Email Dr. Emrick for a copy of the Research Paper
The Zone of Prevention
New Science: Reducing Cancer Risk
Cancer metabolism, the glucose-ketone index, and the evidence for moving cancer control upstream. A critical integrative review of established cancer prevention, metabolic oncology, ketogenic interventions, and the proposed glucose-ketone index framework.
This report is a research synthesis, not medical advice. Any ketogenic intervention for a patient with cancer should be coordinated with the oncology team and a qualified nutrition professional, without delaying or replacing standard treatment.
Purpose, scope, and interpretive standard
A decision-grade synthesis designed for clinical, public health, research, and executive audiences.
Abstract
Cancer prevention warrants greater strategic emphasis. In the United States, approximately 2.115 million new cancers and 626,140 deaths are projected for 2026, even as age-adjusted mortality has fallen 34% since 1991 and five-year relative survival has reached 70%. An estimated 40% of incident cancers and 44% of cancer deaths in 2019 were attributable to evaluated modifiable factors, led by smoking, excess body weight, alcohol, ultraviolet radiation, and physical inactivity. These data support an upstream zone of prevention, but not one organized around a single metabolic ratio.
The glucose-ketone index (GKI), calculated as blood glucose (mmol/L) divided by beta-hydroxybutyrate (mmol/L), is useful for describing the degree of nutritional ketosis. Its proposed use for cancer prevention is not yet clinically validated. The ratio is mathematically sensitive to small changes in the ketone denominator; proposed zones were derived from adherence and mechanistic reasoning rather than population outcome thresholds, and no prospective evidence shows that maintaining a specific GKI prevents cancer incidence or mortality. Human ketogenic-diet studies in oncology primarily demonstrate feasibility, changes in metabolic surrogates, and uncertain efficacy in heterogeneous treatment populations. Recent preclinical studies further show tissue-specific effects, including possible tumor promotion in selected models.
This report recommends an expanded Zone of Prevention encompassing exposure reduction, metabolic resilience, vaccination and infection control, risk-based detection, and structural supports. GKI may be studied as an exploratory exposure or adherence measure within this framework. It should not displace proven prevention, screening, standard oncologic treatment, or individualized medical supervision.
Biological plausibility is not equivalent to clinical validity. A biomarker becomes decision-ready only after demonstrating analytic reliability, prospective association, discrimination, calibration, incremental value, and evidence that acting on it improves outcomes.
Six conclusions
Cancer control must hold two truths at once: absolute burden remains severe, while mortality and survival trends demonstrate that prevention, detection, and treatment already work.
The largest quantified prevention opportunity is multi-factorial. Tobacco, excess body weight, alcohol, ultraviolet exposure, inactivity, infections, and dietary patterns cannot be reduced to glucose and ketones.
Cancer metabolism is a legitimate therapeutic and prevention research domain, but tumors are metabolically heterogeneous and retain functional mitochondria, substrate flexibility, and context-dependent dependencies.
GKI has face validity as a ketosis-adherence ratio. It has not demonstrated criterion validity as a cancer-risk biomarker, nor decision utility as a prevention target.
Ketogenic interventions can reduce glucose, insulin, adiposity, and some symptoms in selected patients. Evidence for cancer prevention, disease control, or survival remains insufficient and heterogeneous.
A credible Zone of Prevention should integrate proven measures, personalized risk, early detection, metabolic health, and the social conditions that determine whether prevention is accessible.
Do not frame the choice as metabolism versus mainstream oncology. The actionable strategy is to strengthen upstream risk reduction, rigorously test metabolic hypotheses, and preserve standards of care as evidence matures.
How the evidence was assessed
A transparent hierarchy prevents mechanistic enthusiasm from outrunning clinical evidence. Evidence was reviewed through July 22, 2026. Priority was given to peer-reviewed systematic reviews, clinical trials, major cohort analyses, national cancer statistics, and consensus prevention guidance. The source video and long-form transcript were treated as claim-generating media, not scientific evidence.
