In recent years, there has been growing interest in the interplay between metabolic dysfunction and female-specific health conditions, including hormone-related cancers and polycystic ovary syndrome (PCOS). PCOS is a prevalent endocrine disorder affecting women of reproductive age and is frequently associated with metabolic disturbances, including obesity, insulin resistance (IR), and elevated androgen levels (hyperandrogenism) (1–3). Chronic hyperinsulinaemia, a hallmark of modern lifestyle-related metabolic factors, has been identified as a central driver in the development of oestrogen-dominant pathologies, such as breast, endometrial, and ovarian cancers, as well as PCOS (4–15). These conditions often coexist with features of metabolic syndrome (MetS), highlighting the intricate link between reproductive and metabolic health in women. Chronically elevated basal levels of insulin, i.e., hyperinsulinaemia, suppresses ketogenesis, which is detectable as hypoketonaemia, where beta-hydroxybutyrate (BHB) ketone body concentrations are for several consecutive days consistently < 0.5 mmol/L before the evening meal at least three hours post-prandially (16, 17).

Sex hormones, particularly oestrogens, androgens, and progesterone, play a significant role in regulating insulin sensitivity and metabolic function, including processes related to energy balance and inflammatory pathways (18). Oestrogen exerts protective effects on metabolic regulation, especially in glucose homeostasis and insulin sensitivity, and modulates immune cell activation and function through multiple mechanisms. Reduced oestrogen levels, commonly observed during menopause and in certain hormonal disorders, have been associated with an increased risk for type 2 diabetes (T2DM) and cardiovascular disease (CVD).

Oestrogen and testosterone circulate in the bloodstream primarily bound to carrier proteins, most notably the liver-synthesised glycoprotein sex hormone-binding globulin (SHBG) and albumin. SHBG is a key regulator of the bioavailability of circulating sex steroids, and its serum concentrations modulated by both hormonal and metabolic factors, especially insulin (1, 19). Emerging evidence suggests that the balance between testosterone and oestrogen may be associated with carbohydrate metabolism and insulin (20). Within this context, lower circulating SHBG concentrations are often considered indicative of a more androgenic hormonal environment. It has been proposed that observed associations between SHBG and glycaemic markers (i.e., insulin, fasting glucose or HOMA-IR) may reflect the underlying steroid milieu, rather than a direct metabolic function of SHBG itself (21). Additionally, decreased SHBG levels have been linked to an increased risk of developing MetS (22), and show inverse correlations with body mass index (BMI) and waist circumference (1). Supporting these findings, cross-sectional studies in both men and women have demonstrated that higher SHBG levels correlate with a lower risk of T2DM, with robust associations observed in women (20). Lower levels of SHBG are also associated with increased incidence of cancer (13, 14, 23, 24). Despite its critical regulatory function, SHBG is rarely measured in clinical practice, even though it plays an essential role in the modulation of sex hormone bioavailability and intracellular signalling modulation. Many cancers, including breast, endometrial, ovarian, prostate, pancreatic, lung and brain tumours are classified as sex hormone-sensitive malignancies (13–15, 25–28). Alterations in SHBG concentrations can therefore affect not only the circulating levels of active sex steroids but also their signalling activity within target tissues, influencing cancer susceptibility and progression.

Recent clinical studies have also highlighted the impact of dietary carbohydrate restriction, particularly ketogenic diets, on SHBG concentrations and metabolic function in women. In both PCOS and non-PCOS populations, euketonaemia (BHB ≥ 0.5 mmol/L - < 5 mmol/L) interventions have been associated with increased SHBG levels, alongside improvements in insulin sensitivity, glycaemic control, and androgen profiles (16–18, 29–35). These effects are mediated by reductions in serum basal insulin, whereas higher levels of insulin that suppress ketogenesis causing hypoketonaemia, and concomitantly inhibit hepatic SHBG synthesis (36). Consequently, SHBG serves as a sensitive biomarker of metabolic-endocrine status, further refining clinicians’ and research scientists’ ability to detect subclinical hyperinsulinaemia (SCHI) and, therefore, the associated increased risk for certain cancers, cardiovascular and neurodegenerative diseases. Earlier detection of SCHI would enable earlier intervention, preventing the development of a serious health condition.

There are limitations with the use of fasting glycaemic and insulin markers. Hyperglycaemia may remain undetected when relying solely on haemoglobin A1c (HbA1c) or fasting glucose measurements, as compensatory hyperinsulinaemia suppresses fasting glucose and masks underlying pathology. In such cases, persistent insulin signalling drives glucose clearance into tissues at an accelerated rate, producing deceptively normal glucose values while concealing chronic metabolic dysfunction (37). This concealed state of hyperglycaemia is compounded by the inflammatory and mitogenic effects of sustained hyperinsulinaemia, which amplify systemic risk despite apparently reassuring glycaemic markers (34, 35, 38–41).

