A serial analysis of United States death certificates found that the age-adjusted, Parkinson's-related mortality rate did not change in a statistically significant way between 1999 and 2025, even though the raw rate rose from 32.74 to 41.03 deaths per 100,000 adults aged 45 and older over that span. The study, published in Frontiers in Neurology, drew on 1,149,731 deaths in which Parkinson's disease (ICD-10 code G20) was listed anywhere on the certificate, using the CDC's Wide-ranging Online Data for Epidemiologic Research, or WONDER, Multiple Cause of Death database. The researchers calculated an average annual percent change of 0.56 percent for the full 26-year window, with a 95 percent confidence interval running from negative 0.10 to 1.22 and a p-value of 0.099, just short of the conventional threshold for significance.

The authors used a log-scale ensemble model, weighted against the model's own past forecasting errors, to project the national rate forward to 41.73 deaths per 100,000 by 2030 (95 percent prediction interval, 33.97 to 51.27). Even though the overall national trend was flat within statistical uncertainty, several subgroups showed increases that did reach significance: adults aged 45 to 64 (average annual percent change of 1.84 percent), deaths in the South census region (1.06 percent), and deaths in non-metropolitan counties (1.28 percent). Men and women separately did not show significant trends, nor did the Northeast, Midwest or West regions.

A substantial part of the recent-year movement, the authors found, was an artifact of the pandemic rather than a change in Parkinson's mortality itself. COVID-19 was co-listed on Parkinson's-related death certificates in 10.15 percent of cases in 2020, a share that fell to under 1 percent by 2025. That co-listing accounted for an estimated 58 percent of the single-year jump in the mortality rate between 2019 and 2020, and for 55 percent of the decline recorded between 2021 and 2024. With a death certificate series stripped of COVID-19 co-listings, the rate was essentially flat across the pandemic years. The authors describe this COVID-unwinding effect as 'now nearly exhausted,' meaning it can no longer be expected to drive the recorded rate down in future years.

The new findings sit alongside an earlier national analysis covering 1999 to 2019, published in Neurology in 2021, which used a narrower definition, Parkinson's listed as the underlying rather than any-mention cause, and reported a statistically significant rise over that shorter window. The difference illustrates how the choice of death-certificate definition and study window can change whether a national mortality trend reads as significant or not; it does not resolve which definition better tracks the disease itself.