The aetiology and trajectory of anabolic-androgenic steroid use initiation: a systematic review and synthesis of qualitative research

Background To our knowledge, there has never been a systematic review and synthesis of the qualitative literature on the trajectory and aetiology of nonmedical anabolic-androgenic steroid (AAS) use. Methods We systematically reviewed and synthesized qualitative literature gathered from searches in PsycINFO, PubMed, ISI Web of Science, Google Scholar, and reference lists of relevant literature to investigate AAS users’ ages of first use and source(s), history prior to use, and motives/drives for initiating use. We adhered to the recommendations of the UK Economic and Social Research Council’s qualitative research synthesis manual and the PRISMA guidelines. Results A total of 44 studies published between 1980 and 2014 were included in the synthesis. Studies originated from 11 countries: the United States (n = 18), England (n = 8), Australia (n = 4), Sweden (n = 4), both England and Wales (n = 2), and Scotland (n = 2). One study each originated from Brazil, Bulgaria, Canada, France, Great Britain, and Norway. The majority of AAS users initiated use before age 30. Sports participation (particularly power sports), negative body image, and psychological disorders such as depression preceded initiation of AAS use for most users. Sources of first AAS were mainly users’ immediate social networks and the illicit market. Enhanced sports performance, appearance, and muscle/strength were the paramount motives for AAS use initiation. Conclusions Our findings elucidate the significance of psychosocial factors in AAS use initiation. The proliferation of AAS on the illicit market and social networks demands better ways of dealing with the global public health problem of AAS use.


Background
Several qualitative investigations have sought to understand the aetiology and trajectory of nonmedical AAS use initiation. However, to our knowledge, there has never been a systematic review and synthesis of the qualitative literature on this important area of nonmedical AAS use. An investigation of this type is important because a global perspective of nonmedical AAS use initiation is necessary for the understanding of this global public health problem [1].
A review and synthesis of the qualitative research on AAS use initiation is also important in light of expressed concern regarding the validity and reliability of survey research on AAS use [2]. Moreover, it has been suggested that the failure of health practitioners and public health officials to appreciate people's perception of antecedents and risk factors is a major hindrance to the success of public health interventions [3,4]. Hence, data on initiation and trajectories of AAS use are important for prevention purposes.
We carried out, as far as we are aware, the pioneering systematic review and synthesis of the qualitative studies presenting data on the initiation of nonmedical AAS use. The United Kingdom's Economic and Social Research Council's manual on the synthesis of qualitative literature [5] indorses the formulation of research questions or hypothesis prior to synthesis. The research questions guiding the present study were: (a) at what age(s) do AAS users have their debut?, (b) what are the psychosocial histories of AAS users prior to the initiation of AAS use?, (c) what are the sources of AAS users' first AAS?, and (d) what are the motives and drives for initiating AAS use?

Search strategy and inclusion criteria
We conducted a comprehensive literature search in Psy-cINFO, PubMed, ISI Web of Science, and Google Scholar. The following keywords: 'anabolic steroid', 'doping', and 'performance enhancing drug', were each used in combination with 'interview', 'focus group', and 'qualitative' for searches in PubMed and ISI Web of Science. Due to unusually high superfluous returns from the above permutation of keywords, 'anabolic steroid + doping + performance enhancing drug + interview + focus group + qualitative' was used in searches in PsycINFO and Google Scholar. The literature search was completed in June 2014. From an initial pool of 10,106 hits, 7,720 articles were evaluated after removing duplicates. In addition, a manual check of reference lists of identified studies was conducted in search of potential unidentified studies. Searches were also conducted in online databases and websites. We identified 4 new articles through this grey literature search. Thus, a total of 7,724 were settled on after eliminating duplicates. After evaluating the 7,724 papers based on titles and abstracts, 95 full-text papers were retrieved for screening.
After initial screening of the 95 full-text papers, 68 papers were identified. Of the 68 papers scrutinized, 35 studies met the following key criteria for inclusion: (a) studies presented original information on the experiences of AAS users (b) studies employed qualitative approaches in data collection (interviews, focus groups, or case studies) and presentation of results, and (c) studies were published in English. Four recent studies [6][7][8][9] and five others [10][11][12][13][14] were later discovered and included in the analysis. We again inspected the characteristics of extracted studies for similarities to curb duplicate extraction and synthesis. Thus, a total of 44 articles were included in the analysis. The literature search strategy adhered to Shaw et al.'s [15] recommendations for finding qualitative research as well as the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [16]. Figure 1 presents the process of the search and selection of relevant studies according to the PRISMA guidelines.

