Abstract Body: Introduction. Eicosanoids (EIC) are lipid mediators that play a systemic role in inflammation and, more specifically, in the brain, contribute to the process of neurovascular coupling. Though EICs have been associated with immediate stroke risk and stroke risk factors, the long-term risk and associations with stroke risk factors have not been previously studied.

Methods. Plasma EICs (214) were measured using non-targeted liquid chromatography mass spectrometry in conjunction with chemical networking of spectral fragmentation patterns among participants of the Atherosclerosis Risk in Communities Study at visit 2 (1990-92). Incident stroke (ischemic and hemorrhagic) was ascertained until the end of 2021 with adjudication by an expert panel. Cox regression was used to estimate the hazard ratio (HR) for incident stroke per doubling of EIC levels, and p-values were FDR corrected. Primary model 1 adjusted for age, sex, race/center and additional models adjusted for stroke risk factors (eGFR, BMI, hypertension, diabetes, current smoking status, total cholesterol, prevalent atrial fibrillation, prevalent heart failure, prevalent coronary heart disease, aspirin use, cholesterol-lowering medications, anticoagulation medication use, time-dependent incident atrial fibrillation and incident anticoagulant use).

Results. Among the 9,444 participants included, the mean age was 57 (5.7 SD), and 43.2% were men. Over a mean of 22.7 (8.5 SD) years of follow-up, 950 participants had an incident stroke. In model 1, higher levels of 9 EICs were associated with a higher risk of incident stroke (11t LTD4 [HR 1.21, p=7.4E-05], 5S HpEPE [HR 1.21, p=0.005], LXA5 [HR 1.15 p=0.006], 13 HpODE_a [HR 1.17 p=0.009], 17S HpDHA_a [HR 1.16, p=0.018], 15 oxoEDE_a [HR 1.29 p=0.03], 11, 12 di-HETrE_a [HR 1.21 p=0.03], Arachidonic Acid_a [HR 1.17 p=0.03], and 5S HpETE_c [HR 1.18 p=0.04]; and 3 EIC analytes were associated with lower risk of incident stroke (PGF1a_b [HR 0.27 p=0.006], 13, 14 dihydroPGF1a [HR 0.37 p=0.02], 13,14 dihydroPGE1_b [HR 0.5 p=0.03]. Associations did not reach FDR-corrected significance after adjusting for risk factors.

Conclusion. Our analysis of this large longitudinal community-based study demonstrates that the risk of incident stroke over decades is associated with alterations in the balance of increasing pro-inflammatory EICs likely interrelated with the presence of stroke risk factors over time and interestingly including an EIC associated with neurovascular coupling.