AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
A Possible Synergistic Association of Obstructive Sleep Apnea and Patent Foramen Ovale on Ischemic Stroke Risk
By Grant B, Elisarraras FX, Negmeldin F et al.
Background: Obstructive sleep apnea (OSA) and patent foramen ovale (PFO) are each independently associated with ischemic stroke. Mechanistically, OSA can induce transient surges in right-sided cardiac pressures, which may cause shifts in atrial gradients, leading to greater right-to-left shunting through a PFO. This may potentially amplify stroke risk when both conditions coexist. Existing evidence is limited to small case series, with no population-level cohort that has tested the interaction. Methods: We performed a one-center retrospective analysis, identifying all adults (≥18 years) with at least one encounter participating in the Los Angeles Data Resource cohort tool (n=4,205,470). Patients were classified by diagnosis and procedure codes (ICD and CPT, respectively) into four groups by OSA and PFO status. Relative risks were calculated based on stratified 2x2 tables. The primary outcome was ischemic stroke (I63 or G46). The impact of OSA on stroke risk was then stratified based on PFO status and other common risk factors for stroke: atrial fibrillation (AF), hypertension (HTN), type 2 diabetes (T2DM), congestive heart failure (CHF), and age. Additive interaction was assessed via the Relative Excess Risk due to Interaction (RERI). Statistical significance was established at p<.05 . Results: Among 142,465 patients with OSA, the presence of a PFO was associated with a 9.04-fold higher stroke prevalence ( p<.01 , 95% CI 8.39-9.75). Among the 8,720 patients with PFO, addition of OSA conferred a 1.30-fold increased risk of stroke ( p<.01 , 95% CI 1.20-1.41) (Figure 1). The interaction was synergistic (RERI +8.66, p<.01 ). When stratified by AF status, in AF-absent patients, the addition of PFO to OSA conferred a RR of 12.40 ( p<.01 , 95% CI 11.25-13.67) and a RERI of +15.66 ( p<.01 ). In AF-present patients, the addition of PFO to OSA patients conferred a RR of 2.99 ( p<.01 , 95% CI 2.66-3.37) and an RERI of -0.17 ( p=0.59 ), (Figure 2). When further stratified by common stroke risk factors (Age, T2DM, CHF, HTN), there is a unique contribution to stroke risk in patients who have both OSA and PFO when excluding these risk factors (Figure 3). Conclusions: Those with both OSA and PFO demonstrated a synergistic interaction on ischemic stroke prevalence. The association was not explained by AF and persisted across common stroke risk factors, though detection bias and the cross-sectional design preclude causal interpretations or relationships. Patient-level data is needed for confirmation, though the presence of a PFO may warrant further investigation as a contributor to stroke risk in patients with OSA.