# Metabolic flexibility, PDH and the Randle cycle

- Canonical: https://nutrifit.health/knowledge/metabolic-inflexibility-pdh-pyruvate-dehydrogenase
- Language: en
- Published: 2026-10-08T21:12:41.829Z
- Updated: 2026-10-08T21:12:41.829Z
- Author: NutriFit Editorial

> How glucose transport, PDH, PDK/PDP, lipid signalling, redox and CrAT coordinate fuel selection. What RER, primary experiments and trials show, and where diagnostic and treatment inferences stop.

Metabolic flexibility is the capacity to adjust fuel use to nutrient availability, hormonal signals and energy demand. A muscle that increases fat oxidation during fasting and carbohydrate oxidation during an insulin stimulus demonstrates one aspect of this capacity. Exercise, prolonged fasting and adaptation to a particular diet probe different aspects. There is no single measurement that captures every tissue and every transition. Pyruvate dehydrogenase is an important control point, but reduced whole-body flexibility does not establish that this enzyme is structurally damaged or completely switched off.

The pyruvate dehydrogenase complex—PDC, often abbreviated PDH—connects mitochondrial pyruvate metabolism with acetyl-CoA production. Its activity responds to products, substrate supply, reversible phosphorylation and tissue-specific gene expression. These mechanisms help explain normal fuel competition and some disturbances in insulin resistance. They also explain why a plausible biochemical mechanism is insufficient to justify a universal supplement regimen or a diagnosis based on fatigue alone.

## 1. Switching fuels is a coordinated physiological response

During the postabsorptive period, lower insulin permits adipose lipolysis and greater delivery of nonesterified fatty acids to other tissues. Skeletal muscle and heart can increase their reliance on fat; liver maintains blood glucose through glycogen breakdown and gluconeogenesis and, with longer fasting, produces ketones. Erythrocytes depend on glycolysis because they lack mitochondria. The brain normally uses substantial glucose but can use ketones during prolonged fasting; it is not permanently an obligate glucose-only organ.

After a carbohydrate-containing meal, insulin suppresses lipolysis, promotes GLUT4-dependent glucose uptake in muscle and adipose tissue, and supports glycogen storage and carbohydrate oxidation. PDE3B-dependent reduction of cAMP contributes to suppression of adipocyte lipolysis, together with regulation of the lipolytic machinery. This is a graded transition. Healthy muscle does not instantaneously stop all fat oxidation or convert every incoming glucose molecule into oxidative ATP. Fuel oxidation, glycogen synthesis and other nonoxidative disposal occur together.

Skeletal muscle is a major site of insulin-stimulated glucose disposal, particularly during experimental clamps. Its contribution depends on the stimulus, muscle mass and measurement method; a percentage from a clamp cannot simply be applied to every meal. The heart continuously uses a mixture of fatty acids, glucose, lactate, ketones and other substrates. Their proportions change with circulation, workload, feeding, disease and oxygen supply. Liver coordinates substrate exchange rather than behaving like a larger version of a GLUT4-dependent muscle cell (Kelley and Mandarino, 2000).

This physiological distinction matters in insulin resistance. Reduced suppression of adipose lipolysis, impaired glucose transport, altered storage, and changes in mitochondrial oxidation can all reduce the observed shift toward carbohydrate. They need not arise from a single primary lesion in PDH. The location of lipid deposition also matters: the discussion of [ectopic and visceral fat in the TOFI phenotype](/knowledge/tofi-phenotype-ectopic-fat-visceral-adiposity) helps distinguish fat quantity from the compartments that influence metabolic signalling.

## 2. What respiratory exchange measures—and what it cannot identify

The respiratory quotient, RQ, describes the stoichiometric ratio of carbon dioxide production to oxygen consumption during oxidation. Respiratory exchange ratio, RER, is measured at the mouth or in a calorimeter. Under a suitable steady state it approximates whole-body RQ; during hyperventilation, intense exercise, changing bicarbonate stores or other nonsteady conditions the two can differ.

$$
\begin{aligned}
\mathrm{RER}&=\frac{\dot V_{\mathrm{CO_2}}}{\dot V_{\mathrm{O_2}}},\\
\mathrm{C_6H_{12}O_6}+6\mathrm{O_2}&\longrightarrow6\mathrm{CO_2}+6\mathrm{H_2O},\qquad \mathrm{RQ}_{\mathrm{glucose}}=1,\\
\mathrm{C_{16}H_{32}O_2}+23\mathrm{O_2}&\longrightarrow16\mathrm{CO_2}+16\mathrm{H_2O},\qquad \mathrm{RQ}_{\mathrm{palmitate}}=\frac{16}{23}\approx0.696.
\end{aligned}
$$

The dotted gas volumes are rates expressed in the same units, for example litres per minute, so their ratio is dimensionless. The palmitate equation represents oxidation of a fatty acid, not the exact composition of all triglycerides. Protein oxidation usually gives an intermediate value near 0.8, with variation among amino acids and nitrogen handling. Estimating nonprotein substrate oxidation requires additional assumptions or information about protein oxidation.

A fasting value near 0.7 suggests a substantial contribution from fat under the measurement conditions, whereas a value near 1 suggests a greater contribution from carbohydrate. Neither value is a universal health target. Recent diet, energy balance, glycogen availability, fasting duration, age, activity and the preceding measurement period influence the result. RER above 1 during vigorous exercise can reflect buffering and ventilation rather than a chemically impossible fuel; during overfeeding, net lipogenesis can also affect gas exchange.

For a defined insulin challenge, flexibility is often operationalised as the change in the measured quotient:

$$
\Delta\mathrm{RER}=\mathrm{RER}_{\mathrm{stimulated}}-\mathrm{RER}_{\mathrm{fasting}}.
$$

The two values must come from a specified protocol. Oral glucose, a mixed meal and a hyperinsulinaemic-euglycaemic clamp are different challenges. Insulin dose, duration, achieved glucose disposal and suppression of fatty acids affect the difference. The frequently repeated claim that everyone healthy must have a change above 0.15–0.20, while a change below 0.05 proves PDH blockade, has no universal clinical validation.

