Health ArticleEducational review — not personal medical advice

The Energy Constraint Theory: A New Way to Understand How Exercise Improves Health

This review proposes a new explanation for why exercise is healthy. Our bodies keep total daily energy expenditure (TEE) within a narrow range.

23 min

Table of Contents

Key Points

  • The Constrained TEE model proposes the body keeps total daily energy expenditure within a narrow range, compensating for extra activity by reducing other functions.
  • In studies lasting 26 weeks or longer, energy compensation averaged about 78%, meaning most exercise calories are offset by reductions in other energy expenditures.
  • Moderate activity is linked to the lowest rates of chronic disease; extremely high workloads may compromise essential functions like immunity, increasing infection risk.
  • Chronic exercise is associated with lower pro-inflammatory markers such as C-reactive protein, and reduced stress-hormone reactivity, in the studies reviewed.
  • The U-shaped curve of health outcomes and the proposed mechanisms are predictions from a review, not proven facts, and require further testing.

Background: Why We Assume Exercise Burns Extra Calories

Exercise is essential for a healthy lifestyle. That much is not in dispute. In clinical and observational studies, greater physical activity has been shown to decrease death from any cause. Observational studies are research that follows people over time without assigning them a treatment. Greater physical activity also lowers the risk of cardiovascular disease (heart and blood vessel disease), Type 2 diabetes, many cancers, mental illness, and cognitive decline (declining memory and thinking skills). Conversely, people who spend more time inactive and sedentary have higher rates of cardiometabolic disease (conditions such as diabetes, obesity, and heart disease) and shorter lifespans.

For decades, researchers have been trying to figure out exactly why exercise produces these benefits. Several mechanisms have been proposed:

  • Cardiovascular effects: Exercise raises cardiac output (the amount of blood the heart pumps per minute). This strengthens the heart muscle and creates shear stress — a frictional force of flowing blood — on vessel walls. That stress increases production of nitric oxide, a molecule that promotes repair of the blood vessel lining and improves vascular elasticity (the ability of arteries to stretch and recoil).
  • Stress-axis effects: Chronic exercise blunts the responsiveness of the hypothalamus-pituitary-adrenal (HPA) axis, the hormone chain that controls the body's stress response. This reduces levels of cortisol, the primary stress hormone, and improves mental health.
  • Cognitive effects: The brain benefits appear to stem largely from increased blood supply and from exercise-induced production of neurotrophins (proteins that support the growth and survival of brain cells).
  • Metabolic effects: Chronic exercise also plays a critical role in regulating energy intake, insulin sensitivity (how well cells respond to insulin, the hormone that moves sugar out of the blood), and glucose trafficking (how the body moves and uses blood sugar).

Running alongside all of this has been a long-standing assumption. That assumption is that increased physical activity leads to a matching increase in total energy expenditure (TEE, measured in kcal per day). The assumption also holds that this extra calorie burn is itself one of the protective benefits of regular exercise.

At first glance, the assumption seems obviously true. In the immediate term, exercise has a profound effect on energy expenditure — metabolic rate rises by an order of magnitude (roughly tenfold) during strenuous exercise. A person may start a regular exercise program or make another lifestyle change. That change adds substantially to their typical workload. TEE usually does increase during the first weeks or months.

Key Finding: Activity and Daily Energy Expenditure Are Only Weakly Linked

Here is where the science gets surprising. Doubly labeled water studies have challenged the idea that high levels of habitual physical activity produce correspondingly high TEE. This technique measures TEE precisely during normal daily life by tracking the body's use of specially labeled water molecules.

Human populations around the globe have similar TEE regardless of lifestyle. Consider the evidence:

  • Luke and colleagues showed that women in a physically active farming community in Nigeria had similar TEE and similar activity energy expenditure (AEE). Women in a more sedentary urban community in the United States also had similar TEE and similar AEE. AEE is the portion of TEE not accounted for by basal metabolic rate (BMR, the calories burned at complete rest) or digestion costs.
  • The author and his colleagues found that TEE and AEE among physically active Hadza hunter-gatherers in northern Tanzania and Tsimane forager-farmers in Bolivia were similar to more sedentary populations. Those more sedentary populations were in the U.S., Europe, and Asia.
  • In an analysis of 332 adults from the Modeling the Epidemiological Transition Study (METS), daily physical activity measured by accelerometry was only weakly correlated with size-adjusted TEE. Above median levels of physical activity, there was no effect on TEE at all.
  • Among non-human primates and other species, captive populations have similar TEE to their wild counterparts.

