Intermittent fasting: between ancestral physiology and modern medicine

What happens when the body goes through hours without food? How does alternating between eating and fasting reorganize metabolism? Why does the indication need to respect physiology, clinical evidence and individual safety?

For decades, the recommendation to eat every three hours was repeated as a strategy to keep the metabolism active. The current literature in human physiology shifts the focus away from meal frequency alone and toward the alternation between periods of energy availability and scarcity.

• The human body responds to cycles, energy sensors, hormones, sleep, body composition, microbiota, muscle and circadian rhythm.
• During the food break, insulin signaling drops, and with it part of the hormonal brake on fat breakdown eases. The liver mobilizes glycogen, lipolysis increases, and as the fast continues, fatty acids and ketone bodies start to play a bigger role in supplying energy.

This transition, called metabolic switching in the literature, explains why intermittent fasting has become relevant in modern metabolic medicine: it reorganizes the biological timing of nutrition and broadens the discussion beyond calorie counting. In practice, fasting should be understood as a judicious clinical tool, not a universal rule.

In clinical terms, intermittent fasting should be understood as a tool for:

• Organizing metabolic alternation between eating, digestive rest, mobilization of reserves, hydration and sleep.
• Reducing continuous food exposure, especially nighttime snacking, wide eating windows, and eating without real awareness of hunger.
• Favoring longer periods of low insulin signaling, as long as the eating window includes protein, fiber, micronutrients and a low load of ultra-processed foods.
• Integrating time, diet composition and circadian rhythm, without turning the strategy into extreme restriction.
• Preserving clinical safety, especially in women, children and adolescents, pregnant and breastfeeding people, people with diabetes, hypoglycemia risk, low weight, frailty, eating disorders, active cancer, or use of blood pressure, heart or glucose medications.

The biochemistry of fasting is consistent. The human outcomes observed in clinical trials are more modest. Fasting strategies can reduce weight and improve cardiometabolic markers like waist circumference, blood pressure, triglycerides, insulin resistance and glycemic control, mainly when they replace unstructured eating, nighttime snacking and frequent exposure to ultra-processed foods. Clinical application depends on the right indication, follow-up and preserved safety.

Intermittent fasting is most useful when it preserves protein, micronutrients, muscle mass, sleep, hormonal health and clinical safety. The protocol should aim for improved metabolic health with adherence, nutritional quality, emotional stability and low risk.

What it is

Intermittent fasting is a way of organizing time: periods of eating alternated with periods of food breaks. This strategy can coexist with different dietary patterns:

• Mediterranean diet;
• a higher-protein diet;
• well-planned vegetarian eating;
• a low-quality dietary pattern.

The clinical result depends on the nutritional quality within the window, including protein, fiber, micronutrients, plant diversity, fermentable foods for the microbiota, and low exposure to ultra-processed foods.

The clinical question broadens the focus from the clock to comprehensive care: which window allows for better eating, preserving muscle, sleeping well, training, controlling hunger, and reducing risk?

The main formats are described by the ratio between hours of food break and hours of eating window. For example, 12:12 means 12 hours of food break and 12 hours of eating window.

• 12:12: the safest entry point for organizing a routine, reducing nighttime eating, and restoring hunger awareness.
• 14:10: a good transition for people who tolerate longer gaps well without cutting protein too much.
• 16:8: a popular time-restricted eating format, useful for some adults when it preserves diet quality, protein and routine.
• 18:6: a more restrictive format, requiring more monitoring of protein intake, bingeing within the window, and training performance.
• 5:2: five days of usual eating and two days of strong, planned dietary restriction.
• Alternate-day fasting, ADF: alternating between eating days and days of fasting or intense restriction. Requires more planning, follow-up and individual tolerance.
• Occasional 24-hour fast: a one-off strategy for adapted adults, with guidance and low clinical risk.
• Prolonged fasts of 36 hours or more: belong to a different level of clinical selection. They can produce a deeper molecular response, but also greater risk of hypotension, hypoglycemia, lean mass loss, bingeing, dehydration, and worsening of pre-existing conditions.

ADA 2026 describes intermittent fasting as an umbrella term for ADF, 5:2 and time-restricted eating. The EASO, in its obesity guidelines, reinforces a logic that should guide any protocol: obesity and metabolic health require personalized assessment, broad clinical goals, and management that goes beyond weight alone. NICE and the American Heart Association (AHA) show the same caution: the method can be useful when integrated with diet quality, physical activity, sleep and safety.

