Personality Traits Associated with the Risk of Exercise Dependence in Ultraendurance Athletes: A Cross-Sectional Study

ise dependence (ED) is common in endurance athletes and can lead to physical and psychological distress with various health effects. We designed a prospective cross-sectional study to investigate the personality traits associated with ED among ultraendurance athletes. A total of 507 participants (41.6 (9.8) years, men: 73.7%) completed (1) a screening questionnaire about sociodemographic data, sporting habits, and healthcare data, (2) the Exercise Dependence Scale- Revised (EDS-R, 21 items scored from 1 (never) to 6 (always), 7 subscales), (3) the Big Five Inventory (BFI), and (4) 2 items of the SCOFF (Sick-Control-One Stone-Fat-Food) questionnaire regarding possible eating disorders. Based on the EDS-R scores, 37 (7.3%) participants were at risk for ED (scores 5/6 on 3 subscales), 366 (72.2%) were nondependent but symptomatic (scores 3/6 on 3 subscales), and 104 (20.5%) were asymptomatic. Participants with ED had a greater training volume and a higher prevalence of possible eating disorders. A higher level of neuroticism was associated with increased EDS-R scores (r = 0.294;p< 0.001), with signi cantly higher scores in the ED group (F =

3 subscales), 366 (72.2%) were nondependent but symptomatic (scores 3/6 on 3 subscales), and 104 (20.5%) were asymptomatic. Participants with ED had a greater training volume and a higher prevalence of possible eating disorders. A higher level of neuroticism was associated with increased EDS-R scores (r = 0.294;p< 0.001), with signi cantly higher scores in the ED group (F = 14.50,p< 0.001). The association between neuroticism and ED was not moderated by the presence of eating disorders. These ndings will help to screen ultraendurance athletes at risk for ED and optimize their care. Keywords:ultraendurance sport; exercise addiction; personality; neuroticism 1. Introduction The practice of regular physical activity, including exercise, has numerous bene ts for mental and physical health [1,2]. Thus, public health authorities and experts encourage increased participation in physical activity and exercise for the prevention of chronic diseases associated with physical inactivity and sedentary behavior [3]. Additionally, regular physical activity is key for the prevention and/or treatment of mental health problems [4–6], and exercise-based interventions have been proposed as a treatment for addiction and substance use disorders [7–9]. However, excessive exercise needs to be distinguished from exercise that occurs at a high frequency and should be distinct from a healthy habit in that excessive exercise occurs to the detriment of physical health, quality of life, and social responsibility [10]. Indeed, exercise dependence (ED) is characterized by behavioral, psychological and physiological symptoms similar to those of substance use disorders or behavioral addictions [11,12] and by engagement in any form of exercise despite negative consequences [13]. ED has been described as a “positive” addiction due to the large bene ts of physical activity and regular exercise [14]; however, there is substantial evidence that ED should be considered a behavioral addiction with detrimental effects and potential coexisting disorders [15]. Int. J. Environ. Res. Public Health2023,20, 1042.

Int. J. Environ. Res. Public Health2023,20, 1042 2 of 17 There is no clear consensus on the de nition of ED, and the most accepted characteri- zation is based on the criteria for common substance use disorders [16,17]. According to the Diagnostic and Statistical Manual of Mental Disorders (5th edition, DSM-5), substance use disorder is characterized by the following symptoms: tolerance (increased amount of exercise to get the desired effect), withdrawal (negative effects in the absence of exercise), loss of control (unsuccessful efforts to control exercise), intention effects (exercising longer or at a higher intensity than intended), duration (a great deal of time dedicated to exercise), a reduction in other activities, hazardous behaviors (harmful exercise for mental and physi- cal health), failure to ful ll obligations, social/interpersonal problems, and continuance (continuing to exercise despite detrimental effects) [18]. However, unlike gambling, ED has not yet been designated a “nonsubstance-related disorder” in the “Substance-related and Addictive Disorders” category of the DSM-5, mostly due to insuf cient peer-reviewed evidence [19]. Additionally, substance dependence is characterized in the International Classi cation of Diseases (11th edition, ICD-11) by the presence of two or more of three composite guidelines among (1) impaired control over substance use, (2) substance use becoming an increasing priority in life, and (3) physiological features as manifested by tolerance, withdrawal symptoms, or repeated use [20]. Additionally, the ICD-11 included non-substance addiction disorders in the grouping of impulse control disorders, which should be de ned by the repeated failure to resist an impulse, drive, or urge to perform an act that is rewarding to the person (at least in the short term), despite longer-term harm [21]. However, ED is not recognized in the ICD-11. There is a large body of literature supporting that ED can be either a primary addiction or a secondary coexisting disorder associated with other mental or medicalconditions [ , especially eating disorders [24,25]. Recent knowledge suggests that the etiology of ED differs depending on the presence or absence of an eating disorder [26], but there is no valid or common conceptualization of compulsory exercise in individuals with eating disorders

