Association patterns of ketone bodies with the risk of adverse outcomes according to diabetes status.

Aim: To investigate the associations between ketone bodies (KB) and multiple adverse outcomes including cardiovascular disease (CVD), chronic kidney disease (CKD) and all-cause mortality according to diabetes status.

Methods: This prospective study included 222 824 participants free from CVD and CKD at baseline from the UK Biobank. Total KB including β-hydroxybutyrate, acetoacetate and acetone were measured by nuclear magnetic resonance. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between KB and adverse outcomes among participants with normoglycaemia, prediabetes and type 2 diabetes, respectively.

Results: During a mean follow-up of 14.1 years, 24 088 incident CVD events (including 17 303 coronary heart disease events, 5172 stroke events and 5881 heart failure [HF] events), 8605 CKD events and 15 813 deaths, were documented. Higher total KB significantly increased the risk of HF among participants with normoglycaemia (HR, 1.32 [95% CI, 1.17-1.49], per 10-fold increase in total KB) and prediabetes (1.35 [1.04-1.76]), and increased the risk of CKD among those with normoglycaemia (1.20 [1.09-1.33]). Elevated KB levels were associated with an increased risk of all-cause mortality across the glycaemic spectrum (1.32 [1.23-1.42] for normoglycaemia, 1.45 [1.24-1.71] for prediabetes and 1.47 [1.11-1.94] for diabetes). Moreover, a significant additive interaction between KB and diabetes status was observed on the risk of death (P = .009), with 4.9% of deaths attributed to the interactive effects.

Conclusions: Our study underscored the variation in association patterns between KB and adverse outcomes according to diabetes status and suggested that KB could interact with diabetes status in an additive manner to increase the risk of mortality.