Keto (Ketogenic Diet)
Keto is a low-carb high-fat dietary pattern (under 50g carb/day standard, under 20g strict) that drives ketosis and BHB above 0.5 mM within 2 to 4 days. The DIETFITS 12-month trial (Gardner 2018, JAMA) found low-carb minus 6.0 kg vs low-fat minus 5.3 kg (NS) with no diet-genotype or diet-insulin-secretion interaction. Real for refractory pediatric epilepsy and Virta-class T2D reversal. Real risk in lean women per Retterstol 2021, SGLT2i users (euglycemic DKA), and undiagnosed FAOD.
Keto (Ketogenic Diet) scored 5.6 / 10 (⚖️ Neutral) on the BioHarmony scale as a Practice / Lifestyle.
What It Is
Keto is a dietary pattern that restricts carbohydrate intake below 50 g/day (standard) or 20 g/day (strict, Westman-Duke) to deplete hepatic glycogen and force the liver to produce ketone bodies from fatty acids. Plasma beta-hydroxybutyrate (BHB) rises above 0.5 mM within 2 to 4 days, defining nutritional ketosis.
The mechanism is broader than substrate switching. BHB is a signaling metabolite: a class I HDAC inhibitor that relieves PGC-1α gene repression and drives mitochondrial biogenesis, an HCAR2 (GPR109A) ligand that suppresses NF-κB inflammation, an AMPK activator, and a SIRT2/FOXO autophagy inducer. Reduced insulin and IGF-1 suppress mTOR. Keto is a metabolic-state intervention rather than a single-target drug, which explains both its breadth of effect and its bioindividual variability.
The diet has many variants. Standard keto runs 70-75% kcal fat, 20-25% protein, 5-10% carb. Westman-Duke strict caps total carb at 20 g. Cyclical keto inserts 1 to 2 day carb refeeds every week to preserve glycogen-dependent exercise and mitigate T3 suppression. Targeted keto (TKD, Ketogains) adds 0.05 g glucose per pound of lean body mass peri-workout. Modified Atkins is more flexible on timing. Mediterranean-keto (Attia) substitutes MUFA-heavy fats (olive oil, avocado, nuts) for the saturated-fat profile typical of LCHF and is preferred for ApoB-monitored chronic use. Classical 4:1 (4 parts fat to 1 part protein+carb by weight) is the pediatric epilepsy clinical protocol.
The condition that defines keto's strongest evidence base is refractory pediatric epilepsy, where classical 4:1 produces 50% seizure reduction in roughly half of unresponsive patients within 3 months. Virta Health's 5-year T2D reversal program is the second-strongest indication. Beyond these two clinical use cases, keto is best understood as an 8 to 12 week elimination diet, after which transition to Mediterranean or carb-cycling is the default sustainable path.
Terminology
- BHB: Beta-hydroxybutyrate, the dominant ketone body. Plasma BHB above 0.5 mM defines nutritional ketosis.
- Ketosis: Metabolic state in which hepatic ketone production exceeds peripheral utilization, achieved within 2 to 4 days of carbohydrate restriction below 50 g/day.
- HDAC inhibitor: Histone deacetylase inhibitor. Class I HDAC inhibition by BHB drives mitochondrial biogenesis via PGC-1α derepression.
- HCAR2: Hydroxycarboxylic acid receptor 2 (GPR109A). BHB and niacin ligand; activation suppresses NF-κB inflammation.
- AMPK: AMP-activated protein kinase. Energy-sensing pathway activated chronically by keto.
- mTOR: Mechanistic target of rapamycin. Suppressed by keto via reduced insulin/IGF-1.
- LMHR: Lean mass hyper-responder. Phenotype on keto with LDL-C ≥200, HDL ≥80, TG ≤70 mg/dL.
- ApoB: Apolipoprotein B. Atherogenic-particle structural protein. More reliable than LDL-C for cardiovascular risk per the AHA 2026 dietary patterns guideline.
- SHBG: Sex hormone-binding globulin. Keto raises SHBG per Svart 2024, reducing free testosterone and estradiol.
- LCHF: Low-carbohydrate, high-fat. Umbrella term covering keto, modified Atkins, animal-based, and Banting variants.
- FAOD: Fatty acid oxidation disorder. Genetic conditions (MCADD, VLCADD, LCHADD, CPT1, CPT2, CACT, primary carnitine deficiency).
- MCT: Medium-chain triglyceride. C8/C10 fatty acids producing ketones rapidly without dietary restriction.
- Exogenous ketones: Ingested BHB salts or ketone esters that raise plasma BHB without dietary carb restriction.
- KETO-CTA: Coronary CT angiography study series in LMHR phenotype. The longitudinal follow-up was RETRACTED.
- Euglycemic DKA: Diabetic ketoacidosis with apparently normal glucose. Catastrophic SGLT2 inhibitor + keto interaction.
- TKD: Targeted ketogenic diet. 0.05 g glucose per lb LBM peri-workout, dextrose or maltodextrin only.
Dosing & Protocols
Dosing information is summarized from published research and community reports. This is not a prescribing guide. Consult a healthcare provider before starting any protocol.
View 5 routes and 5 protocols
Routes & Forms
| Route | Form | Clinical Range | Community Range |
|---|---|---|---|
| Standard keto (continuous) | Dietary pattern, under 50 g total carb/day, 70-75% kcal fat, 20-25% protein | 20-50 g total carb/day; ratio adjustments per protocol | 20-50 g total carb/day |
| Cyclical keto | 5-6 days strict keto + 1-2 day carb refeed (100-300 g carb) | Less RCT data; community-validated | 5-6:1-2 ratio |
| Targeted keto (TKD) | Continuous keto + 0.05 g glucose/lb LBM peri-workout | 0.05 g/lb LBM dextrose or maltodextrin peri-workout | 0.05-0.1 g/lb LBM |
| Modified Atkins / Mediterranean-keto | More flexible carb (30-50 g), MUFA-heavy fat (olive oil, avocado, nuts) | Refractory epilepsy adult use; Attia ApoB-monitored chronic protocol | Variable |
| Classical 4:1 (clinical, pediatric epilepsy) | 4:1 ratio fat:(protein+carb) by weight | Hospital-initiated; weighed meals | Not used outside clinical setting |
Protocols
Virta Health T2D reversal Clinical
- Dose
- Strict keto (20g net carb/day), continuous ketone monitoring (BHB target above 0.5 mM)
- Frequency
- Daily, continuous
- Duration
- Indefinite for sustained remission; 60% T2D remission at 1 year (Hallberg 2018)
Bundled with continuous remote physician coaching, app-based feedback, peer community. Insulin/sulfonylurea dose reduction MANDATORY day 1. 5-year follow-up (Athinarayanan 2024 PMID 39433217) sustained in adherent subset.