The mismatch between habitual physical activity and TEE is not limited to observational studies. In controlled exercise intervention studies, human participants often show a much smaller increase in TEE than would be expected from the exercise workload they were assigned. Similarly, weight loss in exercise interventions is often less than predicted from the exercise workload — suggesting the increase in TEE (and the resulting calorie deficit) was smaller than expected. In many controlled laboratory studies of rodents and birds, TEE simply does not respond to increased physical activity.

The Constrained TEE Model Explained

These observations led the author to propose the Constrained TEE model. This model proposes that humans and other animals share a set of evolved mechanisms that keep TEE within a narrow range. When physical activity changes, the body dynamically compensates to keep daily energy expenditure in check.

This compensation doesn't happen instantly. It may take several weeks, or even months. So acute exercise bouts (single sessions) still raise energy expenditure in the immediate term — exercise does not become "free" of energy cost. Lifestyle changes that affect daily activity workload may also affect TEE until compensation kicks in.

But in the long term, physical activity will have little, if any, measurable effect on TEE. Instead, an organism's TEE is a product of its evolutionary history and, potentially, its developmental environment, and it is matched to the availability of food energy in that environment. TEE can rise or fall over evolutionary time — but habitual TEE does not correspond to habitual daily physical activity.

Energy Compensation: What the Studies Show

Energy compensation, also called "metabolic adaptation," is a commonly observed phenomenon. In this phenomenon, an organism responds behaviorally or physiologically to a change in activity or diet. The impact of that change is blunted. Here's a practical example: a person newly enrolled in an exercise regimen adds 200 kcal/day to their daily activity workload. That person may show an increase in TEE that is smaller than expected. This is an indication that some form of compensation reduced the metabolic impact of the exercise program.

Energy compensation in response to increased physical activity is common in exercise interventions. The degree of compensation varies among participants and studies. This variation has created uncertainty about the size and regularity of the effect.

Riou and colleagues examined energy compensation data from 61 studies involving 928 participants. They found that much of the variation in energy compensation could be attributed to three factors: study duration, and the age and body composition of the participants.

  • Duration matters: For interventions lasting 26 weeks or less, energy compensation was highly variable and generally lower than for longer interventions. For long-duration studies (26 weeks or more), energy compensation approached roughly 80% (mean: 78%; standard deviation: 23%; 16 study cohorts).
  • Body fat matters — but only in short studies: In shorter studies, participants with a greater body fat percentage showed less compensation, and this effect of body fat was strongest in young adults. The author suggests this may reflect a kind of buffering. Greater energy reserves (more fat) might blunt the physiological signals of energy stress. This blunting might slow or reduce the metabolic response.
  • Workload doesn't matter: Exercise workload, measured in kcal per week, had no effect on the degree of compensation.
  • Absolute compensation tracks exercise: In studies lasting 26 weeks or longer, the absolute amount of energy compensation (kcal/week) was strongly correlated with the intervention's exercise energy expenditure. Specifically, compensation equaled 0.70 ± 0.05 of exercise expenditure, with the statistical model adjusted r² = 0.95, p < 0.001 (meaning there is less than a 0.1% probability this result occurred by chance), for 12 degrees of freedom. This correlation was weaker but still statistically significant when shorter studies were included.

To put the compensation percentages in context: at 0% energy compensation, weight loss matches what was predicted from the imposed exercise; below 0% indicates greater-than-expected weight loss. At 100% energy compensation, there is no weight loss at all; above 100% indicates weight gain.

Notably, metabolic compensation appears to be much faster in rodents and birds — appearing in weeks rather than months. This suggests the rate of compensation might be related to body size or mass-specific metabolic rate. The factors affecting the timing and degree of compensation in humans and other species need further investigation.

Caveats and Unknowns in the Research

There is an important caveat to interpreting energy compensation data from exercise-based weight-loss interventions: TEE was not typically measured in these studies. Changes in energy intake and TEE could both contribute to energy compensation, and their relative contributions are usually unclear in this body of work. However, in cases where TEE was measured, changes in TEE often accounted for all or most of the observed compensation. The author calls for more direct measures of TEE in future exercise intervention studies.

