Submitted:
20 October 2026
Posted:
20 October 2026
You are already at the latest version
Abstract
Background: The claim that women inherently require high dietary carbohydrate intake conflates obligatory glucose metabolism with an obligatory intake of exogenous carbohydrate. Before agriculture emerged approximately 12,000 years ago, humans and their ancestors survived in diverse foraging ecologies with variable, sometimes seasonally limited, carbohydrate supplies. Properly formulated ketogenic nutrition is proposed as a broadly applicable strategy engaging elements of this ancestral metabolic flexibility; it is not claimed that all ancestral diets were ketogenic. We hypothesize that some symptoms attributed to ketogenic diets instead reflect inadequate vitamin and mineral intake, altered fluid and electrolyte balance, insufficient protein or energy, medication interactions, or underlying disorders. The 2026 international consensus renamed polycystic ovary syndrome (PCOS) polyendocrine metabolic ovarian syndrome (PMOS). Objective: To evaluate evolutionary and metabolic rationale, evidence in women, differential adverse-event mechanisms, and the hypothesis that intermittent nutritional ketosis may benefit diverse populations. Methods: Critical, nonsystematic narrative synthesis of physiology, evolutionary scholarship, guidelines, clinical studies, reviews, and safety reports available through 20 October 2026; no pooled effect sizes or formal certainty grading were undertaken. Findings: Humans of both sexes can produce glucose endogenously and increase fat and ketone utilization. Dietary evidence in women with PMOS indicates promising short-term changes in insulin-related and menstrual measures, but trials are small, short, and confounded by differences in weight loss. Ancestral dietary patterns included carbohydrate-rich foods in some environments, so evolutionary survival cannot establish superiority of ketogenic nutrition. Nutritional insufficiency and implementation may contribute to symptoms, whereas direct diet effects, lipid responses, performance limitations, and disease- or medication-specific risks also require consideration. Evidence comparing intermittent with continuous ketogenic regimens is inadequate. Interpretation: Integrative Orthomolecular Systems Medicine (I-OM) proposes nutritional adequacy and individualized implementation as modifiable determinants of tolerability; the previously published insulin–cortisol–vitamin C (ICV) model of Cheng, Levy, and Hunninghake (2025) [1] supplies falsifiable, not clinically validated, hypotheses. Conclusion: We hypothesize that properly formulated intermittent ketogenic nutrition can promote metabolic fuel flexibility and may benefit women and men across many adult life stages. This physiological proposition does not establish universal clinical benefit, nor justify generalized use in pregnancy, lactation, childhood, or medically vulnerable groups. Controlled trials should isolate ketosis from energy deficit, nutrient adequacy, and other mechanisms and evaluate meaningful long-term outcomes.
Keywords:
ketogenic diet
; intermittent ketogenic nutrition
; women
; ancestral metabolism
; metabolic flexibility
; PMOS
; PCOS
; insulin resistance
; orthomolecular nutrition
; micronutrients
; insulin–cortisol–vitamin C axis
; safety
1. Introduction: Evolutionary Metabolic Flexibility, Female Physiology, and the ICV Framework
A physiological requirement for glucose does not necessarily imply an obligatory requirement for dietary carbohydrates. Humans possess well-established mechanisms for endogenous glucose production, fatty-acid oxidation, and ketone utilization, enabling adaptation to substantial variations in dietary carbohydrate availability. These metabolic capabilities are shared by women and men and constitute fundamental components of human metabolic flexibility.
For hundreds of thousands to millions of years before the emergence of agriculture approximately 12,000 years ago, humans and their ancestors survived through hunting and gathering under widely varying ecological and nutritional conditions. Although ancestral diets were not uniformly ketogenic and some populations consumed substantial quantities of wild carbohydrate-containing foods, many environments and seasons involved limited or unpredictable carbohydrate availability. Consequently, human metabolic physiology evolved mechanisms supporting transitions between glucose-based and fat-based energy metabolism.
From this evolutionary perspective, properly formulated ketogenic nutrition is proposed as a broadly applicable metabolic strategy that may more closely resemble important aspects of ancestral dietary patterns and metabolic adaptations than modern diets characterized by frequent consumption of refined carbohydrates. Intermittent ketogenic nutrition may be particularly relevant because it allows periodic activation of fatty-acid oxidation and ketone metabolism without necessarily requiring continuous nutritional ketosis.
The proposition that women inherently require higher dietary carbohydrate intake than men deserves critical scientific examination. Female reproductive physiology is influenced by energy availability, insulin signaling, ovarian hormones, stress responses, and nutritional status. However, these physiological differences do not establish an obligatory female-specific requirement for high dietary carbohydrate intake. Furthermore, clinical investigations of ketogenic interventions in women with polyendocrine metabolic ovarian syndrome (PMOS, formerly polycystic ovary syndrome [PCOS]), obesity, and insulin resistance suggest potential metabolic and reproductive benefits, although the available evidence remains heterogeneous and predominantly short-term.
An additional concern is the attribution of adverse effects to ketogenic diets. Symptoms such as fatigue, dizziness, muscle cramps, constipation, hair shedding, menstrual irregularities, and reduced exercise tolerance may arise from multiple overlapping causes. These include insufficient energy or protein intake, inadequate vitamins and micronutrients, electrolyte imbalances, inappropriate fasting, medication interactions, and previously unrecognized endocrine or metabolic conditions. From an Integrative Orthomolecular Systems Medicine (I-OM) perspective, many reported adverse effects may reflect improper dietary implementation or correctable nutritional inadequacies rather than intrinsic harmful effects of nutritional ketosis. This interpretation represents an important clinical and research hypothesis, while acknowledging that direct effects of carbohydrate restriction and individual susceptibility must also be investigated.