Established
Consistent human evidence, major guidelines, surveillance data
Appropriate for current prevention practice or policy
Promising
Human feasibility or surrogate improvement with a coherent mechanism
Appropriate for supervised research or selective adjunctive use
Uncertain
Small, heterogeneous, nonrandomized, or indirect evidence
Hypothesis-generating; avoid broad causal claims
Unsupported
No prospective validation or contradicted by stronger evidence
Do not use as a clinical risk threshold or substitute for care
Limitations
This is not a PRISMA-compliant systematic review, and it does not independently estimate pooled effects. Some 2026 publications are recent and may not yet have accumulated replication or post-publication critique. Cancer prevention trials require long follow-up, so the absence of endpoint evidence for GKI does not prove a lack of effect. It is, however, a reason to avoid treating proposed thresholds as established.
Why prevention belongs at the center
Absolute counts communicate urgency, while age-adjusted trends reveal what intervention has already achieved.
The source video emphasizes an estimated 626,000 United States cancer deaths, roughly 1,715 per day. The number is directionally consistent with the American Cancer Society 2026 projection, but it is incomplete without context. The cancer death rate declined 34% from its 1991 peak through 2023, averting approximately 4.8 million deaths. Five-year relative survival for all cancers combined rose from 49% in the mid-1970s to 70% for diagnoses during 2015 to 2021. The simultaneous rise in absolute burden reflects population growth, population aging, and residual risk, not the simple failure of oncology.
Projected deaths should not be used as a standalone scorecard for scientific progress. Counts answer “how many,” while age-adjusted rates answer “how risk is changing.” Both are necessary for honest communication.
Projected United States cancer deaths by selected site, 2026
Six sites account for a large share of projected deaths. Lung cancer alone approaches the combined total of the next three.
Source: American Cancer Society, 2026. Counts are projections, not observed registry totals.
Five-year relative survival, all cancers combined
The survival trajectory is the evidence that prevention, detection, and treatment already work.
Source: American Cancer Society, 2026. Mid-1970s baseline compared with diagnoses during 2015 to 2021.
of cancers are diagnosed at age 50 or older
of cancers are diagnosed at age 65 or older
Survival remains lower for Black Americans than for White Americans across all cancers combined
Burden is unevenly distributed. A prevention framework that ignores inequities in access, exposure, screening, and treatment will overstate individual responsibility and underperform at the population level.
Stage at diagnosis changes the survival landscape
Early detection is not primary prevention, but it belongs in an upstream cancer-control operating model because stage strongly shapes outcomes.
Five-year relative survival by stage at diagnosis
Select a site to isolate its stage gradient, or view all three together.
Source: American Cancer Society, 2026. Survival reflects diagnosis years 2015 to 2021.
The distance between localized and distant survival is the operational value of detection. For colorectal cancer that distance is 76 percentage points. A cancer-control strategy that funds prevention but not closed-loop follow-up leaves most of that value unclaimed.
Detection outcomes worth measuring
- Screening eligibility and completion, not availability alone
- Abnormal result follow-up and referral closure
- Diagnostic interval from abnormal result to definitive diagnosis
- Stage distribution at diagnosis, stratified by equity dimensions
- Genetic assessment for hereditary syndromes and surveillance for high-risk conditions
What can be changed now?
The attributable burden is substantial, measurable, and broader than metabolic biomarkers.
Attributable share of the total
Evaluated modifiable factors against all other or unevaluated factors.
Islami et al. (2024), United States 2019 data.
Share of incident cases by factor
Categories are not additive because attribution methods account for overlap.
Islami et al. (2024), United States 2019 data.
Population-attributable fractions are not personal predictions. They estimate the proportion of disease that might not have occurred under counterfactual low-risk exposure distributions, assuming causal relationships and model assumptions hold. They do not imply that every case is individually preventable. Genetics, aging, stochastic processes, unmeasured exposures, and imperfect intervention effectiveness remain important.