Because energy intake was also not measured in these studies, the potential effects of calorie restriction on energy compensation remain uncertain. Still, energy compensation and TEE constraint are apparent in people who are weight stable and therefore in energy balance. This indicates that calorie restriction to the point of energy imbalance and weight loss is not necessary to induce compensation. Even so, some degree of food restriction could affect the timing and degree of compensation. For example, participants in weight-loss studies might limit their intake even if not directed to do so. In highly active, non-industrialized societies where food availability may be limited, constraints on energy availability could affect metabolic compensation.

The relative contributions of behavioral versus physiological changes to energy compensation also remain unclear. The evidence for behavioral compensation is mixed: studies in humans suggest that changes in non-exercise physical activity are insufficient to account for the observed degree of energy compensation. Instead, an emerging model suggests that changes in other non-musculoskeletal physiological activity — activity in the immune system, hormone axes, and reproductive system — contribute to energy compensation.

An Evolutionary Perspective on Modern Health

One of Charles Darwin's key insights was that the essential resources for growth, reproduction, and maintenance are limited — all species evolve under conditions of scarcity. For a given species, the allocation and prioritization of resources to various physiological tasks should reflect its evolved life-history strategy.

As a helpful way of thinking about this, the author divides activities into two categories:

  • Non-essential activities: functions that can be delayed or reduced without critical effects on evolutionary fitness (survival and reproduction).
  • Essential activities: functions that cannot be reduced without harming survival or reproduction.

This prioritization is most evident when resources are restricted. In a graded calorie-restriction study, Speakman and colleagues showed that male mice reduce the size of their heart, kidneys, liver, and spleen in response to an energy shortfall. The mice maintain the size of the brain and testes. Preserving reproductive capacity (testes) at the expense of organ maintenance and immune defense (spleen) is consistent with a life-history strategy that prioritizes fertility above longevity.

Energy prioritization during periods of energy shortfall in humans is not well studied but likely differs from short-lived species such as mice. In a recent study among Shuar children in rural Ecuador, Urlacher and colleagues found strong negative correlations between immune activity and linear growth. This suggests humans prioritize immune function over growth rate.

Now consider these changes in energy allocation within the Constrained TEE model, in which TEE is limited and increased physical activity expenditure forces corresponding reductions in other expenditures. Most — perhaps all — species should exhibit an evolved energy compensation strategy that reduces non-essential expenditures first, sparing essential functions as physical activity increases. Once non-essential reductions are exhausted, further increases in physical activity would lead to reductions in essential functions, compromising survival and reproduction.

Throughout the roughly 2-million-year evolutionary history of our genus, high levels of physical activity would have been the norm — just as they are today in small-scale hunter-gatherer and horticulturalist societies. Consequently, our physiology evolved under conditions in which non-essential expenditures were usually held in check, and fully engaged only during limited, sporadic periods of energy abundance and low physical activity. By contrast, in modern industrialized societies, humans live in a perpetual state of energy abundance and low physical activity. Non-essential physiological function is chronically high — a condition unseen in our evolutionary past.

Reducing energy expenditure in non-essential functions may therefore be an important mechanism underlying the health benefits of exercise in industrialized populations today. This hypothesis predicts that physical activity should downregulate (reduce) physiological activity elsewhere in the body, and that these reductions should map onto a continuum of health outcomes:

  1. At moderate levels of physical activity, the reduced non-essential expenditures should be associated with positive health outcomes.
  2. At extremely high levels of physical activity, where essential function is compromised, the result should be negative health outcomes.

The Immune System: A Double-Edged Sword

Immune system response to pathogens (disease-causing organisms such as bacteria and viruses) is essential for survival and involves a wide array of cell types, cytokines (signaling proteins), and other factors. The innate immune response — the non-specific, first-line defense — is thought to be particularly energetically costly. Regulation of immune activity and reactivity, particularly of the innate response, should therefore be sensitive to energetic cues.

In a state of energy abundance, immune response would be greatest, with copious resources allocated to any threat. Under energy stress, immune response would be muted, with a higher threshold for responding and fewer resources (cells, cytokines, and associated metabolic activity) directed toward a perceived challenge. The Constrained TEE model predicts that this energy-related regulation of immune function is mediated by physical activity.