Refractory pediatric epilepsy (classical 4:1) Clinical
- Dose
- 4:1 fat:(protein+carb) by weight
- Frequency
- Daily, continuous
- Duration
- 1-2 years; tapered after 50% seizure reduction
First-line for Glut1 deficiency, tuberous sclerosis. Catastrophic risks: kidney stones (5-6%), reduced BMD (68% pediatric long-term), growth failure. Hospital initiation mandatory.
Short-term elimination diet Mixed
- Dose
- Standard keto (40-50g total carb)
- Frequency
- Daily
- Duration
- 8-12 weeks
Most defensible biohacker use case. Resets glycemic homeostasis, identifies food sensitivities, drives initial fat loss. Exit to Mediterranean or carb-cycling for sustainability.
Cyclical keto (Nick's protocol) Anecdotal
- Dose
- 20-40g total carb 5-6 days; 100-300g clean carb refeed 1-2 days/week
- Frequency
- Weekly cycle
- Duration
- Indefinite; Nick used 2.5 years before endocrine inflection
Most sustainable for active lean adults. Preserves T3, glycogen, and gym performance. Watch SHBG, fasting insulin, fasting glucose at 6-month and 24-month checkpoints.
Mediterranean-keto (Attia) Mixed
- Dose
- 30-50g total carb, MUFA-heavy fat (olive oil, avocado, nuts), ApoB-monitored
- Frequency
- Daily
- Duration
- Indefinite, with quarterly ApoB monitoring
ApoB target below 60 mg/dL. Statin coverage if ApoB rises. APOE4 genotype check. Blends keto's metabolic benefit with reduced atherogenic-particle exposure.
How this score is calculated →
Upside (2.86 / 5.00)
| Dimension | Weight | Score | Visual | Weighted |
|---|---|---|---|---|
| Efficacy | 25% | 3.5 | 0.875 | |
| Breadth of Benefits | 15% | 4.0 | 0.600 | |
| Evidence Quality | 25% | 3.8 | 0.950 | |
| Speed of Onset | 10% | 4.5 | 0.450 | |
| Durability | 10% | 1.5 | 0.150 | |
| Bioindividuality Upside | 15% | 3.5 | 0.525 | |
| Total | 3.550 |
Upside Rationale
Efficacy (3.5 / 5.0). Keto's efficacy is genuinely heterogeneous. For refractory pediatric epilepsy at classical 4:1, 50% seizure reduction in 50% of patients within 3 months represents a transformative effect (NNT roughly 4) where pharmaceuticals have failed. For T2D reversal in the Virta cohort per Hallberg 2018, 60% remission at 1 year and HbA1c minus 1.3% is large by any pharmacological standard. For body composition in the general population, DIETFITS 2018 (Gardner) showed minus 6.0 kg keto vs minus 5.3 kg low-fat at 12 months (NS), and Bueno 2013 meta found just minus 0.91 kg long-term advantage. The averaged efficacy lands moderate because the strongest indications are narrow.
Breadth (4.0 / 5.0). Keto touches multiple systems: neurological (epilepsy seizure reduction, MCI cognition in non-APOE4, mood stabilization in bipolar adjunct trials), metabolic (T2D reversal in Virta cohort, HbA1c reduction, fasting insulin lowering, NAFLD/MASH intrahepatic triglyceride clearance), body composition (initial fat loss, satiety mediated by ketone-driven appetite suppression), cardiovascular (TG reduction, HDL increase, BP reduction in responders), inflammatory (HCAR2/GPR109A-mediated NF-κB suppression), and mitochondrial (PGC-1α derepression, biogenesis, fat-oxidation capacity doubling per Burke FASTER). Saturated-fat keto often raises LDL/ApoB and disrupts gut Bifidobacterium per recent microbiome work, counterweighting the cardiovascular and microbiome benefit.
Evidence (3.8 / 5.0). Keto has a deep RCT base (DIETFITS, Hallberg, Foster 2010, Burke FASTER, Retterstol 2021, Paoli 2022, Svart 2024, recent microbiome work) and multiple meta-analyses (Bueno 2013, recent lipid-panel work lipid panel). Most positive long-term claims (Virta T2D 5-year follow-up, FASTER fat-oxidation, KETO-CTA pre-retraction LMHR safety) come from investigator groups with substantial financial conflicts. The KETO-CTA Longitudinal study retraction is a documented endpoint-switching event. A minus 0.5 evidence integrity adjustment is applied for industry-investigator conflict.
Speed (4.5 / 5.0). Keto produces ketosis (plasma BHB above 0.5 mM) within 2 to 4 days of carb restriction below 50 g/day. Glycemic effects emerge within days as hepatic glycogen depletes and insulin demand drops sharply. Initial weight loss appears within 1 week (largely glycogen-bound water, 1 to 3 kg shift). T2D HbA1c reduction in the Virta cohort per Hallberg 2018 was measurable by week 10. Refractory pediatric epilepsy seizure reduction in classical 4:1 protocols emerges within 3 months. LDL response in lean responders appears within 4 weeks per Retterstol 2021 with universal directional rise across all 30 participants.
Durability (1.5 / 5.0). Durability is keto's weakest dimension. The Foster 2010 2-year follow-up showed about 7 kg loss in both keto and low-fat arms by 24 months: convergence. Adherence is the binding constraint. Most users do not sustain under 50 g carb beyond 6 to 9 months. T2D remission requires continued dietary adherence; Virta 5-year follow-up (Athinarayanan 2024) shows sustained benefit ONLY in the adherent subset. Stop the diet and weight regain is the modal outcome. HDL improvement and triglyceride reduction persist longer than weight changes (Foster plus 23% HDL at 24 mo).
Bioindividuality (3.5 / 5.0). Strong responders cluster in T2D adults (60% Virta remission), refractory-epilepsy children (50% seizure reduction), MCI APOE3 adults (Krikorian, Cunnane MCT trials), fat-adapted ultra-endurance athletes (FASTER), and hyperandrogenic PCOS women per Mavropoulos 2005. Non-responders or harm-cluster: lean active women per Retterstol 2021 (universal LDL rise), APOE4 carriers (MCT/keto cognitive benefit absent or reversed), LMHR phenotype with ABCG5 heterozygous variants, eugonadal lean women (HPA-driven SHBG rise + amenorrhea risk).
