Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Eating Disorders, Body Image, and Relative Energy Deficiency in Sport

Abstract. Athlete eating pathology, body-image evaluation, and relative energy deficiency in sport (REDs) intersect through weight-sensitive performance incentives, dietary restriction, training demands, and the physiological consequences of inadequate energy availability. This critical narrative research monograph synthesizes foundational experiments, original athlete studies, longitudinal and qualitative evidence, systematic reviews, and international clinical consensus. It distinguishes diagnosed eating disorders from disordered eating, body dissatisfaction from objectively insufficient fueling, and problematic low energy availability from the broader multi-system REDs syndrome. It examines causal pathways, epidemiologic measurement, athlete-specific screening, sex and developmental variation, bone and reproductive outcomes, multidisciplinary assessment and prevention, and the limitations of generalizing highly selected athletic samples. The synthesis identifies an evidence-based rationale for early, confidential, multidisciplinary recognition while emphasizing that clinical diagnosis and return-to-sport decisions require qualified professional assessment. Research priorities include prospective causal designs, inclusive measurement, and independent evaluation of sport-environment interventions.

Keywords: eating disorders in athletes; disordered eating; body image; body dissatisfaction; relative energy deficiency in sport; REDs; low energy availability; female athlete triad; male athlete triad; sport psychology; sport nutrition; athlete mental health; clinical screening; athlete safeguarding; eating disorder prevention

Eating Disorders, Body Image, and REDs as an Integrated Research Problem

The scientific object of this dossier is the conjunction of disordered eating, clinically defined eating disorders, bodily self-evaluation, and energy-deficiency-related morbidity in organized sport. These phenomena have overlapping causal pathways, yet they require separate operational definitions. An athlete may have significant body dissatisfaction without nutritional deficiency, may experience clinically consequential low energy availability without an eating disorder, or may meet criteria for an eating disorder while maintaining an outwardly successful athletic career. A defensible explanation therefore follows behavior, subjective experience, training exposure, biological adaptation, and clinical outcomes as distinguishable dimensions. The International Olympic Committee's recent mental-health consensus treats eating disorders and relative energy deficiency as connected clinical concerns while preserving their distinct definitions Reardon et al. (2026). The resulting research task is to explain when and how their relationships emerge, persist, intensify, and become reversible.

Athletic participation creates a distinctive observational setting because the body is simultaneously experienced, trained, evaluated, and publicly judged as a performance instrument. Within a single season, an athlete may receive feedback about strength, movement efficiency, appearance, selection, weight classification, and recovery. Some feedback is task-relevant; other feedback produces appearance surveillance or dietary restriction without demonstrated competitive benefit. Whether perceived pressure becomes harmful depends on the athlete's previous vulnerabilities, the social authority of evaluators, access to nutrition, recovery opportunities, and the consequences attached to body-related judgments. The qualitative synthesis by Fatt et al. (2024) identifies power imbalance, rigidity, identity, shame, and difficulties seeking help as recurring themes across athletes' accounts. These themes are hypotheses about institutional processes and personal experience rather than proof that any one coaching behavior universally causes a diagnosis.

Three explanatory levels must be studied together. At the psychological level, beliefs about achievement, weight, body shape, and self-control guide attention and choices. At the behavioral level, food restriction, compensatory practices, exercise behavior, and help-seeking change the athlete's actual exposures. At the physiological level, prolonged or severe inadequate energy availability can affect neuroendocrine, skeletal, reproductive, and other bodily systems. The 2023 IOC consensus explicitly places problematic low energy availability at the center of its REDs model and recognizes both psychological and physiological impairment Mountjoy et al. (2023). A relationship between two levels may be bidirectional: fatigue and mood changes can influence food-related cognition, while rigid food-related cognition can sustain the conditions that produce fatigue and mood changes.

This monograph is a critical narrative synthesis of published evidence. Its aim is to compare explanations and clarify the reach of the evidence, including experiments on energy availability, sport-specific clinical screening, cohort studies, qualitative research, preventive trials, systematic reviews, and contemporary expert consensus. It makes no claim to have performed original participant recruitment, a registered systematic review, or a meta-analysis. Its central questions concern definition, origins, causal mechanism, measurement, prevalence, diagnosis, intervention, and governance. Statements about clinical care are descriptions and evaluations of published professional guidance; an individual's diagnosis, medical stabilization, and return-to-sport decisions require qualified, context-sensitive assessment NICE (2017/2020) Williams et al. (2026).

Terminology, Diagnostic Boundaries, and Causal Structure

An eating disorder is a recognized psychiatric syndrome assessed using formal diagnostic criteria and a clinical evaluation. Relevant diagnoses include anorexia nervosa, bulimia nervosa, binge-eating disorder, and other specified feeding or eating disorders. Disordered eating denotes harmful or atypical eating-related attitudes and behaviors that may occur below a diagnostic threshold or outside a single named syndrome; its research meaning varies by questionnaire and study. Body image is the multidimensional organization of perceptions, evaluations, feelings, and behavior concerning one's body. Body dissatisfaction is a particular negative appraisal within that broader field. These terms cannot be collapsed into a single prevalence number or interpreted as interchangeable measures of psychopathology. The IOC's 2026 mental-health statement emphasizes precisely these clinical and measurement distinctions in athletic populations Reardon et al. (2026).

Energy availability is a sports-physiology quantity: energy consumed through diet minus exercise-related energy expenditure, divided by fat-free mass, commonly expressed per kilogram of fat-free mass per day. Low energy availability describes a mismatch in which energy remaining for normal bodily functions is insufficient relative to needs. It can arise through intentional dietary restriction, increased training load, inadvertent underfueling, illness, constrained food access, or combinations of factors. Problematic low energy availability describes exposure severe or sustained enough to produce clinically important disruption. The REDs syndrome is defined by resulting impairment of physiological and/or psychological functioning; it is not simply a low questionnaire score or a synonym for an eating disorder. The IOC's updated definitions explicitly differentiate the exposure from the syndrome and distinguish adaptable fluctuations from problematic exposures Mountjoy et al. (2023).

The concepts support a causal diagram with several entry points. Appearance pressure may alter body dissatisfaction, which can lead to restrictive behavior and subsequently to inadequate energy availability. Performance-motivated efforts to change power-to-weight ratio can create the same biological exposure without initial dissatisfaction. An athlete may also develop low availability because training expenditure rises faster than meal planning or food access adapts. The physiology of prolonged deficiency may then alter concentration, sleep, mood, injury vulnerability, and training response, which can change psychological coping and deepen a self-reinforcing cycle. Yet observational association between dietary behavior and a symptom does not establish the direction of causation; competing explanations include prior injury, illness, age, medication, total training load, and socioeconomic constraints Mountjoy et al. (2023) De Souza et al. (2026).

These distinctions matter for research populations. A study using a self-report eating-attitudes threshold measures screened symptoms, while a study using a structured interview can estimate clinical diagnoses according to its diagnostic framework. A study of bone stress injuries measures one consequence that can arise from several causes. A body-image scale measures subjective appraisals rather than caloric intake, energy availability, or endocrine status. Instruments gain meaning through validity evidence in specific athletes and sports, not through the familiarity of their names. The 2026 meta-analysis of disordered eating and eating disorders demonstrates how estimates vary across screening tools Nasser et al. (2026). Consequently this dossier evaluates conclusions according to what the underlying procedure actually identified, instead of treating all positive screens as confirmed cases.

Historical Emergence of Athlete Eating-Disorder Research

The research field grew from clinical interest in anorexia, bulimia, restrictive dietary practices, and menstrual disturbances toward systematic study of athletes whose competition and training conditions shaped symptom presentation. An early difficulty was conceptual: ordinary features of structured training, such as scheduled meals and commitment to exercise, could resemble questionnaire indicators of eating pathology when stripped of context. Conversely, apparently disciplined behavior could conceal clinical symptoms that athletes were unwilling to disclose. Sundgot-Borgen (1993) studied elite Norwegian female athletes using questionnaire and interview methods, helping establish why screening scores and clinical diagnoses should be examined separately. The importance of these early investigations lies in their explicit attempt to operationalize a sport-specific clinical problem and test whether broad assumptions about athletic protection or vulnerability were accurate.

The 2004 Norwegian comparison by Sundgot-Borgen & Torstveit (2004) combined self-reported screening and clinical interviews in elite athletes and a general-population comparison group. It became influential in demonstrating that the relationship between athletic status and eating disorders could be studied against an external comparator rather than inferred from athletes alone. Its methodological advantage also creates boundaries: the sample belongs to particular sports, national institutions, diagnostic conventions, and years of assessment. One cannot simply transfer its estimated prevalence to present-day mixed-gender recreational athletes, to para sport, or to every weight-class competition. Historical research is strongest when interpreted as a sequence of increasingly refined designs, not as one timeless baseline from which all contemporary prevalence figures must be derived.

A parallel physiological tradition investigated the consequences of exercise and dietary energy imbalance for reproductive regulation and skeletal health. In a controlled experimental study, Loucks & Thuma (2003) manipulated energy availability over five-day exposures in 29 regularly menstruating, habitually sedentary young women, then measured luteinizing hormone pulsatility. The experiment provided mechanistic evidence that short-term energy availability can affect reproductive endocrine signals under the studied conditions. Its population and duration are crucial limitations: it did not establish a universal diagnostic cutoff for all athletes, sexes, developmental stages, or durations of exposure. Subsequent sport-medicine frameworks increasingly interpreted menstrual dysfunction and bone outcomes as part of an energy-status continuum rather than as isolated athletic complications.

The clinical and social strands developed at different rates. Eating-disorder psychiatry refined diagnostic evaluation and psychological treatment, while sport physiology built evidence about endocrine regulation, bone health, and training exposure. Sport psychology expanded inquiry into perfectionism, body ideals, power relations, and help-seeking. Studies such as Martinsen et al. (2014a) eventually evaluated prevention interventions within elite sports schools instead of merely describing symptoms after they occurred. Recent qualitative syntheses identify stigma and institutional silence as relevant to the conditions under which athletes seek care Fatt et al. (2024). The historical trajectory thus moves from identifying high-risk groups toward studying modifiable systems: coaching practices, measurement procedures, adequate fueling, confidential pathways to assessment, and evidence-based decisions about continued participation.