To overcome these diagnostic blind spots, alternative strategies are required. A more sensitive approach involves sequential monitoring of capillary glucose and ketone BHB over a minimum of seven consecutive evenings, either pre-dinner or at bedtime, and at least three hours post-prandially (16, 34, 35, 41, 42). This method directly assesses insulin’s suppressive effect on hepatic ketogenesis, providing an early marker of metabolic imbalance. In parallel, fasting venous blood biochemistry should incorporate markers that capture the endocrine, inflammatory, and vascular consequences of insulin excess, including plasminogen activator inhibitor-1, gamma-glutamyl transferase (GGT), leptin, homeostasis model assessment for insulin resistance (HOMA-IR), insulin-like growth factor-1 (IGF-1), vascular endothelial growth factor (VEGF), epidermal growth factor, and monocyte chemoattractant protein-1. Together, these indices provide a more comprehensive profile of the metabolic and vascular stress imposed by hyperinsulinaemia (16, 34, 35, 41).

Conventional standards of care (SOC) define normal glycaemia through fasting glucose and HbA1c values. However, reliance on these static markers fails to capture the metabolic consequences of persistent insulin excess (35, 37, 43). When laboratory results indicate glycaemia within the accepted reference interval yet clinical features such as atherosclerosis, or sleep apnoea are present, SCHI should be considered as a primary aetiological driver (44–54). These conditions share a unifying metabolic pattern in which euglycaemia is maintained at the cost of chronically elevated insulin, accompanied by suppressed ketogenesis (37, 53–55). Failure to recognise this compensatory glucose shunting leads to delayed diagnosis and intervention. The elevation of basal insulin, even within reference intervals, not only accelerates glucose uptake into cells but also produces false-negative results in standard clinical tests, thereby obscuring the detection of subclinical hyperglycaemia and early hyperinsulinaemia. This under-recognition allows metabolic damage to accumulate, such as chronic inflammatory signalling, impaired redox regulation, and endothelial dysfunction progress unchecked. Timely detection requires methods that assess both glucose and insulin dynamics, rather than static glucose endpoints to prevent misclassification and intervene before irreversible pathology develops.

This triad of euglycaemia, SCHI, and hypoketonaemia reflects a state of compensated metabolic dysregulation, in which chronic excess insulin exposure supresses the normal production of ketone bodies despite apparently stable glucose concentrations. Hypoketonaemia, therefore, provides a sensitive indicator of hidden metabolic imbalance, signifying insulin-mediated suppression of hepatic ketogenesis that may coexist with euglycaemia or overt hyperglycaemia. Recognition of this metabolic signature is critical, as the absence of abnormal fasting glucose or HbA1c does not equate to metabolic health when hyperinsulinaemia remains unaddressed.

Hypo-ketonaemia and insulin-compensated euglycaemia (ICE) define a hidden state of insulin and glycaemic pathology direction that is invisible to conventional standards. Insulin concentrations may fall within population reference intervals, but are elevated relative to individual metabolic capacity, thereby masking subclinical hyperglycaemia (fasting glucose ≤ 5.7 mmol/L due to insulin-compensation) and subclinical hyperinsulinaemia (≥ 8 μIU/mL). Hypoketonaemia-ICE is a metabolic state of hidden hyperglycaemia, in which glucose is cleared from the bloodstream, yet the total glycaemic load to the bloodstream is not captured. Where within reference range euglycaemia (fasting glucose ≤ 5.5 mmol/L) and HbA1c are maintained by compensatory insulin activity (hyperinsulinaemia) and revealed by chronic suppression of BHB (hypoketonaemia, < 0.5 mmol/L) most reliably detected through sequential daily evening capillary blood BHB measurement, where chronic hypoketonaemia demonstrates that basal insulin remains pathologically raised (17, 34, 35, 37, 38, 40, 43). A personalised hyperinsulinaemia threshold (PIT) is defined by chronic hypoketonaemia as a marker of insulin excess, which is individual-specific and not reliant on population reference ranges. Hypoketonaemia confirms SCHI even when glucose and serum insulin appear normal within standard reference ranges [37,38,43]. Hypoketonaemia-ICE captures the discordant state in which glycaemic indices appear unremarkable, yet underlying insulin-driven compensation reveals early metabolic disturbances.

A distinct metabolic state exists in which fasting glucose and HbA1c remain within conventional reference intervals, yet underlying hyperglycaemia is concealed by compensatory hyperinsulinaemia. In this condition, insulin levels may also appear ‘normal’ when interpreted against population-based ranges, but are pathologically elevated for the individual, driving excess glucose disposal and preventing hepatic ketogenesis. The most reliable indicator of this state is chronic evening hypoketonaemia, due to persistent insulin suppression of hepatic ketogenesis, defined as persistently low BHB (< 0.5 mmol/L) despite adequate fasting intervals (16, 34, 35, 40–42). This pattern reveals that basal insulin is set too high for the individual’s metabolic capacity, producing an artefactual appearance of normal glycaemia while masking early metabolic dysregulation or disease.

