Risk of dependence on sport in relation to body dissatisfaction and motivation

The aim of the research was getting to know the risk of dependency on physical exercising in individual sportspeople and the relationship with body dissatisfaction and motivation. 225 triathletes, swimmers, cyclists and athleteswith ages going from 18 to 63 years old took part in the research, of which 145 were men (M = 35.57 ±10.46 years) and 80 women (M = 32.83 ±10.31 years). The EDS-R was used to study the dependency on exercising, BSQ to study body dissatisfaction, BREQ-3 to know the motivation of participants and BIAQ to analyse conducts of avoidance to body image. The obtained results show that 8.5% of the subjects had risk of dependency on exercising and that 18.2% tend to have corporal dissatisfaction, without meaningful differences in the kind of sport they practiced. However, there were important differences concerning the dependency on physical exercise (15% vs 4.8%) and body dissatisfaction (31.1% vs 11%) in relation to sex, being the higher percentage referring to women. The introjected regulation and the conduct of food restriction were the predictor variables of the dependency on exercising and corporal dissatisfaction.


Introduction
Nowadays, the multitude of benefits associated with the practice of a sport or physical exercise on a regular basis is known on a physical, psychic, aesthetic and social level [1,2,3,4]. However, in recent decades some new research has emerged on the addiction that sport can create, having negative consequences not only at a physical or physiological level (e.g., abstinence, tolerance) but also psychological (e.g., anxiety, depression) or behavioural (e.g., reduction of other activities) [5,6,7].
First, the term addiction or dependence in relation to physical exercise was not given negative conjectures. Glasser [8] referred to exercise as a positive addiction, in relation to the health benefits it had, stating that it was as addictive as pleasant. But Morgan [9] studied and recognized the negative effects that could appear with sport addiction such as injuries, over-training, social isolation or psychological problems. Therefore, Ogden, Veale and Summers [10] define the dependence on physical exercise (DPE, hereafter) as a combination of biomedical characteristics similar to those of addictions such as withdrawal symptoms and stereotyped behaviour in addition to other psychosocial aspects such as interference with social/family life and positive gratifications. Research in this line, such as that of Sussman, Lisha, and Griffiths [11] has found similarity between sports addiction and drug addiction.
The addiction to sport is manifested by an influence and abuse of it in the day to day leaving aside other factors of life. Whiting [12] concluded that the non-practice of sports during some days generated anxiety, abstinence, as well as physical and moral discomfort in those athletes with symptoms or risk of DPE.
Several studies associate DPE with eating disorders and body disorder or dissatisfaction (BI, hereafter) [7,13,14] in relation to a greater body care or maintaining a good physical appearance, sometimes even exceeding the limits, having athletes to include habits, behaviours or thoughts that distort their body image [15].
An important aspect to take into account in sports practice is motivation and its relationship with the theory of self-determination [16] which can show us how much motivation exists and what kind of motivation leads us to practice sports. This motivation may be related to the DPE, either as a predecessor or as a consequence thereof [17,18].
Sex or age, as well as the sport practiced or the body mass index (BMI, hereafter) are aspects to be pointed out as factors that are related to DEP, CI or motivation [19,20,21]. More and more studies and research are on this addiction to sport and what factor or factors it may be related to. In fitness centres [17] they obtained that 7% of their members could be considered at risk of exercise dependence. Regarding individual sports Ruiz-Juan and Zarauz [22] studied Spanish marathoners, who showed a medium-high addiction to training. Blaydon and Lindner, and also Youngman [13,23] in an investigation with triathletes concluded that they were 30.4% and 19.9% respectively, who could have addiction or DPE. On his behalf, Latorre et al. [24] obtained that the prevalence of DPE in different endurance sports (triathlon, swimming, cycling and athletics) was 13.6%, with triathletes having the highest DPE.
On the other hand, and in relation to sex, the results are not very clear since Guszkowska and Rudnicki [25] obtained that the male sex was more addicted to sport than the female, however, Youngman [23] concluded that girls were who had the highest DPE, but without significant differences.
Nevertheless, despite what we know today, for Forte and Ferreira; Ruiz-Juan and Zarauz [26,27] there are still many gaps to learn more about sports addiction or DPE, as well as which sport carries a higher risk of DPE or BI in case it exists and how these variables interweave with personal and motivational differences. On the other hand, the studies carried out usually take students or gym clients as participants, and the DPE has been a little investigated, for example, with outdoor sports. Therefore, and based on the reviewed studies, this research aims to a) compare the current state of triathlon and the sports that compose it (swimming, cycling and athletics) b) know and study the differences among the DPE, the BI and the motivation of individual sports according to modality and sex; c) establish the relationship between the DPE and the BI together with motivation, conducts of avoidance of the BI, age, BMI and the duration and number of sessions. sessions per week and being active at this time were the parameters of inclusion in the research process.

