Metabolic flexibility: how to measure and build it

It describes the capacity to increase lipid oxidation during fasting or low-intensity work, then shift toward glucose uptake and oxidation when insulin rises or exercise intensity requires rapid ATP production.

David E. Kelley and Lawrence J. Mandarino framed the concept this way in 2000, with skeletal muscle at the center of the model. A metabolically flexible muscle oxidizes more fat in the fasted state and responds to insulin by disposing of glucose efficiently. A metabolically inflexible system may remain disproportionately dependent on carbohydrate at rest, show impaired insulin-stimulated glucose use, or fail to access carbohydrate effectively when high output is required.

This distinction matters because metabolic flexibility is not synonymous with ketosis, fasting tolerance, low fasting glucose, or a stable continuous glucose monitor trace. It is a switching phenotype. The question is not whether the body can use fat or glucose. Nearly every healthy person can do both. The question is whether substrate selection changes appropriately as nutrient availability and energy demand change.

The physiology of metabolic switching

The central measurement problem is substrate oxidation. Cells can derive energy from fatty acids, glucose, lactate, ketones, and—in limited circumstances—amino acids. In day-to-day metabolic physiology, the relevant balance is primarily between fat and carbohydrate oxidation.

After an overnight fast, insulin is relatively low, glucagon signaling is more prominent, hepatic glycogen availability declines, and fatty acid mobilization rises. Skeletal muscle should therefore move toward a greater reliance on lipid oxidation. During a carbohydrate-containing meal or insulin stimulation, the expected direction reverses: glucose uptake and oxidation should increase.

This transition depends on several systems operating together:

• Mitochondrial oxidative capacity. Mitochondria must process fatty-acid-derived acetyl-CoA efficiently at relatively low and moderate workloads. Limited oxidative capacity tends to raise dependence on glycolysis at intensities that should remain predominantly aerobic.
• Insulin signaling and glucose transport. Insulin-stimulated translocation of GLUT4 transporters is essential for glucose disposal in skeletal muscle. Resistance training is relevant here because contractile activity can also increase muscle glucose uptake through insulin-independent pathways.
• Fatty acid transport. Proteins including CD36 and carnitine palmitoyltransferase 1, or CPT1, participate in fatty acid movement and mitochondrial entry. Nutrient status and repeated exercise exposure can modulate these pathways.
• Liver glycogen management. The liver buffers blood glucose between meals. Fasting duration, carbohydrate intake, training volume, sleep, and alcohol intake all influence how rapidly liver glycogen is depleted and replenished.
• Muscle mass. Skeletal muscle is a major glucose-disposal organ. A metabolic strategy that neglects muscle retention may improve a short-term glucose graph while weakening a more consequential long-term variable.

The sequence is chronological rather than mystical. At low energy demand, fat oxidation should contribute substantially. As intensity rises, carbohydrate oxidation becomes more practical because it supplies ATP more rapidly. At very high intensity, glycolytic flux increases and lactate rises. Once effort falls, the system should recover its oxidative behavior.

Metabolic flexibility is not permanent fat burning. It is the ability to use the appropriate fuel without excessive physiological friction.

This is why rigid dietary identities can obscure the subject. A strict ketogenic diet may increase the capacity to oxidize fat in some contexts, but it does not automatically establish efficient carbohydrate handling. Conversely, a high-carbohydrate athlete may dispose of glucose exceptionally well yet show limited fat oxidation at a given low intensity. Neither observation alone describes the entire phenotype.

How to measure metabolic flexibility in a clinical setting

The gold-standard approach combines indirect calorimetry with a euglycemic-hyperinsulinemic clamp. It is technically demanding, expensive, and not designed for routine self-experimentation. But it establishes the logic of the measurements used outside a laboratory.

Indirect calorimetry measures oxygen consumption (VO2) and carbon dioxide production (VCO2). From these values, clinicians calculate the respiratory exchange ratio, or RER:

RER = VCO2 / VO2

An RER of approximately 0.70 indicates predominant fat oxidation. An RER of 1.00 indicates predominant carbohydrate oxidation. Most resting and exercise measurements fall somewhere between these endpoints, with interpretation dependent on nutritional status, exercise intensity, ventilation, and recent training.

During a clamp, insulin is infused while glucose is administered as needed to maintain stable blood glucose. The comparison of baseline and insulin-stimulated RER—the change in RER, or ΔRER—provides a direct view of metabolic switching under controlled conditions. A larger appropriate shift toward carbohydrate oxidation during insulin stimulation generally indicates better metabolic flexibility.