Downside (3.16 / 5.00)
| Dimension | Weight | Score | Visual | Weighted |
|---|---|---|---|---|
| Safety Risk | 30% | 4.0 | 1.200 | |
| Side Effect Profile | 15% | 3.0 | 0.450 | |
| Financial Cost | 5% | 2.0 | 0.100 | |
| Time/Effort Burden | 5% | 3.5 | 0.175 | |
| Opportunity Cost | 5% | 3.0 | 0.150 | |
| Dependency / Withdrawal | 15% | 2.0 | 0.300 | |
| Reversibility | 25% | 1.5 | 0.375 | |
| Total | 2.750 | |||
| Harm subtotal × 1.4 | 3.255 | |||
| Opportunity subtotal × 1.0 | 0.425 | |||
| Combined downside | 3.680 | |||
| Baseline offset (constant) | −1.340 | |||
| Effective downside penalty | 2.340 |
Downside Rationale
Safety (4.0 / 5.0). The 4.0 floor is triggered by two distinct intrinsic catastrophic signals. First, SGLT2 inhibitor + keto = euglycemic DKA per the Kuchkuntla AACE case series. Glucose appears normal at 150-250 mg/dL while pH crashes below 7.30 and bicarbonate below 18 mEq/L; both the drug and diet drive ketogenesis and the drug also reduces renal ketone clearance. Second, undiagnosed fatty acid oxidation disorders (MCADD, VLCADD, LCHADD, CPT1, CPT2, CACT, primary carnitine deficiency) cause metabolic crisis on keto: hypoglycemia, rhabdomyolysis, cardiomyopathy, potentially fatal. Both meet the catastrophic-AE criterion; the floor is anchored at 4.0 per Rule #1 and not stacked further.
Side Effects (3.0 / 5.0). A recent BMC Nutrition systematic review found 43% of keto participants reported at least one adverse event; constipation was the most common at 24%. Keto flu (fatigue, headache, irritability, dizziness) at week 1 to 2, generally resolving with electrolyte correction (Na 3-7 g/d, K 3-4.5 g/d, Mg 300-500 mg/d). Bifidobacterium suppression on long-term keto per recent microbiome work is a microbiome-level signal. T3 suppression with normal TSH per Paoli 2022 drives clinically missed tissue hypothyroidism. Kidney stones at classical 4:1 in pediatric cohorts run 5-6% over 2 years. SHBG rise reduces free sex hormones per Svart 2024.
Cost (2.0 / 5.0). Keto food costs run $0 to $300/month above a standard mixed-diet baseline, depending on protein quality, fat-source preferences (grass-fed vs commodity), and whether organic vegetable sourcing is prioritized. Animal protein, avocado, olive oil, nuts, and low-carb vegetables run higher per-calorie than rice, beans, and oats, but lower than the high-end Mediterranean or paleo grocery basket. Optional ketone meter $30-50 plus strips $1-2 each adds modest monthly variable cost. Optional Virta-class medical coaching $300-400/month is bundled with the T2D reversal protocol. Optional exogenous ketone supplements run $50-200/month if added. Median biohacker cost lands around $50-100/month above baseline.
Effort (3.5 / 5.0). Keto demands sustained meal planning, label reading, restaurant navigation, social-event accommodation, and (in cyclical or targeted variants) macro tracking. Continuous ketone monitoring adds 5-10 minutes daily for users targeting a specific BHB range. Initial keto-adaptation requires 3 to 6 weeks of constrained food choices and electrolyte attention. Travel and dining-out friction is substantial. Most users self-report 30 to 60 minutes per day in cumulative meal planning, shopping, and prep beyond a flexible-eating baseline. The 3.5 score lands above the moderate tier because the cognitive and logistical burden compounds over months.
Opportunity (3.0 / 5.0). Keto crowds out alternative dietary strategies that may serve some users better: Mediterranean (stronger CV evidence base, AHA top-ranked), DASH (best BP-lowering data), GLP-1 agonist therapy (semaglutide and tirzepatide achieve T2D remission and weight loss without dietary restriction), and TRE (autophagy and metabolic flexibility benefits without strict carb limits). For lean active eugonadal women, the opportunity cost is real: Mediterranean or carb-cycling preserves T3, SHBG, and menstrual function while keto compromises them. Opportunity cost is moderate (3.0) because keto remains genuinely first-line for refractory pediatric epilepsy and Virta-class T2D reversal contexts.
Dependency (2.0 / 5.0). Keto has no addiction-type dependency. Stop the diet and ketosis ends within 1 to 3 days; no withdrawal pharmacology, no rebound phenomenon analogous to caffeine withdrawal headache or melatonin suppression. The functional dependency is low: T2D remission requires continued adherence (functional, not addictive); refractory epilepsy seizure control requires continued strict diet. Weight regain on cessation is a metabolic phenomenon (loss of ketone-mediated satiety, return of glycogen-bound water, hyperphagia rebound from extended caloric deficit) rather than a withdrawal syndrome. Compare to GLP-1 agonists (4.0 functional dependency for weight maintenance) where keto's dependency profile is substantially milder.
Reversibility (1.5 / 5.0). Keto is one of the most fully reversible interventions in the BioHarmony index. Stop the diet, eat carbohydrates, glycogen replenishes within 24 to 48 hours, ketosis ends within 1 to 3 days, plasma BHB returns to baseline within a week. LDL response in lean responders per Retterstol 2021 reverses within 4 to 8 weeks of normal-carb eating. T3 suppression reverses within weeks. Microbiome Bifidobacterium recovery on dietary normalization runs roughly 4 to 12 weeks with adequate fiber. Bone density loss in pediatric long-term 4:1 (68% reduction) reverses partially but slowly over months. Endocrine inflection points in long-term users reverse within months of cessation.
Verdict
✅ Best for: T2D reversal under medical supervision via Virta-class protocol with continuous coaching, ketone monitoring, and day-1 medication-dose reduction (insulin and sulfonylureas mandatory). Refractory pediatric epilepsy at classical 4:1 with hospital initiation and clinical supervision (first-line for Glut1 deficiency and tuberous sclerosis). Short-term elimination diet (8 to 12 weeks) for metabolic reset, food-sensitivity identification, and initial fat loss; transition to Mediterranean or carb-cycling thereafter. APOE3 cognitive-aging adults considering Mediterranean-keto with MCT 30-60 g/day augmentation. Hyperandrogenic PCOS women per Mavropoulos 2005 (therapeutic SHBG rise reduces total/free testosterone, improves ovulation). Ultra-endurance fat-adapted athletes who have completed 12+ week adaptation per Burke FASTER.
