Current and projected future economic burden of Parkinson's disease in the U.S.

Parkinson's disease (PD) is a slowly progressive neurodegenerative disorder affecting about one million Americans and is associated with motor and non-motor symptoms (e.g., tremor, bradykinesia, anxiety, depression, cognitive impairment) and significant comorbidities (infections, cardiovascular and gastrointestinal disorders, fall-related injuries). Persons with PD (PWPs) have substantial healthcare needs and experience reduced employment, missed work, and require paid and unpaid caregiving, creating sizable indirect economic burdens. Despite PD being a rapidly growing neurological disorder, comprehensive estimates of its current U.S. economic burden have been limited. This study aims to quantify the total economic burden of PD in the U.S. in 2017 and project its impact through 2037, using a human capital approach to value lost productivity from illness, disability, and premature mortality. The authors synthesize multiple datasets, including nationally representative claims and a dedicated primary survey, to capture both direct medical and broader indirect/non-medical cost components.

The discussion situates findings relative to prior U.S. estimates. Earlier studies (e.g., Kowal et al. 2013; O’Brien et al. 2009; Huse et al. 2005) reported lower prevalence (roughly 0.5–0.65 million) and smaller cost totals, often due to older data, reliance on cost-allocation or limited encounter data, exclusion of certain cost categories (e.g., durable medical equipment, some outpatient services), or extrapolation from non-U.S. prevalence. The current study uses more recent prevalence data (2011–2015 MEPS and 2015 MCBS), claims-based cost differentials with matched controls by age, gender, race/ethnicity, and insurance, and a comprehensive primary survey for indirect and non-medical costs, yielding higher and more comprehensive estimates. Direct medical cost estimates align with recent Medicare-based estimates (Mantri et al.), while other work (Dieleman et al.) allocates comorbidity costs to specific conditions rather than PD, producing smaller PD totals. The paper also compares projections with U.S. and international forecasts, noting broadly consistent growth but with methodological differences driving variation.

Design and perspective: Prevalence-based, societal perspective, human capital approach valuing productivity losses from illness, disability, and premature mortality. Year of analysis: 2017 U.S. dollars; projections presented in real 2017 dollars without inflation adjustments. Data sources: Prevalence from 2011–2015 Medical Expenditure Panel Survey (MEPS) and 2015 Medicare Current Beneficiary Survey (MCBS), applied to 2017 U.S. Census population counts and to 2018–2037 projections. Direct medical costs from 2016 Optum de-identified Normative Health Information (dNHI) private claims (continuous full-year coverage) and 2015 Medicare 5% Standard Analytical File (SAF) claims, supplemented with MCBS for components not in SAF (e.g., Part D drugs not available, long-term care/long-stay SNF, nursing home). Indirect and non-medical costs from a primary PD Impact Survey (n = 4,548 completed responses after quality checks) administered via Fox Insight and UPAC networks, weighted to the national PD population by age and gender; supplemented with CDC WONDER mortality data and Bureau of Labor Statistics earnings data. Case identification: In claims, PWPs identified by ≥1 inpatient or ≥2 non-inpatient PD-related diagnoses, or ≥1 outpatient PD diagnosis plus antiparkinsonian prescription (private claims). A diagnostically inclusive approach captured PD and related parkinsonian disorders (ICD-9/ICD-10 codes: 332.0/G20, 332.1/G21, DLB, PSP, MSA, etc.) to reduce false negatives given diagnostic challenges. In MEPS, PD identified using any ICD-9 332 code in chronic condition files; in MCBS, via self-report/chronic condition screens/facility datasets with de-duplication. Direct medical cost estimation: Excess costs computed as the difference between average annual medical costs of PWPs and matched controls (10:1 matching) on age, gender, race/ethnicity, and insurance. Costs included plan, patient, and third-party payments across hospital inpatient, outpatient/ancillary, physician office, prescription drugs, durable medical equipment, and non-acute institutional care (SNF, nursing home, hospice). For PWPs <65 with non-private, non-Medicare coverage, costs imputed from private claims within age-gender strata, adjusted downward by 28.6% using a GLM (gamma, log link) estimated cost ratio between private and non-private groups from MEPS. Indirect costs: Premature mortality productivity losses estimated by combining PD-attributable premature deaths (CDC WONDER; Medicare SAF) with net present value of future earnings by sex and age (18–74), accounting for labor force participation, mortality rates, 1% productivity growth, and 3% discounting. Additional indirect components (reduced employment, absenteeism, presenteeism, social productivity/volunteer work) and caregiver on-the-job productivity losses derived from PD Impact Survey responses, with national weights. Non-medical and disability costs: From the PD Impact Survey, including paid daily non-medical care, home and motor vehicle modifications, transportation, and other expenses; disability income components include SSI, SSDI, and other sources (e.g., VA, government employee, state or personal disability insurance). Projections: Applied 2017 age- and gender-specific PD prevalence rates to U.S. Census population projections (2018–2037), assuming constant incidence and mortality rates, and multiplied by 2017 per-PWP costs by component to estimate future total burdens in real 2017 dollars.

