Acute ketone monoester ingestion lowers resting cerebral blood flow: a randomized cross‐over trial

Abstract

Abstract

Exogenous ketone monoester (KME) supplements rapidly increase plasma beta‐hydroxybutyrate (β‐OHB) and may impact cerebral blood flow (CBF). However, it is currently unknown how acute KME ingestion impacts resting CBF and whether differences in KME dose have differential effects on CBF regulation. The purpose of this study was to investigate the effect of two separate KME doses on resting CBF in young adults. On separate days and in a double‐blind, placebo‐controlled, cross‐over design, 20 participants (10 females; aged 23 ± 3 years) ingested either: (1) High‐KME (0.6 g kg⁻¹ β‐OHB); (2) Low‐KME (0.3 g kg⁻¹ β‐OHB); or (3) placebo drink, and quietly rested for 120 min. Global CBF (gCBF) was assessed using duplex ultrasound of the internal carotid and vertebral arteries, and transcranial Doppler ultrasound was used to assess middle cerebral artery blood velocity at baseline, 45 min and 120 min post‐ingestion. End‐tidal CO₂ ($P_{\mathrm{ETC}{\mathrm{O}}_2}$) was measured using a gas analyser. β‐OHB was measured in venous blood. At 45 min post‐ingestion, gCBF was significantly reduced by 10.6% in Low‐KME and by 14.6% in High‐KME compared to baseline. At 120 min, gCBF returned towards baseline in Low‐KME, whereas gCBF was further reduced by 19.1% in High‐KME compared to baseline. KME dose‐dependent reductions in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ may have contributed to these reductions in gCBF following KME ingestion. These novel findings provide a foundational characterization of the impact of KME on resting CBF, which prompts further investigation building on these results to isolate underlying mechanisms and develop dosing protocols to mitigate potential CO₂ disruptions.

Key points

• Beta‐hydroxybutyrate (β‐OHB) is a signalling molecule and β‐OHB infusion increases cerebral blood flow (CBF) in humans.

• Ingestion of higher doses of a ketone monoester (KME) supplement have been shown to lower blood pH and arterial CO₂, which are important regulators of CBF.

• This double‐blind and placebo‐controlled cross‐over study tested the effects of two separate KME doses (Low‐KME and High‐KME) on resting CBF, end‐tidal CO₂ and systemic haemodynamics over a 2 h period post‐ingestion in young adults.

• Low‐KME reduced CBF 45 min post‐ingestion and High‐KME reduced CBF at both 45 and 120 min post‐ingestion, which corresponded with dose‐dependent reductions in end‐tidal CO₂.

• The findings from this trial represent a foundational characterization of the effects of KME dose on resting CBF.

Abstract figure legend This double‐blind, placebo‐controlled, randomized cross‐over trial investigated the effects of ketone monoester (KME) ingestion on resting cerebral blood flow (CBF) in young adults. We demonstrate that KME supplementation lowers CBF in a dose‐dependent manner over 2‐houra post‐ingestion, whereas CBF remained unchanged following placebo ingestion. KME ingestion also lowered end‐tidal CO₂ in a dose‐dependent manner. KME ingestion increased heart rate but did not affect mean arterial pressure. KME, ketone monoester; CBF, cerebral blood flow; ICA, internal carotid artery; VA, vertebral artery.

Introduction

Plasma concentrations of the ketone body beta‐hydroxybutyrate (β‐OHB) are increased during periods of energetic stress such as prolonged exercise, fasting or severe carbohydrate restriction (i.e. ketogenic diets) (Cox & Clarke, 2014). In these scenarios, fatty acids are mobilized from adipose tissue and converted into β‐OHB by the liver and are then transported to peripheral tissues including the brain, heart, and skeletal muscle to support metabolism in the presence of low glucose (Puchalska & Crawford, 2017). Elevated plasma β‐OHB acts as a biological starvation signal that exerts pleiotropic effects throughout the body (Newman & Verdin, 2017), including regulating gene expression (Shimazu et al., 2013), mitigating oxidative stress (Kim et al., 2007) and reducing markers of inflammation (Walsh et al., 2021; Youm et al., 2015). Until recently, appreciably elevating circulating β‐OHB was only achievable via prolonged fasting or marked carbohydrate restriction. The recent development of oral exogenous ketone supplements provides a novel method of inducing rapid nutritional ketosis that is safe for human consumption (Soto‐Mota et al., 2019), representing an opportunity to investigate the effects of β‐OHB per se on human physiology.

Emerging evidence supports the potential beneficial effects of elevated β‐OHB for brain health in humans. Infusion of sodium β‐OHB increases cerebral blood flow (CBF) by 39% in young adults, corresponding to peak blood β‐OHB concentrations of ∼2.16 mm (Hasselbalch et al., 1996). Similarly, sodium β‐OHB infusion increases CBF by 30% in middle‐to‐older aged adults, corresponding to blood β‐OHB concentrations of 5.5 mm (Svart et al., 2018). The mechanisms underlying the rapid modulation of CBF with acute β‐OHB exposure are not well understood, but probably stem from a combination of direct vasodilatory effects (McCarthy, Chakraborty et al., 2021) and modifications in neurovascular coupling responses (Mintun et al., 2004; Vlassenko et al., 2006). Elevated plasma β‐OHB lowers cerebral glucose metabolism by 14–33% without altering the rate of oxygen metabolism (Hasselbalch et al., 1996; Svart et al., 2018) as cerebral metabolism shifts towards greater ketone body use (Mikkelsen et al., 2015; Pan et al., 2000). Changes in cerebral substrate utilization alters the cytosolic redox potential within astrocytes, acting as an intracellular signal to stimulate increased CBF (Mintun et al., 2004; Vlassenko et al., 2006). Indeed, this not only alters the sensitivity of neurovascular coupling responses to metabolic stimuli, but also may be contributing to increases in resting CBF with ketone infusion independent of any stimulation (Mintun et al., 2004; Vlassenko et al., 2006).