❌ Avoid if: Taking SGLT2 inhibitors (empagliflozin, dapagliflozin, canagliflozin, ertugliflozin, bexagliflozin); the catastrophic euglycemic DKA risk per the Kuchkuntla case series mandates either drug discontinuation or specialist supervision before initiation. Personal or family history of unexplained metabolic crises, exercise-induced collapse, or Reye-like episodes (undiagnosed FAOD risk). Pregnancy or lactation. T1D without intensive ketone and glucose monitoring. Active eating disorder history. LMHR phenotype without ApoB monitoring and ABCG5 heterozygote genetic check. APOE4 carriers without MUFA-heavy modification. Lean active eugonadal women on long-term strict keto (consistent with Paoli 2022 and Svart 2024 RCT data).
Use Case Breakdown
The overall BioHarmony score reflects the intervention's primary evidence profile. These subratings are independent assessments per use case.
| Use Case | Score | Summary |
|---|---|---|
| 👍 Metabolic Health Primary | 6.5 | Strong short-term insulin sensitivity and HbA1c response across multiple trials. Virta cohort (Hallberg 2018) showed HbA1c minus 1.3% with 60% T2D remission at 1 year. Athinarayanan 2024 extended this to 5 years in the adherent subset. NAFLD/MASH intrahepatic triglyceride clearance is rapid and well-documented. Long-term durability is constrained by adherence: most users do not sustain under 50g carb beyond 6 to 9 months, and metabolic gains revert with dietary regression. For short-term metabolic reset (8 to 12 weeks) and clinically supervised T2D reversal, keto delivers genuine metabolic improvements that few other interventions match. |
| 💪 Blood Sugar / Glycemic Control Primary | 7.0 | Glycemic effects emerge within days as hepatic glycogen depletes and insulin demand drops sharply. Hallberg 2018 showed HbA1c minus 1.3% within 10 weeks in the Virta T2D cohort with 94% insulin elimination at 1 year. The mechanism is direct substrate restriction: without dietary carbohydrate, post-prandial glucose excursions essentially disappear. Mandatory medication adjustment for diabetic patients (insulin and sulfonylurea doses must drop on day 1) to prevent hypoglycemia. For clinical T2D reversal under medical supervision, blood-sugar control is keto's strongest evidence-based use case alongside refractory pediatric epilepsy. GLP-1 agonists like semaglutide and tirzepatide compete here without dietary restriction. |
| ⚖️ Body Composition / Fat Loss Primary | 5.5 | Initial fat loss strong but converges with low-fat diet by 12 months. DIETFITS 2018 (Gardner, JAMA) is the most rigorous head-to-head: low-carb minus 6.0 kg vs low-fat minus 5.3 kg at 12 months, not significant, with no diet-genotype or diet-insulin-secretion interaction. Bueno 2013 meta found just minus 0.91 kg long-term advantage, which the authors called of little clinical significance. Foster 2010 showed convergence to about 7 kg both arms by 24 months. Adherence is the binding constraint, not macronutrient distribution. Best understood as an 8-12 week elimination diet. |
| ⚖️ Cognition / Focus Primary | 5.5 | MCI and cognitive-aging signal in non-APOE4 populations per Cunnane MCT trials and Krikorian's MCI ketone-monoester work. APOE4 carriers may not respond and may be harmed. The signal in healthy adults is weak: ketone-driven brain fuel shift produces subjective focus reports but no large RCTs in healthy cognition. APOE genotype testing should be standard before recommending chronic keto for cognitive aging. Bimodal response by genotype prevents a clean recommendation. For clean cognitive enhancement, modafinil (6.7/10), l-theanine (7.5/10), and phosphatidylserine (7.1/10) carry better population evidence. |
| ○ Longevity / Lifespan Primary | 4.5 | Mechanistic case (mTOR suppression, autophagy induction, mitochondrial biogenesis through BHB-mediated HDAC inhibition) is plausible but no human hard-outcome data exists. The UK Biobank LCHF cohort (Iatan 2024, JACC Adv) showed MACE HR 2.18 (95% CI 1.39-3.43) and HR 6.68 in the LDL≥5 mmol/L subgroup. The NHANES dietary ketogenic ratio cohort (Sci Rep 2024) showed HR 0.76 for all-cause mortality. Both are observational; neither is an RCT. No registered hard-outcome RCT comparing keto to Mediterranean exists. This is a definitive evidentiary ceiling. |
| 💪 Pediatric Use | 7.0 | Refractory pediatric epilepsy at classical 4:1 (4 parts fat to 1 part protein+carb by weight) is keto's strongest indication and the only context earning Top-tier scoring at the use-case level. Decades of clinical evidence support 50% seizure reduction in roughly half of pharmaceutically-unresponsive children within 3 months. First-line for Glut1 deficiency syndrome and tuberous sclerosis complex. Hospital-initiated with weighed meals, vitamin/mineral supplementation, and ongoing labs. Catastrophic risks include kidney stones (5-6% over 2 years), bone density loss (up to 68% reduction in long-term cohorts), and growth failure. Not a biohacker self-experiment. |
| 💪 Liver / Detoxification | 7.0 | Strong NAFLD/MASH evidence with intrahepatic triglyceride reduction within weeks of initiation. Multiple imaging-based trials show keto reduces liver fat content faster than calorie-matched low-fat diets. The mechanism is direct: lipogenic substrate restriction plus increased hepatic fat oxidation. For NAFLD/MASH patients without contraindications, keto is a validated first-line dietary intervention, often more effective than Mediterranean for liver fat clearance specifically. Caveat: liver failure is a contraindication due to impaired ketone metabolism, and SGLT2i + keto risk applies to T2D patients with hepatic comorbidity. |
| 👍 Mitochondrial | 6.0 | BHB-driven HDAC inhibition and PGC-1α derepression provide a credible mechanistic pathway to mitochondrial biogenesis. Burke FASTER 2017 showed fat oxidation doubling to 1.43-1.5 g/min in keto-adapted ultra-endurance athletes, an unprecedented number that demonstrates real mitochondrial reprogramming. AMPK activation chronically through reduced glucose flux supports mitochondrial maintenance. Clinical mitochondrial-disease applications (Glut1 deficiency syndrome) are an FDA-recognized indication for keto. For mitochondrial support, keto carries genuine mechanistic and applied evidence. Methylene blue, SS-31, urolithin A, and MOTS-c compete in the same target space. |
