Diets in Oncology

THE UNDERLYING IDEA

In animal models, short-term starvation protected healthy cells from chemotherapy toxicity while increasing its effect on cancer cells ^[1]. Two variants have been tested in humans: true fasting (several dozen hours on water only, around chemotherapy administration) and the fasting-mimicking diet, a several-day menu very low in calories and protein, designed to make the body respond as it would to a fast ^[1][2].

WHAT THE LARGEST RANDOMISED TRIAL SHOWED

The DIRECT trial (phase 2, 131 patients with HER2-negative stage II/III breast cancer) compared a fasting-mimicking diet used for 3 days before and during neoadjuvant chemotherapy with a regular diet ^[1]. Results:

• no difference in toxicity between the groups, despite dexamethasone being omitted in the fasting arm ^[1];
• a radiologically complete or partial response occurred more often in the diet group (OR 3.168; p=0.039) ^[1];
• in the per-protocol analysis, a Miller-Payne 4/5 pathological response, meaning 90-100% tumour-cell loss, was more likely (OR 4.109; p=0.016) ^[1];
• the diet reduced chemotherapy-induced DNA damage in T-lymphocytes ^[1].

WHY THIS IS STILL NOT ENOUGH

The key pathological-response result comes from a per-protocol analysis rather than from all randomised patients, meaning it counted only those who actually kept to the diet ^[1]. Such an analysis inherently favours the tested method, because patients able to tolerate several days of food restriction are usually in better general condition. An independent 2025 systematic review covering 9 studies and 354 patients concluded plainly that fasting is safe and feasible, but that no effect on treatment outcomes or chemotherapy-related toxicity has been demonstrated ^[2]. The only consistent finding was a fall in insulin and IGF-1, a change in a biological marker rather than in the patient's health ^[2]. The authors explicitly call for large randomised trials before fasting is regarded as an adjunct to therapy ^[2].

WHAT THESE STUDIES DID NOT COVER

DIRECT excluded people with diabetes and those with a BMI below 18 ^[1]. The available studies were dominated by breast cancer patients (258 of 354), generally in good condition and not losing weight ^[2]. This means the data say nothing about fasting in people who are malnourished, losing weight, in cancer cachexia, or with gastrointestinal cancers - and these are precisely the groups most likely to consider restricting food after reading reports about fasting. Absence of data is not the same as evidence of safety.

WHAT THIS MEANS FOR THE READER

Fasting and fasting-mimicking diets remain an investigational approach, not part of standard cancer treatment ^[2]. Maintaining body weight and protein intake during chemotherapy is supported by strong evidence (see the entries on nutritional support and on cachexia), whereas deliberate food restriction is not. Any change to nutrition during cancer treatment is a decision to be made together with the treating team.

Sources

1. de Groot S, et al. — Fasting mimicking diet as an adjunct to neoadjuvant chemotherapy for breast cancer in the multicentre randomized phase 2 DIRECT trial, Nature Communications 2020: europepmc.org ↗
2. Maes J, et al. — Impact of intermittent fasting on patients with cancer undergoing chemotherapy and/or targeted therapies: a systematic review of the literature, Supportive Care in Cancer 2025: europepmc.org ↗

WHY PEOPLE TURN TO IT

The alkaline diet and alkaline water are promoted by the media and by sellers as a way to "de-acidify the body" and thereby hinder cancer growth. The premise is that because cancer cells grow less well in an alkaline environment, the pH of the whole body can be shifted by what a person eats and drinks.

WHAT PHYSIOLOGY SAYS

A systematic review published in BMJ Open documented that dietary changes can alter URINE pH but do not change BLOOD pH. Dietary interventions moved urine pH by 0.2–1.2 units, while blood pH changed by only 0.01–0.02 units, because the kidneys excrete acid effectively and maintain systemic pH balance ^[1]. The premise on which the whole idea rests therefore fails at the level of physiology.

WHAT THE EVIDENCE SAYS

The same review identified 8278 citations and reviewed 252 abstracts. Exactly ONE study met the inclusion criteria ^[1].