Evidence from β‐OHB infusion trials provides compelling evidence that exogenous β‐OHB can increase CBF in humans, and a recent trial demonstrates that 14 days of thrice‐daily ketone monoester (KME) supplementation increases resting CBF by 12% in the fasted, non‐supplemented state in adults with obesity (Walsh, Caldwell et al., 2021). However, it is currently unknown how a single dose of exogenous KME impacts acute CBF regulation, and whether differences in KME dose differentially affect CBF regulation in young healthy adults. Interestingly, recent studies have shown that ingestion of KME supplements can perturb blood pH, which could impact CBF regulation. Indeed, β‐OHB is a weak organic acid and bolus ingestion of higher oral KME doses lowers blood pH (Bone et al., 2024; Dearlove et al., 2019; McClure et al., 2024; Prins et al., 2021). Previous work has shown that disturbances to blood pH following KME ingestion are buffered by the blood bicarbonate buffering system and increased ventilation to restore pH homeostasis (Bone et al., 2024; Dearlove et al., 2019; McCarthy, Bostad et al., 2021). However, hyperventilation from KME ingestion can lower arterial CO₂ ($P_{\mathrm{aC}{\mathrm{O}}_2}$) (Dearlove et al., 2019; McCarthy, Bostad et al., 2021). Alterations in $P_{\mathrm{aC}{\mathrm{O}}_2}$ have a strong influence on CBF regulation, such that reductions in $P_{\mathrm{aC}{\mathrm{O}}_2}$ cause cerebral vasoconstriction and increases in cerebrovascular resistance, which ultimately lowers CBF (Battisti‐Charbonney et al., 2011; Willie et al., 2012). This suggests that different KME doses may exert differential effects on CBF as a result of alterations in $P_{\mathrm{aC}{\mathrm{O}}_2}$ and this warrants rigorous investigation.

The purpose of this study was to investigate the effect of two different KME doses on resting CBF in young adults. We hypothesized that CBF would be increased after ingestion of a relatively lower dose KME (0.3 g kg⁻¹ β‐OHB) and decreased after ingestion of a relatively higher dose KME (0.6 g kg⁻¹ β‐OHB) over a 2 h period compared to ingestion of a placebo drink. This trial is critical to systematically characterize the dose‐dependent effects of a commercially available supplement on resting CBF and establish a foundation of the cerebrovascular responses to acute KME ingestion.

Methods

Ethical approval

This study reports the primary outcomes from a trial registered as ‘Ketone Dose and Cerebral Blood Flow Study (kCBF)’ (ClinicalTrials.gov Identifier NCT06032156). This trial was approved by the Hamilton Integrated Research Ethics Board (Project ID: 15 454) and all participants provided written informed consent prior to commencement of the study. The trial conformed with the standards set by the Declaration of Helsinki and subsequent revisions.

Participants

Participants between the ages of 18–35 years were recruited from McMaster University in Hamilton, ON, Canada, and the greater Hamilton area between September 2023 and April 2024. Exclusion criteria were the presence of obesity (body mass index >30 kg m⁻²), any cardiometabolic disease (i.e. diabetes, hypertension), a history of cardiovascular disease or cardiovascular events requiring hospitalization (i.e. heart attack), a history of concussion(s) with persistent symptoms, or currently following a ketogenic diet or taking ketone supplements. Interested participants who self‐reported being eligible based on the above exclusion criteria were invited for an in‐person familiarization visit. Fig. 1 displays the trial flowchart.

Figure 1. Trial flowchart.

Trial flowchart in the form of the CONSORT flow diagram. Created using biorender.com.

Familiarization and participant characterization

Prior to condition randomization, eligible participants completed a familiarization visit to obtain anthropometric measures of height and body mass, as well as to complete a graded exercise test to exhaustion to measure peak rate of oxygen consumption (${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$). ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ was assessed to characterize participant cardiorespiratory fitness. In addition, resting CBF measurements were obtained using duplex ultrasound (Terason uSmart 3300; Teratech, Burlington, MA, USA) of the internal carotid (ICA) and vertebral arteries (VA) to ensure that viable scans could be obtained before proceeding to the experimental visits.

To measure ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$, participants completed a graded exercise test to volitional exhaustion on a cycling ergometer (Kettler Ergo Race; Kettler, VA, USA). The test consisted of a 3 min warmup at 50 W, followed by increases of 5 W every 10 s until participants were unable to sustain a cadence above 60 revolutions min⁻¹ or volitional exhaustion, upon which the test was terminated. Breath‐by‐breath expired gases and ventilation were recorded using a metabolic cart (Quark CPET Metabolic Cart; COSMED, Rome, Italy) and heart rate (HR) was measured using a chest strap monitor (Polar A3; Polar, Kempele, Finland). Data were averaged over 10 s recording intervals and ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ was determined as the mean of the three highest recording intervals.

Supplement details

The ketone supplement used in the study (Delta G Tactical; TΔS, Oxford, UK) was composed of the KME; [R]‐3‐hydroxybutyl [R]‐3‐hydroxybutyrate. Two KME doses were used in this trial: (1) Low‐KME corresponding to 0.3 g kg⁻¹ β‐OHB; and (2) High‐KME corresponding to 0.6 g kg⁻¹ β‐OHB. The placebo supplement contained 60 ml of the placebo stock and two drops of calorie‐free lemonade flavouring. The placebo stock contained denatonium benzoate (NF; Johnson Mathey, London, UK) and xanthan gum to match the taste and viscosity of the KME. All supplements were dispensed into 60 mL opaque bottles. The remainder of the bottle for Low‐KME and High‐KME was filled with placebo stock and two drops of calorie‐free lemon juice flavouring to reach a total volume of 60 mL. After the supplements were prepared, the bottles were labelled with the participant's ID and visit number by a third‐party researcher who was not involved in data collection to maintain blinding.

Experimental visits

A randomized, placebo‐controlled, double‐blind, counterbalanced, cross‐over design was used to test three conditions: Placebo, Low‐KME (0.3 g kg⁻¹ β‐OHB) and High‐KME (0.6 g kg⁻¹ β‐OHB). Participants were randomly assigned to one of six treatment orders by a third‐party researcher not involved in data collection using a random number generator in a counterbalanced manner. Blinding was maintained until the completion of data collection and analysis.

Participants attended the Brain Exercise Enhancement Laboratory at McMaster University three separate times, separated by at least 72 h to ensure no carry‐over of previous experimental visits. Female participants were asked to report whether they take contraceptives using a menstrual cycle questionnaire. Visits for female participants were conducted during days 0–6 of the follicular phase (or placebo phase for contraceptive users) of their menstrual cycle to control for any potential effects of menstrual cycle phase on the outcomes of this study. Experimental visits occurred between 07.00 h and 14.00 h and at the same time of day for a given individual. Participants were instructed to abstain from consuming alcohol, caffeine, food or drink (excluding water) for at least 8 h prior, and to abstain from vigorous physical activity for at least 24 h prior to each experimental visit. During the first visit, participants completed a dietary recall questionnaire to record their food and drink consumption in the 24 h prior to their first experimental visit. Participants were asked to replicate the same diet in the 24 h prior to each subsequent visit as closely as possible. Participants also filled out the ‘Consensus Sleep Diary’, a questionnaire asking about their sleep duration the night before each visit (Carney et al., 2012), because acute (Elvsåshagen et al., 2019) and chronic (Park et al., 2019) reductions in sleep duration have been associated with reductions in CBF.