The Insulin–Cortisol–Vitamin C (ICV) framework was previously proposed by Cheng, Levy, and Hunninghake (2025) as an integrative model connecting insulin regulation, cortisol-mediated stress physiology, and vitamin C biology in metabolic and hormonal homeostasis [1]. The present manuscript does not introduce or claim novelty for the ICV framework itself. Instead, it explores its potential application to continuous and intermittent ketogenic nutrition, particularly in women, and examines whether interactions among insulin sensitivity, stress physiology, micronutrient adequacy, and metabolic flexibility may help explain variations in dietary responses and reported adverse effects. These proposed applications remain hypothesis-generating and require prospective clinical validation.
The central hypothesis of this review is that properly formulated intermittent ketogenic nutrition may provide metabolic benefits across sexes and life stages by promoting the physiological capacity to transition between glucose utilization, fatty-acid oxidation, and ketone metabolism. We further hypothesize that optimizing nutritional adequacy—including appropriate vitamin and mineral intake—may prevent or alleviate some adverse effects commonly attributed to ketogenic dietary interventions. Broad physiological applicability, however, should be distinguished from demonstrated universal clinical benefit or safety, especially during pregnancy, lactation, childhood, and in patients with specific metabolic disorders or medication-related risks.
Accordingly, this critical narrative review aims to: (1) examine the evolutionary and physiological rationale for ketogenic and intermittent ketogenic nutrition; (2) evaluate available female-specific clinical evidence, including recent developments concerning PMOS; (3) distinguish intrinsic effects of carbohydrate restriction from potentially modifiable nutritional and clinical factors; and (4) extend the previously published ICV framework into testable hypotheses concerning metabolic adaptation, nutritional optimization, and heterogeneous responses to ketogenic interventions.
The objective is not to presume that ketogenic nutrition is universally superior, but to establish a scientifically coherent foundation for investigating its broader metabolic potential, appropriate formulation, and clinical applications.
2. Scope, Evidence Hierarchy, and Methodological Limitations
This is a critical narrative review and hypothesis, not a systematic review. Selected sources were located through focused examination of peer-reviewed studies, clinical guidelines, and reference lists addressing ketosis, women, reproductive outcomes, PMOS/PCOS, exercise nutrition, adverse effects, and metabolic disease, emphasizing evidence available up to 20 October 2026. Controlled human studies and clinical guidance were prioritized for claims of benefit and safety; physiologic studies informed mechanistic interpretation, and case reports were used to identify hazards rather than estimate their frequency. No prospective protocol, exhaustive database search, duplicate screening, risk-of-bias assessment, or meta-analysis was conducted. Consequently, selection bias remains possible and neither absence of reports nor consistent short-term effects establishes long-term safety. For this review, distinctions are maintained among: (A) human comparative interventional findings, (B) systematic synthesis or clinical guidelines, (C) established human physiology, and (D) hypotheses. The review does not claim to be an exhaustive mapping of all available literature, and conclusions about between-diet superiority are provisional where head-to-head adequately powered trials are scarce. The search was targeted rather than exhaustive: source titles, journal bibliographic information, and DOI metadata were checked for selected central and newly added references. Evidence from reviews overlapping in their underlying primary studies was not treated as independent replication. Where available, claims about sex-specific outcomes were linked to female-specific studies rather than generalized from mixed-sex cohorts.
3. Physiologic and Evolutionary Boundaries
3.1. Glucose Requirements Versus Carbohydrate Intake
Erythrocytes require glucose, and the brain retains a glucose requirement during prolonged ketosis. Hepatic and renal gluconeogenesis from lactate, glycerol, and glucogenic amino-acid substrates, together with glycogen turnover and increased ketone oxidation, permit adaptation to reduced carbohydrate supply in humans of both sexes. Fatty acids cannot replace glucose in all tissues; nor does adaptation imply that dietary carbohydrate never has useful physiologic or performance roles. Pregnancy and lactation have special energy and nutrient demands that cannot be inferred safely from nonpregnant adult physiology.
3.2. Evolutionary Context: Hunter-Gatherers, Dietary Variability, and a Testable Hypothesis
The development of agriculture roughly 12,000 years ago is recent relative to the emergence of Homo sapiens about 300,000 years ago and the much longer evolutionary history of the genus Homo. Foraging, hunting, scavenging, and collection of wild plants preceded farming for hundreds of thousands to millions of years. Women as well as men lived and reproduced across ecologies with variable food supplies. These observations are compatible with substantial evolutionary selection for flexible energy regulation, including the ability to mobilize stored fat during episodes of reduced food or carbohydrate availability [18,19].
They do not, however, prove that hunter-gatherers generally consumed ketogenic diets. Ethnographic and archaeological evidence documents considerable use of fruits, honey, tubers, and cooked starches in some populations; energy contributed by plant and animal foods varied markedly by habitat and season [19,20]. Indeed, an influential alternative evolutionary interpretation assigns an important role to cooked starch in meeting hominin energy needs [20]. This counterevidence is central, not incidental: survival without agriculture cannot be used to deduce that prehistoric women lacked dietary carbohydrate, or that modern ketosis necessarily reproduces a single 'natural' ancestral diet. Neanderthal dental calculus has also documented plant microfossils and evidence of consumption of cooked plant foods, further rebutting any assumption of uniformly carbohydrate-free ancestral nutrition [21].