| 👍 Endurance / Cardio | 6.0 | Bimodal response by intensity. Burke FASTER 2017 (n=21 elite race walkers) showed fat oxidation doubling to 1.43-1.5 g/min in keto-adapted athletes, suitable for ultra-endurance fat-dominant work. At race-relevant intensities, absolute oxygen cost was higher (reduced exercise economy), with 1.6% non-significant 10K time impairment. For events above 75% VO2max, keto produces real performance decrement. Targeted ketogenic diet (TKD) protocols using 0.05 g glucose per lb LBM peri-workout bridge the anaerobic gap for resistance training. Ultra-endurance applications score higher; moderate-to-high intensity scores lower. Zone 2 cardio (7.8/10) is the broader-application alternative. |
| ⚖️ Neuroprotection | 5.5 | Refractory pediatric epilepsy mechanism (decades of clinical evidence in classical 4:1) plus MCI signal in non-APOE4 populations support a neuroprotective frame. APOE4 caveat is significant: Cunnane MCT trials show cognitive benefit absent or reversed in ε4 carriers, and chronic high-fat dietary patterns may worsen cerebrovascular risk in this subgroup. TBI ketone-fuel rationale is mechanistically promising with small concussion pilot trials emerging. For broad-population neuroprotection without genotype gating, omega-3 (6.8/10), creatine (8.3/10), methylene blue (6.2/10), and HBOT (7.4/10) are stronger first-line. |
| ⚖️ Anti-Inflammatory | 5.5 | HCAR2 (GPR109A) NF-κB suppression mechanism is well-characterized in cell and animal models, with BHB acting as a niacin-equivalent ligand. Clinical inflammation marker responses are mixed across human trials: hsCRP often improves on weight loss, but specific keto effects beyond weight loss are unclear. recent microbiome work (Dahl) RCT showed Bifidobacterium suppression which counteracts gut-mediated anti-inflammatory tone. Net signal is modestly positive for users who reduce inflammation through caloric/glycemic improvement, less clear for direct keto-specific anti-inflammatory effects. Omega-3 and astaxanthin carry cleaner anti-inflammatory evidence. |
| ⚖️ Cellular Senescence | 5.5 | mTOR suppression and autophagy induction are mechanistically plausible drivers of cellular senescence reduction. Keto suppresses mTOR through reduced insulin/IGF-1 and reduced amino acid sensing, contributing to autophagy induction. No clinical autophagy biomarkers are reported in keto trials directly. The mechanism overlaps with rapamycin (6.5/10) and senolytics (6.8/10), which have stronger direct senescence-targeting evidence. Keto's senescence claim rests on mechanism rather than clinical biomarkers, placing it in the exploratory tier for this use case. |
| ⚖️ Autophagy | 5.5 | Autophagy induction is the most-cited mechanistic claim for keto longevity benefits. SIRT2/FOXO activation through reduced insulin/IGF-1 and ketone-driven AMPK activation drives autophagy at the cellular level in animal models. Clinical autophagy biomarkers are not reported in keto trials. For users prioritizing autophagy induction, time-restricted eating (TRE, 6.5/10) achieves similar autophagic effects without strict carbohydrate restriction. Spermidine (7.0/10) carries cleaner autophagy-specific evidence. Keto's autophagy benefit is real mechanistically but not the cleanest route to that endpoint. |
| ⚖️ Energy / Fatigue | 5.5 | Subjective energy stabilization after 3-6 weeks of adaptation is widely reported in community surveys. The mechanism is likely metabolic flexibility: shifting from glucose-dominant fuel with attendant insulin oscillations to mixed-fuel oxidation with steadier substrate availability. Initial keto-flu (week 1-2) is energy-disruptive for most users. Caffeine sensitivity often increases on keto due to changes in cytochrome P450 activity. For users prioritizing acute energy, creatine (8.3/10), exogenous ketones, and structured caffeine use offer cleaner alternatives without dietary restriction. Adapted keto users report durable stable energy. |
| ⚖️ Mood / Emotional Regulation | 5.0 | Bipolar adjunct trial signal is small and emerging (Sethi 2024 metabolic psychiatry pilot showed promise). Initial week 1-2 keto-flu period is mood-disruptive for most users. Subjective stabilization on full adaptation is anecdotally common but not RCT-confirmed. The mechanism (BHB neuromodulatory effects, GABA-glutamate balance shifts) is biologically plausible. For mood support without dietary restriction, saffron (7.3/10), omega-3, and exercise dominate. Keto's mood signal sits in exploratory territory pending larger RCTs in mood disorders. |
| ⚖️ Sleep Quality | 5.0 | Reports of improved sleep architecture after full adaptation are common in community surveys. RCT data sparse. Initial 1-2 weeks of keto-flu can disrupt sleep onset and continuity. After adaptation, electrolyte balance (Na 3-7g, K 3-4.5g, Mg 300-500mg daily) is critical to avoid leg cramps and nocturnal wakings. The mechanism is unclear: possibly improved insulin sensitivity reducing nocturnal blood sugar swings, possibly direct BHB neuromodulation. Magnesium (8.3/10) and glycine (7.7/10) carry cleaner sleep-specific evidence. |
| ○ Recovery / Repair | 4.5 | Anti-inflammatory mechanism via HCAR2 suppression is plausible for recovery support, but glycogen-dependent recovery is compromised on strict keto. Resistance training recovery suffers from glycogen depletion, and the targeted ketogenic diet (TKD) protocol from Ketogains adds 0.05g glucose/lb LBM peri-workout specifically to address this. For broad recovery support, BPC-157 (7.3/10), TB-500 (7.2/10), creatine (8.3/10), and red light therapy (7.8/10) all carry cleaner recovery evidence without the glycogen tradeoff. Net recovery position depends heavily on training modality. |
| ○ Sleep Architecture (Deep/REM) | 4.5 | Anecdotal reports of increased deep sleep on adapted keto are common but not systematically validated. RCT data on sleep stages (REM, slow-wave, total sleep time) on ketogenic diets is sparse. The mechanism is unclear at the EEG level. For deep and REM sleep architecture support, glycine (7.7/10), magnesium (8.3/10), and consistent sleep timing carry stronger evidence. Keto's effect on sleep architecture remains exploratory pending dedicated polysomnography RCTs. |
