The therapeutic potential of ketones in cardiometabolic disease: impact on heart and skeletal muscle

Abstract

β-Hydroxybutyrate (βOHB) is the major ketone in the body, and it is recognized as a metabolic energy source and an important signaling molecule. While ketone oxidation is essential in the brain during prolonged fasting/starvation, other organs such as skeletal muscle and the heart also use ketones as metabolic substrates. Additionally, βOHB-mediated molecular signaling events occur in heart and skeletal muscle cells, and via metabolism and/or signaling, ketones may contribute to optimal skeletal muscle health and cardiac function. Of importance, when the use of ketones for ATP production and/or as signaling molecules becomes disturbed in the presence of underlying obesity, type 2 diabetes, and/or cardiovascular diseases, these changes may contribute to cardiometabolic disease. As a result of these disturbances in cardiometabolic disease, multiple approaches have been used to elevate circulating ketones with the goal of optimizing either ketone metabolism or ketone-mediated signaling. These approaches have produced significant improvements in heart and skeletal muscle during cardiometabolic disease with a wide range of benefits that include improved metabolism, weight loss, better glycemic control, improved cardiac and vascular function, as well as reduced inflammation and oxidative stress. Herein, we present the evidence that indicates that ketone therapy could be used as an approach to help treat cardiometabolic diseases by targeting cardiac and skeletal muscles.

INTRODUCTION

Ketone bodies (herein referred to as ketones) were first identified in the late 19th century in the urine of patients in diabetic coma, which led to the field viewing ketones as unwanted products of incomplete fat oxidation (1, 2). However, it was not until the 1960s that pivotal studies in starved human participants demonstrated that ketones become the predominant fuel supporting brain ATP production, illustrating that ketones have a major physiological purpose (3). Indeed, the use of ketones as an alternative fuel source for the brain during times of limited carbohydrate availability is essential given the fact that the brain cannot oxidize fatty acids.

While ketones can be synthesized by amino acids such as leucine, ketones are primarily synthesized in the liver as breakdown products of fatty acid oxidation-derived acetyl CoA and consist of β-hydroxybutyrate (βOHB), acetoacetate (AcAc), and acetone. The major regulatory enzyme of ketone synthesis is 3-hydroxymethyglutaryl-CoA (HMG-CoA) synthase (HMGCS2) and is thus necessary for ketogenesis (Fig. 1). Once synthesized, the major ketones βOHB and AcAc are either transported out of hepatocytes via monocarboxylic acid transporters, or passively diffuse through hepatocyte membranes into the circulation to be taken up by extrahepatic organs to serve as a fuel source for oxidative ATP production. When βOHB and AcAc are imported into extrahepatic cells of the body, βOHB dehydrogenase 1 (BDH1) converts d-βOHB to AcAc while reducing NAD⁺ to NADH, following which succinyl-CoA:3-ketoacid CoA transferase (SCOT) will transfer the CoA moiety from succinyl-CoA to AcAc to generate acetoacetyl-CoA (Fig. 1). Finally, acetyl-CoA acetyltransferase (ACAT) converts 1 mol of acetoacetyl-CoA into 2 mol of acetyl-CoA that can then enter the Krebs cycle to support ATP production via the generation of reducing equivalents (i.e., NADH) for the electron transport chain. More comprehensive and detailed discussions on ketone metabolism have been extensively reviewed elsewhere (1, 4–6).

Figure 1.

Overview of ketone metabolism. Hepatic fatty acid β-oxidation during fasting/starvation increases acetyl-CoA supply to support ketogenesis. The synthesized ketones, specifically AcAc and βOHB, are transported to the mitochondria of extrahepatic tissues (i.e. skeletal muscle, heart), where they are oxidized back into acetyl-CoA to fuel the Krebs cycle and subsequent oxidative ATP synthesis. AcAc, acetoacetate; Ac-CoA, acetyl CoA; ACAT, acetoacetyl CoA transferase; AcAc-CoA, acetoacetyl CoA; BDH1, β-hydroxybutyrate dehydrogenase 1; βOHB, β-hydroxybutyrate; FFA, free fatty acids; HMG-CoA, hydroxymethylglutaryl CoA; HMGCL, hydroxymethylglutaryl CoA lyase; HMGCS2, hydroxymethylglutaryl CoA synthase; SCOT, succinyl-CoA:3-ketoacid CoA transferase.

As mentioned, ketone oxidation occurs in extrahepatic organs, with the organs with high metabolic demands typically using more ketones as a source of ATP. These organs include brain, skeletal muscle, and the heart. Indeed, the skeletal muscle comprises nearly 40% of an individual’s total body mass (7) and is thus a major consumer of many metabolic substrates, including ketones. Additionally, the heart is the most metabolically demanding organ on a per gram basis (8), thus, can also become a major consumer of ketones particularly when ketone levels are elevated. Accordingly, the molecular and physiological control of ketone metabolism in skeletal muscle and the heart is a key determinant of optimal skeletal muscle health and cardiac function. Furthermore, if the complex regulation of ketone metabolism becomes disturbed, such as in the presence of underlying obesity, these changes may contribute to cardiometabolic diseases [i.e., type 2 diabetes (T2D) and heart failure (HF)] where metabolic disturbances are often hallmarks of these conditions (8, 9).

Although ketones are generally viewed as a fuel source, it is becoming increasingly recognized that ketones have metabolism-independent functions. Indeed, ketones influence several aspects of cellular signaling and gene expression, which subsequently regulate inflammation, oxidative stress, and even appetite (1, 5). Importantly, all of these biological actions are proposed to be key mediators of cardiometabolic disease. Thus, based on the crucial role that ketones play in contributing to overall health, herein we will review how cardiac and skeletal muscle may be therapeutically targeted by ketone supplementation to modify T2D and cardiovascular disease (CVD) outcomes. We will also discuss the differences between the metabolic effects of ketones and the signaling actions of ketones, while presenting evidence about the potential for ketone therapy to alleviate cardiometabolic disease. In most instances, this review will focus on the effects of βOHB and not AcAc as a therapeutic tool to alter signaling and metabolism.

