
==== Front
BMJ Open Sport Exerc Med
BMJ Open Sport Exerc Med
bmjosem
bmjsem
BMJ Open Sport & Exercise Medicine
2055-7647
BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR

10.1136/bmjsem-2024-002035
bmjsem-2024-002035
Original Research
1506
Exploring the relationship between low energy availability, depression and eating disorders in female athletes: a cross-sectional study
http://orcid.org/0000-0002-1334-4916
Halioua Robin 12robin.halioua@hin.ch

Wasserfurth Paulina 3paulina.wasserfurth@tum.de

Toepffer Désirée 3desiree.toepffer@tum.de

Claussen Malte Christian 14malte.claussen@bli.uzh.ch

http://orcid.org/0000-0002-9618-2069
Koehler Karsten 3karsten.koehler@tum.de

1 Research Group Sports Psychiatry, Center for Psychiatric Research, Department of Adult Psychiatry and Psychotherapy, Psychiatric University Clinic Zurich, University of Zurich, Zurich, Switzerland
2 Praxis Liebestrasse, Winterthur, Switzerland
3 Department Health and Sport Sciences, TUM School of Medicine and Health, Technical University of Munich, Munich, Germany
4 Clinic for Depression and Anxiety, Psychiatric Centre Muensingen, Muensingen, Switzerland
DrRobinHalioua; robin.halioua@hin.ch
None declared.

2024
28 8 2024
10 3 e00203518 7 2024
Copyright © Author(s) (or their employer(s)) 2024. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Abstract

Objective

This cross-sectional study aimed to investigate the role of low energy availability (LEA) in the interplay between depression and disordered eating/eating disorders (DE/EDs) among female athletes. The International Olympic Committee consensus statement on Relative Energy Deficiency in Sport (REDs) identified depression as both an outcome of LEA and a secondary risk factor for REDs. However, the direct link between LEA and depression has yet to be fully established.

Methods

We assessed 57 female athletes participating in weight-sensitive sports at different levels of competition training at least four times a week. Assessment was conducted using laboratory analyses, clinical interviews and the Patient Health Questionnaire-9 questionnaire. Participants were recruited through various channels, including German sports clubs, Olympic training centres, social media platforms and the distribution of flyers at competitions. Indicators of LEA were defined if at least two of the following three physiological indicators were present: menstrual disturbances, suppressed resting metabolic rate and suppressed thyroid hormones. Logistic and linear regression analysis were used to examine the relationship between LEA, depression and DE/ED.

Results

The lifetime prevalence of depressive disorders was 29.6%. 19% of the participants were diagnosed with an ED, and an additional 22.6% exhibited DE.

LEA was not significantly associated with either lifetime prevalence of depressive disorders or current depressive symptoms. However, a significant association was found between depression and DE/ED in terms of both lifetime prevalence and current depressive symptoms. DE/ED increased the probability of lifetime prevalence of depressive disorders by 34% (19%–49%) compared with normal eating behaviour.

Conclusion

We found no evidence that LEA is an independent factor for depression in female athletes. Its association with LEA and REDs appears to occur primarily in the presence of DE/ED.

Depression
Eating disorders
Energy expenditure
Athlete
Robert-Enke-Foundation Bavarian Equal Opportunity Scholarship Program AI use statement: since English is not the first author's first language, AI was used to make linguistic corrections. The use of AI was strictly confined to correcting language errors. No content interpretations or statistical analyses were conducted using AI.
==== Body
pmcWHAT IS ALREADY KNOWN ON THIS TOPIC

Depression is recognised as both a mental health outcome and a secondary risk factor of Relative Energy Deficiency in Sport (REDs). However, the relationship between low energy availability (LEA) and depression has not been fully established, as existing research on depression has neither been based on clinical data nor accounted for potential comorbid eating disorders (EDs).

WHAT THIS STUDY ADDS

We found no evidence that LEA is associated with the lifetime prevalence of depressive disorders or current depressive symptoms in participants with or without disordered eating/EDs (DE/EDs).

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

The association between LEA and depression appears to occur primarily in the presence of DE/ED. Therefore, further investigation is needed before conclusively classifying depression as a mental health outcome and a secondary risk factor of REDs.