ED can be either a primary addiction or a secondary coexisting disorder associated with other mental or medicalconditions [ , especially eating disorders [24,25]. Recent knowledge suggests that the etiology of ED differs depending on the presence or absence of an eating disorder [26], but there is no valid or common conceptualization of compulsory exercise in individuals with eating disorders [27]. Additionally, recent evidence has shown that the risk of ED is associated with coexisting mental disorders such as anxiety and depression or obsessive-compulsive disorders [28] as in other behavioral addictions, but there is uncertainty about the causal factors that may underlie ED [29]. If ED is not restricted to the consideration of high-volume training, it may contribute to extreme volumes and/or intensities of long-term exercise training and be associated with health issues. The consequences of overtraining are widespread, negatively affecting several physiological systems, including neuroendocrine, immunological, neuromuscular and skeletal, and cardiovascular disorders [30,31]. Overtraining could also result in several negative psychological disturbances, such as increased depression, low motivation, anger, and eating disorders [30,31]. Interestingly, Golshani et al. recently conducted a compara- tive study exploring the prevalence of mental health issues among heavy (18–22 h/week) and light exercisers (1–6 h/week) [32]. They found that heavy exercising might be asso- ciated with more mental health issues, and that lower mental toughness and more sleep disturbances could predict worse mental health issues in heavy exercisers. The prevalence and characteristics of ED have been largely studied in endurance sports due to the high volume and intensity of training, with high susceptibility to devel- oping ED [33–35]. Among endurance sports, ultraendurance sports are de ned as events lasting at least 6 h [36], and the sports concerned are mostly trail running, ultramarathons, ultracycling, Ironman triathlons, open-water swimming, and cross-country skiing. The practice of ultraendurance sports has become increasingly popular in recent years, mostly for recreational athletes [37–39]. Additionally, there is extensive evidence that ultraen- durance races and regular ultraendurance training may lead to long-term health problems, including cardiovascular, respiratory, musculoskeletal, renal, immunological, gastrointesti- nal, neurological, and skin problems [40,41]. As ED is characterized by

triathlons, open-water swimming, and cross-country skiing. The practice of ultraendurance sports has become increasingly popular in recent years, mostly for recreational athletes [37–39]. Additionally, there is extensive evidence that ultraen- durance races and regular ultraendurance training may lead to long-term health problems, including cardiovascular, respiratory, musculoskeletal, renal, immunological, gastrointesti- nal, neurological, and skin problems [40,41]. As ED is characterized by continuous exercise despite signi cant and interfering physical problems that are likely to have been caused by intense exercising, it is essential to better understand the continuum and psychological

Int. J. Environ. Res. Public Health2023,20, 1042 3 of 17 and/or interactional mechanisms underlying ED in ultraendurance athletes to prevent both mental health issues and related health problems [42]. ED is a complex and multifaceted behavior encompassing several psychological and physiological characteristics and resulting from the interaction of a multitude of personal and situational factors [43]. To better understand the behavioral tendencies and personality features underlying addiction, numerous studies have explored their association with personality traits. Personality is de ned by an individual's set of stable traits and behaviors, which are the expression of their ways of living, relating to others and perceiving their own person in association with their environment [44]. Personality aspects are grouped into personality traits, which are relatively enduring patterns of thoughts, feelings, and behaviors that re ect a person's tendency to react in certain ways under certain circum- stances and that demonstrate both continuity and change over time [45]. These studies have highlighted speci c and similar personality traits among individuals with substance use disorders or behavioral addictions [46,47]. Most studies have used the Big Five model to describe the broad traits that serve as building blocks of personality in individuals with substance and behavioral addictions [48–50]. These ve primary personality traits are extraversion, agreeableness, openness, conscientiousness, and neuroticism [51]. Individ- uals with substance use disorders or dependence have mostly high neuroticism and low openness and/or agreeableness [49]. Some studies have explored the association between personality traits and ED [52–54]. In 2017, Bircher et al. conducted a systematic review with con icting results and concluded that if personality factors were involved in exercise addiction, the current knowledge is insuf cient to draw a speci c pro le of the personality of a person with exercise addiction, and further studies with more representative samples are needed [55]. Although personality traits have been identi ed in ultraendurance athletes (high extraversion and openness to experience) [56], little is known about the traits associated with ED in this population. Given the prevalence of ED in ultraendurance athletes with potential detrimental impacts on their mental and physical health, the identi