After the completion of intake procedures, participants laid supine and an i.v. catheter was inserted into an antecubital vein for serial blood sampling. Following instrumentation, baseline measures of CBF using duplex (Terason uSmart 3300 NexGen; Teratech) and transcranial Doppler ultrasound (Neurovision; Multigon Industries, NY, USA), haemodynamics, end‐tidal gases and blood were obtained over a 5 min collection window. Upon the completion of baseline measures, participants were provided with one of three supplements in an opaque bottle to ingest. The complete ingestion of the drink was considered the 0 min time marker of the protocol, and a stopwatch was started to track elapsed time. Over the subsequent 2 h period, CBF, end‐tidal CO₂ ($P_{\mathrm{ETC}{\mathrm{O}}_2}$), mean arterial pressure (MAP) and HR were collected during 5 min recording windows at 45 and 120 min post‐ingestion. Blood samples were obtained 30, 45, 60 and 120 min post‐ingestion. After the completion of the last measurements, participants completed a questionnaire regarding gastrointestinal symptoms that may have occurred over the course of the protocol after drink consumption. The questionnaire was a 10‐point, 10 cm visual analogue scale [from ‘no (symptom)’ to ‘worst imaginable (symptom)’] where participants were asked to rate their feelings of nausea, urge to vomit, bloating, belching and cramping (Walsh, Neudorf et al., 2021). Figure 2 displays the study protocol for each experimental visit.

Figure 2. Schematic overview of the experimental protocol.

Baseline measures were obtained in the fasting state and subsequent measures were taken over a 120 min period following ingestion of either Placebo, Low‐KME, or High‐KME. MAP, mean arterial pressure; HR, heart rate; CBF, cerebral blood flow; TCD, transcranial Doppler. Created using biorender.com.

Haemodynamic and respiratory measurements

HR and MAP were continuously monitored using a three‐lead electrocardiogram and non‐invasive finger cuff (Human NIBP Nano System, model INL82; AD Instruments, NSW, Australia), respectively. Non‐invasive blood pressure finger cuff values were calibrated using the mean of three blood pressure measurements collected at the brachial artery using an automated brachial cuff sphygmomanometer (OMRON Healthcare Co. Ltd, Kyoto, Japan). $P_{\mathrm{ETC}{\mathrm{O}}_2}$ was also continuously collected at the mouth using a gas analyser (PowerLab model ML795; ADInstruments) and used to approximate $P_{\mathrm{aC}{\mathrm{O}}_2}$. Participants were fitted with a mouthpiece and nose clip before commencing recordings of breath‐by‐breath gas exchange. For all haemodynamic and respiratory outcomes (HR, MAP and $P_{\mathrm{ETC}{\mathrm{O}}_2}$), recording began at each CBF timepoint (baseline, 45 min and 120 min) and data was collected for 5 min. The mean value for each outcome during the 5 min recording windows was used for analysis, which occurred concurrently with the ICA and VA duplex ultrasound scanning periods.

Middle cerebral artery blood velocity

Middle cerebral artery blood velocity (MCAv) was assessed using a 2 MHz transcranial Doppler ultrasound in accordance with localization and signal optimization procedures previously described (Willie et al., 2011). The ultrasound probe was affixed to the participant's right temporal window using a specialized headband. Recordings included in analysis contained complete waveforms for ≥12 consecutive cardiac cycles. The mean value of MCAv during the longest string of consecutive complete waveforms in each 5 min recording window was used for analysis. The between‐day coefficient of variation for MCAv in this trial was 9.8%.

Cerebral blood flow assessment

Vessel diameter and blood velocity of the ICA and VA were measured using a 4–15 MHz multifrequency linear array duplex ultrasound following recommendations for location and procedure standardization (Thomas et al., 2015). Artery vessel diameter and blood velocity were viewed simultaneously using the B‐mode imaging and pulse‐wave mode features, respectively, of the ultrasound machine. ICA blood velocity and vessel diameter were measured ≥1.5 cm from the carotid artery bifurcation to minimize retrograde flow. Vessel diameter and blood velocity of the VA were measured between C4 and C5 or C5 and C6. The recording location within the vessel was individualized and optimized during each participant's first CBF recording to allow for reliable image acquisition. Recordings were almost exclusively taken from the right ICA and VA, unless image quality was poor, then the left ICA and/or VA was used in substitution. This location was unchanged within participants and between trials. The same insonation angle was used for all recordings (60°). Following acquisition of the first ultrasound image, B‐mode gain and dynamic range were unaltered to minimize variability in arterial wall brightness and/or thickness between recordings. ICA and VA diameter and blood velocity were recorded for at least 1 min using a screen capturing software (Camtasia 2022; TechSmith, East Lansing, MI, USA). Viable recordings included ≥12 cardiac cycles. Offline analysis involved the use of custom‐edge detection and wall tracking software (Woodman et al., 2001; BloodFlow analysis, version 6.1, Australia). This analysis software mitigates observer bias by establishing diameter and velocity traces without the need for manual calliper placement by the researcher (Woodman et al., 2001). Mean beat to beat artery flow was sampled at 30 Hz from the selected recording and calculated automatically from these traces (Woodman et al., 2001). The between‐day coefficients of variation for ICA and VA diameter in this trial were 3.2% and 2.7%, respectively. For measurements of flow, the between‐day coefficients of variation were 11.43%, 13.12% and 9.92% for ICA flow, VA flow and gCBF respectively.

Mean arterial pressure (MAP) was calculated as:

$$\mathrm{MAP}=\mathrm{DBP}+\left(\left(\mathrm{SBP}-\mathrm{DBP}\right)/3\right)$$

Blood flow (Q) was calculated as:

$$\begin{array}{ccc}\hfill \mathrm{Q}& =& \left(\mathrm{mean}\mathrm{envelope}\mathrm{blood}\mathrm{velocity}/2\right)\hfill \\ & & \times \left(\pi \times {\left(0.5\times \mathrm{diameter}\right)}^2\right)\times 60\hfill \end{array}$$

Global cerebral blood flow (gCBF) was calculated as:

$$g\mathrm{CBF}=2\times \left({\mathrm{Q}}_{\mathrm{ICA}}+{\mathrm{Q}}_{\mathrm{VA}}\right)$$

Mean shear rate was calculated as:

$$\mathrm{Shear}\mathrm{rate}=4\times \left(\mathrm{mean}\mathrm{envelope}\mathrm{blood}\mathrm{velocity}/\mathrm{diameter}\right)$$

Cerebrovascular conductance (CVC) was calculated as:

$$\mathrm{CVC}={\mathrm{Q}}_{\mathrm{ICA}},{\mathrm{Q}}_{\mathrm{VA}},\mathrm{gCBF}/\mathrm{MAP}$$

Plasma β‐OHB

A point‐of‐care analyser (FreeStyle Libre; Abbott, Abbott Park, IL, USA) was used to measure plasma β‐OHB in venous blood. If venous blood could not be obtained, capillary blood was used in substitution. Venous blood was drawn in K2 EDTA tubes (BD, Franklin Lakes, NJ, USA). For the determination of plasma β‐OHB, venous blood obtained at baseline, 30, 45, 60 and 120 min post‐ingestion was used. β‐OHB area‐under‐the‐curve (AUC) over the 120‐min period was calculated using Prism, version 10.3.0 (GraphPad Software Inc., San Diego, CA, USA). A researcher not involved in data acquisition or analysis recorded venous β‐OHB concentrations to maintain blinding.