Galgani and colleagues studied 59 people with type 2 diabetes and 42 nondiabetic participants matched for obesity, sex and race. Under their clamp conditions, the average RQ change was approximately 0.06 versus 0.10. The between-group difference disappeared after adjustment for glucose disposal. This finding shows that limited delivery of glucose into cells can explain a smaller oxidative response; it does not exclude tissue-specific mitochondrial regulation. It directly cautions against treating a low delta as proof of an intrinsic PDH defect (Galgani et al., 2008).

## 3. The Randle cycle is physiological competition, not a rule against mixed meals

Randle, Garland, Hales and Newsholme described the glucose–fatty-acid cycle in 1963. In the classic experimental framework, fatty-acid oxidation increases acetyl-CoA relative to free CoA and NADH relative to NAD⁺. These products can constrain pyruvate oxidation through product effects and increased PDK activity. Citrate generated from acetyl-CoA and oxaloacetate can contribute to feedback on cytosolic phosphofructokinase-1. Reduced glycolytic flux can raise glucose-6-phosphate and thereby inhibit hexokinase. The sequence explains substrate competition in the studied conditions; it is not a complete account of human insulin resistance (Randle et al., 1963; Garland et al., 1963).

These effects are distributed across compartments and occur alongside regulation of glucose transport. Citrate export involves the mitochondrial citrate carrier, SLC25A1. PFK-1 acts before production of triose phosphates, so its inhibition is not a block “at the triose-phosphate step.” G6P-mediated inhibition of hexokinase is also distinct from insulin-dependent GLUT4 trafficking. A fall in glucose oxidation cannot by itself reveal which step first limited glucose uptake.

Human experiments demonstrate the importance of that distinction. In a lipid-infusion study with nine healthy participants, Roden and colleagues measured muscle G6P and glycogen during a clamp. The findings supported an early limitation of glucose transport or phosphorylation, rather than the predicted initial accumulation of G6P after PDH inhibition. The classical feedback pathway remains biochemically relevant, but it cannot be imposed as the obligatory causal order in all human tissues (Roden et al., 1996).

The reciprocal side of fuel selection involves acetyl-CoA carboxylase, malonyl-CoA and CPT1. Increased malonyl-CoA inhibits the entry step for long-chain acyl groups through the carnitine shuttle and can restrain fatty-acid oxidation. During contraction, AMPK-dependent ACC phosphorylation can reduce malonyl-CoA and help increase lipid use. This control is tissue- and demand-dependent; neither pathway requires complete exclusion of the other fuel. The Randle cycle does not establish that combining fat and carbohydrate in an ordinary meal is intrinsically toxic.

## 4. The PDH complex: an organised catalytic relay

PDH is a large, dynamic multienzyme assembly primarily located in the mitochondrial matrix. Cytosolic pyruvate crosses the inner mitochondrial membrane through the mitochondrial pyruvate carrier, MPC1/MPC2, before encountering PDH. Transport through MPC and conversion by PDH are different processes: phosphorylation of PDH does not physically seal the carrier. Pyruvate can also support pyruvate carboxylase-dependent anaplerosis and gluconeogenesis, rather than being obligatorily diverted to lactate.

The core organises dihydrolipoamide acetyltransferase, E2/DLAT, and E3-binding protein, E3BP/PDHX. Peripheral E1 is an α₂β₂ heterotetramer composed of PDHA1 and PDHB products; PDHA2 provides a testis-associated α isoform. E3/DLD is a FAD-containing homodimer. E3 participates in other mitochondrial enzyme systems too, so DLD dysfunction is not necessarily an isolated PDH disorder.

Structural work has refined the core architecture. A 2024 human PDH study combined cryo-electron microscopy and mass spectrometry and supported a 60-subunit core with 48 E2 and 12 E3BP copies, capable of associating with 48 E1 heterotetramers and 12 E3 dimers. This replaces an additive interpretation of “60 E2 plus 12 E3BP.” Occupancy, preparation and the flexible peripheral shell affect the mass and dimensions of observed assemblies; a fixed diameter or E1 count is not a universal molecular constant (Zdanowicz et al., 2024).

| Component or cofactor | Function in the relay | Important distinction |
|---|---|---|
| E1, TPP/ThDP and Mg²⁺ | Decarboxylation of pyruvate; transfer of the hydroxyethyl-derived unit | TPP is thiamine diphosphate, a vitamin B1-derived cofactor |
| E2 and covalently bound lipoyllysine | Carries the acetyl group between catalytic sites; transfers it to CoA | The mobile protein domain plus its lipoyl group is more than a fixed short “wire” |
| CoA-SH | Accepts the acetyl group as acetyl-CoA | Pantothenate contributes to CoA biosynthesis; CoA is not vitamin B5 itself |
| E3 with bound FAD | Reoxidises reduced lipoyl groups through its redox machinery | FAD is recycled within E3 rather than released as free FADH₂ |
| NAD⁺ | Final soluble electron acceptor, producing NADH | Niacin-derived NAD metabolism and mitochondrial compartmentation matter |

E1 first forms a TPP-bound intermediate while releasing CO₂. Reductive acetylation transfers the carbon unit onto an E2 lipoyl group. E2 then transfers the acetyl group to CoA-SH. E3 regenerates the oxidised lipoyl group, with NAD⁺ receiving the electrons. The disulfide/dithiol cycling and flexible lipoyl domains make repeated turnover possible without releasing each intermediate into the bulk solution.

$$
\mathrm{Pyruvate}+\mathrm{CoA{-}SH}+\mathrm{NAD^+}\xrightarrow{\mathrm{PDH}}\mathrm{Acetyl\!-!CoA}+\mathrm{CO_2}+\mathrm{NADH}+\mathrm{H^+}.
$$

This conventional biochemical equation uses the usual metabolite notation; detailed proton balance depends on the chosen ionic species and pH convention. The net reaction is strongly favourable under physiological conditions and is effectively irreversible in mammalian metabolism. Consequently, the acetyl carbons of even-chain fatty acids cannot provide net glucose through reversal of PDH or the ordinary TCA cycle. Glycerol from triglycerides and odd-chain fatty-acid carbon are separate cases. Acetyl-CoA also cannot be described as a direct substrate for glycogen synthesis.