| ○ Acute Pain Relief | 4.5 | Anecdotal migraine and inflammatory pain reduction reported across small case series. Mechanism could be via HCAR2-mediated NF-κB suppression or BHB neuromodulation. RCT evidence is sparse and mostly in chronic-pain populations rather than acute. For acute pain management, NSAIDs, celecoxib, and topical analgesics are first-line. Keto's pain-reduction signal exists but is not a primary indication. For chronic inflammatory pain, the anti-inflammatory mechanism may contribute to overall symptom reduction in adapted users. |
| ○ Chronic Pain Management | 4.5 | Same anti-inflammatory mechanism as acute pain. Small case series in fibromyalgia, chronic migraine, and inflammatory arthritis suggest modest improvement. The evidence base remains exploratory. Multiple causal pathways may contribute: weight loss reducing joint load, HCAR2-NF-κB suppression, glycemic stabilization reducing metabolic-driven inflammation, and BHB direct neuromodulation. For chronic pain management, omega-3, celecoxib, PEMF therapy (7.3/10), and lifestyle interventions dominate first-line. |
| ○ Geriatric / Aging Population | 4.5 | Mixed signals in older adults. Sarcopenia risk on protein-restricted variants is real; standard keto provides adequate protein but classical 4:1 does not. Bone density risk amplifies in elderly. Cognitive benefits in MCI populations apply only to non-APOE4 carriers. Cardiovascular risks (LDL/ApoB rise) compound with age-related atherosclerosis. For geriatric metabolic optimization, Mediterranean dietary patterns dominate evidence. Keto in older adults requires intensive monitoring (ApoB, BMD, sarcopenia markers, APOE genotype check) and is rarely first-line. For specific clinical indications (refractory epilepsy, T2D reversal under Virta-class supervision) it remains valid. |
| ○ Cardiovascular | 4.0 | Triglyceride and HDL improvements directionally favorable per multiple trials, but LDL/ApoB rise in responders complicates the cardiovascular ledger. The recent lipid-panel work lipid meta-analysis of 27 RCTs confirmed KD increases total/LDL/HDL cholesterol and decreases triglycerides. AHA 2026 ranks keto LEAST aligned with cardiovascular guidance among reviewed dietary patterns. The KETO-CTA Longitudinal study retraction removed the strongest pro-LMHR safety claim. For users with LMHR phenotype, the Iatan 2024 UK Biobank cohort shows HR 2.18 MACE on long-term LCHF. Cardiovascular evaluation should drive any chronic keto decision. |
| ○ Hormonal / Endocrine | 4.0 | Bimodal by population. Paoli 2022 PLOS One RCT crossover showed significantly lower plasma T3 on keto vs higher-carb diets, mediated by reduced deiodinase activity. Svart 2024 RCT crossover showed keto increased SHBG and decreased free testosterone and free estradiol in both sexes after 3 weeks. In hyperandrogenic PCOS, this is therapeutic per Mavropoulos 2005. In lean eugonadal women, the same SHBG rise plus HPA-axis activation drives amenorrhea and bone loss. Net hormonal score is moderate-to-negative for general population, therapeutic for hyperandrogenic PCOS subset only. |
| ○ Skin / Beauty | 4.0 | Bimodal: acne improvement reported in some users via reduced insulin/IGF-1 sebaceous activity, but keto rash (prurigo pigmentosa) appears in others. The mechanism for prurigo pigmentosa is unclear but ketosis-driven. No large RCTs on skin endpoints. For users with insulin-driven acne, the glycemic effect can be therapeutic. For users prone to prurigo pigmentosa, keto is contraindicated. For broader skin support, collagen peptides, omega-3, astaxanthin, and red light therapy all carry positive evidence. |
| ○ Fertility (Female) | 4.0 | Bimodal by phenotype. Mavropoulos 2005 PCOS keto pilot showed reduced total/free testosterone and improved ovulation in hyperandrogenic PCOS women, a therapeutic effect for that subgroup. In lean eugonadal women, the same SHBG rise plus HPA-axis activation drives amenorrhea and impaired luteal phase per Svart 2024 and Paoli 2022. Net effect splits by population: hyperandrogenic PCOS therapeutic, lean eugonadal women contraindicated. Pregnancy and lactation are categorical avoid-list. Cyclical or modified Atkins approaches may mitigate the eugonadal-women penalty. |
| ○ Gut Health / Microbiome | 3.5 | Bifidobacterium suppression and reduced short-chain fatty acid production are the dominant signals. recent microbiome work (Dahl) RCT confirmed Bifidobacterium reduction at 4 to 12 weeks on a strict ketogenic protocol vs Mediterranean control. Short-chain fatty acid production drops because dietary fiber typically falls below 15 g/day on standard keto. Lower SCFA means reduced colonocyte energy, reduced regulatory T-cell induction, reduced anti-inflammatory tone in the gut. Workarounds: target 15+ g fiber/day from low-net-carb sources (chia, flax, avocado, leafy greens, psyllium). Net negative without aggressive fiber augmentation. |
| ○ Bone / Joint Health | 3.5 | Pediatric 4:1 reduces BMD by up to 68% in long-term cohorts; adult short-term keto shows no significant BMD change in most trials; long-term adult keto bone density is unclear. Calcium and vitamin D supplementation are standard in clinical 4:1 pediatric protocols. The mechanism is multifactorial: chronic mild metabolic acidosis, reduced calcium absorption, possible IGF-1 effects on bone formation. For users prioritizing bone health, vitamin D3 + K2 (8.0/10), creatine, resistance training, and adequate protein dominate. Long-term strict keto carries real bone-density risk, especially in pediatric populations. |
Frequently Asked Questions
Does keto actually work for fat loss long-term?
Keto delivers faster early weight loss but converges with other dietary patterns by 12 months. The DIETFITS trial (Gardner 2018, JAMA, n=609) is the most rigorous head-to-head: low-carb lost 6.0 kg vs low-fat 5.3 kg at 12 months, statistically non-significant, with no diet-genotype or diet-insulin-secretion interaction. The Bueno 2013 meta-analysis showed only 0.91 kg advantage at 12+ months, which the authors called of little clinical significance. Foster 2010 found 7 kg loss in both arms by 24 months. Adherence is the binding constraint: most users do not sustain under 50g carb. Keto works as a SHORT-term elimination diet (8-12 weeks). For sustained fat loss, the dietary pattern that someone can adhere to for years beats any specific macronutrient profile.
