Since decades, perfectionism scientific contributions suggest the existence of a neurotic/negative perfectionism, which refers to those efforts aimed at achieving goals that are out of reach, mainly motivated by an inability to cope with failure, relaunching with the pathology or psychological dysfunction; and a normal/positive perfectionism, referring to the capacity to work for real goals through a positive effort giving rise to feelings of satisfaction or reward, linked to an adequate and adaptive functioning.

Research on perfectionism has been considered a fundamental variable in many psychological disorders (e.g. anxiety, depressive, obsessive) for several decades both sport (Hill et al. 2020a) like general population (Fei and Ma 2017). Some authors point out that there is little consensus on the exact definition of this construct (Hill et al. 2020b). The studies on this topic are based on the idea that it can be considered as a factor of vulnerability in the face of various psychopathologies that appear daily in clinical work (Nogueira et al. 2018).

Apart from that, it is recognized that tendencies to over-exercise exist in a social context, with the presence of socio-cultural pressures that emphasize physical attractiveness or success. Exercise addiction has been classified as a behavioural addiction, similar to other comportamental addictions (e.g., ludopathy) and manifested by reoccurring compulsive urges to exercise and symptoms of dependence (Kovacsik et al. 2019). Normally, the main manifestations of exercise addiction have been associated with a dysfunctional response with body image (Corazza et al. 2019), eating disorders (Trott et al. 2020), distress (Hill et al. 2015) or depression (Weinstein et al. 2015). Griffiths (2005) described the set of six key components (e.g. evidence of repetitive behaviour, mood alteration, increased tolerance, isolation, conflict and relapse) as a biopsychosocial response that generates lasting harmful effects on an individual’s daily life (Weinstein et al. 2015).

Although both are related to diverse and similar dysfunctional responses, there is no extensive literature offering studies on addiction and perfectionism in general or clinical samples, other than those relating them to image disorders (Dahlenburg et al. 2019; Malcolm et al. 2018) or obsessive thoughts (Leone and Wade 2018; Page et al. 2018).

The increase in research on perfectionism in sport and the emergence of different scales to evaluate it with a multidimensional perspective led to the inclusion of both cognitive (e.g. expectations or concerns) and social (e.g. parental criticism or social pressure) components in the definitions. Such components are based either on personal demands (e.g. personal success) or on negative self-assessments (Hewitt et al. 2003). The main investigations qualify perfectionist people as self-critical, with very high goals, with an important negative social perception and with low conformity to make mistakes (Stoeber et al. 2018). At the same time, much of the addiction scales have described the dysfunctional response (cognitive and behavioural) related to the difficulty to control the impulse to practice exercise despite negative physical, social and psychological consequences, associated to this excessive behaviour (Table 1).

Table 1 Summary of studies on exercise addiction and perfectionism
Full size table

Both in general and sporting samples, different concepts have served as synonyms to explain the process of addiction when interacting with perfectionism, which could lead to any of conceptual confusion. While some research speaks of state dependence (Maia et al. 2011; Zuroff et al. 2016) to relate with unhealthy behaviours (e.g. insomnia, binge drinking or drug use), the most prolific links have been with obsessive-compulsive disorder (Egan et al. 2011; Moretz and McKay 2009; Pinto et al. 2017), described as obsessive and insistent thoughts that cause incessant and uncontrolled repetition (compulsion) of a behaviour.

Therefore, the objective of the present study was to carry out a systematic review of the researches in sports contexts that have included the concepts “perfectionism” and “exercise addiction”. The review has the following triple purpose: (a) to identify the most relevant evidence on the constructs in sport contexts, (b) to describe the main measures used of both concepts and (c) to offer information on which are the variables associated so far when both constructs are combined, considering also specific aspects such as sex, type of sport practice and associated variables.