Sample size determination

To date, there have been no previous investigations of the acute effects of KME ingestion on resting CBF, nor has there been a characterization of the potential dose–response effects. Previous infusion trials have reported increased gCBF ranging from 30% to 39% (Hasselbalch et al., 1996; Svart et al., 2018), whereas our previous trial found that 14 days of KME ingestion increased gCBF by 12% in the fasted, non‐supplemented state in adults with obesity (Walsh, Caldwell et al., 2021). At the time of trial pre‐registration (6 April 2023), Walsh et al. (2021) was the only study to demonstrate the effects of KME ingestion on CBF and thus we used the changes in CBF observed in this study to inform our sample size estimation. Despite this trial being a short‐term KME intervention and CBF being assessed in the fasting state, we opted to power our study on the conservative observed effects (12% increase in CBF), rather than the observed effects from infusion trials that have reported increases in CBF between 30% and 40% during ketone infusion (Hasselbalch et al., 1996; Svart et al., 2018). Accordingly, we conducted a sample size estimation based on the hypothesis that dose‐dependent differences in gCBF would occur at the measurement time point corresponding to peak plasma β‐OHB, which was determined as 45 min post‐ingestion based on recent pilot work. A sample size estimation in G*Power, version 3.1 (http://www.gpower.hhu.de) revealed that n = 19 provided 80% power at an alpha level of 0.05 with an effect size partial eta squared of 0.23 using a within‐factors one‐way analysis of variance (ANOVA) (three conditions × one time point). The effect size estimation was based on a 12% change in CBF and the associated measurement error, as based on previous published work (Walsh, Caldwell et al., 2021), as well as pilot data from our lab. To preserve power, 20 adults were recruited and randomized (10 male and 10 female).

Statistical analysis

Statistical analyses were performed using Jamovi, version 2.5.7.0 (https://www.jamovi.org) and R Studio, version 4.3.3 (R Core Team, Vienna, Austria). Q–Q plots and Shapiro–Wilk tests were used to assess normality and skewness. Data are presented as the mean ± SD, or 95% confidence interval, unless otherwise noted. Our a priori sample size was determined using a one‐way ANOVA corresponding to peak β‐OHB concentrations following KME ingestion because we expected baseline CBF values to be the same between conditions. However, we observed some day‐to‐day variability in baseline CBF at the group level, and the one‐way ANOVA approach does not allow for baseline to be included as a covariate given the randomized cross‐over design and impact on causal flow. Accordingly, we used a linear mixed‐effects model (GAMLj3 package; https://gamlj.github.io/) with fixed effects of condition (Low‐KME vs. High‐KME vs. Placebo), time (Baseline vs. 45 min vs. 120 min), as well as time by condition interaction, and a subject random effect to assess cerebrovascular and systemic haemodynamic responses across conditions and time. Baseline was included as a covariate in analysis of all cerebrovascular outcomes to account for potential baseline differences between conditions. P < 0.05 was considered statistically significant for all analyses. A Holm post hoc test was used to evaluate significant main effects of time, condition, or time by condition interaction. Differences in β‐OHB AUC between Low‐KME and High‐KME conditions were assessed using a paired samples t test. For exploratory purposes, a correlation matrix was applied between changes in unadjusted cerebrovascular responses and changes in $P_{\mathrm{ETC}{\mathrm{O}}_2}$, as well as plasma β‐OHB, given their potential influence on CBF regulation. Effect sizes are presented as Cohen's d for a paired samples t test and correlation coefficient (r) for correlation analyses.

Results

Participants

Participant characteristics are displayed in Table 1. In total, 10 males and 10 females were randomized into the study. Participants completed all three conditions, except for one male participant who completed two out of three experimental visits as a result of scheduling issues with the third visit. Eight individuals completed Placebo first, six individuals completed Low‐KME first and six individuals completed High‐KME first (Fig. 1).  Subjectively reported sleep duration prior to experimental visits was not different between visits for individual participants (P = 0.977). Similarly, there were no differences between conditions with subjectively reported nausea, urge to vomit, bloating and cramps from drink consumption (Placebo, Low‐KME and High‐KME; all P > 0.05). There was a condition effect for belching (P = 0.034), driven by High‐KME; however, no major adverse reactions to any supplement occurred. Participants correctly predicted the given supplement at a rate of 39%, demonstrating the effectiveness of condition blinding.

Table 1.

Participant characteristics.

	Female (n = 10)	Male (n = 10)	Group (n = 20)
Age (years)	22 ± 2	24 ± 4	23 ± 3
Body mass (kg)	62.5 ± 9.1	74.6 ± 7.2 *	68.5 ± 10.1
BMI (kg m−2)	23.0 ± 2.5	23.9 ± 1.8	23.4 ± 2.2
${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ (mL kg−1 min−1) (range)	37.95 ± 5.85 (32.39–48.89)	50.91 ± 8.71 * (38.1–64.15)	44.43 ± 9.82 (32.39–64.15)
Contraceptive users (%)	50% (5/10)	NA	NA
Time in education (years)	16 ± 2	18 ± 3	17 ± 3

Data are presented as mean ± SD (range) for ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$. BMI, body mass index; ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$, peak rate of oxygen consumption.

^*Significant difference compared to female, P < 0.05. NA, not applicable.

Plasma β‐OHB

Peak plasma β‐OHB in Low‐KME was 3.0 mm and occurred at 45 min post‐ingestion. In High‐KME, peak plasma β‐OHB was 4.7 mm and occurred at 120 min post‐ingestion. β‐OHB AUC for Low‐KME (285 ± 59 mm × 120 min) was significantly lower than the β‐OHB AUC for High‐KME (458 ± 100 mm × 120 min) (P < 0.001) (Fig. 3B ).

Figure 3. Plasma β‐OHB concentrations after acute ketone monoester supplement (KME) ingestion.

(A) Plasma β‐OHB over time and (B) β‐OHB area‐under‐the‐curve (AUC) in low‐KME (0.3 g kg⁻¹ β‐OHB) and High‐KME (0.6 g kg⁻¹ β‐OHB) over the 120 min intervention period. Data are presented as the mean ± SD. Individual data are represented as open triangles and open squares in (B). *Significant difference compared to baseline within condition (P < 0.05). #Significant difference compared to low‐KME (P < 0.05).