[![PDH reaction stages: E1 with TPP/Mg²⁺ releases CO₂; E2 transfers acetyl to CoA; E3 with FAD regenerates the lipoyl group and reduces NAD⁺. Reaction schematic, not atomic structure or domain attachment map.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-metabolic-inflexibility-pdh-pyruvate-dehydrogenase-relay-en-v3.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-metabolic-inflexibility-pdh-pyruvate-dehydrogenase-relay-en-v3.png)

The orange cycle denotes regeneration of the E2-tethered lipoyl arm by E3 and its return to E1 for another turnover. Electrons leave through NAD⁺; the E3→E1 arc does not mean reverse electron transfer.

The relay diagram shows carbon transfer separately from electron recycling. Enzyme shapes are symbolic and do not represent measured atomic coordinates or the exact number of subunits in the assembled complex.

## 5. Reversible phosphorylation: sites, kinase isoforms and phosphatases

Four PDK isoforms phosphorylate E1α; PDP phosphatases remove those phosphates. Phosphorylation can markedly reduce activity of modified enzyme molecules, but the fraction modified, substrate concentration, tissue state and residual catalytic capacity determine the flux of a tissue. “All PDH is completely off” is usually an unjustified inference from greater PDK4 expression or a phosphoprotein signal.

Two numbering conventions must be kept separate. In the human PDHA1 precursor, the sites are Ser293, Ser300 and Ser232. After removal of the 29-residue mitochondrial targeting sequence, the same residues are Ser264, Ser271 and Ser203. Ser293 and Ser264 are therefore alternative numbers for site 1, not two different sites. A list containing Ser264, Ser271 and Ser293 mixes conventions and omits site 3.

| Regulatory feature | Correct interpretation | Evidence boundary |
|---|---|---|
| Site 1 | Precursor Ser293 = mature Ser264 | Frequently the fastest-modified site; strongly inhibits overall complex activity |
| Site 2 | Precursor Ser300 = mature Ser271 | Modified by all four PDKs in recombinant experiments, at different rates |
| Site 3 | Precursor Ser232 = mature Ser203 | PDK1 phosphorylates this site in the classic isoform comparison; PDK4 is not an all-three-site kinase |
| PDK1–4 | Isoforms differ in tissue distribution, docking and sensitivity to metabolites | No single activity ranking is valid independently of assay conditions |
| PDP1 | Mg²⁺-dependent catalysis; Ca²⁺ promotes interaction with the lipoyl domain and activation | Contraction-associated matrix Ca²⁺ is a physiological signal |
| PDP2 | Also requires Mg²⁺; relatively insensitive to Ca²⁺ | Hormonal regulation varies across tissues and experimental models |

Korotchkina and Patel used recombinant human E1 variants to compare the four kinases. Their results establish site specificity, while also showing that relative phosphorylation rates depend on conditions. Kato and colleagues subsequently studied phosphorylated human E1 variants structurally. Site-1 modification disrupted ordering of phosphorylation loops and impaired interaction with lipoyl domains and reductive acetylation. This mechanism is more specific than an assertion that phosphate simply plugs the pyruvate channel. Activity with an artificial electron acceptor is not equivalent to the complete E1–E2–E3 reaction (Korotchkina and Patel, 2001; Kato et al., 2008).

PDK2 and PDK4 can respond to the state of lipoyl domains and to product-related metabolic signals. Pyruvate inhibits PDK activity with isoform-dependent sensitivity; ADP and the availability of ATP also influence regulation. ATP consumption by a kinase is distinct from ATP production downstream of PDH. Phosphatase activity depends on local ion availability, not merely the amount of magnesium swallowed. Calcium supplements cannot reproduce carefully regulated mitochondrial calcium transients.

## 6. PDK4 induction can be adaptive or maladaptive

Fasting, low carbohydrate availability, glucocorticoid signalling and selected lipid signals can increase PDK4 expression. PPARα/PPARδ, FOXO factors and PGC-1α participate in context-dependent transcriptional regulation. Insulin–Akt signalling can inhibit FOXO1 through phosphorylation and altered localisation. These pathways help conserve carbohydrate when it is scarce. Increased PDK4 during fasting is therefore not, by itself, a disease marker (Huang et al., 2002). (Furuyama, 2003).

With sustained nutrient excess and impaired insulin action, reduced suppression of adipose lipolysis can maintain fatty-acid delivery when it would otherwise fall after feeding. Product inhibition, PDK regulation and disturbed transport can then constrain the oxidative response to glucose. This is a network disturbance, not an irreversible biochemical state. Expression measurements do not show that all kinase molecules are active, that every site is phosphorylated, or that the effect cannot respond to demand.

The ratios acetyl-CoA/CoA-SH and NADH/NAD⁺ are meaningful within defined compartments. Their direction of change can help explain product pressure, but neither must always be “much greater than one” in metabolic inflexibility. Total concentrations measured in a homogenate also do not automatically report the free matrix ratios experienced by an enzyme. There is no universal three- to five-fold activation constant applicable to every tissue and PDK isoform.

[![Glucose through glycolysis and MPC supplies pyruvate to PDH. β-oxidation products can promote PDK-dependent phosphorylation; PDP dephosphorylates PDH. DAG–PKCθ is a parallel insulin-signalling pathway; this is not a diagnostic test.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-metabolic-inflexibility-pdh-pyruvate-dehydrogenase-fuel-en-v1.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-metabolic-inflexibility-pdh-pyruvate-dehydrogenase-fuel-en-v1.png)

The regulation diagram separates mitochondrial fuel competition from parallel impairment of insulin signalling. A reduced PDH reaction flux does not necessarily imply reduced MPC transport, and lipid-mediated signalling disturbances can limit glucose uptake before a mitochondrial enzyme becomes rate-limiting.

## 7. Lipid signalling and mitochondrial stress are related, but distinct

DAG accumulation in relevant membrane compartments can activate novel protein kinase C isoforms. PKCθ is a novel, DAG-sensitive, calcium-independent PKC—not an atypical PKC. In muscle, altered IRS–PI3K–Akt–TBC1D4/AS160 signalling can reduce GLUT4 trafficking. In liver, PKCε and regulation of the insulin receptor are important in different experimental models. Ceramides can interfere with Akt through other mechanisms. They should not be folded into a single “DAG and ceramide both activate atypical PKCθ” pathway. (Baier, 1993).

