Effects of dietary intervention on human diseases: molecular mechanisms and therapeutic potential

This is a narrative review (not a meta-analysis or original study) that synthesises findings from hundreds of published studies on seven distinct dietary interventions:

**Calorie restriction (CR):** 20–40% reduction in total daily energy intake without malnutrition

**Fasting-mimicking diet (FMD):** A low-calorie, low-protein, low-carbohydrate diet (typically ~800–1100 kcal/day) designed to mimic the effects of water-only fasting while allowing some food intake, usually cycled for 3–5 days per month

**Ketogenic diet (KD):** Very low carbohydrate (<50g/day), high fat (70–80% of calories), moderate protein diet that induces ketosis

**Protein restriction diet (PR):** Reducing dietary protein intake, often to <0.8g/kg body weight/day, sometimes specifically restricting certain amino acids (e.g., methionine)

**High-salt diet (HSD):** Consistently high sodium intake (>5g sodium/day)

**High-fat diet (HFD):** Diets where >35% of total calories come from fat, often used as a model of Western dietary patterns

**High-fiber diet (HFD):** Diets providing >25–30g fiber/day from whole grains, legumes, vegetables, and fruits

The review examines how each intervention affects:

Cancer growth, metabolism, and treatment response

Neurodegenerative disease progression (Alzheimer's, Parkinson's)

Autoimmune disease activity (multiple sclerosis, rheumatoid arthritis, inflammatory bowel disease)

Cardiovascular disease risk factors and outcomes

Metabolic disorders (obesity, type 2 diabetes, non-alcoholic fatty liver disease)

The review also explores molecular mechanisms including metabolic reprogramming, immune modulation, autophagy induction, and gut microbiome changes.

Because this is a review of hundreds of studies, the populations vary enormously. The review covers:

**Animal models:** Primarily mice and rats with induced cancers (xenograft, genetically engineered), transgenic Alzheimer's models, chemically induced autoimmune disease models, and diet-induced obesity models

**Human studies:** Ranging from small pilot trials (10–50 participants) to larger clinical trials (100–500 participants), including:

- Cancer patients undergoing chemotherapy or radiation

- Overweight/obese adults with metabolic syndrome

- Patients with type 2 diabetes

- Individuals with mild cognitive impairment or early Alzheimer's

- Patients with multiple sclerosis or rheumatoid arthritis

- Healthy volunteers for short-term mechanistic studies

The review does not provide a single pooled sample size. Individual studies cited range from n=6 mice to n=433 humans.

The review aggregates findings from diverse measurement approaches:

**Cancer outcomes:** Tumor volume (mm³), tumor weight (g), survival time (days), Ki-67 proliferation index, apoptosis markers (cleaved caspase-3, TUNEL assay), glucose uptake (¹⁸F-FDG PET), metabolite levels (LC-MS/MS)

**Metabolic markers:** Blood glucose (mmol/L), insulin (μIU/mL), HbA1c (%), ketone bodies (β-hydroxybutyrate, mmol/L), IGF-1 (ng/mL), leptin (ng/mL), adiponectin (μg/mL)

**Immune markers:** T cell subsets (flow cytometry, %CD8+, %regulatory T cells), cytokine levels (IL-6, TNF-α, IFN-γ, pg/mL), NK cell activity, macrophage polarization markers

**Neurodegenerative markers:** Cognitive scores (MMSE, MoCA), brain amyloid-β load (PET imaging), tau pathology, neuroinflammation markers (GFAP, Iba-1)

**Autoimmune disease activity:** Clinical disease severity scores (e.g., EAE score for multiple sclerosis models, DAS28 for rheumatoid arthritis), inflammatory markers (CRP, ESR), gut barrier integrity (zonulin, LPS levels)

**Cardiovascular markers:** Blood pressure (mmHg), LDL cholesterol (mmol/L), triglycerides (mmol/L), carotid intima-media thickness (mm), endothelial function (flow-mediated dilation, %)

**Microbiome:** 16S rRNA sequencing for bacterial composition, short-chain fatty acid levels (acetate, propionate, butyrate, μmol/g feces)

### Study design

This is a **narrative review**—not a systematic review, meta-analysis, or original study. The authors searched PubMed, Web of Science, and other databases for studies published up to 2024, but they do not report a formal search strategy, inclusion/exclusion criteria, or quality assessment of individual studies. They selected studies they deemed relevant to illustrate the mechanisms and effects of each dietary intervention.