Introduction

Low energy availability (LEA) refers to a state where energy intake is inadequate to sustain optimal functioning of all physiological systems after accounting for exercise-related energy expenditure.1 Problematic LEA represents a maladaptive response to severe and/or prolonged LEA that disrupts various body systems1 and is understood as the underlying cause of Relative Energy Deficit in Sports (REDs), a syndrome of impaired physiological and/or psychological functioning experienced by female and male athletes.24

The impact of problematic LEA on mental health and the influence of psychological factors are not as well understood as the physiological consequences. Eating disorders (EDs) and disordered eating (DE) behaviour are considered significant risk factors for LEA and REDs, as they can be both a cause and consequence and, thus, are considered a primary indicator of REDs.4 5 The International Olympic Committee (IOC) consensus statement also identifies depressive symptoms and affective disorders as adverse mental health outcomes associated with problematic LEA and REDs, considering clinically diagnosed depression a secondary indicator of REDs.4 A narrative review by a subgroup of the IOC consensus on REDs discussed the temporal relationship between exposure to LEA and mental health outcomes.5 It was suggested that mood changes, fatigue and psychological conflict may be expected within days and weeks, whereas the development of depressive symptoms may require longer exposure to LEA.5

However, the current literature on the association between LEA and depressive symptoms and/or clinically diagnosed depression is limited. To our knowledge, only three studies have investigated the relationship between LEA and depressive symptoms.68 Although these studies provide valuable insights, their findings may be subject to interpretation due to methodological limitations. Ackerman et al (2019) reported a positive association between LEA and depressive symptoms in a large-scale (n=1000) cross-sectional study. However, LEA assessment relied on DE/ED screening questionnaires, which may have primarily captured the association between DE/ED and depressive symptoms. Furthermore, diagnoses of ED and depression were based solely on questionnaires without clinical confirmation from diagnostic interviews. In the study by Rogers et al (2021), LEA was established through objective laboratory analyses and defined as a low resting metabolic rate (RMR). And although clinical interviews were conducted, findings from these were not included in the final analysis, which instead solely relied on questionnaire responses that considered depressive symptoms only. Mathisen et al (2019) conducted a longitudinal study and found an increase in depressive symptoms in female fitness athletes preparing for a competition.8 However, the relationship between LEA and depressive symptoms was not directly evaluated, and the assessment of depressive symptoms relied solely on questionnaires. Furthermore, the increased depressive symptoms remained well below clinical significance. Neither study accounted for potential comorbid DE/ED, which could have influenced the observed relationship.

Investigating the association between LEA and depression requires consideration of comorbid DE/ED, which are both a possible cause of LEA and share a well-established comorbidity with depression in non-athletic populations.9 Preliminary findings suggest this comorbidity may also exist in the athletic context.10 In response to the IOC’s call for more studies to explore the reciprocal function of psychological variables with LEA,2 5 we aimed to examine the role of LEA in the interplay between DE/ED and depression more closely. As the study was exploratory in nature, no specific hypotheses were formulated. This allowed for an open investigation into the association between LEA and depression in athletes accounting for DE/ED, as well as an exploration of whether LEA increases the risk of depression in those with ED/DE.

Methods

Study design, setting and participants

This cross-sectional study examined physiological indicators of problematic LEA, depression and DE/ED in female athletes using state-of-the-art assessments: laboratory analyses, clinical interviews and questionnaires. Laboratory analyses were performed at our core facility at the Technical University of Munich, and clinical interviews were conducted via telemedicine by a trained psychiatrist. Participants were recruited through various channels, including German sports clubs, Olympic training centres, social media platforms and the distribution of flyers at competitions. The study was approved by the Technical University Munich’s Ethical Review Board (registration number: 347/21 S) and adhered to the Declaration of Helsinki. All participants provided written informed consent. Recruitment and data collection lasted from November 2021 until April 2023.

Eligibility criteria

The eligibility criteria required participants to be female athletes aged 18–39 years engaged in weight-sensitive sports (cycling, triathlon, long-distance running, swimming, cross-country skiing, biathlon and ballet) who trained ≥4 times weekly. Exclusion criteria encompassed the use of hormonal contraceptives, pregnancy, breastfeeding or chronic illness or acute injury at the time of potential inclusion.

Clinical assessment of depression and eating disorders

The clinical assessment comprised clinical interviews conducted by a board-certified psychiatrist via a secure online video platform, along with the completion of questionnaires. All interviews were conducted following the laboratory analysis and lasted approximately 45 min.

Depression

The clinical interview was conducted using Module A of the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), Clinician Version (SCID-5-CV). The SCID-5-CV is a semistructured interview for diagnosing mental disorders according to DSM-5. Each DSM-5 criterion is associated with specific interview questions to assist the interviewer in assessing the criterion. Both current and past depressive symptoms were assessed. This allowed for an evaluation of both current depressive symptomatology and the lifetime prevalence of depressive disorder.