• No randomised trials were located.
• No study was found that examined alkaline diet or alkaline water for cancer TREATMENT.
• The single included study found no association between dietary acid load and bladder cancer (OR 1.15; 95% CI 0.86–1.55; p = 0.36) ^[1].

WHAT THIS MEANS FOR THE READER

The authors write that despite promotion of the alkaline diet and alkaline water by the media and salespeople, there is almost no actual research to either support or disprove these ideas, and promoting them for cancer prevention or treatment is not justified ^[1]. "No high-quality evidence" is a complete and honest answer here — neither a promise nor silence replaces it.

It is also worth noting what is at stake beyond the water itself: restrictive versions of the alkaline diet limit protein and grain foods, and weight loss and malnutrition during cancer treatment are documented clinical problems ^[2]. Whether and how to modify eating during cancer is decided by the treating physician together with a clinical dietitian.

Sources

1. [1] Fenton TR, Huang T. Systematic review of the association between dietary acid load, alkaline water and cancer. BMJ Open 2016;6:e010438: pmc.ncbi.nlm.nih.gov ↗
2. [2] NCI PDQ — Nutrition in Cancer Care (Health Professional Version): cancer.gov ↗

WHY IT IS USED

Radioiodine (I-131) is taken up by thyroid cells in exactly the same way as dietary iodine - the two forms compete for the same transport mechanism. The reasoning behind the diet is therefore simple: lowering the pool of non-radioactive iodine before administration should direct more of the dose to where it is meant to act. The diet does not treat the cancer; it is a preparation for treatment and applies only for a defined period before the isotope is given, not permanently.

WHAT IS USUALLY AVOIDED

Restrictions concern mainly the richest dietary sources of iodine:

• iodised salt and processed foods salted with it,
• sea fish, seafood and seaweed (nori, wakame, kombu),
• milk and dairy products, and egg yolks,
• the food colouring E127 (erythrosine) and iodine-containing supplements,
• iodine-containing medicines and iodinated contrast agents - the latter matter far more than food.

WHAT THE EVIDENCE SHOWS

A 2010 systematic review covered 8 studies in which the diet lasted from 4 days to 4 weeks. It reduced urinary iodine excretion in all of them and was associated with increased radioiodine uptake in three. The authors noted, however, that not a single study assessed recurrence or mortality, and cautiously supported the practice of a 1-2 week diet before I-131 administration ^[1].

A 2022 systematic review with meta-analysis (59 reports) framed this more critically. Patients with iodine deficiency (urinary iodine below 50 µg/L) had a higher chance of successful ablation than those with iodine excess (at least 250 µg/L): odds ratio 2.63 (95% CI 1.18-5.86; n=283). By contrast, comparing a STRICTER with a less strict diet showed no meaningful difference: odds ratio 0.67 (95% CI 0.26-1.73; n=256). The certainty of both results was rated as very low ^[2].

The 2022 consensus statement of the European Thyroid Association puts it plainly: a low-iodine diet "may be prescribed but its utility is not demonstrated unequivocally", while any iodine-containing drug must be avoided. The document describes the practice of some centres - restriction to no more than 50 µg of iodine per day for 1-2 weeks - noting that evidence on the effect on ablation success is conflicting ^[3].

STRENGTH OF EVIDENCE

Limited. What is well documented concerns surrogate measures: urinary iodine concentration and isotope uptake ^[1][2]. What matters most to the patient - whether the diet increases the chance of cure and lowers the risk of recurrence - was not examined in any study included in the 2010 review ^[1], and the 2022 meta-analysis rates the certainty of the available results as very low ^[2]. This does not mean the diet is useless; it means that a strong recommendation about how strict or how long it should be has no basis in current data.

THE COST TO THE PATIENT - PART OF THE RESULT, NOT A FOOTNOTE. The 2022 review also covered qualitative studies and described the diet's effect on wellbeing: patients called it boring, confusing and overwhelming, and reported anxiety, uncertainty about permitted foods and self-blame over possible lapses - at an already stressful time ^[2]. Because the demonstrated benefit concerns surrogate measures while the cost is real and felt, tightening restrictions "just in case" is not a neutral choice.

