Abstract
Inflammation contributes to cerebrovascular disease, but prospective studies of broad immune marker panels and incident stroke are limited. We aimed to identify plasma immune markers independently associated with incident stroke. Stroke-free Northern Manhattan Study participants aged ≥ 50 years had 60 plasma immune-related proteins measured by Luminex. Least absolute shrinkage and selection operator (LASSO) Cox regression selected markers associated with incident all-cause stroke. Selected markers were tested in Cox models of stroke risk sequentially adjusted for vascular risk, cardiac comorbidities, interleukin-6, interleukin-1β, tumor necrosis factor-α, and white matter hyperintensity volume (WMHV). Ischemic stroke was a secondary outcome, and bootstrap resampling provided internal validation. Among 1175 participants (mean age 70 ± 9 years; 60% women; 67% Hispanic), 130 strokes, 101 (78%) ischemic, occurred over a median of 14 years. LASSO selected only C-X-C motif chemokine ligands 9 and 10 (CXCL9 and CXCL10). Each 1-standard deviation greater CXCL10 was associated with 20% greater all-cause stroke risk after adjustment for vascular risk, cardiac comorbidities, and inflammatory cytokines (HR 1.20; 95%CI 1.04-1.38). The association persisted after WMHV adjustment (HR 1.18; 95%CI 1.02-1.37), was similar for ischemic stroke (HR 1.19; 95%CI 1.01-1.41), and was stable across bootstrap resamples. CXCL9 was not independently associated with stroke. In a multiethnic urban aging cohort followed prospectively for over a decade, baseline plasma CXCL10, also known as interferon-γ-induced protein 10 (IP-10), was independently associated with incident stroke. CXCL10 may mark interferon-γ-responsive immune activity linked to cerebrovascular vulnerability. Whether CXCL10 signals a modifiable pathway for stroke prevention warrants investigation.