The more defensible evolutionary proposition is that contemporary continuous access to refined starches and sugars differs from many historical food environments, whereas periodic shifts toward fatty-acid oxidation and ketogenesis are part of retained human physiology. The ancestral-flexibility hypothesis predicts that suitably designed periods of nutritional ketosis may benefit some modern metabolic phenotypes, particularly those characterized by hyperinsulinemia, while sparing adequate protein, energy, and essential nutrients. Comparative outcome studies—not evolutionary analogy alone—must test whether such cycles outperform other nutrient-dense eating patterns.
3.3. Operational Definitions
For this article, a ketogenic dietary intervention refers to a diet intended to achieve measurable nutritional ketosis, ideally verified using blood β-hydroxybutyrate (BHB), rather than one defined solely by a carbohydrate gram target. A low-carbohydrate diet may not cause sustained ketosis. Intermittent ketogenic nutrition means prescribed alternating phases of ketogenic and nonketogenic intake; intermittent fasting concerns eating timing and can occur with any macronutrient ratio. Study protocols must specify grams and percentages of carbohydrate, energy and protein targets, fat quality, dietary fiber, duration, cycle structure, medication adjustments, and ketone exposure; a universal clinically validated BHB threshold for efficacy has not been demonstrated.
4. Female Endocrine Physiology and Life-Course Considerations
Women are not metabolically homogeneous. Ovarian hormones, body composition, insulin sensitivity, sleep, stress, training load, and energy balance vary with puberty, menstrual cycles, pregnancy, postpartum changes, and menopause. Glucose and fat oxidation can vary across hormonal states, but evidence is insufficient to derive a universal sex-specific carbohydrate minimum or to conclude that physiological ketosis itself causes menstrual dysfunction.
The hypothalamic–pituitary–ovarian axis is responsive to energy deficiency, illness, and training stress. Rapid weight loss, inadequate energy or protein, and persistent low carbohydrate availability may occur together in dietary interventions. An irregular menstrual cycle during a ketogenic intervention therefore requires evaluation of pregnancy, energy availability, weight change, thyroid dysfunction, prolactin disorders, PMOS, exercise, and medications. Conversely, symptom improvement following nutritional repletion does not prove that ketosis was independently safe. Studies should report ovulatory measures, cycle regularity, fertility intent, contraceptive use, reproductive hormones when appropriately timed, and long-term bone outcomes rather than relying on anthropometric changes.
In perimenopausal and postmenopausal women, shifts in body fat distribution, insulin sensitivity, and cardiovascular risk make metabolic interventions relevant, but evidence specifically supporting sustained ketosis over alternative well-formulated dietary patterns remains limited. LDL cholesterol and apolipoprotein B (ApoB), blood pressure, lean mass, and strength should be tracked alongside glycemia and waist circumference.
5. PMOS: The Most Compelling Female-Specific Clinical Test Case
In May 2026 an international consensus introduced polyendocrine metabolic ovarian syndrome (PMOS) to replace the misleading term polycystic ovary syndrome (PCOS), with a staged implementation rather than an immediate change in all clinical codes, guidelines, and trial databases [2]. Both labels are retained here for searchability; historical article titles retain their original PCOS terminology. The name underscores reproductive, endocrine, metabolic, and psychological manifestations but does not imply that all women with PMOS have insulin resistance or require ketogenic therapy.
Systematic reviews and meta-analyses report short-term changes in weight, waist measures, insulin resistance, and several hormonal outcomes following ketogenic interventions in women identified under the earlier PCOS label [3,4,5]. These syntheses often pool differing interventions, including very-low-energy ketogenic therapies, and cannot always disentangle ketosis from energy restriction and weight loss. Changes in surrogate hormonal measurements are not equivalent to demonstrated improvements in live birth or long-term health.
A September 2026 open-label randomized exploratory trial enrolled 60 Chinese women with PCOS, comparing a ketogenic regimen with a simulated Mediterranean diet for 12 weeks. The authors reported greater improvements in insulin-related measures and menstrual regularity in the ketogenic group [6]. However, the study was single-center, short-term, had small groups, involved prescribed energy deficit in both arms, and produced differences in weight loss; these features prevent attribution of outcomes to BHB or carbohydrate restriction alone. Its reported menstrual findings warrant replication with prespecified ovulation and fertility endpoints, independent adjudication, and longer follow-up. The trial also reported exploratory analyses adjusting for weight loss, but residual confounding, limited sample size, and short follow-up preclude attribution to ketones specifically [6].
Countervailing guideline evidence and temporal context. The 2023 International Evidence-based Guideline for PCOS found insufficient evidence to recommend one dietary composition over others for anthropometric, metabolic, reproductive, or psychological outcomes [16]. This remains an important benchmark; newer ketogenic studies published in 2025–2026 may inform future guideline reassessment but do not retrospectively invalidate the earlier appraisal. A 2026 meta-analysis focused on biochemical hyperandrogenism in women with overweight or obesity also underscores the narrow phenotype and surrogate nature of many endpoints [17]. Moreover, a systematic review of eucaloric, normoproteic ketogenic interventions identified only seven eligible gonadal-hormone studies overall, of which just three enrolled women exclusively [15]. Results from selected women with PMOS must not be generalized to all women.
An important trial-specific distinction. The September 2026 randomized trial verified ketosis with urine and blood measures, but prescribed the same nominal 500-kcal/day deficit in each arm and nevertheless observed unequal weight-loss outcomes [6]. Thus the randomized comparison evaluates a dietary strategy, not a pure ketone effect. A claim of ketone-specific reproductive benefit requires mediation analysis or a differently controlled experiment.