Cerebral blood flow and cerebrovascular outcomes

All cerebrovascular measures are presented in Table 2. There was a significant time by condition interaction for gCBF (P < 0.001), with gCBF decreasing from baseline over the 2 h period in both Low‐KME and High‐KME (Fig. 4). Post hoc analysis revealed that gCBF was significantly lower at 45 min compared to baseline in Low‐KME (P = 0.002), corresponding to a 10.6% reduction. Similarly, gCBF was significantly lower at 45 min relative to baseline in High‐KME (P < 0.001), corresponding to a 14.6% reduction. At 120 min post‐ingestion, gCBF was not statistically different in Low‐KME (P = 0.393), whereas gCBF was significantly lower in High‐KME compared to baseline (P < 0.001), corresponding to a 19.1% reduction. Moreover, gCBF was significantly lower in High‐KME compared to both Low‐KME (P < 0.001) and Placebo (P < 0.001) at 120 min post‐ingestion. gCBF remained unchanged over the entire intervention period in the Placebo condition. Additionally, there was a significant time by condition interaction effect for gCBF CVC (P = 0.001).

Table 2.

Cerebrovascular measures.

	Baseline	45 min	120 min
gCBF (mL min−1)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	522 ± 110 (503, 540)	499 ± 104 (480, 518)	520 ± 103 (501, 538)
Low‐KME	521 ± 105 (502, 540)	466 ± 72 * (447, 485)	488 ± 87 (469, 507)
High‐KME	528 ± 125 (510, 547)	451 ± 90 * , ** (433, 470)	427 ± 86 * , ** , *** (408, 446)
ICA flow (mL min−1)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	202 ± 48 (193, 210)	194 ± 46 (185, 203)	204 ± 46 (196, 213)
Low‐KME	201 ± 50 (192, 210)	183 ± 37 (174, 192)	187 ± 37 (178, 196)
High‐KME	206 ± 54 (197, 214)	174 ± 41 * , ** (165, 183)	167 ± 37 * , ** , *** (158, 176)
VA flow (mL min−1)	Time, P < 0.001; Condition, P = 0.011; Interaction, P < 0.001
Placebo	59 ± 21 (56, 62)	53 ± 19 (50, 56)	55 ± 18 (52, 58)
Low‐KME	59 ± 22 (56, 62)	50 ± 18 * (47, 53)	57 ± 24 (54, 60)
High‐KME	59 ± 19 (56, 62)	51 ± 18 * (49, 55)	47 ± 16 * , ** , *** (44, 50)
MCA velocity (cm s−1)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	69 ± 9 (67, 72)	66 ± 9 (63, 68)	69 ± 7 (66, 72)
Low‐KME	70 ± 13 (67, 72)	63 ± 12 * (60, 65)	66 ± 9 (63, 68)
High‐KME	69 ± 14 (67, 72)	63 ± 14 * (61, 66)	57 ± 13 * , ** , *** (55, 60)
ICA velocity (cm s−1)	Time, P < 0.001; Condition, P = 0.002; Interaction, P = 0.040
Placebo	40 ± 8 (38, 41)	39 ± 9 (35, 38)	39 ± 8 (35, 38)
Low‐KME	40 ± 10 (38, 41)	37 ± 7 (35, 38)	36 ± 7 (35, 38)
High‐KME	40 ± 10 (38, 41)	36 ± 8 * (35, 38)	35 ± 8 * , ** (33, 37)
VA velocity (cm s−1)	Time, P < 0.001; Condition, P = 0.380; Interaction, P = 0.036
Placebo	19 ± 4 (18, 20)	17 ± 4 (16, 18)	18 ± 3 (17, 19)
Low‐KME	19 ± 5 (18, 20)	17 ± 4 (16, 18)	19 ± 4 (18, 20)
High‐KME	19 ± 4 (18, 20)	18 ± 5 (17, 19)	16 ± 4 * , *** (16, 19)
ICA diameter (cm)	Time, P = 0.009; Condition, P < 0.001; Interaction, P = 0.007
Placebo	0.464 ± 0.040 (0.458, 0.470)	0.461 ± 0.038 (0.455, 0.467)	0.470 ± 0.039 (0.464, 0.476)
Low‐KME	0.463 ± 0.040 (0.457, 0.469)	0.461 ± 0.041 (0.454, 0.467)	0.467 ± 0.039 (0.461, 0.473)
High‐KME	0.465 ± 0.043 (0.459, 0.472)	0.451 ± 0.040 * (0.445, 0.457)	0.452 ± 0.041 * , ** , *** (0.446, 0.458)
VA diameter (cm)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	0.358 ± 0.043 (0.353, 0.362)	0.357 ± 0.041 (0.352, 0.362)	0.361 ± 0.041 (0.356, 0.366)
Low‐KME	0.358 ± 0.045 (0.353, 0.363)	0.348 ± 0.041 (0.343, 0.353)	0.354 ± 0.047 (0.349, 0.359)
High‐KME	0.358 ± 0.040 (0.353, 0.362)	0.346 ± 0.038 * , ** (0.341, 0.350)	0.342 ± 0.041 * , ** , *** (0.338, 0.347)
ICA shear rate (1/s)	Time, P < 0.001; Condition, P = 0.033; Interaction, P = 0.294
Placebo	345 ± 81 (330, 359)	340 ± 89 (325, 355)	338 ± 80 (323, 353)
Low‐KME	347 ± 100 (332, 362)	323 ± 82 (308, 338)	314 ± 71 (299, 330)
High‐KME	346 ± 94 (331, 360)	324 ± 80 (310, 339)	312 ± 71 (297, 327)
VA shear rate (1 s−1)	Time, P = 0.003; Condition, P = 0.617; Interaction, P = 0.157
Placebo	214 ± 47 (203, 226)	193 ± 43 (182, 205)	200 ± 46 (189, 211)
Low‐KME	212 ± 47 (201, 224)	199 ± 39 (187, 210)	209 ± 42 (198, 221)
High‐KME	213 ± 49 (202, 224)	206 ± 51 (194, 217)	192 ± 46 (181, 203)
gCBF CVC (ml/min/mmHg)	Time, P < 0.001; Condition, P < 0.001; Interaction, P = 0.001
Placebo	6.44 ± 1.55 (6.18, 6.69)	5.87 ± 1.25 * (5.62, 6.12)	6.20 ± 1.29 (5.96, 6.46)
Low‐KME	6.48 ± 1.29 (6.23, 6.74)	5.56 ± 0.86 * (5.31, 5.81)	5.75 ± 1.05 * (5.49, 6.01)
High‐KME	6.56 ± 1.50 (6.31, 6.82)	5.43 ± 1.06 * (5.18, 5.68)	5.23 ± 0.92 * , ** (4.96, 5.50)
ICA CVC (mL min−1 mmHg−1)	Time, P < 0.001; Condition, P = 0.001; Interaction, P = 0.005
Placebo	2.48 ± 0.67 (2.36, 2.59)	2.29 ± 0.53 (2.18, 2.41)	2.43 ± 0.56 (2.31, 2.54)
Low‐KME	2.50 ± 0.64 (2.39, 2.62)	2.16 ± 0.47 * (2.05, 2.28)	2.19 ± 0.50 * (2.07, 2.31)
High‐KME	2.54 ± 0.67 (2.42, 2.65)	2.08 ± 0.47 * (1.96, 2.19)	2.04 ± 0.42 * , ** (1.92, 2.16)
VA CVC (mL min−1 mmHg−1)	Time, P < 0.001; Condition, P = 0.101; Interaction, P = 0.017
Placebo	0.743 ± 0.283 (0.703, 0.783)	0.644 ± 0.253 * (0.604, 0.683)	0.677 ± 0.224 (0.637, 0.716)
Low‐KME	0.740 ± 0.275 (0.700, 0.780)	0.615 ± 0.214 * (0.575, 0.655)	0.685 ± 0.281 (0.644, 0.727)
High‐KME	0.743 ± 0.197 (0.703, 0.783)	0.636 ± 0.191 * (0.596, 0.676)	0.587 ± 0.189 * , ** , *** (0.546, 0.628)
MCAv CVC (cm s−1 mmHg−1)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	0.848 ± 0.140 (0.815, 0.882)	0.759 ± 0.116 * (0.725, 0.793)	0.819 ± 0.069 (0.785, 0.854)
Low‐KME	0.861 ± 0.148 (0.829, 0.893)	0.743 ± 0.116 * (0.709, 0.776)	0.760 ± 0.075 * (0.723, 0.797)
High‐KME	0.843 ± 0.146 (0.810, 0.875)	0.742 ± 0.145 * (0.709, 0.776)	0.689 ± 0.170 * , ** (0.652, 0.726)