Method

Eligibility Criteria

For the literature search on addiction and perfectionism, any work that is related to both topics, and any variable in the sports context, was selected. Once the first search was completed, the studies that met that the following inclusion criteria were selected: (a) typology of the document source (articles published in impact journals); (b) typology of the scientific proposals (theoretical proposals, empirical designs, relationship with variables or populations in sports contexts); and (c) documents that include perfectionism or addiction in the title, abstract or key words. Studies were excluded from the review if they did not include the following: (i) a specific instrument to assess perfectionism or addiction; (ii) samples in non-sporting contexts; (iii) no English, French or Spanish language; and (iv) they were a thesis, a book chapter or congress dissertations.

Information Sources and Search

A systematic literature search was conducted to identify documents that follow the PRISMA guidelines for systematic reviews of scientific studies (Urrútia and Bonfill 2010). The inclusion criteria (document typology, study typology, links with sports contexts and language) followed the PICOS methodology (Moher et al. 2010).

Studies were identified by searching electronic and unrestricted access databases. The electronic databases and platforms used were Web of Science, Scopus, ProQuest, MEDline and EBSCO-HOST, and included major resources such as PSY Articles, PsycINFO, LWW, ERIC, SportDISCUS, PubMed, ERIC, Dialnet, PubMed, ISOC, the Cochrane Library and Google Scholar.

The literature process of search was conducted during April 2020. The search terms selected were as follows: “addiction”, “perfectionism” and “exercise addiction”, combining each one of them with the word “sport”; (perfectionism OR addiction) AND exercise OR sport; (perfectionism AND addiction) AND exercise OR sport; (perfectionism OR/AND exercise addiction).

Study Selection and Data Collection Processes

After the first literature search, the next step was to check the title, abstract and keywords of each study, as a filter for the papers, and then to further assess the resulting studies that could potentially be included in the final selection. The entire process of study selection can be consulted in the PRISMA data flow chart (see Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram of the systematic review process of the Exercise Addiction and Perfectionism in sports settings

Full size image

Results

Study Selection

A total of 754 studies (Google Scholar = 592; ProQuest = 39; Scopus = 35; PubMed, n = 35; Web of Science, n = 24; EBSCO, n = 9; PSYarticles = 6; PsycInfo = 4; Dialnet = 4; MEDLINE = 4; SportDISCUS = 4; Dialnet = 4) were identified in the initial search process. After checking the duplicate studies (57) and re-analysing the title, abstract and keywords of each study (404), a total of 461articles were excluded. In this case, 293 studies were selected for eligibility phase. In this phase, a total of seven studies were excluded for language issues, including another 264 papers, as they did not use a sport context sample or specific instruments to assess perfectionism or addiction. Following these procedures, 22 eligible empirical studies were chosen for evaluation in the present review.

Sport Context in which the Data Was Collected

The major part of the sports sample analysed in the included studies is identified as fitness/gym users (n = 5; Costa et al. 2016b; Grandi et al. 2011; Hill et al. 2015; Monasterio et al. 2014) or habitual exercisers (n = 4; Costa et al. 2016a; Miller and Mesagno 2014; Şenormancı et al. 2019; Taranis and Meyer 2010). Among these studies, one of them (Lichtenstein et al. 2014), in addition, divided the sample in two groups (addiction and non-addiction) and include football teams as members of the sample.

Four studies used a sample that included different sport modalities (Egan et al. 2017). They included a wide diversity of sports (running, triathlon, ball sports, gym, cycling, dance, yoga, athletics, rowing, swimming); Hauck et al. (2020) used a sample with sportsmen and sportswomen of endurance sports (running, triathlon, cycling), fitness users and team sports. On their behalf, the study by Gulker et al. (2001) has participants who are runners, weightlifters and aerobics exercisers, and finally, Davis (1990) included runners, swimmers, cyclers and aerobic exercisers, and one group of non-exercisers.

Three of the selected papers analysed addiction and perfectionism in endurance athletes, in particularly, two in middle-distance runners (Hall et al. 2007, 2009). One study did not specify the distance and it divided the sample in obligatory and non-obligatory runners (Coen and Ogles 1993). Finally, of the six remaining studies, five are reviews (Bircher et al. 2017; Gonçalves Baptista et al. 2019; Hausenblas and Downs 2002; Lichtenstein et al. 2017; Petit and Lejoyeux 2013), and one is a qualitative study (Macfarlane et al. 2016).