- Prevalence: Estimated 1.04 million individuals with diagnosed PD in the U.S. in 2017; ~89% Medicare-eligible; higher prevalence in ≥65 (16.9 per 1,000) vs. <65 (0.7 per 1,000); more prevalent in men (595,000) than women (443,000). - Total economic burden (2017): $51.9 billion. - Direct medical costs: $25.4 billion (per PWP $24,439). Approximately 90% borne by Medicare-eligible populations; 7% private; 3% other coverage. Largest service categories: hospital inpatient (28.4%), non-acute institutional care (28.2%), outpatient/ancillary (21.7%). - Indirect and non-medical costs: $26.5 billion (per PWP $25,558), comprising: - Indirect costs (PWP + caregivers): $14.2 billion, including premature death, reduced employment, absenteeism, presenteeism, and social productivity losses; caregiver absenteeism/presenteeism exceeded PWP on-the-job losses due to higher caregiver employment rates. - Non-medical costs: $7.5 billion (paid daily non-medical care; home and vehicle modifications; other expenses). - Disability income: $4.8 billion (SSI, SSDI, other disability payments). - Per-PWP indirect and non-medical costs (PWP + caregiver) slightly exceed per-PWP direct medical costs ($25,558 vs. $24,439). - Projections: PD prevalence projected to rise from 1.04 million (2017) to ~1.64 million by 2037, with disproportionate growth in ages ≥75. Total economic burden projected to increase from $51.9B (2017) to ~$79.1B (2037). From 2017 to 2037, direct medical costs and social productivity loss increase by ~52%; disability income, paid care, other non-medical, and caregiver productivity loss by ~50%; premature death, reduced employment, absenteeism, and presenteeism by ~9–15%.

The study provides a comprehensive, up-to-date assessment of PD’s economic burden in the U.S., revealing a larger societal impact than previously recognized. By integrating multiple representative data sources and a large PD-specific survey, the analysis captures cost components often missed in earlier work (e.g., detailed non-medical expenses, caregiver productivity losses). The dominance of Medicare in bearing direct costs reflects the age distribution of PD, while substantial indirect and non-medical burdens highlight impacts on patients, caregivers, employers, and communities. Comparisons with prior studies suggest that methodological differences—more recent prevalence data, claims-based matched-control cost differentials, and inclusion of broader indirect and non-medical costs—drive higher totals. The projected growth in prevalence and costs, driven by population aging, underscores the urgency of interventions that reduce incidence (e.g., addressing environmental risks), delay progression (disease-modifying therapies), improve symptom management (falls, cognition), and expand access to specialist care, all of which may reduce future utilization, improve outcomes, and mitigate costs.

This study estimates that diagnosed PD affected about 1.04 million Americans in 2017 with a total economic burden of $51.9 billion, split roughly evenly between direct medical and indirect/non-medical costs, and projects growth to ~1.64 million affected individuals and ~$79.1 billion in costs by 2037. Contributions include the use of comprehensive, recent data sources and a large primary survey to quantify often underappreciated components such as caregiver productivity losses and non-medical expenditures. Policy and research priorities include preventive strategies addressing environmental and other modifiable risks, development and dissemination of therapies that slow progression, improved management of disabling symptoms, and enhanced access to specialist care. Future research should refine estimates for underrepresented subgroups (e.g., Medicaid populations), better capture long-term care and undiagnosed PD, and explore geographic and demographic disparities to inform targeted interventions.

- Undiagnosed PD (estimated 12–42% of PWPs) not captured, likely understating prevalence and costs. - Case identification limitations: MEPS uses 3‑digit ICD-9 code 332, which may include PD mimics; reliance on any single PD diagnosis in MEPS for <65 may introduce false positives/negatives; inclusive diagnostic approach increases uncertainty but reduces false negatives. - Small sample sizes for prevalence sources (MEPS n≈245 PWPs; MCBS n≈229) may reduce precision and preclude subgroup analyses (e.g., by race/ethnicity). - Imputation for PWPs <65 with non-private, non-Medicare coverage based on adjusted private claims may misestimate costs if service use differs materially. - Potential underestimation of long-term care costs due to limited Medicaid-specific data; MCBS used for Medicare beneficiaries but may not fully capture differences in the “Other” insured group. - Primary survey (PD Impact Survey) is convenience-based and subject to selection and recall biases, though weighting mitigates some concerns. - Projections assume constant incidence/mortality and no changes in treatment intensity, utilization, or prices (real 2017 dollars), potentially understating or overstating future burdens.