Specific IRS serine positions depend on species and context; mouse Ser307 and human Ser312 cannot be listed as independent human residues proving an invariant mechanism. Insulin resistance generally means impaired responsiveness rather than complete extinction of the pathway. Neutral triglyceride storage is also not equivalent to the concentration or location of bioactive lipid intermediates. Endurance-trained muscle can contain substantial intramyocellular triglyceride while retaining good insulin sensitivity.

Fatty-acid oxidation supplies electrons through different routes. NADH communicates with complex I. Electrons from acyl-CoA dehydrogenases pass through ETF and ETF:ubiquinone oxidoreductase to the ubiquinone pool; they do not pass obligatorily through complex II. Complex II is the succinate dehydrogenase pathway. High membrane potential and a highly reduced quinone pool can favour reverse electron transport and ROS formation in particular conditions, but a universal membrane-potential threshold of 180–200 mV does not diagnose lipid-induced PDH dysfunction. (Christensen, 1984).

Koves and colleagues developed an influential model in which excessive delivery to β-oxidation becomes mismatched with TCA-cycle and respiratory demand. In experimental systems, metabolite profiling and manipulation of fatty-acid entry linked incomplete oxidation with impaired substrate switching. This differs from the simple assertion that insulin resistance always reflects insufficient fat oxidation. Both limited oxidative capacity and excessive fuel delivery may be relevant in different settings (Koves et al., 2008).

Acylcarnitines reflect this network, but individual circulating species are not direct photographs of matrix congestion. Their levels depend on tissue production, exchange, diet, fasting, renal handling and inherited metabolic disorders. C2 relates to acetyl-group metabolism; C3 and several C5 isomers can also reflect amino-acid catabolism. Long-chain species can increase in multiple conditions, and their biological effects depend on concentration and context. An elevated panel alone cannot prove that they caused mitochondrial destruction or establish a general PDH block.

## 8. Mitochondrial architecture, nuclear metabolism and immune cells

MFN1/MFN2 participate in outer-membrane fusion, OPA1 in inner-membrane organisation and fusion, and DRP1 in division. ER–mitochondrial contacts influence calcium and lipid exchange. Mfn2 loss in experimental models can disturb insulin signalling and cellular stress responses, supporting an interaction between architecture and metabolism. However, its effects on contacts depend on model and method; it is not a universal physical tether whose threefold reduction explains every case of inflexibility. Cristae organisation also involves proteins beyond the fusion–fission pair. Division can participate in adaptation and quality control rather than always meaning injury (Sebastián et al., 2012).

PDH activity is not confined to a simple mitochondrial-only description. Sutendra and colleagues detected functional nuclear PDC in mammalian cell models and linked it to local acetyl-CoA generation, histone acetylation and cell-cycle progression. The finding establishes a metabolic–epigenetic connection in those systems. It does not demonstrate that cytosolic and nuclear acetyl-CoA can never exchange, or that common metabolic inflexibility inevitably silences protective genes through nuclear PDH failure. ACLY- and ACSS2-dependent acetyl-CoA generation and spatial regulation provide additional context (Sutendra et al., 2014).

Macrophage activation illustrates why more glucose oxidation is not always the only desirable state. Experimental inflammatory activation often increases glycolysis and modifies TCA-cycle use; cytokines, nitric oxide, HIF signalling and metabolite accumulation influence the outcome. PDK1 knockdown in mouse macrophage experiments reduced selected M1-associated responses and enhanced early M2-associated oxidative activation. Other studies implicated PDK2/PDK4. The conventional M1/M2 labels simplify a spectrum of states, and these experiments do not establish that all immune cells in people with obesity are locked into a single PDH-phosphorylated phenotype (Tan et al., 2015; Min et al., 2019).

## 9. Tissue consequences and the limits of symptom-based inference

In type 2 diabetes, impaired transport and storage of glucose coexist with altered lipid delivery and hepatic glucose production. Chronic hyperglycaemia can increase polyol and hexosamine pathway activity, methylglyoxal formation and glycation. These processes contribute to complications but cannot all be attributed to a preceding complete PDH block. ATP supply may be maintained while fuel selection is disturbed; the phrase “cellular starvation amid abundance” is a metaphor rather than a measured universal energy deficit.

In diabetes, cardiac fatty-acid reliance can increase and glucose oxidation can become less responsive to insulin or workload. The heart nevertheless continues to use multiple fuels. Glucose oxidation generally yields more ATP per unit of oxygen than palmitate oxidation under conventional assumptions, but exact percentage differences depend on ATP accounting, shuttle use, coupling and workload. Actual mechanical efficiency additionally reflects ion cycling, contractile work and uncoupling; fuel stoichiometry alone does not demonstrate chronic myocardial hypoxia.

During ischaemia, glycolysis can continue without oxygen, while oxidation of either pyruvate or lactate ultimately requires respiratory oxygen. Lactate oxidation cannot be presented as an anaerobic rescue pathway. PDH regulation can influence coupling of glycolysis to glucose oxidation during reperfusion, but oxygen delivery, respiratory integrity and proton handling remain essential. SERCA2a is an ATP-driven calcium pump, not a calcium channel. Intracellular acidosis, calcium overload and cell death arise from several interacting mechanisms, not one fixed pH threshold after PDH phosphorylation.

In MASLD, lipid delivery from adipose tissue, dietary fat, DNL, oxidation and lipoprotein export jointly determine liver fat. Insulin resistance can coexist with continued lipogenesis and excess hepatic glucose output. Pyruvate carboxylase and PDH partition pyruvate differently, and acetyl-CoA can activate pyruvate carboxylase. This does not make PDH suppression the universal initiating cause of steatosis or fibrosis. The [fructose–ketohexokinase and DNL pathway](/knowledge/fructose-ketohexokinase-de-novo-lipogenesis-nafld) gives complementary detail on carbohydrate-derived lipid synthesis, including the roles of energy surplus and food form.