WHAT THESE RESULTS DO NOT UNDERMINE

They do not undermine the need to avoid iodinated contrast agents and iodine-containing medicines before radioiodine treatment - these deliver iodine on a completely different scale from a meal, and the European statement lists avoiding them as an unconditional requirement ^[3].

WHAT TO DISCUSS WITH THE TREATING TEAM

The scope and duration of the diet, when to start it and how to proceed after contrast-enhanced imaging are decided by the treating team - practice differs between centres and also depends on how much iodine the national diet contains. Diagnostic and therapeutic decisions are made solely by a specialist physician.

WHY THIS ENTRY STANDS ON ITS OWN

Most nutrition content in oncology concerns maintaining weight and preventing malnutrition. In sellar region tumours the problem can be the reverse, and transferring the general principle of "eat more so you do not lose weight" here is not merely inapt but potentially harmful.

WHAT THE SOURCE SAYS

• in craniopharyngioma, obesity may occur that can be life-threatening, together with metabolic syndrome including non-alcoholic fatty liver disease ^[1];
• obesity develops in children after both complete and subtotal resection, which the source says suggests that a predilection to obesity is a component of the disease itself rather than solely a consequence of direct hypothalamic injury ^[1];
• posterior hypothalamic involvement by the tumour, or operative injury to it, appeared to be a key factor in the development of severe obesity ^[1];
• separately described are hormone deficiencies — growth hormone, thyroid and cortisol — and the almost universal need for lifelong replacement of multiple pituitary hormones ^[1].

WHY THIS IS NOT ORDINARY OBESITY

The hypothalamus regulates satiety and energy expenditure. When this region is involved by tumour or damaged during treatment, the satiety signal and metabolic rate cease to track the amount of food eaten. The result is weight gain that cannot be explained by eating pattern — which is why attributing it to "lack of discipline" is both unjust and clinically misleading.

A SECOND LAYER - HORMONES

Hypothyroidism and growth hormone deficiency in themselves alter body composition and energy expenditure, and cortisol deficiency is a state of danger. Assessment of body weight after treatment of a tumour in this region therefore cannot be separated from endocrine assessment; that is a task for an endocrinologist, not a question of choosing a diet.

WHAT THIS ENTRY DOES NOT SAY

We give no diet, no calorie figure and no plan of action. We know of no high-quality evidence that any particular pattern of eating reverses hypothalamic obesity; the cited source does not supply it, and we do not substitute conjecture. The general NCI recommendation for survivors of childhood cancer is stated broadly: healthy dietary practices — a diet rich in plant foods and moderate in animal foods — and an active lifestyle, to reduce treatment-related metabolic and cardiovascular complications ^[2].

THE DECISION BELONGS TO THE DOCTOR

Nutritional management and hormone replacement after surgery for a sellar region tumour are decided by the treating team — neurosurgeon, endocrinologist and clinical dietitian. This page describes the state of knowledge and is not advice for any individual.

Sources

1. [1] National Cancer Institute — Childhood Craniopharyngioma Treatment (PDQ), Health Professional Version (otyłość zagrażająca życiu i zespół metaboliczny z NAFLD; otyłość po resekcji doszczętnej i niedoszczętnej jako możliwa składowa samej choroby; zajęcie/uszkodzenie tylnego podwzgórza jako kluczowy czynnik ciężkiej otyłości; niedobory GH, tarczycy i kortyzolu; dożywotnie uzupełnianie hormonów przysadki): cancer.gov ↗
2. [2] National Cancer Institute — Late Effects of Treatment for Childhood Cancer (PDQ), Health Professional Version (ogólne zalecenie: dieta bogata w produkty roślinne i umiarkowana w produkty zwierzęce oraz aktywny tryb życia w celu zmniejszenia metabolicznych i sercowo-naczyniowych następstw leczenia): cancer.gov ↗

WHY THIS LOCATION IN PARTICULAR

NCI states explicitly that nutrition complications are usually most notable and severe with tumors involving the digestive tract or head and neck, owing to mechanical obstruction or dysfunction ^[1]. For the oesophagus and stomach NCI lists dysphagia and odynophagia as typical consequences, and describes tumour-induced weight loss as occurring frequently in solid tumours of the lung, pancreas and upper gastrointestinal tract, and less often in breast cancer or lower gastrointestinal cancer ^[1]. Surgery adds a further mechanism: NCI notes that procedures including oesophagectomy and gastrectomy may produce early satiety, a premature feeling of fullness ^[1].