addiction, and further studies with more representative samples are needed [55]. Although personality traits have been identi ed in ultraendurance athletes (high extraversion and openness to experience) [56], little is known about the traits associated with ED in this population. Given the prevalence of ED in ultraendurance athletes with potential detrimental impacts on their mental and physical health, the identi cation of personality traits associated with the risk of ED could help physicians and healthcare workers better identify athletes before they experience symptoms of dependence. The present study aimed to investigate the personality traits associated with ED among ultraendurance athletes and identify the symptoms associated with the signi - cant personality traits. The secondary objectives were to assess the prevalence of ED, test participant characteristics, and screen for eating disorders associated with ED in ultraendurance athletes. Regarding our main objective, the hypotheses were that ED could be associated with personality traits similar to those of other behavioral addictions or SUD, with high neuroticism and low openness and/or agreeableness. Additionally, we hypothesized that the coexistence of possible eating disorders would interfere with the strength and level of association between personality traits and ED. 2. Materials and Methods 2.1. Design This prospective cross-sectional study was conducted in an academic medical center from 1 April 2022 to 30 May 2022. The local research committee granted approval (Caen University's Health Research and Ethics Committees COTECH: Ref. 3363), and the study protocol was registered in an international database (ClinicalTrials.gov NCT05348798). The study was conducted in accordance with the Declaration of Helsinki, and informed consent was obtained from all participants. 2.2. Population The study population was all volunteer adult athletes (aged 18 years or older) who spoke French, had participated in an ultraendurance event, and had prepared for at least one ultraendurance event in the current year (e.g., events lasting at least 6 h). The disciplines

Int. J. Environ. Res. Public Health2023,20, 1042 4 of 17 considered were trail running, Ironman triathlons, ultracycling, ultradistance running or ultramarathons, ultradistance open-water swimming, swim-run races, multisport raids, and cross-country skiing. Participants were recruited via federal associations, ultraendurance event organizers, and social networks. The list of federations/event/social media groups targeted for the survey is presented in a Supplementary File (Table S1). 2.3. Measures All participants completed a secure and anonymous web-based questionnaire that took approximately 20 min to complete. The questionnaires included a screening ques- tionnaire about demographic data, healthcare, and sporting characteristics, the Exercise Dependence Scale-Revised (EDS-R), the Big Five Inventory (BFI), and two items of the SCOFF questionnaire to screen for eating disorders. A website version of the questionnaire was developed using the Google Forms platform. An introductory page provided all information regarding the study, participants' rights, and ways to contact the researcher. All participants were informed that participation was voluntary and anonymous. A clear statement indicated that consent was assumed after checking a mandatory “I agree” button that provided access to the survey. The survey platform forced the completion of all ques- tions before the respondent could move on, and all of the collected questionnaires were complete. Personal data (including names, dates of birth, and email addresses) were not collected within the questionnaire or on the survey platform. For data analysis, Google Forms generated an Excel spreadsheet with the answers to the questionnaires. 2.3.1. Screening Questionnaire The screening questionnaire collected the following: (1) sociodemographic data, in- cluding sex, age, marital status, and current employment status; (2) sporting data, including the type of ultraendurance sport(s) practiced, age at the rst ultraendurance event, the number of weekly hours of training, training preferences (with a group and/or alone), the number of ultraendurance events per year, sports club participation (“yes”/“no”), and the consumption of food supplements (“yes”/“no”); and (3) healthcare data, including the type of physician consulted for the medical certi cate of tness and whether the physician had previously discussed ED with the participant (“yes”/“no”). 2.3.2. Exercise Dependence Scale-Revised (EDS-R) ED was investigated via the French version of the Exercise

of ultraendurance events per year, sports club participation (“yes”/“no”), and the consumption of food supplements (“yes”/“no”); and (3) healthcare data, including the type of physician consulted for the medical certi cate of tness and whether the physician had previously discussed ED with the participant (“yes”/“no”). 2.3.2. Exercise Dependence Scale-Revised (EDS-R) ED was investigated via the French version of the Exercise Dependence Scale-Revised (EDS-R) [57,58]. The EDS-R is a 21-item self-administered questionnaire with 7 subscales based on the DSM diagnostic criteria (tolerance; withdrawal effects; intention effects; lack of control; time; reductions in other activities; and continuance) [59]. For each of the 21 items (“statements”), the rating is given on a 6-point Likert scale, ranging from 1 (never) to 6 (always) according to what best corresponds to the participant. The total score of exercise dependence symptoms is calculated out of a possible 126 (higher scores indicate greater dependence). Additionally, the scale allows participants to be categorized into either at risk for ED (ED+, i.e., 5–6 on the Likert scale for 3 subscales), nondependent symptomatic (ED /S+, i.e., 3–4 on the Likert scale for 3 subscales and failing to meet the criteria of ED+), or nondependent asymptomatic (ED /S , failing to meet the criteria of ED+ and/or ED /S+) groups according to the scores on 3 or more of the subscales [59]. 2.3.3. Big Five Inventory (BFI) To assess personality traits, participants completed the French version [60] of the Big Five Inventory (BFI) [61]. The BFI is a valid and reproducible self-administered question- naire for classifying individuals according to ve major personality traits [62]. The BFI is a 44-itemquestionnaire, and each item is a statement (“I am someone who. . .” followed by the item statement). The rating is given on a 5-level Likert scale from 1 (strongly disap- prove) to 5 (strongly approve). Items are combined into ve scales, Extraversion (8 items), Agreeableness (9 items), Conscientiousness (9 items), Neuroticism (8 items), and Openness

