Association Between Cerebral Amyloid Angiopathy and Nontraumatic Subdural Hemorrhage.

BACKGROUND AND OBJECTIVES: Cerebral amyloid angiopathy (CAA) is a common cause of intracerebral hemorrhage (ICH) in older patients. Whether CAA is associated with isolated subdural hemorrhage (SDH), without an accompanying ICH, remains unclear. We, therefore, tested this relationship in a large, heterogeneous sample of patients across the United States.

METHODS: We performed a retrospective cohort study using administrative claims data from all admissions to nonfederal acute care hospitals in 11 states in the United States between 2016 and 2021. Among hospitalized patients, we included only those aged 50 years or older, a threshold necessary to meet Boston criteria v2.0 for CAA. We divided this population into 3 groups: those with a diagnosis of CAA, those with other cerebrovascular diseases (CVDs) but without CAA, and those with neither CAA nor other CVDs. The main outcome was a first-documented, isolated, nontraumatic SDH; we did not count SDH cases with a concurrent traumatic brain injury. The exposures and outcome were identified using previously validated ICD-10-CM diagnosis codes. Using Cox regression analyses, we compared the risk of incident SDH among the 3 groups after adjustment for demographics and comorbidities. In prespecified sensitivity analyses, patients with a baseline diagnosis of dementia were excluded.

RESULTS: Among 8.5 million hospitalized patients aged 50 years or older, 2,335 had CAA and 600,646 had other CVDs. During a median follow-up of 2.0 years (interquartile range 1.0-3.9), incident SDH occurred in 34 patients with CAA (1.5%), 3,552 patients with other CVDs (0.6%), and 35,425 patients without CAA or other CVDs (0.4%). In adjusted Cox regression analysis, there was an increased risk of incident SDH seen with CAA (hazard ratio [HR] 3.1; 95% CI 2.2-4.4) and with prevalent CVD (HR 1.4; 95% CI 1.3-1.5). Findings were similar in sensitivity analyses excluding patients with dementia.

DISCUSSION: In a large, heterogeneous cohort, we found that CAA was associated with a 3-fold heightened risk of SDH, higher than the increased risk seen in patients with other CVDs. These findings support the emerging hypothesis that CAA is a risk factor of isolated nontraumatic SDH.