Relative Energy Deficiency in Sport (REDs): Endocrine Manifestations, Pathophysiology and Treatments

Abstract

Research on lean, energy-deficient athletic and military cohorts has broadened the concept of the Female Athlete Triad into the Relative Energy Deficiency in Sport (REDs) syndrome. REDs represents a spectrum of abnormalities induced by low energy availability (LEA), which serves as the underlying cause of all symptoms described within the REDs concept, affecting exercising populations of either biological sex. Both short- and long-term LEA, in conjunction with other moderating factors, may produce a multitude of maladaptive changes that impair various physiological systems and adversely affect health, well-being, and sport performance. Consequently, the comprehensive definition of REDs encompasses a broad spectrum of physiological sequelae and adverse clinical outcomes related to LEA, such as neuroendocrine, bone, immune, and hematological effects, ultimately resulting in compromised health and performance. In this review, we discuss the pathophysiology of REDs and associated disorders. We briefly examine current treatment recommendations for REDs, primarily focusing on nonpharmacological, behavioral, and lifestyle modifications that target its underlying cause-energy deficit. We also discuss treatment approaches aimed at managing symptoms, such as menstrual dysfunction and bone stress injuries, and explore potential novel treatments that target the underlying physiology, emphasizing the roles of leptin and the activin-follistatin-inhibin axis, the roles of which remain to be fully elucidated, in the pathophysiology and management of REDs.

In the near future, novel therapies leveraging our emerging understanding of molecules and physiological axes underlying energy availability or lack thereof may restore LEA-related abnormalities, thus preventing and/or treating REDs-related health complications, such as stress fractures, and improving performance.

Graphical Abstract

relative energy deficiency in sport, female athlete triad, low energy availability, pathophysiology, endocrinological disorders, therapeutic approaches

Essential Points

• Relative Energy Deficiency in Sport (REDs) refers to a frequently overlooked constellation of disorders stemming from low energy availability (LEA) in diverse exercising populations of either biological sex.

• REDs encompasses neuroendocrine abnormalities in central hypothalamic-pituitary axes, mainly the reproductive, thyroid, adrenal, and somatotropic axes as well as bone, hematological, immunological, gastrointestinal, cardiovascular, neuromuscular, and psychological abnormalities.

• Basic, translational, and clinical research during energy deprivation indicates the adipokine leptin as a promising regulator of neuroendocrine function downregulated during LEA, which might be associated with several REDs-related disorders.

• Currently, the front-line treatment approach targeting the underlying causes of LEA entails the incorporation of nutritional, behavioral, and lifestyle alterations.

• Leptin replacement therapy is currently not recommended by consensus guidelines for individuals with hypothalamic amenorrhea or during low energy availability because its safety and effectiveness, including potential benefits like improved neuroendocrine function, menstrual and reproductive health, growth factors, and bone health, require further investigation for confirmation.

• The ACTH-cortisol and activin-follistatin-inhibin axes have also been shown to be implicated as potential regulators of a variety of biological processes, including development, reproduction, body composition, and energy metabolism, independently of leptin; however, the full spectrum of diagnostic and therapeutic relevance of the activin-follistatin-inhibin signaling pathways in REDs is under investigation.

• Because more studies describing the diagnosis, clinical relevance, and potential treatments of REDs are being published, compounds in development, targeting the physiological pathways involved, could help improve therapeutic outcomes in energy-deficient populations suffering from REDs.

Over the past 3 decades, the fundamental role of energy availability (EA) in athletes’ health and overall well-being has attracted increasing attention from the medical community, international sports committees and federations, as well as sports professionals and scientists in general (1).

In sports science and nutrition, EA is defined as the difference between total daily energy intake (EI) and exercise energy expenditure (EEE) divided by the fat-free mass (FFM) (ie, EA = (EI – EEE)/FFM, expressed in kcal/kg of FFM per day) (2-6). EA represents the total energy at the body's disposal to support fundamental physiological functions (such as thermoregulation, growth, maintenance of cellular homeostasis, immune system functions, and reproduction) that underpin optimal performance, adaptation to training, and ultimately health (3, 4, 7). It should be emphasized that EA indicates the total available energy after deducting the energy expended during exercise, thereby highlighting the possibility that excessive energy expenditure during exercise can divert energy away from other critical physiological functions (8). Moreover, energy balance (EB) is the difference between EI and total energy expenditure (TEE) (ie, EB = EI – TEE). In a broader sense, EB is the amount of energy remaining from EI after subtracting total energy used by the body, including the energy consumed for the function of all physiological systems (and is thus either positive or negative) (4). Therefore, it is essential to acknowledge that EA and EB are not interchangeable but reflect 2 different parameters and should be clearly differentiated to identify athletes at risk of or athletes in a pathophysiologically abnormal state.

Both research evidence and clinical experience show that athletes may adapt to situations of energy deficiency, particularly during prolonged periods of severely low energy availability (LEA). This adaptation occurs by suppressing energetically costly physiological processes (such as neuroendocrine function, and growth), reducing the athlete's overall energy requirements, and altering physiological functions such as reproductive and bone health or insulin resistance status and lipolysis. These changes aim to promote survival in environments of limited energy resources (4, 7, 9).

Considerable research, commencing several decades ago, has been performed to elucidate the underlying causes of menstrual dysfunction and low bone mineral density (BMD) leading to poor bone health (eg, osteoporosis and stress fractures [SF]) observed mainly in high-performance female athletes with or without a disordered eating (DE) behavior (9). Low energy availability has been acknowledged as the cornerstone for these unfavorable health outcomes recognized as the Female Athlete Triad (10-13). Since its inception in 1992 (10) and reevaluation by the American College of Sports Medicine in 1997 (12), the term “Female Athlete Triad” involves any 1 of the following 3 interrelated components: (1) LEA with or without DE, (2) menstrual dysfunction, and (3) low BMD (9, 12, 14).

Furthermore, over the past several decades, animal experiments and clinical research (mainly observational studies and survey data) have revealed that some male athletes and soldiers may also experience similar LEA-induced physiological abnormalities and health disruptions indicative of those identified in the Female Athlete Triad (15-21). Thus, a similar model was defined as the “Male Athlete Triad,” and it consists of 3 interrelated conditions, namely the LEA (as the key etiologic factor) and perturbation of bone and reproductive health (alterations of the hypothalamic-pituitary-gonadal [HPG] axis) (22, 23). However, he available evidence in males is still more limited, chiefly focused in military populations and based mainly on epidemiological information, but contradictory data still exist (17, 21, 24, 25). Therefore, further research efforts focusing on the underlying physiological mechanisms and long-term effects of the Male Athlete Triad on health and performance are required.

In 2014, the International Olympic Committee (IOC) expert working group introduced a broader term, coined Relative Energy Deficiency in Sport (REDs) syndrome (26), referring but not limited to the LEA-related abnormalities (including also endocrine, hematological, immunological, cardiovascular, cardiovascular, and psychological health disorders) affecting athletic performance, physiological function, and health status in physically active men and women (21, 26-31). As awareness of REDs increases, regulatory organizations in the field of sports have begun outlining the clinical evaluation, diagnostic workup, and basic treatment of the syndrome (26, 30). There has also been extensive research on DE as well as debate on the definition of REDs as a distinct clinical entity, juxtaposed against overlapping features of the Overtraining Syndrome (32) and, importantly, the Female Athlete Triad (33). Thus, a “pathway for progress” has been suggested to facilitate cooperation between the Female Athlete Triad and REDs investigators, calling for quality studies with clinically relevant outcomes and aiming to establish causality and address previous research gaps (34). In recent years, significant scientific progress has advanced our understanding of the underlying physiology and psychology related to REDs, culminating in the publication of the IOC consensus statement on REDs in 2023 (6). Therefore, according to the revised 2023 definition, REDs is identified as a syndrome resulting from prolonged and/or severe LEA (problematic), which affects athletes of both sexes, leading to impaired physiological and/or psychological functioning, including detrimental health issues, compromised well-being, performance challenges, and elevated injury risks (6).

The management of REDs should be multidisciplinary and carried out by a team of sports and exercise professionals. Current practice guidelines and consensus statements (14, 19, 35) highlight nutritional, behavioral, and lifestyle changes as the recommended primary treatment strategies for both the Female and Male Athlete Triad, as well as for REDs. To date, there is a lack of effective pharmacological remedies to counter the negative consequences of REDs. The adipokine leptin, the circulating levels of which reflect primarily the size of energy stores in adipose tissue and secondarily acute energy deprivation, may be considered a potential contributory factor to the pathophysiology of the syndrome because of its involvement in regulating neuroendocrine function and energy homeostasis. Leptin orchestrates an intricate response system to conserve energy by suppressing nonessential functions vs deficits, aiming to ensure the survival of the organism through limiting procreation and preserving energy during LEA (36-41) by influencing several neurohormonal axes affected in REDs. Administration of recombinant methionyl human leptin (r-metHuLeptin) to lean individuals with hypoleptinemia in replacement doses to normalize the falling levels in response to energy deprivation, confers neuroendocrine benefits (42-44), including reversing the clinical manifestations of hypothalamic amenorrhea (HA) (45-47). Furthermore, leptin may help protect against LEA-induced low BMD and SF in the long term (48-52). However, leptin replacement therapy is not currently recommended pending further research on its effectiveness and potential risks in addressing the syndrome's pathophysiology. The activin-follistatin-inhibin (AFI) axis is an additional hormonal system known for regulating reproductive function and has secondary roles in maintaining muscle mass, altering bone mass, and influencing insulin sensitivity and thus lipolysis and glycemia (ie, provision of energy) (43, 53, 54). The role of the AFI axis in REDs is starting to emerge and may also pinpoint to novel therapeutic strategies, as is the stress hormone axis of ACTH-cortisol, which is affected by both the leptin and the AFI axis. Consequently, our recent and evolving understanding of the underlying and mediating pathophysiological mechanisms is anticipated to change the way we approach, diagnose, and treat this significant condition.

Military training may negatively affect well-being and sports performance, especially among personnel who experience energy deficits during training and field exercises, which can be linked to the Triad and REDs because of the high energy expenditures and reduced energy intake (21).

Therefore, this review summarizes available research on the definition and recognition of LEA/REDs in athletic and military cohorts. It further explores the characteristics of REDs and discusses the pathophysiology of endocrine conduits behind REDs, including leptin and the AFI axis. Additionally, current and potential treatments for REDs, the Female Athlete Triad, and related disorders are discussed.

LEA and the Manifestations of REDs

Definition of LEA

As mentioned, EA is calculated by subtracting the EEE from EI and dividing it by the FFM (6). Severe and/or extended LEA has been associated with several detrimental effects, with impaired reproductive function and bone health being the most widely described. However, a consensus threshold for the minimum EA required to sustain all the physiological functions necessary for optimal health and sports performance has not been achieved. Therefore, a precise threshold of LEA remains undetermined, with ongoing debates in the scientific community. This uncertainty in defining a clear-cut threshold could also be attributed to the length of LEA exposure, type of exercise, body size and composition, differences between males and females and interpersonal variability, dietary macronutrient composition, the lack of previous models to consider jointly acute energy changes and chronic energy availability and the fact that the type of methods for EA assessment vary across studies and are not universally standardized (eg, self-reported questionnaires) (2, 4, 5, 55).

In more detail, based on findings from early randomized clinical trials investigating gonadotropin pulsatility in regularly menstruating, habitually sedentary, young females of normal body composition, LH pulse frequency and amplitude were considered not disrupted when EA levels were set at 45 kcal/kg FFM/day (56). Moreover, it was found that LH pulsatility did not follow a linearly proportional to EA, whereas females with shorter luteal phases were more susceptible to disruption of LH pulsatility by restricted EA (56). Therefore, an EA of at least 45 kcal/kg FFM/day is typically recommended as the optimal EA to ensure sufficient energy for healthy physiological function (4, 30). Likewise, EA <30 kcal/kg of FFM/day, even for a short period, has been associated with substantial perturbation of various body systems (mainly including alterations in circulating thyroid hormones, disruptions in LH pulse frequency, and increases in bone resorption) (56-58). Furthermore, our comprehension of the LEA threshold or the range within which males manifest REDs-related symptoms remains relatively limited. Nonetheless, current evidence suggests that this threshold is likely lower in males compared with females (59). Furthermore, it is crucial to recognize that the manifestations of REDs in males and the thresholds for LEA are less understood and appear to differ from those in females, indicating a need for gender-specific research in this area.

Therefore, an EA threshold lower than 30 kcal/kg FFM per day has been considered a potential LEA threshold, mainly based on short-term controlled clinical trials in females. However, because of limited epidemiological information on free-living individuals, paucity of long-term studies, and divergent findings among research studies, fail to establish a standardized and robust threshold for LEA, which continues to be a topic of active discussion (2, 5, 17, 24, 25, 30, 60, 61).