From the Female Athlete Triad to the REDs Framework

The Female Athlete Triad linked three interrelated domains: inadequate energy availability, menstrual or reproductive disturbances, and reduced bone health. Its clinical contribution was to show that a deficit in nutritional energy available for body functions can produce consequences beyond visible weight change. The 2014 coalition consensus developed a spectrum model and risk-based guidance for screening, treatment, and return to play De Souza et al. (2014). Importantly, low availability could occur with or without overt eating-disorder symptoms. This distinction already challenged the assumption that identifying a psychiatric syndrome was necessary before examining physiological risk. It also placed medical decision-making within an interdisciplinary framework, since eating behavior, reproductive endocrinology, training, and skeletal health cannot be understood by one specialist or one test.

The IOC introduced the term Relative Energy Deficiency in Sport in 2014 to describe a broader constellation of impairments and explicitly include male athletes Mountjoy et al. (2014). Its 2018 update reviewed an expanding research base across health and performance systems Mountjoy et al. (2018). In the 2023 revision, the acronym REDs was adopted and the model more clearly separated low energy availability as an exposure from the syndrome of impairment associated with problematic exposure Mountjoy et al. (2023). This evolution has a theoretical consequence: a diagnostic framework must be able to evaluate energy deficit and multi-system outcomes independently of the individual's motives for eating and exercise. The psychological experience of pressure, fatigue, or fear may be clinically important, but it does not exhaust the physiological explanation.

The 2025 Female Athlete Triad Coalition update, published in a 2026 journal issue, further revised the energy-status continuum to recognize energy deficiency and metabolic compensation. The companion research review introduced a model specifically addressing adolescent athletes and incorporated bone stress injury more explicitly into skeletal assessment De Souza et al. (2026). Its clinical companion revised screening, assessment, treatment, and return-to-play recommendations Williams et al. (2026). These developments should be read alongside the REDs framework, since they share interest in energy-related reproductive and skeletal consequences while maintaining different scopes and clinical tools. Neither framework makes all body-image concerns equivalent to underfueling, and neither permits a single laboratory value or body-mass threshold to stand in for a complete clinical assessment.

The Male Athlete Triad consensus formalized an analogous concern for energy deficiency, suppression of the hypothalamic–pituitary–gonadal axis, and impaired bone health in adolescent and young adult men, especially in some endurance and weight-category contexts Nattiv et al. (2021). The associated clinical guidance addresses diagnosis, management, and return to participation Fredericson et al. (2021). These extensions reveal why the history cannot be reduced to a change in terminology: the population under study, endocrine indicators, vulnerability to missed diagnosis, and evidentiary assumptions all expanded. A coherent contemporary monograph must retain the specificity of reproductive pathways and life-course stages while recognizing that nutritional, psychological, and training exposures can produce consequences across sexes and types of sport.

Evidence Architecture and Methods of Scientific Synthesis

Evidence concerning athlete eating disorders and REDs is assembled from fundamentally different designs. Controlled physiology experiments test short-term responses to manipulated energy availability. Cross-sectional athlete surveys estimate symptom distributions but are vulnerable to selection, disclosure, and instrument effects. Diagnostic interview studies provide a closer link to clinical criteria while remaining constrained by case definition and sampling. Longitudinal cohorts can establish temporal ordering between exposure and outcome more convincingly than cross-sectional comparisons, although they still require careful handling of confounding. Qualitative interviews reveal meanings, institutional pressures, and barriers to care that biological markers alone cannot capture. Preventive trials examine whether altering an environment changes subsequent outcomes. Consensus statements organize clinical expertise and varied studies into guidance, but consensus cannot substitute for direct experimental evidence on every mechanism Mountjoy et al. (2023) Fatt et al. (2024).

The research synthesis follows a question-specific evidence hierarchy. To describe a mechanism, priority is given to directly manipulated exposure and measured physiological outcomes, with attention to ecological limits. To describe frequency, priority is given to clearly defined populations and validated assessment, with separate reporting of clinical diagnoses and positive screens. To describe athletes' experiences, qualitative work is assessed for sampling breadth, transparency, context, and thematic coherence. To evaluate prevention, controlled studies are considered for allocation procedure, baseline imbalance, cluster effects, follow-up, treatment fidelity, and generalizability. To evaluate management, authoritative clinical guidelines are compared against the evidence they cite and the population for which they were developed Martinsen et al. (2014a) NICE (2017/2020).

The review gives particular attention to measurement invariance and differential detection. A self-report scale developed in one language, sex, age group, or sport may perform differently in another; a muscularity-oriented concern may be missed by a thinness-oriented questionnaire. Intense training can be routine while a compulsion to exercise despite injury or severe distress may signal pathology; context is necessary to interpret behavioral frequency. Similarly, a missed menstrual cycle, low bone density, or stress fracture has a differential diagnosis beyond REDs. The design of a robust study must make its inclusion criteria, clinical verification, observed outcomes, exposure period, and comparison group explicit Martinsen et al. (2014b) Melin et al. (2014) Williams et al. (2026).

This is a structured critical narrative review conducted for a scholarly monograph on November 3, 2026. The selected literature includes original experiments, athlete interview and questionnaire studies, a school-based preventive randomized trial, quantitative and qualitative reviews, clinical guidance, and the 2023 and 2026 IOC consensuses. The source set is purposive, organized by independent research questions, and assessed against study design and relevance. It does not claim exhaustive database retrieval, duplicated independent screening, a registered protocol, formal risk-of-bias scoring of every eligible study, or pooled effect estimation. Those would be necessary features of a formal systematic review or new meta-analysis. The resulting conclusions are therefore calibrated to the directly reviewed evidence, with uncertainties preserved rather than masked by the scale of the narrative.

Athletic Body Image and the Meaning of the Sporting Body

Athletic body image is a set of perceptions, attitudes, affects, and self-evaluations directed toward the body in a setting where bodily capabilities are also measures of occupational or competitive competence. Researchers must distinguish the athlete's appreciation of functional capacity from appearance dissatisfaction, perceived mismatch with a sport ideal, anxiety about assessment, and self-worth invested in shape or weight. These dimensions can coexist. A competitor may appreciate strength and endurance while feeling intense dissatisfaction about visible muscularity or body fat; another may report a positive appearance image while following a nutritionally inadequate training regimen. Body image has no one-to-one correspondence with eating-disorder diagnosis or energy availability. The review by Burgon et al. (2023) found that adult athletes, on average, reported fewer body-image concerns than non-athlete comparison groups in the available literature, with important measurement and group-comparability limitations.

The aggregate finding is intellectually important because sport can expose athletes both to sources of corporeal competence and to unusual standards of appearance. Participation may improve perceived functionality, belonging, bodily efficacy, and familiarity with bodily variation. At the same time, selection in some disciplines is organized through visible physiques or body-mass categories, and the athlete's physique can be framed as an object requiring continuous improvement. Cross-sectional comparisons can conceal these different pathways. Athletes with stronger baseline body appreciation may self-select into sport, while those exposed to frequent evaluative comments may develop problems after sustained participation. Without measuring baseline attitudes and following athletes over time, lower average dissatisfaction among participants does not establish that sport protects every individual or that its protective effect applies to all competitive environments Burgon et al. (2023).

The meaning of body change also varies by task. A change in mass may affect weight-class eligibility, a change in muscular strength may alter performance demands, and changes during puberty may affect how a young athlete experiences selection. These are material dimensions of sport, but they do not logically require appearance-focused humiliation or inappropriate nutritional restriction. The central psychological question is the interpretation of the body as evidence of worth, control, discipline, or belonging. When self-evaluation becomes excessively dependent on a narrow body standard, otherwise ordinary performance feedback can acquire the power to destabilize identity or eating behavior. The qualitative themes of hyperfocus on body and food, rigid rules, and overvaluation of simplified performance metrics identified by Fatt et al. (2024) make such mechanisms empirically plausible, while their qualitative origin limits claims about prevalence or direct causality.

A comprehensive measurement program should capture appreciation of function, dissatisfaction about appearance, shape and weight concern, muscularity-oriented attitudes, fears about changing body composition, internalization of sport ideals, and the consequences of body-related evaluation. Measures should specify whether the athlete is responding as an everyday person or as a competitor under a sport-specific criterion. Researchers should also distinguish body dissatisfaction from self-reported confidence and from objective body composition; correlations among these variables require separate tests. The value of a multidimensional conception is preventive and explanatory: it allows a study to investigate protective contexts, risky assessment practices, and symptoms on their respective terms. It also avoids converting a group mean into a stereotype about the psychological character of any sport or athlete Burgon et al. (2023) Reardon et al. (2026).

Body Ideals, Social Comparison, and Gendered Presentation

Sports create local standards of an ideal body through coaching discourse, judging practices, teammate comparison, recruitment decisions, uniforms, weight classifications, and representations of successful performers. An ideal may privilege leanness, muscularity, compactness, or an appearance associated with a particular technical style. The same athlete can encounter several standards simultaneously, including ideals from teammates, family, mainstream culture, and digitally mediated sporting communities. Body evaluation becomes psychologically consequential when these standards acquire authority over participation, selection, and self-respect. Social comparison may provide information about technique and training while also encouraging excessive monitoring of shape or body weight. The reviewed literature supports treating sport-specific pressures as candidate exposures to measure, not as presumptive determinants of illness in all athletes Fatt et al. (2024) Reardon et al. (2026).