Abbreviations: ADF = alternate-day fasting; ADA = American Diabetes Association; EASO = European Association for the Study of Obesity; NICE = National Institute for Health and Care Excellence; AHA = American Heart Association.

Hours of fasting

The physiology of fasting changes gradually. The body moves from the fed state to the post-absorptive state, reduces insulin, increases glucagon, mobilizes liver glycogen, expands lipolysis, and, with more time, increases production of ketone bodies.

As the fast progresses, the shift in energy substrate also changes cellular signals. The reduction in available glucose, the drop in insulin signaling, and the gradual rise in fatty acids and ketone bodies create a metabolic environment more geared toward mobilizing reserves, conserving energy, and adapting to cellular stress.

Beta-hydroxybutyrate works as both fuel and a signaling molecule, modulating pathways like NF-kB, BDNF, FGF21, NAD+, sirtuins, PARP1 and CD38. At the same time, the drop in glucose, amino acids and insulin reduces mTOR and activates AMPK, connecting fasting to autophagy, mitophagy and mitochondrial biogenesis.

This whole picture is what’s called metabolic flexibility: the ability to switch between glucose, fatty acids and ketone bodies according to energy availability, physical demand and hormonal state. In people with insulin resistance, visceral adiposity and a sedentary lifestyle, this alternation tends to become less efficient. Fasting can train part of this transition within a plan that also includes exercise, visceral fat reduction, diet quality and sleep.

The intensity of these effects depends on fasting duration, prior diet, muscle mass, physical activity, sleep, age, sex, body composition, insulin resistance, medications, and overall health. A 16-hour fast in a sedentary person with low protein intake offers a very different stimulus than a complete metabolic protocol.

Insulin is essential for life, protein synthesis, energy storage and glycemic control. The clinical problem arises from the combination of chronic energy excess, insulin resistance, ultra-processed eating, poor sleep, visceral adiposity and compensatory hyperinsulinemia. Eating intervals help part of this picture when inserted into broad metabolic care.

Evidence and guidelines

A network meta-analysis published in BMJ in 2025, is one of the most comprehensive clinical syntheses on the topic. It brought together 99 randomized clinical trials and 6,582 adults, comparing time-restricted eating, alternate-day fasting, whole-day fasting, continuous caloric restriction and free eating. Fasting strategies and continuous caloric restriction reduced weight compared to free eating, with similar average results across most outcomes.

Alternate-day fasting showed a small advantage over continuous caloric restriction, around 1.29 kg more weight loss. For HbA1c and HDL, results were similar between fasting, continuous caloric restriction and free eating.

Fasting works best when it changes eating behavior sustainably:

• it reduces spontaneous food intake while preserving nutritional quality;
• it eliminates nighttime snacking and impulsive eating;
• it improves adherence because it simplifies timing;
• it reduces ultra-processed foods within the window;
• it preserves protein and resistance training;
• it fits into sleep and real-life routine.

Reading the guidelines and the main studies is more useful when separated by clinical function:

• ADA 2026. ADF, 5:2 and time-restricted eating can produce a loss of 3% to 8% of initial weight in short studies, with results close to continuous caloric restriction. In obesity management, losses of 5% to 7% already improve glucose and intermediate risk factors; above 10%, the benefits tend to be greater.
• NICE 2025 and AHA 2021. Fasting can be considered individually, without replacing the main pillars of care: quality dietary pattern, physical activity, sleep, safety and adherence.
• EASO 2024. Obesity should be treated as a chronic, adiposity-based disease, with assessment of complications, function and quality of life. The practical question becomes: which strategy improves metabolic health, preserves muscle, reduces risk, and fits the patient’s life?
• NEJM. The mechanistic basis organizes the concept of metabolic switching: ketones as fuel and signal, AMPK, mTOR, sirtuins, autophagy and mitophagy. The clinical trial with an 8-hour eating window showed weight loss and metabolic improvement similar to caloric restriction with free timing. The mechanism is plausible; the clinical result depends on context, adherence and diet quality.
• Nature. The studies help separate a specific fasting protocol from casual fasting: cycles of the fasting-mimicking diet were associated with liver and blood changes consistent with lower biological risk; seven days of fasting preserved maximal strength, with a drop in peak VO2, reduced glycogen, and lean mass loss; in breast cancer, the translational pathway remains restricted to research or a specialized oncology team.

