Effects of ketone supplements on blood β-hydroxybutyrate, glucose and insulin: A systematic review and three-level meta-analysis.
Level 1 - systematic review of randomized trials
Systematic review and three-level meta-analysis of randomized controlled trials
PubMed 37327753 · doi:10.1016/j.ctcp.2023.101774
What was done
Authors conducted a systematic review and three-level meta-analysis of randomized crossover and parallel studies from Medline, Web of Science, Embase, and Cochrane Central through November 2022. They quantified the acute effects of exogenous ketone supplementation versus placebo on blood β-hydroxybutyrate (BHB), glucose, and insulin using Hedges' g. Fractional polynomial regressions were applied to examine dose-response relationships and time effects.
What was found
Based on 327 data points from 30 studies comprising 408 participants, ketone supplementation significantly increased blood BHB (Hedges' g = 1.4994, 95% CI [1.2648, 1.7340]) and decreased blood glucose (Hedges' g = -0.3796, 95% CI [-0.4550, -0.3041]). Insulin rose significantly in healthy non-athletes (Hedges' g = 0.1214, 95% CI [0.0582, 0.3011]), but showed insignificant change in populations with obesity and prediabetes. Dose-response and time-course analyses demonstrated nonlinear dynamics for BHB and insulin at specific time intervals, and dose-dependent linear reductions in glucose past 120 minutes.
Why it matters
This meta-analysis demonstrates that acute exogenous ketone ingestion reliably lowers blood glucose while elevating circulating BHB. Crucially, the glucose-lowering effect occurred without raising insulin levels in individuals with obesity or prediabetes.
Limits
The overall sample size was small relative to the number of studies (averaging ~14 participants per study). The analysis only evaluated acute post-ingestion responses, precluding conclusions about chronic use or long-term clinical outcomes. Differences between specific ketone formulations (e.g., ketone esters versus ketone salts) were not distinguished in the abstract.