Is keto-based T2D reversal real?
T2D reversal via keto is real, but the Virta evidence base is non-randomized and the coaching effect is inseparable from the diet effect. Hallberg 2018 (n=262) showed HbA1c minus 1.3%, 60% remission, 94% insulin elimination at 1 year. Athinarayanan 2024 extended this to 5 years with sustained remission in adherent subset, but the Virta protocol bundles continuous remote physician coaching, daily blood ketone tracking, app-based feedback, and a peer community. Disentangling diet from intensive support is impossible without a randomized comparison. Newer GLP-1 agonist data (semaglutide, tirzepatide) achieves similar HbA1c reductions without dietary restriction. Insulin and sulfonylurea doses must drop on day one. Medical supervision is mandatory.
Is the LMHR phenotype safe to ignore on keto?
The LMHR phenotype (lean, metabolically healthy, on keto, with LDL-C ≥200, HDL ≥80, TG ≤70) is not safely ignored. The cross-sectional KETO Trial 2024 (Norwitz, JACC Adv) found ApoB and LDL did not predict plaque burden cross-sectionally, but the longitudinal KETO-CTA follow-up study was retracted after the pre-registered primary endpoint, change in non-calcified coronary plaque volume, came in at plus 18.9 mm³ (plus 42.8% relative) and was buried in favor of percent atheroma volume. The retraction is the relevant signal. The UK Biobank LCHF MACE cohort (Iatan 2024) found HR 2.18 at 11.8 years and HR 6.68 in the LDL≥5 mmol/L subgroup. ApoB monitoring plus ABCG5 heterozygote screening should be standard.
What about APOE4 carriers and keto?
APOE4 carriers may NOT benefit from keto and may be harmed. Cunnane MCT trials and Krikorian's MCI ketone-monoester work showed cognitive benefit in non-ε4 carriers, but the same trials showed the benefit was absent or reversed in ε4 carriers. APOE4 alters lipid trafficking and brain ketone metabolism in ways that interact with high-fat dietary patterns and may worsen cerebrovascular risk. The population most concerned about Alzheimer's prevention (ε4 carriers) is the population for whom MCT and chronic keto may not work. APOE4 carriers considering keto should run lipid panels (especially ApoB and oxidized LDL) before and during, prefer MUFA-heavy Mediterranean-keto over saturated-fat keto, and consider carb-cycling. APOE genotype testing is a $99 add-on.
Does keto wreck endurance performance?
Keto helps ultra-endurance fat-dominant work but impairs moderate-to-high intensity exercise even after full adaptation. Burke FASTER 2017 (n=21 elite race walkers) showed fat oxidation in keto-adapted athletes doubled to 1.43-1.5 g/min, an unprecedented number, but at race-relevant intensities the absolute oxygen cost was higher, indicating reduced exercise economy. The 10K time was 1.6% slower (non-significant in this small n). For ultra-endurance events at low intensity, keto-adapted athletes can theoretically run nearly indefinitely on fat stores. For events requiring sustained intensities above 75% VO2max, keto produces real performance decrement. The targeted ketogenic diet (TKD) protocol from Ketogains uses 0.05g glucose per lb lean body mass peri-workout to bridge this gap for resistance training.
Does keto disrupt female hormones?
Keto can disrupt female hormones, particularly in lean active women on long-term protocols. Paoli 2022 PLOS One RCT crossover showed significantly lower plasma T3 on keto vs higher-carb diets, mediated by reduced deiodinase activity. Tissue hypothyroidism with normal TSH is the clinically missed pattern: fatigue, cold intolerance, hair loss, menstrual disruption. Svart 2024 showed keto increased SHBG and decreased free testosterone and free estradiol in both sexes after 3 weeks. In hyperandrogenic PCOS, this is therapeutic per Mavropoulos 2005, keto reduced total and free testosterone and improved ovulation. In lean eugonadal women, the same SHBG rise plus HPA-axis activation drives amenorrhea and bone loss.
Should I cycle keto or stay strict?
Cyclical keto wins on adherence, hormonal preservation, and exercise performance, while strict keto wins for refractory epilepsy and aggressive T2D reversal. Strict (under 20g net carb continuous) maintains BHB above 1 mM and is the protocol for pediatric epilepsy and Westman-Duke T2D work. Cyclical (5-6 days strict + 1-2 day carb refeed) preserves glycogen-dependent exercise performance, mitigates T3 suppression per Paoli 2022, and is more sustainable beyond 3 months. Targeted keto (TKD, 0.05g glucose per lb LBM peri-workout) is a third path for resistance trainees. Nick's personal protocol shifted from strict 40g total to strict 20g net to cyclical at the 6-month mark for exactly these reasons. The keto-purity framing is not supported by efficacy data.
What about the KETO-CTA retraction?
The KETO-CTA Longitudinal study was retracted last year after the pre-registered primary endpoint (change in non-calcified coronary plaque volume) was buried and percent atheroma volume (a secondary endpoint) was foregrounded. The pre-registered ΔNCPV came in at plus 18.9 mm³ (plus 42.8% relative increase over the follow-up period), a directionally adverse finding inconsistent with the trial's framing. Endpoint switching is a documented research-integrity pattern: when the registered primary fails, secondary endpoints get promoted to headline status. The investigator group has substantial financial conflicts. The retraction does not prove keto causes plaque progression in LMHR, but it removes the strongest pro-LMHR longitudinal safety claim. The original cross-sectional KETO Trial 2024 remains valid but is not a longitudinal safety endpoint.
What does keto do to the gut microbiome?
Keto suppresses Bifidobacterium and reduces short-chain fatty acid production. The Cell Reports Medicine microbiome RCT (Dahl) confirmed Bifidobacterium reduction at 4 to 12 weeks on a strict ketogenic protocol vs Mediterranean control. Short-chain fatty acid production drops because dietary fiber typically falls below 15 g/day on standard keto. Lower SCFA means reduced colonocyte energy, reduced regulatory T-cell induction, reduced anti-inflammatory tone. Workarounds: target 15+ g fiber/day from low-net-carb sources (chia, flax, avocado, leafy greens, psyllium), include fermented foods (sauerkraut, kimchi, kefir if dairy-tolerant), consider acacia or partially hydrolyzed guar gum prebiotic. The microbiome signal does not appear in shorter (4-week) trials and is not catastrophic, but it is a real opportunity-cost penalty.
