Marras · NPJ Parkinson's disease 2018 · Multi-cohort epidemiological meta-analysis and prevalence modeling study · n=5 studies

Prevalence of Parkinson's disease across North America.

Cited 1027 times in the scientific literature.

Level 3 - non-randomized controlled study

Meta-analysis and synthesis of multi-region observational cohort studies and administrative datasets.

PubMed 30003140 · doi:10.1038/s41531-018-0058-0 · record verified 2026-08-27

What was done

Researchers synthesized epidemiological data from five separate cohort studies across California (two cohorts), Minnesota (one cohort), Hawaii (one cohort), and Ontario, Canada (one cohort) alongside complementary US Medicare program data. Case ascertainment methods included health-care record reviews (three cohorts), active clinical ascertainment across facilities and neurology practices (one cohort), and longitudinal population follow-up (one cohort). Age- and sex-specific meta-estimates derived from California, Minnesota, and Ontario were applied to 2010 US population structure data and US Census Bureau projections to estimate current and future Parkinson's disease prevalence.

What was found

The estimated overall prevalence of Parkinson's disease among North Americans aged 45 years and older was 572 per 100,000 (95% confidence interval, 537 to 614). This represented approximately 680,000 individuals aged 45 years and older with Parkinson's disease in the United States in 2010. Projections estimated this number would increase to approximately 930,000 in 2020 and 1,238,000 in 2030. Estimates from the Hawaii cohort were consistently lower across age groups, and regional variations persisted across cohorts and Medicare data.

Why it matters

This study provides multi-region population-based estimates of Parkinson's disease burden across North America to guide healthcare infrastructure planning. The projected increase to more than 1.2 million US cases by 2030 emphasizes the growing need for specialized neurological care resources.

Limits

Total individual participant count was not stated in the abstract. Meta-estimates excluded the Hawaii cohort due to lower prevalence rates, leaving unexplained geographic heterogeneity. Ascertainment methodologies varied substantially across cohorts (administrative records, active clinic ascertainment, and longitudinal follow-up), and the authors note that these figures likely represent minimum prevalence estimates. Estimates were restricted to populations aged 45 years and older.

Cited by