The Patient Health Questionnaire-9 (PHQ-9) questionnaire was administered in addition to the clinical interview. The PHQ-9 corresponds to the Depression Module of PHQ (PHQ-D) and consists of nine questions related to depressive symptoms. It was developed as a screening instrument for diagnosing depression for routine use in somatic medical settings. To minimise potential bias, the interviewer was kept blinded to the participants’ questionnaire outcomes throughout the data collection process.

Eating disorders and disordered eating

The clinical interview was conducted using the German version of the Eating Disorder Examination by Fairburn and Cooper, a structured interview for assessing the specific psychopathology of ED.11 12 It assesses eating behaviour across four scales: ‘Restraint’, ‘Eating Concern’, ‘Weight Concern’ and ‘Shape Concern’. In addition to these scales, 14 additional diagnostic items enable clinical diagnosis of ED.

DE was only identified in cases where clear pathological behaviours and/or distorted body image were present, without meeting the diagnostic criteria for an ED. The assessment took into account the context of sports and the resulting demands on eating behaviour.10 The study set a high threshold for DE, placing it closer to ED than to normal, functional eating behaviour.

Assessment of LEA

Classification of menstrual status

Participant menstrual status was determined retrospectively via phone interviews, with information collected on menstruation within the last 3 months. Based on their cycle lengths, participants were classified as eumenorrheic (28±7 days), oligomenorrheic (cycle length >35 days), primary amenorrheic (no menarche after 15 years of age) or secondary amenorrheic (absence of ≥3 consecutive periods or no menses for ≥90 days).13

Laboratory examinations

For participants with regular or irregular periods, laboratory examinations were scheduled during the early follicular phase (days 3–5 after period onset), while for those with amenorrhoea, the examination was conducted at the earliest opportunity. Prior to measurement, participants refrained from exercise, caffeine and alcohol for 24 hours and arrived at the laboratory after a 12-hour overnight fast in a rested state.

Anthropometry and body composition

Participants wore only underwear or tightly fitting swimwear for measurements. Height was measured to the nearest 0.1 cm using a stadiometer (seca 217, SECA, Hamburg, Germany). Body composition was assessed using air displacement plethysmography (ADP; BOD POD, COSMED, Rome, Italy), with body mass being measured using a built-in scale. The device was calibrated before each measurement session. Participants wore a swim cap and remained still during the measurement. Fat mass and fat-free mass were determined using the Siri equation based on ADP-measured body density.14

Resting metabolic rate

RMR was assessed using open-circuit indirect calorimetry with a canopy (Q-NRG, COSMED, Rome, Italy) at room temperature (20–25 °C).15 The device was calibrated with external gas and room air prior to each measurement, and ethanol burning tests were conducted monthly. Participants rested supine for ~10 min before data collection, which lasted at least 30 min. RMR (kcal/day) was calculated using the Weir equation from steady-state data, which was defined as a coefficient of variation ≤10% in steady-state oxygen consumption and carbon dioxide production.16

Blood samples

Blood samples were drawn from the antecubital vein and free triiodothyronine (fT3) was analysed by an external laboratory (Labor Becker MVZ GbR, Munich, Germany).

Classification of LEA

Following the 2023 IOC REDs consensus statement,4 cases of LEA were operationally defined if at least two of the following three physiological indicators were present: menstrual disturbances (amenorrhoea and oligomenorrhoea), suppressed RMR (<30 kcal/kg/fat-free mass (FFM)) and suppressed thyroid hormones (fT3 below the IOC-recommended laboratory reference ranges: within or below the lowest quartile). Suppressed RMR was selected as an indicator due to its association with LEA,17 without being directly related to DE/ED.

Statistical analysis

Sample size

Based on the existing literature on LEA prevalence,7 a sample size of 50 was determined, allowing for approximately 30% of LEA cases. To account for a 15%–20% dropout rate between measurements, we aimed for a total of 60 participants.

Descriptive analysis

We calculated frequencies, means and 95% CIs for demographic data and assessment instruments.