Instruments
The "Exercise Dependence Scale-Revised" (EDS-R) [28], Spanish version of Sicilia and González-Cutre [29] was used. The scale is composed of 21 items that allow to obtain a global score in dependence (as this score increases there is a greater risk of dependence) and a scoring for each of the seven symptoms that define it. The questionnaire is headed by the statement "In the fitness centre ..." and the responses are presented in the Likert format from 1 (never) to 6 (always The scale also allows to classify the practitioners of physical exercise in three groups: at risk of dependence (scores 5-6 in at least three of the seven criteria), symptomatic non-dependent (scores 3-4 in a minimum of three criteria, or scores 5-6 combined with scores 3-4 in three criteria, but without meeting the conditions of at risk) and asymptomatic non-dependent (scores 1-2 in at least three criteria, but without actually meeting the conditions of the symptomatic non-dependent).
The Body Shape Questionnaire (BSQ) [30] was also used, and adapted to the Spanish population by Raich et al., [31] for the analysis of body satisfaction. It is a questionnaire formed by 34 items that are evaluated by the following scale (1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Very often, 6 = Always), so that the test range is 34-204. Following Cooper and Taylor (1988), from the total BSQ score of> 80 BI risk is established. In this study we obtained an Alpha value of Cronbach = .96 To analyse the motivation towards sport, we used the Behaviour Regulation in Exercise Questionnaire (BREQ-3) [32] and in its Spanish version of González-Cutre, Sicilia and Fernández, [33], composed of 23 items which are divided into 6 factors. The questionnaire is headed by the statement "I do physical exercise ..." and the answers are presented in a Likert format from 0 (not true) to 4 (totally true

Process
The participants signed an informed consent in which they were briefly informed of the type of study that was being carried out, that the answers were anonymous and confidential, the voluntary participation and that the results were available to them at the end of the investigation. The EDS-R, the BSQ, BREQ-3, BIAQ and a data questionnaire referring to different sociodemographic variables were provided. The data was collected in different competitive events at the Andalusian level (Andalucía triathlon championships (medium and Olympic distance), Andalusian cycling circuit, open water swimming events, popular races. Games-Howell in those that had different variances, carrying out in this case a robust Welch analysis.

Analysis of data
In addition, the percentage of practitioners who could be considered as asymptomatic nondependent (AND), symptomatic non-dependent (SND), and at risk of dependence (RD), with the total sample and with each of the sports was calculated. Similarly, the sample was studied to determine the risk of BI. On the other hand, the same analysis of variance with a segmentation by sport and sex for its study. We also used a Pearson correlation analysis between the variables and a multiple linear regression with the DPE and the BI as dependent variables. The level of significance was established at p <.05. The statistical analysis of the data was performed through the statistical program SPSS., V.24.0 for Windows, (SPSS Inc., Chicago, USA). Table 1 describes the results obtained from the relationship among the different sociodemographic variables and the practiced sport modality. It is observed that there are significant differences in age, swimmers being the youngest; in occupation, where the triathletes have their job as occupation and the level of studies, being the triathletes the ones who have higher percentage in higher studies; years of training with swimmers with a greater percentage of more than 10 years of training, even being the youngest and cyclists; the condition of federated, being greater in swimmers and triathletes than in the other two modalities, in the same way swimmers and triathletes are those who dedicate more sessions being the cyclists who more time spend in training. Finally, swimmers are those who have a coach in relation to other sports.   Table 2 shows the results obtained from the EDS-R, BSQ, BREQ-3 and BIAQ questionnaires corresponding to each of the analysed sport modalities. Being analysed the EDS-R, the cyclists are those who present higher score, but without significant differences with respect to the rest of sports.

Results
Yes, there are significant differences in the time of exercise in relation to the triathletes against the athletes and at the same time the triathletes with respect to the swimmers. There are no significant differences among the different sports modalities in the BSQ total scores. However, we do find significant differences in the BREQ-3. Specifically, in the higher intrinsic regulation in cyclists than in athletes and even more in cyclists in comparison to swimmers; integrated regulation being in triathletes greater than in swimmers; in identified regulation obtaining a higher score cyclists against swimmers; and external regulation being higher in swimmers compared to cyclists.
Conducting a study by sex in endurance sports there were significant differences (t (223) =  In addition, the differences among sports by sex were observed through the results obtained in the EDS-R, BSQ, BREQ-3 and BIAQ. In the male sex, significant differences were found in the    428; p <.01) and checking behaviours (r = .316; p <.01). Table 4 and 5 show the linear regression of the factors that predict the BI and the DPE respectively. As a common predictive factor, we find the introjected regulation and the tendency to negative conducts to Body Image.