Gendered norms influence how symptoms are expressed and detected. Some screening instruments emphasize fear of weight gain and striving for thinness, which can capture important but incomplete patterns. In men and in athletes whose preferred physiques emphasize muscularity, a desire for greater muscle size or a tightly controlled leanness-to-muscularity ratio can structure body dissatisfaction differently. Eating pathology may present through rigid food rules, cycling of intake, compensatory exercise, or the persistent belief that training success demands a particular physique. The 2026 quantitative synthesis explicitly identifies the underrepresentation of muscularity-related concerns in popular screening tools such as the EAT-26 Nasser et al. (2026). Measurement can therefore under-detect certain presentations even when total survey coverage appears extensive. Clinical inquiry must follow the individual's behaviors and consequences rather than assume that one gendered aesthetic is universal.

The qualitative evidence documents interactions between authority and disclosure. Coaches and senior athletes may supply norms about body assessment, and sporting hierarchies can make challenging those norms costly. An athlete whose scholarship, team place, or reputation depends on meeting a body target may conceal restriction or dismiss fatigue. In the meta-synthesis of 38 qualitative studies, Fatt et al. (2024) identified themes including power imbalance, shame, fear, stigma, and conflicts between athlete identity and personal needs. Such findings are valuable for identifying mechanisms that standardized questionnaires may omit, particularly the meanings attached to perceived obligations. They also have limits: samples were concentrated in particular countries and predominantly female cohorts, and qualitative accounts cannot independently estimate how many athletes experience a specific pressure.

Researchers should specify the level of comparison when testing effects of body ideals. Between-sport comparisons may be confounded by age, competitive level, training volume, and the selection of athletes who enter each sport. Within-team comparisons can identify differences in feedback or coaching relationships while controlling some institutional factors, though they remain vulnerable to individual predispositions. Longitudinal within-person data can examine whether escalating weight surveillance precedes eating concerns or follows earlier symptoms. Interventions that reduce appearance-based evaluation should track both psychological outcomes and downstream nutritional and clinical consequences, without presuming that changes in attitudes automatically normalize energy availability. A coherent sport-environment account makes gender and sport type potential moderators of causal pathways and tests them rather than relying on broad categories alone Burgon et al. (2023) Martinsen et al. (2014a).

Weight Regulation, Performance Incentives, and the Sport Ecology

Weight and body composition enter sporting decision-making through distinct pathways. Weight-class events use categorical eligibility rules, endurance disciplines may reward aspects of physiological economy, aesthetic competitions may privilege visual conventions, and strength-oriented sports can emphasize particular dimensions of muscle development. These conditions influence what athletes believe will improve performance. A coaching instruction to reduce body mass, however, is not equivalent to evidence that a given individual's health or performance will benefit. Energy intake, training adaptation, recovery, and developmental requirements must be considered in the actual context of competition. The IOC REDs consensus identifies attempts to improve power-to-weight ratios, extreme leanness, physique demands, and high exercise expenditure as settings that can lead to problematic low energy availability Mountjoy et al. (2023).

Exposure should be studied as a sequence of demands and decisions. A training block raises expenditure; dietary intake fails to rise sufficiently; performance may initially remain acceptable; symptoms and biological disruption develop over time. Another athlete begins deliberate restriction after being told a certain physique is necessary for selection. A third experiences food insecurity or logistical barriers when traveling and misses meals despite no weight-loss intention. These pathways converge on insufficient energy available for normal function but differ in psychological antecedents and suitable institutional responses. The 2026 IOC mental-health consensus explicitly acknowledges intentional and unintentional underfueling Reardon et al. (2026). A theory that labels all low availability as deliberate self-starvation will miss organizational, economic, and educational determinants that are measurable and potentially modifiable.

Performance pressure also changes what counts as evidence inside a team. A short period of successful results after dietary restriction may be interpreted as proof that restriction was effective, even when training improvements, maturation, technical learning, or external conditions offer alternative explanations. Outcome evaluation must distinguish an observed change in competitive result from the causal effect of an eating regimen and from delayed health consequences. The REDs literature emphasizes that negative health and performance outcomes may emerge across different timescales and domains rather than simultaneously Mountjoy et al. (2023). An adequate research design records exposure trajectories and measures athletic performance separately from medical indicators. Anecdotal performance gains cannot justify the inference that a sustained energy deficit is safe or optimal.

Body-composition monitoring warrants especially careful analysis because its measurement process can itself become a psychological exposure. The device, frequency, privacy of results, explanation of uncertainty, and authority of the person requesting measurement can alter the significance attached to a number. Measurements designed for scientific comparison are vulnerable to false precision when used as individualized commands. Repeated testing without a clear clinical or performance question can encourage self-surveillance, and an expectation of public disclosure may discourage honest reporting. The sport-specific evidence reviewed in Reardon et al. (2026) supports safeguarding against problematic monitoring and weight-regulation practices. Future intervention studies should test policies that protect confidentiality, require professional interpretation, and separate medically justified assessment from routine appearance policing.

Psychological Vulnerability, Perfectionism, Identity, and Control

Person-level vulnerability is best modeled as a pattern of interacting traits, histories, and current contingencies. Perfectionistic standards, elevated concern about errors, rigid self-evaluation, prior eating concerns, anxiety, and a strong need for control can become relevant when combined with intense competition and appearance-focused feedback. Many high-performing athletes pursue demanding goals without developing clinical pathology. The research problem is to identify when a motivational style becomes inflexible, disproportionately tied to body-related achievement, or accompanied by clinically meaningful distress and behavioral impairment. In the qualitative research synthesized by Fatt et al. (2024), rigidity and athlete identity emerged as recurring themes; these observations justify longitudinal hypotheses about how psychological inflexibility and perceived external expectations interact, but cannot establish a universal personality profile of eating disorders.

Athletic identity organizes time, social relationships, and future opportunity. When participation becomes the principal source of personal meaning, injury, deselection, or weight-related evaluation may threaten a larger sense of continuity. Restrictive eating or compulsive training can then acquire a function beyond nutrition or physical conditioning: it may appear to restore predictability, demonstrate commitment, or protect membership in a valued group. This interpretation requires empirical testing against alternative explanations such as affect regulation, habit formation, or practical misunderstanding of nutritional needs. The athlete's subjective account is evidence about lived meaning, not evidence that one unobserved process caused a biological outcome. Qualitative studies can develop conceptual categories for such testing and identify questions that broad symptom surveys have not asked Fatt et al. (2024).

The distinction between persistence and compulsion is central. Repeated training undertaken for a valued goal may be flexible, coordinated with recovery, and responsive to medical advice. Exercise becomes concerning when a person experiences it as obligatory despite injury, fear, or serious harm, or uses it as a means of compensating for eating in a broader pathological pattern. The frequency of training alone cannot determine this distinction. Clinical inquiry should examine distress when prevented from exercising, the rationale for the behavior, flexibility in the face of changing circumstances, and consequences for health and social functioning. The IOC mental-health consensus addresses exercise addiction and related symptoms within the broader evaluation of eating disorders and REDs Reardon et al. (2026).

A robust causal theory must also avoid assigning responsibility solely to individuals. Psychological vulnerability can amplify a harmful environment, while institutional practices can generate pressure in individuals who had no previous disorder. Evaluation of coaching culture, access to professional services, athletic contracts, competition rules, and peer norms is therefore part of assessment, not merely background description. Athletes should not be required to demonstrate a specific trait profile before physical or psychological symptoms merit clinical attention. The prevention trial by Martinsen et al. (2014a) provides a model for changing aspects of the educational and organizational environment, although the effectiveness of any particular component requires further evaluation. Research should test interactions between personal and structural exposures through prospective multilevel designs.

Eating Behavior, Compulsive Exercise, and Symptom Continuity

Eating behavior in sport ranges from flexible planning around training to restrictive patterns that undermine physical health, psychological well-being, and social functioning. A clinical evaluation considers cognition about food, shape, and weight; dietary restriction and avoidance; episodes of loss of control; compensatory behaviors; distress; impairment; and the relevant diagnostic criteria. Dietary structure itself is insufficient for classification, since athletes frequently organize meals around competition and recovery. At the same time, a superficially performance-oriented explanation does not establish that a behavior is harmless. The professional guidance summarized by American Psychiatric Association (2023) and NICE (2017/2020) requires a wider assessment of nutritional and psychiatric risk than can be obtained from weight or a single self-report score.

The spectrum of disordered eating should be treated as a research continuum only with care. The presence of a symptom does not imply inevitable progression to a formal disorder, and the absence of a formal diagnosis does not eliminate risk of malnutrition or problematic low energy availability. Subthreshold syndromes, episodic restriction, rigid food beliefs, and compensatory behaviors can be clinically consequential in athletes, particularly during periods of heavy expenditure. Conversely, low energy availability can emerge without psychological symptoms at all. The conceptual independence of these pathways is a central finding of the Triad and REDs frameworks De Souza et al. (2026) Mountjoy et al. (2023). A meaningful study therefore measures exposure, beliefs, and biological outcomes separately rather than creating a single severity ladder from attitude to diagnosis.

Athletic behaviors can complicate classification. Exercise is a legitimate component of sport but can function as a compensatory behavior within an eating disorder. Aesthetic or weight-class demands can normalize dietary manipulation without determining whether the individual meets clinical criteria. Symptoms may change as an athlete moves between intensive training, injury rehabilitation, travel, off-season periods, and competition. A one-time questionnaire is poorly suited to capturing such temporal instability, and disclosure may change according to whether assessment is confidential or linked to team selection. Athlete-specific tools were developed partly to improve detection in this context; Martinsen et al. (2014b) tested the Brief Eating Disorder in Athletes Questionnaire against interview-based information in a selected female elite-athlete sample, illustrating both the value and population limits of targeted measurement.

The meaning of symptoms should be tested against consequences and longitudinal patterns. Researchers can examine whether increased food rigidity predicts reduced availability or whether worsening availability is accompanied by affective, cognitive, and behavioral changes after accounting for training load and injury. Repeated interviews can track how rules about eating become embedded in team culture or personal identity, while medical observations establish whether the body shows physiologic disruption. The qualitative meta-synthesis by Fatt et al. (2024) suggests that shame and fear of consequences can delay recognition and disclosure. This makes confidential, nonpunitive assessment methods a scientific necessity as well as a safeguarding concern: the observed prevalence of a hidden behavior depends partly on the conditions under which the athlete is asked to report it.