Main analysis

To examine the role of LEA in the interplay between depression and DE/ED, logistic regression and linear regression analyses were conducted. The lifetime prevalence of diagnosed depressive disorder was chosen as the outcome for logistic regression, while the outcome for linear regression was current depressive symptoms according to PHQ-9 scores. In both analyses, LEA was used as a predictor variable, with DE/ED serving as a control variable. Given the high threshold applied for DE in this study, DE and ED were examined as a combined variable. In both regression analyses, DE and ED were also inserted as separately coded variables, aiming to explore the potential differences in outcomes compared with analyses combining DE/ED. Interactions between LEA and DE/ED were examined by the likelihood ratio test.

The OR and Average Marginal Effects (AME) were calculated to interpret the effect of LEA and DE/ED. In this case, AMEs reflect the average change in predicted probability for each observation in factorial variables while holding the other variable constant.18 Predicted probabilities for lifetime depressive disorder and predicted margins for depressive symptoms were also calculated.

Missing data

We conducted Multiple Imputation by Chained Equations to analyse 10 multiply-imputed datasets, addressing incomplete variables using the mi impute command in Stata. Rubin’s combination rules were applied to adjust estimates and their standard errors for imputation uncertainty. Additionally, we performed a complete case analysis for comparison.

Statistical software

Stata Statistical Software: Release V.16.1. College Station, Texas was used to analyse the data.

Patient and public involvement

As our study focuses on female athletes in weight-sensitive sports, no patients were involved. Participants were first engaged in the study during the recruitment phase. Some participants subsequently recruited additional participants from their own circles. Results of laboratory measurements and clinical interviews were shared with participants. Assistance was provided to one participant in arranging therapy for her ED. Results and new insights will be provided to the participants.

Equity, diversity and inclusion statement

Our study is on female athletes in weight-sensitive sports. The author group consists of junior, mid-career and senior researchers from different disciplines in Switzerland and Germany. Participants were recruited through various channels to ensure a broad spectrum of socioeconomic backgrounds. Nevertheless, our patient cohort predominantly represents the middle to upper socioeconomic strata. Clinical interviews were conducted via telemedicine to enhance accessibility.

Results

Participants

A total of 93 athletes provided written informed consent to participate in the study, with 57 completing on-site measurements. Demographics, laboratory assessments and clinical diagnoses of depressive disorder and ED are presented in table 1.

Table 1 Descriptive characteristics of the study sample

Variables	Controls (39)	LEA (14)	Total (57)	
Age (years), mean±SD	28.4±5.7	27.3±4.2	27.9±5.3	
Height (cm), mean±SD	168.7±5.8	169.5±5.5	168.8±5.5	
Weight (kg), mean±SD	62.0±6.5	55.8±5.3	60.5±7.1	
BMI (kg/m²), mean±SD	21.7±1.7	19.4±1.3	21.2±2.1	
Education (%)				
 Academic	61.5	100	70.2	
 High school	25.6	0	21.1	
 Non-academic	2.6	0	1.8	
 Secondary school	10.3	0	7	
Training level (%)				
 National team	7.7	0	5.3	
 Professional	5.1	14.3	7	
 Club	41.0	21.4	36.8	
 Individual with competition	38.5	57.1	43.9	
 Individual without competition	7.7	7.1	7	
Menstrual status (%)				
 Amenorrhoea	7.7	64.3	21.1	
 Oligomenorrhoea	20.5	21.4	21.1	
 Eumenorrhoea	71.8	14.3	57.9	
RMR (kcal/kg FFM), mean±SD	30.9±2.6	27.4±1.7	30.0±2.8	
fT3 (ng/L), mean±SD	2.7±0.3	2.2±0.5	2.6±0.4	
Depressive disorder (%)				
 No depressive disorder	71.1	61.5	68.5	
 Mild depressive episode	m	m	1.9	
 History of one depressive episode in the past	10.5	15.4	11.1	
 Recurrent depressive disorder, current episode in remission	15.8	15.4	14.8	
 Recurrent depressive disorder, current episode mild	2.6	7.7	3.7	
 Lifetime prevalence	28.9	38.5	29.6	
PHQ-9 (0–28), mean±SD	5.8±3.6	5.8±3.6	5.8±3.5	
Eating disorder (%)				
 Normal eating behaviour	60.5	46.2	58.5	
 Disordered eating	23.7	23.1	22.6	
 Bulimia nervosa	2.6	7.7	3.8	
 Anorexia nervosa	0.0	7.7	1.9	
 Bulimia nervosa of low frequency/limited duration	7.9	0.0	5.7	
 Atypical anorexia nervosa	2.6	7.7	3.8	
 Purging disorder	2.6	7.7	3.8	
Eating disorder or disordered eating (%)	39.5	53.8	41.5	
Unless otherwise specified, mean and standard deviationSD are reported for continuous variables, and proportions are reported for categorical data.

