Does bariatric surgery change the risk of acute ischemic stroke in patients with a history of transient ischemic attack? A nationwide analysis

Background: Stroke is the second leading cause of death worldwide and fifth in the United States, and it represents the major cause of disability in older adults.

Objective: We aimed to determine the risk of acute ischemic stroke (AIS) in individuals with obesity with a history of transient ischemic attack (TIA) compared with patients with a history of bariatric surgery.

Setting: Academic hospital, United States.

Methods: Using the Nationwide Inpatient Sample (NIS) database from 2010 to 2015, we retrospectively identified patients with obesity and past medical history of TIA and divided them into 2 groups: a treatment group of patients who underwent bariatric surgery, and a control group of patients with obesity. We compared incidence of new AIS in both groups using a univariate analysis and multivariate regression model. Covariates included were lifestyle (smoking status, alcohol habits, cocaine use), family history of stroke, co-morbidities (diabetes, hypertension, hyperlipidemia, atrial fibrillation) and long-term medical treatment (antiplatelet/antithrombotic treatment).

Results: A total of 91,640 patients met inclusion criteria, of which treatment patients were 12.3% (n = 11,284) and control patients 87.6% (n = 80,356). The average age of the treatment group was 62.9 ± 17.08 years, and the average of the control was 59.6 ± 12.74 years. The rate of AIS in the treatment group was significantly lower compared with the control group (2.8% versus 4.2%, P < .0001). After adjusting for covariables, the risk difference of AIS was still significant between groups (odds ratio = 1.33, P < .0001), showing that patients in the treatment group were less likely to have AIS compared with the control group.

Conclusions: After analyzing nationwide information, we conclude bariatric surgery helps decrease risk of AIS in patients with a history of TIA. However, this comparison is limited by the nature of the database; further studies are needed to better understand these results.