Insulin-Compensated Euglycaemia (ICE):

Moving from population-based thresholds to individual metabolic capacity as the diagnostic anchor, the personalised hyperinsulinaemia threshold (PIT) refers to the detection of insulin excess relative to an individual’s metabolic capacity, rather than against population-derived reference intervals (Table 1). PIT is identified through the hypoketonaemia-ICE framework, where chronic evening hypoketonaemia (minimum 3 h post-prandial BHB < 0.5 mmol/L) reveals that basal insulin levels are elevated, even when fasting insulin remains within conventional laboratory ranges (17, 35). This approach recognises that insulin-induced mitochondrial, metabolic and endocrine dysregulation can occur below standard thresholds and provides a more sensitive and specific means of identifying early metabolic disease (16, 38, 40, 43, 56). Through the lens of hypoketonaemia-ICE, subclinical and clinical hyperinsulinaemia may be detected to within personalised thresholds, where the key determinant is not absolute serum insulin within population intervals, but the functional suppression of ketogenesis (hypoketonaemia).

We have previously published the effects of suppressing ketosis in the Ketosis Suppression and Ageing (KetoSAge) cohort, demonstrating changes in biomarkers associated with ageing and chronic disease, including insulin, HOMA-IR, BHB, leptin, IGF-1, thyroid hormone levels, and GGT (16, 34, 35, 39, 41). Building on this foundation, the present study focuses on female-specific endocrine responses, by examining the impact of long-term sustained nutritional ketosis (NK), euketonaemia (BHB ≥ 0.5–5 mmol/L), and the suppression of ketosis (SuK) and hypoketonaemia (BHB < 0.5 mmol/L), on key reproductive and hormonal biomarkers. Specifically, we investigated changes in circulating concentrations of SHBG, oestrogen, progesterone, testosterone, luteinising hormone (LH) and follicle-stimulating hormone (FSH), providing novel insights into the metabolic regulation of female hormonal health. SHBG may further refine the detection of SCHI and aid clinicians in determining risk and prognosis in chronic diseases.

The female reproductive system is closely regulated by nutritional status and overall energy balance. Metabolic conditions, such as hyperinsulinaemia and insulin resistance can profoundly impact female hormone levels and reproductive function, particularly in conditions like PCOS (4, 59, 60, 68). Despite growing interest, the complex pathways by which metabolic health and nutritional signals influence reproductive function remain incompletely understood. Recent studies have investigated the potential role of SHBG, a hepatic glycoprotein, in metabolic dysfunction, due to its inverse relationship with hyperinsulinaemia, insulin resistance and hepatic fat accumulation (58–60). Lower circulating SHBG levels have been correlated with markers of metabolic dysregulation, suggesting its possible utility as a biomarker for hyperinsulinaemia, insulin resistance and metabolic dysfunction-associated steatosis liver disease (MASLD). The present study examined the dynamics of key female hormone biomarkers across three menstrual cycle phases in a cohort of healthy premenopausal women who followed a ketogenic lifestyle with sustained NK (mean duration: 3.9 ± 2.3 years). A more precise understanding of how metabolic state influences reproductive hormone regulation is crucial to advancing women’s health across the lifespan.

We observed a significant decrease in SHBG levels following suppression of ketosis (P2, hypoketonaemia for 21 days), with a subsequent marked return to baseline values upon re-establishment of NK (P3, euketonaemia). Linear mixed-effects modelling further revealed strong associations between SHBG and several metabolic and hormonal markers, including insulin, HOMA-IR, GKI, leptin, IGF-1, GLP-1, free T3 and GGT. These findings are consistent with the role of SHBG as a marker responsive to both insulin signalling and hepatic function (15, 23, 60–67), two systems known to be modulated by long-term ketosis, euketonaemia (16, 67, 68). Observational studies have consistently shown an inverse relationship between SHBG and insulin, with lower SHBG concentrations linked to higher insulin levels and an increased risk of T2DM, independent of circulating sex steroid levels in both men and women (66). This inverse relationship is particularly relevant in the context of insulin resistance and hepatic dysfunction, both of which are central features of MetS and are ameliorated through carbohydrate restriction. The return of SHBG to baseline levels upon reintroduction of ketosis further supports the metabolic responsiveness of SHBG and highlights its potential utility as a dynamic biomarker of metabolic-endocrine status in women.

