Document resource
Introduction Broad-spectrum antibiotic exposure after acute ischemic stroke (AIS) may influence long-term neurologic recovery through microbiome-mediated inflammatory pathways; however, infection severity may confound these associations. We evaluated three-year cognitive and clinical outcomes across four mutually exclusive post-AIS exposure phenotypes defined by pneumonia and antibiotic use. Early post-stroke pneumonia was used as a clinically relevant infection indicator to distinguish infection-related from antibiotic-related risk.Methods Adults age 18 years or older with AIS (2010-2023) and no baseline dementia or cognitive impairment were identified using the TriNetX federated research network (110 healthcare organizations). Pneumonia and exposure to broad-spectrum antibiotics within 30 days defined four cohorts: pneumonia with antibiotics, pneumonia without antibiotics, antibiotics without pneumonia, and neither exposure. Outcomes from day 1 through 3 years included cognitive impairment, delirium, recurrent ischemic stroke, and all-cause mortality. Pairwise 1:1 propensity score matching balanced demographics, vascular and cardiovascular comorbidities, prior cerebrovascular disease, chronic kidney disease, nicotine dependence, peripheral vascular disease, antithrombotic and antiplatelet therapy, and stroke severity and treatment proxies, including mechanical ventilation, critical care services, and endovascular thrombectomy.Results Matched cohorts included 34,498 patients with pneumonia and antibiotics, 15,887 with pneumonia without antibiotics, 120,076 with antibiotics without pneumonia, and 1,031,366 with neither exposure. Compared with neither exposure, pneumonia with antibiotics was associated with higher rates of cognitive impairment (18.1 percent vs 14.3 percent; OR 1.32), delirium (3.1 percent vs 2.0 percent; OR 1.58), and mortality (43.5 percent vs 17.1 percent; OR 3.72). Pneumonia without antibiotics was associated with higher mortality (30.1 percent vs 14.4 percent; OR 2.55) but lower cognitive impairment (10.5 percent vs 11.9 percent; OR 0.88). Antibiotics without pneumonia were associated with increased cognitive impairment (14.3 percent vs 13.1 percent; OR 1.11), delirium (2.2 percent vs 1.6 percent; OR 1.40), and mortality (27.6 percent vs 13.8 percent; OR 2.38). Within the pneumonia cohort, antibiotic exposure was associated with increased cognitive impairment (15.9 percent vs 11.0 percent; OR 1.53), delirium (2.6 percent vs 1.1 percent; OR 2.52), and mortality (42.2 percent vs 31.0 percent; OR 1.63). Among antibiotic-exposed patients, concurrent pneumonia increased cognitive impairment (17.7 percent vs 16.0 percent; OR 1.13) and mortality (44.2 percent vs 33.3 percent; OR 1.58). Direct comparison of antibiotics without pneumonia versus pneumonia without antibiotics showed higher cognitive impairment (13.3 percent vs 10.3 percent; OR 1.33) with similar mortality (29.2 percent vs 30.0 percent; OR 0.96).Conclusions In this large propensity-matched analysis, post-stroke pneumonia was the primary driver of mortality, whereas broad-spectrum antibiotic exposure was independently associated with adverse neurocognitive outcomes, including cognitive impairment and delirium, even in patients without pneumonia. Patients with both pneumonia and antibiotic exposure experienced the worst overall recovery profile, suggesting additive effects from infection-related systemic inflammation and potential microbiome disruption. These findings support the need for prospective trials evaluating antimicrobial stewardship, microbiome-preserving strategies, and targeted neuroprotective interventions following AIS.Disclosures M. Costa: None. S. O’Leary: None. C. Young: None.