Dietary and lifestyle pattern evidence
The 2025 World Cancer Research Fund synthesis supports an integrated pattern rather than reliance on a single nutrient or metabolic value. Healthy weight, physical activity, whole grains and fiber, vegetables and fruits, lower intake of red and processed meats and sugar-sweetened beverages, and avoidance of alcohol and smoking form a coherent pattern of cancer prevention. Evidence is especially persuasive for colorectal and postmenopausal breast cancer. The American Cancer Society similarly recommends 150 to 300 minutes of moderate activity, or 75 to 150 minutes of vigorous activity weekly, with the upper range offering greater benefit.
Intervention domains and what to do now
| Intervention domain | Current evidence | Action now |
|---|---|---|
| Tobacco | Causal for multiple cancers; the largest attributable burden | Prevent initiation; support cessation; strengthen smoke-free policy |
| Adiposity and activity | Consistent associations and plausible mechanisms | Support a healthy weight and 150 to 300 minutes of moderate activity weekly |
| Alcohol | Causal for at least seven cancer sites | Avoid or minimize exposure |
| Vaccination and infection | HPV and HBV prevention; H. pylori and hepatitis treatment reduce risk | Vaccinate, screen, and treat according to guidelines |
| UV and occupational exposure | Established preventable carcinogenic exposures | Protection, regulation, and workplace controls |
| Screening | Site-specific mortality benefit when evidence-based | Age-appropriate and risk-appropriate screening with equitable follow-up |
A valid field, not a single-cause doctrine
Cancer metabolism is central to tumor behavior, but its pathways are adaptive, heterogeneous, and genetically regulated.
Two accounts of the same biology
The panels below hold the strong-form hypothesis alongside the more defensible synthesis. Use the switch to move between them.
The strongest version of the metabolic hypothesis argues that impaired mitochondrial respiration is the initiating and defining cause of cancer, with fermentation of glucose and glutamine sustaining growth. It is a single-cause chain, and its clarity is part of its rhetorical appeal.
A more defensible synthesis is that metabolic reprogramming is a core hallmark that interacts bidirectionally with genetic alterations, epigenetic state, tissue of origin, immune pressure, nutrient supply, and the tumor microenvironment. Causation runs in both directions, and the relative weight of each input varies by tumor and by stage.
Warburg metabolism and beyond
Many tumors display elevated glucose uptake and aerobic glycolysis, the classic Warburg phenotype. This supports biosynthesis, redox balance, and hypoxia adaptation. Yet aerobic glycolysis does not mean mitochondria are universally defective. Contemporary evidence shows that many cancers depend on oxidative phosphorylation, glutamine metabolism, fatty-acid oxidation, lactate exchange, or combinations that change across disease stage and treatment pressure. Metastatic cells may use different substrates than the primary tumor. Intratumoral subclones can also occupy different metabolic niches.
“Cancer has metabolic vulnerabilities” is well supported. “Cancer is one metabolic disease caused by mitochondrial failure” is not an adequate account of current evidence.
Systemic metabolism still matters. Hyperinsulinemia, insulin resistance, adipose inflammation, sex hormone changes, immune dysregulation, and altered growth-factor signaling provide plausible pathways linking obesity and inactivity to cancer risk. These mechanisms justify metabolic health as one of the pillars of prevention. They do not establish ketosis, or a low GKI, as a universal protective state.
Claim by claim assessment
| Claim | What supports it | What limits it | Assessment |
|---|---|---|---|
| Tumors often increase glucose use | FDG imaging, glycolytic signatures, and mechanistic studies | Not universal; substrate flexibility is common | Established but incomplete |
| Mitochondria matter in cancer | Biosynthesis, redox control, apoptosis, oxidative metabolism | Function varies by tumor and state | Established |
| Lower insulin and adiposity may reduce risk | Epidemiology and biological pathways | Effects differ by cancer and intervention | Supported prevention target |
| Ketosis selectively starves cancer | Preclinical models and metabolic rationale | Human tumors may use ketones or fatty acids; context-dependent | Uncertain |
| A single ratio defines prevention | Proposed GKI zones | No prospective cancer-risk validation | Unsupported for current use |
The glucose-ketone index
GKI describes the relationship between circulating glucose and beta-hydroxybutyrate. Its clinical meaning depends on context.