Energy Availability, Energy Deficiency, and Measurement Uncertainty

Energy availability expresses the relationship between dietary intake, exercise expenditure, and fat-free mass. In its conventional research formulation, energy availability equals dietary energy intake minus exercise energy expenditure, divided by fat-free mass over a defined interval, usually a day. This construct became influential because it describes energy remaining to support bodily functions after exercise expenditure. It is a derived quantity that depends on the accuracy of each underlying estimate. Dietary records are vulnerable to reporting error; expenditure estimates vary with methods and sport; and body-composition estimates introduce their own uncertainty. Mountjoy et al. (2023) treats low availability as an exposure and problematic low availability as a condition requiring clinical interpretation rather than an automatically observed diagnosis. A single calculated number cannot establish the duration, severity, or clinical effects of an athlete's energetic state.

Short-term experiments provide an important mechanistic foundation. In the controlled study by Loucks & Thuma (2003), 29 habitually sedentary, regularly menstruating young women experienced specified energy-availability conditions for five days, followed by detailed measurements of luteinizing hormone pulsatility. The investigators observed reproductive endocrine changes at lower experimental availability levels, supporting causal sensitivity to the manipulated energetic exposure under their conditions. The experiment cannot establish a biologically universal threshold across competitive athletes, individuals with different hormone profiles, adolescents, or longer exposure windows. Energy intake, expenditure, body composition, and endocrine regulation change over time; adaptation and individual variability matter. Contemporary clinical consensus therefore warns against reducing energy deficiency to a single absolute threshold derived from a narrow historical experimental model.

The 2025 Triad update reframed energy status as a continuum involving energy deficiency and metabolic compensation, with calculated low energy availability treated as one possible indicator alongside weight change and other evidence De Souza et al. (2026). The 2023 IOC consensus similarly recognizes a continuum between relatively adaptable, transient energetic mismatches and problematic deficiency accompanied by impairment Mountjoy et al. (2023). These developments resolve an earlier conceptual tension: laboratory definitions can be precise within an experiment while clinical diagnosis requires integration of imperfect observations in a changing person. Body mass may remain apparently stable during metabolic compensation, and changes in body mass alone may not reveal the underlying balance of energy needs, training, growth, and recovery.

Research on energy availability should report the operational calculation, observation period, units, measurement method, uncertainty, and contextual training demands. It should distinguish measured intake from self-reported restriction and actual workload from estimated exercise expenditure. Researchers should predefine whether their outcome is estimated exposure, a metabolic indicator, a specific injury, menstrual or gonadal alteration, or a multidisciplinary diagnosis of REDs. A physiological marker may be informative without being diagnostically specific. The clinical application of this science lies in pattern recognition across independent observations rather than a single cutoff value; this principle is shared by Mountjoy et al. (2023) and the later Triad clinical guidelines Williams et al. (2026).

Mechanistic Pathways from Underfueling to Systemic Dysfunction

The REDs model begins with the body allocating finite energetic resources among physical activity and physiological maintenance. Prolonged or severe inadequacy can prompt compensatory changes in energy metabolism and affect endocrine signaling, immune function, recovery, hematologic function, and other systems. Different systems need not respond at the same threshold or timescale, and the observed pattern varies with sex, developmental status, exposure severity, training load, and previous health. The 2023 IOC consensus assembles evidence across these domains while emphasizing that strength of evidence differs by outcome Mountjoy et al. (2023). It is appropriate to recognize a common exposure pathway while resisting the claim that every fatigue episode, mood symptom, or performance decline in an athlete is caused by low energy availability.

An endocrine pathway can be tested experimentally by manipulating energy availability and measuring hormone dynamics. The work of Loucks & Thuma (2003) is a direct demonstration for luteinizing hormone pulsatility in its specific population. Other pathways are often inferred from observational associations among dietary practice, training, metabolic markers, and clinical outcomes. Observational data may be affected by changes in sleep, psychological stress, illness, travel, medications, or underlying endocrine disorders. Mechanistic explanations should therefore specify whether an effect was demonstrated in a controlled exposure, associated with a clinical syndrome, or proposed by biological analogy. Conflating these levels produces an appearance of certainty unsupported by the research design.

Psychological functioning may change alongside physiological adaptation. A prolonged mismatch between expenditure and intake can be accompanied by tiredness, irritability, concentration difficulties, sleep disruption, or mood changes, but these symptoms also occur in overreaching, depression, anxiety, and other common conditions. Conversely, psychological illness can contribute to deficiency through restriction or compulsive exercise. The IOC 2026 mental-health consensus explicitly presents mental-health conditions as possible antecedents or consequences of REDs, which makes causal direction an empirical question Reardon et al. (2026). Longitudinal studies with repeated clinical, psychological, and metabolic measurements are needed to identify whether physiological deterioration precedes changes in affective regulation or follows an already established eating pathology.

The performance pathway is equally complex. An athlete may maintain competitive execution for a period while energy reserves, tissue repair, or endocrine function deteriorate. Performance measures are sensitive to equipment, competitors, coaching, technique, motivation, and environmental conditions, so short-term results do not function as direct biological tests. Negative outcomes can appear through increased injury, slower recovery, reduced adaptation to training, or inconsistent availability for competition rather than an immediate decline in a single result. Mountjoy et al. (2023) documents this multi-system model, and De Souza et al. (2026) emphasizes physiological adaptation and differential rates of injury and reproductive recovery. The most persuasive future research will align repeated exposure estimates with objective health and sport outcomes over clinically meaningful intervals.

Endocrine, Bone, and Reproductive Consequences across Sexes

The classical Triad framework linked energy status, menstrual function, and bone health in girls and women. This organization remains clinically important because reproductive endocrine changes and skeletal vulnerabilities can emerge in athletes whose training and nutritional demands are not adequately matched. Menstrual disturbance is a sign requiring evaluation of many possible causes; it should not be assumed to be a benign effect of athletic participation. Likewise, bone stress injury has multiple determinants, including loading history, biomechanics, nutrition, skeletal maturation, and clinical conditions. De Souza et al. (2014) established an influential framework for risk recognition, while the updated science synthesis De Souza et al. (2026) gives greater explicit attention to metabolic compensation, ovarian steroid exposure, and the relationship between energy status and bone stress injury.

The revised Triad consensus discusses differing rates of onset and recovery across systems. Changes in energy intake may alter physiologic markers and menstrual function before skeletal deficits fully recover. The return of regular menstrual bleeding does not necessarily indicate the immediate restoration of ovulation or all ovarian hormonal function, an issue emphasized by the 2025 clinical update published in 2026 Williams et al. (2026). Bone health also has a different temporal scale because bone accrual and recovery depend on developmental stage, loading, nutrients, endocrine status, and prolonged tissue remodeling. This temporal mismatch has major implications for research design: a short follow-up showing symptom improvement is insufficient to demonstrate complete long-term recovery of skeletal outcomes.

Male athletes may experience energy-related suppression of gonadal function and problems with bone health even when standard menstrual-based detection pathways obviously do not apply. The Male Athlete Triad definition considers energy deficiency, functional hypothalamic hypogonadism, and reduced bone health as interrelated conditions, particularly relevant in certain endurance and weight-class settings Nattiv et al. (2021). The corresponding clinical consensus describes age-appropriate assessment, differential diagnosis, and individualized return-to-participation decisions Fredericson et al. (2021). These guidelines expand the population addressed by energy-deficiency medicine while underscoring that markers, normative ranges, and symptom presentations cannot simply be transposed from women to men.

Sex-related patterns must be investigated without flattening variation within groups. Hormonal contraception can complicate interpretation of menstrual patterns, while sex hormones, age, puberty, medical conditions, and previous energy availability alter what an endocrine observation signifies. Researchers should record relevant therapies and contextual factors, use appropriate laboratory interpretation, and distinguish biological sex-related outcomes from assumptions about gender identity or behavior. The scientific question concerns demonstrable functional disturbance in a person who trains, regardless of whether an obvious appearance change has occurred. The contemporary IOC framework includes female and male athletes, and the updated Triad literature provides more specialized models for particular reproductive and developmental pathways Mountjoy et al. (2023) De Souza et al. (2026).

Adolescent Development, Growth, and Vulnerable Transitions

Adolescence creates distinct nutritional and psychological conditions for athletic participation. Growth, maturation, bone accrual, education, changes in social identity, and variable training loads all place demands on energy and recovery. Competition can coincide with rapid body changes and heightened sensitivity to feedback from coaches, peers, and family. A dietary plan that neglects growth may become inadequate even when it appears to cover exercise expenditure. The 2025 Triad science update introduced an adolescent-specific model precisely because the timing of skeletal development, reproductive maturation, and energy conservation differs from adult trajectories De Souza et al. (2026). Assessments in young athletes must consider growth patterns and developmental context alongside current symptoms.

Early specialization and appearance-based selection can amplify the significance of bodily change. An adolescent may interpret normal pubertal development as evidence of competitive decline if training culture treats one physique as ideal. The resulting psychological pressure can coexist with practical underfueling due to school schedules, travel, intensive training, and limited autonomy over meals. Some athletes may have neither overt food-related distress nor knowledge of their energetic needs. The 2026 IOC mental-health consensus identifies early sport specialization and problematic body-composition practices among sport-specific exposures relevant to eating disorders Reardon et al. (2026). Each proposed pathway needs testing against developmental and socioeconomic factors rather than assuming that adolescent status is itself a disorder.