THE SIGNALING EFFECTS OF KETONES

While ketones are an important fuel source, their numerous nonmetabolic signaling effects have been reported within the last 2 decades and are also now well recognized. For instance, βOHB can regulate signaling via serving as a ligand for G protein-coupled receptors (GPCRs), which includes activating GPR109a to both inhibit adipocyte lipolysis (10) and mitigate atherogenesis (11), and it also inhibits GPR41 to reduce sympathetic nervous system activity (12). AcAc can also act on GPCRs, and their distinct signaling differences have been reviewed elsewhere (4, 5).

In addition to GPCR signaling, ketones can also regulate epigenetic modifications. Specifically, βOHB directly inhibits class I histone deacetylases (HDACs), thereby increasing resistance to reactive oxygen species (ROS) via upregulating superoxide dismutase 2 (SOD2) and catalase (13). Furthermore, βOHB-mediated inhibition of class I HDACs also manifests in protective cardiac, spinal, and hepatic effects in healthy and disease models such as diabetes, spinal cord injury, and nonalcoholic (metabolic dysfunction-associated) fatty liver disease (14–17). In other cases, such as the retina and diabetic kidney, βOHB can also confer similar protection against oxidative stress by modulating upstream effectors, such as activating the transcription factor nuclear factor erythroid 2–related factor 2 (Nrf2), which results in HDAC inhibition and subsequently induces antioxidant enzymes such as NAD(P)H-quinone oxidoreductase 1, heme oxygenase 1, and glutathione S-transferases (18–20). While these findings connect βOHB-mediated HDAC inhibition to defense against oxidative stress, HDAC inhibition may also have other favorable effects in cardiometabolic disease, such as T2D and HF. Other epigenetic modifications via βOHB that are now recognized include β-hydroxybutyrylation, a primarily concentration-dependent posttranslational modification. For context, β-hydroxybutyrylation is 10- to 40-fold greater than the ∼2-fold increase in acetylation when assessed in vitro as well as in the livers of fasted or diabetic mice (21). Since β-hydroxybutyrylation competes with acetylation binding sites, the beneficial effects of ketone therapy may also arise from reducing the hyperacetylation observed in cardiometabolic disease (22). However, the impact of these signaling mechanisms has yet to be fully explored.

A rapidly growing area of interest in ketone signaling is the ability of ketones to modulate inflammation by regulating inflammatory cytokine production and subsequent signaling cascades. For instance, βOHB can potently inhibit the nucleotide-binding domain-like receptor protein 3 (NLRP3) inflammasome activation independent of effects on oxidative metabolism, energy status, glycolysis, GPR109a, HDAC inhibition, ROS, or autophagy (23). While these other mechanisms are not necessary for βOHB to inhibit NLRP3, these additional mechanisms can still modulate NLRP3 inhibition in different disease states and therefore should not be excluded when considering the mechanisms by which ketones exert benefit (11, 24–27). By inhibiting the NLRP3 inflammasome, βOHB decreases interleukin (IL)-1β production and secretion (23). This is particularly important as IL-1β plays a role in the pathogenesis and progression of cardiometabolic diseases (28–31). Indeed, reducing IL-1β production or inhibiting IL-1β and its downstream signals has been shown to improve outcomes in preclinical and clinical studies investigating cardiometabolic disease (28–31). Although ketones themselves can initiate signaling cascades, the effects of ketone signaling on inflammation are not necessarily independent of ketone metabolism, as evidenced by the finding that ketone metabolism is necessary for optimal CD8⁺ T-cell and macrophage function (32, 33). Therefore, it is likely that both ketone metabolism and ketone signaling are important for reducing inflammation and inflammatory cytokine production.

Taken together, ketones have a wide array of signaling effects that can act on several GPCRs, regulate epigenetic modifications, enhance the response to oxidative stress, and inhibit NLRP3-mediated inflammation (Fig. 2). While many of these effects can occur independent of ketone metabolism, ketone metabolism is important in various cell populations and is likely needed to sustain the beneficial effects of ketones in these specific cell types. Moreover, it should be noted that βOHB exists as either a d-βOHB or l-βOHB enantiomer, and while both share signaling actions in many cases, they may also have differing signaling effects as the target proteins may be stereoselective. The general differences between these enantiomers have already been described (34), and the differences in enantiomer-specific effects on cardiometabolic diseases are only beginning to be defined. Nonetheless, the majority of βOHB in mammalian systems exist as D-βOHB, whereas l-βOHB is not normally found in high levels in the systemic circulation (34).

Figure 2.

The cellular signaling effects of ketones. Ketones regulate an array of cellular signaling pathways. They suppress lipolysis, atherogenesis, and sympathetic activity through the activation of GPCRs, specifically GPR109a and GPR41. Additionally, ketones facilitate vasodilation via potassium channels, modulate chromatin architecture to enhance the expression of various antioxidant genes, promote autophagy, and reduce inflammation by inhibiting the NLRP3 inflammasome. AC, acetylation; BHB, β-hydroxybutyrylation; CAT, catalase; GPCRs, G protein-coupled receptors; HDAC, class I histone deacetylases; GST, glutathione S-transferases; HMOX-1, heme oxygenase 1; NLRP3, nucleotide-binding domain-like receptor protein 3; Nrf2, nuclear factor erythroid 2–related factor 2; ROS, reactive oxygen species; SOD2, superoxide dismutase 2.