Inflammation — an important and pervasive component of the innate immune response — is known to respond to physical activity.

  • During exercise: Inflammation increases in a dose-dependent manner with exercise intensity — in other words, the harder you exercise, the greater the immediate inflammatory response.
  • After exercise: The body enters an anti-inflammatory state. Chronic exercise and physical fitness are associated with lower levels of pro-inflammatory cytokines, including C-reactive protein, in both cross-sectional and longitudinal (exercise intervention) studies.

In a large cross-sectional study of 3,638 healthy U.S. adults, Abramson and Vaccarino found that men and women reporting four or more bouts of physical activity in the previous month had lower plasma levels of C-reactive protein and fibrinogen, and lower white blood cell counts, than those reporting three or fewer bouts.

The mechanisms behind exercise's anti-inflammatory effects are increasingly well understood:

  • Exercising muscle produces the pro-inflammatory cytokine interleukin (IL)-6. The transient (about 1 hour) elevation in IL-6 in turn stimulates production of the anti-inflammatory IL-10 and IL-1 receptor antagonist (IL-1A).
  • Exercise downregulates the number of pro-inflammatory monocytes (a type of white blood cell) and their expression of pro-inflammatory Toll-like receptors (TRL1, TRL2, and TRL4).
  • Cortisol and norepinephrine levels rise during exercise, which also has anti-inflammatory effects.
  • Exercise may reduce both the amount of visceral fat (fat stored around internal organs) and macrophage infiltration into fat tissue. This reduces production of pro-inflammatory cytokines and adipokines (signaling molecules released by fat tissue).

Lower inflammation has been proposed as a mechanism underlying the health benefits of exercise for some time. A wide range of chronic diseases have been linked to chronic inflammation, including cardiovascular disease, metabolic disease, some cancers, and cognitive dysfunction. Exercise can also help alleviate inflammation and symptoms associated with rheumatic autoimmune diseases. Rheumatic autoimmune diseases are conditions in which the immune system attacks the body's own tissues. The evidence for exercise as an autoimmune therapy is mixed. High levels of physical activity might even contribute to the low prevalence of allergy and asthma seen in small-scale, non-industrialized populations — a pattern usually attributed to childhood exposure to pathogens.

But the immune-suppressive effects of chronic exercise are not limited to the innate immune response; they can also affect acquired immune activity, such as T-cell and B-cell numbers and activity. At high levels of physical activity, downregulation of immune function can be severe enough to have negative consequences for health, diminishing the body's ability to fight infection. Compromised immune function is a well-documented component of "overtraining syndrome" in elite athletes. Large workloads of high-intensity exercise are associated with reduced white blood cell count, including reduced neutrophils, monocytes, and lymphocytes (T cells).

The contrast between the clear health benefits of moderate exercise and the increased incidence of illness at high exercise workloads has been called the "elite athlete paradox" — but the author notes it follows directly from the Constrained TEE model:

  • At moderate exercise workloads, physical activity expenditure leads to decreases in non-essential immune system activity (such as inflammation), producing reduced reactivity and a muted response to pathogens.
  • At higher workloads, physical activity expenditure reduces essential function, compromising the immune response and increasing the likelihood and severity of infection.

Stress Hormones and the Nervous System

The HPA axis and the sympathetic nervous system (SNS) are the body's primary mediators of physiological readiness and response to physical and psychological stressors. They increase energy availability by breaking down energy stores (catabolizing them) and by raising heart rate and blood pressure.

Acute activation of the SNS and HPA axis — the classic "fight-or-flight" response — is an ancient and essential evolved mechanism for meeting short-term challenges, and it promotes survival. However, chronic activation of these systems is associated with a range of negative health outcomes, including poor cardiovascular health, obesity, and mental illness.

HPA and SNS activity require energy to break down energy stores and maintain a higher heart rate and blood pressure. The Constrained TEE model predicts that HPA and SNS activity and reactivity should be sensitive to physical activity and energy availability. As with inflammation, cortisol (HPA) and norepinephrine (SNS) levels rise in a dose-dependent manner during exercise. But chronic exercise and greater physical fitness are associated with reduced reactivity and lower baseline levels of HPA and SNS activity.