Our previous research highlights the importance of not only considering the acute daily energy intake but also longer term caloric reserves, specifically the amount of energy stored in adipose tissue. We have found that the hormone leptin acts as a “thermostat” by reflecting both acute changes in energy intake (ie, leptin levels drop by approximately 50% of baseline in response to acute and absolute caloric deprivation and by only ∼10% of baseline after 3 days of acute and absolute caloric deprivation even before any changes in total fat mass and body weight occur) (62). The decreased leptin levels inform the brain of these changes and activate molecular pathways that lead to the physiological changes depicted in the graphical abstract. Nevertheless, it is important to note that reductions in circulating leptin in relation to LEA have been reported in earlier studies (63).

No consensus has been developed to date concerning a targeted LEA threshold between optimal and impaired physiological mechanisms, highlighting the complexity of accurately identifying individuals at risk, and this needs to be a priority for research in the future. Therefore, the currently used EA cutoff points (ie, 30 kcal/kg FFM/day) serve as a preliminary guide and should not be used diagnostically because of the complex interplay of factors affecting LEA and REDs. However, EA thresholds might be helpful to direct systematic research on the mechanistic underpinnings of LEA and provide a gross tool to identify athletes at risk of LEA and subsequent performance decline and REDs development.

In conclusion, because of individual variability, including gender differences as previously described, and the outlined limitations, a precise threshold of LEA remains undetermined. Thus, it is imperative to acknowledge the ongoing debates and contradictions, which highlight the complexities and inherent challenges in the evidence-based classification and precise definition of both LEA and REDs. This underscores the need for further research to refine our understanding and methods, which is essential for advancing both scientific knowledge and clinical applications related to this condition.

LEA: Causative Factor of REDs

It is crucial to acknowledge that LEA is the primary etiological key factor of the REDs, which expands the concept of the Female and Male Athlete Triad to encompass a wide range of potentially LEA-related severe health consequences beyond those included in the aforementioned Triad models (26, 56, 57, 64-66) and negatively affect the sport performance and well-being of individuals at risk (67, 68). The underlying conditions underpinning the development of LEA are multifactorial and complex, including, but not limited to, an intentional or unintentional reduction in caloric intake (eg, cultural eating habits, restriction, binging, fasting, anorexia nervosa, other eating disorder/DE disorders) (69), an increase in daily exercise energy expenditure (eg, prolonged acute exercise, intense and extended duration training demands), and a poor awareness or misconceptions about the syndrome resulting in an inappropriate matching of intake energy to the exercise energy demands (9, 26, 30, 70-72). Interestingly, psychological impairments can either contribute to LEA and precede REDs (as an exposure variable) or result from REDs (as an outcome) (26, 67).

However, it should be noted that LEA may not always disrupt physiological processes and result in the clinical consequences described in the REDs, a fact that is more apparent, especially in males (30, 55, 72). Consequently, LEA may be considered adaptable, particularly in the short term and in the context of transient and reversible conditions, in which LEA might be associated with a necessary and valuable activity in sport (eg, body composition management and intensified training) (6). On the contrary, LEA could be considered problematic when it persists over an extended duration or exhibits specific characteristics indicative of a maladaptive response (6). LEA exposure, in conjunction with other moderating factors, may lead to impaired physiological processes. Consequently, normal physiology can transition into pathophysiology by disrupting physiological systems, ultimately resulting in poor performance and health issues from inadequate energy to support optimal functions (8). The characteristics of problematic LEA and the potential moderating factors may differ among individuals and body systems (eg, sex, genetic characteristics, preexisting medical conditions, and most importantly, energy reserves such as overall fat mass at baseline, type of sport, exercise training duration and characteristics). Moreover, emerging evidence suggests that besides the amount of caloric consumption, the dietary patterns followed (eg, low carbohydrate consumption) may exert a negative impact on athletes’ health, but this remains to be studied further and in conjunction with underlying neuroendocrine pathways activated or inactivated (73, 74).

However, additional research is warranted in this area. More rigorous study designs are essential to improve the overall quality of evidence regarding the LEA definition and the Female Athlete Triad and REDs-related outcomes (75). Therefore, more prospective long-term studies considering the impact of potential moderating and confounding factors are needed to elucidate further the causality and timing of the effects of energy deficiency on the Female Athlete Triad and REDs development. It is also crucial to consider all of the aforementioned factors jointly rather than independently of each other.

Key Abnormalities of REDs

The physiological consequences of REDs stem from problematic LEA (graphical abstract) and encompass most neuroendocrine systems in the human body (Table 1). However, several short- and long-term outcomes and related signs and symptoms are not REDs-specific, and these consequences can be observed in the absence of LEA under the prism of different health disorders. Therefore, the differential diagnosis should always be considered and carefully conducted based on a comprehensive assessment and thorough evaluation of the patient's clinical presentation, imaging findings, and laboratory examinations, as appropriate.

Table 1.

Most commonly reported clinical and biochemical manifestations reported in milieus of REDs

Physiological axis		Manifestation	Most common study type
Endocrine	Hypothalamic-pituitary-gonadal	Functional HA (12, 36, 61, 65, 68, 76-79)	Systematic review Society guidelines Consensus statements Prospective cohort Cross-sectional
		Diminished gonadotropin (LH, FSH) concentrations and pulses (56, 80-84)	Open-label clinical trial Prospective cohort
		Decreased testosterone concentrations (81-83, 85-91)	Prospective cohort Cross-sectional
		Decreased motile sperm count (86, 92) Abnormal sperm morphology (93)	Cross-sectional
	Hypothalamic-pituitary-thyroid	Reduced thyroid hormone concentrations (42, 57, 85, 94-97)	Case-control Prospective cohort Cross-sectional
		Blunted TSH pulse and response to TRH (42, 96)	Cross-sectional Prospective (short-term)
	Hypothalamic-pituitary-somatotropic	Accentuated GH pulse peaks (short-term) (98, 99)	Prospective cohort
		Decreased IGF-1 levels (56, 85, 99-103)	Prospective cohort
		Increased IGFBP-1 levels (104)	Clinical trial
		Decreased IGFBP-4, IGFBP-6, GHBP levels (104)	Clinical trial
	Hypothalamic-pituitary-adrenal	Elevated cortisol concentrations (105-109)	Review Prospective cohort
		Blunted diurnal salivary cortisol pulse (42, 110, 111)	Prospective cohort
	Leptin, insulin, and ghrelin	Reduced insulin levels (112)	Prospective (short-term)
		Elevated ghrelin levels (113)	Cross-sectional
		Reduced leptin levels (42, 45, 46, 104, 112, 114-116)	Clinical trial
Bone	Loss of BMD and premenopausal osteoporosis (17, 24, 25, 58, 70, 117, 118)		Randomized controlled trials Reviews, Society guidelines, consensus statements
	↑ NTX, CTX, bone resorption markers (58, 70, 119)		Randomized and nonrandomized trials Cross-sectional
	Stress fractures (120-136)		Systematic review Epidemiological Prospective cohort Case-control
Immunological	Diminished lymphocyte and increased neutrophil counts (137-139)		Review Randomized controlled trial Prospective cohort
	Increased white blood cell counts (73)		Prospective cohort
	Increased IL-6 levels (140-142)		Prospective cohort
	Decreased monocytes and T-cell subpopulation (143)		Prospective cohort
	T-cell hyporesponsiveness (144)		Clinical trial
	Neutrophil hyporesponsiveness and decreased phagocytic activity (145, 146)		Prospective cohort
	Decreased IgA levels (64, 147)		Prospective cohort Cross-sectional
	Increased likelihood of infections and illnesses (64, 147-151)		Cross-sectional
Muscle and performance	Decreased muscle mass and strength, increased risk of injuries (152-154)		Randomized controlled trial Cross-sectional Prospective
	Decreased explosive power (155)		Prospective
	Dysregulation of myokines (156, 157)		Review
	Reduced muscle glycogen content (158)		Crossover randomized controlled trial
	Decreased muscle protein synthesis (159)		Clinical trial
	Diminished activin and increased follistatin concentrations (43, 53, 111, 160, 161)		Clinical trial Prospective
	Impaired sports performance (162)		Review
Hematological	Iron deficiency anemia (163, 164)		Cross-sectional
	Abnormal levels of ferritin and transferrin, elevated levels of hepcidin (73, 142, 165-167)		Cross-sectional Prospective Case-control
Cardiovascular	Endothelial dysfunction concurrent with HA—association with increased cardiovascular risk (168)		Cross-sectional
Gastrointestinal	Nausea, regurgitation, upper abdominal bloating, diarrhoea, delayed gastric emptying and constipation (169, 170)		Systematic review
	Increased intestinal permeability (170, 171)		Systematic review

Abbreviations: BMD, bone mineral density; CTX, collagen type 1 C-telopeptide; GHBP, GH-binding protein; HA, hypothalamic amenorrhea; IGFBP, IGF binding protein; NTX, N-terminal telopeptide.

Substrate Metabolism and Food Intake

The biochemical fingerprint of short-term LEA can be observed in the context of acute nutrient deprivation, which is well known to trigger a classical physiological cascade. Fasting- and energy-deficiency induced hypoglycemia (172, 173) can be partially prevented, owing to reduced insulin production and elevated glucagon, epinephrine, and cortisol secretion secondary to starvation (174, 175). However, carbohydrate and glycogen stores are soon depleted, and gluconeogenesis takes over using amino acids and glycerol for 24 to 48 hours, followed by ketogenesis (174). After a few days, a branched-chain amino acid surge triggers the shift to lipid metabolism and ketogenesis, as attested by elevated cortisol, free fatty acids, β-hydroxybutyrate, and ketone body concentrations (173, 176, 177). Once adipose tissue (AT) reserves are depleted, the lipolytic metabolic shift is followed by protein catabolism. Caloric deprivation importantly produces changes in circulating adipokines, the prime sensors of energy deficiency and appetite and energy intake controlling hormones and, thus, calibrators of adipose tissue energy reserves and function. Acute energy deficits have a major effect on body composition, energy balance, and hormone concentrations, including leptin (42, 85, 178) but exerts little effect on adiponectin (179). A sustained energy deficit leading to diminished AT reserves decreases leptin production and alters resting metabolic rate (RMR) and neuroendocrine profile to prolong survival (180). In the long term, LEA can dysregulate the secretion of orexigenic peptides (ie, ghrelin and PYY), with studies reporting both increases and decreases in circulating levels (181, 182).

Neuroendocrine Effects

Hypothalamic-pituitary-gonadal axis

Suppression of the HPG axis is a major adaptive mechanism in states of energy deprivation. This adaptation occurs because procreation may not be prioritized when the survival of the organism is at risk because of LEA, which may not provide sufficient energy to support the creation and development of a new organism. LEA is associated with suppression of GnRH signaling and pulsatility, reduced pituitary gonadotropins, primarily LH levels (56, 80), and gonadal steroid hormones, including decreased estrone-1-glucuronide (principal metabolite of estradiol) and pregnanediol glucuronide (primary metabolite of progesterone), resulting in menstrual disturbances (60) and functional hypothalamic amenorrhea (FHA). Past studies have indicated that up to 50% of exercising women experience menstrual irregularities (76), whereas FHA in itself constitutes 1 of the most common manifestations of REDs (36, 77, 78). The presence of FHA depends on the duration and severity of LEA (30) and was recently reviewed (79). The highest prevalence for primary amenorrhea was reported in rhythmic gymnastics at 25%, for secondary amenorrhea in cycling at 56%, and for oligomenorrhea in boxing at 55% (79). However, study and group heterogeneity, as well as menstruation reporting, should be taken under careful consideration (77, 183).

There are conflicting results from studies investigating the occurrence of hypogonadism in male athletes, with some investigations underlining differences from female manifestations and varying forms of gonadal disruption in duration, quality, and reversibility, encompassing a vast array of hypogonadism forms that occur in elite male athletes (184). Likewise, albeit studied to a lesser extent, elite male endurance athletes may display notable downregulation of gonadotropin concentrations, blunted gonadotropin pulses, suppressed testosterone levels, and impaired spermatogenesis (30, 81-83, 86). In military milieus, a series of surveys on male Norwegian cadets under a 5-day arduous training course has demonstrated downregulated gonadotropin and sex steroid hormone levels (87-89). Numerous other studies in energy-deprived male soldiers have proven similar changes in testosterone and sex steroid hormones (185-188), even reaching “castrate levels” following prolonged ranger training courses (85).

Hypothalamic-pituitary-thyroid axis

The hypothalamic-pituitary-thyroid axis also adapts to energy deprivation (31) by implementing mechanisms that promote and prolong survival through energy conservation (78). Short-term energy deprivation exercise studies, including previously untrained regularly menstruating females, showed that LEA induces a nonlinear reduction of T3 and free T3 levels, which is associated with decreased metabolism and energy expenditure in states of reduced energy availability, and an increase of reverse T3 levels (ie, the inactive form of thyroid hormone) (57, 94). Downregulation of T3 has also been consistently observed in female athletes with LEA and FHA compared with eumenorrheic individuals (95-97).

