The term has become a sales pitch: ketone strips, breath analyzers and apps promise to “measure your metabolic flexibility” every morning. None measures what researchers mean by the term.
Metabolic flexibility is the ability to adjust fuel use to the situation: more fat during fasting or moderate exercise, more carbohydrates after a meal or as intensity increases. It is neither ketosis nor posting a high “fat-burning” reading one morning.
You cannot calculate it with a consumer formula. In the laboratory, several protocols measure a change between two standardized states. At home, a watch, glucose sensor or breath analyzer provides only a fragment of the picture.
The practical decision is therefore less spectacular: do not chase the lowest-carb score. Instead, build a body that can respond to a meal, an overnight fast and different exercise intensities, while monitoring the clinical markers that actually matter.
What is metabolic flexibility?
The review by Goodpaster and Sparks defines it as the ability to respond or adapt to changes in energy demand and conditions. In the classic model, muscle shifts from a greater contribution of fatty acids during fasting toward more glucose oxidation and storage in response to a meal and insulin.
This switch involves several organs. Fat tissue releases fatty acids when insulin is low, then slows that release after a meal. The liver maintains blood glucose between meals and stores or produces glucose depending on the context. Muscle sharply increases its energy needs during exercise and uses a fuel mix determined by intensity, duration, training and available stores.
Three common mix-ups blur the concept:
- oxidizing fat describes energy flow at a given moment
- being in ketosis describes increased ketone body production by the liver
- being flexible describes an appropriate response in several directions and contexts
A low respiratory quotient after a week of very low carbohydrate intake primarily shows that the body is using what it is given. It does not prove that it will respond well to a glucose load. Conversely, using more carbohydrates during an intense interval is not a failure: it is the expected response to a rapid demand for ATP.
Which signs cannot establish a diagnosis?
Fatigue between meals, cravings, sleepiness after lunch, irritability when fasting or difficulty losing weight are real experiences, but they are nonspecific. They may also reflect an overly large meal, an energy deficit, insufficient sleep, anemia, a thyroid disorder, medication or many other causes.
The clues often presented as a diagnosis are not enough:
- going a long time without eating
- producing lots of ketones
- having “perfect” morning blood glucose
- losing energy after a carbohydrate-rich meal
- burning a high proportion of fat in Zone 2
- maintaining a flat stomach
Insulin resistance deserves a clear distinction. It describes a reduced response by certain tissues to a given insulin concentration. It often contributes to a smaller shift toward glucose use, but the two concepts are not interchangeable. The NIDDK notes that direct tests of insulin resistance are mainly used in research and that prediabetes is assessed using HbA1c, fasting blood glucose or an OGTT, depending on the context.
If fatigue, unusual thirst, frequent urination, unintentional weight loss or episodes of feeling unwell persist, do not try to fix them by fasting. The article on insulin resistance and useful tests explains when to seek a clinical assessment.
How is it measured in the laboratory and at home?
Indirect calorimetry measures oxygen consumption and CO₂ production. Their ratio, measured at the mouth, is the RER. Under steady-state conditions, it approximates the respiratory quotient of the tissues: a value near 0.70 corresponds to a high contribution from fat, and a value near 1.00 to a high contribution from carbohydrates. Protein, ketogenesis, hyperventilation and the absence of a steady state complicate this interpretation.
In a classic protocol, the researcher measures fasting RER, infuses insulin while keeping blood glucose stable, then repeats the measurement. Flexibility is then expressed as follows: ΔRER = RER under insulin minus fasting RER. This calculation describes only that specific response. It has no universal cutoff applicable across all laboratories.
The test chosen changes the information obtained:
| Measurement | Information obtained | Main limitation |
|---|---|---|
| Hyperinsulinemic-euglycemic clamp + calorimetry | Change in RER and glucose use under insulin | Expensive research protocol, dependent on insulin dose and baseline RER |
| Standardized meal + calorimetry | Postprandial response to a defined mix | Result depends on the meal, time of day and preceding days |
| Calorimetric chamber | Adaptation over 24 h to food intake, rest, sleep and activity | Very limited access and sensitive to energy balance |
| Exercise test with gas analysis | Changes in fat and carbohydrate oxidation as intensity increases | No universal protocol for establishing “good” flexibility |
| Breath acetone | Trend in ketogenesis | Measures neither RER nor glucose use |
| CO₂-based consumer device | Proprietary estimate of a substrate mix | Without simultaneous oxygen measurement, this is not complete clinical calorimetry |
| Blood glucose, CGM or blood ketones | Availability or concentration of a fuel | A concentration does not reveal which fuel tissues are oxidizing |
Even ΔRER requires control of food intake and energy balance. The review by Galgani and colleagues shows that a high-fat diet or an energy deficit lowers baseline RER and can artificially enlarge the available range of change. Two results are comparable only if fasting, previous meals, activity, timing and protocol are also comparable.
A breath analyzer can therefore serve a narrowly defined experiment, not calculate flexibility. The guide to metabolic breath analyzers details the difference between acetone, CO₂ and respiratory quotient.