ROLE OF KETONES IN OBESITY AND TYPE 2 DIABETES

Ketones and Ketogenic Diets

Because insulin inhibits both adipose tissue lipolysis and fatty acid oxidation rates, insulin is a key indirect regulator of hepatic ketogenesis (1, 2). Based on this, dysregulated ketone metabolism is often viewed as a readout of impaired insulin action and worsening diabetes. Indeed, a population-based study of 9,398 Finnish nondiabetic men and men with newly diagnosed T2D [metabolic syndrome in men (METSIM)] observed that increases in fasting blood glucose levels were positively associated with elevated levels of βOHB and AcAc (35). In addition, follow-up examinations over 5 yr in the nondiabetic men from METSIM revealed that elevations in βOHB and AcAc predicted worsening of oral glucose tolerance and a new diagnosis of T2D, further highlighting the important link between circulating ketone levels and T2D. More recently, observations from the Prevention of REnal and Vascular ENd-stage Disease (PREVEND) longitudinal study of 3,307 participants from northern Netherlands also reported that increases in fasting ketone levels were positively associated with incident T2D over 7.3 yr (36). Thus, because elevated ketone levels are correlated with T2D, the concept of elevating ketone levels as a therapeutic approach to treating T2D has not received much attention.

Based in part on the studies described above, it is now becoming increasingly recognized that changes in ketone metabolism may directly contribute to the pathogenesis of obesity and/or T2D. However, several studies have investigated the impact of very high-fat and low-carbohydrate ketogenic diets on weight loss in overweight/obese individuals and the regulation of glycemia. A meta-analysis of individuals adhering to a ketogenic diet for 1–2 yr reported significant decreases in body weight, although the overall mean weight loss was 0.91 kg and was not significant in the studies that followed their participants over 2 yr (37). Thus it appears as though ketogenic diets are more effective in shorter timeframes (38, 39). Several hypotheses have been put forward to explain why such a dietary strategy may be beneficial. As the carbohydrate-insulin model of obesity argues that elevations in postprandial insulin promote body fat accumulation due to the anabolic actions of insulin (40), the low-carbohydrate nature of ketogenic diets would counteract this. Furthermore, the low basal circulating insulin levels associated with adherence to a ketogenic diet would result in elevated rates of adipose tissue lipolysis.

While ketogenic diet-induced ketosis has also been shown to promote satiety and decrease appetite (41), a recent study in normal-weight subjects demonstrated that consumption of exogenous ketone esters (KEs) to induce ketosis may suppress appetite by lowering circulating ghrelin levels (42). In addition to potential weight-loss actions, the low-carbohydrate nature of ketogenic diets is a major contributor to reduced glycemia (even in the absence of weight loss), which leads to less stimulation of islet β-cells to constantly produce and secrete insulin, thereby indirectly improving β-cell function (43). Notwithstanding this, in some preclinical and clinical studies, ketogenic diets induce variable responses. For instance, some studies demonstrated reductions in glycemia and body weight (38, 44–46), whereas others have reported no improvements in these parameters (47, 48). Nevertheless, despite the positive correlation between elevated circulating ketone concentrations and T2D, elevating ketones may benefit patients with T2D by stimulating weight loss and/or improving glycemic control.

Ketone Metabolism and Signaling in Skeletal Muscle

Although many beneficial effects of a ketogenic diet are mediated via weight loss and improvements in glucose handling, there is growing evidence that alterations in skeletal muscle ketone oxidation may also contribute to insulin resistance/T2D. For example, a study of 47 lean and 47 age-matched obese women reported that rectus abdominus muscle biopsy homogenates had decreased βOHB oxidation rates in obese women (49), demonstrating perturbations in muscle ketone metabolism in obesity. Although this may suggest that impaired muscle ketone oxidation contributes to obesity-related insulin resistance, this decrease was related to a marked reduction in mitochondrial content and/or function, as citrate synthase activity was reduced by ∼50%. In contrast, incubation of isolated mouse soleus muscles with 5 mM βOHB decreased insulin-stimulated Akt phosphorylation and 2-deoxyglucose uptake (50). Moreover, the mRNA/protein expression and enzyme activity of SCOT are increased in gastrocnemius muscles from male mice subjected to high-fat diet-induced obesity (51). This would suggest that elevations in muscle ketone oxidation may actually contribute to obesity-related insulin resistance/T2D. In support of this, pharmacological inhibition of SCOT improved glycemia in obese male mice (51, 52). Similarly, skeletal muscle-specific deletion of SCOT in mice also resulted in a robust improvement in glycemia and glucose tolerance in response to experimental obesity or T2D (51).

Mechanisms explaining why decreasing skeletal muscle ketone oxidation would protect against insulin resistance/T2D remain elusive but may simply arise from substrate competition, whereby restricting the use of ketones as fuel leads to a corresponding increase in carbohydrate utilization. In support of this, pharmacological inhibition of SCOT failed to improve glycemia in obese mice with a genetic modification selectively preventing carbohydrate oxidation in skeletal muscle (51). While this may explain why decreasing skeletal muscle βOHB oxidation can alleviate insulin resistance/T2D, contrasting findings also suggest that increases in ketones improve insulin sensitivity via other mechanisms. By interacting with GPR109A on adipocytes and inhibiting lipolysis, βOHB reduces the release of free fatty acids, thereby decreasing hepatic gluconeogenesis and ameliorating the fatty acid-induced antagonism of insulin-mediated glucose uptake. These effects ultimately result in enhanced insulin sensitivity (10). Concurrently, ketones exert an anorectic effect that can pivot the energy balance toward a deficit, causing weight loss and enhancing peripheral insulin sensitivity (42).

Because SCOT inhibition would also increase the intracellular accumulation of AcAc and βOHB, as well as circulating ketone levels, it is possible that the signaling aspects of ketones described in the signaling effects of ketones play a role in regulating glycemia. As previously stated, increases in βOHB have been linked to inhibition of class I HDACs (13), and pharmacological inhibition of class I HDACs has been reported to improve glucose homeostasis in genetically diabetic mice, which was associated with increases in skeletal muscle oxidative metabolism (53). However, it remains to be determined whether inhibiting ketone metabolism per se decreases HDAC signaling specifically in the skeletal muscle.