THE SCALE OF THE PROBLEM AFTER CHEMORADIOTHERAPY

A secondary analysis of a prospective cohort of 260 patients with oesophageal squamous cell carcinoma treated with concurrent chemoradiotherapy between 2008 and 2024 found that cachexia developed after treatment in 60.7% of patients ^[2]. Independent predictors of cachexia were greater tumour length and a larger tumour-occupying proportion of the lumen, with cut-offs of at least 6 cm in length and at least 70% of the luminal area ^[2].

PRETREATMENT FEEDING STOMA — WHO ACTUALLY BENEFITED. This is the crux of the topic and the part most often oversimplified. Across the cohort as a whole, creating a gastrostomy or jejunostomy before chemoradiotherapy was NOT associated with longer progression-free or overall survival ^[2]. Benefit appeared only in the subgroup whose tumour occupied at least 70% of the lumen: less cachexia, better treatment response (odds ratio 2.78; P=0.028), longer median progression-free survival (6 vs 4 months; P=0.012) and longer median overall survival (11 vs 9 months; P=0.009) ^[2]. In patients with a smaller tumour-occupying proportion no benefit was found ^[2]. The authors draw a cautious conclusion: the tumour-occupying proportion may serve as a selection criterion for this invasive procedure, rather than as an argument for performing it in everyone ^[2].

STRENGTH OF EVIDENCE

Limited. In favour: the data come from a large, prospectively collected cohort with long follow-up, and tumour measurements were taken from standardised endoscopic images ^[2]; the statements on nutritional mechanisms and on early satiety after surgery rest on NCI's review ^[1]. Against full certainty: the key finding on feeding stomas comes from a SUBGROUP ANALYSIS of a non-randomised study — treating clinicians decided who received a stoma, so differences between groups may partly reflect that choice. The study concerns squamous cell carcinoma; evidence of comparable strength for oesophageal adenocarcinoma is lacking.

WHAT THIS ENTRY DOES NOT SAY

It does not indicate whether any individual patient should have a gastrostomy or jejunostomy, when, or in what form. That decision is made by the treating team — medical oncologist, radiation oncologist, surgeon and clinical dietitian — on the basis of endoscopic findings, nutritional status and the treatment plan. The service does not interpret individual clinical situations or test results.

Sources

1. [1] National Cancer Institute — PDQ: Nutrition in Cancer Care (Health Professional Version): cancer.gov ↗
2. [2] Pretreatment feeding-stoma placement in advanced esophageal cancer: defining optimal patient selection criteria. Esophagus 2026 (PMID 42165951, PMC13319426): europepmc.org ↗

WHAT HAPPENS AND WHY

Taste and smell changes are among the most frequently reported complaints during cancer treatment. Food may seem to have no flavour, meat may taste bitter or metallic, and food odours can become hard to tolerate. NCI lists as causes the treatment itself (chemotherapy, radiotherapy to the head and neck, immunotherapy), dental problems, and the cancer itself ^[1].

WHAT THE EVIDENCE SHOWS

A scoping review of 47 studies in head and neck cancer found that all 37 studies assessing taste reported some taste alteration after treatment. Agreement ends there: which taste qualities (sweet, sour, salty, bitter) were impaired, whether function returned to baseline, and which treatments were responsible varied across studies. For smell the findings were even less consistent — some studies found measurable impairment, others found none ^[2].