The randomized trial by Martinsen et al. (2014a) is a central intervention study because it evaluated prevention among students attending Norwegian elite sports high schools. Its school-level intervention and control structure created an opportunity to examine whether changing education and support during adolescence could affect later eating-disorder outcomes. The trial reported beneficial results in female students, including fewer new diagnosed cases in intervention schools at a specified follow-up, while results for male students were less conclusive. The study also demonstrates the difficulty of evaluating institutional interventions: allocation occurred in clusters, and findings are tied to particular educational and cultural settings. Its value is evidence for a feasible preventive strategy, not proof that an identical program will work across all countries, ages, and sports.

A developmental approach should track transitions as potential stress points: joining a national training program, moving away from family, changing coaches, entering college sport, recovering from injury, and leaving competition. During these transitions, nutritional provision, identity, social comparison, and access to confidential care can change rapidly. The clinical review should assess growth and health when appropriate, while psychological research should measure perceived body expectations, autonomy, support, and help-seeking. The 2025 Triad clinical update calls for age-sensitive risk assessment and interdisciplinary management, including adolescent pathways Williams et al. (2026). Longitudinal designs can test whether institutional transitions explain new symptoms beyond baseline vulnerabilities and ordinary developmental change.

Injury, Recovery, Performance, and Longitudinal Dynamics

Injury offers a natural context in which psychological, nutritional, and physiological processes interact. Training expenditure may fall abruptly while an athlete fears changes in body composition, loses routine, and worries about selection. Others may enter rehabilitation with an established history of inadequate fueling that affects skeletal health or recovery. Bone stress injury is especially relevant because it can reflect interactions among loading, energy status, endocrine changes, and tissue adaptation. The 2025 Triad update explicitly incorporates bone stress injury into its model and differentiates clinically important skeletal risk from an isolated low body-weight observation De Souza et al. (2026). Research should establish the event sequence rather than conclude that any injury is evidence of REDs.

A performance-centered culture may delay care when the athlete's current results remain high. The REDs consensus recognizes that impairment may accumulate over time and across domains, including recovery, injury vulnerability, cognition, and motivation Mountjoy et al. (2023). This creates a longitudinal problem of delayed manifestation: performance remains acceptable at one assessment, while markers of health decline and later constrain athletic availability. Studies that select only athletes who continue competing may underrepresent those with severe symptoms who have already withdrawn. Prospective designs should therefore follow the entire eligible cohort, account for loss to follow-up, and record injury-related withdrawal as an outcome rather than an inconvenience in analysis.

Clinical recovery is multidimensional. Improvement in food intake or reduced fear of weight gain may precede resolution of endocrine or skeletal consequences, while normalization of a laboratory measure does not automatically imply that psychological symptoms have remitted. Decisions about training modification, restrictions, and return to sport depend on the diagnosed problem, the risk profile, and the availability of multidisciplinary follow-up. The IOC 2023 REDs model and the updated Triad clinical recommendations supply structured approaches to assessment and risk stratification Mountjoy et al. (2023) Williams et al. (2026). These are tools for qualified clinicians; their clinical judgments cannot be reduced to a single published traffic-light category applied by athletes, coaches, or an automated score alone.

A high-quality longitudinal study should define time zero, the exposure window, and outcomes at separate recovery intervals. It should distinguish first incidence, recurrence, symptom persistence, and recovery, and it should account for training cycles and seasonality. Repeated measures of eating behavior, body-image pressure, injury history, growth or reproductive status, training workload, and relevant biomarkers would allow more informative models of temporal interaction. Independent performance outcomes should be analyzed alongside medical and psychological indices, because one domain can improve while another deteriorates. Such designs would move the field beyond static cross-sectional labels toward trajectories of risk and adaptation that are directly useful for prevention and professional management Reardon et al. (2026) De Souza et al. (2026).

Epidemiology and Interpretation of Prevalence Estimates

The prevalence of eating disorders in athletes depends on what is counted, who is sampled, and how an observation is verified. A formal clinical disorder, elevated questionnaire score, intermittent disordered behavior, and inadequate energy availability are four different outcomes. Even within a named screening instrument, changes in cutoff, language, administration, and sport context can produce different prevalence estimates. Early epidemiological studies by Sundgot-Borgen (1993) and Sundgot-Borgen & Torstveit (2004) helped establish the need for interview-based confirmation and comparison populations. Their findings are historically influential, but differences in diagnostic categories, selection criteria, and sporting environments make direct comparison with later surveys difficult. Epidemiology must therefore state denominators and definitions before it states percentages.

The large quantitative synthesis by Nasser et al. (2026) searched four databases through December 2024 and included 190 studies comprising 58,335 active athletes in its systematic review. Its meta-analysis included 127 selected studies with 43,006 athletes and estimated a pooled prevalence of eating disorders or disordered eating of 20.0%, with a 95% confidence interval of 17.0% to 23.0%. The numerator in that summary combined outcomes defined by clinical diagnosis, interview, or screening thresholds, making the result an estimate of the broad ED/DE construct rather than a rate of confirmed psychiatric diagnoses. The underlying studies used multiple instruments and sampled different sporting populations. The scientific interpretation is that eating-related concerns are widespread enough to warrant systematic attention, while the exact rate of diagnosed illness cannot be read directly from the pooled estimate.

Instrument-dependent variation is substantive. The same 2026 meta-analysis reported lower pooled estimates for studies using the Eating Disorder Examination and higher estimates for studies using the Eating Disorder Inventory, despite finding no statistically significant subgroup difference when measures were broadly classified as interview/diagnosis versus self-report. Its meta-regression associated a greater proportion of women and more recent publication year with higher prevalence estimates Nasser et al. (2026). These findings may reflect differences in symptom expression, screening sensitivity, athlete recruitment, changing awareness, and reporting conditions. They do not prove that a particular sport necessarily causes the observed difference or that disorder incidence has increased by a known amount over time. Uncontrolled confounding and differences in study quality remain alternative explanations.

Prevalence research has the additional problem of disclosure. An athlete who fears being removed from a team may underreport symptoms; another may interpret ordinary performance nutrition as pathological when completing a generic questionnaire. Studies drawn from health clinics are influenced by referral patterns, while voluntary online surveys can attract participants with strong personal interests in the topic. Competitive level, sex, age, nationality, body-related rules, and injuries should be recorded and stratified without constructing stereotypes. A strong future prevalence program would use probability-based or clearly defined cohort samples, repeat validated measurement, provide confidential interview confirmation, and report clinical diagnoses separately from broad positive screens. This is particularly important given the questionnaire limitations discussed in Martinsen et al. (2014b) and the population heterogeneity identified by Nasser et al. (2026).

Screening Instruments, Diagnostic Interviews, and Clinical Evaluation

Screening seeks to identify people whose symptoms or experiences warrant further assessment. Diagnosis establishes a clinical condition after structured evaluation of psychological, behavioral, and physical evidence, considering alternative explanations. Screening has value when the instrument detects relevant presentations accurately in the population being tested, when confidentiality is protected, and when a credible referral pathway exists. It also carries risks of missed cases and false positives. The National Institute for Health and Care Excellence explicitly advises against using instruments such as SCOFF alone to determine whether a person has an eating disorder NICE (2017/2020). In sport, generic instruments can overclassify disciplined behaviors while missing presentations shaped by muscularity, sport-specific dieting, and a desire to remain eligible for competition Nasser et al. (2026).

Athlete-targeted tools have attempted to address these problems. Martinsen et al. (2014b) developed the Brief Eating Disorder in Athletes Questionnaire through screening and eating-disorder interviews in an adolescent female elite-athlete sample. Its validity evidence is linked to the population and clinical reference method used in the study; transferring the same score to adult male athletes, para athletes, or different sporting cultures requires independent testing. The Low Energy Availability in Females questionnaire, described by Melin et al. (2014), was designed to screen selected female athletes for indications relevant to the Triad, including menstrual and injury histories. It is an indirect risk instrument and cannot by itself determine energetic exposure or diagnose REDs. The titles of these instruments are reminders of their initial population scope, not permissions to generalize them beyond validation data.

An adequate clinical assessment integrates confidential interview, eating-disorder symptom evaluation, nutritional and training history, relevant medical examination, and investigations selected according to symptoms and risk. A clinician may need to examine weight trajectory, vital signs, physical complications, endocrine function, bone health, and mental-health comorbidities. No single variable captures the entire syndrome. The 2026 IOC statement recommends coordinated evaluation by sports medicine, nutrition, and mental-health professionals for athletes with eating disorders or REDs Reardon et al. (2026). NICE guidance similarly advises assessment of physical compromise, mental-health conditions, compensatory behaviors, and urgent safety concerns rather than relying on one body-mass measure NICE (2017/2020). Recognition of these components provides a scientific account of assessment; individual investigations belong to qualified professional judgment.

Screening also operates in an institutional environment. Athletes may need assurance that sensitive information will be handled by clinicians with clear confidentiality and safeguarding responsibilities rather than automatically forwarded to a coach or selector. An assessment program linked to punishment without accessible care can distort responses and worsen concealment. The qualitative evidence synthesized by Fatt et al. (2024) identifies shame, stigma, and power imbalance among obstacles to help-seeking. Instrument calibration should therefore be evaluated jointly with the conditions of administration, follow-up, and professional response. The validity of a screen in practice includes whether it leads to accurate, timely clinical evaluation and appropriate care, not simply whether a questionnaire's internal consistency is high.

REDs Clinical Risk Stratification and Differential Diagnosis

Relative Energy Deficiency in Sport requires clinical interpretation of a pattern of exposure and bodily consequences. The 2023 IOC consensus introduced a revised Clinical Assessment Tool, IOC REDs CAT2, for structured recognition and risk stratification Mountjoy et al. (2023). The tool uses multiple indicators and contextual modifiers rather than declaring a diagnosis from an estimated energy-availability number alone. Its color-graded approach concerns clinical severity, potential risk, and appropriate participation decisions in a defined framework. It expressly requires expert interpretation and warns against use in isolation. Treating the output as an automated permission slip would erase the uncertainty of measurements and the individualized decisions necessary for an athlete whose health, competition calendar, and medical history interact.