Discussion
First of all, we have to mention the scarce literature found about the relationship among our different variables, including the variety of instruments that are used for their study, making the results of the different studies differ among them, as well as the conceptual problems to define addiction to sport. Therefore, we must be cautious in the discussion with other investigations in relation to our study.
The first objective was to study the sociodemographic differences that could exist among the different studied individual sports. There are not many studies that have investigated it, however, Latorre et al. [24], agree that triathletes are the ones who in greatest number federate and perform the most sessions a week, although our study also reveals that swimmers have a similar percentage in both cases. Likewise, it coincides with the cyclists in the duration of the sessions. Similarly, in their study there are significant differences in the BMI, with the runners having the lowest BMI. Finally, there are also differences in the total of the sample and in all sports, except in swimming, with respect to BMI being higher in men than in women. The second objective of this study was to analyse the DPE and the BI. Sussman et al. [35], observed that the DPE in the general population of the USA was between 3% and 5%. Reche et al. [36], with university students obtained as a result that there was a DPE in 8% in individual sports.

Preprints
For Harris et al. [37], it was between 16% and 36% depending on which degree they studied. Studies like those of Latorre et al. [24], show that in individual sports there is a DPE of 13.6%, existing significant differences among different sports, being the triathlon the largest with 29.6% as opposed to our study where swimmers (10%) have had the highest DPE. Magee, Buchanan and Barrie [38], and Valenzuela and Arriba-Palomero [39], obtained an RD of 8.6% in a study with male triathletes, while Blaydon and Linder (2002) [13], analysed the triathletes showing between 25-30% symptoms of sport addiction, the same way as Youngman and Simpson [40], who obtained that approximately 20% of the triathletes were at risk of sport addiction.
Taking into account the comparison of sexes, there are several studies and conflicts among them.
Ruiz-Juan, Zarauz and Flores-Allende [41], studied the Negative Addiction to Running (ANC) in endurance runners (half marathon and marathon) finding a mean score higher than the mean without significant differences between sexes. Like Ortiz and Arbinaga [42], and González-Cutre and Sicilia [17], in a study with individual sports and conditioning centre they did not find significant differences in the DPE in relation to sex, as Modolo et al. [20], where they obtained that 28% of women and 38% of men had DPE symptoms without finding significant differences. Bingol and Bayansalduz [43], concluded that 19.2% of women and 13.15% of men suffered RD in a multitude of studied sports.
Mayolas-Pi et al. [44], in a study with cyclists (female and male) showed that 17% of men and 16% of women were at risk of sport addiction. We have found a few studies with significant difference in DPE being higher in women than our results have been.
In the study of DPE and its dimensions compared to each sport modality Latorre et al. [24], found differences in the total score being higher in triathletes and cyclists than in runners, a fact that did not result in our study. However, there were coincidences in the time dimension of exercise being higher in triathletes than in athletes, although in our study it was also reflected with triathletes greater than swimmers. Likewise, Harris et al. [37], conducted a study with university students where differences were found in the subscale "exercise time" among students of Sports Science and other university degrees.
Like our research, Ortiz and Arbinaga (2016) [42] did not find significant differences in the subscales of the EDS-R comparing the two sexes. However, Gonzalez-Cutre and Sicilia [17] in fitness centres did find significant differences between men and women in all the subscales of the DPE, being higher in the male sex, except for the abstinence factor. It is important to note that the sample in Ortiz and Arbinaga [42] was much smaller.
As far as BI is concerned and its relation to sport, in this case individual sports, Fortes, Almeida In the association of these two variables there is also a significant relationship as we have seen it in the results in agreement with other studies where there is a strong association between DPE and BI (Davis [47], Cook and Hausenblas [14], Weinstein and Weinstein [15]). ]; Zmijewski and Howard [19]) in a study with university students of Physical Activity and Sports Sciences concludes that girls with a higher score in the DPE showed higher scores in the BI, while in boys this did not happen, confirming the significant differences that we found in our study regarding greater HF in women than in men. Hausenblas and Fallen (2002) [7], even obtain as a result that at a general level the subscale of lack of control predicted the BI. These same authors, in men, the intention subscale and time of exercise were predictors of BI.
The next objective was to establish the relationship between the DPE and the BI in relation to the motivation, in particular with the regulation conducts of motivation.
In this regard, several studies have linked introjected regulation as a major predictor of DPE (Edmunds, Ntoumanis and Duda [48], Fortier and Farrell [49]). On the other hand, González-Cutre and Sicilia [50], established in their study that not only was DPE related to non-self-determined motivations such as external and introjected regulation, but also to some self-determined motivation such as integrated regulation. Our study coincides with its results adding demotivation as a factor that prevails in the DPE. These authors conclude that the positive or negative character of a high DPE could be in how the person constitutes their behaviours within their lifestyle. Latorre et al. [24], find the integrated, identified, and introjected regulation as associated factors to the DPE.
Zarauz, Ruiz-Juan and Arbinaga [51] studied the sport addiction and the running compromise (EAG) in athletes and they found that in females scoring high in two self-determined motivations (psychological and life -self-esteem objectives) predicted the EAG. However, in males, the EAG could be predicted in a significant way by its high scores in the three more self-determined motivations and lower scores in the non-self-determined motivations.
Sicilia et al. [52] indicate that the practice of exercise in a multitude of occasions comes preceded by external, internalised or not pressure by the people, that takes them to do exercise to lose body weight and maintain a good body image. Regarding this, Latorre et al. [24], obtained that only the introjected regulation is associated with BI. On the other hand, in our study they were the three nonself-determined regulations (introjected, external and demotivation) as a negative association with the intrinsic regulation.
However, although the physical and health motivations can be associated with the positive consequences of physical exercise for people who have a low body dissatisfaction, greater support for the motivations of both physical fitness and health, like appearance and weight, are associated with a higher dissatisfaction of the state body in women, categorised as high dissatisfied body (Sampasa-Kanyinga et al. [53]).
In the same way, Gonzalez-Cutre and Sicilia [17], include the integrated regulation and external regulation as predictors of the DPE, coinciding in this study only with the integrated regulation, being Latorre et al. [24], the identified regulation what predicts the DEP. According to other variables, a multitude of studies and instruments have found that the prevalence in sport addiction decreases with the age, so it correlates inversely (Szabo [54]; Ruiz-Juan, Zarauz y Arbinaga [55]; Costa et al. [56] There are studies that make reference not only to the training or the duration itself, but also to the fact that there are differences in relation to the tests that are carried out (Guszkowska and Rudnicki [25]). It is not the same to prepare a sprint as an ironman in a triathlon (Youngman and Simpson [40]), or a half marathon as a marathon (Ruiz-Juan et al. 2016). It was also a predictor of the DPE finding an inversely significant relationship with the age (Latorre et al. [24]). Social activities were one of the conducts that prevail the DPE. As to the BI, in addition to the previously mentioned, it was obtained as a result that sex is an associated factor to the CI. In this way, it coincides with our study, where the sex is an important factor in the BI (Harris et al. [37]; Hausenblas and Fallon [7]).
To know which variables or factors predicted the DPE and the BI showed by the analysis of multiple linear regression, it was the introjected regulation that predicted them as equal as in other studies (Latorre et al. [24]). In any of the analysed studies was a comparison carried out with the conducts of avoidance, being predictor the food restriction in our study for both variables.
The biggest difficulty of this study was to find women who practiced or competed in cycling. In addition to it was to underestimate the data collection, since I thought it would be easier to perform.
Another important aspect to take into account in the sociodemographic analysis is not only the occupation, but the working hours per week, as well as the family burden, since having children or not is an important factor. Another important limitation is the great distances and kilometres made to get the sample, all of which was paid by me, including a larger investment of time. Finally, and in relation to the questionnaires, the extension of them was highlighted, especially in the BSQ questionnaire where most of the participants issued a complaint of being too long, this led some questionnaires to be considered as experimental death for not being completely completed.

