Back

National trends, demographic disparities, and phenotypic shifts in encephalopathy-associated mortality in the United States: A CDC WONDER analysis, 1999-2024

Hamed, K. J. A.; Sayah, M. A.; Bundid, R. M.

2026-09-27 epidemiology
10.64898/2026.09.21.26363584 medRxiv
Show abstract

Encephalopathy is a lethal manifestation of global cerebral dysfunction spanning hypoxic, metabolic, toxic, and vascular etiologies, yet its population burden is difficult to quantify because it predominantly appears as a contributing rather than underlying cause of death. Here, we analyzed US multiple-cause death records from 1999 through 2024 using harmonized CDC WONDER datasets. A longitudinally consistent core definition of 10 ICD-10 codes identified 953,846 encephalopathy-associated deaths. The national age-adjusted mortality rate increased 28.4%, from 9.143 per 100,000 in 1999 to 11.738 in 2024, after peaking at 14.391 in 2021. Segmented log-linear regression identified four temporal phases: 1999-2004 (annual percent change [APC],-1.76%), 2004-2018 (APC, 2.57%), 2018-2021 (APC, 6.54%), and 2021-2024 (APC, -5.98%). Encephalopathy was listed as the underlying cause in only 21.6% of deaths in 2024. Unspecified encephalopathy surged more than 3.6-fold, while anoxic brain injury remained stable. Mortality remained higher in males than females (13.973 versus 9.778 per 100,000), and rates among non-Hispanic Black individuals were 1.90-fold those of non-Hispanic White individuals. Geographically, mortality clustered in the Midwest and South (13.144 and 13.120 per 100,000 in 2024), with persistent state-level dispersion across the study period. COVID-19 accounted for 10.9% of associated deaths in 2021 before falling to 1.4% in 2024. These findings reveal a nonlinear trajectory shaped by diagnostic shifts, pandemic disruption, and persistent demographic and geographic disparities.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.