5-level Likert scale from 1 (strongly disap- prove) to 5 (strongly approve). Items are combined into ve scales, Extraversion (8 items), Agreeableness (9 items), Conscientiousness (9 items), Neuroticism (8 items), and Openness

Int. J. Environ. Res. Public Health2023,20, 1042 5 of 17 (10 items), with mean scores based on a 5-point scale for each trait (higher scores indicate higher personality traits). 2.3.4. Screening for Eating Disorders The screening for possible eating disorders was explored using two items of the SCOFF questionnaire [63] about body image, namely, Fat (“Do you believe yourself to be Fat when others say you are too thin?”) and preoccupation with food, namely, Food (“Would you say Food dominates your life?”). If the participant answers “yes” to both items, possible eating disorders are considered [64]. 2.4. Statistics The results are presented as means (standard deviations (SDs)). The normality as- sumption for quantitative data was veri ed using a Kolmogorov–Smirnov test. Differences among independent groups (ED+, ED /S+, and ED /S ) were tested using a one-way analysis of variance (ANOVA) with Tukey post hoc analysis for quantitative variables and a chi-square test for categorical variables, with a post hoc analysis of Pearson residu- als. Regression analysis with an interaction effect was used to check for the confounding impact of possible eating disorders on the signi cant personality traits associated with ED. Pearson's correlations were used to determine the strength of the associations be- tween the scores for personality traits and ED and interpreted as 0.11 r 0.30 = small, 0.31 r 0.49 = moderate , or r 0.50 = large [65]. The sample size calculation was based on the results of Di Lodovico et al., who found an ED prevalence of 14.2% in endurance athletes [33]; thus, for a target ofn> 30 in the ED+ group, a minimum sample size of 212 participantswas needed. Statistical analyses were performed using SPSS software ver. 25.0 (IBM, Armonk, NY, USA) for Windows, and the signi cance level was set atp< 0.050. 3. Results 3.1. Characteristics of Participants at Risk for ED A total of 507 participants (41.6 (9.8) years, 374 men; 73.7%) who participated in a mean of 3.5 (2.1) ultraendurance events each year were included. The mean score was 59.1 (18.7) out of 126 on the EDS-R, andn= 37 (7.3%) were

and the signi cance level was set atp< 0.050. 3. Results 3.1. Characteristics of Participants at Risk for ED A total of 507 participants (41.6 (9.8) years, 374 men; 73.7%) who participated in a mean of 3.5 (2.1) ultraendurance events each year were included. The mean score was 59.1 (18.7) out of 126 on the EDS-R, andn= 37 (7.3%) were classi ed as at risk for ED (ED+ group.) The characteristics of participants, with differences between the ED+ and ED groups, are presented in Table. Between the groups, there was no signi cant difference regarding the demographic characteristics (p> 0.050). Participants in the ED+ group trained for a greater duration each week (p< 0.001), with 35.2% practicing for at least 15 h per week, whereas this proportion was only 7.5% in the ED group. There was no difference regard- ing the type of sport, the age at the rst ultraendurance event, the training preference (alone and/or with a group), the number of ultraendurance events per year, the consumption of food supplements, or registration in any sports federation (p> 0.050). Table 1. (a) Characteristics of participants and differences between individuals at risk of ED and those not at risk (n= 507). (b) Characteristics of participants and differences between individuals at risk of ED and those not at risk (n= 507). (a) No exercise dependence Exercise dependence Differences between groups n= 470 n= 37 2 ort-testp-value Age (years, (SD)) 41.6 (9.7) 41.4 (11.1) 0.155 0.877 Age at rst ultra (years, (SD)) 33.4 (8.3) 34.7 (8.4) 0.89 0.375 Sex (n(%)) Male 351 (74.7%) 23 (62.2%) 2.78 0.096 Female 119 (25.3%) 14 (37.8%) Marital status (n(%)) Single 107 (22.8%) 13 (35.1%) 2.91 0.088 Married 363 (77.2%) 24 (64.9%)

The study explores personality traits linked to exercise dependence in ultraendurance athletes.