### What this design can and cannot prove

**What it can do:**

Provide a broad overview of the current state of research across multiple disease areas

Identify common molecular pathways (e.g., IGF-1 signaling, AMPK activation, mTOR inhibition) that may explain how different diets produce similar effects

Highlight gaps in knowledge and areas needing further research

Generate hypotheses for future studies

**What it cannot do:**

Cannot quantify the overall effect size of any intervention (no meta-analysis was performed)

Cannot assess publication bias or the quality of individual studies systematically

Cannot resolve contradictions between studies (the authors may selectively cite studies that support their narrative)

Cannot provide a definitive recommendation for clinical practice

The review is inherently subjective—the authors chose which studies to include and how to interpret them

### Major methodological weaknesses

**No systematic search protocol:** Unlike a Cochrane review or PRISMA-guideline systematic review, the authors do not specify their search terms, databases, date ranges, or how many studies were screened versus included

**No quality assessment:** Studies with small sample sizes, short durations, or weak designs are presented alongside rigorous RCTs without differentiation

**Heterogeneity not addressed:** The review combines animal studies, human observational studies, and human clinical trials without clearly distinguishing the strength of evidence from each

**No quantitative synthesis:** Effect sizes, confidence intervals, and p-values from individual studies are not pooled or compared systematically

**Potential for cherry-picking:** Without a pre-registered protocol, readers cannot know whether the authors omitted studies that contradict their conclusions

### Duration of studies cited

The human studies cited range from:

**Short-term:** 3–7 days (e.g., fasting before chemotherapy)

**Medium-term:** 4–12 weeks (e.g., ketogenic diet for weight loss or cognitive function)

**Longer-term:** 6–24 months (e.g., calorie restriction for metabolic health)

Most animal studies lasted 4–16 weeks

### Calorie restriction (CR)

**Cancer:** In mouse models, 20–40% CR reduced tumor growth by 25–60% across multiple cancer types (breast, prostate, colon, lung). In humans, 2–6 months of CR (25% reduction) decreased serum IGF-1 by 20–40%, which is associated with reduced cancer risk

**Neurodegeneration:** In Alzheimer's mouse models, CR improved cognitive performance by 30–50% on maze tests and reduced amyloid-β plaque burden by 20–35%

**Metabolic:** In humans, 12–24 months of 20–25% CR reduced body weight by 7–10%, improved insulin sensitivity (HOMA-IR decreased by 25–40%), and reduced blood pressure by 5–10 mmHg systolic

**Longevity:** In rodents, lifelong CR extends maximum lifespan by 30–50%. Human data is limited to surrogate markers

### Fasting-mimicking diet (FMD)

**Cancer treatment:** In a phase II trial of 101 breast cancer patients, 5 days of FMD per month for 6 months reduced chemotherapy-induced DNA damage in healthy cells by 30–50% while enhancing tumor cell death (measured by circulating tumor cell counts, which decreased by 40–60%)

**Immune modulation:** In mice, 3 cycles of 4-day FMD increased CD8+ T cell infiltration into tumors by 2–3 fold and reduced regulatory T cells by 40–60%

**Metabolic health:** In a human RCT of 71 healthy adults, 3 cycles of 5-day FMD per month for 3 months reduced body weight by 5.7 kg (vs 1.2 kg in control), decreased IGF-1 by 22%, and reduced blood pressure by 6 mmHg systolic

**Autoimmune:** In multiple sclerosis mouse models, FMD cycles reduced clinical disease severity scores by 50–70% and promoted remyelination

### Ketogenic diet (KD)

**Cancer:** In mouse glioblastoma models, KD reduced tumor growth by 30–50% and extended survival by 20–40% when combined with radiation. Human pilot studies (n=10–20) show mixed results—some patients show stable disease for 6–12 months, others progress

**Neurodegeneration:** In Alzheimer's patients, 6–12 weeks of KD improved cognitive scores (ADAS-Cog) by 2–4 points compared to controls. In Parkinson's, KD improved motor symptoms (UPDRS scores) by 20–30% in small trials (n=20–40)

**Epilepsy:** Well-established—KD reduces seizure frequency by >50% in 30–50% of drug-resistant epilepsy patients, with some becoming seizure-free

**Metabolic:** KD consistently reduces HbA1c by 0.5–1.0% in type 2 diabetes patients over 3–6 months, and reduces triglycerides by 20–40%

### Protein restriction (PR)

**Cancer:** In mice, 50–80% protein restriction (particularly methionine restriction) reduced tumor growth by 40–60% and enhanced chemotherapy efficacy. Human data is limited to small pilot studies

**Metabolic:** In humans, 4 weeks of protein restriction (<0.8g/kg/day) reduced IGF-1 by 25–30% and improved insulin sensitivity (HOMA-IR decreased by 20–30%)

**mTOR inhibition:** PR reduces mTORC1 activity by 50–70% in animal tissues, mimicking the effects of the drug rapamycin

### High-salt diet (HSD)

**Autoimmune:** In mouse models of multiple sclerosis, HSD (4–8% NaCl in diet) worsened disease severity by 50–100% and increased inflammatory Th17 cell counts by 2–3 fold