Inclusion Criteria to Be Considered as Habitual Exercisers

The studies established different criteria to determine who was considered a regular exerciser participant or not. Two studies chose as inclusion criterion, practice exercise twice or more per week (Egan et al. 2017; Miller and Mesagno 2014); three, at least 3 h/week (Davis 1990; Grandi et al. 2011; Gulker et al. 2001); one, the participants should have been exercising for more than 6 months (Costa et al. 2016b); one, at least 4 h/week (Hauck et al. 2020); one, at least 1 h/3 times week during the past 4 weeks (Taranis and Meyer 2010); and one, subjects had to have exercised regularly over the period of a year (Costa et al. 2016a).

On the other hand, three research used the scores of the instruments to determinate the inclusion in the sample. In this way, one study (Lichtenstein et al. 2014) considered EAI scores (high, medium and low) together to other markers such as continuing to train despite being injured; two considered EDS-R scores (Hall et al. 2009; Monasterio et al. 2014), as a criterion, while one of the studies (Coen and Ogles 1993), selected as a criterion, has participated in a race (5 K, 10 K, half-marathon and marathon). Finally, three studies do not describe a specific criterion; they made the choice through a questionnaire (Hall et al. 2009; Hill et al. 2015; Şenormancı et al. 2019).

Sex, Age and Time/Week

The total number of the participants in the 22 reviewed studies was 3751, in which 1702 (45.40%) were female and 1917 (54.60%) were male exercisers. Two of the studies (Coen and Ogles 1993; Davis 1990) did not provide sex information, while the proposal of Monasterio et al. (2014) only recruited male participants (n = 56), and the proposal of Taranis and Meyer (2010) was composed of female (n = 97). All of the studies found use general or amateur exercisers. In particular, more symptoms of exercise addiction in male exercisers were seen, except when this behaviour is associated with eating disorders, women are most at risk (Costa et al. 2016b; Miller and Mesagno 2014; Monasterio et al. 2014; Şenormancı et al. 2019). In these cases, females tend to show high indicators of self-orientated perfectionism (r2 = 0.102) and socially prescribed perfectionism (r2 = 0.137), but not other-orientated perfectionism (Miller and Mesagno 2014).

According to the age, the mean of the sample was 29.80 years, calculated with the data of all studies except to three of them that did not specify it (Egan et al. 2017; Gulker et al. 2001; Monasterio et al. 2014), along with the five reviews and the qualitative study (Macfarlane et al. 2016).

Finally, the average of time the participants in these studies invest in doing exercise goes from a minimum of 4.52 h to a maximum of 12 h/week, where the most repeated mean time was around 6.6 h/week (n = 5; Coen and Ogles 1993; Costa et al. 2016a; Grandi et al. 2011; Hall et al. 2007, 2009). Three of the studies showed mean times below 6.6 h/week (4.52 h, Hill et al. 2015; 5.84 h, Taranis and Meyer 2010; 6 h, Miller and Mesagno 2014), and four indicated over mean times over (7.30 h, Costa et al. 2016a; 7.85 h, Hauck et al. 2020; 8.4 h Lichtenstein et al. 2014 and 12 h, Şenormancı et al. 2019).

Country and Prevalence

Two countries share the largest number of articles included in this review (see Fig. 2), with six studies of the 22 researches conducted in the USA (Coen and Ogles 1993; Costa et al. 2016b; Egan et al. 2017; Gulker et al. 2001; Hauck et al. 2020; Hausenblas and Downs 2002) and five in the UK (Bircher et al. 2017; Hall et al. 2007, 2009; Hill et al. 2015; Taranis and Meyer 2010). Following these countries, we can find Italy with three studies (Costa et al. 2016a; Costa et al. 2016b; Grandi et al. 2011) followed by Denmark (Lichtenstein et al. 2014; Lichtenstein et al. 2017) and Australia (Egan et al. 2017; Miller and Mesagno 2014) with two. The rest of the countries only have been represented by one research, Brazil (Gonçalves Baptista et al. 2019), Canada (Davis 1990), Finland (Egan et al. 2017), France (Petit and Lejoyeux 2013), Germany (Hauck et al. 2020), Hungary (Bircher et al. 2017), New Zeland (Macfarlane et al. 2016), Turkey (Şenormancı et al. 2019) and Venezuela (Monasterio et al. 2014).