Data are the estimated marginal mean ± SD (95% confidence interval). gCBF, global cerebral blood flow; ICA, internal carotid artery; VA, vertebral artery; MCAv, middle cerebral artery blood velocity; CVC, cerebrovascular conductance.

^*Significant difference compared to baseline within condition (P < 0.05).

^**Significant difference compared to Placebo at the same time point (P < 0.05).

^***Significant difference compared to Low‐KME at the same time point (P < 0.05).

Figure 4. Global CBF (gCBF) responses to acute ketone monoester supplement (KME) ingestion.

(A) gCBF responses adjusted for baseline over the 120 min intervention between conditions. Data are presented as the estimated marginal mean ± SD. Unadjusted individual gCBF responses to Placebo (B), Low‐KME (C)and High‐KME (D). Group means are indicated by solid black circles. *Significant difference compared to BSL (baseline) (P < 0.05). #Significant difference compared to Placebo (High‐KME) (P < 0.05). &Significant difference compared to Low‐KME (High‐KME) (P < 0.05).

Extracranial and intracranial artery blood flow responses are displayed in Table 2. High‐KME reduced ICA flow compared to baseline (time × condition interaction; P < 0.001). Similarly, there was a significant time by condition interaction effect observed for ICA CVC (P = 0.005), velocity (P = 0.040) and diameter (P = 0.007). A main effect of condition for ICA shear rate (P = 0.033) was observed; however, post hoc analyses revealed no significant differences between conditions. We also observed no time by condition interaction effect for ICA shear rate (P = 0.294). VA flow was significantly lower in both Low‐ and High‐KME compared to baseline (time × condition interaction; P < 0.001). Additionally, there was a significant time by condition interaction effect observed for VA CVC (P = 0.017), velocity (P = 0.036) and diameter (P < 0.001), but not shear rate (P = 0.157). Finally, Low‐KME and High‐KME ingestion lowered MCAv relative to baseline (time × condition interaction; P < 0.001) and a significant time by condition interaction effect was observed for MCA CVC (P < 0.001). There were no changes in the majority of blood flow outcomes over the 2 h intervention period in Placebo; however, there were statistically significant reductions in gCBF CVC, VA CVC and MCA CVC at 45 min post‐ingestion in Placebo.

Haemodynamic and respiratory responses

Haemodynamic and respiratory measures are displayed in Table 3. Relative to baseline, KME ingestion significantly elevated HR by 8 ± 6 beats min⁻¹ and 11 ± 8 beats min⁻¹ at 45 min post‐ingestion in Low‐ (P < 0.001) and High‐KME (P < 0.001), respectively. HR remained significantly elevated at 120 min post‐ingestion in both Low‐KME (5 ± 5 bpm; P = 0.017) and High‐KME (12 ± 5 bpm; P < 0.001) compared to baseline. HR did not change over the 2 h period in Placebo. Additionally, MAP was unaltered by KME ingestion (time × condition interaction; P = 0.747).

Table 3.

Haemodynamic and respiratory responses.

	Baseline	45 min	120 min
$P_{\mathrm{ETC}{\mathrm{O}}_2}$ (mmHg)	Time, P < 0.001; Condition, P < 0.001; nteraction, P < 0.001
Placebo	38.4 ± 3.7 (37.0, 39.8)	38.3 ± 3.3 (36.9, 39.7)	38.6 ± 3.8 (37.2, 40.0)
Low‐KME	38.5 ± 2.6 (37.1, 39.9)	36.3 ± 2.7 * , ** (34.9, 37.8)	36.8 ± 2.9 * , ** (35.3, 38.2)
High‐KME	38.8 ± 2.6 (37.4, 40.2)	35.4 ± 3.0 * , ** (34.0, 36.8)	34.9 ± 2.7 * , ** , *** (33.5, 36.3)
HR (bpm)	Time, P < 0.001; Condition, P < 0.001; Interaction, P < 0.001
Placebo	61 ± 7 (57, 64)	61 ± 9 (57, 64)	60 ± 6.0 (55, 63)
Low‐KME	61 ± 6 (57, 65)	69 ± 9 * , ** (65, 73)	66 ± 7 * , ** (62, 70)
High‐KME	61 ± 7 (57, 65)	72 ± 12 * , ** (68, 76)	73 ± 9 * , ** , *** (69, 76)
MAP (mmHg)	Time, P < 0.001; Condition, P = 0.080; Interaction, P = 0.747
Placebo	83 ± 9 (79, 86)	86 ± 8 (83, 89)	85 ± 8 (82, 88)
Low‐KME	81 ± 75 (78, 85)	84 ± 6 (81, 87)	84 ± 7 (81, 88)
High‐KME	82 ± 6 (78, 85)	84 ± 6 (80, 87)	84 ± 7 (81, 88)

Data are estimated marginal means ± SD (95% confidence interval). $P_{\mathrm{ETC}{\mathrm{O}}_2}$, end‐tidal carbon dioxide; HR, heart rate; MAP, mean arterial pressure.