The relevant clinical inference is potential benefit for selected women with PMOS, not evidence that all PMOS phenotypes are metabolically identical or that ketogenic diets outperform all other nutrition interventions.
5.1. Applicability of the 2026 Name Change
The PMOS label was agreed through a global stakeholder process but is being implemented progressively [2]. It does not create a new diagnosis, independently validate a new biochemical test, or alter historical participants’ inclusion criteria. The older term PCOS should be retained in database search strings, prior trial titles, and contexts that describe the diagnostic standards in force when evidence was collected. In this paper, PMOS denotes the same syndrome now being renamed; no result is attributed to the terminology change itself.
6. Obesity, Type 2 Diabetes, and Cancer: Differing Evidence Thresholds
6.1. Obesity and Type 2 Diabetes
Carbohydrate restriction can improve short-term glycemic control and may facilitate reduced diabetes medication use in selected adults; outcomes depend on diet definition, weight change, comparator diet, support intensity, and adherence [7]. The 2026 American Diabetes Association (ADA) Standards recognize carbohydrate reduction as an option for some adults with diabetes while emphasizing individualized medical nutrition therapy, long-term reassessment, and medication safety [8]. The strongest glycemic evidence is not uniformly female-specific. Claims about ketone-mediated effects should therefore be separated from trials demonstrating benefits of carbohydrate reduction without measured ketosis.
Insulin and sulfonylureas may require prompt dose adjustment to avoid hypoglycemia. Very-low-carbohydrate diets should be avoided with sodium–glucose cotransporter 2 (SGLT2) inhibitors because of ketoacidosis risk under ADA 2026 guidance [8]. This is an interaction between diet and treatment, not an instruction to stop prescribed medication independently.
6.2. Oncology
Altered glucose utilization and insulin signaling motivate research into ketogenic interventions in oncology, but tumors are metabolically heterogeneous, and normal tissues also require nutrition. There is no basis for calling a strict ketogenic diet a generally proven anticancer treatment or for substituting it for evidence-based cancer care. Cancer-associated weight loss, cachexia, impaired intake, and treatment toxicity can make dietary restriction inappropriate. Women with cancer should only consider intensive dietary intervention in an oncology and nutrition context with adequate monitoring. Proposed anticancer mechanisms and clinical efficacy must be treated as separate evidentiary questions.
7. Female Athletes: Energy Availability is not the Entire Explanation
Relative energy deficiency in sport (REDs) is a crucial alternative explanation for menstrual, immune, bone, and performance problems during restrictive eating. However, the 2023 International Olympic Committee (IOC) consensus also reviews research suggesting that low carbohydrate availability can modify selected bone, immune, and iron-related biomarkers partly independently of total energy availability [9]. Some evidence derives from small, short-term athlete studies and is not equivalent to demonstrated long-term clinical harm in every woman following a ketogenic diet. Nevertheless, the evidence rejects the simple assertion that all exercise-related adverse findings disappear once calories are matched.
Trials should cross energy availability (adequate versus low) with carbohydrate availability (moderate versus low), adjust for training load, and assess menstrual and ovulatory function, ferritin and hepcidin when indicated, bone turnover and density over appropriate time frames, and sport-specific performance. Observational symptoms should not be reduced to either low calories alone or ketosis alone.
8. Adverse-Event Attribution: A Causal-Discrimination Framework
The core contribution of this review is an explicit distinction among (i) effects resulting from carbohydrate restriction or ketosis, (ii) effects resulting from the way an intervention is implemented, (iii) effects resulting from underlying conditions and therapies, and (iv) effect modification by patient characteristics. These pathways can coexist and should not be treated as mutually exclusive.
8.1. Keto-Induction, Sodium Balance, and Hydration
Lower insulin exposure with carbohydrate restriction can promote natriuresis and shifts in water balance. Headache, dizziness, reduced exercise capacity, orthostatic symptoms, fatigue, and cramps during diet initiation are plausible consequences; direct individual-level mechanistic attribution remains incomplete [10]. Assess hydration, blood pressure, dietary intake, kidney function, and indicated electrolytes. Empirical high-dose sodium or potassium advice is inappropriate for patients with kidney or heart disease, hypertension, or relevant medication exposure.
8.2. Calories, Protein, and Reproductive Function
Reduced appetite or restrictive food lists can produce unintended energy deficit. In physically active women, inadequate energy availability may compromise reproductive function and recovery. Prospective measurement of total energy and protein intake, weight-loss rate, training, menstrual status, and body composition is essential. If symptoms persist, evaluate relevant clinical diagnoses rather than recommending continuation on the assumption that adaptation will necessarily resolve them [9].
8.3. Vitamins, Minerals, and Protein: Orthomolecular Determinants of Implementation
A properly formulated ketogenic diet is not synonymous with a monotonous menu of fats and protein, nor does carbohydrate restriction itself automatically produce micronutrient deficiency. Restrictive food selection, pre-existing insufficiency, inadequate food volume, and dietary displacement can nevertheless reduce intakes of magnesium, folate, thiamine, iron, iodine, selenium, potassium, calcium, and vitamin C. Their contribution to symptoms should be investigated explicitly rather than assuming that fatigue, cramps, hair shedding, constipation, or reduced exercise tolerance demonstrate an adverse effect of ketone bodies. A cross-sectional study found lower thiamine intakes and/or status indicators in adults following low-carbohydrate/ketogenic diets, offering a concrete example of potential nutrient inadequacy while not demonstrating that ketosis caused deficiency [22]. Adequate vitamin C can be supplied within a ketogenic food pattern through appropriate foods or supplementation if intake is insufficient; high-dose vitamin C has not been shown in controlled trials to prevent ketogenic adverse effects. Nutrient status, clinically appropriate correction, and dose-related safety should be specified prospectively.