Fig. 2
figure 2

Studies that integrate perfectionism and exercise addiction (by countries)

Full size image

It is important to note that four of the studies were conducted collaboratively among more than one country: Egan et al. (2017), Australia, USA and Finland; Costa et al. (2016a), Italy and USA; Bircher et al. (2017), Hungary and UK, and finally Hauck et al. (2020), Germany and USA.

Regarding the exercise of addiction prevalence, few studies have reported these data. In this sense, two of the reviews, Lichtenstein et al. (2017), report an interval between 3 and 42%, and Gonçalves Baptista et al. (2019) point out a wide interval between 1.44 and 51.7%. Only three of the other studies informed about these data, Hausenblas and Downs (2002) placed these values in the 9.68% of their sample, while Grandi et al. (2011) report an exercise dependence prevalence of 36.4% and 29.9% regarding primary exercise dependence. Most recently, Hauck et al. (2020) obtained a prevalence of 30.5%.

Operationalization of Exercise Addiction and Perfectionism

Different terms are used to refer addictive behaviours related to exercise (Fig. 2). In this review, the distribution was as follows: dependence (n = 12; Costa et al. 2016a; Costa et al. 2016b; Gonçalves Baptista et al. 2019; Grandi et al. 2011; Gulker et al. 2001; Hall et al. 2009; Hauck et al. 2020; Hausenblas and Downs 2002; Hill et al. 2015; Miller and Mesagno 2014; Monasterio et al. 2014; Şenormancı et al. 2019) addiction (n = 5; Bircher et al. 2017; Davis 1990; Lichtenstein et al. 2014; Macfarlane et al. 2016; Petit and Lejoyeux 2013); compulsive (n = 3; Egan et al. 2017; Lichtenstein et al. 2017; Taranis and Meyer 2010) and obligatory (n = 2; Coen and Ogles 1993; Hall et al. 2007).

In this sense, the most selected to register addictive behaviour joint perfectionism with a total of five studies was the Exercise Dependence Scale Revised (EDS-R; Symons Downs et al. 2004), selected by Costa et al. (2016a); Costa et al. (2016b); Hill et al. (2015); Miller and Mesagno (2014); Monasterio et al. (2014); followed by the Exercise Dependence Questionnaire (EDQ; Ogden et al. 1997) (n = 2; Grandi et al. 2011; Hall et al. 2009), the Compulsive Exercise Test (CET; Taranis et al. 2011) (n = 2; Egan et al. 2017; Taranis and Meyer 2010) and the Obligatory Exercise Questionnaire (Pasman and Thompson 1988) chosen by Coen and Ogles (1993) and Gulker et al. (2001).

Two measures of perfectionism have been the most used in relation to exercise addiction. The first of these is the Multidimensional Perfectionism Scale (Frost et al. 1990) which has also shown a link with eating disorders (Costa et al. 2016a; Costa et al. 2016b; Egan et al. 2017), goal orientations and perceived ability (Hall et al. 2007), and compulsive exercise (Taranis and Meyer 2010). The second assessment tool is the Multidimensional Perfectionism Scale (Hewitt and Flett 1991), which focuses on descriptively connecting dysfunctional response (Coen and Ogles 1993), body image (Hill et al. 2015), self-acceptance (Hall et al. 2009), exercise dependence (Monasterio et al. 2014) and narcissism (Miller and Mesagno 2014).

Other used perfectionism scales have been Perfectionistic Self-presentational Styles (Hill et al. 2015), the Multidimensional Inventory of Perfectionism in Sport (Hauck et al. 2020) and the Dysfunctional Attitude Scale Turkish short form (Şenormancı et al. 2019).