The evidence supports this:

  • In an early meta-analysis (a study that pools results from many studies), Crews and Land demonstrated that the stress response was reduced among physically fit subjects.
  • In a recent comprehensive review, Silverman and Deuster describe a consistent body of evidence. That evidence shows that individuals with greater physical fitness exhibit less HPA and SNS reactivity to acute stressors. Greater physical fitness means greater long-term levels of daily physical activity.
  • Rimmele and colleagues reported that trained, physically fit men showed a smaller increase in cortisol and heart rate in response to a psychological stressor called the Trier Social Stress Test. Heart rate is a measure of SNS activity. Untrained, less fit counterparts showed a larger increase. Notably, the trained and untrained men had similar baseline cortisol levels and heart rates.
  • Intense physical training can also blunt the cortisol response to submaximal exercise.

Findings are mixed as to whether chronic exercise also blunts the cortisol awakening response (CAR) — the natural spike in cortisol that occurs shortly after waking. The effect may only be evident with large exercise workloads. A recent within-subjects study design (in which the same people are compared under different conditions) in recreational runners found a decrease in cortisol awakening response with greater exercise workload.

Why does this matter for patients? Blunted SNS and HPA reactivity are associated with better cardiovascular, metabolic, and mental health. One exercise intervention study by Nabkasorn and colleagues, involving 49 women with mild to moderate depression, nicely demonstrates the connections between exercise, energy compensation, and improved mental health outcomes.

Reproductive Function

The review also examines the reproductive system as a third area where energy compensation appears to operate. Low energy availability suppresses ovarian function in healthy women. In physically active, rural Nepalese farmers, seasonal increases in physical activity workload were associated with diminished ovarian function. Women in more rural, less economically developed populations have lower luteal phase progesterone levels. These populations are presumably more physically active. Progesterone is the hormone that prepares and maintains the uterus for pregnancy. In female athletes, large exercise workloads are associated with irregular menstrual cycling and amenorrhea (absence of menstruation).

These patterns are consistent with the core prediction of the Constrained TEE model. When energy is scarce or activity is high, the body reduces expenditures that are not immediately essential to survival. These expenditures include reproduction. This keeps total energy expenditure within its narrow, evolved range.

Clinical Implications: What This Means for Patients

The Constrained TEE model reframes how we should think about exercise and health. Rather than viewing exercise primarily as a way to "burn calories," the model suggests that the health benefits of exercise come largely from what the body stops doing when it reallocates energy.

The author describes the expected relationship between physical activity and health as a U-shaped curve:

  • Sedentary people carry the highest burden of chronic disease.
  • Moderately active people enjoy the lowest rates of chronic disease and the best health outcomes.
  • Extremely active people — for example, elite athletes — may face increased infection and other maladies, because essential functions are being compromised.

This also helps explain a common frustration among patients: why starting an exercise program often produces less weight loss than expected. In long-duration exercise studies, energy compensation averaged around 78%, meaning that most of the calories burned through exercise are offset by reductions in other energy expenditures. This is not a failure of willpower — it is an evolved, physiological response.

For patients with inflammatory or autoimmune conditions, this framing is also relevant. Exercise reliably lowers markers of chronic inflammation such as C-reactive protein. Chronic inflammation is linked to cardiovascular disease, metabolic disease, some cancers, and cognitive decline. However, the evidence for exercise as a therapy for rheumatic autoimmune diseases remains mixed, so it should be discussed with a treating clinician.

Limitations: What the Study Could Not Prove

This is a review article — a synthesis of existing evidence and a proposed theoretical model — not a new clinical trial. Several important limitations should be kept in mind:

  • TEE was rarely measured directly in weight-loss intervention studies. Because energy intake and TEE could both contribute to compensation, the relative contribution of each is usually unclear. Where TEE was measured, changes in TEE accounted for all or most of the compensation, but more direct measures of TEE are needed.
  • Energy intake was generally not measured in these studies, so the potential effects of caloric restriction on metabolic compensation remain uncertain.
  • Behavioral versus physiological compensation is unresolved. Evidence suggests changes in non-exercise physical activity are insufficient to explain the observed degree of compensation, but the exact contributions of each are not yet known.
  • Human data on energy prioritization during energy shortfall are sparse. Most of what we know about resource prioritization comes from short-lived species such as mice, which may differ from humans.
  • The effect of chronic exercise on the cortisol awakening response is inconsistent across studies and may only appear at large exercise workloads.
  • The rate of compensation varies enormously between species — weeks in rodents and birds versus months in humans — and the reasons for this are not established.
  • The hypotheses about the U-shaped curve of health outcomes are predictions, not proven facts. The author explicitly frames them as predictions that require further testing.