In contrast to the studies that show that restricting skeletal muscle ketone utilization improves glycemia in obesity/T2D, increases in circulating ketones secondary to treatment with KEs also appear to be beneficial. For example, KE treatment has been shown to improve glycemia in mice and humans with obesity/T2D (54–57). While KEs can acutely stimulate insulin secretion (57–59), it was observed in isolated perifused islets of lean and obese mice that the effect of ketones on insulin secretion was mild at best (60). Despite these mild actions on insulin secretion, acute elevations in circulating ketone levels following oral KE ingestion improved glucose tolerance in both obese mice and obese mice with skeletal muscle-specific deletion of SCOT (60). The latter illustrates that KE-mediated glucose lowering is independent of muscle ketone oxidation, suggesting that the signaling actions of ketones outweigh any potential benefit attributed to a reduction in their oxidation (51, 52).

Of relevance to skeletal muscle health, ketones also have an anticatabolic effect in muscle, which is mediated by signaling pathways such as Akt/FoxO3a and mammalian target of rapamycin/eIF4E-binding protein-1 (mTOR/4E-BP1) (61). The anabolic potential of βOHB was demonstrated in studies where its infusion in healthy men led to a decrease in leucine oxidation and an approximate 10% increase in muscle protein synthesis (62). The previously mentioned mild actions on insulin secretion would also indirectly increase muscle protein synthesis. In addition, βOHB significantly diminishes phenylalanine efflux from the muscle by up to 70% in the setting of lipopolysaccharide-induced muscle inflammation (63, 64). It is noteworthy that the anticatabolic effects of βOHB are minimally increased by hyperinsulinemia, suggesting that its anabolic role is independent of elevated insulin levels. In T2D where the anabolic efficacy of insulin on skeletal muscle is impaired (anabolic resistance) and there is an attenuated rate of muscle protein synthesis and elevated muscle protein breakdown, ketone therapy may represent a novel strategy to improve muscle mass/quality and insulin sensitivity.

Another important aspect of skeletal muscle ketone metabolism that needs to be considered relates to observations that increases in ketone oxidation secondary to acute elevations in circulating ketones have been linked to increases in exercise performance in humans (65). Mechanisms that may explain these improvements in performance may be linked to increases in muscle mitochondrial function and oxygen consumption (66), whereas other studies have reported increases in muscle glucose uptake and glycogen synthesis (67), both of which would benefit overall insulin sensitivity. Nonetheless, the impact of ketones on performance appears to be linked not only to the timing of the ketone intake but also to the specific nature and length of the physical activity [reviewed elsewhere (68)]. While the existing literature mostly focuses on healthy or physically trained individuals, there have been contrasting results, with some studies suggesting improvements in endurance and others showing limited or negative impacts on high-intensity exercise (69, 70, 111). Thus the effects of ketone supplementation on exercise performance in humans have yet to be fully elucidated.

Taken together, it appears that skeletal muscle ketone metabolism influences whole body glucose homeostasis in obesity/T2D (Fig. 3), while increases in muscle ketone metabolism may improve exercise performance and subsequent metabolic health. However, as ketone oxidation and signaling may have contrasting roles in the heart and other organs involved in CVD (as discussed below) versus the muscle, further work is necessary before strategies aimed at decreasing ketone oxidation can be considered as a glucose-lowering strategy.

Figure 3.

Both increasing and decreasing ketone availability to skeletal muscle paradoxically improves glycemia in T2D/obesity. Decreases in skeletal muscle ketone oxidation have glucose-lowering effects in T2D, yet providing exogenous ketones also lowers glucose in T2D, independent of skeletal muscle ketone metabolism. This may involve both an increase in glucose oxidation and/or the signaling effects of ketones. However, the signaling pathways by which ketones improve glycemia are incompletely understood but are suspected to outweigh the impacts of reducing ketone oxidation. T2D, type 2 diabetes.

ROLE OF KETONES IN CARDIAC AND VASCULAR INJURY

Myocardial Infarction and Ischemia

As NLRP3 inflammasome activation is a key mediator of myocardial ischemia/reperfusion (I/R) injury (71–73), the ability of βOHB to inhibit NLRP3 activation (23) suggests that it may have clinical utility in the treatment of myocardial infarction (MI) and/or acute coronary syndromes. While ketone-mediated inhibition of NLRP3 may be attributed to multiple systemic effects, βOHB still induces cardioprotection in ex vivo hearts, indicating a direct cardiac effect. For instance, the addition of βOHB to isolated Langendorff-perfused mouse hearts subjected to I/R results in reduced NLRP3 inflammasome activation and improved cardiac function (74). Similarly, 24 h of subcutaneous βOHB administration in mice post-I/R reduced infarct size and apoptosis while preserving cardiac function (75). These beneficial effects were also associated with improvements in autophagy, mitochondrial function, and oxidative and endoplasmic reticulum stress (75). While intravenous infusion of βOHB before an MI has been shown to reduce infarct size and decrease cardiomyocyte apoptosis in mice (76), as well as reduce oxidative stress and cardiac dysfunction in pigs (77), whether intravenous infusion of βOHB is also effective in a post-MI setting is unknown. Despite some reports identifying a positive correlation between circulating βOHB and MI (78, 79), these other data suggest that further augmenting circulating βOHB beyond the physiological response is cardioprotective.

In support of ketone therapy having beneficial effects following MI and ischemia, investigations using oral KEs in larger animal models and patients have been conducted. Yurista et al. (80) tested whether pigs subjected to I/R and then treated with KE for 3 days would benefit in a post-MI setting. In this study, the KE-treated pigs showed notably lower cardiac NLRP3-related and macrophage-mediated inflammation, oxidative stress, and apoptosis (80), consistent with a previous study that demonstrated that an elevated circulating βOHB reduced cardiac inflammation (74). In addition, in a small trial of patients with cardiogenic shock due to acute MI or decompensated HF, acute enteral KE administration improved cardiac output and peripheral perfusion beyond the effects of the existing vasoactive treatment for cardiogenic shock (81). Taken together, these studies suggest that elevating circulating levels of βOHB via KE supplementation can reduce cardiac injury following hypoxia, ischemia, and MI, and thus lay the foundations for further exploration of ketone therapy in acute coronary syndromes, MI, and decompensated HF.