The clamp also quantifies insulin sensitivity through the glucose infusion rate required to maintain euglycemia. In research protocols, an insulin infusion rate of 40 mU/m²/min is commonly used. This is not a consumer test. Its value is precision, not convenience.

RER answers one question; lactate answers another

RER is informative, but it is not the only useful marker. Exercise lactate profiling provides a practical window into oxidative metabolism under load.

In work published by Iñigo San-Millán and George A. Brooks in 2018, blood lactate accumulation during exercise was strongly inversely correlated with fat oxidation. The overall correlation was reported at approximately r = -0.76, becoming substantially stronger in endurance-trained cohorts. In plain terms: at a matched low-to-moderate workload, an earlier or higher lactate rise often indicates lower fat oxidation and a greater glycolytic contribution.

That relationship is mechanistically plausible. When mitochondrial oxidation cannot meet ATP demand at a given workload, glycolytic throughput rises and lactate production increases. Lactate itself is not metabolic waste; it is a valuable fuel and signaling molecule. The relevant variable is the workload at which lactate begins to accumulate disproportionately.

A useful assessment sequence is therefore layered rather than dependent on a single wearable metric:

1. Establish the baseline metabolic context. Fasting glucose, glycated hemoglobin, triglycerides, HDL cholesterol, liver enzymes, waist circumference, blood pressure, sleep pattern, and training status shape interpretation. No isolated RER value can compensate for absent context.

2. Measure exercise physiology where possible. A graded test with lactate sampling can identify the workload at which lactate begins to rise and can help define an individually plausible low-intensity aerobic range.

3. Use indirect calorimetry for a direct substrate measurement. This is most useful when the result will alter a clinical or training decision rather than merely satisfy curiosity.

4. Treat CGM trends as supplementary data. Glucose variability, meal responses, and nocturnal patterns may reveal behavioral triggers. They cannot establish that an individual is oxidizing fat efficiently.

5. Repeat under comparable conditions. A post-travel test after poor sleep and a well-rested test after a standardized evening meal are not interchangeable observations.

The most common analytical error is to infer metabolic flexibility from a single fasting glucose value or a “flat” glucose curve. Both can be compatible with very different substrate-use patterns.

Zone 2 and the mitochondrial component

Zone 2 training is often presented as an independent metabolic cure. The evidence does not support that framing. It is a useful stimulus because it permits substantial aerobic work with relatively low glycolytic flux. It is not the only exercise mode relevant to metabolic health.

In broad practical terms, Zone 2 is frequently approximated at 60% to 70% of maximum heart rate. That range is only a population estimate. Heart-rate formulas have wide individual error, and medications, heat exposure, hydration, fitness level, and autonomic state can shift heart rate independent of metabolic intensity.

Lactate-guided testing is more specific. The aim is to identify an intensity below the point at which lactate rises persistently, while maintaining a workload high enough to stimulate oxidative adaptation. At this intensity, the muscle repeatedly recruits mitochondrial pathways and fatty acid oxidation without relying heavily on high-rate carbohydrate metabolism.

Repeated aerobic exposure can support:

• increased mitochondrial density and oxidative enzyme activity;
• improved capacity for fatty acid transport and oxidation;
• a higher workload at a given lactate concentration;
• greater tolerance for prolonged low-to-moderate output;
• improved recovery of oxidative metabolism after short higher-intensity efforts.

The intervention is not inherently dramatic. Consistency matters more than occasional exhaustive sessions. A person who performs two long aerobic sessions per month is not producing the same signal as a person who accumulates repeated, tolerable low-intensity volume across several weeks.

However, aerobic work alone is incomplete. Resistance training preserves or increases muscle mass, which supports glucose disposal. Higher-intensity intervals can improve maximal oxygen uptake and carbohydrate utilization under demand. A flexible system needs both oxidative capacity and glycolytic competence.

The target is not to suppress carbohydrate metabolism. The target is to avoid being metabolically trapped in either fuel state.

For those using lactate testing, trends are generally more useful than isolated threshold labels. If the same cycling power or running pace produces less lactate over time, with comparable pre-test conditions, that suggests improved oxidative efficiency. It does not prove a universal improvement in insulin sensitivity, but it is a meaningful physiological signal.

Fasting and carbohydrate cycling: useful tools, limited claims

Metabolic flexibility fasting protocols typically involve a 12- to 16-hour overnight food-free interval. This duration can lower the insulin-to-glucagon ratio, reduce liver glycogen availability, and create conditions that favor fatty acid mobilization and oxidation. Fasted low-intensity exercise may amplify this signal in some individuals.