A clinical trial showed that an early eating window, ending at 3pm, improved insulin sensitivity, blood pressure, oxidative stress and appetite in men with prediabetes, even with stable weight. This result suggests that meal timing can influence metabolic markers. For many patients, having dinner with family, training later in the day, or protecting adherence also influences the choice of protocol.

Abbreviations: BMJ = British Medical Journal; NEJM = New England Journal of Medicine; TRE = time-restricted eating; HbA1c = glycated hemoglobin; HDL = high-density lipoprotein; FMD = fasting-mimicking diet; peak VO2 = maximal oxygen consumption at peak effort.

Autophagy and repair

Autophagy is a cellular recycling system. The cell identifies damaged or dysfunctional components, wraps this material in autophagosomes, and directs it to lysosomes for degradation and reuse. Mitophagy is the version focused on mitochondrial quality control. It removes damaged mitochondria, reduces the buildup of dysfunction, and takes part in cellular energy efficiency.

Fasting, dietary restriction and exercise can modulate these pathways through AMPK, mTOR, sirtuins and energy availability. The review by de Cabo and Mattson in NEJM describes this axis clearly: less available glucose, fewer available amino acids, and less insulin reduce mTOR and activate AMPK, favoring responses of conservation, repair and adaptation.

This same review describes that cells exposed to fasting cycles can trigger a coordinated adaptive response, with increased antioxidant defenses, DNA repair, protein quality control, mitochondrial biogenesis, autophagy, and lower inflammatory signaling. The clinical magnitude in 12- or 16-hour fasts depends on the tissue, fasting duration, metabolic state, and the person’s context.

The clinical interpretation separates mechanism from demonstrated outcome. The magnitude in humans, especially in short fasts, varies a great deal and is hard to measure directly in the tissues of interest.

Longer fasts and fasting-mimicking diets enter different territory. Preclinical studies showed, in animal models, signs of hematopoietic regeneration, reduced IGF-1, and modulation of stem cells. In humans, cycles of the fasting-mimicking diet were associated with improved insulin resistance, liver fat, and blood markers linked to disease risk. These findings belong to specific protocols, with planned restriction, participant selection, and clinical evaluation.

The Yilmaz group at MIT showed in Nature that stem-cell-mediated intestinal regeneration happens mainly during refeeding, not during the fast itself. Upon resuming eating after a 24-hour fast in mice, mTORC1 was intensely activated in intestinal stem cells, increasing protein synthesis via polyamine metabolism and driving proliferation and repair of the epithelium. This finding reinforces the logic of metabolic alternation: fasting mobilizes reserves and activates conservation pathways; refeeding triggers repair. The post-fast period isn’t a neutral moment: cells are dividing faster, metabolically more active, and more sensitive to stimuli, both regenerative and potentially harmful. In mice, activating an oncogene during this proliferation window increased the formation of pre-cancerous polyps, indicating that post-fast regeneration is a phase of both benefit and greater genetic vulnerability.

The correct message for the patient is: fasting can take part in repair and metabolic flexibility pathways. The degree of clinical benefit depends on context. Exercise, sleep, protein intake, diet quality, and visceral fat reduction remain the determining factors.

Exercise and muscle

During fasting, growth hormone, GH, can increase. This increase is mainly adaptive: it helps mobilize fat, preserve glucose, and get through the period of low energy intake. Hypertrophy depends on mechanical stimulus, amino acids, energy, and recovery.

Muscle mass depends on three pillars:

• mechanical stimulus from progressive resistance training;
• essential amino acids, especially enough protein spread through the eating window;
• energy and recovery, including sleep, micronutrients and rest.

During fasting, muscle protein synthesis tends to be limited by the absence of amino acids. The body can preserve strength for short periods, while protecting lean mass requires protein, resistance training, and enough energy within the eating window. A study published in Nature Communications helps calibrate the interpretation: after seven days of fasting, volunteers maintained maximal strength, but lost lean mass and fat; peak VO2 dropped 13%, muscle glycogen was cut in half, and capacity for prolonged, high-intensity exercise worsened.

In practical terms, training while fasted can work for walking, mobility, light exercise, or moderate aerobic activity for a short time, especially in people who are already adapted. When training involves heavy weightlifting, HIIT, running, cycling, long endurance sessions, athletes, women with an irregular cycle, or patients losing weight, the priority shifts: preserving performance, protein intake, recovery, and lean mass.