Fatigue and poor exercise tolerance have broad differential diagnoses: anaemia, cardiopulmonary disease, deconditioning, sleep disorders, endocrine disorders, medication effects, infection and neurological or metabolic disease. They do not establish a PDH lesion. Physiological lactate production also occurs under adequately oxygenated conditions, and lactate is a transferable fuel. Neither a walking speed nor an isolated postexercise lactate value defines common metabolic inflexibility. Persistent or substantial lactate elevation requires interpretation in its clinical setting.

Ageing can change muscle mass, fitness, mitochondrial turnover and substrate responses. Human measurements of phosphocreatine kinetics and CrAT activity support associations with metabolic inertia, but activity and training history modify the phenotype. Age is not a diagnosis of obligatory PDH failure, and a narrow RER response does not alone explain frailty or sarcopenia (Mancilla et al., 2023).

## 10. Assessing flexibility without inventing a diagnostic panel

Research calorimetry can quantify substrate responses when fasting, activity, gas calibration, steady-state conditions and the challenge are standardised. Clamps additionally quantify insulin-stimulated glucose disposal. Tissue enzyme assays, phosphorylation measurements, isotope tracing and magnetic-resonance methods address different levels of metabolism. Their results are related but not interchangeable.

| Measurement | What it can contribute | Why it is not a standalone PDH diagnosis |
|---|---|---|
| Resting and stimulated RER, ΔRER | Integrated change in whole-body oxidative fuel selection | Depends on recent diet, balance, protocol and glucose delivery; no universal clinical cutoff |
| Clamp glucose disposal | Quantifies insulin-stimulated disposal under specified conditions | Combines oxidation and storage; cannot localise the first limiting step alone |
| Glucose, HbA1c, insulin or HOMA-IR | Describes glycaemia or a model-based estimate of insulin resistance | HOMA values depend on assay and population, and are not PDH measurements |
| Nonesterified fatty acids | Gives information about circulating lipid availability and suppression | Sampling conditions and adipose regulation matter; a concentration does not identify matrix flux |
| Acylcarnitine profile | Helps evaluate defined inherited disorders and research metabolic patterns | Species are not specific to incomplete oxidation or a universal PDH block |
| Lactate and pyruvate, with molar ratio | Supports evaluation of selected lactic-acidosis and mitochondrial disorders | Sampling quality, lactate concentration, oxygenation and other causes affect interpretation |
| β-Hydroxybutyrate and acetoacetate | Reflects ketogenesis and redox-related influences in context | Plasma ratios do not directly measure every tissue's mitochondrial NADH/NAD⁺ |
| Tissue or genetic testing | Can establish specific enzyme defects or inherited disease in selected cases | Requires indication and expert interpretation; not a routine screening panel for fatigue |

LDH links lactate and pyruvate to the free cytosolic redox couple. The equilibrium relation includes proton activity; at specified pH it can be incorporated into an apparent equilibrium constant. The concentrations forming a ratio must be expressed in the same molar units.

$$
\begin{aligned}
\mathrm{Pyruvate}+\mathrm{NADH}+\mathrm{H^+}&\rightleftharpoons\mathrm{Lactate}+\mathrm{NAD^+},\\
\frac{[\mathrm{Lactate}]}{[\mathrm{Pyruvate}]}&=K'_{\mathrm{LDH}}\frac{[\mathrm{NADH}]}{[\mathrm{NAD^+}]},\\
\mathrm{Acetoacetate}+\mathrm{NADH}+\mathrm{H^+}&\rightleftharpoons\beta\text{-hydroxybutyrate}+\mathrm{NAD^+}.
\end{aligned}
$$

These are compartment-specific biochemical relationships, not equations permitting a blood test to reconstruct all intracellular redox ratios. Pyruvate is particularly sensitive to collection and processing. An elevated lactate/pyruvate ratio does not specifically identify PDH inhibition. Primary PDH deficiency often raises lactate and pyruvate proportionally, leaving the ratio low or within the usual range near 10–20. Debray and colleagues showed that diagnostic discrimination among causes of congenital lactic acidosis depends strongly on lactate concentration. Those paediatric diagnostic findings must not be converted into a threshold for lifestyle-associated metabolic inflexibility (Debray et al., 2007).

Likewise, fixed fasting β-hydroxybutyrate/acetoacetate thresholds of 5–8 do not establish a universal matrix “redox collapse.” Ketone concentration, production, utilisation and sample stability matter. It is more informative to specify the clinical question and obtain appropriate measurements than to order an unvalidated eight-marker panel and interpret every deviation as the same enzyme defect.

## 11. Carnitine acetyltransferase buffers acetyl groups in both directions

CrAT catalyses reversible exchange of acetyl groups between acetyl-CoA and carnitine:

$$
\mathrm{Acetyl\!-!CoA}+\mathrm{Carnitine}\rightleftharpoons\mathrm{Acetylcarnitine}+\mathrm{CoA{-}SH}.
$$

The reaction can regenerate free CoA and permit acetyl-carbon exchange when acetyl supply exceeds immediate use. In the opposite direction, acetylcarnitine can provide an acetyl reserve as demand rises. This buffering role differs from the long-chain CPT1–CACT–CPT2 shuttle: CPT1 forms long-chain acylcarnitines, CACT exchanges them across the inner membrane, and CPT2 generally regenerates matrix acyl-CoA. Describing CPT2 as a universal emergency exporter of all incompletely oxidised products reverses the usual direction of that transport system.

Muscle-specific Crat deletion in mice, experiments in human myocytes and limited human supplementation observations support a role in substrate switching. Seiler and colleagues further investigated acetyl buffering during exercise. These studies provide a mechanistic rationale, not proof that every person with fatigue needs grams of carnitine, that circulating C2 measures CrAT deficiency, or that oral acetyl-L-carnitine instantly removes PDK activity (Muoio et al., 2012; Seiler et al., 2015).

## 12. Cofactors: correcting deficiency is different from forcing flux

Thiamine deficiency can impair PDH and other thiamine-dependent enzymes and can contribute to serious neurological or cardiovascular disease and lactic acidosis. Its detection and treatment are clinically important in appropriate circumstances. However, cofactor dependence does not mean that a person with adequate intake benefits from saturating every enzyme with pharmacological doses. Some inherited PDH variants have particular thiamine responsiveness; that is a specialised disease context, separate from common insulin resistance.