Conclusions
Once having studied the literature, many authors have come to the conclusion that the increase in social relationships, economic stability, type of job, but especially weight loss, feeling better with your body, pleasure, achieving motivating challenges or achievements are the main factors that lead to a continuity in training or exercising, that is, physical and psychic components.
But this practice of exercise can become a problem or even an addiction, which not only has an effect on loss of control, abstinence or reduction of time in other activities, but also refers to body image or food behaviours in a negative way. In view of our results we have observed how 8.5% was declared in RD, but ascending up to 48.4% the AND, whose subjects in many cases had a score higher than 5 in two dimensions or total scores of the EDS-R much higher than the average. This makes us see that in this group many of them can be in danger of DR, without being very different between women and men. In our study the RD has been greater in women, standing out in swimmers being in 1 out of 4.
In addition, we can conclude that in the analysed individual athletes does not exist a great BI, but we must take into account the difference between HF in men and women, a fact that many studies have corroborated in all types of people (athletes or not athletes).
In our study, it was found that there are several validated instruments for the DPE not giving the same results or even results that differ among them even researchers using the same instruments, which makes us see that perhaps the addiction manifests itself in different ways and the context in which the investigation is carried out must be taken into account. Finally, we have seen in our study how DPE and BI correlate each other, and both can be preceded by a high motivation that is not self-determined, highlighting intrinsic regulation as a predictor of both. Together with this regulation, food restriction is the other predictor variable found between the two variables, existing scarce literature that studies these conducts of avoidance in relation to the DPE.
All this, together with the number of studies that have been done, shows us how we can identify the profile of endurance athletes, in order to create strategies to prevent and treat (in its case) addiction in sports.