Differential diagnosis is essential because menstrual changes, fatigue, low bone mineral density, sleep disturbance, gastrointestinal symptoms, recurrent injury, and mood symptoms can arise from several causes. A sports physician must consider the individual's medical background, medications, training and nutrition, endocrine conditions, and other relevant illnesses before attributing signs to REDs. Psychological diagnosis introduces a separate differential process: dietary behavior may involve a recognized eating disorder, another psychiatric condition, sport-contextualized habits, or inadvertent inadequate fueling. REDs and eating disorders can coexist, but neither condition logically requires the other. The IOC framework places independent exposure and clinical findings at the center of reasoning Mountjoy et al. (2023), and the contemporary mental-health consensus specifically recommends assessing for REDs even without an eating disorder Reardon et al. (2026).

The Female and Male Athlete Triad frameworks complement the general REDs assessment with population-relevant evidence concerning reproductive and skeletal systems. The updated guidance for women addresses adolescent developmental factors, clinical risk categories, and return-to-play decisions Williams et al. (2026). The male-athlete consensus identifies different markers and considerations for gonadal function, skeletal health, and exposure patterns Fredericson et al. (2021). These guidelines should be interpreted according to their specified age and sex-related evidence base. The presence of a tool for one group does not establish its transportability to another, and applying a clinician's risk scheme as a universal statistical classifier would overstate validation.

A reliable clinical and research protocol should document how the conclusion was reached: the primary concern, symptoms and their temporal pattern, exposure assessment, medical and psychiatric alternatives considered, additional tests ordered, and reasoning behind any change in training participation. It should also make room for revision as new information becomes available. High-consequence decisions require protection against conflicts of interest when coaches, clubs, sponsors, or selectors stand to gain from an athlete's continued competition. The 2023 IOC statement describes contextual modifiers and individualized medical judgment in participation guidance Mountjoy et al. (2023). The research implication is to evaluate not only whether tools classify risks but also whether their use improves clinical outcomes, avoids missed cases, and maintains athletes' access to independent care.

Treatment Principles and Coordinated Clinical Care

The treatment evidence base spans specialized eating-disorder care and management of energy-related physiological impairment. For diagnosed eating disorders, clinical guidelines organize psychological therapy, nutritional rehabilitation, monitoring of medical complications, and coordinated mental-health assessment according to condition, age, severity, and comorbidity. The American Psychiatric Association practice guideline describes evidence-based treatment within a psychiatric framework American Psychiatric Association (2023). NICE provides detailed recommendations for early assessment, age-appropriate services, clinical monitoring, and treatment selection NICE (2017/2020). These recommendations cannot be reduced to a sport-specific meal rule or motivational coaching technique; they are clinical approaches requiring licensed professionals who can assess psychiatric symptoms and physical safety together.

Treatment of problematic low energy availability addresses the energetic mismatch and resulting physiological concerns, including the possible need to modify training exposure. This general principle does not specify an individual caloric prescription, safe weight trajectory, or timetable for return to sport. The 2023 IOC consensus recommends a multidisciplinary, individualized approach and risk assessment guided by clinical expertise Mountjoy et al. (2023). The 2025 Triad clinical update elaborates on the evaluation of reproductive and skeletal outcomes and the requirement for sustained monitoring Williams et al. (2026). When an eating disorder is present, restoration of energy availability and psychological treatment may need to proceed together; when an eating disorder is absent, nutritional education, workload management, and assessment of practical barriers can still be necessary.

Care must address the athlete's identity, autonomy, and social environment. A treatment plan may become difficult to sustain if team culture rewards continued restriction, if a coach demands rapid changes in appearance, or if the athlete sees time away from sport as an existential loss. Effective coordination requires medical professionals, dietitians with sport-specific competence, and qualified mental-health practitioners to share an appropriate plan while maintaining confidentiality and consent. The 2026 IOC statement emphasizes interdisciplinary management, comorbid anxiety and mood conditions, and the need for suicide-risk assessment when clinically indicated Reardon et al. (2026). A scientific monograph can establish the rationale for this structure without offering treatment protocols to individual readers or claiming specialist clinical approval that has not occurred.

Outcome evaluation should follow more than the disappearance of one symptom. Relevant domains include eating-related distress, flexibility around food and exercise, sufficient energy for health, functional medical recovery, absence of serious complications, mental-health functioning, and durable participation choices. The updated Triad literature emphasizes that endocrine or skeletal recovery may proceed at different rates Williams et al. (2026). Observing an improved training result or a single menstrual cycle is therefore insufficient evidence of full recovery. Clinical studies should report patient-relevant outcomes, adverse events, follow-up duration, recurrence, and the extent to which participants were able to resume valued activity safely. Evidence for specific treatments in non-athlete clinical populations may guide practice, but athlete-specific effectiveness and contextual barriers require separate evaluation.

Prevention, Coaching, Organizational Culture, and Safeguarding

Prevention begins with identifying modifiable conditions that shape exposure and behavior before clinical problems become entrenched. These conditions include inaccurate beliefs about body composition and performance, inadequate access to fueling, frequent appearance-focused feedback, rapid weight manipulation, secretive measurement, and unclear routes to independent care. An organizational approach can examine the authority of coaches and selectors, the timing of weight-related decisions, and whether athletes have freedom to report symptoms without automatic humiliation. The 2026 IOC consensus recommends attention to body dissatisfaction, unhealthy dieting, compulsive exercise, and problematic low energy availability as interacting prevention targets Reardon et al. (2026). Prevention should be assessed through independently measured changes in harm and clinical outcomes rather than the mere delivery of educational material.

The Norwegian intervention by Martinsen et al. (2014a) provides one of the strongest direct prevention tests in this field. Sixteen elite sport high schools participated, with nine intervention schools and seven control schools; 465 first-year athletes were followed across the study. Among female athletes, the authors reported no new eating-disorder cases in intervention schools at the relevant later assessment, compared with eight new cases in control schools. At an earlier follow-up, the odds ratio for reporting symptoms favored intervention schools, although the effect attenuated at a later follow-up. These results support the potential of structured prevention in a selected educational setting. Their interpretation must also account for school-cluster allocation, small numbers of incident clinical cases, limited evidence of benefit among male athletes, and the cultural specificity of the program.

Prevention can operate at several connected levels: educating athletes and staff about nutrition and health, evaluating the necessity and confidentiality of body-composition assessment, ensuring professional access when concerns arise, and establishing rules for respectful body-related communication. A strong intervention changes organizational contingencies as well as individual knowledge. For example, if a team's selection system continues to reward hazardous short-term weight practices, isolated psychoeducation may be undermined. The qualitative evidence on power imbalance and stigma provides a plausible account of such implementation failure Fatt et al. (2024). Future trials should document which program components are actually delivered, how staff behavior changes, whether athletes perceive safety, and whether effects persist across seasons and coaching turnover.

Safeguarding is also an issue of institutional accountability. Clear referral channels, privacy protections, age-appropriate consent, and transparent separation of medical decisions from competitive interests help reduce the incentives for concealment. These principles are especially important for minors and for athletes dependent on scholarships, national-team selection, or professional contracts. Professional clinical bodies advise early recognition and appropriate care rather than relying on a single low-weight threshold or delaying evaluation until a serious complication occurs NICE (2017/2020). The scientific research question is whether safeguarding structures improve detection, reduce delayed care, and lower rates of eating disorders and energy-deficiency complications. Such outcomes must be tested through rigorous, ethically designed studies rather than assumed from the presence of written policies.

Athlete Diversity, Para Sport, and Research Generalizability

The athlete population is heterogeneous in its bodily capacities, training demands, social environments, and exposure to weight-related norms. Studies drawn largely from young, female, elite, or European cohorts have produced important knowledge, but their results cannot be treated as population-invariant. Para athletes may encounter differences in body composition, health status, mobility, adaptive equipment, classification systems, injury exposure, and barriers to accessible clinical services. Athletes from resource-constrained settings may face food insecurity or limited access to qualified sports nutritionists. An athlete's gender identity and cultural background may also shape the meaning of body ideals and willingness to disclose distress. The IOC's 2026 mental-health consensus includes para-athlete considerations and emphasizes the need for context-sensitive screening Reardon et al. (2026).

Measurement generalizability requires more than translating a questionnaire. An instrument built around the thinness ideal may fail to detect muscularity-oriented distress, and one requiring menstrual-history responses cannot be interpreted identically for every athlete. A measure derived from selected female elite athletes should be validated independently before being used to estimate prevalence in men, adolescents, para athletes, or culturally different populations. The developmental studies of Martinsen et al. (2014b) and Melin et al. (2014) offer clear examples of instruments whose original validation samples constrain claims about transportability. Large meta-analytic samples can improve precision while still reproducing systematic biases if the component studies rely on instruments that omit particular symptom presentations Nasser et al. (2026).

Competitive level is another source of heterogeneity rather than a simple ordering of risk. Recreational participation, collegiate competition, national-team sport, and professional careers differ in funding, surveillance, choice, time commitments, access to recovery, and personal stakes of selection. A population with higher average training volume may have improved nutrition services or more restrictive performance incentives; the net effect cannot be inferred from competition level alone. The 2026 prevalence synthesis reported no statistically significant subgroup difference by training level in its aggregate comparisons Nasser et al. (2026). That result is bounded by the included studies and their classification methods. It should motivate more detailed investigation of sport-specific exposures rather than the assumption that all competitive levels are psychologically equivalent.

Ethically robust research should involve athletes in selecting meaningful outcomes and identifying harms from data collection. In a small team, reporting extremely detailed symptoms or demographic characteristics may make individual participants identifiable even when names are removed. Researchers should specify safeguards for privacy, independent clinical referral, and participants' right to withdraw. Qualitative work can reveal experiences and meanings absent from conventional measurement, while quantitative designs can test whether proposed mechanisms generalize to broader populations. The qualitative synthesis by Fatt et al. (2024) identified major gaps in representation of men, para athletes, and ethnically diverse groups. Closing those gaps requires participatory design, appropriate language adaptations, and reporting that distinguishes absence of data from absence of risk.