Exercise, diet and sleep: useful strategies
Exercise has the most direct evidence
Exercise increases energy demand rather than merely changing fuel supply. Endurance training improves oxidative capacity, intervals demand a shift toward greater carbohydrate use, and strength training maintains the muscle mass that stores and uses a large share of postprandial glucose.
In a controlled study involving 18 men with type 2 diabetes and 20 matched controls, twelve weeks combining two cycling sessions and one strength session per week improved flexibility measured by ΔRER in the diabetic group. It did not change significantly in controls, a reminder that an effect depends on the starting point. The study was not a randomized trial comparing Zone 2, HIIT and strength training.
The most honest classification is therefore functional:
- easy-to-moderate cardio to accumulate oxidative training volume you can recover from
- intervals to challenge high aerobic power and carbohydrate availability
- strength training to preserve strength, muscle and glycogen storage capacity
To schedule these intensities without turning Zone 2 into a metabolic test, see the comparison Zone 2 or HIIT · FR.
Diet should support adaptation, not manufacture a score
A very low-carbohydrate diet lowers RER and raises ketones. In a metabolic trial involving 17 men with overweight or obesity, switching to an isocaloric ketogenic diet lowered RQ by about 0.11 without accelerating body fat loss. The signal mainly showed a change in available fuels.
Conversely, following a so-called healthy diet for six weeks did not improve ΔRER or insulin sensitivity compared with a Western diet in a small trial of 40 adults with overweight or obesity, despite a greater reduction in fasting insulin and HOMA-IR. Several authors were affiliated with Unilever’s innovation center. A single marker therefore does not sum up a diet’s quality.
The defensible foundation is simpler: intake matched to expenditure, minimally processed foods, fiber, sufficient protein and carbohydrate intake suited to activity. Fasting and ketosis are not necessary.
Sleep mainly affects the underlying metabolic conditions
Direct trials using flexibility as the primary outcome are lacking. However, a meta-analysis of 41 randomized trials in healthy adults associated sleep restriction with reduced insulin sensitivity and an average increase in energy intake. That supports protecting sleep, not promising that an eight-hour night “repairs” a ΔRER.
Should you alternate carbohydrates and fats?
Not to train some supposed switch. Alternating a ketogenic day with a carbohydrate refeed mainly causes changes in glycogen, water, hunger and RER. This shows that oxidation follows fuel availability, not that a clinical benefit emerges.
A small randomized trial published in 2026 illustrates this in 24 male recreational runners. Four weeks at 50 g of carbohydrates or less per day increased fat oxidation. After four weeks of high carbohydrate intake, substrate use moved back toward its initial profile, with no clear performance advantage between groups. The small sample, self-reported diet and athletic population prevent this from becoming a general recommendation.
Periodization can nevertheless make sense in sport: more carbohydrates before an intense or long session, fewer when demand falls. It then serves training quality and recovery, not a flexibility score. “Train low” strategies remain advanced and can reduce power or contribute to insufficient energy availability if repeated.
The useful test goes both ways. Can you sustain an easy session without constant refueling, then complete a properly fueled intense session without treating increased carbohydrate use as a failure? This functional question makes more sense than constantly chasing ketones.
A cautious four-week observation protocol
This protocol does not measure metabolic flexibility. It checks whether a stable routine improves accessible outcomes without imposing fasting, ketosis or a sensor. Four weeks is enough to assess tolerability, not to establish a lasting physiological transformation.
Use this framework:
- Week 1, baseline: keep your usual habits and record wake-up time, sleep duration, hunger before meals, energy two hours afterward and performance during a standardized easy session
- Week 2, sleep: set a consistent wake-up time and allow a sleep window suited to your needs, without deliberately changing macronutrients
- Week 3, movement: add two easy sessions of 30 to 45 minutes and two full-body strength sessions of 10 to 20 minutes, with at least one recovery day between demanding efforts
- Week 4, fuel: keep meals regular and place a usual portion of carbohydrates around your most demanding session, without compensating with a ketogenic day
- Reassessment: compare weekly medians for hunger, energy, sleep, and pace or power at a comparable perceived effort
If you are already trained, replace one of the easy sessions with controlled, not maximal, intervals. If you are returning after a long period of inactivity, stick to conversational intensity throughout these four weeks.
Stop the experiment if you feel unwell or experience chest pain, abnormal breathlessness, unusual palpitations, compulsive eating behavior or a persistent drop in performance. Seek medical advice before increasing intensity if you have diabetes, cardiovascular, respiratory or kidney disease, are pregnant, take medication affecting blood glucose or experience symptoms during exercise.
The Verdict
Metabolic flexibility is a useful physiological framework, but a poor consumer score. Rigorous measurement compares controlled states. A morning acetone, glucose or CO₂ reading is not enough.
My editorial choice is to prioritize a progressive combination of endurance, intensity and strength, then stabilize sleep and diet. Fasting, ketogenic diets and carb cycling are context-dependent options, not prerequisites. If the question is medical, screen for metabolic risk with validated tests rather than trying to diagnose “inflexibility” at home.
