
| SS13 | Dysphagia | 0.55 (0.20-1.55) | .26 | 2.47 (1.12-5.44) | .02 | 1.73 (0.66-4.52) | .26 | 1.40 (0.41-4.78) | .59 | |
| SS14 | Speech difficulty | 2.33 (1.12-4.84) | .02 | 2.54 (1.02-6.32) | .04 | 1.48 (0.44-5.01) | .52 | 2.03 (0.56-7.29) | .28 | |
| SS15 | Pain | 1.15 (0.67-1.97) | .61 | 2.09 (1.06-4.13) | .03 | 1.09 (0.56-2.13) | .80 | 1.45 (0.65-3.22) | .36 | |
| SS21 | Weakness in arms | 1.77 (1.04-3.02) | .04 | 1.77 (0.95-3.31) | .07 | 1.68 (0.84-3.37) | .14 | 1.35 (0.60-3.07) | .47 | |
| SS23 | Weakness in legs | 1.25 (0.73-2.15) | .42 | 2.12 (1.14-3.97) | .02 | 2.20 (1.11-4.35) | .02 | 1.13 (0.50-2.52) | .77 | |
| SS25 | Fatigue | 0.85 (0.49-1.46) | .55 | 1.43 (0.76-2.69) | .26 | 2.03 (1.01-4.09) | .05 | 1.22 (0.55-2.67) | .62 | |
| SS27 | Imbalance | 2.50 (1.42-4.39) | .002 | 2.70 (1.42-5.15) | .003 | 1.55 (0.77-3.10) | .22 | 2.77 (1.24-6.20) | .01 | |
Abbreviation: OR, odds ratio.
Odds of item responses for cases as the numerator and controls as the denominator. A response of 1 was used as the reference score for binary variables, and the lowest score was used for continuous variables. Adjusted for age group, gender, and proxy status.
Fear of falling, speech difficulty, and inability to stand independently from kneeling or to lift a heavy object overhead were also associated with PD prior to diagnosis, in addition to anhedonia, inability to walk 6 blocks or up 20 steps, dysphagia, pain, leg weakness, fatigue, and imbalance. Three years prior to diagnosis, symptoms remaining associated with PD included ability to walk 6 blocks (odds ratio [OR], 0.34; 95% CI, 0.15-0.82), stand independently from kneeling (OR, 0.30; 95% CI, 0.11-0.85), lift a heavy object overhead (OR, 0.36; 95% CI, 0.15-0.87), and imbalance (OR, 2.77; 95% CI, 1.24-6.20).
Discussion
We found that responses to questions addressing mobility and strength were associated with PD not only at diagnosis but also up to 3 years prior to diagnosis. While symptoms at baseline may reflect parkinsonism,1 their presence years prior to diagnosis supports literature suggesting that MPSs are also symptomatic4,9 and the argument that PPD should be recognized as a disease stage in its own right, for which identification will be important for disease-modifying intervention.10,11,12 Alternatively, it is possible that responses represented undiagnosed PD, which would indicate that in the US, diagnosis is delayed by years. Prior work has suggested that individuals presenting with gait dysfunction have longer diagnostic delays compared with those with tremor or other appendicular signs.13 Regardless, the results of this study corroborate those of a case-control study nested within an urban population study in the Netherlands,5 in which the authors also found greater imbalance in cases less than 4 years before diagnosis. While fall risk in that study was higher less than 2 years prior to diagnosis, which was not observed in the current study, this may reflect the fact that their questionnaire assessed falls in the past 12 months instead of 1 month.5
Strengths and Limitations
A strength of this study is that, while studies of PPD often rely on symptom recall or are limited to at-risk subgroups, our approach benefited from prospectively collected data from a national sample. However, there are several limitations. Evaluation of multiple associations may have caused spurious results; we did not perform a Bonferroni correction given the study’s exploratory nature and concerns for a type II error. Outcome misclassification is possible with retrospective analysis, and coding errors may influence results of claims-based studies; since ICD-9-CM and ICD-10-CM coding does not distinguish between PD and idiopathic parkinsonism, certain participants coded as such may not have had PD. Notably, prevalence of PD was higher than reported by a study across North America, which likely reflects those authors’ application of stricter diagnostic criteria and different case-ascertainment methods.14 While PPD subtyping is of growing interest,15 functional differences between potential subtypes could not be evaluated.