SUPPLEMENTS: WHAT WAS NOT CONFIRMED. Zinc sulfate was not found to be an effective treatment for taste impairment; a liposomal spray showed some potential, which the authors describe as a direction for research rather than a ready recommendation ^[2]. This should be distinguished from a different use of zinc with which it is often confused. A meta-analysis of five randomised trials (332 patients) addressed prevention of radiation-induced oral mucositis, not taste changes, and found the result depends on the route: systemic oral zinc does not provide a reliable prophylactic benefit, while topical zinc mouthwashes show an encouraging protective trend ^[3].

AT THE TABLE — NCI SUGGESTIONS ^[1].

• when food tastes metallic or bitter: use plastic, silicone, wooden or bamboo utensils instead of metal ones
• when meat has become unpalatable: plant-based protein sources such as tofu, nuts or beans
• when food seems to have no flavour: marinades for meat and fresh or dried herbs (basil, oregano, thyme)
• when odours are the problem: keep food covered until it is served and avoid strong smells such as brussels sprouts or fish

WHAT WE DO NOT KNOW

There is little research pointing to an effective treatment for taste and smell impairment — of the 47 studies in the review only three addressed treatment at all, and the authors name this explicitly as a gap ^[2]. Nor has it been established how far these changes worsen a patient's nutritional status. Persistent or worsening taste and smell changes are worth reporting to the treating team: they may have a cause requiring separate management, such as a dental problem.

Sources

1. NCI PDQ — Nutrition in Cancer Care (patient version): taste and smell changes: cancer.gov ↗
2. Smell and Taste Impairments in Head and Neck Cancer Patients — A Scoping Review, Nutrients 2025;17(6):1087: doi.org ↗
3. The Route of Administration Determines the Efficacy of Zinc in Preventing Radiation-Induced Oral Mucositis: A Systematic Review and Meta-Analysis, Curr Oncol 2026;33(6):371: doi.org ↗

WHERE THIS ADVICE CAME FROM

Low-fibre diets have historically been recommended during pelvic radiotherapy despite a lack of evidence — that is how the authors of the randomised trial in the American Journal of Clinical Nutrition describe the starting point ^[1]. The reasoning was mechanistic: fewer residues in the bowel, less stimulation of an irritated mucosa. For years the assumption was never tested under random allocation.

WHAT THE TRIAL THAT TESTED IT SHOWED

• 166 patients undergoing radical pelvic radiotherapy were randomly assigned to three arms: low-fibre (≤10 g non-starch polysaccharides/day), habitual-fibre (control) and high-fibre (≥18 g/day), each with individualised dietary counselling ^[1];
• fibre intakes did differ significantly between arms (P < 0.001), so the intervention worked as intended ^[1];
• THE PRIMARY ENDPOINT WAS NOT MET: the between-group difference in the change of the IBDQ-B bowel score from baseline to nadir was not significant (P = 0.093) ^[1];
• in further analyses the change from start to end of radiotherapy was smaller in the high-fibre arm (−3.7 ± 12.8) than in the habitual-fibre arm (−10.8 ± 13.5; P = 0.011), and at one year (n = 126) the scores were +0.1 ± 14.5 versus −8.4 ± 13.3 (P = 0.004) ^[1];
• no significant differences were found in stool frequency or form, nor in faecal short-chain fatty acid concentrations ^[1];
• significant reductions in energy, protein and fat intake occurred in the low-fibre and habitual-fibre arms ONLY ^[1].

HOW TO READ THIS HONESTLY

The primary endpoint was not met, so the trial does not prove that a high-fibre diet reduces bowel toxicity — the authors' conclusion rests on secondary measures and needs confirmation. What is robust is something else: the trial found no benefit from restricting fibre, and it did show a cost of that restriction, namely a fall in energy and protein intake in people already at risk of malnutrition. Hence the closing sentence of the paper: restrictive, non-evidence-based advice to reduce fibre intake in this setting should be abandoned ^[1].