The rest of the tools were only selected in one occasion: Şenormancı et al. (2019) selected the Exercise Dependence Scale (EDS, Hausenblas and Downs 2002), Lichtenstein et al. (2014) used the Exercise Addiction Inventory (EAI; Terry et al. 2004), Hauck et al. (2020) applied the Questionnaire to diagnose exercise dependence in endurance sports (Schipfer 2015) and Hall et al. (2009) used the Obligatory Running Questionnaire (Pasman and Thompson 1988). The last study, Davis (1990) chose the Eysenck Personality Questionnaire (Addictiveness) (Eysenck 1975).

The studies included in the systematic review among exercise addiction and perfectionism point out the perfectionism as a mediator variable that promotes or boost the emergence of exercise addiction (Egan et al. 2017; Grandi et al. 2011; Hauck et al. 2020; Hall et al. 2007, 2009; Lichtenstein et al. 2014; Şenormancı et al. 2019; Taranis and Meyer 2010).

In addition, some of these studies showed more specific results, for instance Costa et al. 2016a and Costa et al. 2016b indicated the maladaptive perfectionism, while Hall et al. 2009 and Miller and Mesagno 2014 noted the role of self-orientated and socially prescribed perfectionism, as determinants of the risk of exercise addiction. On their behalf, Monasterio et al. (2014) and Hill et al. (2015) only highlighted a positive relationship with one of the perfectionism dimensions, socially prescribed and self-orientated perfectionism, respectively. Furthermore, in the last-mentioned study, they pointed out how perfectionistic self-presentational styles are also associated with exercise dependence symptoms.

Other Variables/Topics Associated

Almost all of the studies included in the review looked at exercise addiction and perfectionism, along with other variables. Of all these, eating disorders (Costa et al. 2016a; Costa et al. 2016b; Egan et al. 2017; Gonçalves Baptista et al. 2019; Grandi et al. 2011; Lichtenstein et al. 2014; Lichtenstein et al. 2017) and personality traits (Bircher et al. 2017; Gonçalves Baptista et al. 2019; Grandi et al. 2011; Lichtenstein et al. 2014; Lichtenstein et al. 2017; Petit and Lejoyeux 2013; Şenormancı et al. 2019) were the most used, appearing in seven studies each one.

Regarding the first topic, exercise addiction is related to the symptoms of eating disorders, but without reaching a clinical level. Besides these, other disorders associated with these behaviours were obsessive-compulsive disorder (Gulker et al. 2001; Lichtenstein et al. 2017), body image disturbance (Gulker et al. 2001), depression (Grandi et al. 2011), muscle dysmorphia (Lichtenstein et al. 2017), dysfunctional attitudes (Şenormancı et al. 2019) and food addiction (Hauck et al. 2020).

In personality traits, the studies point out the possible existence of different personality types associated with a higher risk of exercise addiction, and where perfectionism together with narcissism seem to be two modulating variables and possible triggers for the appearance of this type of behaviour.

In second place is self-esteem (Hall et al. 2009; Gonçalves Baptista et al. 2019; Şenormancı et al. 2019) and anxiety (Coen and Ogles 1993; Grandi et al. 2011) is the topic that would occupy the third position in our study.

In the case of self-esteem studies, it seems to indicate that depending on the profile of the sample, the effect of this variable can modify considerably. In the situations that this relationship was significant, self-esteem was positioned as a mediating variable for exercise addiction, oriented to achieve a goal regulated by perfectionism thoughts. Finally, exercise addicts showed higher levels of anxiety and a high correlation between trait anxiety and several of the perfectionism subscale scores, and were the last variables which appear in more than one study, specifically in two.

The rest were only used in one study: anger, curiosity, ego identity and social desirability (Coen and Ogles 1993); deprivation (Hausenblas and Downs 2002); goal orientations and perceived ability (Hall et al. 2007); self-criticism (Taranis and Meyer 2010); attitudes regarding the self (e.g. self-approval, self-condemn), hostility and somatization (Grandi et al. 2011); doping and other addictions (Petit and Lejoyeux 2013); attachment styles and self-report of quality of life (Lichtenstein et al. 2014); body mass (Hill et al. 2015); parental psychological control (Costa et al. 2016a); psychological needs satisfaction (autonomy, competence and relatedness) and thwarting (Costa et al. 2016b); psychological characteristics (motivation) (Gonçalves Baptista et al. 2019).