More complex findings regarding T4 and free T4 changes have been observed. An increase of T4 has been described (94), suggesting a potential mechanism related to decreased conversion of T4 to T3, which is the active thyroid hormone. Other studies either identified a decrease in T4 levels (96) or did not detect a difference (57). Similar findings have been noted in soldiers, wherein marked downregulations of T3, T4, and TSH have been noted through longitudinal evaluations of training male cadets and conscripts (85, 90, 186, 188). Supporting literature on male athletes is limited, and in contrast to females, either a decrease or a neutral effect on T3 levels has been observed, emphasizing the role of baseline energy stores and sex (as indicated by varying leptin levels in males and females) (112, 189).

LEA association with thyroid hormone alterations is likely to be subject to several parameters, including participant's characteristics (within and between participant variability, sex, menstrual status), the underlying causes of LEA (dietary restriction, increased exercise energy expenditure), and the type, duration, and volume of athletic training. Therefore, the full consequences of LEA on thyroid-binding proteins and other components of the axis remain to be fully investigated. This research should be conducted in the context of the effects of those factors on both acute and chronic energy availability (ie, acute changes in energy balance in relation to the baseline energy stored in adipose tissue).

Hypothalamic-pituitary-somatotropic axis

In the anterior pituitary gland, GH is released by somatotroph cells and binds to GH receptors in multiple tissues, including the liver, which is the main organ responsible for IGF-1 secretion. GH may stimulate the production of IGF-1 via binding to GH receptor, activating Janus kinase 2 (JAK2) and subsequently the phosphorylation of signal transducer and activator of transcription 5 (STAT5), which plays an essential role in IGF-1 gene expression (190, 191).

LEA accelerates GH pulse frequency and accentuates GH peaks (98, 99), whereas it reduces IGF-1 production, potentially reflecting peripheral resistance or reduced sensitivity to growth signals through diminished liver production of IGFs and downregulated hepatic and peripheral somatogenic receptor expression (56, 99-102).

LEA in the form of energy deprivation culminating in hypothalamic amenorrhea, leads to increased levels of IGF binding protein 1 (IGFBP-1) and diminished levels of IGFBP-4, IGFBP-6, and growth hormone-binding globulin compared with healthy individuals (104), potentially indicating a consistent downregulation of hepatic GH receptors and reduced bioactivity of growth factors, hampering overall circulating GH activity. Moreover, resistance exercise, a potent anabolic stimulus, increased GH response and decreased IGF-1 levels in female and male participants during a short-term, calorie-restricted LEA state while on a postexercise protein or carbohydrate supplementation (103). Significant and consistent effects have been observed in male soldiers, wherein increased energy expenditure leads to an upregulation of IGFBP-1 and an intermediate downregulation of IGFBP-3 (186, 192, 193). These results suggest that a catabolic state may also limit the effectiveness and/or the potential benefits of resistance exercise during the calorie-restricted state.

Finally, GH may exert effects on carbohydrate and lipid metabolism and help to maintain euglycemia via the mobilization of fat stores in an IGF-1-independent way (194, 195). Therefore, GH resistance could be considered an adaptive response to LEA. The decreased IGF-1 levels may help to conserve energy, and through a decreased negative feedback at the pituitary, the subsequent increased circulating GH levels may help to maintain euglycemia under LEA conditions (190).

Hypothalamic-pituitary-adrenal axis

Increased serum and cerebrospinal fluid cortisol concentration (105-107), reduced cortisol response to intense exercise (106), as well as accentuated cortisol pulse amplitude compared with eumenorrheic exercisers (108) have all been established in women with HA. The diurnal pulsatility of salivary cortisol appears to be abolished in elite gymnasts, which may indicate an adaptation to prolonged periods of LEA (196). In the long term, however, cortisol levels may be increased, reflecting an increasingly stressful situation from prolonged energy deprivation. Malnutrition and anorexia, on the extreme spectrum of energy deficiency, as well as chronic excessive exercise, directly related to REDs, have been well-established as effectuators of acute and chronic stress and activate the hypothalamic-pituitary-adrenal (HPA) axis (109, 197). In turn, this effect entails alterations in almost all hypothalamic-pituitary axes and may lead to reductions in muscle mass and BMD (Fig. 1). This increase in cortisol levels is a mechanistically multifactorial process, which depends not only on decreasing leptin levels, as all the previously described neuroendocrine changes, but also on alterations in the AFI axis and other proinflammatory molecules in the periphery, as well as changes in neuropeptide levels centrally (Figs. 1-3). The elucidation of energy deficiency-induced changes in the HPA axis is further evident in studies of both male and female soldiers, consistently reporting an upregulation of cortisol (187, 188, 198, 199), which has been shown to be more pronounced in females (200).

Figure 1.

Acute and chronic exercise-induced stress responses of neuroendocrine axes. Acute and chronic exercise, particularly in excess, constitute important stress effectors. Acutely, this response activates the locus coeruleus-norepinephrine system and induces the secretion of catecholamines (sympathetic response), while also boosting glucocorticoid production, which are known to acutely produce a spike in GH. Chronic stress, on the other hand, also produces sustained increases in glucocorticoids through activation of the hypothalamic-pituitary-adrenal axis. The parent hormone, CRH, is known to suppress GnRH through β-endorphins, as well as TRH, and increase somatostatin, which in turn inhibits GH and TRH. Meanwhile, increased glucocorticoid production suppresses the concentrations of thyroid hormones, GH, and GnRH. These effects, when compounded and persistently accentuated in chronic stress, which entails excessive exercise, may lead to lower muscle mass and BMD, as well as to increased visceral adiposity risk. Abbreviations: BMD, bone mineral density; CRH, corticotropin-releasing hormone; GnRH, gonadotropin releasing hormone; IGFBP, IGF binding protein; LC/NE, locus coeruleus-norepinephrine; SNS, sympathetic nervous system; TRH, thyroid releasing hormone.

Figure 2.

Endocrine effects underlying REDs-associated states across tissues and systems, highlighting the independent functions of leptin and the AFI axis. Arrows with plus signs indicate positive relationships or feedback in states of hypoleptinemia; arrows with minus signs indicate negative feedback; up and down arrows represent changes (increase or decrease, respectively) occurring under energy deprivation. Leptin is a central regulator of the HPG axis and can simultaneously influence bone health both directly and indirectly through modulations in the reproductive system. In states of LEA, leptin levels are decreased, thus leading to downregulation of the reproductive system and its hormones, bringing about long-term directly or indirectly bone loss; these changes can be reversed through leptin supplementation. However, the AFI axis functions independently of leptin. LEA upregulates follistatins and decreases activins. Follistatins thus render the reproductive system inactive, help preserve precious muscle tissue, which is achieved through follistatins blunting the catabolic functions of activins and myostatin. In the liver, follistatin concentrations are inversely correlated with insulin sensitivity and improved glucose metabolism, thus leading to de novo lipogenesis and glycemia. Emerging evidence demonstrates that lower activins/follistatin ratios are also linked to osteopenia and osteoporosis. Abbreviations: ACTRII, activin receptor type 2; AMH, anti-müllerian hormone; BMD, bone mineral density; fT3, free triiodothyronine; FSTL3, follistatin-like 3; HPA, hypothalamic-pituitary-adrenal; HPG, hypothalamic-pituitary-gonadal; HPT, hypothalamic-pituitary-thyroid; IGF-1, insulin-like growth factor; IGFBP, IGF binding protein; TG, triglyceride.

Figure 3.

Neuroendocrine regulation of reproduction, bone health, growth, thyroid, and adrenal function by leptin. Leptin concentrations are directly proportional to AT energy reserves, yet obese patients demonstrate significant resistance to its hormonal effects. In the brain, leptin can act directly on dopaminergic neurons or bind to its receptor in the hypothalamus, leading to upregulation of proopiomelanocortin and the downregulation of the orexigenic NPY and agouti-related protein (AgRP) neurons, thus exerting a holistic control over appetite and food intake. Leptin directly influences the SNS in animals but not in humans. The hypothalamic effects of leptin influence the pituitary downstream, affecting several vital endocrine axes in both animal models and human cohorts. Through kisspeptin-1 and other neurons, GnRH secretion triggers LH and FSH production in a pulsatile manner directly influenced by leptin's circadian variations. Thyroid and corticoid hormone adaptations, driven by perturbations in TRH, TSH, and CRH, respectively, are also influenced by leptin, especially in animal models but also in humans in lesser degrees. Leptin also exerts control over the growth factor IGF axis, with administration in humans effectuating increases in IGF-1 and IGFBP-1. Last, leptin constitutes a key conduit of bone health, either directly or indirectly by influencing estrogen and other trophic hormone concentrations. Abbreviations: AgRP, agouti-related protein; CNS, central nervous system; CRH, corticotropin-releasing hormone; E2, estradiol; IGF, insulin-like growth factor; Kiss1, kisspeptin-1; NPY, neuropeptide-Y; POMC, proopiomelanocortin; SNS, sympathetic nervous system; TRH, thyrotropin-releasing hormone.

Immunological Effects

As attested by experiments in mice, mainly because of falling leptin levels in response to energy deprivation, an abnormal Th1/Th2 balance, upregulated production of proinflammatory cytokines (ie, IL-6 and TNF-α), and cortisol is observed with LEA and may increase the risk of specific infections (201). In the 2018 and 2023 consensus statements (30, 202), the IOC noted that LEA might impair the immune system, highlighting the increased rates of upper respiratory and gastrointestinal tract infections based on observational data from amenorrheic elite female distance runners (64) and Olympic athletes assessed as being at risk for LEA (using the low energy availability in females questionnaire) (148, 149). Findings of various studies have suggested adverse effects of energy restriction and weight loss on immunological parameters, such as impaired humoral and cell-mediated immune function and subsequent increased risk of potential infections in athletes. In particular, 2 observational studies exploring the effects of intensive training and rapid weight loss on immunological markers revealed suppressed mucosal immunity indicated by the reduced levels of salivary Ig A and an increased upper respiratory tract infection incidence in male and female elite taekwondo athletes (147, 150). Decreased salivary Ig A levels and increased upper respiratory tract infection symptoms were also observed in amenorrheic elite female distance runners compared with the eumenorrheic runners (203). Furthermore, weight loss (reduced energy intake) was associated with impaired cell-mediated immune function (ie, decreased counts of specific monocytes and T-cell subpopulations) and high susceptibility to upper respiratory tract infection symptoms in an observational study of judo athletes (143). In addition, weight reduction resulted in decreased cytokine production (ie, interferon-g), proliferation of T cells (144), and neutrophil phagocytic activity (145, 146) in case-control studies in amateur wrestlers and judo athletes. However, in a randomized control trial, calorie restriction during high-intensity exercise was associated with an increase in IL-6 and TNF-a levels and a significant variation in lymphocyte, leukocyte, and neutrophil counts (204). Of note, a large cross-sectional evaluation of 2247 elite athletes has shown markedly downregulated white blood cell and neutrophil counts in athletes trained for aerobic vs skill-based sports, indicating a negative immune adaptive response which is distinct for individuals at risk for LEA (205). On the other hand, male soldiers also display reportedly increased circulating levels of IL-6 and TNF-a (140-142) as well as hepcidin, also outlined in the hematological effects section that follows and can display downregulated in vitro T-lymphocyte responses and increases in infection rates (151, 206). Further research is needed to more precisely identify the relationship between LEA, immunological parameters, immune functions, and susceptibility to infectious diseases. However, these changes should be seen in the broader context of energy deprivation leading to immune dysfunction in a direct but also in a potentially indirect way through alterations in leptin, AFI, and cortisol levels. Pathophysiologically, similar but more pronounced changes occur in other states of more severe energy deprivation, such as in starving populations in developing countries.

Hematological Effects

Lower hemoglobin concentrations and iron deficiency have been observed in both male and female professional athletes (163, 164, 207). LEA has been associated with iron deficiency and impaired ferritin levels, which may result in iron deficiency anemia commonly seen in young female athletes (208, 209). Iron deficiency has been suggested as a hematological concern of interest in the context of REDs, according to the updated 2018 IOC consensus statement, and a potential driver for other REDs-related health complications (30). The prevalence of iron deficiency has also been frequently reported in both male and female soldiers, both throughout and following training (210-214), alongside elevated hepcidin and ferritin concentrations (141, 142).

LEA may impact the iron balance in both direct and indirect ways. Innately, an impaired reproductive axis in females results in menstrual dysfunction and amenorrhea. Therefore, LEA may improve iron status by reducing iron losses from menstrual bleeding. However, elevated acute phase reactants and proinflammatory molecules, including hepcidin concentrations (a known iron-regulating protein associated with reduced dietary iron absorption and impaired iron metabolism) and abnormal ferritin levels, and low-grade inflammation-induced iron deficiency have been observed in athletes with a high-intensity workload (165). Moreover, LEA accompanied by conditions with high energy expenditure may also trigger an immune response and increase hepcidin levels, which may induce potential declines in iron status (142, 166). However, the duration and type of exercise might also have different effects on iron concentrations (167). In addition, an indirect effect of LEA on iron levels may be secondary to undernutrition or decreased macronutrient (ie, low carbohydrate availability) and micronutrient intake combined with increased iron demands and a substantial reduction in bioavailable iron in athletes (215).