* Depression Module of the Patient Health Questionnaire.

BMIbody mass indexfT3free triiodothyronineLEALow energy availabilitym, missingPHQ-9Patient Health Questionnaire-9RMRresting metabolic rate

Missing data

The assessment and clinical diagnosis of depression and DE/ED were completed for 53 (92.9%) participants. Four participants did not participate in the clinical interviews, one of whom did not provide a blood sample. Three participants reported taking thyroid medication, which was treated as missing values for low fT3 and RMR <30 kcal/kg/FFM. Non-participation in clinical interviews was treated as missing values for lifetime depressive disorder and DE/ED. The missing blood analysis was treated as a missing value for low fT3. There were a total of five missing values for LEA/Controls, four for lifetime depressive disorders and four for DE/ED. Multiple imputation and complete case analysis lead to similar results; therefore, we present only the former.

Main results

Logistic and linear regressions

Table 2 summarises the results of the logistic regression analysis on lifetime prevalence of depressive disorders and the linear regression analysis on current depressive symptoms according to the PHQ-9. LEA was not statistically associated with either lifetime prevalence of depressive disorders or current depressive symptoms. DE/ED showed statistical significance in both regression analyses. There was no significant interaction between LEA and DE/ED.

Table 2 Results from the logistic and linear regressions

Logistic regression on lifetime prevalence of depressive disorders	
Variable	OR	(95% CI)	SE	P value	
LEA*	1.3	(0.29 to 5.75)	0.99	0.73	
DE/ED†	7.59	(1.93 to 29.82)	5.29	<0.01	
Constant	0.14	(0.05 to 0.43)	0.08	<0.01	
Linear regression on current depressive symptoms according to the PHQ-9	
Variable	β	(95% CI)	SE	P value	
LEA*	−0.23	(−2.33 to 1.87)	1.04	0.83	
DE/ED†	3.09	(1.33 to 4.85)	0.88	<0.01	
Intercept	4.62	(3.42 to 5.82)	0.6	<0.001	
* Low energy availability.

† Disordered Eating / Eating disorder.

DE/EDdisordered eating/eating disorderLEAlow energy availabilityPHQ-9Patient Health Questionnaire-9

DE/ED increased the probability of lifetime prevalence of depressive disorders by 34% (19%–49%) compared with normal eating behaviour. The predicted probabilities for female athletes with DE/ED and concurrent LEA were 57.9% (25.3%–90.5%) and 51.5% (27.7%–75.3%) for those without concurrent LEA. The difference of 6.4% was statistically not significant (p=0.73). Those with a DE/ED and concurrent LEA yielded a predicted PHQ-9 score of 7.5 (5.5–9.4), and those without concurrent LEA yielded a score of 7.7 (6.2–9.2). The difference was statistically not significant (p=0.71). Figure 1 presents the predicted probabilities for lifetime prevalence of depressive disorders, while figure 2 displays the predicted scores of the PHQ-9.

Figure 1 Predicted probabilities for lifetime prevalence of depressive disorders. For those without LEA (controls), the predicted probabilities for lifetime prevalence of depressive disorders were 12.4% (0.2%–24.5%) with normal eating behaviour and 51.5% (27.7%–75.3%) with DE/ED. For those with LEA and normal eating behaviour, the probability was 14.3% (0%–29%) compared with 57.9% (25.3%–90.5%) with concurrent DE/ED. Whiskers represent the 95% CI. DE/ED, disordered eating/eating disorder; LEA, low energy availability.

Figure 2 Predicted scores of the PHQ-9. For those without LEA (controls), the predicted PHQ-9 score was 4.6 (3.4–5.8) with normal eating behaviour and 7.7 (6.2–9.2) with DE/ED. For those with LEA and normal eating behaviour, the predicted score was 4.4 (2.4–6.4) compared with 7.5 (5.5–9.4) with concurrent DE/ED. Whiskers represent 95% CI. DE/ED, disordered eating/eating disorder; LEA, low energy availability; PHQ-9, Patient Health Questionnaire-9.

Discussion

The aim of this study was to examine the role of LEA in the interplay between depression and DE/ED in an exploratory investigation. Our findings indicate that the previously established association between LEA and depression may be mediated by comorbid DE/ED. We found no evidence that indicators of LEA are associated with the lifetime prevalence of depressive disorders or current depressive symptoms in participants regardless of DE/ED status. However, a strong association was found between DE/ED and depression, in terms of both lifetime prevalence and current symptoms.