GKI = glucose (mmol/L) divided by beta-hydroxybutyrate (mmol/L). Convert glucose from mg/dL to mmol/L by dividing by 18.
GKI calculator and denominator sensitivity test
Enter a glucose and a beta-hydroxybutyrate value. The calculator returns the ratio and then does what the paper argues matters more: it shows how far the ratio moves when the ketone denominator shifts by a typical measurement increment.
A fixed glucose value can produce radically different GKI values
Glucose held constant at 5.0 mmol/L (90 mg/dL). Only the ketone denominator changes.
GKI = glucose (mmol/L) divided by beta-hydroxybutyrate (mmol/L). This analytic demonstration is not a clinical risk classification.
The instability is structural, not incidental
The ratio is a hyperbola. As beta-hydroxybutyrate approaches zero, GKI rises without bound. Your calculator value is plotted as the traveling marker.
Where the index came from
The GKI was introduced as a practical calculator for monitoring metabolic therapy in brain cancer research. In 2026, Lee and colleagues proposed extending it to chronic disease prevention and management. The proposal is explicit that prevention zones are hypotheses, subject to revision as future data accumulate. This distinction is essential. The publication provides a research roadmap, not validated cut points.
Ratio-specific limitations
Near-zero beta-hydroxybutyrate values can produce very large and unstable ratios.
Fasting duration, recent exercise, alcohol, acute illness, medications, diabetes status, and device performance can change either component.
The same GKI can arise from clinically different glucose and ketone combinations.
A ratio can obscure whether change came from improved glycemia, increased ketones, or both.
Repeated within-person measurement may describe adherence better than a single value, but prevention validity still requires prospective outcomes.
Construct validity audit
A biomarker becomes decision-ready only after clearing every gate below. Select a domain to see the question it answers and where the evidence currently stands.
Analytic reliability
Partially establishedDo devices produce reproducible glucose and BHB measurements?
Partially established for component meters. The error propagates through the ratio, which compounds rather than cancels the uncertainty of the two inputs.
Construct validity
Reasonable with caveatsDoes GKI measure the intended metabolic state?
Reasonable for the degree of ketosis, but denominator-sensitive. The index describes what it claims to describe, within limits.
Criterion validity
Not demonstratedDoes a threshold correspond to future cancer incidence?
Not demonstrated. This is the gate at which a descriptive ratio would become a risk marker, and it has not been cleared.
Discrimination
Not demonstratedDoes GKI distinguish those who will and will not develop cancer?
Not demonstrated. No published discrimination statistics exist for cancer incidence.
Calibration
Not demonstratedDo predicted risks match observed risks across groups?
No validated prediction model exists, so calibration cannot yet be assessed.
Incremental value
Not demonstratedDoes GKI improve prediction beyond age, smoking, adiposity, HbA1c, family history, and screening?
Not demonstrated. The ratio must outperform the simpler component models before it earns a place in a risk equation.
Clinical utility
Not demonstratedDoes targeting GKI improve cancer outcomes with acceptable harms and costs?
Not demonstrated. Utility is the final gate, and it requires randomized evidence that acting on the measure changes outcomes.
Use “proposed GKI prevention zone” or “exploratory metabolic exposure.” Avoid “safe zone,” “cancer-free zone,” or any statement that implies a validated probability of cancer prevention.
Human evidence is early and outcome-limited
Feasibility and metabolic change are more established than antitumor efficacy or prevention.
Reported reductions in fat mass, visceral fat, insulin, blood glucose, fatigue, and insomnia among cancer patients receiving ketogenic diets, with increased beta-hydroxybutyrate as expected.
Documented weaknesses: heterogeneous cancer types and diet protocols, inclusion of nonrandomized studies, small sample sizes, high heterogeneity across several outcomes, and an emphasis on surrogate or quality-of-life outcomes rather than incidence, progression, or mortality.