Benfotiamine is a thiamine derivative, not a guarantee of direct matrix delivery or protection of a serine residue against PDK. Studies of thiamine derivatives have often focused on transketolase, vascular outcomes or diabetic complications, rather than directly measuring human PDH flux. A small crossover trial in 12 hyperglycaemic participants investigated 300 mg/day thiamine for six weeks and reported a change in glucose tolerance. Its size and outcome do not validate a PDH-reactivation protocol or a universal dose (Alaei Shahmiri, 2013).

Human enzyme lipoylation uses a specialised mitochondrial assembly pathway involving LIPT2, LIAS and LIPT1. Free α-lipoic acid sold as a supplement must not be assumed to rebuild E2 lipoyl arms through a human equivalent of bacterial lipoate salvage. Biochemical work on the human pathway and model experiments distinguish lipoylation from the antioxidant and signalling effects of administered lipoic acid. Evidence from those other actions is not evidence of direct repair of PDH lipoylation (Cao et al., 2018).

Riboflavin, niacin and pantothenate contribute to FAD, NAD and CoA metabolism. Providing a precursor does not establish that a deficient matrix pool was the limiting step. NR/NMN supplementation is particularly easy to overinterpret: an increase in a blood NAD-related marker is not necessarily an increase in PDH flux. In a randomised trial in obese, insulin-resistant men, 12 weeks of NR did not improve insulin sensitivity or whole-body glucose metabolism; related muscle analyses did not demonstrate improved mitochondrial respiration or morphology (Dollerup et al., 2018, 2020).

Magnesium is required for phosphatase catalysis and many other processes. Correcting a recognised deficiency can be appropriate, but an in-vitro Mg²⁺ dependence is not a basis for prescribing 400–600 mg of supplemental elemental magnesium to everyone. Oral dose does not directly specify free matrix Mg²⁺; renal function and gastrointestinal tolerability matter. Calcium signalling during contraction is likewise not reproduced by indiscriminate supplementation. Pantethine, zinc, chromium and omega-3 products have separate research literatures, but the combined “PDH rescue stack” lacks demonstrated efficacy as a unit. Neither chromomodulin models nor zinc's general biological roles establish a need for high-dose chromium or zinc to protect PDH thiols.

## 13. Drugs and experimental peptides: match the claim to the evidence

Dichloroacetate inhibits PDK and can lower lactate in defined experimental and clinical contexts. The isoforms differ in sensitivity; reducing kinase activity permits existing phosphatases to act rather than directly removing phosphate. Lower lactate is a biochemical endpoint, not proof of improved survival, exercise function or cardiometabolic health. Peripheral neurotoxicity is a material limitation. In a randomised MELAS trial involving 30 participants, neuropathy contributed to early termination without a demonstrated clinical benefit under those conditions. That disease-specific result should neither be ignored nor described as a trial proving efficacy in common insulin resistance (Kaufmann et al., 2006).

PS10 is also frequently misdescribed. The compound developed by Tso and colleagues targeted the ATP-binding pocket and inhibited all four PDK isoforms; it was not a selectively PDK4-only agent. Increased PDH activity, improved glucose tolerance and less hepatic steatosis were reported in obese mice. These findings support a research target, not a human clinical protocol or an absence of toxicity in long-term use (Tso et al., 2014).

| Intervention or model | What was actually examined | What remains unproven |
|---|---|---|
| DCA in MELAS, Kaufmann et al., 2006 | Randomised disease-specific treatment; trial stopped early with peripheral nerve toxicity | General benefit for obesity-related inflexibility or chronic fatigue |
| PS10, Tso et al., 2014 | PDK inhibition, tissue PDH changes and metabolic outcomes in obese mice | Human efficacy, a PDK4-only action, and long-term clinical safety |
| Thiamine, 2013 crossover trial | Glucose tolerance in a small hyperglycaemic human sample | Direct PDH normalisation or a universal cofactor regimen |
| NR, Dollerup et al., 2018 | Randomised human supplementation with metabolic outcomes | Reliable improvement of insulin sensitivity or mitochondrial flux |
| Early eating window, Sutton et al., 2018 | Small controlled crossover feeding study in men with prediabetes | A mandatory 16/8 schedule or PDH-specific rescue |
| MOTS-c, Lee et al., 2015 | Cellular metabolic signalling and mouse interventions | A validated injectable treatment for human PDH dysfunction |

SGLT2 inhibitors and GLP-1 receptor agonists have established clinical uses in appropriate disease settings, but their benefits cannot be reduced to a shared, proven direct suppression of PDK4 in every human tissue. SGLT2 inhibition changes glucose loss, insulin–glucagon balance and substrate availability; increased ketone availability is one proposed contributor to cardiac effects. Ketones provide acetyl-CoA through pathways that bypass PDH, but still require TCA-cycle and respiratory function. They are not a universally superior oxygen-independent fuel. GLP-1 therapies can improve metabolic outcomes through effects on appetite, weight, glycaemia and several organ systems; direct hepatocyte GLP-1R mechanisms require careful model-specific interpretation.

Metformin similarly cannot be described as a universal direct PDH dephosphorylator. Established indications and clinical outcomes guide treatment. Mechanistic signals in cells or animals are useful hypotheses but should not replace evidence about benefits and adverse effects in the relevant population.

MOTS-c is a 16-amino-acid peptide associated with an open reading frame within mitochondrial 12S rRNA. Lee and colleagues linked it to inhibition—not activation—of folate-linked de novo purine synthesis, AICAR accumulation and AMPK signalling in cell models, with metabolic effects in mice. This does not prove selective PDK4 suppression as its universal mechanism. Humanin, associated with a mitochondrial 16S rRNA sequence, has experimental cytoprotective actions. The described tripartite receptor includes CNTFRα, WSX-1 and gp130; WSX-1 is itself IL-27Rα, so listing both as different receptor subunits duplicates the same component. Peptide length and translation context also require care. These research observations do not establish routine peptide treatment for metabolic inflexibility (Lee et al., 2015; Hashimoto et al., 2009).