In the last one of the studies in this review, the qualitative study (Macfarlane et al. 2016), the participants pointed out three main components that could be an important role in the exercise addiction model, learned (negative perfectionism), behavioural (obsessive–compulsive drive), and hedonic (self-worth compensation and reduction of negative affect and withdrawal). Emphasizing the positive relationship among the obsessive behaviours, perfectionism and exercise addiction acts as a variable that enhances the manifestation of these dysfunctional behaviours.

Discussion

This systematic review had the goal to identify the studies that analysed the possible association between perfectionism and exercise addiction (EA). In addition, of this information and after revising each one of the documents included in the final selection, extracted data about the sample studied, that are the criteria for inclusion, sex and age of the samples, the terms and the tools most frequently used, and the other variables/topics are chosen to try to explain this addictive behaviour.

The main objective of the present review was to select the research that analysed the relationship between exercise addiction and perfectionism, and after revising all the sources included in the systematic review, it is enough evidence that the existence of a strong relationship among this two concepts, as proved Lichtenstein et al. (2014), who found that how a group of identified as addiction exercisers, scored higher on perfectionism than the control group or Coen and Ogles (1993), in a group of obligatory runners.

The results found in regular exercisers are very similar because the conclusions showed by the studies pointed out as a perfectionism predicted exercise dependence symptoms. People who are addicted to exercise seem to try to develop a number of mechanisms in order to overcome their limits (tolerance) in search of the greatest possible goals. This produces enormous gratification (e.g. excitement), which in turn serves as a natural strategy in the face of both physical and emotional pain (Fuss et al. 2015; Raichlen et al. 2012). Hence, in this desire or impulse to reach the highest levels (compulsive exerciser—obsessive/compulsive traits), perfectionism has a strengthening role, since the closer to excellence one can get, the more gratification one will receive (Coen and Ogles 1993; Miller and Mesagno 2014).

In particular, some studies had described more precisely the role of perfectionism, for instance, Miller and Mesagno (2014), affirm that both dimensions of perfectionism, self-orientated perfectionism and socially prescribed perfectionism, were positively related to exercise addiction. Highlighting that this relationship is stronger, self-orientated perfectionism is combined with narcissism. In the same way, Hall et al. (2009) mentioned that this association is produced by different psychological mechanisms, possibly caused by the involvement of motivation.

Hill et al. (2015) confirmed the previous results, but they add that self-oriented perfectionism is the most important predictor of exercise dependence, due to the importance to the people to try to manage or control the feelings related to achieving the best version of oneself where excessive practice of exercise could act like a coping strategy to reduce the possible negative feelings derived from not being able to achieve the eager perfection. Besides confirming this relationship, some studies pointed out that the presence of the perfectionism trait behaves as a risk factor to the presence of eating disorders (Costa et al. 2016b; Gulker et al. 2001; Hauck et al. 2020; Lichtenstein et al. 2014). Eating disorders are known to be one of the most common disorders that occur along with exercise addiction (Freimuth et al. 2011).

On the other hand, perfectionism has been identified as one of the four factors in the transdiagnostic model of eating disorders (Fairburn et al. 2003); in addition to its proven relationship with compulsive exercise (Taranis and Meyer 2010), people who practice sport seeking higher levels of perfection, together with an imperative desire for control (Alcaraz-Ibáñez et al. 2018), increase the number of mechanisms that favour the appearance of eating disorders, especially among the female population (Egan et al. 2017). However, it also appears that the effect of perfectionism will also be reflected in eating disorders, without the need for excessive exercise behaviour (Costa et al. 2016b).