WHAT THE SYSTEMATIC REVIEWS SAY

• a review of fibre modification in women with gynaecological cancers included 4 studies (89 participants in total) with quality ratings of neutral or negative; because of risk of bias, inconsistency, indirectness and imprecision, the GRADE certainty of evidence was rated VERY LOW and the body of evidence was judged insufficient to form specific recommendations for clinical practice ^[4];
• a broader review of dietary counselling during pelvic radiotherapy covered 11 randomised trials prescribing the consumption or avoidance of fats, fibre, lactose, protein and FODMAPs; three of them reported an improvement in the incidence of diarrhoea, results varied between studies, and the authors call for higher-quality trials ^[3].

WHEN RESTRICTING RESIDUE IS JUSTIFIED

This entry does not argue against a low-residue diet where it is a clinical indication — only against applying it ROUTINELY to every patient receiving radiotherapy. Situations in which the treating team advises a diet restricted in residue include narrowing or threatened obstruction of the gastrointestinal tract, the early period after bowel surgery and after stoma formation, and severe diarrhoea requiring immediate management. These are individual decisions based on examining the patient.

WHY A SPECIALIST SHOULD ASSESS THE SYMPTOMS

The 2025 British Society of Gastroenterology practice guidance (103 recommendations) stresses that long-term physical side effects of cancer therapy impinge on quality of life in up to 25% of those treated, and that gastrointestinal side effects are the most common and troublesome among them ^[2]. The guidance also makes two points that matter directly to patients: individual symptoms or symptom clusters are poor at distinguishing the underlying cause, and patients frequently have several coexisting gastrointestinal causes, all of which need to be diagnosed and optimally treated for symptoms to resolve ^[2]. The choice of management — dietary management included — belongs to the treating team, not to a general recommendation found online.

WHAT THIS ENTRY DOES NOT SAY

It contains no dietary recommendation for any individual and does not state how much fibre a patient should eat during radiotherapy. It states only where the evidence stands: automatic withdrawal of fibre is not supported by trials and may be paid for with poorer nutrition, while the evidence for the opposite direction is limited and requires confirmation. All the studies cited concern radiotherapy covering the pelvis and do not apply to irradiation of other body regions.

Sources

1. [1] Wedlake L i wsp. — Randomized controlled trial of dietary fiber for the prevention of radiation-induced gastrointestinal toxicity during pelvic radiotherapy, Am J Clin Nutr 2017 (166 chorych, trzy grupy; główny punkt końcowy p = 0,093; PMID 28679552, NCT01170299): europepmc.org ↗
2. [2] Andreyev J i wsp. — British Society of Gastroenterology practice guidance on the management of acute and chronic gastrointestinal symptoms and complications as a result of treatment for cancer, Gut 2025 (103 zalecenia; PMID 40068855): europepmc.org ↗
3. [3] The effect of nutritional interventions involving dietary counselling on gastrointestinal toxicities in adults receiving pelvic radiotherapy — a systematic review, J Med Radiat Sci 2021 (11 badań randomizowanych; PMID 34288532): europepmc.org ↗
4. [4] The Efficacy of Dietary Fiber in Managing Gastrointestinal Toxicity Symptoms in Patients with Gynecologic Cancers undergoing Pelvic Radiotherapy: A Systematic Review, J Acad Nutr Diet 2021 (4 badania, 89 uczestniczek, GRADE: bardzo niska pewność dowodów; PMID 33127328): europepmc.org ↗

WHY THIS ENTRY EXISTS

Throughout this service the message recurs that malnutrition in cancer must not be dismissed. This page addresses the other side of the same problem: bringing someone out of prolonged starvation is a separate medical task, and haste in it can be more dangerous than the deficit itself. It is one of the few situations in clinical nutrition where faster does not mean better.

WHAT THE PHENOMENON IS

During prolonged starvation the body switches to burning fat, and stores of phosphate, potassium and magnesium become depleted — even though their blood concentrations may appear normal ^[3]. When feeding resumes, rising glucose stimulates insulin secretion, and insulin drives phosphate, potassium and magnesium from the blood into cells; at the same time increased glucose metabolism consumes thiamine (vitamin B1) stores ^[3]. The classic laboratory feature is a fall in phosphate, but the picture also includes disturbed sodium and fluid balance and deficiency of thiamine, potassium and magnesium ^[1][2].