On the other hand, iron deficiency may also interact with multiple physiological systems, resulting in exacerbated energy deficiency, impaired reproductive functions, and poor bone health (the 3 pillars of the known Female Athlete Triad), which are REDs-related health consequences (30, 216). In more detail, iron deficiency may be associated with reproductive endocrine dysfunction, hyperprolactinemia, infertility issues, reduced metabolic efficiency, and worsened energy status via impaired thyroid function (209). In addition, iron deficiency may induce neuroendocrine abnormalities by impairing the levels and function of dopamine, serotonin, and norepinephrine neurotransmitters (217). These neuropeptide alterations might contribute to traits associated with anxiety-related disorder symptoms and disturbed appetitive and eating behaviors (218, 219). In addition, iron deficiency per se may induce negative feelings of anxiety, depression, and low perceived quality of life (220). Therefore, it has been proposed that iron deficiency may also affect the risk for DE by increasing anxiety-related behaviors and disrupting eating behaviors via neuroendocrine aberrations (209). Iron deficiency may also compromise bone health through imbalances in growth factors (eg, GH and IGF-1 suppression), hypoxia-induced bone resorption, and thyroid dysregulation (209).

The available research is limited. Future investigation is required to elucidate further the relationships between iron deficiency and LEA and its components and explore the potential effects of dietary interventions, supplementation, neuroendocrine regulation, and restoration of menstrual function on iron metabolism and REDs outcomes.

Bone marrow fat actively contributes to various metabolic processes such as energy storage, endocrine function, and bone metabolism. The most abundant type of cells in the bone marrow cavity and hematopoietic microenvironment are bone marrow adipocytes. These cells play a crucial role in maintaining the balance between the proliferation and differentiation of hematopoietic stem cells, thereby supporting normal blood cell formation (221). Bone marrow (BM) adipose tissue (BMAT) is regulated in a distinct manner that reflects visceral fat stores in obesity and does not shrink under caloric deprivation (221). BMAT possesses both metabolic-, bone-, and immune-related properties and is considered the third-largest fat storage after subcutaneous and visceral fat, accounting for up to 70% of BM volume and ∼10% of total adipose mass. It derives from marrow adipogenic lineage precursors, which are cells that possess adipocyte markers but do not contain lipid droplets and function by preserving marrow vasculature and preventing the differentiation of mesenchymal progenitor cells into osteoblasts (222).

BMAT, a metabolically active, insulin-sensitive tissue, correlates with the amount of visceral fat in obesity, whereas obesity and aging induce ectopic adipocyte accumulation in the BM (223, 224). BMAT accumulates in individuals with chronic caloric restriction, such as in anorexia nervosa and possibly REDs, or during skeletal growth periods, localized principally within the tibia or femur (225). Studies in mice have shown how BMAT fails to mobilize its lipid reserve during moderate starvation (226) and is resistant to β-adrenergic lipolysis stimulation.

Cardiovascular Effects

In young endurance athletes, a link between amenorrhea and endothelial dysfunction has been suggested (168). Low estrogen levels have also been associated with unfavorable lipid profiles and increased cardiovascular risk as well as compromised orthostatic and cardiac sympathetic responses (168, 227).

Heart rate variability (HRV), reflecting autonomic nervous system function, is a metric of fitness and performance in sports and the military (228). Elite endurance athletes who participated in a 4-year longitudinal study were classified as experiencing fatigue or not based on their scoring on a specific questionnaire designed and validated by the consensus group on overtraining of the French Society of Sports Medicine (229). Athletes classified in the “fatigue state” (score exceeded 20 negative items of 54) experienced a lower and more variable HRV than their counterparts not classified in the fatigue state (229). However, direct evidence regarding cardiovascular health and HRV in REDs in athletic populations and the military is still limited; more research is warranted (33).

Gastrointestinal Effects

The redistribution of blood and nutrients toward metabolically active tissues during exercise can lead to gastrointestinal disturbances, alterations in the gut microbiome, and changes in markers of gut health, damage, and permeability in healthy exercising athletes (171, 230, 231). REDs has been associated with gastrointestinal symptoms in athletes, such as gut damage, increased intestinal permeability, delayed gastric emptying, constipation, and worsened sphincter control (30, 171). Several gastrointestinal disturbances (increased intestinal permeability, diarrhea, irritable bowel, chronic pain, and constipation) (140, 232, 233) and even relevant clinical entities (eg, Gulf War syndrome) (234) have been reported in soldiers, yet these have not been adequately delineated in the context of REDs.

Muscle and Performance Effects

The beneficial effects of exercise include reductions in body fat and inflammation and improvements in strength and endurance (235, 236). Moreover, adequate carbohydrate and protein intake are essential to maintain blood glucose levels, replenish muscle glycogen, repair and synthesize muscle tissue, promote muscle protein synthesis (237) and muscle adaptation to training (238, 239), and maximize training outcomes. In addition, it has been suggested that resistance exercise combined with increased postexercise protein availability may promote muscle protein synthesis in a dose-dependent way during a short-term energy deficiency state and ultimately preserve lean (muscle) mass in the long term (159).

Therefore, energy deficits, often accompanied by inadequate carbohydrate or protein intake (240), may lead to lower muscle glycogen stores (56, 158), which may result in unfavorable health outcomes (240) and sports performance (74). In addition, incomplete recovery and increased fatigue further impede athletic or military goals (21, 30). Importantly, as shown by a recent investigation in elite male endurance athletes, reduced performance and explosive power as a result of reduced energy availability may precede hormonal changes (ie, reductions in IGF-1 levels) (155), whereas previous studies on Norwegian soldiers have demonstrated a progressive deterioration of muscle damage, soreness, and blunted performance (241), and a marked and persistent drop in performance even 1 week after hormonal imbalances were normalized following refeeding (185).

Bone Health and Stress Fractures

Along with decreased muscle mass and reduced performance, LEA may result in compromised bone health linked to increased risk of SF in athletes and military personnel as well as long-term consequences (ie, osteoporosis) (21, 118, 242). Moreover, a negative dose-response association between EA and bone turnover was found in a randomized control trial in which the energy availability (ie, 10, 20, or 30 kcal/kg FFM/day) was controlled in 29 young, regularly menstruating exercising women (58). Potential sex differences have been described, with females displaying increased bone resorption and decreased bone formation (70). In male athletes, loss of BMD is often characterized as a direct or indirect effect of LEA, particularly in weight-restrictive, lean endurance sports (eg, cycling, rowing, distance running), leading to prolonged periods of LEA (71).

Both short-term diet-induced LEA and exercise-induced LEA have been associated with reduced bone formation in eumenorrheic women. Further evidence suggests that the type of diet may also play a role in bone formation. In particular, a low-carbohydrate diet has been found to be associated with a more profound reduced bone formation, whereas sufficient energy, including adequate carbohydrate support, may mitigate the exercise-induced unfavorable bone turnover responses (74). Inadequate nutrition, regarding both macronutrients (eg, carbohydrates, protein) and vitamins and micronutrients (eg, vitamin D and calcium, iron), as well as hormonal disturbances (eg, sex steroid deficiency, GH resistance, hypercortisolism, mainly from hypoleptinemia and disturbances in other adipokines and the AFI axis) may all directly and indirectly contribute to increased bone resorption and low BMD. Additional SF risk factors among athletes and military personnel include pronounced mechanical stress without sufficient recovery (120).

Early case reports in athletes indicate that the most common injuries are localized in the tibia (49.1%), tarsals (25.3%), and metatarsals (8.8%), as well as the femur, fibula, and pelvis (243). Furthermore, low weight and BMD have been associated with a higher risk of bone stress injuries in trabecular-rich compared with cortical-rich locations (121). Military personnel is also heavily burdened by SF reaching a cumulative incidence of 5.69 per 1000 person-years in large-scale epidemiological studies (122). These numbers differ between men and women (reportedly 2.05 vs 7.47 per 1000 person-years in male and female soldiers, respectively, representing >5.2 million person-years of risk (123); other reports delineate an incidence of 19.3 vs 79.9/1000 recruits respectively in general SF (124) and a prevalence of 2.76 vs 5.78 per 1000 person-years respectively for ankle-foot complex SFs) (125). Apart from epidemiology, soldiers are known to display elevated levels of bone turnover markers and parathormone, which are, however, higher in male soldiers (119) who similarly display a lower total bone mineral content (BMC) compared with female counterparts (126). SFs impose an estimated cost of $100 million annually to the US military, not including lost duty (244).

More prolonged (58) studies with larger sample sizes are needed to refine the dose-response relationships between chronic restrictions of EA and bone turnover and the underlying mechanisms and determine the optimal nutritional interventions to restore bone health (117).

Pathophysiology and Mechanistic Pathways Underlying REDs-associated Disorders

In states of LEA, the body strives to preserve the function of systems necessary for immediate survival, diverting energy from nonessential systems (eg, reproduction). From an endocrine point of view, exercise and inflammation both induce and are influenced by various hormonal changes, particularly in adipokines, which constitute the main cellular and metabolic signals of AT and are prime regulators of energy balance (245).

Adipose Tissue

White AT (WAT) stores energy, brown AT (BAT) is considered a key mediator of thermogenesis. A third type of AT termed “beige” or “brite” (brown-in-white) fat comprises recruitable thermogenic brown adipocytes within WAT depots, as demonstrated through in vitro and preclinical experiments (246). BAT was first investigated as a key thermogenetic tissue in small mammals (247). In humans, less than half of BAT deposits are activated under cold exposure, and BAT appears less activated in obese individuals compared with lean counterparts (248). Exercise can induce “browning” of WAT, causing AT to switch from storing white adipocytes to beige adipocytes, and has been hypothesized to be an adaptive process (249). Interestingly, BAT thermogenic activity may be lower in male endurance athletes and chronically exercising women compared with nonexercising controls (250, 251). BAT activity is negatively correlated with amenorrhea duration (251), potentially indicating compromised metabolic activity of BAT in REDs to preserve energy, but these data remain to be confirmed, especially given that BAT is extremely low to start with in normal adults. BMAT has also been implicated in energy regulation, bone health, and hematopoiesis, and its role as a mediator in REDs-related disorders is emerging.

Leptin

Leptin physiology

Leptin is a hormone secreted by adipose tissue. It is an adipokine that plays a crucial role in regulating energy metabolism and adipose tissue. Leptin helps maintain the overall physiological balance of the body and regulates important functions related to metabolism (252). The circulating levels of leptin display sexual dimorphism, being higher in women than in men, and reflect mainly acute changes in energy intake (with a precipitous drop in acute energy deprivation) as well as long-term energy stores (with decreasing levels reflecting LEA). Leptin binds and activates its receptors not only in the brain but also in peripheral organs, thus mediating the effects of dysregulated energy homeostasis in most REDs-related hormonal subsystems (Fig. 3) (1, 27, 37, 253). Therefore, leptin is a reliable indicator of energy availability and, by extension, energy deprivation in animals and humans (47, 62, 253).

In humans, the leptin gene is expressed primarily in WAT and in much lower concentrations in other tissues such as the stomach, placenta, mammary gland, and the immune system (254). Leptin displays a pulsatile pattern of secretion, which is notably accentuated in women and produces elevated basal leptin concentrations in the female sex (255). On a molecular level, leptin can cross the blood-brain barrier through a facilitated system and directly (256) binds in the brain but also in other organs to its receptor LEPR, which belongs to the JAK-STAT family of transcription factors, primarily in the hypothalamus and more specifically in areas such as the arcuate nucleus of the hypothalamus (256-258). In the context of neuroendocrine regulation, falling leptin levels in the context of energy deprivation activate orexigenic NPY/agouti-related protein neurons (258) and inhibit anorexigenic proopiomelanocortin activity in the brain, thus constituting the afferent signal of an AT-mediated negative feedback loop regulating energy availability by influencing food intake and altering physiology (37, 259). Leptin regulates the hypothalamic-pituitary-peripheral hormonal axes, as mentioned previously, and has secondary actions in peripheral tissues, including the immune system, bones, muscle, and adipose tissue, all aiming at maintaining energy homeostasis under normal physiological conditions but leading to pathophysiological abnormalities and diseases such as SF when the physiologically adaptive processes last longer than they should.