The liver plays a central role in the regulation of systemic insulin sensitivity and sex steroid bioavailability, primarily through its modulation of SHBG synthesis in response to insulin signalling (65, 69). Notably, women with both low SHBG and elevated hepatic fat content have been shown to exhibit the highest insulin concentrations (70). This inverse relationship between SHBG and insulin is particularly pronounced among women with greater liver fat accumulation (62, 70), suggesting a compounded metabolic risk in the presence of both hepatic steatosis and decreased SHBG. These findings highlight the liver’s central role in integrating metabolic and reproductive functions, a relationship that becomes increasingly relevant in midlife women as oestrogen levels begin to decline. Consistent with this, lower endogenous SHBG levels have been robustly associated with increased risk for cardiometabolic disorders and non-alcoholic fatty liver disease across both sexes and age groups (24, 71, 72). In vivo animal-model studies support a protective role of SHBG in hepatic metabolism. Higher SHBG expression downregulates hepatic ATP production and inhibit lipogenic enzymes, such as acetyl-CoA-carboxylase and fatty acid synthase, thereby reducing hepatic lipid accumulation (63, 73). Further supporting this, SHBG overexpression in transgenic mice protects against high-fat (with carbohydrate) diet (HFD)-induced obesity and insulin resistance, improving glucose tolerance, lowering insulin levels, and attenuating increases in leptin and resistin levels (63).

Importantly, the significant differences in SHBG concentrations observed between NK and SuK phases suggest that SHBG may serve as a sensitive marker of metabolic-endocrine interactions within the female hormonal axis. These findings raise important questions about whether fluctuations in SHBG, for instance as those observed from P1 (NK) to P2 (SuK) and subsequently to P3 (return to NK), reflect a compensatory response (homeostatic adaptation) or an active regulatory mechanism in hormone availability in response to altered energy metabolism, particularly under ketogenic conditions or states of metabolic dysfunction. In line with emerging literature, our data support the role of SHBG as both a proxy marker of insulin sensitivity and exposure, hepatic function, as well as a potential diagnostic biomarker and therapeutic target in the management of PCOS, MASLD, and breast cancer (13, 14). Further research is warranted to elucidate the mechanistic role of SHBG in these contexts and to evaluate its clinical relevance across broader populations.

Dietary carbohydrate restriction, particularly when inducing sustained nutritional ketosis, exerts a profound influence on SHBG concentrations and broader metabolic-endocrine signalling. Across both PCOS and non-PCOS female cohorts, ketogenic interventions that achieve euketonaemia have been consistently associated with elevations in SHBG, accompanied by improvements in glycaemic regulation, insulin sensitivity, and androgen homeostasis (18, 29–32, 68). These outcomes are driven primarily by reductions in basal circulating insulin concentrations. As insulin is the principal suppressor of hepatic ketogenesis, any degree of hyperinsulinaemia, whether overt or subclinical, can suppresses endogenous ketone generation and simultaneously inhibits SHBG synthesis within the liver (16, 34–36).

SHBG is a tightly regulated hepatic glycoprotein that binds and transports sex hormones while also serves as a sensitive biomarker of metabolic homeodynamics. Its synthesis is acutely responsive to insulin levels, thyroid hormones, and hepatic fatty acid oxidation status (19, 66, 70, 73). The restoration of higher SHBG levels through nutritional ketosis reflects a shift toward improved metabolic homeodynamics, characterised by sufficiently reduced insulin levels to allow normalisation of hepatic processing and endocrine transport dynamics. Conversely, low SHBG levels are indicative of metabolic dysfunction and actively reflect a state of insulin-excess, often preceding disturbances in reproductive hormones, glucose metabolism, and vascular tone. Thus, SHBG functions as a peripheral surrogate of chronic insulin over-exposure, with reductions in SHBG representing one of the earliest physiological responses to sustained hyperinsulinaemia and hypoketonaemia.

In this context, SHBG offers a unique window into subclinical disease states. The detection of persistently low levels of SHBG in euglycaemic individuals may signal occult hyperinsulinaemia, particularly when ketone levels remain low despite carbohydrate restriction. Such a pattern reveals hepatic insulin resistance before the onset of dysglycaemia, positioning SHBG as a frontline biomarker for the early detection of stage-1 type-2 diabetes and hypoketonaemia-ICE (38). This concept aligns with the broader framework in which chronic insulin elevation precedes and drives pathophysiological cascades implicated in cardiovascular disease, oestrogen-dependent cancers, PCOS, and neurodegeneration (35, 74–77).

When oestrogen is coupled with SHBG and both activate their respective receptors on target cells, the resulting intracellular signalling cascade differs from that initiated by oestrogen alone. Many tissues express SHBG membrane receptors, and the binding of SHBG to its receptor induces cyclic adenosine monophosphate (cAMP) activation, functioning as a signal transduction modulator. Furthermore, SHBG molecules can enter cells and exert direct intracellular effects (13, 14). SHBG downregulates the proto-oncogenes c-Myc, B-cell lymphoma-2 (Bcl-2), and growth factor receptors associated with cancer cell growth, proliferation and survival (42, 78–80). In contrast, a low SHBG setting results in cancer cells exhibiting increased levels of c-Myc expression, driving transcriptional programmes that promote cellular proliferation, aerobic fermentation, and dedifferentiation. c-Myc upregulates the glucose transporter GLUT1, enabling insulin-independent cellular glucose uptake (13, 14, 16, 74, 81–84). This enhanced capacity for glucose uptake supplies cancer cell dependence on aerobic fermentation to facilitate their bioenergetic needs via cytosolic substrate-level phosphorylation, independent of insulin mediated glucose uptake signalling (42, 80, 81, 85).

