Association of Assisted Reproductive Technology and Stroke During Hospitalization for Delivery in the United States

Abstract

Background: Infertility treatment with assisted reproductive technologies (ART) has been associated with adverse vascular events in some, but not all previous studies. Endothelial damage, prothrombotic factor release, and a higher prevalence of cardiovascular risk factors in those receiving ART have been invoked to explain this association. We sought to explore the relationship between ART and stroke risk using population-level data. Methods: We conducted a retrospective cohort study using data from the National Inpatient Sample (NIS) registry from 2015-2020, including all delivery hospitalizations for patients aged 15-55 years. The study exposure was use of ART. The primary endpoint was any stroke defined as ischemic stroke (IS), subarachnoid hemorrhage (SAH), intracerebral hemorrhage (ICH), or cerebral venous thrombosis (CVT) during index delivery hospitalization. Individual stroke subtypes (IS, SAH, ICH, and CVT) were evaluated as secondary endpoints. Standard International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) algorithms were used to define study exposure, comorbidities, and prespecified endpoints. In addition to reporting population-level estimates, propensity score (PS) adjustment by inverse probability weighting (IPW) was used to mimic the effects of randomization by balancing baseline clinical characteristics associated with stroke between ART and non-ART users. Results: Among 19,123,125 delivery hospitalizations identified, patients with prior ART (n = 202,815, 1.1%) experienced significantly higher rates of any stroke (27.1/100,000 vs. 9.1/100,000), IS (9.9/100,000 vs. 3.3/100,000), SAH (7.4/100,000 vs. 1.6/100,000), ICH (7.4/100,000 vs. 2.0/100,000), and CVT (7.4/100,000 vs. 2.7/100,000) in comparison to non-ART users (all p < 0.001 for all unadjusted comparisons). Following IPW analysis, ART was associated with increased odds of any stroke (aOR 2.14 [95% CI 2.02-2.26]; p < 0.001). Conclusion: Using population-level data among patients hospitalized for delivery in the United States, we found an association between ART and stroke after adjustment for measured confounders.