Mice with deleted leptin genes, now identified as the ob/ob phenotype, as well as humans with similar genetic defects, are innately obese and display impaired thermogenesis, diminished energy expenditure, and endocrine and metabolic abnormalities (246, 260). Leptin replacement in these mice and humans reduces fat mass without affecting lean body mass and improves neuroendocrine function and metabolism (261). In addition, females tend to have increased leptin production and pulse amplitude, potentially indicating an innate relative leptin resistance (255, 262). However, in LEA, short-term fasting induces a profound and rapid decrease in serum leptin concentrations, potentially through increased leptin clearance (263), which precede and are disproportionate to changes in fat mass (42, 44, 264). Moreover, the weight loss attained during fasting with concomitant leptin administration in physiologic, supraphysiologic, or pharmacologic doses is independent of dose or circulating leptin levels achieved (263). Long-term energy deprivation through caloric restriction likewise results in markedly downregulated leptin concentrations (265). Women with DE and amenorrhea are also known to have reduced circulating leptin levels and blunted pulses (107, 266, 267). In this energy-deficient milieu of low basal leptin, human studies show notable neuroendocrine benefits in response to both physiological and supraphysiological doses of r-metHuLeptin (42, 45-47, 114) (Fig. 2).

Neuroendocrine effects of leptin

The pivotal role of leptin as key regulator of neuroendocrine axes, whose dysfunctions constitute components of REDs, is physiologically distinct between animals and humans.

Hypothalamic-pituitary-gonadal axis

Leptin stimulates GnRH release from incubated neuronal cells and accentuates GnRH pulsatility, but not amplitude, in hypothalamic neurons (268). Low leptin concentrations, common in LEA, downregulate GnRH secretion from the hypothalamus, through either direct action or indirect changes in neuropeptide systems, such as neurokinin or kisspeptin-mediated pathways (269).

Ob/ob mice, which are sterile, regain reproductive capacity following leptin administration (270). In calorically deprived mice with low leptin levels, leptin administration has also been shown to reverse the LEA-induced reductions in LH and testosterone (271). Likewise, r-metHuLeptin replacement stimulated gonadotropin pulsatility, culminating in normal LH and FSH secretion in a seminal study on a child with congenital leptin deficiency (CLD) (272) and another case report of leptin replacement in a female child with CLD for 24 months observed restoration of LH pulsatility and pubertal development, including menarche (273). In a case series of 3 adults with CLD, leptin replacement increased LH and testosterone levels, associated with the development of secondary sex characteristics, in a male patient and induced ovulation in 2 female patients (274).

Disruptions in circulating leptin and reproductive hormone profiles are evident in LEA and have been observed in healthy individuals after short-term starvation (88, 275). In 8 healthy males, 72 hours of fasting altered LH pulsatility and markedly decreased testosterone. Concomitant r-metHuLeptin administration maintained normal LH pulsatility and testosterone levels (42). In a pilot study of 14 adult females with hypothalamic amenorrhea resulting from vigorous exercise or low weight, the administration of recombinant leptin (r-metHuLeptin) appeared to improve reproductive function. In particular, women who received r-metHuLeptin demonstrated increases in LH pulsatility, estradiol concentrations, ovarian activity, and the number and size of follicles (46). Another randomized placebo-controlled trial in women with HA, 11 of whom were assigned to r-metHuLeptin and 9 to placebo, reported elevated estradiol and progesterone levels in the leptin-treated women, in addition to improvements in other neuroendocrine abnormalities after 36 weeks (45). Treatment with r-metHuLeptin resulted in the resumption of menses in a significant proportion of adhering participants (114). During the entire study period, the r-metHuLeptin dose was reduced in 5 of 11 metreleptin-treated participants who lost more than 5% of their baseline body weight, whereas 2 of the participants withdrew from the study because of weight loss when circulating leptin levels exceeded the physiologically normal range. Besides local injection site reactions developed by 1 participant, no other clinically significant drug-related adverse events were documented (114). Thus, leptin administration at replacement doses, sufficient to maintain circulating leptin levels within the normal range, normalizes neuroendocrine axes impairments from energy deprivation. However, when leptin is administered in supraphysiological doses, the side effect of further suppression of fat mass and weight loss is observed (apparently as part of a feedback loop to limit the endogenous contributions to circulating leptin when the exogenously administered leptin is too high).

Hypothalamic-pituitary-thyroid axis

Leptin stimulates thyrotropin-releasing hormone production in hypothalamic cells of fasted rodents (276) and prevents the fasting-induced downregulation of TSH pulsatility and T4 secretion, albeit partially (271). In humans, LEPR mutations result in low T4 levels but normal basal TSH levels and sustained TSH responses to thyrotropin-releasing hormone (277). In leptin-deficient children, leptin replacement increased free T3 and free T4 concentrations (273).

In lean healthy men, r-metHuLeptin administration in replacement doses prevented the energy derivation-induced suppression of TSH pulsatility and increased the concentrations of free T4 but did not affect T3 and reverse T3 (42). However, in 7 women following a similar protocol, r-metHuLeptin did not prevent fasting-induced fluctuations in thyroid hormones (44). Women with HA receiving high-dose r-metHuLeptin for 3 months had increased TSH pulse frequency and amplitude and transiently increased free T3 and free T4 levels (46), whereas treatment over 36 weeks resulted in increased free T3 but not free T4 or TSH levels, compared with placebo (45). Thus, although the precise effects of leptin on the thyroid axis remain to be fully elucidated, findings to date are consistent with the hypothesis that fasting-induced thyroid hormone changes are mediated through falling leptin levels due to energy deprivation. It is however established that among overweight or obese individuals, who are leptin resistant, long-term leptin replacement therapy to push circulating leptin levels to even higher supraphysiologic levels does not produce any changes in circulating TSH or any of the active thyroid hormones, both total and free, indicating a neuroendocrine milieu of leptin resistance (278).

Hypothalamic-pituitary-somatotropic axis

Incubation of human pituitary cells in leptin increases GH secretion (279). and in mice, leptin prevents the fasting-induced suppression of GH and IGF-1 and partially corrects GHRH mRNA expression (280). In adults with CLD, r-metHuLeptin, in physiological replacement doses, elevated IGFBP-1 and 2, but not IGF-1 (274).

When administered in healthy men (42) or women (44) undergoing a 72-hour complete energy deprivation, leptin in replacement doses does not markedly influence GH pulsatility or IGFBP levels but blunts the starvation-induced decrease (104) of total, but not free, IGF-1 secretion. In contrast, in a 3-month pilot study of women with HA treated with r-metHuLeptin, IGF-1 increased in the first month and returned to baseline levels, whereas IGFBP-3 was markedly increased during months 2 and 3 (46). A longer study of female athletes with HA also found that r-metHuLeptin treatment for 36 weeks significantly increased total IGF-1 and tended to elevate free IGF-1 and IGFBP-3 levels compared with controls (104). These changes in women with HA are evident with up to 2 years of r-metHuLeptin treatment (114). Overall, leptin administration seems to have beneficial effects on the GH-IGF1 axis in states of long-term energy deficiency rather than short-term starvation.

Hypothalamic-pituitary-adrenal axis

LEA increases ACTH and corticosterone in mice, and leptin administration inhibits corticotropin-releasing hormone from hypothalamic cells and reverses the LEA-induced activation of the HPA axis (281). However, humans with LEPR mutations do not present with HPA abnormalities (277), indicating a key interspecies difference.

Short-term pharmacological leptin administration in healthy fasted lean men (42) and women (44) does not affect fasting-induced changes in plasma or urine cortisol concentrations, cortisol pulsatility, or the HPA axis overall. Three months of r-metHuLeptin treatment in women with HA failed to induce changes in cortisol or ACTH levels (46); however, long-term administration of replacement-dose leptin of 36 weeks to 2 years in women with HA results in marked declines in cortisol levels (42, 45, 46, 114). Overall, leptin administration seems to have beneficial effects on the ACTH-cortisol axis in states of long-term energy deficiency rather than short-term starvation.

Sympathetic nervous system

Leptin also differentially affects components of the sympathetic nervous system (heart rate, blood pressure, catecholamines) in animals and humans. In mice, leptin increases BAT sympathetic tone in the dorsomedial hypothalamus and activates sympathetic nerves (282-284). However, leptin administration in short-term studies in healthy lean men and women and long-term studies in women with HA do not demonstrate any significant changes in resting metabolic rate, body temperature, resting heart rate, blood pressure, or urinary catecholamine levels (285).

Immune and hemopoietic systems

LEPR is expressed in almost all immune cells (286, 287), and its long form, which is considered to be fully capable of JAK/STAT signaling, is expressed on natural killer cells and activated T-lymphocytes (287). Incubation with leptin has been shown to activate B cells and induce the expression of cytokines and signaling molecules in macrophages and mononuclear cells (288). When administered in fasted mice, leptin reversed the starvation-induced suppression of T-lymphocyte responses by increasing Th1 cells and downregulating Th2 cytokine production (289). In humans, leptin administration in children with congenital leptin deficiency boosted T-cell responsiveness and increased CD4+ T-lymphocytes (273), and in adults it boosted signals that stimulate chemotaxis and increase CD4+ cell counts (290, 291).

In a study of 14 women with HA (6 treated with placebo and 8 with replacement-dose r-metHuLeptin for 36 weeks) and 13 healthy controls, individuals with HA had lower basal leptin levels and T-cell counts compared with controls. Treatment with r-MetHuLeptin improved CD4+ cell counts and their proliferative responses following in vitro stimulation. Leptin additionally boosted the expression of cell survival and hormonal response genes and blunted apoptosis-related genes. Finally, leptin activated cell proliferation and growth pathways within CD4+ T-lymphocytes, such as AMPK, STAT3, and mTOR (291).

Leptin is also known to promote the proliferation, differentiation, and activation of marrow hemopoietic cells, induce bone marrow colonies, and interact with granulocyte precursors (292-294). More recently, the presence of LEPR on hematopoietic stem cells was identified as an indicator of hematopoietic stem cell engrafting quality and capacity (295).

Bone Health

Loss of BMD constitutes 1 of the prime characteristics of the Female Athlete Triad and is a pivotal feature of REDs. The status of bone tissue is a direct reflection of human nutritional status, including but not limited to calcium and vitamin D intake, and is regulated by PTH and other hormonal conduits, including leptin, particularly in energy-insufficiency states (296) (Fig. 4).

Figure 4.

Functions and pathways of leptin in bone health. Arrows with plus signs indicate positive relationships or feedback; arrows with minus signs indicate negative feedback. Physiologically, RANKL is critical for the differentiation of osteocytes into osteoclasts and is inhibited by osteoprotegerin (OPG). Meanwhile, sclerostin binds to LRP receptors and antagonizes WNT signaling, which is key for osteoblast activation. Leptin's role in bone physiology can be discerned into indirect and direct. Leptin indirectly influences bone status by producing changes in hormone levels such as estrogens and growth factors, especially IGF-1. Studies in animals and humans receiving leptin have demonstrated positive effects on bone mass through downregulation of bone resorption and increased osteoblast formation and activity. Leptin may also act through CART to deactivate osteoblasts while it additionally increases osteoprotegerin expression, which binds RANKL and thus deactivates osteoclasts in a manner similar to estrogen. This is evident in human studies showcasing a downregulation of CTX, a marker of bone catabolism. On the other hand, leptin upregulates a series of important bone formation factors, including TGF-B, collagen 1a, bone-specific alkaline phosphatase, and osteocalcin. Leptin may additionally influence PTH secretion, while it also regulates vitamin D metabolism and activates FGF-23 in animal models, which is central to bone and mineral metabolism, yet does not impact sclerostin, and only tends to increase osteoprotegerin (OPG) and decrease RANKL levels in humans. Intracellularly, leptin drives osteoblast proliferation through cyclin D1, activated through the same molecular clock the hormone oversees. Overall, leptin demonstrates a tendency to improve cortical bone integrity and thus may pinpoint a strategy to reduce SF risk in athletes and military personnel. Abbreviations: AP-1, activator protein 1; BAP, bone-specific alkaline phosphatase; CART, cocaine and amphetamine regulated transcript; Col1A, collagen type I alpha 1; CTX, C-terminal telopeptide of collagen; CY, cytochrome; E2, estradiol; FGF23, fibroblast growth factor-23; LPR, low-density lipoprotein receptor-related protein; OPG, osteoprotegerin; RANKL, receptor activator of nuclear factor kappa-Β ligand; VDR, vitamin D receptor.

When elucidating the inner workings of bone regulation, it is important to mention the roles of receptor-activator of nuclear factor κΒ ligand (RANKL) pathway signaling, osteoprotegerin (OPG), and sclerostin, among others, in regulating the osteoblast-osteoclast equilibrium. Osteoclasts, responsible for bone resorption, are activated through RANKL ligand signaling, which is naturally inhibited by OPG, itself secreted by osteoblasts (297). On the other hand, sclerostin, which is also produced by osteocytes, binds to low-density lipoprotein receptor-related protein receptors, which are involved in WNT signaling and the regulation of bone mass (298). Normally, WNT activation induces osteoblast activation, whereas sclerostin inhibits lipoprotein receptor-related protein and WNT signaling, thus reducing bone formation (298). Sex steroid hormones and estrogen directly regulate OPG and RANKL production (299) and may influence RANKL-induced osteoclast differentiation (300), whereas in animals, estrogen targets and suppresses RANKL expression in bone lining cells (301).