WHY IT MATTERS

The reported consequences can be severe and affect several systems at once ^[3]:

• heart — arrhythmias, low blood pressure, in extreme cases cardiac arrest;
• respiratory system — respiratory failure, pulmonary oedema;
• nervous system — features of thiamine deficiency (confusion, memory impairment, disturbed balance and eye movements).

For this reason, feeding is restarted under laboratory monitoring rather than by impression.

WHOM IT CONCERNS IN ONCOLOGY

The risk relates not to the diagnosis itself but to the DURATION AND DEPTH of reduced intake ^[1]. In oncological practice this usually means situations in which eating has been physically difficult or impossible for weeks: narrowing of the oesophagus or stomach, bowel obstruction, severe dysphagia after treatment of head and neck cancer, intractable vomiting, and advanced cancer cachexia. Additional risk factors include low body mass index, unintentional weight loss, previous starvation, past alcohol misuse and low baseline electrolyte levels ^[3].

WHAT THE TREATING TEAM DOES

The ESPEN guideline describes increasing nutritional provision slowly over several days and supplying vitamin B1 together with a balanced micronutrient mixture before and during repletion; potassium, phosphate and magnesium are monitored and replaced as needed ^[1][2]. We deliberately give NO doses or threshold values here — these are set individually from body weight, laboratory results and route of administration, and quoting them outside the clinical context would amount to advice for an individual rather than a description of knowledge.

HOW STRONG IS THE EVIDENCE

The ESPEN recommendation to advance nutrition slowly carries a strong grade but rests on a LOW level of evidence and on expert consensus ^[1][2]. The mechanism is well described physiologically and clinically ^[3], whereas trials comparing different refeeding schedules are few — which is why we tagged this entry as based on limited evidence. That does not mean the hazard is doubtful; it means there is no single schedule proven in trials.

WHAT THIS ENTRY DOES NOT SAY

It does not say after how many days without eating the risk appears in a given person, nor how many calories may be given on the first day. Nor is it a reason to postpone nutritional treatment — on the contrary: malnutrition in cancer requires treatment, and knowledge of refeeding syndrome exists so that it can be treated safely.

THE DECISION BELONGS TO THE DOCTOR

Whether a person is at risk of refeeding syndrome, and how fast to advance feeding, is decided by the treating specialist together with the nutrition team. If a patient has eaten very little for a prolonged period, it is worth telling the treating team BEFORE intensive refeeding begins — including when it takes place at home with oral nutritional supplements.

Sources

1. [1] Muscaritoli M i wsp. — ESPEN practical guideline: Clinical Nutrition in cancer. Clin Nutr 2021;40(5):2898-2913 (PMID 33946039) — zalecenie powolnego, rozłożonego na kilka dni zwiększania żywienia po długotrwale zmniejszonym przyjmowaniu pokarmu; siła zalecenia wysoka, poziom dowodów niski (konsensus): pubmed.ncbi.nlm.nih.gov ↗
2. [2] ESPEN — ESPEN practical guideline: Clinical Nutrition in cancer (pełny tekst wytycznych, espen.org) — hipofosfatemia jako klasyczna cecha zespołu; profilaktyczna podaż witaminy B1 i mieszanki mikroelementów; monitorowanie potasu, fosforanów i magnezu: espen.org ↗
3. [3] Persaud-Sharma D, Saha S, Trippensee AW — Refeeding Syndrome. StatPearls, NCBI Bookshelf (NBK564513) — definicja, mechanizm (wyrzut insuliny, przesunięcie fosforanów/potasu/magnezu do komórek, zużycie tiaminy), czynniki ryzyka i następstwa sercowe, oddechowe i neurologiczne: ncbi.nlm.nih.gov ↗

WHAT IT IS AND WHERE THE IDEA COMES FROM. The ketogenic diet sharply restricts carbohydrate in favour of fat, moving the body into nutritional ketosis. The biological rationale is real and named in the literature: many tumour cells, and glioblastoma in particular, show the Warburg effect — a dependency on aerobic glycolysis for energy ^[1]. The authors of a 2026 review in Neurological Sciences describe this as a "metabolic vulnerability" the diet is meant to exploit ^[1]. A biological rationale, however, is not the same thing as clinical proof, and this entry is precisely about that distinction.

