Abstract
BackgroundInfections have been associated with a greater risk of Alzheimer’s disease and related dementias (ADRD), but it is unclear how infections influence structural brain patterns over time, and whether post-infection brain atrophy can accelerate cognitive decline and molecular changes underlying dementia.MethodsUsing the Baltimore Longitudinal Study of Aging (BLSA; n = 793; mean age = 70.1), we examined how infections relate to longitudinal changes in machine learning-derived, 3 T-MRI neuroimaging signatures, and leveraged the UK Biobank (UKB; 1,120; mean age = 62.9 yrs) to externally validate infection-brain atrophy relationships. Using the BLSA, we also asked if infection history and infection-related brain volume loss were associated with cognitive decline, amyloid-beta PET, and ADRD plasma biomarker trajectories (Aβ42/40, pTau-181, NfL, GFAP).ResultsWe detected accelerated parieto-temporal atrophy in BLSA participants with a history of upper respiratory tract, bacterial, and urinary tract infections (p < 0.05), as well as influenza and skin/subcutaneous infections (FDR p < 0.05). After demonstrating their associations with longitudinal neuroimaging signatures in the UKB and prevalent dementia in the BLSA, we found that infections were related to a greater burden of ADRD plasma biomarkers and accelerated rates of cognitive decline in BLSA participants. Integrating longitudinal brain scans, cognitive assessments, and plasma biomarker measurements, we identified infection-related changes in verbal memory and NfL that were more prominent among BLSA participants who experienced greater post-infection brain atrophy.ConclusionAlong with demonstrating that infections mediate clinically relevant brain atrophy patterns, these findings highlight the consequences of post-infection brain volume loss on longitudinal neurocognitive outcomes and extend our understanding of the biological basis by which infections may contribute to neurodegeneration.