Energy availability

LEA alone did not increase the prevalence of lifetime depressive disorder or current depressive symptoms in those with or without DE/ED. Our results indicate that DE/ED as an underlying cause of LEA has a more significant impact on depression than LEA itself. LEA may result from intentional or unintentional undereating.19 In the case of intentional undereating, it is important to distinguish DE/ED from controlled eating for performance enhancement and other functional reasons. Although both can result in LEA, they are driven by different motivational factors and seem to play a decisive role in whether LEA is associated with depression or not. Given that DE/EDs typically have a long course and that LEA was not associated with depression in those with DE/EDs, it seems unlikely that there is a temporal relationship between prolonged exposure to LEAs and the development of depressive symptoms. Taken together, depression as a direct mental health outcome resulting from a physiological causal chain involving exposure to LEA seems unlikely.

This contradicts the findings of previous studies by Ackermann6 and Rogers,7 who found a positive association between LEA and depression. The difference in results with Ackermann’s study can be attributed to their use of the DE/ED screens as a marker for LEA, essentially confirming the association between LEA and DE/ED observed in our study. Our study’s more precise operationalisation of LEA, which incorporates several physiological indicators such as low RMR, low T3, and menstrual dysfunction, may explain the difference from Rogers’ findings, which identified LEA only using low RMR. Furthermore, Rogers’ study did not consider potential comorbid DE/ED in their analysis. Our results are in line with those in the non-athletic context, which found no correlation between nutritional status and depressive symptoms in individuals with anorexia nervosa.20 21

Depression

When compared with the general population, participants in the present study showed higher levels of depression, both in terms of lifetime prevalence and current symptoms. The overall lifetime prevalence was 29.6%, surpassing the 10%–15% range observed in the general population.22 23 Additionally, the average PHQ-9 score was elevated at 5.8, exceeding the normative value of 3.624 and aligning with scores observed in representative patient samples from general hospitals in Germany.25 The findings are consistent with Schaal et al (2001),26 who found a lifetime prevalence of 30% in women who participate in aesthetic sports. Only 5.6% of the participants in our sample exhibited clinically diagnosed depression during the clinical interviews. Therefore, the score from the PHQ-9 should be interpreted as mainly subclinical. However, in light of the higher lifetime prevalence, it may have clinical significance as it may indicate a higher overall burden of depressive symptoms.

DE/ED

The prevalence of ED (19%) and DE (22.6%) in our sample was higher than that of the general population,27 which is consistent with the findings of similar studies.28 29 The majority (13.3%) classified as Other Specified Feeding and Eating Disorder. Participants with LEA primarily displayed atypical anorexia nervosa and purging disorder, whereas those with adequate EA showed bulimia of low frequency/limited duration. Notably, individuals with LEA were twice as likely to have an ED (30.8%) when compared with those with adequate EA (15.7%); however, DE occurred at the same frequency (23 %) in both groups. This underscores the necessity to consider LEA and DE/ED as distinct entities. While DE/ED is a prevalent underlying cause of LEA, a notable proportion (46.2%) of those with LEA did not exhibit an associated DE/ED. Furthermore, 40% of individuals without LEA still exhibited a DE/ED.

Interplay between depression and DE/ED

A strong association was found between DE/ED and depression, in terms of both lifetime prevalence and current depressive symptoms. The presence of DE/ED increased the likelihood of experiencing depression at least once in one’s lifetime by 35% and current depressive symptoms by three points on the PHQ-9 scale. Participants with DE/ED had a lifetime prevalence of depressive disorder ranging from 51.5% (without concurrent LEA) to 57.9% (with concurrent LEA). This is in line with findings in the general and athletic population, which indicate a high rate of comorbidity between depression and ED, suggesting a bidirectional relationship.10 27 30 Our findings indicate that the previously established association between LEA and depression may be mediated by comorbid DE/ED. Additionally, this comorbidity may explain the higher prevalence of depression in weight-sensitive samples compared with other sports.26 31

Clinical implications and future research

REDs is conceptualised as an etiological syndrome based on problematic LEA. Within this framework, depression is both considered a mental health outcome and a secondary risk factor of REDs.4 From a theoretical perspective, there is insufficient evidence to conclude that LEA is in fact a causative factor for depression. Our data demonstrate that the association between LEA and depression appears to occur primarily in the presence of DE/ED. Therefore, further investigation is needed before conclusively classifying depression as a mental health outcome and a secondary risk factor of REDs.