^*Significant difference compared to baseline within condition (P < 0.05).

^**Significant difference compared to Placebo at the same time point (P < 0.05).

^***Significant difference compared to Low‐KME at the same time point (P < 0.05).

KME ingestion significantly decreased $P_{\mathrm{ETC}{\mathrm{O}}_2}$ (time × condition interaction; P < 0.001). In Low‐KME, $P_{\mathrm{ETC}{\mathrm{O}}_2}$ decreased by 5.7% (∆ −2.2 ± 1.7 mmHg) at 45 min (P < 0.001) and by 4.4% (∆ −1.7 ± 1.4 mmHg) at 120 min post‐ingestion (P = 0.003) relative to baseline. In High‐KME, $P_{\mathrm{ETC}{\mathrm{O}}_2}$ decreased by 8.8% (∆ −3.5 ± 1.8 mmHg) at 45 min (P < 0.001) and 10.0% (∆ −3.9 ± 1.9 mmHg) at 120 min post‐ingestion compared to baseline (P < 0.001). Furthermore, the decrease in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ in High‐KME was significantly greater than the decrease in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ in Low‐KME at 120 min post‐ingestion (P = 0.001). $P_{\mathrm{ETC}{\mathrm{O}}_2}$ did not change at either 45 min or 120 min post‐ingestion in Placebo (P = 1.0).

Exploratory correlations

The absolute changes in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ relative to the absolute changes in gCBF over the 2 h intervention are displayed in Fig. 5. We explored whether higher plasma β‐OHB was related to changes in $P_{\mathrm{ETC}{\mathrm{O}}_2}$, given that β‐OHB is a weak acid. In support of this, we found a negative correlation between the absolute change in plasma β‐OHB and the percentage change in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ following KME ingestion for pooled data (r = −0.704; P < 0.001) (Fig. 6A ). We also explored whether plasma β‐OHB concentrations were related to gCBF responses following KME ingestion. There was a negative correlation between the absolute change in plasma β‐OHB and the percentage change in gCBF (r = −0.462; P < 0.001) (Fig. 6B ). Correlation analyses separated by condition revealed negative correlations between the absolute change in plasma β‐OHB and the percentage change in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ in Low‐KME (r = −0.541; P < 0.001) and High‐KME (r = −0.632; P < 0.001). Furthermore, there was a negative correlation observed between the absolute change in plasma β‐OHB and percentage change in gCBF in High‐KME (r = −0.534; P < 0.001), but not Low‐KME (r = −0.118; P = 0.495).

Figure 5. Relationship between end‐tidal CO₂ and global CBF (gCBF).

$P_{\mathrm{ETC}{\mathrm{O}}_2}$ plotted against gCBF in Low‐KME (A)and High‐KME (B)over the 2 h intervention. Data are presented as the estimated marginal mean ± SD. BSL, baseline.

Figure 6. Exploratory correlations.

(A) Correlation between absolute change in plasma β‐OHB and percentage change in end‐tidal CO₂ ($P_{\mathrm{ETC}{\mathrm{O}}_2}$) in Low‐KME (blue circles) and High‐KME (red circles).(B) Correlation between absolute change in plasma β‐OHB and percentage change in global CBF (gCBF) in Low‐KME (blue circles) and High‐KME (red circles).

Discussion

The purpose of this trial was to characterize the effect of different doses of an oral KME supplement on resting CBF over a 2 h period in young adults. We hypothesized that ingestion of a lower KME dose would increase CBF, whereas ingestion of a higher KME dose would reduce CBF. The main findings were that: (1) ingestion of Low‐KME and High‐KME reduced CBF at 45 min post‐ingestion; (2) CBF returned towards baseline values with Low‐KME, whereas reductions were sustained with High‐KME at 120 min post‐ingestion; and (3) KME ingestion reduced $P_{\mathrm{ETC}{\mathrm{O}}_2}$ in a manner consistent with β‐OHB dose, which may have contributed to reductions in CBF. Collectively, these findings suggest that acute KME ingestion lowers resting CBF in young adults.

β‐OHB infusion and oral KME supplements can both rapidly induce nutritional ketosis and are purported as nutritional tools to enhance aspects of brain health, including CBF. Indeed, acute elevations in plasma β‐OHB via infusion increases CBF by 30 to 39%, corresponding to peak plasma β‐OHB of 5.55 and 2.19 mm, respectively (Hasselbalch et al., 1996; Svart et al., 2018). However, despite achieving similar β‐OHB concentrations in the present study, there are clear, discrepant effects of KME ingestion on resting CBF regulation compared to β‐OHB infusion (Hasselbalch et al., 1996; Svart et al., 2018). Our findings stand in direct opposition to infusion trials because ingestion of either Low‐KME or High‐KME significantly reduced resting gCBF within a relatively short time frame and was related to changes in plasma β‐OHB. At 45 min post‐ingestion in Low‐KME, gCBF was 10.6% lower compared to baseline, corresponding to plasma β‐OHB of 2.97 mm. Similarly, gCBF was 14.6% lower compared to baseline in High‐KME at 45 min post‐ingestion, corresponding to plasma β‐OHB of 4.07 mm. At 120 min post‐ingestion, gCBF returned closer to baseline values, which was paralleled by a reduction in plasma β‐OHB to 2.37 mm in Low‐KME. By contrast, High‐KME led to further reductions in gCBF (∆ −19.1%) and increases in plasma β‐OHB (4.72 mm) compared to baseline.

An important consideration with ketone infusion trials is the potential influence of sodium overload and haemodilution on CBF regulation. Infusion trials utilize a ketone salt (i.e. sodium β‐OHB), which induces a slight rise in blood pH (Hasselbalch et al., 1996; Miles et al., 1981, 1983). Additionally, the haemodilution from infusion directly reduces O₂ content (C_aO₂) per given mL of blood volume (Todd et al., 1994). Collectively, these two effects would stimulate cerebral vasodilatation and increased CBF independent of β‐OHB (Caldwell et al., 2021; Todd et al., 1994). It is therefore plausible that these factors may be contributing to the discrepancies between our findings and those of previous infusion studies.

Our data support previous findings demonstrating that KME ingestion lowers blood pH and cannot be fully compensated for by blood bicarbonate buffering. Indeed, consumption of both lower (Dearlove et al., 2019; Stubbs et al., 2017) and higher dose KME at rest (Bone et al., 2024; McClure et al., 2024; Prins et al., 2021) significantly lowers blood pH despite similar reductions in blood bicarbonate concentrations. Elevated arterial H⁺ stimulates chemoreceptors, which increase ventilation to reestablish blood pH homeostasis (Tan et al., 2007). KME ingestion has been shown to increase ventilation during submaximal exercise compared to placebo (McCarthy, Bostad et al., 2021), whereas no studies have assessed ventilation at rest over longer periods following KME ingestion (e.g. 2 h or more). We speculate that prolonged elevations in plasma β‐OHB are causing the observed reductions in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ as a result of increased ventilation via sustained chemoreflex activation, as evidenced by the observed strong negative correlation between $P_{\mathrm{ETC}{\mathrm{O}}_2}$ and plasma β‐OHB in this trial. However, given that ventilation and blood pH were not measured in the present study, follow‐up trials are needed to confirm this speculation via direct measurement of blood pH, ventilation and chemoreflex manipulation.