Conclusions
This case-control study showed that in nationwide longitudinal surveys of daily living, functional impairment associated with impending neurodegenerative diagnoses can be identified. Specifically, individuals with PPD (or unrecognized PD) may have greater motor dysfunction at least 3 years before diagnosis compared with the general population. Earlier identification of disease may be important not only for initiating symptomatic intervention but also for identifying populations for neuroprotective studies.
eMethods
eReference
eTable 1. Univariable Associations Between Positive or Higher Item Response and PD Diagnosis From Year –3 to Year 0
eTable 2. Questions and Response Options for Each Item Included From NHATS
References
- 1.Postuma RB, Berg D, Stern M, et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov Disord. 2015;30(12):1591-1601. doi: 10.1002/mds.26424 [DOI] [PubMed] [Google Scholar]
- 2.Mahlknecht P, Seppi K, Poewe W. The concept of prodromal Parkinson’s disease. J Parkinsons Dis. 2015;5(4):681-697. doi: 10.3233/JPD-150685 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Heinzel S, Berg D, Gasser T, Chen H, Yao C, Postuma RB; MDS Task Force on the Definition of Parkinson’s Disease . Update of the MDS research criteria for prodromal Parkinson’s disease. Mov Disord. 2019;34(10):1464-1470. doi: 10.1002/mds.27802 [DOI] [PubMed] [Google Scholar]
- 4.Prasuhn J, Piskol L, Vollstedt EJ, et al. Non-motor symptoms and quality of life in subjects with mild parkinsonian signs. Acta Neurol Scand. 2017;136(5):495-500. doi: 10.1111/ane.12760 [DOI] [PubMed] [Google Scholar]
- 5.Darweesh SK, Verlinden VJ, Stricker BH, Hofman A, Koudstaal PJ, Ikram MA. Trajectories of prediagnostic functioning in Parkinson’s disease. Brain. 2017;140(2):429-441. doi: 10.1093/brain/aww291 [DOI] [PubMed] [Google Scholar]
- 6.Kasper JD, Freedman VA. Findings from the 1st round of the National Health and Aging Trends Study (NHATS): introduction to a special issue. J Gerontol B Psychol Sci Soc Sci. 2014;69(suppl 1):S1-S7. doi: 10.1093/geronb/gbu125 [DOI] [PubMed] [Google Scholar]
- 7.Szumski NR, Cheng EM. Optimizing algorithms to identify Parkinson’s disease cases within an administrative database. Mov Disord. 2009;24(1):51-56. doi: 10.1002/mds.22283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393. doi: 10.1136/bmj.b2393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Louis ED, Tang MX, Schupf N, Mayeux R. Functional correlates and prevalence of mild parkinsonian signs in a community population of older people. Arch Neurol. 2005;62(2):297-302. doi: 10.1001/archneur.62.2.297 [DOI] [PubMed] [Google Scholar]
- 10.Ntanasi E, Maraki M, Yannakoulia M, et al. Frailty and prodromal Parkinson’s disease: results from the HELIAD Study. J Gerontol A Biol Sci Med Sci. 2021;76(4):622-629. doi: 10.1093/gerona/glaa191 [DOI] [PubMed] [Google Scholar]
- 11.Devos D, Hirsch E, Wyse R. Seven solutions for neuroprotection in Parkinson’s disease. Mov Disord. 2021;36(2):306-316. doi: 10.1002/mds.28379 [DOI] [PubMed] [Google Scholar]
- 12.Postuma RB, Berg D. Prodromal Parkinson’s disease: the decade past, the decade to come. Mov Disord. 2019;34(5):665-675. doi: 10.1002/mds.27670 [DOI] [PubMed] [Google Scholar]
- 13.Breen DP, Evans JR, Farrell K, Brayne C, Barker RA. Determinants of delayed diagnosis in Parkinson’s disease. J Neurol. 2013;260(8):1978-1981. doi: 10.1007/s00415-013-6905-3 [DOI] [PubMed] [Google Scholar]
- 14.Marras C, Beck JC, Bower JH, et al. ; Parkinson’s Foundation P4 Group . Prevalence of Parkinson’s disease across North America. NPJ Parkinsons Dis. 2018;4:21. doi: 10.1038/s41531-018-0058-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Berg D, Borghammer P, Fereshtehnejad SM, et al. Prodromal Parkinson disease subtypes—key to understanding heterogeneity. Nat Rev Neurol. 2021;17(6):349-361. doi: 10.1038/s41582-021-00486-9 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods
eReference
eTable 1. Univariable Associations Between Positive or Higher Item Response and PD Diagnosis From Year –3 to Year 0
eTable 2. Questions and Response Options for Each Item Included From NHATS