Limitations

First, while ensuring that participants represented the primary target audience of REDs, our inclusion criteria of a minimum of four training sessions per week may have excluded participants with severe depressive symptoms, as they may not have been able to maintain such a sports regimen at this frequency. Second, despite our efforts to represent a broad socioeconomic spectrum, the generalisability is limited by the fact that the participants are predominantly from middle to upper socioeconomic strata with educational backgrounds. Third, the use of state-of-the-art methodology involving lab measurements and clinical interviews resulted in a relatively small sample size. As such, caution is necessary when interpreting effect sizes as they may be susceptible to inflation. Fourth, by using cross-sectional data, we are precluded from drawing causal conclusions from the associations we found.

Acknowledgements

We would like to thank Mona Saller, Lena Fladerer, Josephine Ruf and Vanessa Corrêa Mascarenhas for their assistance. We would further like to thank all volunteers who participated in the study.

Data availability statement

Data are available upon reasonable request.

Funding: This study was funded by the Robert-Enke-Foundation. PW is supported by a fellowship by the Bavarian Equal Opportunity Scholarship Program.

Data availability free text: The datasets used and/or analysed during the current study are available from the first author on reasonable request.

Patient consent for publication: Consent obtained directly from patient(s).

Ethics approval: This study involves human participants and was approved by the Ethical Review Board of the Technical University Munich (registration number: 347/21 S). Participants gave informed consent to participate in the study before taking part.

Provenance and peer review: Not commissioned; externally peer reviewed.

Patient and public involvement: Patients and/or the public were involved in the design, conduct, reporting or dissemination plans of this research. Refer to the Methods section for further details.
==== Refs
References