Other studies mentioned how other variables can also function as mediating variables in the relationship perfectionism-exercise addiction, mainly self-esteem and attitudes regarding to the self (Bircher et al. 2017; Grandi et al. 2011; Hall et al. 2009; Şenormancı et al. 2019; Taranis and Meyer 2010), or narcissism (Bircher et al. 2017; Grandi et al. 2011; Lichtenstein et al. 2014; Lichtenstein et al. 2017; Miller and Mesagno 2014). The narcissistic personality (multiplying their effect joint to perfectionism) can be identified and associated with obsessive behaviours, focusing primarily on the need and desire to be better than others, especially when being evaluated externally (Nogueira et al. 2019; Wallace and Baumeister 2002).

As can be seen from the studies included in the review (Costa et al. 2016a; Costa et al. 2016b; Davis 1990; Egan et al. 2017; Grandi et al. 2011; Gulker et al. 2001; Hauck et al. 2020), what is not clear is what dose could be identified as that line that marks regular practice, perfectionism and addictive behaviour. Given the variability of time references used, the most important perfectionistic features are self-criticism (seems to be the main indicator of the dysfunctional perfectionism associated with compulsive exercise) (Taranis and Meyer 2010), socially prescribed perfectionism and self-oriented perfectionism (Miller and Mesagno 2014).

In the case of the research proposed by Lichtenstein et al. (2014), people addicted to exercise were driven by a perfectionist pursuit of high goals and excitement that, despite pain and overuse injuries, showed more willingness to continue training. This behaviour seems to be associated to the commitment feeling with the practice and to the feelings of guilt when exercise cannot be performed (Egorov and Szabo 2013; Szabo et al. 2016).

Sex, Age, Time Practising in the Relationship Exercise Addiction-Perfectionism

Sex, age and time practising some kind of exercise seem to be three of the control variables (Egan et al. 2017; Lichtenstein et al. 2017) with more impact in this dyad (Hill et al. 2015; Marques et al. 2019). The literature shows that the radiography of the exercise addict athlete could be a man, about 30 years old, who practices intense exercise about 6.6 h/week.

In general, it seems to be that the youngest are the most commonly categorized as being at risk of addiction (Bruno et al. 2014; Cabrita et al. 2018; Lichtenstein et al. 2014).

Regarding sex, the literature shows quite controversial results. While there are studies that found that women had a higher risk of suffering from addiction in sports such as triathlon or running (Cook et al. 2013b; Youngman and Simpson 2014), justifying that it was women who presented the greatest risk because their main motivation was related to body image (Dumitru et al. 2018; Lichtenstein and Jensen 2016), another important group of research observed the differences in favour of men (Cook et al. 2013a; Cunningham et al. 2016) (considering they could be associated with the current trend on the cult and care of the male body motivations, more oriented towards competing and winning).

The studies included in this systematic review highlight that when exercise is used as a means of reducing psychological distress, it is men who are at greatest risk of addiction (Monasterio et al. 2014; Szabo 2010). When exercise has a function more related to physical image, it is women who show higher values, with eating problems acting as a mediating variable (Costa et al. 2016b; Grandi et al. 2011).

As for perfectionism, for female, perfectionist seems to be the motivational trigger exercise addiction, due to the problems in handling failure. For men, the combination of socially prescribed perfectionism and contingent self-esteem appears to pose the greatest risk of exercise dependence (Hall et al. 2007, 2009; Miller and Mesagno 2014).

Finally, in the time dedicated to the practice of sports, like the previous variables, the results are varied and show a strong link to the type of activity, where the bibliography seems to point out a positive and direct relationship between risk of exercise addiction and the number of hours that people spend training (Lukacs et al. 2019; Román et al. 2016).

Conclusions

The analyses of the present systematic review confirm the relationship between exercise addiction and perfectionism, where the personality trait works as a risk factor and that encourages the emergence of exercise addictive behaviours. In addition, when this occurs along with other personality traits such as narcissism, the manifestation of addictive behaviours is greater. On the other hand, perfectionism was also shown as a bridging feature between exercise addiction and eating disorders, due to the compulsive nature of these variables. These results are important as they provide information for further research to discover an explanatory model for exercise addiction.