Low SHBG levels commonly observed in hyperinsulinaemic and hypoketonaemic states, result in increased c-Myc expression and activity. In cancer cells, higher levels of c-Myc expression drives increased transcription and activity of key glycolytic enzymes, including lactate dehydrogenase (LDH) and hexokinase 2 (HK2) (14, 42, 78, 80, 81). LDH plays a central role in sustaining aerobic glycolysis in proliferative tumour phenotypes by catalysing the reduction of pyruvate to lactate (80, 86–90). This reaction regenerates cytosolic nicotinamide adenine dinucleotide (NAD⁺), a redox cofactor essential for the maintenance of glycolytic flux (38, 40, 43, 56). The regeneration of NAD⁺ permits uninterrupted substrate-level phosphorylation, enabling ATP synthesis independently of mitochondrial oxidative phosphorylation (91–93). This metabolic re-routing underpins the Warburg effect, whereby cancer cells ferment glucose despite sufficient oxygen availability (42, 80). By diverting pyruvate away from mitochondrial entry, LDH functionally suppresses oxidative metabolism, concomitantly enforcing aerobic fermentation of glucose and mitochondrial substrate level phosphorylation of glutamine. This not only sustains ATP production under conditions of mitochondrial dysfunction or high biosynthetic demand but also supports anabolic precursor generation and redox homeostasis necessary for rapid cellular proliferation. The accumulation of extracellular lactate further modifies the tumour microenvironment, promoting immune evasion, angiogenesis, and stromal reprogramming.

HK2 is one of four glucokinase isoforms and catalyses the first committed step of glycolysis by phosphorylating glucose to glucose-6-phosphate, thereby trapping it within the cell. HK2 is predominantly expressed in insulin-responsive tissues, such as skeletal muscle, myocardium, and adipose tissue, where it supports postprandial glucose clearance and energy storage. However, HK2 expression is markedly upregulated in a wide range of malignant tumours, reflecting its dual regulation by insulin signalling and oncogenic drivers (94–96). Activation of the SHBG receptor decreases c-Myc levels and activity, leading to a decrease in HK2 expression. Conversely, low SHBG levels result in increased c-Myc expression, leading to increased HK2 and LDH levels (14, 81). HK2 binds directly to the outer mitochondrial membrane, where it utilises ATP generated within the mitochondrial matrix to catalyse the phosphorylation of glucose to glucose-6-phosphate. This spatial proximity, typically mediated via the voltage-dependent anion channel (VDAC), enables HK2 to bypass cytosolic ATP competition and gain privileged access to mitochondrial ATP pools. The mitochondrial anchoring of HK2 permits a high-flux glycolytic state by coupling residual oxidative phosphorylation-derived ATP to the first committed step of glucose metabolism. This arrangement not only accelerates glycolytic throughput but also insulates cancer cells from ATP depletion under fluctuating nutrient conditions. By drawing on mitochondrial energy to support cytosolic glycolysis, HK2 acts as a gatekeeper of bioenergetic partitioning, reinforcing the Warburg phenotype and promoting proliferation under oncogenic and insulin-activated signalling contexts (94, 95, 97). Hyperinsulinaemia enforces glucose oxidation over beta-oxidation and ketolysis, decreasing the cellular pool of NAD⁺. Concurrently, the mitochondrial derived ATP consumed by HK2 imposes additional mitochondrial demand of NAD⁺, as a result, mitochondrial oxidative phosphorylation is compensatorily reduced to drive up aerobic glycolysis in order to refurnish the cytosol with NAD⁺ (38, 43, 93).

Hyperinsulinaemia decreases SHBG levels, which increases c-Myc expression in cancer cells. Increased c-Myc activity, in turn, upregulates transcription and activity of LDH and HK2 (14, 81). Mitochondrial binding of HK2 inhibits Bax-induced cytochrome c release and subsequently inhibits mitochondrial death-receptor-mediated apoptosis (98). This strategic positioning enables cancer cells to integrate anabolic demands with survival pathways, reinforcing the glycolytic phenotype characteristic of aggressive, proliferative states (96).

Lower levels of SHBG receptor signalling, whilst cells receive oestrogen receptor activation, also increases Bcl-2 levels. Bcl-2 preserves cellular viability by suppressing mitochondrial apoptosis pathways at the level of the outer mitochondrial membrane, preventing the membrane permeabilisation (MOMP) required for cytochrome c release into the cytosol, thereby maintaining the integrity of the mitochondrial intermembrane space and inhibiting caspase activation (99). Bcl-2 achieves this through direct interaction with the pro-apoptotic effectors Bax and Bak, sequestering them in inactive conformations and preventing pore formation within the membrane. Through this mechanism, Bcl-2 sustains mitochondrial membrane potential (Δψm), redox balance, and ATP production, enabling the continued survival of metabolically stressed or oncogenically transformed cells.