High cortisol may additionally directly lead to increased bone resorption by means of reducing OPG and increasing RANKL expression, promoting osteoclast survival and decreased bone formation via inducing osteoblast apoptosis, and by preferentially driving BM stromal cell differentiation to adipocytes over osteocytes (253, 302-304). Moreover, cortisol concentrations have been positively correlated with collagen type 1 C-telopeptide (CTX), a bone resorption marker, and inversely associated with LH pulsatility and procollagen type 1 N-terminal propeptide (P1NP), a marker of bone collagen deposition (110).

PTH aims to maintain and mobilize circulating calcium from the bones, kidneys, and gut, while reducing the renal reabsorption of phosphate. Vitamin D, on the other hand, stimulates both calcium and phosphate metabolism, providing adequate minerals for bone formation (305). In exercising humans, short-term exercise often increases circulating PTH, whereas conflicting results are present in long-term exercising cohorts, even displaying decreasing trends (306). On the other hand, vitamin D deficiency is a frequent occurrence in athletes and often correlated with increased fracture risk (307), precipitating therapeutic recommendations for supplementation in amenorrhea with osteoporosis.

Estrogen promotes osteoblast proliferation and blocks osteoclast differentiation (308). Thus, hypoestrogenism results in increased bone resorption with a relative deficit in bone formation and is associated with low BMD (309-312). Furthermore, androgens play important antiresorptive and anabolic roles in bone for men and women by decreasing osteoclastogenesis and stimulating osteoclast apoptosis while inhibiting osteoblast apoptosis (313).

Low IGF-1 levels and activity are also culprits of compromised BMD in LEA. IGF-1 has been positively correlated with bone mass (314) and shown to stimulate osteoblast proliferation and activity (315), is considered essential for the differentiation of mesenchymal osteoprogenitors (316), has been associated with bone resorption, formation, and healing, and linked, alongside GH administration, to rapid clinical improvements in tibial fractures (317). IGFBPs also possess an important anabolic role within the skeleton by directing the actions of IGFs and particularly IGF-1 within bone tissue (314).

The HPA axis, which responds to stress and LEA by excess cortisol production, influences bone health as well. Cortisol can inhibit osteoblast and osteoclast growth, hinder the anabolic effects of growth hormone, and suppress IGF-1 synthesis in skeletal cells (99, 318-320). Excess cortisol is also known to suppress the HPG axis, which is pivotal for bone regulation (12, 26, 36, 321).

Role of leptin in bone health

Leptin serves as a link between AT and bone both directly by stimulating osteoblasts, chondrocytes, and BM stromal cells and indirectly through alterations in neuroendocrine activity (49, 322). At the molecular level, leptin regulates osteoblasts by upregulating Ap-1 gene expression, which activates osteoblast proliferation through cyclin D1 (323). Leptin has also been shown to promote osteoblast differentiation by upregulating TGF-β, IGF-1, collagen-1α, alkaline phosphatase, and osteocalcin mRNA in human iliac crest osteoblasts (324). Osteocalcin, in particular, is known to be affected by leptin via the hypothalamus and constitutes a notable orchestrator of bone health, insulin sensitivity, and energy balance (325, 326). Moreover, leptin may also promote osteoblastogenesis by upregulating CYP27B1, CYP27A1, and VDR, which are mediators in vitamin D metabolism (327). In rodent models, leptin also increases FGF-23 expression in bone, leading to suppression of renal 1α-hydroxylase expression (328). It is now accepted that leptin represses osteoclast differentiation through cocaine- and amphetamine-regulated transcript and acts in a similar way to estrogen, increasing OPG levels, which binds to RANKL (329) and thus blunts osteoclast activity. Finally, leptin also exerts beneficial effects on bone through an overall amelioration of metabolism and neuroendocrine health (49). In lean hypoleptinemic women, r-metHuLeptin administration for 36 weeks markedly downregulated intact PTH and RANKL to OPG ratio did not alter sclerostin, dickkopf-1, and FGF-23 levels, but it nevertheless tended to increase serum OPG and decreased serum RANKL (330). This indicates differences in mechanisms of action between humans and animal models.

Low leptin levels have also been observed among female athletes with elevated bone turnover as indicated by levels of CTX (331). Even though early animal models displayed inverse relationships with leptin concentrations and BMD, reporting predominantly antiosteogenic pathways through the central hypothalamus as a result of intracerebroventricular leptin infusions (284, 332), such pathways are not activated in humans (42, 114). Although leptin levels have not been linked to estrogens in healthy individuals with sufficient leptin levels (333), leptin has an intricate relationship with the regulation of sex steroid hormone production through its paramount influence on the HPG and other axes, particularly when circulating leptin levels are low as in cases characterized by acute or chronic energy deprivation (334, 335).

In vitro, leptin directly activates osteoblasts, chondrocytes, and bone marrow stromal cells and promotes osteoblast differentiation (49). Leptin-deficient mice present with increased trabecular volume and BMD in axial bones, but diminished length and mineralization, along with increased marrow adiposity, in the peripheral skeleton (322, 336). Intracerebroventricular leptin administration in ob/ob mice inhibits bone formation and stimulates bone resorption (284) to reduce vertebral trabecular bone volume but increase femur length and bone volume, normalizing the bone phenotype (322). In several animal models, leptin administration exerts positive effects on bone mass except at markedly elevated doses, which may decrease body weight (49, 322, 337).

A retrospective analysis of 1193 Japanese postmenopausal women found that low leptin concentrations were associated with a higher risk of long-bone osteoporotic fractures (hazard ratio 0.70) (48), confirming the results of an older report in men and women (51). Leptin replacement in leptin-deficient individuals may improve bone health directly and indirectly via restoration of the HPG axis, increased IGF-1 activity, and decreased cortisol concentrations (46). In a randomized controlled trial of 19 women with HA, 36 weeks of r-metHuLeptin treatment upregulated osteocalcin levels, a marker of bone formation, and prevented an increase in urinary N-terminal telopeptide-to-creatinine ratio, a marker of bone resorption, compared with placebo (45). Lumbar BMC and BMD improved with r-metHuLeptin treatment (114). The extension of treatment in 6 subjects for 1 additional year increased BMC by 1.4% to 6.5% and BMD by 2.2% to 10.8% from baseline at the lumbar spine and maintained low levels of CTX, another marker of bone resorption (114). These r-metHuLeptin-treated women with HA also had increases in estradiol concentrations and resumed menses, which would also contribute to decreased bone resorption (114). In another randomized placebo-controlled trial, leptin administration in women with hypothalamic amenorrhea for 36 weeks did not influence circulating sclerostin or FGF23 levels but markedly downregulated intact PTH and tended to downregulate serum RANKL and increase OPG, while also significantly decreasing the RANKL-to-OPG ratio (330), indicating reduced bone resorption and osteoclastic activity, which are beneficial to bone integrity.

The AFI Axis

Leptin significantly influences the energy-consuming process of reproduction, as attested by the downregulation of the HPG axis in energy-deprived hypoleptinemic states, leading to HA and infertility (43). However, only 60% of patients with HA improve their reproductive function with leptin replacement, suggesting the presence of additional regulators. We have proposed that the AFI axis may have a role in reproduction, as well as muscle mass and metabolic processes, in energy-deficiency states, that is independent of leptin (43, 338) (Fig. 2).

Activins and inhibins are members of the TGF-β superfamily, chiefly functioning through SMAD signaling. They are implicated in reproduction, muscle tissue growth and differentiation, and energy metabolism (339). Under normal circumstances, activins and inhibins bind to their receptors on pituitary, ovarian, and testicular cells. Principally, inhibins suppress whereas activins upregulate FSH production from the anterior pituitary, thus maintaining an equilibrium (339-341). Activins also induce muscle atrophy and limit muscle mass alongside myostatin (53, 342).

Follistatin and follistatin-like 3 (FSTL3), which increase with LEA, suppress the bioactivity of activins by irreversibly binding to them (339, 340). By neutralizing activins’ role in reproduction, follistatins shunt energy away from the reproductive system (43, 54, 340). Follistatins also block the catabolic function of activins and myostatin to preserve muscle mass (339, 341).

Follistatins also mediate liver metabolic processes, lipid homeostasis, and glucose metabolism (54). FSTL3 knockout mice have reduced visceral fat, decreased insulin resistance, increased pancreatic islet number and size, and improved glucose tolerance (343). In humans, follistatin concentrations are inversely correlated with insulin sensitivity and positively associated with body mass index, fat mass, lipid profiles, and blood pressure (43, 53, 54, 111).

Activin A and B concentrations are elevated in physically active individuals (without known LEA), and exercise tends to increase the levels of activins, but also follistatin, which acts as a “binding protein” for activins and FSTL3 (53). In contrast, acute energy deprivation by fasting in healthy males results in decreased activin A and increased follistatin concentrations (37). In the context of chronic energy deficiency, women with HA also have higher follistatin levels and lower circulating levels of activins but lower FSTL3 levels compared with healthy women (43). These changes would lead to suppression of reproduction, preservation of muscle mass, and insulin resistance, mainly at the level of the liver, which would lead to higher availability of circulating levels of glucose and lipids (ie, energy sources). Importantly, 72-hour crossover experiments in lean men and women, as well as long-term placebo-controlled leptin replacement in women with hypothalamic amenorrhea attaining resumption of menses and reductions in body fat, have shown that LEA-induced alterations of the AFI axis are not influenced by leptin administration and thus mostly function independently of leptin (53, 54, 111, 338, 344) (Fig. 2).

Management of REDs

The management of REDs should be multidisciplinary and undertaken by a team of sports and exercise professionals, including nutritionists, physicians, psychologists, psychiatrists, physiotherapists, and physiologists. However, REDs is often misdiagnosed by health professionals, coaches, and athletes, with studies demonstrating insufficient education, poor management, and prioritization of performance over health (345-347). Initial recommendations start with specialized dietary plans, nutritional education, and supplementation, followed by mitigation of exercise, before proceeding with specialized hormonal or pharmaceutical treatments.

Nonpharmacological Strategies

Nonpharmacological therapy is the preferred mainstay strategy, which frequently results in the successful resolution of most cases. The treatment of REDs should involve a multifaceted approach, taking into account the complexity and individual variability in each athlete's nutritional needs and recovery processes. Therefore, it is crucial to recognize the significance of addressing the underlying cause of REDs rather than solely focusing on treating its symptoms. Current consensus statements and practice guidelines (6, 14, 19, 35) emphasize nutritional, behavioral, and lifestyle changes, including an appropriate exercise regimen and the importance of increasing dietary energy intake and/or reducing exercise energy expenditure to improve EA for the prevention and treatment of the clinical consequences of LEA and therefore of the Female and Male Athlete Triad and REDs.

The IOC panel of experts also highlights the need for educational programs on nutritional options as well as the risks and consequences in health from inadequate dietary patterns (26). Besides, they describe the REDs clinical treatment as based on realistic health-promoting goals for weight and body composition (26). Patients are usually advised to increase their daily calorie intake by 300 to 600 kcal (26) or to set energy intake at a minimum of 2000 kcal/day, depending on exercise energy expenditure (14), but there is no established protocol for gradually increasing calorie consumption (26). In addition, it is suggested to restore body weight at levels that are correlated with the resumption of regular menstrual cycles (14). It has been observed that resumption of menses may occur after achieving approximately 90% of the ideal body weight for age and height (348). However, the timeframe for recovery typically spans several months, although it can vary significantly and may even extend to years in some cases (348, 349).

Moreover, it is crucial to emphasize that the relationship between body weight gain and the restoration of menstrual function is not a linear process, and additional factors may play a role (350). Recovery can be complex, particularly in the context of REDs, as intense physical activity may lead to the persistence of amenorrhea (6, 30). Therefore, it is essential to acknowledge both the complexity and individual variability in the resumption of the menstrual cycle while emphasizing the importance of personalized goals and targets in clinical practice to facilitate menstrual function recovery. This highlights that achieving a specific weight is not a universal solution, and a “1-size-fits-all” approach may not be appropriate and adjustments may be required over time.

Besides observational data (351), recently, the first randomized controlled trial (ie, REFUEL study) showed that increased energy intake 20% to 40% above baseline energy requirements was associated with an improved menstrual function in exercising women with oligomenorrhoea or amenorrhoea (352). It is also important to focus on diet quality, ensuring a proper balance of macronutrients and appropriate consumption of micronutrients (such as iron, calcium, and vitamin D) (14). In particular, vitamin D and calcium supplementation, as needed, may promote bone health and prevent SF with a potential benefit during the recovery phase (127, 296, 353-356), whereas an optimal daily intake of 1000 to 1300 mg of calcium and vitamin D concentrations within the range of 32 to 50 ng/mL are recommended (14). Additional factors such as the diversity of food choices, individual taste, and the practicality of food availability (practical aspects of food availability) should be considered while designing a meal plan (14). As a result, potential challenges of correcting the underlying LEA by altering energy intake and/or exercise energy expenditure in individual athletes or soldiers (from difficulties in changing the antecedents that underpin LEA or noncompliance) should also be acknowledged. Finally, the European College of Sport Science and the American College of Sports Medicine consensus also highlight resting, adequate sleeping, adjusting dietary patterns, and decreasing training as important treatment factors (357).

