Effectiveness of sodium-glucose co-transporter 2 inhibitors in stroke prevention among patients with heart failure: a propensity-score matched cohort study
J Tang, H He, V Yan, C Wong, E ChanAbstract
Background
Ischemic and haemorrhagic strokes among patients with heart failure were associated with higher risk of mortality and disability. In heart failure management, although randomised clinical trials demonstrated the effectiveness of sodium-glucose co-transporter 2 inhibitors (SGLT2i) in decreasing risk of cardiovascular mortality/heart failure hospitalisation, whether they may be effective in reducing risk of ischemic and haemorrhagic strokes, has been less discussed.
Purpose
To evaluate the effectiveness of SGLT2i on stroke prevention in heart failure management.
Methods
Using an electronic medical record database, this study identified adult patients (aged 18 years or above) (1) with heart failure; (2) who received standard-of-care therapy (concurrent prescription of angiotensin converting enzyme inhibitors/angiotensin-receptor blockers/angiotensin receptor-neprilysin [ACEI/ARB/ARNI], beta-blocker, and mineralocorticoid receptor antagonist [MRA]) between January 1, 2013, and December 31, 2022; and (3) without ischemic/haemorrhagic stroke at baseline (index date for SGLT2i users/pseudo index date of non-users). The index date was defined as the date of SGLT2i initiation for the SGLT2i users and a pseudo index date for the non-users was randomly assigned in the standard-of-care therapy period. Baseline covariates included age, sex, comorbidities (atrial fibrillation/flutter, vascular disease, hypertension, diabetes, neoplasms, obesity, and coronary procedure), medication use (calcium channel blockers, lipid lowering drugs, nitrates, cardiac glycosides, antiplatelets, oral anticoagulants, parenteral anticoagulants, insulins, and antidiabetic drugs), and laboratory test results of lipid profile. Propensity score matching (with a ratio of 1:1) was applied to reduce possible bias from confounding introduced by baseline covariates. Cox proportional hazards regression model was used to measure the association of the use of SGLT2i with incident ischemic and haemorrhagic strokes, respectively.
Results
After propensity score matching, 1,949 SGLT2i users and 1,949 non-users were included, with a standard mean difference (SMD) lower than 0.1 for all baseline covariates. In the 5-year follow-up, the incidence rate was 1.04 and 0.41 per 100 person-years for ischemic and haemorrhagic strokes among the SGLT2i users, respectively. Among the non-users, the incidence rate within the same follow-up was 1.86 and 1.02 per 100 person-years for ischemic and haemorrhagic strokes, respectively. Compared with the non-users, the SGLT2i users had a significantly lower risk of incident ischemic and haemorrhagic strokes, with a hazard ratio (HR) of 0.55 (95% confidence interval [CI]: 0.38, 0.79) and 0.39 (95% CI: 0.22, 0.68), respectively.
Conclusion
SGLT2i was associated with reduced risks of ischemic and haemorrhagic strokes among patients with heart failure, which may enrich stroke prevention strategies in heart failure management.Study designFor image description, please refer to the figure legend and surrounding text.