Heart Failure

Current pharmacological therapies used to treat HF are vast and are continuing to expand (82). While these therapies [such as sodium glucose cotransporter 2 (SGLT2) inhibitors, glucagon-like peptide 1 receptor agonists, angiotensin-converting enzyme inhibitors, angiotensin receptor/neprilysin inhibitors, soluble guanylate cyclase stimulators, β-blockers, etc.] have helped improved cardiac mortality in people with HF, the overall prevalence of HF is still increasing (82). Thus new therapeutic approaches are needed to help treat this devastating condition. One such therapeutic option to help treat HF may be ketone therapy, where this approach may be most effective as an adjunct therapy. The rationale for this potential approach originated with the findings of two independent groups who reported increases in myocardial ketone utilization during HF with reduced ejection fraction (HFrEF) (83, 84). These reports coincided with data from SGLT2 inhibitor cardiovascular outcome trials, which suggested that the cardiovascular benefit of SGLT2 inhibition could be partly due to elevations in ketones (85, 86). As such, investigations into the therapeutic potential of ketones on the failing heart grew considerably.

βOHB is a more oxygen-efficient substrate than fatty acids and can theoretically improve energy production efficiency in the failing heart, which is known to have defects in metabolic processes that normally allow for proper ATP production necessary to maintain contractile function (4, 87). The importance of ketones for cardiac function is exemplified by the findings that a cardiac-specific SCOT-deficient mouse model accelerated the progression of pressure overload-induced cardiac dysfunction and remodeling (88), while another genetically modified mouse with increased circulating levels of βOHB prevented the development of HFrEF (74). Furthermore, various studies have linked the protective effects of ketones in HF with the increase in βOHB utilization/metabolism (89, 90). However, despite this earlier hypothesis, more recent work has challenged this concept and has shown that this protective effect does not appear to be due to an improved cardiac energy efficiency (91, 92). Thus it appears as though improvements in oxygen efficiency are likely not the primary factor driving improved cardiac function and remodeling in HF in the presence of elevated βOHB, and other mechanisms may play more important roles.

Based on the findings discussed above, several studies have investigated other potential mechanisms explaining the beneficial effects of ketones in HF. Indeed, many reports focusing on the pleiotropic signaling effects of ketones have shown that the benefits of ketones in HF could occur via reduced NLRP3 activation, inflammation, and oxidative stress, as well as alterations to mitochondrial function, mitophagy, and autophagy (56, 74, 75, 89, 90, 93, 94). For example, when humans or animals with HFrEF or HF with preserved ejection fraction (HFpEF) are supplemented with exogenous βOHB both acutely or chronically (either through KE, a ketogenic diet, or βOHB injection/infusion), most studies observe an alleviation of cardiac dysfunction and pathological remodeling (22, 80, 90, 91, 95–97). This cardioprotection also exists in mice with T2D, where KE treatment prevents the decline of both systolic and diastolic function (56), similar to the effects observed when on a ketogenic diet (98). Furthermore, treatments that elevate βOHB directly reduce systemic vascular resistance and improve myocardial contractility and blood flow (91, 93, 99, 100), indicating a vascular benefit to treatment with βOHB. While we discuss some of the mechanisms that have been shown to contribute to the beneficial effects of elevated ketones in HF (Fig. 4), it is quite likely that other mechanisms are also involved and have yet to be elucidated.

Figure 4.

Pleiotropic actions by which ketones attenuate cardiovascular dysfunction in cardiometabolic disease. The net therapeutic benefits of ketones are due to a complex interplay of both nonmetabolic cellular signaling and metabolic ATP-producing oxidation. ATP, adenosine triphosphate.

One important distinction of the beneficial effects of ketones in HF is that the salutary effects of βOHB do not appear to be consistent across all forms of HF. While the therapeutic effects of βOHB in HFrEF may occur due to a combination of augmenting both βOHB signaling and oxidation (80, 91, 95, 96), enhanced βOHB signaling and decreased βOHB oxidation appear to confer the beneficial effects observed in mice with HFpEF (22). Importantly, while KE treatment did not reduce body weight in mice with HFpEF, it did ameliorate their hypertension and improve their exercise tolerance (22), thereby demonstrating that ketones can improve HFpEF independent of weight loss. Furthermore, myocardial ketone uptake is decreased in HFpEF patients versus HFrEF patients (101), suggesting differential perturbations of ketone handling or metabolism depending on the type of HF. Although myocardial ketone uptake is generally concentration-dependent, this alone cannot explain the differences in relative extraction among various HF pathologies (101–105). In addition, even though the differences in ketone utilization, kinetics, signaling, and metabolism between different forms of HF remain incompletely understood, ketones still have therapeutic benefits in HFpEF (22, 97). This may be due to the fact that IL-1β is required for HFpEF pathogenesis (28, 106, 107), and since βOHB reduces IL-1β production via NLRP3 inhibition, the provision of KEs may attenuate the development and/or progression of HFpEF via this pathway. This is also consistent with Deng et al. (22) who reported that KE treatment reduced cardiac NLRP3 activation and subsequent IL-1β and IL-18 production. Additionally, ketone therapy may contribute to rescuing diastolic function in HFpEF by inhibiting HDAC activity, which has been shown to govern diastolic dysfunction (108). That said, despite the distinct differences in HFrEF and HFpEF (109), these syndromes still exist on a dynamic continuum (110) and more studies are needed to clarify what specific patient phenotypes may best benefit from ketone therapy. The current clinical trials using ketone therapies for HF and other cardiometabolic diseases are shown in Table 1, and these trials may help to address which diseases are most responsive to ketone therapy.

Table 1.