Digital Media, Monitoring Technologies, and the Artificial Era

Digitally mediated sport introduces new channels through which the body becomes an object of comparison and measurement. Images of elite performers, individualized performance analytics, wearable-derived expenditure estimates, body-composition dashboards, and nutrition-tracking applications may influence what athletes believe they must change. These technologies can support communication and access to health services when designed for a defined scientific or clinical purpose. They can also increase the frequency of body surveillance, introduce inaccurate numerical certainty, and place sensitive information under institutional control. A measurement that is useful for longitudinal group research may be unsuitable as an immediate individual directive. The broader literature on sport-related body pressure and REDs provides the necessary empirical context for examining digital systems Reardon et al. (2026) Mountjoy et al. (2023).

An artificial system that estimates training demands or summarizes clinical evidence has a limited epistemic role. Its inputs may contain wearable sensor error, incomplete dietary logs, inaccurate energy expenditure, and biased training data; output confidence cannot exceed the validity of these underlying observations. A prediction of energy deficiency is not identical to a clinical diagnosis of REDs, and an algorithmic risk category cannot substitute for the qualified professional assessment required by IOC guidance Mountjoy et al. (2023). Responsible research should separately test measurement validity, prediction calibration, subgroup performance, response to missing data, and the consequences of false reassurance or unnecessary alarm. Systems should also disclose how data are stored, who can access results, and whether performance decision-makers can use health information outside the purpose for which it was collected.

The Artificial Era adds a governance question to sport psychology: who controls the interpretation and circulation of intimate information about athletes' bodies? A team may gain valuable longitudinal evidence from carefully administered monitoring while an athlete loses practical autonomy if those data are used for unreviewed selection decisions. Scientific research should therefore assess consent, access, audit trails, fairness, and the separation of clinical from commercial or competitive authority. Privacy and safeguarding become properties of the whole sociotechnical process, extending from the individual sensor to the institution that interprets the output. The athlete's experience of being monitored should be measured alongside any gains in accuracy or efficiency. This is an extension of established sport-environment concerns, not evidence that a new technology already causes or prevents eating disorders Fatt et al. (2024).

Prospective evaluations of digital interventions should compare them against credible non-digital alternatives and document the potential for unintended consequences. Improved knowledge about fueling may be beneficial, while repeated appearance scoring could intensify body dissatisfaction in a susceptible subgroup. Researchers should monitor behavioral change, clinical referral, wellbeing, and adverse effects, and predefine whether the system is intended for education, screening, or licensed professional decision support. The earlier IOC statement on elite-athlete mental health established the importance of integrating psychological expertise into sporting care Reardon et al. (2019), and the 2026 update integrates contemporary concerns about athlete-specific eating disorders and REDs Reardon et al. (2026). Within this continuity, Artificial can contribute validated tools for scientific observation only through transparent methods and accountable clinical oversight.

Conflicting Evidence and Methodological Limitations

Several apparent contradictions in this field reflect differences in what studies measure. The 2023 body-image synthesis found that athletes overall reported less body-image concern than non-athletes across its included adult samples Burgon et al. (2023). A separate literature documents clinically important eating-disorder risk among some athletic populations, and the 2026 meta-analysis pooled a substantial frequency of positive eating-disorder or disordered-eating outcomes Nasser et al. (2026). Both findings can coexist because body-image comparisons and eating-pathology prevalence use different constructs, comparison groups, and recruitment procedures. Furthermore, a favorable average athlete body-image score may conceal susceptible individuals, while a broad eating-pathology screen can count symptoms without establishing a formal diagnosis. Resolving the tension requires measurement of both constructs within the same representative, longitudinal sample.

Mechanistic certainty also varies. Controlled studies of short-term energy availability show causal effects on selected endocrine signals under specified conditions, as in Loucks & Thuma (2003). Clinical REDs outcomes arise over longer intervals and interact with developmental and training factors that are difficult to manipulate experimentally. Observational evidence is indispensable but susceptible to confounding and reverse causation. If a study associates fatigue and low energy availability, it must consider whether illness, changes in training, dietary underreporting, or another condition produced the association. Likewise, observed menstrual disturbance requires a differential diagnosis before attribution to nutritional deficiency. The updated IOC and Triad consensuses provide useful syntheses while acknowledging the limits of single indicators and general thresholds Mountjoy et al. (2023) De Souza et al. (2026).

The prevention literature illustrates the difference between promise and broad effectiveness. The Norwegian school intervention reported benefit for selected female outcomes Martinsen et al. (2014a), but the trial did not determine the isolated causal effect of each educational component or demonstrate equivalent results for all athletes. Interventions may interact with local rules, staff adherence, resources, and patterns of athlete disclosure. General psychiatric clinical guidelines offer a basis for treating recognized eating disorders American Psychiatric Association (2023) NICE (2017/2020), yet the question of how treatment should adapt to individual sport contexts still requires research. Claims of successful prevention or rehabilitation should specify diagnosis versus self-report outcome, duration of follow-up, adherence, comparison intervention, and the relevant population.

Publication bias, missing data, and measurement heterogeneity constrain synthetic estimates. Studies showing dramatic risk in a particular sport may be more visible than null findings, while participants reluctant to disclose symptoms can be systematically missing from studies. Different questionnaires may be validated against different reference interviews or clinical standards. Meta-analytic pooling yields useful summaries but cannot repair all underlying validity problems. Qualitative studies offer richer explanatory accounts yet can overrepresent participants ready to tell their stories. The 2026 meta-analysis Nasser et al. (2026) and 2024 qualitative synthesis Fatt et al. (2024) provide complementary evidence whose strengths and limitations should be judged together. Scientific certainty should attach to a specified claim and research design, not to the number of citations collected beneath an expansive label.

A Research Agenda for Causal and Translational Progress

The first priority is prospective, multisite research that separates psychological antecedents, energetic exposures, biological changes, and clinical outcomes. Cohorts should establish baseline body-image attitudes, previous psychiatric and injury histories, sport rules, access to nutrition, and training loads before follow-up begins. Repeated measurements can then test whether changes in coaching behavior, body-related surveillance, or training expenditure predict subsequent restriction, symptoms, low availability, and medical impairment. Appropriate comparisons, transparent missing-data procedures, and preregistered causal hypotheses would improve inference. The field should distinguish between predictors that support early identification and mechanisms that change when manipulated. The conceptual model of Mountjoy et al. (2023) and empirical gaps identified by Nasser et al. (2026) provide a foundation for this work.

A second priority is cross-population validation of assessment tools. Studies should compare symptom questionnaires against confidential structured interviews and assess the correspondence of energy-deficiency indicators with clinically evaluated REDs outcomes. Measurement invariance should be examined across sex, gender, adolescence, para sport, language, sport type, and competition level. Sensitivity, specificity, calibration, and positive predictive value should be reported in the actual settings where an instrument will be used. Researchers should document the practical effect of screening on referral and care, since a technically accurate tool can be harmful or ineffective if athletes fear punitive consequences. The development histories of Martinsen et al. (2014b) and Melin et al. (2014) show why careful validation is integral to interpretation rather than a final administrative check.

A third priority is rigorous evaluation of organizational prevention. Cluster trials can examine training for coaches, confidential clinical access, changes to body-composition measurement policies, and equitable provision of food and recovery resources. Intervention packages should be described in enough detail for replication, with attention to staff behavior, athlete perceptions, adverse effects, and durability. Where randomization is infeasible, credible natural experiments and interrupted time-series designs may clarify what changed after rule revisions. Outcomes should include clinical diagnosis and objectively meaningful health indicators as well as self-reported attitudes. Martinsen et al. (2014a) supplies an important early example; contemporary REDs and athlete mental-health guidance identifies additional organizational questions that have not yet received equivalent controlled testing Reardon et al. (2026).

A fourth priority is translation across the entire period of clinical risk and recovery. Research needs to establish the relative time courses of psychological symptoms, reproductive function, metabolic compensation, bone repair, and sustained participation after intervention. Multidisciplinary studies can test how medical and mental-health care interact while preserving participant safety and avoiding ethically unjustified exposure to harmful energy deficits. The 2025 Triad updates emphasize that physiologic systems can recover at different rates De Souza et al. (2026) Williams et al. (2026). New observational and intervention protocols should therefore specify long-term endpoints rather than assuming that an early improvement in mood, eating behavior, or one laboratory indicator completes the recovery process. Transparent data definitions and external replication are necessary to convert promising findings into dependable clinical knowledge.

Conclusion

Eating disorders, body image, and Relative Energy Deficiency in Sport form an interconnected field whose explanatory strength depends on maintaining the precision of its constituent constructs. An eating disorder is a psychiatric diagnosis established through clinical assessment; disordered eating is a broader set of potentially harmful attitudes and behaviors; body image concerns the evaluation and experience of the body; and REDs is a syndrome of physiological and/or psychological dysfunction associated with problematic low energy availability. These phenomena can overlap, reinforce one another, or arise along independent pathways. The modern IOC definition of REDs and the revised Triad literature supply a basis for organizing their relationships without making any single behavioral or bodily indicator definitive Mountjoy et al. (2023) De Souza et al. (2026).

The historical evidence shows the development of a field through successive methodological improvements: early athlete surveys and interviews, physiology experiments, Triad frameworks, expanded male-athlete and IOC consensus models, qualitative accounts of institutional pressure, prevention trials, and systematic synthesis. Each contribution answers a different type of question. The experimental work of Loucks & Thuma (2003) provides evidence for a specific short-term physiological response. The preventive trial of Martinsen et al. (2014a) investigates whether changing an educational setting can alter subsequent clinical outcomes. The 2026 prevalence meta-analysis by Nasser et al. (2026) summarizes a broad ED/DE burden while showing how tool choice affects estimation. The logic of scientific interpretation requires that the method, studied population, and outcome definition accompany every claimed finding.