Seventeen evaluable patients with newly diagnosed glioblastoma completed a supervised 16-week 3:1 ketogenic diet alongside standard chemoradiation. The trial met its primary safety and feasibility objectives.
The absence of randomization prevents causal efficacy inference. A multicenter randomized phase 2 study is underway.
What each evidence stream can and cannot establish
| Evidence stream | Representative finding | What can be concluded | What cannot be concluded |
|---|---|---|---|
| Mechanistic | Lower glucose and insulin; higher ketones; altered substrate availability | Biologically active metabolic intervention | Universal tumor starvation or prevention |
| Animal models | Antitumor effects in some settings | Context-specific proof of principle | Human efficacy or safety across cancers |
| Human feasibility | Adherence is achievable with intensive support | Selected patients can implement the diet | Routine scalability or long-term adherence |
| Human surrogates | Changes in weight, insulin, glucose, BHB, and symptoms | Metabolic and symptom effects are plausible | Reduced incidence or mortality |
| Clinical endpoints | Small and heterogeneous studies; limited randomized evidence | Efficacy remains an open question | Replacement of standard treatment |
| Primary prevention | No prospective threshold trial for GKI | The research gap is clear | A validated prevention zone |
Countervailing preclinical signals
Recent animal research reinforces the need for caution when treating tumors and tissues. A 2026 Nature study found that a ketogenic, high-fat diet increased small-intestinal tumorigenesis in genetically susceptible mice through lipid oxidation and PPAR-linked stem-cell proliferation, while suppressing tumors in the colon. Earlier preclinical breast cancer work found slower primary tumor growth but increased lung metastasis under a ketogenic diet through BACH1-related mechanisms. These models do not establish harm in humans, but they directly challenge universal benefit claims and justify monitoring for unintended effects.
Ketogenic diets should not replace evidence-based cancer therapy. Patients with cancer, diabetes, kidney or liver disease, pregnancy, eating-disorder risk, or rare disorders of fat metabolism require individualized medical assessment and monitoring.
The Zone of Prevention, rebuilt
The most useful version is a layered risk-reduction system, not a narrow biochemical target.
01. Exposure reduction
Prioritize tobacco prevention and cessation, alcohol reduction, ultraviolet protection, radon and occupational controls, air-quality improvement, and avoidance of known carcinogens. These interventions address large, causal, and often policy-sensitive risks.
02. Metabolic resilience
Promote healthy weight trajectories, physical activity, sleep, dietary quality, and prevention or management of diabetes and insulin resistance. Metabolic measures may support personalization, but no single value should be treated as a cancer shield. In selected research settings, GKI can be collected alongside glucose, insulin, HbA1c, lipids, body composition, inflammatory markers, and diet quality.
03. Vaccination and infection control
Scale HPV and hepatitis B vaccination, identify and treat hepatitis C, manage Helicobacter pylori when indicated, and preserve access to HIV prevention and care. These are direct cancer-prevention interventions with clear causal pathways.
04. Risk-based detection
Apply evidence-based screening by age and risk, genetic assessment for hereditary syndromes, surveillance for high-risk conditions, and rapid diagnostic evaluation of warning symptoms. Early detection is not primary prevention, but it belongs in an upstream cancer-control operating model because stage strongly shapes outcomes.
05. Structural and policy supports
Design environments in which the low-risk choice is feasible. Food access, paid time, safe spaces for activity, clean air and water, insurance coverage, transportation, health literacy, and culturally competent care determine whether prevention recommendations translate into population benefit.
Lower preventable risk, earlier detection, and equitable access to proven interventions.
Risk can be reduced, not eliminated. A prevention framework must avoid blame, acknowledge uncertainty, and provide pathways for people who develop cancer despite healthy behavior.
What can leaders do now?
Implementation should be staged according to evidence strength, clinical risk, and equity impact.
Establish the baseline and stop the overclaim
Baseline tobacco, alcohol, activity, weight, vaccination, screening, and access gaps. Align messaging. Stop using unvalidated GKI risk labels.