Clinical trials using exogenous ketones as a therapeutic approach for cardiometabolic disease

Patient Characteristics	Intervention	Aim of Investigation	Clinical Trial Identifier
20 HFrEF patients	Acute KE vs. placebo; crossover trial	Effect on exercise performance and skeletal muscle PCr (via 31 P-MRS) during exercise	NCT05348460
20 HFpEF patients	Acute KE vs. placebo; crossover trial	Effect on exercise performance and cardiovascular function	NCT04633460
10 HFpEF patients with T2D	14 days of KE vs. placebo; crossover trial	Effects on lipolysis rate, and protein/glucose kinetics	NCT05159570
20 HFrEF patients	Acute ketone salt vs. carbohydrate supplement; crossover trial	Effect on cardiac function (via cardiac MRI) and blood biomarkers	NCT05651529
24 patients with T2D and HFpEF	14 days of KE vs. placebo; crossover trial	Effect on cardiac function and exercise capacity	NCT05236335
10 HFrEF patients	14 days of KE vs. placebo; crossover trial	Effects on lipolysis rate and protein and glucose kinetics	NCT05161676
18 nonischemic HFrEF patients	10 days of KE vs. ketogenic diet vs. placebo	Effect on cardiac function (via MRI) and biomarkers	NCT04921293
8 HFrEF patients	Acute KE vs. ketone salt vs. placebo; crossover trial	Effect on cardiac function	NCT04443426
8 HFrEF patients	Acute weight-adujusted dose of KE vs. placebo; crossover trial	Effect on cardiac function	NCT04594265
50 HFrEF patients	12 weeks of KE vs. placebo	Effect on cardiac function and creatine (via MRI)	NCT05924802
24 HFrEF patients	14 days of KE vs. placebo; crossover trial	Effect on cardiac function and exercise capacity	NCT05161650
20 HFrEF patients	Acute KE (startified by SGLT2 inhibitor use)	Safety of KE and effect on blood pH, βOHB, glucose, and pressure	NCT05757193
12 patients hospitalized with HF	Acute KE vs. placebo (maltodextrin)	Effect on cardiac function	NCT04442555
16 HFrEF patients	2 weeks of KE vs. placebo; crossover trial	Effect on cardiac function and exercise performance	NCT04370600
24 patients with acute HFrEF decompensation	Acute KE vs. placebo	Effect on cardiac and hemodynamic function and blood biochemistry	NCT04698005
10 comatose survivors of out-of-hospital cardiac arrest	48 h of KE (nasogastric)	Effect on blood biochemistry, biomarkers, and acid-base status	NCT03226197
90 patients with T2D, HF (±T2D), and healthy controls	14 days of KE	Effect on cardiac function and energetics (PCr via MRS) during rest and dobutamine stress test	ISRCTN24885065
12 HFrEF patients	Acute 1,3-butanediol vs. placebo; crossover trial	Effect on cardiac function	NCT05768100
15 healthy male adults	Acute KE vs. placebo; crossover trial	Effect on diurnal and nocturnal blood pressure and glucose homeostasis	NCT05794802
40 elder adults (>50 years)	Acute KE vs. placebo; crossover trial	Effect on nocturnal blood pressure in aging adults	NCT05888506
11 elder adults (50–75 years)	2 weeks of KE vs. placebo; crossover trial	Effect on vascular function, blood metabolites, and cognition	NCT04236388
78 T2D patients with EF <50%	6-hour dose-dependent βOHB infusion vs. time control	Effect on cardiac function and myocardial glucose uptake (via MRI)	NCT03560323
14 T2D patients	Acute pre-prandial KE vs. Na-βOHB (in water) vs. water; crossover trial	Effect on postprandial glucose and metabolites, lipid and glucose kinetics, and metabolic hormones	NCT05263401
20 T2D patients	90 days of concentrated βOHB acid with 1,3-butanediol vs. placebo	Feasibility and safety of chronic ketone supplenmtation; glycemic control, weight, and blood biochemistry/biomarkers	NCT05983562
20 T2D patients	90 days of βOHB acid with 1,3-butanediol vs. placebo	Feasibility and safety of chronic ketone supplenmtation; glycemic control, weight, and blood biochemistry/biomarkers	NCT05477368
12 T2D patients	Acute KE vs. placebo; crossover design	Effect on hepatic glucose production and postprandial glycemia	NCT05518448
30 T2D or obese patients and healthy controls	Acute KE vs. placebo; crossover trial	Effects on hemodynamic function, appetite, and cognition	NCT05651243
18 T2D patients	Acute KE vs. placebo; crossover trial	Effect on glycemia, lipids, inflammation, and blood pressure	NCT04194450
15 T2D patients	14 days of pre-prandial KE vs. placebo; crossover trial	Effect on glycemia, vascular function, and cognition	NCT05155410
16 T2D patients	Acute KE vs. placebo; crossover design	Effect on cardiac function, exercise capacity, metabolic rate, and systemic inflammation	NCT04854330
T1D patients	Acute KE vs. palcebo; crossover trial	Effect on blood pH and metabolites	NCT04487678

³¹P, 31 phosphorus; EF, ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; KE, ketone ester; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; PCr, phosphocreatine; SGLT2, sodium-glucose cotransporter 2; T2D, type 2 diabetes.

As previously mentioned, studies on athletes and fit individuals have suggested that ketone supplementation can improve exercise performance (65, 111, 112). In HF, exercise intolerance is a particularly important symptom contributing to patient morbidity due to reduced exercise and functional capacity (113). Despite this, to date, there have not been any direct investigations to determine whether ketone therapy during HF can improve exercise capacity or symptoms of exercise intolerance, and thus is an area that remains to be explored.

Atrial Fibrillation

Atrial fibrillation (AF) is a common comorbidity in cardiometabolic disease (114, 115) and an area where ketone therapy remains understudied. Since SGLT2 inhibitors are becoming better recognized as an approach to improve AF and arrhythmias (116, 117), and because ketones are believed to mediate some of the cardioprotective effects of SGLT2 inhibitors, it is possible that ketones may also improve AF. Furthermore, since NLRP3 inflammasome activation (and downstream inflammation) is important in the development of AF (118, 119), and because βOHB inhibits NLRP3 activation (23), ketone therapy may be a clinically relevant adjunct for patients with AF. While some studies suggest no correlation between circulating βOHB and new-onset AF (120), other studies show that cardiac βOHB is elevated in AF patients (121) and that circulating βOHB levels are elevated in arrhythmogenic cardiomyopathy patients (105), suggesting the potential for high βOHB concentrations to contribute to AF. While it is unknown how this may occur, βOHB enantiomers themselves have different effects on cardiomyocyte action potential peak current and calcium transients alone or dependent on the presence of other metabolites (122). However, βOHB can augment cardiomyocyte contraction and may improve excitation-contraction coupling (123). Despite these positive correlations/indications, direct investigation of whether βOHB administration ameliorates or exacerbates AF should be explored not only to identify a potential adjunct therapy but also to determine whether ketone therapy may be contraindicated in patients with AF.