**Cardiovascular:** HSD (>5g sodium/day) increases blood pressure by 5–15 mmHg in salt-sensitive individuals and increases cardiovascular event risk by 15–30% in epidemiological studies

**Gut microbiome:** HSD reduces beneficial Lactobacillus species by 50–90% in mice and humans, and increases gut permeability

### High-fat diet (HFD)

**Cancer:** HFD (35–60% fat) promotes tumor growth in multiple mouse models—colon tumor incidence increases by 50–100%, breast cancer growth accelerates by 30–60%

**Metabolic:** HFD induces insulin resistance within 1–4 weeks in mice (HOMA-IR increases 2–4 fold), and in humans, 4–8 weeks of HFD reduces insulin sensitivity by 20–30%

**Neurodegeneration:** HFD worsens cognitive performance in mice by 20–40% and increases amyloid-β deposition by 30–50% in Alzheimer's models

### High-fiber diet (HFD)

**Cancer:** In epidemiological studies, high fiber intake (>25g/day) is associated with 15–30% reduced risk of colorectal cancer. In intervention studies, 4–8 weeks of high-fiber diet increased gut butyrate levels by 50–200%

**Immune modulation:** High-fiber diets increase regulatory T cell populations in the gut by 30–60% and reduce systemic inflammation (CRP decreases by 20–40% in some trials)

**Metabolic:** High-fiber diets improve glycemic control—HbA1c decreases by 0.3–0.5% in type 2 diabetes patients over 3–6 months

**Calorie restriction:** A 20–40% reduction in calories produces effects roughly comparable to metformin for insulin sensitivity (HOMA-IR improvement of 25–40%), but with additional benefits on blood pressure (5–10 mmHg drop) and IGF-1 (20–40% reduction). The weight loss effect (7–10% over 12 months) is similar to what many commercial weight loss programs achieve.

**Fasting-mimicking diet:** The 5.7 kg weight loss over 3 months (3 cycles) is substantial—roughly equivalent to a very low-calorie diet (800 kcal/day) but with less hunger and better adherence. The 22% reduction in IGF-1 is comparable to what you'd see with a 30–40% daily calorie restriction.

**Ketogenic diet:** The 0.5–1.0% reduction in HbA1c is clinically meaningful—comparable to adding a second diabetes medication. The 20–30% improvement in Parkinson's motor symptoms is similar to what you'd expect from low-dose levodopa.

**High-salt diet:** The 5–15 mmHg blood pressure increase in salt-sensitive individuals is clinically significant—comparable to the effect of gaining 10–20 kg of body weight.

**High-fiber diet:** The 15–30% reduced colorectal cancer risk from epidemiological studies is modest but consistent—similar to the risk reduction from regular aspirin use.

### Limitations acknowledged by the authors

The molecular mechanisms linking diet to disease outcomes are "largely unexplored" and "complex"

Most evidence comes from animal models, not human clinical trials

The optimal timing, duration, and composition of dietary interventions remain unclear

Individual variability in response to dietary interventions is poorly understood

Long-term safety data for many interventions (especially ketogenic diet and fasting) is limited

### Additional critical limitations

**No systematic review methodology:** The authors did not follow PRISMA guidelines, did not register a protocol, and did not assess risk of bias in individual studies

**Animal-to-human translation gap:** Many findings come from mouse models with genetically identical animals, controlled environments, and precisely controlled diets—conditions that do not reflect human real-world variability

**Short duration of human studies:** Most human trials lasted 3–12 months; long-term effects (>2 years) are unknown for most interventions

**Small sample sizes:** Many human studies cited had fewer than 50 participants, making them underpowered to detect moderate effects or rare adverse events

**Confounding factors:** Dietary interventions often produce weight loss, which independently improves metabolic health—it's difficult to separate the specific effects of the diet from the effects of weight loss

**Adherence issues:** Long-term adherence to restrictive diets (ketogenic, calorie restriction) is poor in real-world settings—dropout rates of 30–50% are common in trials lasting >6 months

**Publication bias:** Studies with positive results are more likely to be published, and this narrative review may overrepresent positive findings

**Industry funding:** Some studies on ketogenic diets and commercial fasting-mimicking diets were funded by companies with financial interests in the products

**No discussion of harms:** The review focuses on benefits but gives limited attention to potential adverse effects (e.g., ketoacidosis risk in diabetes, nutrient deficiencies with prolonged restriction, eating disorder risk with fasting)

For someone running their own n=1 experiment:

### What to test (specific intervention and dose)

**Option 1: Intermittent fasting / Fasting-mimicking diet**

Test: 5 consecutive days per month of a low-calorie (~800–1100 kcal/day), low-protein, low-carbohydrate diet (e.g., vegetable soups, small portions of nuts, olive oil)

Alternative: 16:8 time-restricted eating (16 hours fasting,