The protocol should protect muscle. In modern metabolic medicine, muscle is an endocrine organ, a functional reservoir, a determinant of insulin sensitivity, and protection against frailty. Loss of lean mass reduces the quality of the clinical outcome.

Women and the cycle

In women, fasting needs to be analyzed more carefully. The clinical risk concentrates in the combination of fasting, dietary restriction, low protein intake, low body fat percentage, excessive training, stress, poor sleep, and chronic low energy availability.

Low energy availability means insufficient energy to sustain training, basal metabolism, thermoregulation, immunity, hormone synthesis, and reproductive function. The IOC’s consensus on REDs describes the consequences of prolonged low energy availability in both sexes, including menstrual changes, bone health, immunity, metabolism, cardiovascular function, mood, and performance.

For women of reproductive age, warning signs include:

• cycles that become longer, irregular, or disappear;
• worsening PMS, sleep, irritability, or bingeing;
• excessive cold sensitivity, hair loss, persistent fatigue;
• reduced performance or increased injuries;
• intense cravings for food after the window;
• low libido or a feeling of exhaustion;
• a history of eating disorders or a rigid relationship with food.

The review by Mao, Liu and Zhang describes a relevant point: in women with overweight, obesity, or PCOS, some fasting protocols, like 5:2 or TRF, can improve free androgen index and SHBG, with possible favorable impact on menstrual regularity. In pregnancy, low weight, or the absence of metabolic excess or energy vulnerability, greater caution is warranted.

So the clinical decision considers: which woman, at what life stage, with what body composition, what cycle, what training, what sleep, what protein intake, what history, and what goal?

Pregnant and breastfeeding people, adolescents, women with amenorrhea, low weight, eating disorders, very irregular cycles under investigation, athletes under heavy load, or patients with symptoms of low energy require clinical evaluation before any fasting protocol.

Clinical contexts

In obesity and insulin resistance, intermittent fasting can be useful when it organizes eating behavior, reduces nighttime intake, improves adherence, and facilitates an energy deficit while preserving nutritional quality. The EASO framework reinforces that obesity is a chronic, adiposity-based disease and should be managed with personalized assessment, function, complications, quality of life, and metabolic health. In this context, fasting is a possible tool within broader care.

In prediabetes and type 2 diabetes, the main issue is safety. When insulin, sulfonylureas, or other drugs with hypoglycemia risk are in use, fasting requires adjustment and monitoring. In users of SGLT2 inhibitors, there’s additional concern about euglycemic ketoacidosis, especially in scenarios of low intake, acute illness, dehydration, or intense carbohydrate reduction. The IDF-DAR guidelines for fasting during Ramadan are useful because they treat fasting as a clinical situation requiring risk stratification, education, hydration, and clear stopping criteria.

In cardiovascular risk, fasting needs to go hand in hand with what has the strongest evidence: a cardioprotective dietary pattern, fewer ultra-processed foods, blood pressure control, lipids, glucose, sleep, physical activity, and visceral fat reduction. People using medications for blood pressure, arrhythmia, heart failure, coronary disease, or anticoagulation deserve individualized guidance, because fasting can change hydration, blood pressure, exercise tolerance, medication timing, and symptom risk.

In oncology, the assessment requires nutritional rigor and clinical selection. Studies with fasting, short-term fasting, and fasting-mimicking diets investigate modulation of IGF-1, insulin, glucose, leptin, adiponectin, inflammatory pathways, and therapeutic sensitivity in specific models. A study published in Nature in 2026, shows mechanisms involving the fasting-mimicking diet, endocrine therapy, and hormone-receptor-positive breast cancer, with data from animal models and clinical samples.

Active cancer is a high nutritional risk context. Involuntary weight loss, sarcopenia, cachexia, inflammation, nausea, mucositis, chemotherapy, surgery, and radiotherapy completely change the risk-benefit balance. The EASO position on nutrition therapy in overweight, obesity, and cancer reinforces that the evidence for strategies like intermittent fasting is still limited and inconsistent, requiring personalized plans.

In oncology, fasting should be considered only in research protocols or with an experienced oncology and nutrition team, with rigorous selection, protein preservation, and monitoring of weight, lean mass, symptoms, tests, and ongoing treatment.

Children, adolescents, pregnant and breastfeeding people, frail older adults, people with diabetes, individuals at risk of hypoglycemia, people using cardiovascular medications, and patients with greater clinical vulnerability require an extra layer of protection. In these groups, the decision depends on diagnosis, medications, nutritional status, hydration, symptoms, dietary routine, and the real possibility of follow-up.