Magnesium is relevant to neuromuscular function and many ATP-dependent reactions, although cramps are nonspecific and a deficiency should not be assumed without assessment. Sodium and potassium need particular attention during the initial renal water and sodium shifts; supplementation is not automatically appropriate in kidney disease or with renin–angiotensin drugs and diuretics. Thiamine and other B vitamins support cellular intermediary metabolism and neurological function. Iron and folate warrant attention where menstruation, pregnancy potential, or restrictive intake changes risk. Iodine and selenium contribute to thyroid physiology but excessive supplementation can be harmful. Vitamin C is required for collagen synthesis and contributes to carnitine biosynthesis and antioxidant function; inadequate intake may occur when vitamin-C-containing foods are displaced, but neither high-dose vitamin C nor an ICV-specific vitamin C target has been shown to prevent ketogenic intolerance.
Protein, calories, and dietary quality are coequal determinants. Insufficient energy or protein can worsen recovery, lean-mass preservation, and hypothalamic–pituitary–ovarian function. Evidence-based correction of demonstrated insufficiencies and a practical nutrient-rich food plan are preferable to automatically attributing symptoms to ketosis. From an orthomolecular perspective, individually tailored supplementation merits controlled investigation, including dose-response and adverse-event monitoring; routine megadoses should not be presumed protective.
An informative test would randomize participants using the same ketogenic macronutrient prescription to standard nutritional support versus individualized assessment and correction of deficiencies, tracking symptom incidence, biomarkers, adherence, and prespecified safety endpoints. A second factorial design could compare ketogenic with nonketogenic diets at matched energy, protein, and micronutrient adequacy. These designs would test the 'improper implementation' hypothesis without circularly defining all negative results as improper use.
8.4. Medication Interactions, Lipids, and Clinical Red Flags
Medication-associated hypoglycemia or ketoacidosis, diuretic-related volume changes, thyroid disease, anemia, eating disorders, and postpartum states may mimic diet-related intolerance [8,11]. Some people experience marked LDL-C and ApoB increases on high-fat ketogenic diets; such responses require attention even when triglycerides or glucose improve. Severe vomiting, progressive weakness, abdominal pain, respiratory distress, syncope, or altered mental status warrant urgent medical assessment, not routine “keto flu” reassurance. A meta-analysis of three randomized trials in normal-weight adults found increased LDL-C and ApoB with ketogenic diets versus controls, emphasizing that favorable glycemic or triglyceride responses do not negate possible atherogenic lipoprotein changes [23].
8.5. Environmental Exposures: Narrow, Testable Hypothesis
Lipophilic persistent organic pollutants stored in adipose tissue can increase in circulation with substantial weight loss from various causes [12]. This does not establish that pollutants explain typical ketogenic symptoms or that chelation, unspecified detoxification, or routine broad toxin screening improves outcomes. Only credible exposure histories or prespecified research settings justify exposure-specific laboratory evaluation. Weight-loss-related mobilization is a candidate modifier for research, not a default clinical diagnosis.
8.6. Causal Testing
Well-controlled trials should quantify dietary carbohydrate, BHB exposure, calories, protein, fatty-acid composition, fiber and micronutrients, sodium and fluid intake, medication changes, exercise, sleep, and symptom timing. Causal attribution is strongest when energy/protein-matched diets, prespecified adverse events, objective biomarkers, and blinded outcome assessment where feasible are combined. Failure to detect a particular adverse event in a short underpowered trial must not be described as proof of long-term safety.
9. Evidence-to-Practice Matrix
| Clinical setting | Observed evidence | Principal uncertainty | Implication |
| PMOS with overweight or insulin resistance | Short-term improvements in metabolic and some menstrual endpoints [3,4,5,6,17] | Weight loss, durability, live birth, cardiovascular effects | Consider as one monitored dietary option, not proven first-line superiority |
| Type 2 diabetes in adults | Short-term glycemic improvement with carbohydrate restriction [7,8] | Female-specific comparative durability and medication changes | Individualize and adjust glucose-lowering therapies clinically |
| Healthy reproductive-age women | Sparse direct long-term comparative trials [15] | Menstruation, fertility, energy adequacy | Do not infer general superiority or lifetime safety |
| Female athletes | Low energy and low carbohydrate availability may both matter [9] | Long-term bone and reproductive risk and sport-specific performance | Match nutrition to training and monitor symptoms |
| Postmenopausal women | Limited direct ketogenic-versus-comparator evidence | ApoB, bone, lean mass, adherence | Prioritize comparative trials and risk monitoring |
| Pregnancy or lactation | Inadequate efficacy/safety trials; lactation ketoacidosis case reports [8,13,14] | Maternal/infant safety | Do not generalize nonpregnant adult findings |
| This matrix is a qualitative editorial interpretation, not a formal GRADE certainty assessment. | |||
10. Special Populations and Contraindications
Nutritional ketosis is distinct from diabetic or starvation ketoacidosis; the latter are potentially dangerous acid-base disorders. Even so, rare but life-threatening lactation-associated ketoacidosis has been documented with strict carbohydrate restriction, sometimes accompanied by intercurrent infection or other stressors [13,14]. These cases do not establish population incidence, but make a universal claim of safety untenable. The ADA 2026 advises against very-low-carbohydrate plans in pregnancy, lactation, children, people with kidney disease, or people with or at risk for disordered eating, and advises avoidance in users of SGLT2 inhibitors [8]. Specific inborn errors of fat oxidation or carnitine metabolism and selected other metabolic disorders can also contraindicate ketogenic treatment [11].
