Cerebral amyloid angiopathy, brain iron concentrations, and cognitive decline in older people

Abstract

Cerebral amyloid angiopathy (CAA) is a common brain pathology in older people and has been recently recognized as a major risk factor for amyloid-related imaging abnormalities during anti-amyloid antibody therapy. CAA pathophysiology may involve iron released from ruptured vessels, but the association between postmortem CAA and brain iron is unclear. This study investigates the association between CAA and brain iron and whether elevated iron modifies the association between CAA and cognitive decline. We studied 626 Rush Memory and Aging Project decedents (mean age at death = 90 [SD = 6.1] years, 70% women) who completed baseline and longitudinal cognitive assessments and underwent detailed neuropathologic evaluation for CAA, Alzheimer’s disease neuropathologic changes (ADNC), and other brain pathologies. Brain iron content was assessed from the inferior temporal cortex using Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Linear regression and mixed-effects models were used for analysis. CAA was common: 266 (42%) had mild, 153 (24%) had moderate, and 75 (12%) had severe CAA. In analyses adjusted for demographics, intermediate/high ADNC, and other pathologies, the presence and severity of CAA were associated with elevated cortical brain iron (Est = 0.033, SE = 0.011, p = 0.002; Est = 0.017, SE = 0.004, p < 0.001, respectively). When examining associations with nonlinear cognitive change before death (mean follow-up = 7.7 [SD = 3.9] years), both CAA and elevated iron were independently associated with faster annual rates of decline in global cognition and semantic memory (all p < 0.04). Elevated iron was also associated with faster declines in episodic and working memory and perceptual speed (all p < 0.02). When exploring whether brain iron modulates the association of CAA with cognitive decline, we found that CAA had steeper decline in perceptual speed when elevated iron was present compared to when low iron was present (p = 0.03). Together, these findings suggest that brain iron may contribute to the clinical impact of CAA in older age.