1 Redman LM Loucks AB Menstrual disorders in athletes Sports Med 2005 35 747 55 10.2165/00007256-200535090-00002 16138785
2 Mountjoy M Sundgot-Borgen J Burke L et al The IOC consensus statement: beyond the Female Athlete Triad--Relative Energy Deficiency in Sport (RED-S) Br J Sports Med 2014 48 491 7 10.1136/bjsports-2014-093502 24620037
3 Mountjoy M Sundgot-Borgen J Burke L et al International Olympic Committee (IOC) Consensus Statement on Relative Energy Deficiency in Sport (RED-S): 2018 Update Int J Sport Nutr Exerc Metab 2018 28 316 31 10.1123/ijsnem.2018-0136 29771168
4 Mountjoy M Ackerman KE Bailey DM et al 2023 International Olympic Committee’s (IOC) consensus statement on Relative Energy Deficiency in Sport (REDs) Br J Sports Med 2023 57 1073 98 10.1136/bjsports-2023-106994 37752011
5 Pensgaard AM Sundgot-Borgen J Edwards C et al Intersection of mental health issues and Relative Energy Deficiency in Sport (REDs): a narrative review by a subgroup of the IOC consensus on REDs Br J Sports Med 2023 57 1127 35 10.1136/bjsports-2023-106867 37752005
6 Ackerman KE Holtzman B Cooper KM et al Low energy availability surrogates correlate with health and performance consequences of Relative Energy Deficiency in Sport Br J Sports Med 2019 53 628 33 10.1136/bjsports-2017-098958 29860237
7 Rogers MA Appaneal RN Hughes D et al Prevalence of impaired physiological function consistent with Relative Energy Deficiency in Sport (RED-S): an Australian elite and pre-elite cohort Br J Sports Med 2021 55 38 45 10.1136/bjsports-2019-101517 33199358
8 Mathisen TF Heia J Raustøl M et al Physical health and symptoms of relative energy deficiency in female fitness athletes Scand J Med Sci Sports 2020 30 135 47 10.1111/sms.13568 31593622
9 Godart NT Perdereau F Rein Z et al Comorbidity studies of eating disorders and mood disorders. Critical review of the literature J Affect Disord 2007 97 37 49 10.1016/j.jad.2006.06.023 16926052
10 Tan JOA Calitri R Bloodworth A et al Understanding Eating Disorders in Elite Gymnastics Clin Sports Med 2016 35 275 92 10.1016/j.csm.2015.10.002 26832977
11 Fairburn C Wilson G Binge eating: nature, assessment, and treatment J Nerv Ment Dis 1993 183
12 Hilbert A Tuschen-Caffier B Karwautz A et al Eating Disorder Examination-Questionnaire Diagn 2007 53 144 54 10.1026/0012-1924.53.3.144
13 Elliott-Sale KJ Minahan CL de Jonge XAKJ et al Methodological Considerations for Studies in Sport and Exercise Science with Women as Participants: A Working Guide for Standards of Practice for Research on Women Sports Med 2021 51 843 61 10.1007/s40279-021-01435-8 33725341
14 Siri WE Body composition from fluid spaces and density: analysis of methods. 1961 Nutrition 1993 9 480 91 8286893
15 Compher C Frankenfield D Keim N et al Best practice methods to apply to measurement of resting metabolic rate in adults: A systematic review J Am Diet Assoc 2006 106 881 903 10.1016/j.jada.2006.02.009 16720129
16 Weir JBDB New methods for calculating metabolic rate with special reference to protein metabolism J Physiol 1949 109 1 9 10.1113/jphysiol.1949.sp004363 15394301
17 Melin A Tornberg ÅB Skouby S et al Energy availability and the female athlete triad in elite endurance athletes Scand J Med Sci Sports 2015 25 610 22 10.1111/sms.12261 24888644
18 Long JS Freese J Regression Models for Categorical Dependent Variables Using Stata, 3rd Edn College Station, Texas Stata Press 2014
19 Wasserfurth P Palmowski J Hahn A et al Reasons for and Consequences of Low Energy Availability in Female and Male Athletes: Social Environment, Adaptations, and Prevention Sports Med Open 2020 6 44 10.1186/s40798-020-00275-6 32910256
20 Mattar L Huas C Godart N et al Relationship between affective symptoms and malnutrition severity in severe Anorexia Nervosa PLoS ONE 2012 7 e49380 10.1371/journal.pone.0049380 23185320
21 Mattar L Thiébaud M-R Huas C et al Depression, anxiety and obsessive-compulsive symptoms in relation to nutritional status and outcome in severe anorexia nervosa Psychiatry Res 2012 200 513 7 10.1016/j.psychres.2012.04.032 22703719
22 Streit F Zillich L Frank J et al Lifetime and current depression in the German National Cohort (NAKO) World J Biol Psychiatry 2023 24 865 80 10.1080/15622975.2021.2014152 34870540
23 Nixdorf I Frank R Hautzinger M et al Prevalence of Depressive Symptoms and Correlating Variables Among German Elite Athletes J Clin Sport Psychol 2013 7 313 26 10.1123/jcsp.7.4.313
24 Gräfe K Zipfel S Herzog W et al Screening psychischer Störungen mit dem «Gesundheitsfragebogen für Patienten (PHQ-D)»: Ergebnisse der deutschen Validierungsstudie. [Screening for psychiatric disorders with the Patient Health Questionnaire (PHQ) Results from the German validation study.] Diagn 2004 50 171 81 10.1026/0012-1924.50.4.171
25 Rief W Nanke A Klaiberg A et al Base rates for panic and depression according to the Brief Patient Health Questionnaire: a population-based study J Affect Disord 2004 82 271 6 10.1016/j.jad.2003.11.006 15488257
26 Schaal K Tafflet M Nassif H et al Psychological balance in high level athletes: gender-based differences and sport-specific patterns PLoS One 2011 6 e19007 10.1371/journal.pone.0019007 21573222
27 Keski-Rahkonen A Mustelin L Epidemiology of eating disorders in Europe: prevalence, incidence, comorbidity, course, consequences, and risk factors Curr Opin Psychiatry 2016 29 340 5 10.1097/YCO.0000000000000278 27662598
28 Sundgot-Borgen J Torstveit MK Prevalence of eating disorders in elite athletes is higher than in the general population Clin J Sport Med 2004 14 25 32 10.1097/00042752-200401000-00005 14712163
29 Byrne S McLean N Elite athletes: effects of the pressure to be thin J Sci Med Sport 2002 5 80 94 10.1016/s1440-2440(02)80029-9 12188089
30 Puccio F Fuller-Tyszkiewicz M Ong D et al A systematic review and meta-analysis on the longitudinal relationship between eating pathology and depression Int J Eat Disord 2016 49 439 54 10.1002/eat.22506 26841006
31 Gorczynski PF Coyle M Gibson K Depressive symptoms in high-performance athletes and non-athletes: a comparative meta-analysis Br J Sports Med 2017 51 1348 54 10.1136/bjsports-2016-096455 28254747
