Abstract
OBJECTIVE: Sodium-glucose cotransporter 2 (SGLT2) inhibitors are guideline-recommended agents for treating heart failure (HF), but the role of metabolomic biomarkers in underlying mechanisms, particularly acylcarnitines, remains unclear. This study examined the associations of acylcarnitines with SGLT2 inhibition and incident HF. METHODS: This subcohort study included 2178 participants from the prospective China Kadoorie Biobank without cardiovascular disease, diabetes or cancer at baseline. Plasma levels of 40 acylcarnitines were quantified using targeted mass spectrometry-based platforms. The impact of genetically predicted random plasma glucose (RPG) via SGLT2 inhibition on acylcarnitines was assessed with Mendelian randomization (MR). The associations of acylcarnitines with HF risk were assessed using Cox proportional hazards models. Acylcarnitines were classified into short-, medium- and long-chain groups and analysed individually or summed as scores. RESULTS: Of the 2178 participants, the mean (SD) age was 53.2 (9.8) years. 13 incident HF cases occurred during a median follow-up of 10.5 years. SGLT2 inhibition was associated with higher levels of acylcarnitines, while higher levels of acylcarnitines were associated with reduced HF risk. An unweighted acylcarnitines score was associated with SGLT2 inhibition (β, 2.04 (0.29, 3.79) SD increase per 1 mmol/L lower genetic RPG via SGLT2 inhibition) and HF risk (HR, 0.97 (0.93, 0.99) per 1-SD higher of the score). Glucokinase activation, another antidiabetic agent used for comparison, showed weaker associations with acylcarnitines. CONCLUSION: MR analysis indicated SGLT2 inhibition showed associations with acylcarnitines, which are also associated with HF risk. Our findings highlighted the potential involvement of acylcarnitines in the mechanisms between SGLT2 inhibitors and HF.