Protocols and safety

This section is for educational purposes. The sequence below is a practical editorial synthesis by InovaSaude, built from reading NICE 2025, ADA 2026 and IDF-DAR. It doesn’t replace medical evaluation and shouldn’t be applied on your own: your doctor needs to assess your health status, tests, medications, routine, body composition, symptoms, and goals to decide whether fasting is a suitable, beneficial tool for you. These references support the principles of individualization, glycemic safety, preserving diet quality, symptom monitoring, and clear criteria for stopping the fast; they don’t publish a single protocol with these phases.

Phase 1: 12:12 for 7 to 14 days. Reduce nighttime eating, set a time for the last meal, maintain hydration, and preserve a first meal with protein. This phase already resolves an important part of snacking.

Phase 2: 14:10 for 2 to 4 weeks. Assess hunger, sleep, mood, bowel movements, bingeing, training, menstrual cycle, and protein intake. A drop in protein, worsening sleep, or compensatory eating within the window indicate a need to adjust the protocol.

Phase 3: 16:8 on 2 to 5 days a week. Use when there’s good tolerance, complete meals, stable energy, no bingeing, and no hypoglycemia symptoms. Weekly frequency should be individualized.

Phase 4: occasional 24 hours. A strategy for adapted adults, with a stable routine, guidance, and low clinical risk.

36 hours or more. A strategy requiring high clinical selection, with individual assessment and a clear goal. For most patients, increasing diet quality and resistance training offers a better benefit-risk ratio.

During the eating window, the priority should be:

• enough protein to preserve lean mass;
• vegetables, legumes, fruit, and fiber;
• good-quality fats;
• hydration and electrolytes when needed;
• a low load of ultra-processed foods;
• a post-workout meal when there’s intense training;
• preserved sleep.

Fasting should be stopped immediately in the face of mental confusion, tremors, cold sweat, palpitations, intense weakness, vomiting, fainting, chest pain, symptomatic hypotension, or hypoglycemia symptoms. In people with diabetes, blood glucose below 70 mg/dL requires stopping and correction according to clinical guidance.

The following require professional evaluation before starting fasting:

• pregnant and breastfeeding people;
• children and adolescents;
• frail older adults or those with sarcopenia;
• people with low weight or involuntary weight loss;
• current or past history of an eating disorder;
• type 1 diabetes;
• diabetes treated with insulin, sulfonylureas, or SGLT2 inhibitors;
• use of medications for blood pressure, heart, arrhythmia, or anticoagulation;
• advanced kidney disease;
• significant liver disease;
• active cancer;
• recent postoperative period;
• post-bariatric surgery;
• symptomatic hypotension or syncope;
• migraine triggered by fasting;
• use of medications that require food intake;
• amenorrhea, very irregular cycles, or suspected low energy availability.

Conclusion

Intermittent fasting is best understood as a metabolic alternation strategy. Periods without energy intake activate pathways of conservation, fat mobilization, and adaptation. Periods of adequate eating allow synthesis, rebuilding, training, recovery, and lean mass preservation.

The mechanisms described include a drop in insulin, glycogen mobilization, increased lipolysis, ketogenesis, signaling by beta-hydroxybutyrate, AMPK, mTOR, sirtuins, autophagy, and mitophagy. In human clinical trials, the average benefits on weight and cardiometabolic markers are modest. When compared to continuous dietary restriction, the average results tend to be similar.

This difference between mechanism and outcome defines fasting’s place. Intermittent fasting can reduce snacking, improve hunger awareness, organize routine, favor lower post-meal exposure, and support fat loss in selected people. The right indication protects the menstrual cycle, energy, training, lean mass, and glycemic safety.

The best protocol is the one that’s most sustainable, safe, and integrated with sleep, exercise, protein, micronutrients, hormonal health, and diet quality. The central strategy is to restore a physiological alternation between nutritious eating and metabolic rest.

At InovaSaude, intermittent fasting is presented as a possible clinical tool within a bigger plan for metabolism, autonomy, and individualized care.

This article is for educational purposes and serves as support for a qualified clinical conversation. Fasting protocols should be defined with follow-up from specialized health care teams, especially for people with chronic diseases, medication use, hypoglycemia risk, pregnancy, breastfeeding, frailty, eating disorders, active cancer, or any condition of greater clinical vulnerability.