Deploy integrated pathways
Deploy integrated prevention pathways, embed risk assessment, improve food and activity supports, and create supervised metabolic research protocols.
Evaluate, then scale what works
Evaluate effectiveness, costs, implementation, and disparities. Scale interventions with outcome evidence.
Measurement architecture
Primary prevention
Tobacco exposure, vaccination, adiposity trajectory, activity, alcohol, diet quality, and infection treatment.
Detection
Screening eligibility, completion, abnormal result follow-up, diagnostic interval, and stage distribution.
Metabolic
Fasting glucose, HbA1c, insulin where appropriate, lipids, blood pressure, body composition, and exploratory GKI.
Safety
Nutritional adequacy, hypoglycemia, dyslipidemia, renal stones, gastrointestinal symptoms, weight loss, disordered eating, and treatment interactions.
Equity
Participation, completion, outcome, and harm by race, ethnicity, rurality, income, insurance, disability, and language.
Say that tumors use multiple fuels; GKI thresholds are proposed; evidence adoption depends on quality, replication, benefit-harm balance, and feasibility; and a substantial fraction of cancer burden is preventable, although individual risk cannot be eliminated.
Zone of Prevention readiness assessment
Fifteen items across the five domains of the expanded model. This evaluates how completely your organization operates the prevention system. It is an organizational readiness instrument, not a personal health risk score. Rate each item: 0 not in place, 1 emerging, 2 established, 3 systematic.
01Exposure reduction
Tobacco cessation is systematically offered, documented, and tracked across our care settings.
Alcohol, ultraviolet, radon, and occupational exposure risks are addressed in our prevention messaging and workflows.
Smoke-free and exposure-control policies are enforced and monitored, not simply published.
02Metabolic resilience
Weight, activity, sleep, and glycemic health are managed as cancer prevention targets, not only as chronic disease targets.
Diabetes and insulin resistance are identified and treated proactively in our attributed population.
Metabolic measures support personalization without being communicated to patients as cancer risk thresholds.
03Vaccination and infection control
HPV and hepatitis B vaccination rates are measured, reported, and actively improved.
Hepatitis C and Helicobacter pylori are identified and treated according to guideline.
Infection-related cancer prevention has a named accountable owner in the organization.
04Risk-based detection
Screening is delivered by age and risk with measured completion, not availability alone.
Abnormal results have closed-loop follow-up with tracked diagnostic intervals.
Hereditary risk assessment and high-risk surveillance pathways exist and are used.
05Structural and policy supports
Food access, transportation, and time barriers are addressed as prevention infrastructure.
Participation, completion, and outcomes are stratified by equity dimensions and acted on.
Prevention communication avoids blame and avoids unvalidated risk labels.
How to test the prevention-zone hypothesis
Validation should proceed in sequence, with preregistered thresholds and clinically meaningful endpoints.
Measurement and natural history
- Standardize fasting state, device calibration, sampling time, medication and exercise documentation, and calculation rules.
- Measure within-person variability and reproducibility across capillary and laboratory assays.
- Characterize distributions by age, sex, diabetes status, dietary pattern, adiposity, race and ethnicity, and medication use.
- Compare the ratio against its separate components to determine whether it adds information.
Prospective association and prediction
Embed repeated glucose and beta-hydroxybutyrate measurements in large prospective cohorts with adjudicated cancer outcomes. Prespecify cancer sites, lag periods, confounders, and competing risks. Evaluate discrimination, calibration, nonlinearity, missingness, and incremental value beyond established predictors. Replicate externally before proposing thresholds.
Intervention and utility
If association and prediction are credible, randomize eligible participants to a safe, supported metabolic intervention versus an evidence-based comparison program. The target should initially be modifiable metabolic risk, not an arbitrary ratio. Cancer incidence may require large pragmatic trials or validated intermediate endpoints. Monitor nutritional status, cardiovascular risk, quality of life, adherence, adverse events, and inequitable exclusion.