## 14. Exercise, diet and circadian coordination

At the start of contraction, calcium signalling, greater ADP production, pyruvate availability and altered product pressure can increase active PDH and oxidative flux. The precise timing depends on intensity, fibre type, training and prior diet. Insulin-independent glucose transport and regulation of fatty-acid entry occur alongside this response. There is no single 65–70% VO₂max boundary at which every muscle switches completely from fat to carbohydrate.

Repeated training can increase oxidative capacity, improve insulin action and change mitochondrial turnover and substrate handling. Endurance and resistance exercise address complementary aspects of health and function. The usefulness of a programme depends on adherence, progression, recovery and disease context; fixed claims of 40–60% mitochondrial expansion or a mandatory combination of Zone 2 and HIIT are not universal biological rules. Severe exertional symptoms require assessment rather than automatic escalation of exercise to “force PDH on.”

Diet influences substrate supply and the measured response. Lower carbohydrate intake can induce PDK4 and reduce carbohydrate oxidation as an adaptation to scarcity. This does not prove irreversible pathological inflexibility or inevitable diabetes. Conversely, chronic energy surplus can increase lipid delivery and ectopic deposition regardless of whether the surplus is assigned to one macronutrient. Before a clinical glucose-tolerance test, standard preparation matters because preceding carbohydrate restriction can alter the response.

Time-restricted eating can help some people organise intake, and selected controlled studies report improvements in insulin-related outcomes. Sutton and colleagues tested an early six-hour eating window against a twelve-hour schedule in men with prediabetes; this was not a trial of a universally required 16/8 pattern or a direct measure of PDH restoration. Total intake, dietary quality, weight change, timing and adherence must be separated when interpreting benefits. Timing carbohydrate around exercise may serve sports-specific goals, but ordinary carbohydrate meals need not be restricted to a narrow period before or after training (Sutton et al., 2018).

Muscle clocks coordinate metabolic programmes. In muscle-specific Bmal1-knockout mice, Dyar and colleagues found changes in glucose handling and the regulation of PDH-related enzymes. This supports a link between circadian biology and substrate metabolism. It does not establish a universal human PDK4 peak at 03:00–05:00, a minimum at breakfast, or persistent maximal PDK4 expression in everyone exposed to evening screens. Sleep, shift work and feeding schedules influence metabolism through several routes; precise clock times require evidence in the relevant human setting (Dyar et al., 2014).

## 15. A practical interpretation, with clear evidence limits

The useful clinical sequence is to identify the actual problem—dysglycaemia, excess adiposity, exercise limitation, persistent lactate elevation or a suspected inherited disorder—and investigate it with appropriate methods. Established cardiometabolic measures, clinical assessment and targeted testing generally answer more than a purported universal “PDH panel.” Review medications, nutritional risk, activity, sleep and other causes of symptoms. Treat a demonstrated deficiency or disease on its own evidence; do not infer that every symptom needs simultaneous manipulation of five cofactors, PDK, peptides and feeding windows.

Metabolic flexibility remains a valuable concept precisely because it describes adaptation to context. PDH regulation, insulin signalling, substrate transport, lipid storage, redox balance and tissue architecture all contribute. A smaller respiratory response is informative within its protocol; it does not by itself establish an enzyme lesion, and maximal carbohydrate oxidation is not the correct target in every physiological state.

This article is a narrative synthesis with targeted verification of molecular numbering, substrate pathways, selected structural studies, human diagnostic observations and intervention claims. Cell experiments, animal models, human physiological studies and clinical trials are distinguished throughout. It is not a systematic review, a GRADE assessment, an external peer review or a validated clinical treatment protocol. The illustrative pathways simplify a dynamic network, and the absence of a universal dose or cutoff is a substantive evidence limit rather than a missing step in an established regimen.

Evidence access was checked on 8 October 2026. Targeted primary bibliographic records and indexed abstracts were available; selected full-text sections of Kato and colleagues’ structural paper were examined for the phosphorylation mechanism. Abstract or record access for other studies is not presented as a full-text audit.

## FAQ

### Does a low RER prove a PDH defect?

No. RER depends on conditions, diet and substrate availability; a small change does not locate the limitation in PDH.

### Is PDK4 always pathological?

No. Fasting or carbohydrate restriction can induce it to conserve substrates; tissue and context matter.

### Must fats and carbohydrates be separated?

The Randle cycle describes physiological competition and does not establish toxicity of ordinary mixed meals.

### Can a supplement rebuild lipoyl arms?

Human lipoylation uses LIPT2, LIAS and LIPT1. Antioxidant effects of free lipoic acid do not prove E2 repair.

### Are universal cofactors or peptides justified?

Correcting a confirmed deficiency differs from forcing enzyme flux. A clinically validated universal PDH-restoration stack is absent.