It is well‐established that cerebrovascular tone is sensitive to changes in arterial CO₂ (Battisti‐Charbonney et al., 2011; Willie et al., 2012) because reductions in arterial CO₂ cause cerebral vasoconstriction (Boedtkjer, 2018; Swietach et al., 2013). Past trials in humans demonstrate that manipulating blood pH at the same time as maintaining $P_{\mathrm{ETC}{\mathrm{O}}_2}$ does not impact CBF, suggesting that the observed reductions in $P_{\mathrm{ETC}{\mathrm{O}}_2}$ in this trial may have a stronger influence on CBF regulation than reductions in blood pH from KME ingestion (Lambertsen et al., 1961). However, given that the impact of arterial CO₂ was not experimentally isolated in our current design, future trials are needed to confirm the mechanistic influence of changes in $P_{\mathrm{aC}{\mathrm{O}}_2}$ on reductions in CBF following KME ingestion.

This trial also assessed the impact of KME ingestion on HR and MAP given their influence on CBF regulation (Lucas et al., 2010; Ogoh et al., 2005). We observed that HR increased by 6 beats min⁻¹ in Low‐KME and 11 beats min⁻¹ in High‐KME, on average. This is consistent with a previous study demonstrating that a single KME dose (0.714 g kg⁻¹) increases HR by 9 beats min⁻¹ in young adults (Selvaraj et al., 2022). Mechanistically, β‐OHB has direct signalling effects on cardiac pacemaker cells and acute KME ingestion (0.483 g kg⁻¹) increases cardiac output by 2.0 L min⁻¹, primarily through increases in HR (∼8 beats min⁻¹), in young adults (Oneglia et al., 2023). By contrast, KME ingestion did not impact MAP in this trial, which is consistent with findings from a recent meta‐analysis reporting that acute exogenous ketone supplementation (KME and ketone salts) has no effect on either systolic or diastolic blood pressure (Marcotte‐Chénard et al., 2024). The absence of a change in MAP despite increases in HR following KME ingestion could be explained by β‐OHB decreasing total peripheral resistance (Homilius et al., 2023; Nielsen et al., 2019; Selvaraj et al., 2022). However, any indirect effects of increased HR on CBF regulation would probably be masked by the sustained hypocapnia following KME ingestion in the current trial.

Strengths, limitations and future directions

This double‐blinded, cross‐over, counterbalanced, placebo‐controlled trial addresses crucial knowledge gaps regarding CBF regulation following acute KME ingestion. This trial provides a critical foundational characterization of the relationship between KME dose and CBF regulation at rest. However, this study has limitations worth addressing. Resting CBF was the only measure of cerebrovascular physiology in this study, which does not fully represent functional aspects of cerebrovascular health. Additional studies are needed to uncover how acute KME ingestion impacts cerebrovascular function, such as with cerebrovascular reactivity to CO₂ or visual stimulation (i.e. neurovascular coupling) tasks. Furthermore, it is unclear how CBF is impacted by lower KME doses that do not perturb $P_{\mathrm{aC}{\mathrm{O}}_2}$ and this warrants further investigation. Finally, the observed relationship between $P_{\mathrm{ETC}{\mathrm{O}}_2}$ and CBF in this experimental design does not definitively establish the causal link between changes in $P_{\mathrm{aC}{\mathrm{O}}_2}$ and changes in CBF. Protocols involving end‐tidal clamping of CO₂ or chemoreflex manipulation should be used to corroborate evidence from this trial to quantify the involvement of arterial CO₂ in the mediation of CBF responses to acute KME ingestion. Future studies should also focus on developing strategies to prevent perturbations to acid–base homeostasis with oral KME ingestion, aiming to determine whether the same oral KME doses from this study can increase CBF. Despite these limitations, this rigorous and well‐controlled study design provides confidence in the observed effects from this trial.

Conclusions

In conclusion, acute KME ingestion lowered resting CBF over a 2 h period in young adults, and the responses differed between higher and lower KME doses. Reductions in CBF were probably a result of disruptions to $P_{\mathrm{aC}{\mathrm{O}}_2}$ that were dependent on β‐OHB dose. However, further mechanistic investigations are necessary to consolidate the contribution of $P_{\mathrm{aC}{\mathrm{O}}_2}$ and other regulators of CBF on cerebrovascular responses to KME ingestion. This research advances the field by providing a foundational characterization of the effect of KME dose on resting CBF, providing a platform for future mechanistic work to fully elucidate the impact of KME ingestion on CBF. This is especially important given that KME supplements are commercially available and promoted for their beneficial effects on brain health.

Competing interests

The authors declare that they have no competing interests.

Author contributions

AR, GC, BKA and JW contributed to the conception and design of the work. AR, CY, AO, JN and JB contributed to acquisition of data for the work. AR and JW completed data analysis and interpretation. AR and JW prepared the original manuscript, and AR, CY, GC, AO, JN, JB, BKA and JW contributed to revising the manuscript critically for important intellectual content.

Funding

This trial was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant (#202 103 522; PI: JW) and by a Canadian Foundation for Innovation John R. Evans Leaders Fund (CFI‐JELF) Infrastructure Grant (#41 305; PI: JW). AR was supported by an NSERC Canadian Graduate Student Master's Award and the Brain Canada Foundation Dr. Hubert van Tol Travel Fellowship. The Dr Hubert van Tol Travel Fellowship has been made possible by the Canadian Brain Research Fund (CBRF), an innovative arrangement between the Government of Canada (through Health Canada) and Brain Canada Foundation, and by the Seger‐van Tol family. The views expressed herein do not necessarily represent the views of the Minister of Health or the Government of Canada.

Supporting information

Peer Review History

Biography

Aedan Rourke is a PhD student working under the supervision of Dr. Jeremy Walsh in the Department of Kinesiology at McMaster University. This present work was conducted as a part of his MSc degree in Kinesiology from McMaster. His current research focuses on investigating mechanisms of cerebral blood flow regulation and cerebral metabolism in both healthy adults and neurodegenerative disease. Aedan's overarching future professional goal is to establish a research career developing lifestyle‐based interventions to protect brain health and function across the lifespan.

Data availability statement

The data that support the findings of this trial are available from the corresponding author upon reasonable request.