WHAT THE GLIOBLASTOMA DATA SHOW

A PRISMA-compliant systematic review (Neurological Sciences 2026, PROSPERO CRD420251232650) included 41 studies published up to September 2025 — from randomised trials to case series and conference abstracts ^[1]. Adherence was high: more than 75% of participants maintained nutritional ketosis ^[1]. In adherent cohorts, median overall survival was 29.4 months versus 14.6 months in historical controls, with a 66.7% three-year survival rate ^[1]. The authors themselves conclude that standardised phase 3 trials are needed — and that is the most important sentence in the review ^[1].

WHY THOSE NUMBERS CALL FOR CAUTION

First, the comparator is a HISTORICAL control: survival reported in other studies and other years, not a group randomly allocated to standard care ^[1]. Second, the review included case series and abstracts, a literature in which favourable results are the ones that get published ^[1]. Third — the caveat a patient will not find in headlines — people able to sustain a restrictive diet for many months are by definition in better general condition, and better performance status is itself associated with longer survival in glioblastoma. From such a comparison it is not possible to say how much of the gap between 29.4 and 14.6 months belongs to the diet and how much to who was able to follow it ^[1].

WHAT IS KNOWN BEYOND GLIOBLASTOMA

An umbrella review in Integrative Cancer Therapies (2026) screened 615 articles and examined 24 systematic reviews and meta-analyses of the ketogenic diet in cancer ^[2]. The findings cut both ways: the diet improved metabolic markers — glucose and triglycerides — and showed benefits for body composition and quality of life, while evidence on treatment-related complications, including radiotherapy side effects, is described by the authors as limited and heterogeneous ^[2]. Their conclusion is that the diet shows promise as a safe and effective ADJUNCTIVE therapy and that more evidence is needed before firm conclusions can be drawn ^[2]. There is no claim of tumour control ^[2].

SAFETY. In the glioblastoma review the diet was well tolerated: adverse events were limited to mild gastrointestinal symptoms and fatigue, with no grade 3 or 4 diet-related toxicity reported ^[1]. That is meaningful, but it describes trial participants — people who were selected and monitored. A separate and unresolved issue is the risk of weight loss in patients already at risk of wasting; a ketogenic diet by design restricts energy from one macronutrient, and unintended weight loss is a marker of poor prognosis in oncology (see our entry on nutrition in cancer cachexia).

WHAT THE REFERENCE SOURCE SAYS

The National Cancer Institute, on its page about complementary approaches, states plainly that "there is no food or special diet that has proved to control cancer", and advises telling your doctor about any complementary approach you use ^[3]. That sentence does not rule out studying the ketogenic diet as supportive care — it rules out treating it as cancer therapy.

THE SCOPE OF WHAT WE WRITE HERE

We do not give macronutrient ratios, protocols for inducing ketosis, or target ketone levels. In cancer, a ketogenic diet affects body weight, glucose metabolism and treatment tolerance, so it belongs in the hands of the treating team and a clinical dietitian — particularly for people with diabetes, kidney disease, malabsorption or ongoing weight loss. We also found no high-quality data on interactions between the ketogenic diet and specific chemotherapy regimens, and we do not fill that gap with conjecture.

Sources

1. Firdous J, et al. — Efficacy and safety of ketogenic diet in glioblastoma: an updated systematic review and meta-analysis, Neurological Sciences 2026 (PMID 42032215): europepmc.org ↗
2. Kamali M, et al. — The Effect of a Ketogenic Diet on Cancer: Evidence From Systematic Reviews and Meta-Analyses, Integrative Cancer Therapies 2026 (PMID 41782309): europepmc.org ↗
3. National Cancer Institute — Complementary and Alternative Medicine (CAM) in Cancer Treatment, cancer.gov 2024: cancer.gov ↗