Repeated GKI provides independent and reproducible prediction of site-specific cancer incidence beyond glucose, beta-hydroxybutyrate, HbA1c, adiposity, smoking, alcohol, activity, age, sex, family history, and screening behavior. This hypothesis must outperform the simpler component models before the ratio is considered useful.
Design threats and mitigations
| Design threat | Why it matters | Mitigation |
|---|---|---|
| Reverse causation | Occult cancer may change weight, glucose, appetite, or ketones | Lag analyses; exclude early cases; repeated measures |
| Diet confounding | Low GKI may correlate with many health behaviors | Detailed dietary and lifestyle adjustment; negative controls |
| Denominator instability | Small BHB values create extreme ratios | Model components separately; robust transforms; assay standards |
| Cancer heterogeneity | Associations may differ by tissue and subtype | Prespecified site-specific analyses; interaction tests |
| Selection and equity | Intensive diets may select affluent, healthy participants | Pragmatic recruitment, access supports, transportability analysis |
| Outcome substitution | Surrogate improvement may not translate to fewer cancers | Retain clinical endpoints and long-term follow-up |
Prevention is the strategy. Precision is the discipline.
The source video effectively directs attention to prevention and metabolism, two areas that deserve sustained investment. It is also rhetorically stronger than the evidence when it implies that a single metabolic account resolves cancer causation or that a GKI zone currently defines cancer protection. The modern cancer landscape is more complex: mortality rates have fallen, survival has improved, preventable burden remains large, tumors are metabolically heterogeneous, and ketogenic interventions have not yet demonstrated primary prevention efficacy.
A scientifically mature Zone of Prevention should therefore be built on convergent evidence. It should reduce tobacco and carcinogenic exposures; support healthy weight, activity, diet quality, sleep, and metabolic health; prevent infection-related cancers; deliver risk-appropriate screening; and address structural barriers. GKI can have a place within this model as an exploratory, context-sensitive measure. It should earn, not assume, its status through analytic validation, prospective prediction, randomized utility testing, and independent replication.
The enduring insight is not that cancer has one preventable fuel. It is that cancer control improves when action moves upstream, risk is measured honestly, and promising mechanisms are translated through rigorous human evidence.
Adopt the Zone of Prevention as a systems framework. Do not adopt GKI zones as validated cancer-risk categories. Fund studies to determine whether the ratio adds value beyond its components and beyond established prevention science.
Claim-to-evidence matrix
A compact audit trail for presentation and discussion. Filter by verdict to isolate what is supported and what is not.
Cancer burden remains severe
SupportedUrgency is justified, but trends show substantial progress.
Cancer is fundamentally metabolic
Partly supported, overly reductiveMetabolism is central; genetics, epigenetics, immunity, and tissue context remain causal.
Lower glucose and higher ketones can alter tumor biology
Supported mechanistically; variable clinicallyAppropriate research hypothesis, not a universal treatment rule.
GKI measures nutritional ketosis
Supported with measurement caveatsUseful for adherence monitoring in selected contexts.
A GKI zone prevents cancer
Not validatedDo not communicate as a clinical threshold.
Ketogenic diets improve cancer survival
Insufficient evidenceFeasibility and surrogate effects do not establish a survival benefit.
Mainstream oncology ignores prevention
Contradicted as a generalizationPrevention guidance is extensive; implementation and investment gaps remain.
Prevention should receive greater strategic emphasis
SupportedExpand integrated, equitable prevention and implementation research.
Appendix B: GKI worked examples
| Glucose | BHB | GKI | Interpretation |
|---|---|---|---|
| 90 mg/dL = 5.0 mmol/L | 0.1 mmol/L | 50.0 | Low ketone state; ratio inflated by small denominator |
| 90 mg/dL = 5.0 mmol/L | 0.5 mmol/L | 10.0 | The nutritional ketosis boundary is used in some contexts |
| 90 mg/dL = 5.0 mmol/L | 2.0 mmol/L | 2.5 | Deeper ketosis; not a validated cancer-prevention probability |
| 72 mg/dL = 4.0 mmol/L | 2.0 mmol/L | 2.0 | Example discussed in the proposed therapeutic literature |
References
Filter by evidence category. Digital object identifiers link to the source record where available.
