An exogenous ketone supplement — an orally-ingested ketone monoester or ketone salt — is a distinct exposure from the ketogenic diet (see Ketogenic Diet and Endurance Performance) and from carbohydrate restriction (see Carbohydrate Restriction and Type 2 Diabetes Remission): it raises blood beta-hydroxybutyrate (BHB) acutely «without prolonged periods of fasting or adoption of a ketogenic diet» (Falkenhain et al., 2022). The single held source is a gold SR/MA (43 trials, 586 participants, 58 comparisons, published 2016-2021) — the first systematic evaluation of what these supplements do to blood glucose. It is read here against the wiki’s own four-evidence-states discipline and the surrogate-vs-outcome rule, not at the source’s own enthusiasm (see The source caveat below).
The decision this changes. For someone considering a ketone supplement for glucose control / metabolic health, the honest state is: a robustly demonstrated movement of a surrogate (acute glucose lowering of about 0.5 mM) sitting on top of an unbridged transmission gap to every patient-important outcome, with the evidence drawn almost entirely from the wrong stratum (healthy, mostly young athletic men) and over the wrong horizon (single acute doses). Sorted by evidence state:
| Claim | State | Basis |
|---|---|---|
| Acute BHB elevation after ingestion | established (the exposure works) | within-group +1.73 mM; vs placebo +1.98 mM; monoester ~5x the salt |
| Acute blood-glucose lowering (a surrogate) | benefit — on the surrogate only | vs placebo -0.47 mM, I2=0% (consistent); within-group -0.54 mM (I2=94%) |
| Durable glycaemic control (HbA1c, mean daily glucose) | insufficient evidence | only 5 prolonged trials; fasting glucose unchanged; CGM/HbA1c signals in tiny studies |
| Any patient-important outcome (T2D incidence/progression, CV, mortality, function) | insufficient evidence — not measured | no trial reports one; the surrogate->outcome link is untested |
| GI harm (bloating, diarrhoea, nausea) | harm — dose/form-dependent | ~85% of salt comparisons vs ~40% of monoester comparisons reported AEs |
(inferred from Falkenhain et al., 2022)
The established surrogate movement — acute glucose lowering
The robust finding is the between-condition (vs a matched comparator) result: pooled glucose MD = -0.47 mM (95% CI: -0.57, -0.36; P < 0.001) with low, non-significant heterogeneity (I2 = 0%), a prediction interval of -0.90 to -0.04 mM (Falkenhain et al., 2022). The within-group (after vs before ingestion, fasted) estimate is larger, -0.54 mM (95% CI: -0.68, -0.40), but carries very high heterogeneity (I2 = 94%) (Falkenhain et al., 2022). In magnitude terms, -0.5 mM is about -9 mg/dL — a modest acute dip.
Form dominates dose. «ketone monoesters increased blood BHB approximately 5 times higher and lowered glucose approximately 3 times more than ketone salts» (Falkenhain et al., 2022): between-condition glucose -0.61 mM for monoesters vs -0.24 mM for salts (Table 2). What tracks the glucose fall is the achieved BHB concentration (a significant study-level moderator, average-BHB-vs-glucose R2 = 59.2%), not the administered dose — supplement dosage did not significantly moderate the glucose effect in meta-regression (P = 0.86 monoester, P = 0.27 salt), though a coarse low-vs-high split did show a larger fall at higher doses (low-dose MD -0.33 mM vs high-dose -0.63 mM across all comparisons, P < 0.01). So the dose-response curve is not cleanly characterised: the lever is the BHB level reached, and the supplement form is the main determinant of that.
Why the surrogate does not carry the decision
Blood glucose is a surrogate, and the transmission to a patient-important outcome is untested here. Per Surrogate Outcomes, a target is legitimate only if its causal transmission to a named patient-important outcome is itself evidenced — never assumed. Falkenhain measures the surrogate and is explicit that the outcome link is not established: «more research is needed to determine if supplementing with exogenous ketones is a viable therapeutic option for improving glycemic control» (Falkenhain et al., 2022). Three gaps make the acute surrogate finding decision-inert as it stands.
Wrong stratum. The population is overwhelmingly healthy-weight (and largely young, male, athletic): «Fifty-one comparisons were conducted with participants at a healthy weight, 3 with participants with overweight or obesity, 2 with participants with prediabetes, 1 with participants with T2D, and 1 with participants with heart failure» (Falkenhain et al., 2022). The stratum where lowering glucose could plausibly matter (prediabetes / T2D) contributes 3 of 58 comparisons — a transportability hole, not a finding.
Wrong horizon. «The majority of evidence was obtained in acute, single-dose studies conducted in controlled laboratory-based environments, and there was limited evidence on prolonged supplementation with exogenous ketones, limiting our ability to draw conclusions about longer-term or clinically relevant effects» (Falkenhain et al., 2022). Of the 5 prolonged (>1 day) trials: «no differences in fasting glucose following a period of exogenous ketone supplementation were found, but trials assessing glucose via continuous or intermittent scanning glucose monitoring found statistically significant decreases in mean daily glucose during thrice-daily supplementation with ketone monoesters, which was also reflected in statistically significant decreases in HbA1c in 2 studies and a decrease in fructosamine in 1 study» (Falkenhain et al., 2022). That is an insufficient-evidence signal on a downstream surrogate (HbA1c), not a patient-important-outcome result.
Net effect vs intended effect. The glucose-lowering «does not appear to be fully mediated by an increase in insulin secretion, as infusion of ketones in the absence of insulin secretory function lowers blood glucose as well»; the proposed mechanisms include suppressed hepatic glucose output and BHB inhibiting lipolysis via GPR109A/HCAR2 (Falkenhain et al., 2022).
Harm, cost, and the realistic substitution
GI distress is the consistent harm, and it is form-dependent — «Of all comparisons using a ketone salt and assessing adverse events, approximately 85% of comparisons reported adverse side effects, in contrast to approximately 40% comparisons using a ketone monoester»; 3 participants across all studies withdrew for supplement-related side effects (Falkenhain et al., 2022). Salts additionally deliver a sodium/calcium/potassium load.
The source caveat
Read the source’s optimistic framing against its own limitations. The between-condition glucose result
is genuinely consistent (I2 = 0%), but «Across analyses, we observed a large degree of heterogeneity
alongside wide prediction intervals … This suggests uncertainty around the effect estimate, which
could indicate that certain populations may not experience the observed (BHB-raising and
glucose-lowering) effect» (Falkenhain et al., 2022). And
industry entanglement is pervasive: «many of the included studies may have had potential for a conflict
of interest (e.g., due to affiliation of authors to the ketone industry or provision of ketone
supplement free of charge to study authors)» (Falkenhain et al., 2022) — the senior author also discloses founder shares in a metabolic-monitoring company.
confidence: low — single gold SR; the patient-important verdict is insufficient evidence, and the
surrogate result, though robust, is drawn from a population and horizon that do not transport to the
decision.
Open threads (gaps)
- G — needs a prolonged RCT in the hyperglycaemic stratum on HbA1c and, ideally, a patient-important endpoint: does the acute glucose dip compound into durable glycaemic improvement in prediabetes/T2D, or is it an acute fuel-substitution artefact? No gold source on the outcome link is held (directional-only lever; no specific paper identified).
- G — the ketogenic-diet vs ketone-supplement contrast on the same glycaemic target is not yet a filed comparison; both raise BHB by different routes with different adherence and net-effect profiles.