Hypertension and Vascular Function

Vascular dysfunction is a major contributor to CVD (110), and thus the effects of ketones on vessel function have been studied. Clinical and preclinical findings recently demonstrated that βOHB reduces vascular resistance and alleviates high-salt-induced hypertension (91, 93, 100, 124, 125), the latter of which was associated with reduced renal inflammation and NLRP3 cascade activation (125). In addition, 2 wk of KE treatment in obese adults also improves vascular function, as measured by flow-mediated dilation (54), further supporting a beneficial vascular effect as a result of elevated circulating ketone levels. While βOHB-mediated vasodilation likely occurs via numerous mechanisms, their direct effects involve small (SK) and intermediate (IK) conductance calcium-activated potassium channels as well as the sodium/potassium ATPase (93). In contrast, ATP-sensitive potassium channels were not involved in mesenteric artery βOHB-dependent vasodilation (93), although they are activated by βOHB and reduce firing in neuronal tissue (126–129). This latter finding may be important in reducing the excess sympathetic drive typically present in hypertension as well as HF. Furthermore, βOHB-dependent vasodilation is not mediated by nitric oxide or cyclooxygenase products in healthy rat mesenteric vessels and also not through GPR41 or GPR109a (93). This lack of involvement of GPR109a was also shown in HFrEF patients where activating GPR109a did not elicit a vasodilatory response like the one that was induced with βOHB (130), suggesting that the vasodilatory mechanisms of βOHB are not mediated by these GPCRs. Regardless of the mechanisms, the ability of βOHB to induce vasodilation may be important in numerous diseases involving restricted blood flow. Indeed, newer lines of investigation have suggested that βOHB may also be of therapeutic utility in pulmonary arterial hypertension and potentially other subtypes of pulmonary hypertension. For example, recent preclinical and clinical evidence demonstrates that increasing circulating βOHB in pulmonary arterial hypertension improves right ventricular function and reduces pulmonary vascular resistance, inflammation, fibrosis, and macrophage activation/infiltration (124, 131). Thus the beneficial effects of ketone therapy may be important for several CVDs.

Ketones also appear to have other favorable vascular effects. This includes a ketone-mediated regulation of cardiac endothelial cell proliferation, migration, and sprouting capacity, the combination of which contributes to preserving vascularization in the hypertrophied mouse heart (132). Likewise, ketones also promote proliferative effects in lymphatic endothelial cells to functionally enable lymphangiogenesis after MI in mice (133). When considering that KE supplementation in healthy males subjected to endurance training increases angiogenesis and skeletal muscle capillarization (112), these data suggest that ketones contribute to vascular remodeling in the organs that have increased energetic demand in both physiological and pathological conditions. Conversely, βOHB signaling also promotes quiescence of endothelial cells, vascular smooth muscle cells, and muscular stem cells (134, 135). However, the reasons for these contrasting proliferative and quiescent effects have yet to be understood. Collectively, ketones prevent vascular and endothelial aging and preserve vessel density in lymphatic and blood vessels, demonstrating the favorable effects of ketones in the vasculature and in vessel injury.

LIMITATIONS AND ADDITIONAL CONSIDERATIONS OF KETONE THERAPY

The growing body of preclinical and clinical evidence supports the expanding use of ketone therapy to treat cardiometabolic diseases. However, ketone therapy is not without its pitfalls and there may be concerns with its use in some clinical contexts. One of the most apparent safety concerns of ketone therapy is diabetic ketoacidosis (DKA), a condition that is predominantly a risk in T1D and rare in T2D patients (136). Although the body can adaptively respond to an elevation in blood ketones up until ∼10 mM βOHB, pathological levels may occur at 10–20 mM or higher and can result in ketoacidosis and detrimental organ effects (137, 138). In most studies though, KEs or other methods to augment circulating ketones result in nonpathological elevations of βOHB. Thus, while ketone therapy may be a potential concern in those with poorly controlled T2D, it does not appear to be of notable concern for those with well-controlled T2D. For example, T2D patients who received 4 wk of KEs did not have concerning changes in acid-base status or a risk of DKA and, instead, had an improvement in glycemic control (55). Furthermore, KE use in these patients also did not result in hypoglycemia, renal dysfunction, electrolyte imbalances, or hypertension (55). Similar to other healthy or diseased individuals using KEs, the potential adverse effects in T2D or obese patients are mild and transient, but further research on the long-term risks of KEs in patients with cardiometabolic disease is still needed. It is also unknown whether KE use may further predispose T2D patients to euglycemic DKA, a syndrome now associated with the use of SGLT2 inhibitors, although preliminary data suggest that this predisposition is unlikely (55). Furthermore, while T1D patients are at risk of DKA, or even euglycemic ketoacidosis, the use of ketone therapy in the context of T1D remains understudied. However, with the increasing popularity and therapeutic uses of exogenous ketones, it will be important to determine the safety of ketone therapy use in patients with well-controlled T1D (Table 1).