How This Score Could Change
BioHarmony scores are living assessments. New research, regulatory changes, or personal context can shift the score up or down. These are the most likely scenarios that would change this intervention's rating.
| Scenario | Dimensions changed | New score |
|---|---|---|
| Registered hard-outcome RCT (keto vs Mediterranean) shows MACE benefit | Evidence 3.8→4.5, Efficacy 3.5→4.0, Safety 4.0→3.5 | 6.7 / 10 (👍 Worth trying) |
| New large-cohort data confirms LMHR phenotype longitudinal CV harm | Safety 4.0→4.5, Bioindividuality 3.5→3.0, Evidence 3.8→4.0 | 4.8 / 10 (⚖️ Neutral / lower edge) |
| GLP-1 + keto combination trials show synergy without DKA | Efficacy 3.5→4.2, Safety 4.0→3.8 | 6.2 / 10 (👍 Worth trying) |
| Microbiome research validates fiber-augmented keto preserves Bifidobacterium | Side Effects 3.0→2.5, Breadth 4.0→4.3 | 5.9 / 10 (👍 Worth trying lower edge) |
| Endpoint-switching audit retracts 2-3 additional pro-keto trials | Evidence 3.8→3.0, Safety 4.0→4.2 | 4.6 / 10 (⚠️ Caution) |
Key Evidence Sources
- Gardner DIETFITS 2018 (JAMA, n=609, 12 mo). RCT. Most rigorous modern head-to-head. Low-carb minus 6.0 kg vs low-fat minus 5.3 kg, NS. No diet-genotype or diet-insulin-secretion interaction. Defining null finding for biohacker keto framing.
- Hallberg/Virta 2018 (n=262, 1 yr). Non-randomized prospective cohort. T2D HbA1c minus 1.3%, 60% remission. Bundled with intensive coaching. Coaching effect inseparable from diet effect.
- Athinarayanan Virta 5-year 2024. Long-term follow-up. Sustained T2D remission in adherent subset. Same coaching-vs-diet confound as Hallberg 2018.
- Bueno 2013 meta-analysis (≥12 mo trials). Meta-analysis. Minus 0.91 kg LCKD vs LFD at ≥12 mo. Authors call this of little clinical significance. Methodological war with Mansoor 2016 (≥6 mo cutoff captured peak divergence).
- Foster 2010 (AIM 2-yr follow-up). RCT. Convergence to about 7 kg both arms by 24 months. Persistent HDL plus 23% on low-carb arm. Adherence is the binding constraint.
- Iatan 2024 UK Biobank LCHF (JACC Adv, n=305, 11.8 yr). Observational cohort. MACE HR 2.18 (95% CI 1.39-3.43); LDL≥5 subgroup HR 6.68. Single 24h dietary recall is a severe limitation but real long-term follow-up. Contradicts Sci Rep 2024 NHANES.
- NHANES DKR cohort 2024 (Sci Rep, n=43,776). Observational cohort. Higher dietary ketogenic ratio: HR 0.76 all-cause mortality, no CV mortality increase. Directly contradicts UK Biobank Iatan. Neither is an RCT.
- Retterstol 2021 (n=30 healthy young women, 4 wk). Controlled feeding trial. 4% carb / 77% fat / 19% protein. LDL-C rose in EVERY participant; treatment effect plus 1.82 mM (p below 0.001), ApoB rose, small-dense LDL rose. Lean healthy women = biohacker target population.
- Paoli 2022 (PLOS One, RCT crossover). RCT crossover. Plasma T3 significantly lower on KD vs HC/LF. Deiodinase impairment mechanism. Tissue hypothyroidism with normal TSH especially in women on levothyroxine.
- Svart 2024 (RCT crossover, n=24, 3 wk). RCT crossover. KD increased SHBG, decreased free T and free E2 in BOTH sexes. Therapeutic in hyperandrogenic PCOS; harmful in lean eugonadal women.
- Burke FASTER 2017 (n=21 elite ultra-endurance). Crossover trial. Fat oxidation 1.43-1.5 g/min in keto-adapted (unprecedented). Increased absolute O2 cost at race intensities, 1.6% non-significant 10K impairment. Ultra-endurance fat-dominant work suitable; moderate-high intensity impaired even after adaptation.
- Norwitz KETO Trial 2024 (JACC Adv, n=80, cross-sectional). Cross-sectional. ApoB and LDL did NOT predict plaque burden in LMHR phenotype. Longitudinal follow-up KETO-CTA RETRACTED 2025 after pre-registered primary endpoint (ΔNCPV plus 18.9 mm³, plus 42.8% relative) buried in favor of percent atheroma volume.
- Kuchkuntla SGLT2i + keto euglycemic DKA case series. Case series. Catastrophic interaction. Glucose 150-250 mg/dL while pH below 7.30, bicarbonate below 18 mEq/L. Both drug and diet drive ketogenesis; drug also reduces renal ketone clearance. Mandatory screen.
- Mavropoulos 2005 PCOS keto pilot. Small pilot trial. PCOS hyperandrogenic women: keto reduced total/free testosterone, improved insulin and ovulation. Therapeutic in this subpopulation. Generalization to eugonadal women NOT supported.
Other interventions for Metabolic Health
See all ratings →📊 How BioHarmony scoring works
BioHarmony translates a weighted expected-value calculation into a reader-facing 0–10 score. 5.0 is neutral (benefits and risks balance). Above 5 = benefits outweigh risks; below 5 = risks outweigh benefits.
Harm-type downsides (safety risk, side effects, reversibility, dependency) carry a 1.4× precautionary multiplier. Harm weighs more than benefit. Opportunity-type downsides (financial cost, time/effort, opportunity cost) are subtracted at face value.
Use case subratings are independent assessments of how well the intervention addresses specific health goals. They are not components of the overall score. Each subrating reflects the scorer's judgment based on use-case-specific evidence, safety, and effect sizes.
Every dimension is evaluated on a 1–5 scale, and the baseline (1) is subtracted before weighting. A perfect intervention with zero downsides contributes zero penalty rather than a residual floor, so top-tier scores are actually reachable.
EV = Upside − Downside
EV = 2.860 − 3.160 = -0.300
EV ranges from −5 to +5. Adding 7 shifts to 2–12, dividing by 12 normalizes to 0–1, then ×10 gives the 0–10 score.
Score = ((-0.300 + 7) / 12) × 10 = 5.6 / 10