Aggressively proliferating cancer cells overexpress Bcl-2, leading to mitochondrial resistance to apoptosis which underpins tumour persistence and therapeutic resistance. Chronic hyperinsulinaemia decreases SHBG, leading to increased Bcl-2 and c-Myc levels. c-Myc expression in cancer cells increases LDH, GLUT1, HK2, and alters transcriptional programming in transformed cells to be more similar to embryonic cells, which is a dedifferentiated, hyperplastic and hyper-prolific state (82). This convergence of metabolic and developmental reprogramming reflects mitochondrial pathophysiological transformation and the dominance of insulin–glucose–growth factor signalling loops.

Given insulin’s suppressive control over hepatic SHBG synthesis (15, 100), its measurement offers a mechanistically grounded and clinically accessible metric of metabolic endocrine status. Unlike glucose or HbA1c, which reflect late-stage dysregulation, SHBG, along with euketonaemia monitoring, provides an indirect yet precise index of the insulin axis activity and its broader metabolic-endocrine consequences. This is particularly relevant for women, in whom hypo-SHBG states are predictive of hyperandrogenism, ovulatory dysfunction, and infertility, as well as increased long-term risk for breast and endometrial cancer (1, 13, 14, 61). Importantly, SHBG levels increase rapidly in response to dietary carbohydrate restriction, reflecting improved hepatic insulin sensitivity and restoration of fatty acid oxidation (16, 34, 35, 83, 84).

Early restoration of SHBG levels via ketogenic metabolic therapy (KMT) at therapeutic levels, therefore, serves dual functions: as a marker of mechanistic reversal and as a mediator of reduced downstream hormonal and metabolic dysregulation. Its modulation is causally linked to the insulin–ketone–liver axis and should be integrated into personalised metabolic-endocrine assessments. A shift upwards in SHBG trajectory in the context of euketonaemia and reduced fasting insulin constitutes evidence of improving metabolic-endocrine health, offering an inexpensive, widely accessible surrogate for direct insulin quantification, which remains underutilised and poorly standardised in routine clinical practice (16).

SHBG plays a central regulatory role in endocrine–oncogenic signalling, with robust mechanistic evidence showing its direct modulation of bioactive sex hormone availability and intracellular receptor activation (61). Its suppression by insulin via downregulation of HNF4α links hepatic metabolic status to systemic hormonal control, establishing SHBG as a key mediator at the intersection of metabolic dysfunction, hyperinsulinaemia, and cancer risk (61). Reduced SHBG levels increase free testosterone and oestradiol, promoting tumour proliferation in hormone-responsive cancers, while elevated SHBG inhibits growth and metastasis even in receptor-negative phenotypes (13, 14, 101). These integrated findings fulfil the criteria for a central regulatory molecule and support the conclusion that SHBG functions as a protective metabolic–endocrine marker with system-wide relevance.

Thus, SHBG emerges as a pivotal indicator in the modern metabolic framework. In women, its diagnostic and prognostic relevance spans reproductive, cardiovascular, oncological, and neurological domains (19, 23, 59–61, 63, 71). In both metabolic research and clinical intervention alike, its early normalisation through maintaining euketonaemia reflects therapeutic efficacy and the re-establishment of hepatic metabolic control, preceding irreversible disease progression.

A significant elevation in circulating leptin levels was observed following the 21-day suppression of ketosis during Phase 2 (P2), characterised by sustained hypoketonaemia-ICE. Upon re-establishment of nutritional ketosis in Phase 3 (P3), leptin concentrations returned markedly toward baseline in parallel with restored euketonaemia. Linear mixed-effects modelling confirmed a strong inverse association between leptin and SHBG across phases, highlighting the shared sensitivity of both markers to shifts in insulin signalling. These findings are consistent with the role of leptin as a dynamic metabolic-endocrine biomarker of insulin tone and adipose–liver axis regulation.

Leptin, the adipocyte-derived cytokine encoded by the Ob gene, functions far beyond its canonical role in energy homeostasis, exerting pro-tumorigenic effects across multiple malignancies through mitogenic, anti-apoptotic, metabolic, and angiogenic mechanisms. Leptin promotes endothelial tube formation and increases vascular permeability by synergising with VEGF and fibroblast growth factor 2 (FGF-2), thereby facilitating tumour vascularisation and the establishment of a permissive vascular–stromal interface for cellular invasion (102–104). Within the glioblastoma microenvironment, leptin secreted by tumour cells directly stimulates endothelial growth, mirroring the effects of VEGF and sustaining tumour expansion under hypoxic and nutrient-depleted conditions (104, 105).