The mechanistic case is reasonable. Lower insulin signaling and reduced hepatic glycogen availability can increase expression or activity of pathways associated with fat transport and oxidation, including CD36 and CPT1. But the interpretation requires restraint.

A longer fasting window is not necessarily better. In people with high training loads, poor sleep, low energy availability, menstrual disturbances, diabetes medications, a history of disordered eating, or clinically significant stress, fasting may be counterproductive or unsafe. The metabolic response is not separable from the broader endocrine environment.

A metabolic flexibility diet plan is similarly best understood as a structure for matching substrate availability to demand rather than a fixed menu. Carbohydrate cycling is one example. Higher carbohydrate intake can be placed closer to demanding resistance, interval, or endurance sessions. Lower-carbohydrate intake may be more compatible with rest days or lower-intensity aerobic days.

A practical comparison looks like this:

Low-carbohydrate days are sometimes structured around roughly 50 to 75 grams of carbohydrate. That range is not a universal target. Body size, lean mass, insulin sensitivity, sport requirements, medication use, total energy intake, and food preference all materially alter the equation.

The relevant error is not consuming carbohydrate. The error is applying a dietary rule without asking what the day’s energy demand, training objective, and measured response actually are. Chronic carbohydrate restriction may reduce exposure to postprandial glucose excursions, yet it can also impair high-intensity training capacity in some individuals. Chronic high carbohydrate intake, meanwhile, may be entirely compatible with favorable metabolic function in a highly active person—but not in a sedentary person with impaired insulin sensitivity and excess energy intake.

Food quality remains operationally relevant. Fiber-rich carbohydrate sources, adequate protein, unsaturated fats, and microbiome diversity tend to support a more stable nutritional environment than highly refined, low-satiety foods. But no individual food category substitutes for adequate muscle mass, aerobic capacity, sleep, and energy balance.

What the newer monitoring tools can—and cannot—show

Continuous glucose monitors have changed the self-quantification landscape. They can identify repeated post-meal glucose patterns, differences between meals, the effect of exercise timing, and nocturnal glucose behavior. For selected users, particularly those with diabetes or clinically relevant dysglycemia, their utility is clear.

For healthy users, the interpretation is less settled. A CGM cannot measure fat oxidation. It cannot calculate RER. It cannot determine whether a person has shifted effectively from fasting lipid oxidation to insulin-stimulated glucose oxidation. A modest glucose rise after a carbohydrate meal is not automatically a sign of dysfunction; it may reflect meal composition, glucose appearance rate, recent exercise, stress, circadian timing, and sensor noise.

Continuous lactate monitors may eventually add a more relevant exercise signal. Lactate is closely tied to the balance between glycolytic flux and oxidative clearance during exercise. A wearable that tracks lactate continuously could help users identify workload transitions outside the laboratory and observe whether those transitions move over time.

Commercial development is moving in that direction, but the evidence base remains early. Long-term clinical trial data for continuous lactate monitors in healthy consumer populations are not yet sufficient to treat them as diagnostic instruments. Their likely near-term value is longitudinal pattern recognition, not clinical certainty.

Consumer breath devices occupy a similar position. They may be useful for observing repeated changes under standardized conditions, particularly when paired with training and dietary logs. Their agreement with laboratory indirect calorimetry across clinical populations remains uncertain. A device can be directionally useful without being diagnostically equivalent to a metabolic cart.

A sober interpretation of the evidence

The signs of metabolic flexibility are not a particular ketone level, a low-carb identity, or an unusually flat glucose chart. More persuasive evidence is convergent: appropriate fasting substrate use, preserved glucose disposal, low lactate at a meaningful aerobic workload, improving exercise capacity, and the ability to tolerate both lower- and higher-carbohydrate periods without major loss of function.

Metabolic flexibility is therefore best treated as a phenotype to characterize, not a score to chase. Indirect calorimetry and clamp studies remain the clearest research tools. Lactate profiling offers a practical exercise-based proxy. CGMs and emerging wearables can provide context but should not be asked to answer questions they cannot measure.

The current evidence supports Zone 2 work, resistance training, appropriately dosed higher-intensity exercise, overnight fasting windows for suitable individuals, and strategic carbohydrate availability as plausible ways to modulate the underlying physiology. It does not establish one optimal fasting duration, one carbohydrate ratio, or one consumer device protocol for everyone.

The more rigorous position is also the less marketable one: metabolic flexibility improves through repeated physiological exposure, sufficient recovery, and measured adaptation—not through permanent restriction of either fat or carbohydrate.