## Sources

- [Randle PJ, Garland PB, Hales CN, Newsholme EA. The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet. 1963.](https://pubmed.ncbi.nlm.nih.gov/13990765/)
- [Kelley DE, Mandarino LJ. Fuel selection in human skeletal muscle in insulin resistance: a reexamination. Diabetes. 2000;49:677–683.](https://pubmed.ncbi.nlm.nih.gov/10905472/)
- [Garland PB, Randle PJ, Newsholme EA. Citrate as an intermediary in the inhibition of phosphofructokinase by fatty acids, ketone bodies, pyruvate, diabetes, and starvation. Nature. 1963;200:169–170.](https://pubmed.ncbi.nlm.nih.gov/14073034/)
- [Roden M et al. Mechanism of free fatty acid-induced insulin resistance in humans. J Clin Invest. 1996.](https://pubmed.ncbi.nlm.nih.gov/8675698/)
- [Galgani JE et al. Metabolic flexibility in response to glucose is not impaired in people with type 2 diabetes after controlling for glucose disposal rate. Diabetes. 2008;57:841–845.](https://pubmed.ncbi.nlm.nih.gov/18285553/)
- [Zdanowicz R et al. Stoichiometry and architecture of the human pyruvate dehydrogenase complex. Sci Adv. 2024;10:eadn4582.](https://pubmed.ncbi.nlm.nih.gov/39018392/)
- [Korotchkina LG, Patel MS. Site specificity of four pyruvate dehydrogenase kinase isoenzymes toward the three phosphorylation sites of human pyruvate dehydrogenase. J Biol Chem. 2001;276:37223–37229.](https://pubmed.ncbi.nlm.nih.gov/11486000/)
- [Kato M et al. Structural basis for inactivation of the human pyruvate dehydrogenase complex by phosphorylation: role of disordered phosphorylation loops. Structure. 2008;16:1849–1859.](https://pubmed.ncbi.nlm.nih.gov/19081061/)
- [Huang B, Wu P, Bowker-Kinley MM, Harris RA. Regulation of pyruvate dehydrogenase kinase expression by peroxisome proliferator-activated receptor-alpha ligands, glucocorticoids, and insulin. Diabetes. 2002;51:276–283.](https://pubmed.ncbi.nlm.nih.gov/11812733/)
- [Koves TR et al. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab. 2008;7:45–56.](https://pubmed.ncbi.nlm.nih.gov/18177724/)
- [Sebastián D et al. Mitofusin 2 links mitochondrial and endoplasmic reticulum function with insulin signaling and is essential for normal glucose homeostasis. Proc Natl Acad Sci USA. 2012.](https://pubmed.ncbi.nlm.nih.gov/22427360/)
- [Sutendra G et al. A nuclear pyruvate dehydrogenase complex is important for the generation of acetyl-CoA and histone acetylation. Cell. 2014.](https://pubmed.ncbi.nlm.nih.gov/24995980/)
- [Tan Z et al. Pyruvate dehydrogenase kinase 1 participates in macrophage polarization via regulating glucose metabolism. J Immunol. 2015;194:6082–6089.](https://pubmed.ncbi.nlm.nih.gov/25964487/)
- [Min BK et al. Pyruvate dehydrogenase kinase is a metabolic checkpoint for polarization of macrophages to the M1 phenotype. Front Immunol. 2019;10:944.](https://pubmed.ncbi.nlm.nih.gov/31134063/)
- [Mancilla RF et al. Skeletal muscle mitochondrial inertia is associated with carnitine acetyltransferase activity and physical function in humans. JCI Insight. 2023;8:e163855. Online 2022.](https://pubmed.ncbi.nlm.nih.gov/36413408/)
- [Debray FG et al. Diagnostic accuracy of blood lactate-to-pyruvate molar ratio in the differential diagnosis of congenital lactic acidosis. Clin Chem. 2007.](https://pubmed.ncbi.nlm.nih.gov/17384007/)
- [Muoio DM et al. Muscle-specific deletion of carnitine acetyltransferase compromises glucose tolerance and metabolic flexibility. Cell Metab. 2012;15:764–777.](https://pubmed.ncbi.nlm.nih.gov/22560225/)
- [Seiler SE et al. Carnitine acetyltransferase mitigates metabolic inertia and muscle fatigue during exercise. Cell Metab. 2015;22:65–76.](https://pubmed.ncbi.nlm.nih.gov/26154055/)
- [Alaei Shahmiri F, Soares MJ, Zhao Y, Sherriff J. High-dose thiamine supplementation improves glucose tolerance in hyperglycemic individuals: a randomized, double-blind cross-over trial. Eur J Nutr. 2013;52:1821–1824.](https://pubmed.ncbi.nlm.nih.gov/23715873/)
- [Cao X et al. Protein moonlighting elucidates the essential human pathway catalyzing lipoic acid assembly on its cognate enzymes. Proc Natl Acad Sci USA. 2018.](https://pubmed.ncbi.nlm.nih.gov/29987032/)
- [Dollerup OL et al. A randomized placebo-controlled clinical trial of nicotinamide riboside in obese men: safety, insulin-sensitivity, and lipid-mobilizing effects. Am J Clin Nutr. 2018.](https://pubmed.ncbi.nlm.nih.gov/29992272/)
- [Dollerup OL et al. Nicotinamide riboside does not alter mitochondrial respiration, content or morphology in skeletal muscle from obese and insulin-resistant men. J Physiol. 2020;598:731–754. Epub 2019.](https://pubmed.ncbi.nlm.nih.gov/31710095/)
- [Kaufmann P et al. Dichloroacetate causes toxic neuropathy in MELAS: a randomized, controlled clinical trial. Neurology. 2006.](https://pubmed.ncbi.nlm.nih.gov/16476929/)
- [Tso SC et al. Structure-guided development of specific pyruvate dehydrogenase kinase inhibitors targeting the ATP-binding pocket. J Biol Chem. 2014;289:4432–4443.](https://pubmed.ncbi.nlm.nih.gov/24356970/)
- [Lee C et al. The mitochondrial-derived peptide MOTS-c promotes metabolic homeostasis and reduces obesity and insulin resistance. Cell Metab. 2015.](https://pubmed.ncbi.nlm.nih.gov/25738459/)
- [Hashimoto Y et al. Humanin inhibits neuronal cell death by interacting with a cytokine receptor complex or complexes involving CNTF receptor alpha/WSX-1/gp130. Biochem Biophys Res Commun. 2009.](https://pubmed.ncbi.nlm.nih.gov/19703422/)
- [Sutton EF et al. Early time-restricted feeding improves insulin sensitivity, blood pressure, and oxidative stress even without weight loss in men with prediabetes. Cell Metab. 2018.](https://pubmed.ncbi.nlm.nih.gov/29754952/)
- [Dyar KA et al. Muscle insulin sensitivity and glucose metabolism are controlled by the intrinsic muscle clock. Mol Metab. 2014.](https://pubmed.ncbi.nlm.nih.gov/24567902/)
- [Furuyama T et al. Forkhead transcription factor FOXO1 (FKHR)-dependent induction of PDK4 gene expression in skeletal muscle during energy deprivation. Biochem J. 2003;375:365–371.](https://pubmed.ncbi.nlm.nih.gov/12820900/)
- [Baier G et al. Molecular cloning and characterization of PKC theta, a novel member of the protein kinase C (PKC) gene family expressed predominantly in hematopoietic cells. J Biol Chem. 1993;268:4997–5004.](https://pubmed.ncbi.nlm.nih.gov/8444877/)
- [Christensen E, Kølvraa S, Gregersen N. Glutaric aciduria type II: evidence for a defect related to the electron transfer flavoprotein or its dehydrogenase. Pediatr Res. 1984;18:663–667.](https://pubmed.ncbi.nlm.nih.gov/6433313/)