Leptin receptor (ObR) signalling also plays a central role in the metabolic reprogramming of cancer cells. In breast, colon, pancreatic, and endometrial cancers, leptin promotes proliferation by activating PI3K–Akt, JAK2–STAT3, MEK–ERK, and JNK signalling cascades (106–110). These cascades converge on cell cycle regulators such as cyclin D1 and Mcl-1, inhibit mitochondrial apoptotic checkpoints that are regulated by Bcl-2 and pro-apoptotic effectors (e.g., Bax, Bak), and support the preferential reliance on aerobic glycolysis over oxidative phosphorylation, driven by mitochondrial adaptation, dysfunction, and redox dysregulation. Chronic excess leptin, insulin, and IGF-1 upregulate anti-apoptotic proteins Bcl-2 and Mcl-1, whilst also inhibiting Bax/Bak oligomerisation. This stabilises the Δψm, preventing the mitochondrial outer MOMP, which inhibits the mitochondrial checkpoint that governs apoptosis initiation (109, 111). In hypothalamic astrocytes, leptin regulates nutrient transporter expression, including GLUT1 and glutamate transporters, providing further evidence of its influence on nutrient uptake and metabolic dysregulation (112).

In hormone-sensitive tissues, leptin interacts with insulin and IGF-1 signalling, forming an endocrine-oncogenic triad that supports tumour progression. Hyperglycaemia and insulin excess potentiate leptin-mediated signalling, particularly via IGF1R–Akt–mTOR activation, further accelerating cell cycle progression and biomass accumulation in breast and mammary epithelial cells (106, 113). In prostate and endometrial cancers, leptin enhances differentiation, proliferation, and invasiveness through transcriptional upregulation of proto-oncogenes and metabolic enzymes (111, 114). Long-term exposure to leptin increases tumour cell viability and shifts the mitochondrial phenotype toward apoptosis evasion and aerobic glycolysis, enabling anabolic growth in nutrient-variable microenvironments (115–119).

Leptin’s role in metastasis is also increasingly recognised. In pancreatic and ovarian cancers, leptin–ObR signalling promotes extracellular matrix remodelling, migration, and invasion through upregulation of matrix metalloproteinase-13, an enzyme that breaks down the extracellular matrix, particularly collagen, and downstream effectors (110, 120). In ovarian cancer cells, leptin maintains stem-like characteristics and drives a more aggressive transcriptional phenotype, explaining the poor prognosis observed in obese patients with leptin resistance or chronic hyperleptinaemia (121, 122). Clinical data further associate high leptin levels with increased prostate cancer risk and unfavourable outcomes in ovarian cancer (122–124).

Leptin and its receptor ObR are overexpressed in malignant brain tumours and display a strong positive correlation with histopathological grade, with the highest levels consistently observed in glioblastoma (GBM), a WHO grade IV astrocytic neoplasm (105). This overexpression actively contributes to tumour pathophysiology. Leptin engages proliferative, anti-apoptotic, and migratory signalling via PI3K–Akt and JAK2–STAT3 cascades, thereby supporting cellular survival within the tumour microenvironment. These signalling axes converge with broader patterns of metabolic reprogramming, redox mal-adaptation, and downregulation of mitochondrial oxidative metabolism, enabling sustained proliferation even under hypoxic or nutrient-depleted conditions (104, 105, 107, 111, 113, 115, 116, 125–129).

Functioning additionally as a pro-angiogenic cytokine, leptin promotes the expression of VEGF and stimulates endothelial tube formation, driving neovascularisation in support of tumour expansion (34, 35, 103, 104, 130). This angiogenic phenotype is especially prominent in GBM, where vascular proliferation ensures perfusion and also establishes a permissive vascular–stromal interface for cellular migration and infiltration. Notably, GBM cells possess the capacity for autocrine leptin production, allowing locally secreted leptin to activate ObR on neighbouring or self-same tumour cells. This autocrine loop amplifies survival signalling, suppresses mitochondrial apoptotic checkpoints, and facilitates the autonomous maintenance of ATP production, redox cycling, and anabolic flux, independent of systemic endocrine input (104, 105, 131).

Within the broader metabolic–endocrine context, hyperinsulinaemia and suppression of SHBG frequently co-occur with leptin resistance, wherein elevated leptin reflects compensatory hypersecretion rather than receptor activity (132–135). In glioblastoma, the leptin–ObR axis actively contributes to tumour metabolism, regulating glutamate and glucose transporters, reinforcing reliance on aerobic glycolysis and glutaminolysis over oxidative phosphorylation, driven by mitochondrial adaptation and dysfunction (105, 112, 131). These endocrine–oncogenic dynamics support the therapeutic rationale for KMT for cancer. By raising SHBG, improving leptin sensitivity, and decreasing excess insulin-mediated signalling, nutritional and therapeutic ketosis may attenuate this endocrine, paracrine and autocrine growth axis and re-sensitise tumour cells to apoptotic initiation (16, 34, 35). Targeting the metabolic dependencies that sustain tumour bioenergetics offers a viable adjunct to existing therapeutic modalities. Based on KMT’s multiple mechanisms of action, this provides a rational explanation of how KMT would improve patient responses to standard of care (SOC) oncology treatments, including chemotherapy, immunotherapy, checkpoint inhibitors, receptor inhibitors, immune cell-based vaccines and radiotherapy.