Regarding eating disorders, treatments for medically stable and unstable patients differ but also require multidisciplinary treatment. Although cognitive behavioral therapy is a first-line treatment option, each patient evaluation should consider the need for an inpatient hospitalization, outpatient care, or day programs based on their stability, associated complications, and comorbidities (358). Specifically, for anorexia nervosa, it can be attempted to be treated with cognitive behavioral therapy, which has been expanded to females with HA and shown to normalize leptin, TSH, and cortisol concentrations (probably because body weight and fat mass increased) (353, 359). More precisely, its first-line management should include weight restoration with nutritional rehabilitation. It should include restoring a structured meal routine, supervised meals, and psychotherapy to address dysfunctional thoughts and behaviors (358, 360). Of note, nutritional repletion might restore several related endocrine abnormalities but is not commonly sufficient by itself (361, 362).

Several ongoing (Table 2) and recently completed trials have investigated risk assessment for REDs, as well as pharmacological and nonpharmacological lifestyle modification strategies for REDs and related disorders. One study enrolled 481 female college athletes who were assigned to a comprehensive lifestyle modification termed the Female Athlete Body Project, a behavioral eating disorder risk factor reduction program, or a control arm, and found benefit in body perception-related outcomes through the Female Athlete Body Project (363). Under the same principles, the Male Athlete Body Project has also been investigated up to 1 month follow-up (364); yet, these studies chiefly pertain to DE and do not specifically address outcomes on endocrine REDs markers. Another more recent study implementing a food and nutrition learning program intervention of online lectures applied to a multinational cohort alongside a control intervention, demonstrated improvements in REDs symptoms, particularly menstrual function and, to a lesser extent, gastrointestinal issues, in athletes who underwent the FUEL intervention, with these benefits persisting up to 12 months after intervention (365). Another study in healthy females under low energy availability led to decreased P1NP levels indicative of bone resorption, which were prevented through the incorporation of high-impact jumping, indicated by stable β-CTx levels, suggesting a protective effect on bone health (366). On the other hand, in male cyclists, 4 weeks of intensive endurance interval training was shown to improve aerobic performance and testosterone but produced reductions in RMR, T3, and increases in cortisol, all of which constituted risk markers of REDs (367).

Table 2.

Ongoing studies, as of November 2023, investigating the pathophysiology and treatment of LEA-, REDs-, and the Female Athlete Triad-associated conditions in athletic, training, healthy, and energy-deficient populations

Registered trials investigating prevalence, risk or mechanisms, and characteristics of LEA, REDs, and the Female Athlete Triad
Identifier	Design	Participants	Arms	Outcomes	Time frame
NCT05649267	Prospective/cross-sectional	10 Judoka athletes, aged 18-30 y, training ≥4 times/week for at least 2 h	Year 1—Energy deficit, dietary risk Years 2 and 3—effect of carbohydrates and protein on function and performance	Body weight and composition including BMD Food intake	3 y
NCT04821076	Prospective RCT	Females, aged 18-30 y, BMI 18.5-30, normally menstruating: 30 trained 30 recreationally active	Trained population: Low-calorie diet + exercise Energy-balanced diet + exercise Recreationally active population: Low-calorie diet Energy-balanced diet	Changes in muscle protein synthesis rate Muscle and adipose tissue biopsies, anthropometrics, ergometric assessment, hormones, inflammatory markers, cardiovascular indices	10 d
NCT02858336	Crossover	150 females ages 18-28 y, BMI 18.5-27, normal menstruation	Neutral energy availability (45 kcal/kg lean body mass/d) Deficient energy availability (20 kcal/kg lean body mass/d)	LH pulse frequency Levels of hormones, adipokines, assessment of glucose homeostasis	5 d
NCT04910724	Prospective	Individuals, ages 18-35 y, BMI < 30, stable weight for 2 months	Energy deficit resulting from a combination of exercise and diet: Energy deficit (20% of daily requirements) Energy deficit (40% of daily requirements) Energy deficit (60% of daily requirements)	Whole body protein balance (leucine isotope) Muscle protein synthesis (phenylalanine isotope) Carbohydrate oxidation (glucose isotope)	2 d
NCT04900701	Prospective	36 individuals, aged 35-65 y	Hypocaloric energy status Energy balance Hypercaloric energy status	Muscle protein fractional synthetic rate Muscle morphology	6 d
NCT05259969	Prospective cohort	25 competitive male athletes aged 18-40 y	Observational study examining a series of markers across the sport seasons to assess the risk of LEA in male adult athletes. Preparation phase (peak exercise workload) Competition phase transition phase (lowest exercise workload)	Energy intake, expenditure and fat-free mass Resting metabolic rate, body mass and composition, iron status markers, thyroid markers, testosterone, cortisol, LEAM-Q, eating attitude and disorder questionnaires, inflammation markers, ghrelin, GH, gut microbiota, metabolites and performance markers, lipid profiles, leptin concentrations, IGF-1	1 y, (visits on baseline, 6 mo, 12 mo)
NCT03593382	Cross-sectional	Female endurance athlete exercising at least 5 times/week, 18-39 y	Energy availability	Resting energy metabolism Secondary outcomes include: work efficiency, substrate utilization, bone mineral content and density, hormonal measurements, lipid profile, vital signs, dietary and activity questionnaires	6 wk to 6 mo after inclusion
NCT04254900	Prospective	Athletes from a national team in wheelchair sports, 18-60 y	Energy availability evaluation	Energy intake, exercise energy expenditure and fat-free mass and using the following formula: energy availability = (energy intake – exercise energy expenditure)/fat-free mass)	7 consecutive days during preseason
NCT05587270	Prospective	Swedish climbers on elite or subelite levels of competition, >13 y	Web-based surveys and DXA	Eating disorders based on the Eating Disorders Examination-Questionnaire Basic symptoms indicative of REDs, questionnaires on meals, menstruation, injuries Sleep quality; depression, anxiety, and stress assessment; body image; compulsive exercise and perfectionism behaviors; overuse injuries	3 y

Registered interventional trials investigating treatments for energy-deficiency related disorders in healthy and unhealthy leanness, athletes, and soldiers
Identifier	Design	Phase	Participants	Arms	Outcomes	Time frame
Behavioral modifications and dietary supplementation
NCT04748250	Prospective RCT	N/A	60 female adolescent gymnasts with diagnosed Female Athlete Triad, particularly osteoporosis	Acupuncture Soy phytoestrogens Acupuncture + soy phytoestrogens	Bone mineral density—DXA	3 mo
NCT05390346	Cross-sectional, prospective	N/A	100 male or female college athletes	Evaluation of subjects’ energy availability and risk stratification Intervention through nutrition education	Energy availability REDs-CAT risk scores, improvements in energy availability after nutrition education	Up to 21 mo
NCT04766203	Prospective nonrandomized	N/A	2000 athletes or para-athletes >15 y	REDs screening REDs screening (blood draw) + follow-up REDs screening (DXA scan, RMR and exercise testing) + follow-up Treatment of REDs using a holistic treatment arm through individualized nutrition plans and counselling	REDs CAT, menstrual function, sex hormone levels, biochemical blood tests, RMR and ergometric capacities, IGF-1, BMD, cardiovascular parameters Anthropometrics, complete blood counts, glucose homeostasis, hormones, vitamin D, questionnaires	6 mo Treatment arm: 3-6 mo
NCT04823156	Prospective RCT	N/A	45 females, ages 18-36 y, BMI 18-30, stable weight over 2 mo	Daily calcium + vitamin D supplementation LEA (45 kcal/kg lean body mass/d for 7 days proceeding to 15 kcal/kg lean body mass/d thereafter) + vitamin D LEA + calcium + vitamin D	Urine calcium isotopes At rest and during load carriage: calcium isotopes, P1NP, calcium metabolism, reproductive and metabolic hormones, muscle strength, anthropometrics, BMD	2 wk
NCT03963128	Prospective RCT	N/A	4450 male soldiers, ages 16-32 y	Vitamin D3 supplementation (800 IU/d) Placebo	SF risk reduction Skin, soft tissue, and respiratory infection, vitamin D and hormone levels, bone turnover markers	32 wk
NCT05341700	Cross-sectional	N/A	15 females, ages 18-30 y, with regular menstrual cycles, exercising at least 5 days per week	Impact loading exercises No impact load exercises	Changes in osteocalcin, sclerostin, CTX, PTH, estrogen, IGF-1, hepcidin, insulin, thyroid hormones, 24-h glucose patterns Running economy—oxygen consumption, change in body weight	5 d
NCT05709678	Randomized	N/A	Physically active (training, competition-level) adults aged 18-40 y	General nutrition and sport nutrition with athlete testimonials on impacts of REDs on health Control group	Platform to Evaluate Athlete Knowledge of Sports Nutrition Questionnaire (Peak-Nq) Carbohydrate intake Food frequency questionnaire	3 wk
NCT05589077	Interventional	N/A	Females aged 18-25 y, that participate in organized sport	Behavioral: Education on Female Athlete Nutritional Concerns	Change in Knowledge, Confidence, and Impact Scores	Cross-sectional
NCT05709639	Prospective	N/A	8 tier 4 and 5 elite Olympic and paralympic athletes, aged ≥ 18 years old	Individualized counselling in nutrition	1) Nutrition knowledge 2) Change in carbohydrate intake	12 wk
NCT06116097	Interventional	N/A	100 competitive female endurance athletes aged 14-18 y training minimum 6 h weekly (not taking a break from sports for more than 3 mo)	Experimental: Nutrition education (behavioral) intervention group No intervention: Control group	Energy availability LEAF-Q Sports nutrition knowledge questionnaire (SNKQ), Eating Attitude Test, Dietary intake, Energy Expenditure, BMR, body composition	6 mo
NCT06220240	Interventional	N/A	Professional female basketball players aged 18-35 y with a BMI 18.5-25, energy deficient.	Eccentric muscle contraction modified training program Routine training program	Energy intake and availability, RED-S CAT, biomechanics, running speed, activity Hormones (gonadal, thyroid, adrenal)	2 wk
Pharmacological options for BMD loss in soldiers and athletes
NCT04196855	Prospective RCT	3	136 individuals undergoing military training, aged 18-40 y, with lower limb SF	Teriparatide 20 μg/d No intervention, standard SF care	Radiological healing of SF Time to healing, rehabilitation, pain quantification, quality of life, adverse effects, P1NP, CTX	8, 10, 12, 14, 16, 20, 24 wk
NCT04589819	Prospective RCT	4	183 soldiers, diagnosed with tibial SF	Teriparatide 20μg/d Placebo	Time to return to full activity Long-term side effects, injury recurrence, effects on bone stress injuries	3 y
NCT05382026	Prospective RCT	N/A	114 adolescent athletes aged 12-17 y, having reached menarche and performing resistance training 3x/wk	Chocolate milk supplement Pea-based beverage supplement Placebo supplement	Changes in lean mass (DXA) Changes in fat mass, lumbar BMD, hip BMD, bench press strength, squat strength	6 mo

Abbreviations: BMD, bone mineral density; BMI, body mass index; BMR, basal metabolic rate; DXA, dual-energy X-ray absorptiometry; LEA, low energy availability; LEAF-Q, low energy availability in females questionnaire; N/A, not applicable; P1NP, pro-peptide of type 1 collagen; RCT, randomized controlled trial; REDs CAT, relative energy deficiency in sport clinical assessment tool; RMR, resting metabolic rate; SF, stress fracture; β-CTX, β-carboxyl-terminal cross-linked telopeptide of type 1 collagen.

Pharmacological Strategies

Targeting underlying causes of LEA incorporating nutritional, behavioral, and lifestyle alterations is the front-line treatment approach. However, pharmacological management could be considered, especially in specific cases regarding the treatment of problematic LEA-related symptoms. The nonpharmacological approach should continue along with the initiation of pharmacological treatment.

Despite recommendations for treatment of FHA with estrogen with or without calcium and vitamin D, a recent meta-analysis on estradiol treatment in premenopausal women with FHA and low BMD found no significant improvements in lumbar BMD overall, though there may be a benefit with transdermal estradiol (309). A recent randomized controlled study in young, amenorrheic female athletes demonstrated robust improvements in spine, femoral neck, and hip BMD with 12 months of transdermal estradiol treatment, compared with oral estradiol or no treatment (368). Unlike transdermal estrogen, oral formulations of estrogen can decrease IGF-1 activity (369), reduce free testosterone concentrations (by increasing SHBG levels), and as a result, the potential anabolic testosterone-induced effects on bone (370, 371). The Endocrine Society suggests against using oral contraceptives to improve BMD if the underlying energy deficit is not addressed (78). However, the 2014 Female Athlete Triad Coalition recommended considering transdermal estradiol if BMD Z-scores are ≤−2.0, there are ongoing risk factors, and an adequate response has not been achieved after 1 year of lifestyle modification (ie, BMD loss or new fracture) (14).

