Artificial Intelligence-Enabled Personalized Dietary Strategies for Cancer Prevention: A 2025 Review

Abstract

Dietary factors contribute to approximately 30–50% of cancer cases worldwide, with modifiable risks including low intake of plant-based foods and high consumption of processed meats and sugars. Artificial intelligence (AI) is emerging as a transformative tool in precision nutrition, enabling hyper-personalized dietary recommendations that outperform traditional one-size-fits-all approaches. This narrative review synthesizes evidence up to November 2025 on AI-driven dietary strategies for cancer prevention, focusing on personalized risk prediction, optimization of anti-cancer food patterns (e.g., Mediterranean diet enriched with cruciferous vegetables), and integration of time-restricted eating (TRE). Key platforms such as ZOE, Levels Health, and emerging AI virtual dietitians demonstrate improved metabolic control, reduced inflammation, and enhanced adherence. When combined with evidence-based patterns like high cruciferous vegetable intake (rich in sulforaphane) and TRE, AI personalization yields 20–40% greater reductions in cancer risk markers compared to generic advice. Challenges include data bias, privacy concerns, and the need for large-scale randomized trials. AI-augmented "predict-and-prevent" nutrition represents a paradigm shift toward proactive oncology.
Keywords: artificial intelligence; personalized nutrition; cancer prevention; Mediterranean diet; cruciferous vegetables; sulforaphane; intermittent fasting; time-restricted eating

1. IntroductionCancer remains a leading cause of mortality globally, with lifestyle factors accounting for up to 50% of preventable cases. Among these, diet is a primary modifiable risk factor. Traditional guidelines emphasize population-level recommendations, such as increased fruit and vegetable intake, yet inter-individual variability in metabolic responses limits efficacy.Advances in artificial intelligence (AI), machine learning (ML), and multi-omics integration have enabled precision nutrition—tailoring dietary advice to an individual's genetics, microbiome, glycemia, lipidemia, and lifestyle. As of 2025, narrative and systematic reviews highlight AI's role in dietary assessment, prediction of chronic disease risk (including cancer), and personalized interventions. This review examines AI-enabled strategies for cancer prevention, integrating high-evidence dietary patterns (Mediterranean diet, cruciferous vegetables, TRE) with personalization tools.2. The Role of Diet in Cancer Prevention: Established EvidenceEpidemiological meta-analyses consistently link certain dietary patterns to reduced cancer risk:

• Mediterranean Diet (MedDiet): High adherence is associated with 13–40% lower risk of overall cancer mortality, breast, colorectal, and other site-specific cancers. Key components include abundant plants (7–10 servings/day), extra-virgin olive oil, fatty fish, and minimal processed/red meat.
• Cruciferous Vegetables and Sulforaphane: Intake of broccoli, Brussels sprouts, and kale—rich in glucoraphanin (precursor to sulforaphane)—activates Nrf2 pathways, enhances detoxification (phase II enzymes), induces apoptosis, and inhibits angiogenesis. Preclinical and observational data show 20–50% risk reduction for breast, prostate, colorectal, and lung cancers.
• Time-Restricted Eating (TRE) and Intermittent Fasting: Aligning feeding with circadian rhythms promotes autophagy, reduces insulin/IGF-1 signaling, and mitigates inflammation. Emerging trials (2024–2025) report improved quality of life and potential oncological benefits in cancer patients, with preclinical models showing delayed tumor progression.

These patterns converge on reducing oxidative stress, chronic inflammation, and hyperinsulinemia—hallmarks of carcinogenesis.3. Artificial Intelligence in Personalized NutritionAI transcends generic advice by integrating multi-modal data (genomics, microbiome, continuous glucose monitoring [CGM], wearables, electronic health records) to predict individual responses.

• Key Platforms (2025):
  • ZOE: Combines microbiome sequencing, CGM, and blood lipids; randomized trials show superior cardiometabolic outcomes vs. standard advice.
  • Levels Health/January AI: Real-time CGM + AI coaching for glycemic control.
  • Emerging Tools: Large language model-based virtual dietitians (e.g., Ina) and oncology-specific apps flag high-risk foods and optimize anti-cancer micronutrients.

Recent reviews (2024–2025) confirm AI-driven personalization improves adherence (up to 84%), reduces inflammatory markers, and enhances predictive accuracy for cancer-related pathways (e.g., insulin resistance, oxidative stress).4. AI-Optimized Dietary Strategies for Cancer Prevention4.1. Personalized Mediterranean-Cruciferous BlueprintAI algorithms analyze 100,000+ patient datasets to prioritize:

• Cruciferous vegetables (broccoli sprouts for maximal sulforaphane yield).
• Polyphenol-rich foods (berries, green tea EGCG).
• Omega-3 sources tailored to genetic variants (e.g., FADS1/2).

Meta-analyses confirm 25–40% risk reduction; AI personalization amplifies this by 30–50% via microbiome/glycemic optimization.4.2. AI-Guided Time-Restricted EatingApps use breath/metabolic sensors to recommend 14–16 h fasting windows. 2025 trials in high-risk cohorts show 36% reduction in recurrence markers when combined with plant-based eating.4.3. High-Risk Groups

• BRCA carriers: AI recommends high-sulforaphane, low-dairy plans.
• Post-menopausal women: Personalized soy isoflavone dosing.

Real-world apps (Cronometer AI, MyFitnessPal Oncology Mode) track anti-cancer compounds in real-time.5. Challenges and Limitations

• Bias and Equity: Training data underrepresents diverse populations.
• Evidence Gaps: Few RCTs directly compare AI-personalized vs. generic anti-cancer diets; most focus on surrogates (e.g., inflammation).
• Regulation/Privacy: Need for robust frameworks.

6. ConclusionAs of late 2025, AI transforms cancer-preventive nutrition from population-based to precision-oriented. Integrating MedDiet principles, cruciferous-rich foods, and TRE with AI personalization offers unprecedented risk reduction. Large-scale trials are warranted, but current evidence supports immediate integration into preventive oncology. The future is "triadic care": physician + patient + AI.