American Cancer Society. (2026). Cancer facts & figures 2026. American Cancer Society.
Amaral, L. J., et al. (2025). A phase 1 safety and feasibility trial of a ketogenic diet plus standard of care for patients with recently diagnosed glioblastoma. Scientific Reports, 15.https://doi.org/10.1038/s41598-025-06675-6
Du, H., Xu, T., Yu, S., et al. (2025). Mitochondrial metabolism and cancer therapeutic innovation. Signal Transduction and Targeted Therapy, 10, 245. https://doi.org/10.1038/s41392-025-02311-x
Hanahan, D. (2022). Hallmarks of cancer: New dimensions. Cancer Discovery, 12(1), 31 to 46. https://doi.org/10.1158/2159-8290.CD-21-1059
Islami, F., Marlow, E. C., Thomson, B., McCullough, M. L., Rumgay, H., Gapstur, S. M., Patel, A. V., Soerjomataram, I., & Jemal, A. (2024). Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States, 2019. CA: A Cancer Journal for Clinicians, 74(5), 405 to 432. https://doi.org/10.3322/caac.21858
Lee, D. C., Duraj, T., Cooper, I. D., Maroon, J. C., Smith, K., Abdel-Hadi, W., Omene, E., Evangeliou, A. E., & Seyfried, T. N. (2026). The glucose ketone index: A proposed quantitative biomarker to support cancer and chronic disease prevention and management. Frontiers in Science, 4, 1763395. https://doi.org/10.3389/fsci.2026.1763395
Meidenbauer, J. J., Mukherjee, P., & Seyfried, T. N. (2015). The glucose ketone index calculator: A simple tool to monitor therapeutic efficacy for metabolic management of brain cancer. Nutrition & Metabolism, 12, 12. https://doi.org/10.1186/s12986-015-0009-2
Rock, C. L., Thomson, C., Gansler, T., et al. (2020). American Cancer Society guideline for diet and physical activity for cancer prevention. CA: A Cancer Journal for Clinicians, 70(4), 245 to 271. https://doi.org/10.3322/caac.21591
Siegel, R. L., Kratzer, T. B., Wagle, N. S., Sung, H., & Jemal, A. (2026). Cancer statistics, 2026. CA: A Cancer Journal for Clinicians.https://doi.org/10.3322/caac.70043
Sung, H., Ferlay, J., Siegel, R. L., et al. (2021). Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide. CA: A Cancer Journal for Clinicians, 71(3), 209 to 249. https://doi.org/10.3322/caac.21660
World Cancer Research Fund International. (2025). Dietary and lifestyle patterns for cancer prevention: Evidence and recommendations from the Continuous Update Project Global.
Zhang, M., Zhang, Q., Huang, S., Lu, Y., & Peng, M. (2025). Impact of ketogenic diets on cancer patient outcomes: A systematic review and meta-analysis. Frontiers in Nutrition, 12, 1535921. https://doi.org/10.3389/fnut.2025.1535921
Zong, X., et al. (2026). The ketogenic diet mediates intestinal tumorigenesis through lipids, not ketones. Nature.https://doi.org/10.1038/s41586-026-10779-y
The Diary of a CEO. (2026). Have doctors been ignoring the data? [YouTube Short]. YouTube. https://www.youtube.com/shorts/CllmfUX-7CU
Note: “et al.” is retained for space in selected high-author-count references. A manuscript submitted to a journal should export complete author lists from a reference manager in the target journal’s required APA implementation.
The Zone of Prevention: New Science, Reducing Cancer Risk
Research and analysis by Kelly Emrick, DHSc, PhD, MBA, BSRT(ARRT)R. A critical integrative review. Evidence reviewed through July 22, 2026.
This dashboard is a research synthesis, not medical advice. Any ketogenic intervention for a patient with cancer should be coordinated with the oncology team and a qualified nutrition professional, without delaying or replacing standard treatment.