PTH analogs are used for the treatment of osteoporosis and are extremely effective at reducing vertebral fractures. Teriparatide, an injectable synthetic recombinant human parathyroid hormone analog, has been shown to markedly increase BMD in women with anorexia nervosa after 6 months of treatment (372) and is stipulated by the Endocrine Society's guidelines for HA-associated delayed fracture healing and low BMD (78). Two ongoing clinical trials (Table 2) are currently investigating its use in military populations to prevent SF.

Romosozumab, an anti-sclerostin antibody, has proven fracture reduction benefits in postmenopausal women with osteoporosis and improve bone profiles in men, albeit with some potential cardiovascular side effects, chiefly myocardial infarction and stroke (373). Currently, there is only 1 small trial examining BMD changes in premenopausal women with idiopathic osteoporosis (NCT04800367). Denosumab has also been studied for 12 and 24 months in 32 women with premenopausal idiopathic osteoporosis following a treatment course of 24-month teriparatide, attaining marked increases in lumbar, hip, and femoral BMD (374). These medications have yet to be studied in energy-deficient cohorts of athletes or soldiers. Moreover, it should be noted that the Endocrine Society does not recommend bisphosphonates, denosumab, testosterone, or leptin for bone health based on limited evidence, and these therapies are only reserved for individuals who have failed or have contraindications to estrogen replacement. Recombinant parathyroid hormone is suggested in cases of delayed fracture healing and markedly low BMD (78).

Given the reduced GH activity associated with LEA, GH and IGF-1 therapies have been tried in women with anorexia nervosa. Although treatment with GH does not seem effective because of GH resistance, recombinant IGF-1 treatment with oral contraceptives has been shown to modestly increase spine BMD in women with anorexia (195, 375). Growth hormone secretagogues are currently being investigated, but not in cohorts with REDs. LUM201, also known as ibutamoren or MK-677, is a GH secretagogue currently being investigated for pediatric GH deficiency (NCT04614337). MK-677 has been shown to increase serum levels of GH, IGF-1, and IGFBP-3, as well as lean body mass in a 2-month study in obese patients (376). MK-677 additionally increased both markers of bone formation and resorption, elevating CTX by 23% and procollagen III by 28% in early stages of treatment and osteocalcin at later stages (377). A randomized controlled trial in 165 older adults with hip fractures found marked increases in IGF-1 levels but no improvements in functional or performance metrics (378). In another trial of elderly patients with hip fracture, MK-677 was found to improve gait speed but not other performance measures and did not significantly reduce falls (379). However, it is important to clarify that GH, IGF-1, and GH secretagogues are considered forms of doping and therefore are prohibited for use by competitive athletes.

Bimagrumab, an antibody blocking the activin type II receptor, has been shown to increase lean body mass in older men and women with sarcopenia and in older patients with recent hip fractures (380, 381). However, no improvements were seen in physical performance. Another randomized controlled trial in adults with type 2 diabetes and obesity found that bimagrumab over 48 weeks decreased fat mass by a mean of 20.5% (7.5 kg), waist circumference, and hemoglobin A1c (382). Whether bimagrumab may have a role in maintaining muscle mass in lean, energy-deficient cohorts remains to be seen.

On the basis of improving performance, testosterone administration has been recently assessed in a controlled study of 48 young female athletes. Therein, daily application of testosterone cream led to marked enhancements in aerobic capacity and total lean mass, but not anaerobic performance or muscle strength. Although this trial assessed circulating testosterone and performance-specific characteristics, it did not specifically address potential improvements in REDs-related markers, warranting further research (383).

According to findings from studies on college athletes, weight gain was identified as the most important indicator of normal menstrual function restoration (351, 384). The severity of energy deficiency and the duration of the menstrual disturbances affect the timeline of the menstrual cycle resumption (385). Although not REDs-specific, according to the 2017 Endocrine Society Clinical Practice Guidelines, the administration of GnRH analogs is considered the first line of treatment for patients with FHA who wish to conceive, followed by gonadotropin therapy and ovulation induction with clomiphene citrate (78). Short-term use of transdermal estradiol with cyclic oral progesterone is suggested for those who have not resumed menses after 6 to 12 months of lifestyle modifications, but all these have a limited effect in normalizing only a small part of the big picture (78).

Leptin as a Potential Approach for REDs

Current pharmacological options for the Female Athlete Triad and REDs fail to address the complete physiological picture. GnRH analogs, estradiol, bone active agents, and GH analogs, as outlined previously, only contend with the endocrine changes downstream of leptin. Based on available data regarding the efficacy of leptin in reversing FHA and remedying LEA-induced neuroendocrine perturbations in experimental and small clinical settings, leptin could hypothetically constitute an effective candidate treatment for the broader family of REDs-related disorders, especially in individuals with clinically low concentrations of leptin. Thus, from a pharmacological perspective, administering leptin, which is suppressed under LEA conditions (113, 115, 116, 386, 387), could potentially serve as a complementary treatment strategy that could address problematic LEA-related pathophysiological alterations (eg, FHA, compromised bone health). However, leptin replacement therapy is currently not recommended by consensus guidelines for individuals with hypothalamic amenorrhea or during low energy availability because its safety and effectiveness require further investigation for confirmation.

Leptin administration in energy-deficient individuals overrides the energy conservation milieu that hinders the HPG axis in REDs (36, 45, 46), reversing gonadotropin pulsatility in males (42) and inducing ovulation, restoring menstruation, and thus resolving HA in lean energy-deficient females as discussed previously (36, 45-47). Moreover, leptin replacement in energy-deprived amenorrheic women increased osteocalcin levels, decreased CTX levels, and increased lumbar BMC and BMD (45, 114) while also decreasing intact PTH levels and tending to diminish the RANKL/OPG ratio indicative of osteoclast differentiation and activity (330). Notwithstanding favorable results, because no large clinical trials beyond our proof-of-concept studies have been performed, the safety and efficacy of leptin have not been adequately evaluated and thus leptin cannot be recommended at this time as a definitive treatment for HA. A major physiological caveat to the use of leptin in hypoleptinemic individuals is weight loss from overtreatment, which can lead to supraphysiological circulating leptin levels (62, 388). Although leptin treatment reduces caloric intake when administered in supraphysiological levels in lean individuals, it does not affect resting metabolic rate or physical activity, nor does it prevent the metabolomic shift from carbohydrate to fatty acid metabolism (285). At this time, it is unknown if antibodies with any neutralizing effects may develop in subjects with REDs in response to therapy with anti-r-metHuLeptin or leptin analogues currently in development (388, 389). Large, randomized phase 3 clinical trials investigating the effects of leptin in women and men with energy-deficiency-related disorders are warranted. Thus far, r-metHuLeptin has only been approved for patients with CLD or congenital or acquired generalized lipodystrophy (390, 391).

In contrast to its effects to normalize immune and neuroendocrine function when leptin administration leads to normalization of the low leptin levels seen in energy deprivation states, leptin might exert anorexigenic properties and therefore suppress appetite when administered in supraphysiological doses and only when circulating levels exceed the upper physiological range (392-394). Moreover, according to studies conducted on women with FHA, recombinant leptin administration has been associated with significant weight loss and fat mass reduction when doses exceeded the upper limit of normal (45, 46, 114). Therefore, as previously highlighted, leptin is not currently recommended for treating REDs-related health consequences because randomized clinical trials on leptin administration with doses within the normal range (ie, beyond the proof of concept studies discussed previously) have not been performed. The potential role of leptin in specific athletic and military populations experiencing problematic LEA as a pharmacological approach should be explored and confirmed in larger clinical trials involving both females and males with problematic LEA and REDs-related health consequences.

Conclusions

REDs is a frequently overlooked constellation of disorders stemming from LEA in diverse exercising populations of either biological sex, resulting in various physiological consequences and impaired health and overall performance. Addressing the underlying causes of LEA by incorporating nutritional, behavioral, and lifestyle modifications is currently the primary treatment approach.

As the principal arbiter of energy homeostasis, leptin may constitute a key neuroendocrine conduit for several REDs-related disturbances. The AFI system also seems to contribute to the pathophysiology of REDs. Decades of investigation have offered valuable insight into the potential therapeutic implications of leptin and the AFI axis, which exert a wide array of pro-metabolic effects (37, 254), including direct and indirect osteoprotective mechanisms (49).

Furthermore, the practicality and feasibility of routinely measuring leptin levels or AFI in clinical practice for REDs remains a significant consideration. However, it is important to recognize that REDs is a multifaceted disorder involving various physiological and psychological factors, underscoring the need for a more comprehensive understanding of REDs, which includes exploring other contributing factors. This has paved the way for new observational studies, prospective cohorts, and randomized controlled trials in athletes and military personnel, which should aim to encompass a broader range of factors impacting REDs for a more holistic approach to treatment and understanding.

Additional Information

Review criteria: The authors consulted the relevant position statements of the American College of Sports Medicine and the IOC. Each author queried PubMed, Clinicaltrials.gov, trial registries accepted by the International Committee of Medical Journal Editors, and Google Scholar using the terms “energy deficiency,” “RED-S,” “REDs,” “female athlete triad,” “anorexia,” “low energy availability,” “hypothalamic amenorrhea,” “bone,” “stress fractures,” “leptin,” “activin,” “follistatin,” “adipose,” “soldier,” and “military.” The authors also reviewed relevant references cited within retrieved articles. Cumulative results were presented as summaries in tabular format, to be screened, assessed, and selected for inclusion in the current manuscript based on topic relevance, conciseness, and study quality by the authors. Of note, for the completeness of this review, the authors searched the literature for both Triad and REDs syndromes. The use of each term inside the manuscript depends on the findings of the original paper.

Author Contributions

C.S.M. conceptualized the review. K.S. and C.S.M. performed the initial search and wrote the first draft. A.M.A. performed extensive review of the first draft and A.M.A. and K.S. wrote the manuscript. All coauthors (A.M.A., K.S., S.H.C., L.V.V., D.G.G., K.D., K.N., C.B., P.T., A.K., H.A.P., and C.S.M.) performed literature searches, wrote sections of the text, contributed to the discussion of the content, edited, corrected, extended, and reappraised all versions of the manuscript. All authors have approved the final version of the submitted manuscript.

Disclosures

No funding sources or conflicts of interest pertaining to the present work were reported from the authors. This review is not funded, and none of the authors have received any funding or nonfinancial support relevant to its completion. Other unrelated disclosures are as follows: A.K. reports grants and advisory services for Novo Nordisk, advisory services for Eli Lilly, Bausch Health, Sanofi-Aventis, MSD, AstraZeneca, Elpen Pharmaceuticals, Boehringer-Ingelheim, Galenica, Ethicon, and Epsilon Health. None is related to the work presented herein. C.S.M. reports grants through his institution from Merck, Massachusetts Life Sciences Center, and Boehringer Ingelheim, has been a shareholder of and has received grants through his institution and personal consulting fees from Coherus Inc. and AltrixBio; he reports personal consulting fees and support with research reagents from Ansh Inc., collaborative research support from LabCorp Inc., reports personal consulting fees from Genfit, Lumos, Amgen, Corcept, Aligos, Intercept, 89 Bio, Madrigal, and Regeneron, reports travel support and fees from TMIOA, Elsevier, and the Cardio Metabolic Health Conference. None is related to the work presented herein. The other authors have nothing to disclose.

Abbreviations

 
• AFI

  activin-follistatin-inhibin

 
• AT

  adipose tissue

 
• BAT

  brown adipose tissue

 
• BM

  bone marrow

 
• BMAT

  bone marrow adipose tissue

 
• BMC

  bone mineral content

 
• BMD

  bone mineral density

 
• BMI

  body mass index

 
• CLD

  congenital leptin deficiency

 
• DE

  disordered eating

 
• EA

  energy availability

 
• EB

  energy balance

 
• EEE

  exercise energy expenditure

 
• EI

  energy intake

 
• FFM

  fat-free mass

 
• FHA

  functional hypothalamic amenorrhea

 
• FSTL3

  follistatin and follistatin-like 3

 
• HA

  hypothalamic amenorrhea

 
• HPA

  hypothalamic-pituitary-adrenal

 
• HPG

  hypothalamic-pituitary-gonadal

 
• HRV

  heart rate variability

 
• IOC

  International Olympic Committee

 
• JAK

  Janus kinase

 
• LEA

  low energy availability

 
• OPG

  osteoprotegerin

 
• P1NP

  pro-peptide of type 1 collagen

 
• r-metHuLeptin

  recombinant methionyl human leptin

 
• RANKL

  receptor-activator of nuclear factor κΒ ligand

 
• REDs

  Relative Energy Deficiency in Sport

 
• RMR

  resting metabolic rate

 
• SF

  stress fracture

 
• STAT

  signal transducer and activator of transcription

 
• TEE

  total energy expenditure

 
• WAT

  white adipose tissue

Author notes

Angeliki M. Angelidi and Konstantinos Stefanakis have contributed equally to this work and share co-first authorship.