Recommendations for Patients

Based on this review, several practical messages emerge for patients trying to use exercise to improve their health:

  1. Aim for moderate activity, not extreme workloads. The health benefits of exercise appear to be greatest at moderate levels of physical activity, where the body reduces non-essential expenditures such as inflammation and stress reactivity.
  2. Don't expect exercise alone to produce dramatic weight loss. In studies lasting 26 weeks or more, energy compensation averaged about 78%. Exercise remains valuable for health, but weight change is a poor way to judge whether it is "working."
  3. Be aware of the risks of overtraining. Large workloads of high-intensity exercise are associated with reduced white blood cell counts, reduced neutrophils, monocytes, and T cells, and an increased risk of infection. If you are training heavily and getting sick more often, that may be a signal, not a coincidence.
  4. Give exercise programs at least six months. Energy compensation is inconsistent before roughly six months, so early results may not reflect the long-term picture.
  5. Talk to your clinician about your specific situation. If you have an autoimmune or rheumatic condition, cardiovascular disease, metabolic disease, or are pregnant or trying to conceive, the balance of exercise and energy availability may need individual guidance.

Frequently Asked Questions

Why doesn't exercise make me burn many more calories each day?

The Constrained TEE model proposes that the body keeps total daily energy expenditure within a narrow range. When you exercise more, it compensates by reducing other energy uses, so long-term physical activity has little measurable effect on total daily energy expenditure. In studies lasting 26 weeks or longer, energy compensation averaged about 78%.

Does this mean exercise is pointless for losing weight?

Exercise remains valuable for health, but weight change is a poor way to judge whether it is working. In long-duration studies, energy compensation averaged around 78%, meaning most calories burned through exercise are offset by reductions in other energy expenditures. This is an evolved physiological response, not a failure of willpower.

How long should I give an exercise program before judging results?

Energy compensation is inconsistent before roughly six months, so early results may not reflect the long-term picture. In studies lasting 26 weeks or more, compensation approached about 80%. Give an exercise program at least six months before drawing conclusions about its effects.

Can too much intense exercise be harmful?

Yes. Large workloads of high-intensity exercise are associated with reduced white blood cell counts, including reduced neutrophils, monocytes, and T cells, and an increased risk of infection. If you are training heavily and getting sick more often, that may be a signal, not a coincidence. The model predicts a U-shaped curve of health outcomes.

What does the U-shaped curve of exercise and health mean?

The model predicts that sedentary people carry the highest burden of chronic disease. Moderately active people enjoy the lowest rates of chronic disease and the most favorable health outcomes. Extremely active people may face increased infection and other maladies because essential functions are compromised. These are predictions, not proven facts.

Does exercise lower inflammation?

Chronic exercise and physical fitness are associated with lower levels of pro-inflammatory cytokines, including C-reactive protein. In a cross-sectional study of 3,638 healthy U.S. Among adults, those reporting four or more bouts of physical activity in the previous month had lower plasma levels of C-reactive protein and fibrinogen. These adults also had lower white blood cell counts than those reporting three or fewer bouts.

Can exercise affect my menstrual cycle or fertility?

Low energy availability suppresses ovarian function in healthy women. In physically active rural Nepalese farmers, seasonal increases in physical activity workload were associated with diminished ovarian function. In female athletes, large exercise workloads are associated with irregular menstrual cycling and amenorrhea. If you are pregnant or trying to conceive, discuss exercise and energy availability with your clinician.

If I'm training heavily and getting sick more often, when should I seek a second opinion about my exercise plan?

Large workloads of high-intensity exercise are associated with reduced white blood cell counts, including reduced neutrophils, monocytes, and T cells, and an increased risk of infection. This pattern is a documented component of overtraining syndrome. A second opinion can help assess whether your training load is compromising essential immune function, and whether your symptoms reflect overtraining rather than another cause. If you have an autoimmune or rheumatic condition, cardiovascular disease, metabolic disease, or are pregnant or trying to conceive, the balance of exercise and energy availability may need individual guidance. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

  • Original article title: Energy Constraint as a Novel Mechanism Linking Exercise and Health