For type 1 diabetes, pregnancy intention, undernutrition, major comorbidity, and complex pharmacotherapy, specialist assessment is required and ketogenic intervention may be unsuitable. These cautions are clinical safety boundaries rather than evidence that female sex itself makes all ketogenic patterns intrinsically unsafe.
11. Intermittent Ketogenic Nutrition: A Proposed Route to Metabolic Flexibility
Intermittent ketogenic nutrition is central to the author's broader hypothesis. A recurring intervention could involve, for example, defined multiweek ketogenic phases alternated with balanced nonketogenic phases; this is not equivalent to eating within a restricted daily time window. Fatty-acid oxidation occurs even outside ketosis, and ketosis is not an obligatory marker of metabolic flexibility. The proposed advantage of intermittency is repeated exposure to differing fuel environments without permanent carbohydrate exclusion, not proof that cycles are safer or more effective. Because protocols vary, it is currently unwarranted to assert that intermittent ketosis consistently improves adherence, preserves female endocrine function, or lowers adverse-event incidence compared with continuous ketosis. Each pattern should be named and reported independently. Evidence for intermittent fasting provides related background on metabolic switching but should not be interpreted as direct evidence for alternating ketogenic and nonketogenic food composition, which remains a distinct intervention [24]. Evidence about time-restricted eating or fasting is mechanistically adjacent but does not constitute direct clinical evidence for periodic dietary ketosis [24].
Future interventions should define cycle length, minimum and maximum carbohydrate intake, energy balance, protein adequacy, fat quality, measured time spent in ketosis, reproductive-phase timing, wash-in/wash-out criteria, and medications. The relevant comparator is not only continuous ketosis but also an equally supported, energy- and protein-matched nonketogenic diet. The predicted benefit of periodic ketosis—greater flexibility with reduced cumulative restriction—is a hypothesis requiring trial confirmation, not an established protective mechanism.
12. Application of the Previously Published ICV Framework: Mechanisms and Falsifiable Predictions
The ICV framework previously proposed by Cheng, Levy, and Hunninghake (2025) [1] proposes interactions among insulin regulation, stress-related cortisol physiology, and vitamin C biology [1]. Its present use is exploratory. There is no validated ICV biomarker score, clinical treatment algorithm, or trial demonstrating that raising vitamin C concentrations makes ketogenic diets safer for women. Cortisol has circadian, pulsatile, and stress-dependent dynamics; a single uncontextualized measure is not a reliable index of “adrenal balance.”
Hypothesis 1 (baseline metabolic phenotype): Women with higher insulin resistance may show larger short-term glycemic improvements with carbohydrate restriction than metabolically healthy women, after adjustment for weight loss. Falsifier: no meaningful interaction between baseline insulin resistance and randomized diet assignment on prespecified metabolic endpoints.
Hypothesis 2 (energy/stress context): Low energy availability and sustained psychosocial or exercise stress may modify reproductive or sleep responses to carbohydrate restriction. Falsifier: no diet-by-energy-availability or diet-by-stress interaction on prospectively defined outcomes when confounding is controlled.
Hypothesis 3 (orthomolecular nutritional adequacy): Baseline micronutrient insufficiency, including vitamin C insufficiency where present, may predict some nonspecific symptoms regardless of assigned diet. Falsifier: no predictive value of measured baseline deficiency or no improvement from evidence-based correction in a suitable randomized design. This does not justify empirical high-dose vitamin C or claim vitamin C-specific protection from ketosis-related harms.
Hypothesis 4 (intermittency): A defined periodic ketogenic protocol may achieve clinically useful glycemic effects with a different adherence, lipid, symptom, and reproductive profile from continuous ketosis. Falsifier: no prespecified differences in a sufficiently powered trial with objective ketone and dietary exposure data.
These predictions permit empirical testing of ICV-related propositions while explicitly allowing null or contradictory findings.
13. Proposed Research Program
Study A: PMOS phenotype-stratified trial. Randomize adult women with PMOS and insulin resistance to (1) continuous well-formulated ketogenic nutrition, (2) a prospectively defined intermittent ketogenic regimen, or (3) an energy/protein-matched moderate-carbohydrate comparator. Stratify by adiposity, fertility intention, and medication exposure. Follow for 12 months with a longer observational extension. Prespecify HOMA-IR or clamp-derived insulin sensitivity where feasible, menstrual and biochemically confirmed ovulatory endpoints, patient-reported quality of life, ApoB, nutrient adequacy, and adverse events. Pregnancy attempts require separate clinical and ethical planning.
Study B: female athlete factorial trial. In appropriately screened premenopausal adults, vary carbohydrate availability independently of energy availability with standardized training, adequate supervision, and predefined stop rules; assess menstrual endpoints, iron regulation, performance, and suitable bone markers. A short biomarker study cannot resolve long-term fracture risk.
Study C: postmenopausal cardiometabolic feasibility trial. Compare defined continuous and intermittent ketogenic interventions with a high-quality moderate-carbohydrate diet, measuring glycemia, lean mass, physical function, LDL-C/ApoB, blood pressure, adherence, and adverse events. Screen for high baseline cardiovascular and nutritional risk.
Study D: adverse-event mechanism and micronutrient optimization substudies. Nest prospective diet records and structured symptom assessment within trials. Test diet-related natriuresis, low energy availability, medication changes, and deficiencies as competing or joint mediators. Randomize documented-insufficiency correction strategies where ethical and compare nutritionally matched ketogenic and nonketogenic arms. Assess magnesium and vitamin C intake/status when justified, avoiding an unvalidated universal supplementation protocol. Environmental chemical assays should be restricted to exposure-specific or prespecified substantial-weight-loss subgroups; toxicant causality must not be presumed.