Leptin, thus, acts as a pleiotropic mitokine, coordinating energy-sensing, redox adaptation, apoptosis resistance, and vascular remodelling within the tumour microenvironment. Its downstream effects reinforce the oncogenic landscape established by chronic hyperinsulinaemia, hypoketonaemia and low SHBG. This amplifies proliferative and invasive capacity while inhibiting apoptosis. Therapeutic strategies targeting leptin–ObR signalling may restore mitochondrial sensitivity to redox stress and re-establish endocrine regulation, particularly when combined with ketogenic endocrine metabolic oncology (KEMO) therapies that decrease insulin signalling, increase SHBG and euketonaemia (16, 34, 35, 41, 42, 136).

Beyond SHBG and leptin, we also investigated the effects of ketosis suppression on key reproductive hormones, including oestrogen, progesterone, testosterone, LH, and FSH. While no significant changes were detected in circulating levels of oestrogen, progesterone, testosterone, LH, or FSH across the study phases and in linear effects models, FSH concentrations varied significantly by menstrual cycle phase, with elevated levels during ovulation compared to the follicular phase. This aligns with the physiological role of FSH in follicular development and emphasises the importance of accounting for cycle phase in hormonal-endocrine research. These findings are particularly relevant, given the increasing interest in how dietary strategies, such as low-carbohydrate or ketogenic diets, influence female reproductive physiology. While carbohydrate restriction has shown therapeutic benefits in PCOS, characterised by hyperandrogenism and ovulatory dysfunction, its effects in healthy, reproductive-age women remain underexplored. The present data contribute to this emerging field by demonstrating the apparent hormonal resilience of the reproductive axis in response to short-term changes in metabolic state.

This study represents the first controlled investigation into the effects of long-term sustained nutritional ketosis, and its deliberate suppression, on SHBG and related metabolic–reproductive signalling in healthy premenopausal women. These findings establish SHBG as a dynamic biomarker and regulator within the insulin–liver axis, integrating metabolic and reproductive signalling and demonstrating responsiveness to changes in dietary carbohydrate load. Restoration of euketonaemia, achieved through carbohydrate restriction, whilst consuming adequate natural healthy fats, consistently increased SHBG levels, indicating improved hepatic insulin sensitivity and reversal of insulin-driven endocrine suppression. Importantly, this effect occurred without destabilising reproductive hormone profiles, supporting the safety of carbohydrate restriction in premenopausal women. The elevation of SHBG through therapeutic ketosis holds particular relevance for oncology. In multiple tumour contexts, low SHBG states are associated with increased c-Myc and Bcl-2 expression and activity, upregulation of glycolytic effectors GLUT1, LDH and HK2, promoting mitochondrial resistance to apoptosis.

Chronic hyperinsulinaemia suppresses hepatic SHBG synthesis, thereby disinhibiting these oncogenic pathways, driving increased aerobic glycolysis and substrate level phosphorylation dependence, proliferative signalling, and apoptosis resistance. Elevated leptin, under insulin resistance and hyperinsulinaemia, further reinforces this oncogenic landscape by activating mitogenic, anti-apoptotic, and pro-angiogenic pathways (JAK2-STAT3 and PI3K-Akt) and maintaining cancer stem-like phenotypes in many tumour types. Elevation of SHBG through euketonaemia, acts as a counter-regulatory signal that downregulates substrate level phosphorylation dependence and restores apoptotic capacity. These findings support the therapeutic rationale for KMT in cancer, where correction of hyperinsulinaemia and induction of euketonaemia restricts tumour fuel supply, thus reprogramming endocrine-oncogenic signalling through SHBG modulation. Further translational studies are warranted to determine whether SHBG normalisation may serve as both a biomarker and a mediator of therapeutic response in KMT-treated cancer patients. This paper identifies SHBG as a central regulator of endocrine–oncogenic signalling, with elevated SHBG functioning as a protective metabolic-endocrine marker. Insulin thus drives the cancer cell phenotype via transcriptional reprogramming and downregulation of mitochondrial apoptosis capability. Leptin, likewise, regulated by insulin and nutritional status, reinforces this oncogenic landscape. Leptin levels rise under insulin resistance and hyperinsulinaemia, amplifying proliferative signalling and further decoupling tumour growth from endocrine restraint.

Finally, this work also advances a refined diagnostic framework for hyperinsulinaemia. We propose that persistent suppression of BHB below 0.5 mmol/L before the evening meal, despite normoglycaemia and reference-range insulin, indicates hypoketonaemia-Insulin-Compensated Euglycaemia (ICE), a subclinical state of pathological insulin excess. This functional phenotype, linked to the Personal Hyperinsulinaemia Threshold (PIT), is routinely missed by conventional glucose-centric assessments. By integrating SHBG with fasting insulin, leptin, and ketone-based metabolic phenotyping, both research and clinical frameworks can achieve earlier detection of SCHI. SHBG and leptin, when interpreted in context, provide mechanistically grounded biomarkers that support timely metabolic intervention before the onset of pathology such as cancer progression.