The integrated evidence locates vulnerability at the meeting point of individual history, bodily experience, training demands, performance incentives, and institutions with authority over athletic participation. Body-image research rejects a universal characterization of athletes as dissatisfied with their bodies; important group-level protective experiences coexist with localized or individual risks Burgon et al. (2023). Qualitative evidence describes power imbalances, rigid norms, stigma, and barriers to help-seeking Fatt et al. (2024). REDs research establishes that inadequate fueling may be unintentional and can be clinically important without a formal eating disorder Reardon et al. (2026). These findings justify multidisciplinary science and care that examine both individual needs and environmental conditions rather than reducing all risks to motivation or discipline.

Future knowledge will depend on rigorous prospective designs, independent validation of screening instruments, inclusive athletic samples, intervention trials that modify institutional practices, and long-term study of recovery across psychological and biological systems. Clinical guidelines currently support early assessment and coordinated professional response where there are concerning symptoms or health indicators NICE (2017/2020) Williams et al. (2026). This monograph's scientific synthesis places the protection of health and autonomy alongside the study of athletic performance. In the Artificial Era, analytic and monitoring systems can strengthen research only when their measurements are valid, their limitations visible, and the interpretation of clinical risk accountable to qualified professionals. The enduring scientific objective is to explain the conditions under which athletic striving and bodily health can develop together, and to identify those under which their relation becomes hazardous.

References

  • American Psychiatric Association. (2023). The American Psychiatric Association practice guideline for the treatment of patients with eating disorders (4th ed.). American Psychiatric Association. https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines/eating-disorders
  • Burgon, R. H., Beard, J., & Waller, G. (2023). Body image concerns across different sports and sporting levels: A systematic review and meta-analysis. Body Image, 46, 9–31. https://doi.org/10.1016/j.bodyim.2023.04.007
  • De Souza, M. J., Nattiv, A., Joy, E., Misra, M., Williams, N. I., Mallinson, R. J., Gibbs, J. C., Olmsted, M., Goolsby, M., & Matheson, G. (2014). 2014 Female Athlete Triad Coalition consensus statement on treatment and return to play of the Female Athlete Triad. Clinical Journal of Sport Medicine, 24(2), 96–119. https://doi.org/10.1097/JSM.0000000000000085
  • De Souza, M. J., Williams, N. I., Misra, M., Nattiv, A., Joy, E., Barrack, M., Ricker, E. A., Gorrell, S., Koltun, K. J., O’Donnell, E., Mallinson, R. J., Salamunes, A. C. C., Woodruff, K., Fredericson, M., & Plessow, F. (2026). 2025 update to the Female Athlete Triad Coalition consensus statement Part 1: State of the science and introduction of a new adolescent model. Sports Medicine, 56(2), 327–373. https://doi.org/10.1007/s40279-025-02333-z
  • Fatt, S. J., George, E., Hay, P., Jeacocke, N., Day, S., & Mitchison, D. (2024). A systematic review and meta-synthesis of qualitative research investigating disordered eating and help-seeking in elite athletes. International Journal of Eating Disorders, 57(8), 1621–1641. https://doi.org/10.1002/eat.24205
  • Fredericson, M., Kussman, A., Misra, M., Barrack, M. T., De Souza, M. J., Kraus, E., Koltun, K. J., Williams, N. I., Joy, E., & Nattiv, A. (2021). The Male Athlete Triad: A consensus statement from the Female and Male Athlete Triad Coalition Part II: Diagnosis, treatment, and return-to-play. Clinical Journal of Sport Medicine, 31(4), 349–366. https://doi.org/10.1097/JSM.0000000000000948
  • Loucks, A. B., & Thuma, J. R. (2003). Luteinizing hormone pulsatility is disrupted at a threshold of energy availability in regularly menstruating women. Journal of Clinical Endocrinology & Metabolism, 88(1), 297–311. https://doi.org/10.1210/jc.2002-020369
  • Martinsen, M., Bahr, R., Børresen, R., Holme, I., Pensgaard, A. M., & Sundgot-Borgen, J. (2014). Preventing eating disorders among young elite athletes: A randomized controlled trial. Medicine & Science in Sports & Exercise, 46(3), 435–447. https://doi.org/10.1249/MSS.0b013e3182a702fc
  • Martinsen, M., Holme, I., Pensgaard, A. M., Torstveit, M. K., & Sundgot-Borgen, J. (2014). The development of the Brief Eating Disorder in Athletes Questionnaire. Medicine & Science in Sports & Exercise, 46(8), 1666–1675. https://doi.org/10.1249/MSS.0000000000000276
  • Melin, A., Tornberg, Å. B., Skouby, S., Faber, J., Ritz, C., Sjödin, A., & Sundgot-Borgen, J. (2014). The LEAF questionnaire: A screening tool for the identification of female athletes at risk for the Female Athlete Triad. British Journal of Sports Medicine, 48(7), 540–545. https://doi.org/10.1136/bjsports-2013-093240
  • Mountjoy, M., Sundgot-Borgen, J., Burke, L., Carter, S., Constantini, N., Lebrun, C., Meyer, N., Sherman, R., Steffen, K., Budgett, R., & Ljungqvist, A. (2014). The IOC consensus statement: Beyond the Female Athlete Triad—Relative Energy Deficiency in Sport (RED-S). British Journal of Sports Medicine, 48(7), 491–497. https://doi.org/10.1136/bjsports-2014-093502
  • Mountjoy, M., Sundgot-Borgen, J. K., Burke, L. M., Ackerman, K. E., Blauwet, C., Constantini, N., Lebrun, C., Lundy, B., Melin, A. K., Meyer, N. L., Sherman, R. T., Tenforde, A. S., Torstveit, M. K., & Budgett, R. (2018). IOC consensus statement on Relative Energy Deficiency in Sport (RED-S): 2018 update. British Journal of Sports Medicine, 52(11), 687–697. https://doi.org/10.1136/bjsports-2018-099193
  • Mountjoy, M., Ackerman, K. E., Bailey, D. M., Burke, L. M., Constantini, N., Hackney, A. C., Heikura, I. A., Melin, A., Pensgaard, A. M., Stellingwerff, T., Sundgot-Borgen, J. K., Torstveit, M. K., Jacobsen, A. U., Verhagen, E., Budgett, R., Engebretsen, L., & Erdener, U. (2023). 2023 International Olympic Committee's (IOC) consensus statement on Relative Energy Deficiency in Sport (REDs). British Journal of Sports Medicine, 57(17), 1073–1097. https://doi.org/10.1136/bjsports-2023-106994
  • Nasser, S. N., Martins, C. B., Cavalcante, G. M., Marmirolli, F. A. D. P., Claudino, A. M., & Fidalgo, T. M. (2026). Prevalence of eating disorders and disordered eating in athletes: A systematic review and meta-analysis. British Journal of Sports Medicine, 60(14), 1051–1063. https://doi.org/10.1136/bjsports-2025-110678
  • National Institute for Health and Care Excellence. (2017; updated 2020). Eating disorders: Recognition and treatment (NICE guideline NG69). NICE. https://www.nice.org.uk/guidance/ng69/chapter/Recommendations
  • Nattiv, A., De Souza, M. J., Koltun, K. J., Misra, M., Kussman, A., Williams, N. I., Barrack, M. T., Kraus, E., Joy, E., & Fredericson, M. (2021). The Male Athlete Triad: A consensus statement from the Female and Male Athlete Triad Coalition Part I: Definition and scientific basis. Clinical Journal of Sport Medicine, 31(4), 335–348. https://doi.org/10.1097/JSM.0000000000000946
  • Reardon, C. L., Hainline, B., Aron, C. M., Baron, D., Baum, A. L., Bindra, A., Budgett, R., Campriani, N., Castaldelli-Maia, J. M., Currie, A., Derevensky, J. L., Glick, I. D., Gorczynski, P., Gouttebarge, V., Grandner, M. A., Han, D. H., McDuff, D., Mountjoy, M., Polat, A., ... Engebretsen, L. (2019). Mental health in elite athletes: International Olympic Committee consensus statement (2019). British Journal of Sports Medicine, 53(11), 667–699. https://doi.org/10.1136/bjsports-2019-100715
  • Reardon, C. L., Gouttebarge, V., Kroshus-Havril, E., Miller Aron, C., Bahr, R., Blauwet, C., Castaldelli-Maia, J. M., Cheng, C., Currie, A., Derevensky, J. L., Edwards, C., Fussek, S., Gorczynski, P., Grandner, M., Han, D. H., Hitchcock, M. E., Lu, F., Massey, A., McDuff, D., ... Hainline, B. (2026). Mental health in elite athletes: International Olympic Committee consensus statement (2026). British Journal of Sports Medicine, 60(15), 1083–1130. https://doi.org/10.1136/bjsports-2025-111514
  • Sundgot-Borgen, J. (1993). Prevalence of eating disorders in elite female athletes. International Journal of Sport Nutrition, 3(1), 29–40. https://doi.org/10.1123/ijsn.3.1.29
  • Sundgot-Borgen, J., & Torstveit, M. K. (2004). Prevalence of eating disorders in elite athletes is higher than in the general population. Clinical Journal of Sport Medicine, 14(1), 25–32. https://doi.org/10.1097/00042752-200401000-00005
  • Williams, N. I., De Souza, M. J., Misra, M., Nattiv, A., Joy, E., Barrack, M., Ricker, E. A., Gorrell, S., Koltun, K. J., O’Donnell, E., Mallinson, R. J., Salamunes, A. C. C., Woodruff, K., Fredericson, M., & Plessow, F. (2026). 2025 update to the Female Athlete Triad Coalition consensus statement Part 2: Clinical guidelines for screening, diagnosis, treatment, and return to play for adolescents and adults. Sports Medicine, 56(2), 375–427. https://doi.org/10.1007/s40279-025-02332-0