14. Prioritized Mechanistic and Reporting Standards
A compelling hypothesis must specify what would not support it. Interventional studies should preregister a primary endpoint, minimum clinically important difference, safety monitoring, and a statistical plan for multiple comparisons. Recruitment and analysis should report reproductive stage, cycle phase, hormonal contraception, PMOS phenotype, medication changes, race/ethnicity where appropriate, baseline nutrient status, and training load. Reports should use intention-to-treat analyses, quantify missing data, and distinguish investigator-assessed adverse events from self-reported symptoms. Nutritional ketosis should be measured prospectively, not inferred from diet labels alone.
Separating causal pathways. An isocaloric, protein-matched randomized comparison estimates the impact of assigning dietary strategies when adherence differs. It cannot by itself attribute outcomes to β-hydroxybutyrate, insulin reduction, reduced energy intake, altered food selection, or shifts in lipid species. Objective exposure, repeated measures, mediation models, and—in a subset—controlled feeding can help, although mediational claims require strong assumptions. An intermittent regimen should specify whether periods outside ketosis permit carbohydrate repletion without overall energy compensation.
Ethics and stopping criteria. Women planning pregnancy, those with significant malnutrition or eating-disorder risk, and patients vulnerable to hypoglycemia or ketoacidosis require individualized protocols or exclusion as appropriate. Prospective research should specify stopping rules for prolonged amenorrhea, excessive weight loss, clinically important ApoB elevation, symptomatic hypoglycemia, or other serious adverse outcomes.
15. Clinical Interpretation and Conclusion
The categorical claim that women require high dietary carbohydrate merely because they are women lacks a biochemical basis. Human evolutionary history, including prolonged reliance on variable foraged resources before agriculture, supports a capacity for metabolic fuel switching. It cannot establish a universally ketogenic ancestral diet, still less an optimal diet for every modern human. The author advances a positive, testable I-OM thesis: properly formulated intermittent ketogenic nutrition may be a broadly applicable method of engaging an evolutionarily conserved capacity for fat oxidation and ketone utilization, particularly where modern metabolic disease is accompanied by impaired insulin regulation.
This position does not require asserting that dietary ketosis is intrinsically free of risk. Indeed, different mechanisms must be separated: nutritional deficiencies or low energy availability due to poor formulation; predictable renal electrolyte shifts induced by carbohydrate restriction; disease and drug interactions; LDL-C/ApoB responses; and effects of carbohydrate availability on performance or endocrine outcomes. The conjecture that a substantial proportion of nonspecific adverse symptoms reflects modifiable implementation is plausible and clinically important, but is not yet quantified or proven. A fully nourished participant who experiences an adverse outcome cannot simply be reclassified post hoc as improperly treated.
In women with PMOS and insulin resistance, new trials justify rigorous follow-up, but pregnancy, lactation, childhood, eating disorders, and medication-associated ketoacidosis risks preclude blanket prescriptions. The decisive agenda is now empirical: define intermittent protocols, measure ketosis and fuel oxidation, match calories and micronutrients, report adverse events regardless of attributed cause, and evaluate reproductive, cardiovascular, functional, and patient-centered outcomes. ICV-related predictions should be preregistered and falsifiable. A shift from theoretical capacity to demonstrated general clinical benefit requires direct comparative evidence; until then, broad applicability remains a hypothesis rather than an established recommendation. The research agenda should explicitly test both the broad-benefit hypothesis and the possibility that healthy subgroups derive no net benefit, or experience unfavorable outcomes, despite documented ketosis.
Evidence triangulation: trials of carbohydrate-reduced diets, studies of objectively measured ketosis, and intermittent-fasting experiments examine related but nonidentical exposures. A nutrient-deficiency association does not establish that correcting it prevents diet-related symptoms. Historical dietary diversity supports metabolic adaptability, not an invariably ketogenic ancestral diet. These distinctions limit universal clinical claims.
Strengths and Limitations
Strengths include explicit separation of inherited fuel-switching physiology from dietary efficacy, inclusion of the 2026 PMOS nomenclature and recent randomized evidence, serious consideration of countervailing guidelines and ancestral starch use, and a prospectively testable framework for adverse-event mechanisms. Limitations include nonsystematic literature selection, absence of quantitative evidence grading, limited studies of genuinely intermittent nutritional ketosis, heterogeneity of dietary implementations, and sparse long-duration female-specific outcomes. These constraints preclude estimates of universal benefit or attribution of adverse events to a single mechanism.
Declarations
Author Contributions
Richard Z. Cheng—conceptualization, literature evaluation, clinical interpretation, critical review, and responsibility for the final scientific content.
Funding
This research received no external funding.
Conflicts of Interest
The author declares no financial conflicts of interest. The author is the founder of I-OM.ai and is professionally involved in integrative orthomolecular medicine education and clinical activities; these interests are disclosed for transparency.
Institutional Review Board Statement
Not applicable; no new human or animal research was conducted.
Informed Consent
Not applicable.
Data Availability
No new datasets were generated or analyzed.
Use of Generative AI
I-OM.ai and ChatGPT (OpenAI) assisted with literature discovery, manuscript structuring, drafting, editorial revisions, and document formatting. The author assumes responsibility for verification of scientific claims and references, interpretation, and final approval. No original experimental data were generated.
Acknowledgments
None reported.