We have used ketone therapy to refer to ketogenic diets, KEs, and ketone salts, although there are still other methods of elevating circulating ketones (139). However, unlike KEs or ketone salts, ketogenic diets should not necessarily be considered an equivalent therapeutic option. Indeed, ketogenic diets stimulate endogenous ketosis by virtually eliminating dietary carbohydrate intake. This alone can result in significant changes to the body’s metabolic profile (140). Thus the resulting positive or negative effects cannot directly be attributed to the addition of ketones. Conversely, increases in circulating ketones by KEs or ketone salts do not deprive the body of carbohydrates nor affect the body’s metabolic profile to the same degree. Moreover, in the case of HF, many reports have indicated that the failing heart is metabolically inflexible and has an increased reliance on glucose to meet its energetic demands (9). Thus limiting glucose availability could impair the heart from producing adequate ATP levels, despite whatever beneficial effects may come from ketones. In addition, some reports have observed that ketogenic diets do not improve cardiac function in HF or I/R (84, 90, 141). One also needs to consider the very high-fat nature of ketogenic diets, which may lead to other adverse effects that are not likely to be observed with KEs or ketone salts. One potential adverse effect worth noting with ketogenic diets is that they are frequently associated with increases in circulating low-density lipoprotein (LDL) cholesterol levels (142, 143). However, T2D patients receiving a ketogenic diet compared to the standard of care for 1 yr had favorable improvements in parameters such as high-density lipoprotein (HDL) cholesterol and 10-yr risk of atherosclerotic CVD despite an increase in LDL cholesterol. While it is unknown if sustained elevations in LDL cholesterol levels may contribute to worsening cardiovascular outcomes with more chronic ketogenic diet adherence, this finding is consistent with the notion that the correlation of LDL cholesterol with CVD is oversimplified and incomplete (144). Thus, in general, while the long-term effects of ketogenic diets in T2D/obesity and HF are more ambiguous (37), the preliminary evidence with KEs seems to be more promising (54–56, 60).

With regards to the optimal doses for ketone therapy, elevating circulating βOHB can be therapeutic anywhere from ∼0.5 mM to 5 mM (139, 145), although it stands to reason that a modest, but prolonged, elevation would be preferred to achieve consistency of effects. However, it remains unknown whether pulsatile or consistently elevated levels of circulating βOHB yield greater beneficial effects, as endogenous ketone levels exhibit a diurnal fluctuation (146, 147). Moreover, different KEs can also affect whether or not circulating ketone levels are elevated during the day (95), suggesting that the modality of ketone supplementation should also be considered in determining the ideal dosing times and intervals.

With the various strategies to achieve ketosis (139), the metabolism and the kinetics of all these compounds are different. It also needs to be considered that KEs have a bitter taste and are not palatable. Because of this, and to prolong the duration of ketosis, there are continued efforts to create new KEs that may be more advantageous than the primary ketone monoester (R)-3-hydroxybutyl (R)-3-hydroxybutyrate and the ketone diester 1,3-butanediol acetoacetate diester (148–150). Before the widespread use of KEs and ketone salts, 1,3-butanediol (1,3-BD) was another commonly known ketone precursor used to induce ketosis. While a low dose of 1,3-BD has generally been reported to be safe, using a dose of 1,3-BD that elicits the same βOHB concentrations as with a 24-h fast is deleterious and toxic in rats (151). Interestingly, a low dose of 1,3-BD may improve cardiometabolic parameters independent of its conversion to βOHB (151), suggesting that the beneficial effects of 1,3-BD might occur via an alternate mechanism. While evidence demonstrates improvements in cardiometabolic parameters in up to 4 wk of KE use in humans, there are relatively limited data with longer periods of KE use, and this is needed to better understand the efficacy of chronic ketone therapy. In any case, it should be noted that the use of KEs is likely superior to using salt-based preparations (i.e., sodium-βOHB) as ketone salts have a shorter half-life and may cause potentially dangerous sodium loads (139), even though both are classified as Generally Recognized as Safe supplements by the Food and Drug Administration.

One particularly striking theme of almost all preclinical studies using ketone therapy is the lack of female animals. The lack of study in females is a major shortcoming in translating preclinical findings to use in humans. Interestingly, ketone metabolism appears to decline at a greater rate in male than female mice as they age (152), although more research on female animals and humans is needed to determine what effects this has in the context of cardiometabolic disease. Of relevance, HFpEF is more common in women (153), and thus further investigation of ketone therapy and its potential sexual dimorphic benefits in cardiometabolic disease is warranted.

CONCLUDING REMARKS

There is a growing body of evidence demonstrating the benefits of ketone therapy for cardiometabolic disease, particularly with regard to obesity, T2D, and CVD. The numerous beneficial effects of ketone metabolism and signaling in the heart also extend to the vasculature and skeletal muscle, although which component contributes the most to attenuating cardiometabolic diseases is still unclear. Furthermore, the goal(s) of ketone therapy may require contrasting approaches for improving functional outcomes in the skeletal muscle versus the heart, which will also need to be carefully considered. It is also imperative that future studies continue to explore the benefits of using ketones as a potential adjunct therapy, while also determining whether it may pose certain risks in patients with cardiometabolic disease. As it stands to date, it is likely that ketones may correct metabolic and signaling perturbations to improve inflammation, oxidative stress, and cardiac and vascular function. In addition, ketone therapy can potentially stimulate weight loss and improve glycemic control to improve metabolic outcomes. With further advancements in our understanding of the multifaceted actions of ketones as a fuel source and a signaling molecule, ketone therapy may become an important component of the approach to treating cardiometabolic diseases.

GRANTS

J.R.B.D. receives funding from the Canadian Institutes of Health Research (CIHR), the Heart and Stroke Foundation of Canada, and Diabetes Canada. J.R.U. is supported by CIHR and Diabetes Canada. SS is supported by the CIHR Canada Graduate Doctoral Scholarship and the Izaak Walton Killam Memorial Scholarship. S.A.T.D. is supported by a postdoctoral fellowship from the CIHR (MFE-186352). J.R.U. is a Canada Research Chair in Pharmacotherapy of Energy Metabolism in Obesity. J.R.B.D. is a Canada Research Chair in Molecular Medicine.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.S., S.A.T.D., J.R.U., and J.R.B.D. prepared figures; S.S., S.A.T.D., J.R.U., and J.R.B.D. drafted manuscript; S.S., S.A.T.D., J.R.U., and J.R.B.D. edited and revised manuscript; S.S., S.A.T.D., J.R.U., and J. R.B.D. approved final version of manuscript.