References
- Cheng RZ, Levy TE, Hunninghake R. The Insulin–Cortisol–Vitamin C Axis: A Missing Regulatory Framework in Metabolic and Hormonal Homeostasis—A Narrative Review. Preprints. 2025;2025120217, v3. [CrossRef]
- Teede HJ, Bahri Khomami M, Morman R, et al.; Global Name Change Consortium. Polyendocrine metabolic ovarian syndrome, the new name for polycystic ovary syndrome: a multistep global consensus process. Lancet. 2026;407(10545):2329–2339. [CrossRef]
- Cannarella R, Rubulotta M, Leonardi A, et al. Effects of ketogenic diets on polycystic ovary syndrome: a systematic review and meta-analysis. Reprod Biol Endocrinol. 2025;23:74. [CrossRef]
- Tosatti JAG, Magalhães FMV, Gomes KB. Effects of the very low-carbohydrate ketogenic diet in women with polycystic ovary syndrome: a systematic review with meta-analysis of clinical trials. Br J Nutr. 2026;135:178–193. [CrossRef]
- Arsenaki E, Stathi D, Triantafyllidis KK, et al. The effects of ketogenic diet on polycystic ovary syndrome: a systematic review and meta-analysis. Clin Nutr. 2026;56:106535. [CrossRef]
- Wang M, Xie L, Ma L, et al. Efficacy of ketogenic diet on glycemic-lipid metabolism and menstrual regularity in women with polycystic ovary syndrome: a randomized controlled trial. Front Nutr. 2026;13:1920489. [CrossRef]
- Goldenberg JZ, Day A, Brinkworth GD, et al. Efficacy and safety of low and very low carbohydrate diets for type 2 diabetes remission: systematic review and meta-analysis of published and unpublished randomized trial data. BMJ. 2021;372:m4743. [CrossRef]
- American Diabetes Association Professional Practice Committee. Facilitating Positive Health Behaviors and Well-being to Improve Health Outcomes: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl 1). [CrossRef]
- Mountjoy M, Ackerman KE, Bailey DM, et al. 2023 International Olympic Committee's (IOC) consensus statement on Relative Energy Deficiency in Sport (REDs). Br J Sports Med. 2023;57:1073–1098. [CrossRef]
- Skartun O, Smith CR, Laupsa-Borge J, Dankel SN. Symptoms during initiation of a ketogenic diet: a scoping review of occurrence rates, mechanisms and relief strategies. Front Nutr. 2025;12:1538266. [CrossRef]
- Dyńka D, Rodzeń Ł, Rodzeń M, et al. The ketogenic diet is not for everyone: contraindications, side effects, and drug interactions. Ann Med. 2026;58(1):2603016. Published online 4 January 2026. [CrossRef]
- Jansen A, Lyche JL, Polder A, Aaseth J, Skaug MA. Increased blood levels of persistent organic pollutants (POP) in obese individuals after weight loss—a review. J Toxicol Environ Health B Crit Rev. 2017;20:22–37. [CrossRef]
- Nnodum BN, Oduah E, Albert D, Pettus M. Ketogenic diet-induced severe ketoacidosis in a lactating woman: a case report and review of the literature. Case Rep Nephrol. 2019;2019:1214208. [CrossRef]
- Liu MC, Bertsch RA. Case report: lactation ketoacidosis can complicate the ketogenic diet. Perm J. 2021;25:20.162. [CrossRef]
- Lima M, Fabe J, Grigolon RB, et al. Effects of the ketogenic diet on gonadal hormones: a systematic review. Nutr Rev. 2026;84:1129–1139. [CrossRef]
- Teede HJ, Tay CT, Laven JJE, et al. Recommendations from the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. J Clin Endocrinol Metab. 2023;108:2447–2469. https://pmc.ncbi.nlm.nih.gov/articles/PMC10477934/.
- Maseroli E, Alfaroli C, Cirillo M, et al. Ketogenic diet as a therapeutic intervention for biochemical hyperandrogenism in PCOS women with obesity or overweight: a meta-analytic study. J Endocrinol Invest. Published online 13 June 2026. [CrossRef]
- Cahill GF Jr. Fuel metabolism in starvation. Annu Rev Nutr. 2006;26:1–22. [CrossRef]
- Pontzer H, Wood BM, Raichlen DA. Effects of evolution, ecology, and economy on human diet: insights from hunter-gatherers and other small-scale societies. Annu Rev Nutr. 2021;41:363–385. [CrossRef]
- Hardy K, Brand-Miller J, Brown KD, Thomas MG, Copeland L. The importance of dietary carbohydrate in human evolution. Q Rev Biol. 2015;90:251–268. [CrossRef]
- Henry AG, Brooks AS, Piperno DR. Microfossils in calculus demonstrate consumption of plants and cooked foods in Neanderthal diets (Shanidar III, Iraq; Spy I and II, Belgium). Proc Natl Acad Sci USA. 2011;108(2):486–491. [CrossRef]
- Churuangsuk C, Catchpole A, Talwar D, Welsh P, Sattar N, Lean MEJ, Combet E. Low thiamine status in adults following low-carbohydrate/ketogenic diets: a cross-sectional comparative study of micronutrient intake and status. Eur J Nutr. 2024;63:2667–2679. [CrossRef]
- Joo M, Moon S, Lee YS, Kim MG. Effects of very low-carbohydrate ketogenic diets on lipid profiles in normal-weight (body mass index <25 kg/m2) adults: a meta-analysis. Nutr Rev. 2023;81(11):1393–1401. [CrossRef]
- de Cabo R, Mattson MP. Effects of intermittent fasting on health, aging, and disease. N Engl J Med. 2019;381(26):2541–2551. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.