Data extraction and synthesis
The first author scrutinized and selected studies. Smith et al.'s [17] Interpretative Phenomenological Analysis (IPA) was used to analyse the studies because it facilitates in-depth exploration of the meanings of experiences [18]. Each full-text paper was regarded as a transcript. The first author read through the full-text papers several times, gaining an overall sense of the themes in the studies through this process. These themes were then highlighted. We developed a standardized data extraction form unto which the first author and another reviewer independently extracted the following data from the included studies: author name and publication year, country, study type, type of AAS users involved in the study, and recruitment site or mode. These characteristics are presented in Table 1. The first author independently coded the full-text papers according to the presence or absence of the following themes: (a) age(s) of first use, (b) history prior to use, (c) source(s) of first AAS, and (d) motive(s)/drive(s) for initiating use. These characteristics are presented in Table 2. Statistical inferences have little meaning in qualitative synthesis. However, the presence of a theme in multiple studies may be evidence of the validity of the theme [4]. In this regard, we have presented all the studies that fall under each theme.

Quality of extraction, included studies, and synthesis
To assess the quality of the extraction, we calculated inter-reviewer reliability for the two reviewers in SPSS version 20 (IBM Corp.) [54]. Sensitivity analysis is conducted in the synthesis of qualitative research to examine the effect of the exclusion of high or poor quality studies on the overall findings. We assessed the relevance of the included papers according to the four themes: (a) age(s) of first use, (b) history prior to use, (c) source(s) of first AAS, and (d) motivation(s) for use (see Table 2). Each theme was scored '1' thus yielding a possible total score of '4'. Subsequently, we excluded studies that scored ≤ 2 out of 4 on the themes and investigated the effect of the exclusion on our synthesis and results. Moreover, as most of the included studies were conducted in the United States, we excluded the United States studies to investigate the effect of the exclusion on the quality of our synthesis and results.

Strength of extraction, included studies, and synthesis
The inter-reviewer reliability for the reviewers was found to be Kappa = 0.82 (p < 0.001) indicating very good agreement between the two reviewers [55]. Consensus was reached on discrepant extractions through further review and discussion. Thirty-eight (38) of the 44 studies scored ≥ 3 out of 4 on the themes and were thus deemed to be of high relevance. Six studies [31,36,37,41,46,52] scored ≤ 2 out of 4 on relevance and were therefore excluded in the quality analysis. However, when we removed the study characteristics generated from these studies in the sensitivity analysis, our themes or results did not change. Consequently, we retained them in the final analysis. Similarly, the removal of the study characteristics generated from the studies originating from the United States did not affect the quality of our themes or results. Thus, they were also retained in the final analysis. The sensitivity analysis therefore indicated a strong synthesis of included studies.

Narrative synthesis
We found that majority of studies had participants initiating use before they were 30 years old. In addition, histories of negative body image, psychological disorders such as mood and depressive disorders, and participation in power sports preceded initiation of AAS use for most persons. We also found that sources of first AAS were mainly users' immediate social networks and the illicit market. Furthermore, we found that motives for AAS use were mainly enhanced sports performance, appearance, and muscle or strength.

Age of AAS use initiation
Of the 24 studies that presented the ages at which participants initiated AAS use, initiation ages ranged from 14 to 54 years. However, only 5 of the 24 studies presented participants that initiated AAS use after age 30 consistent with evidence that about 80% of AAS users initiate use before age 30 [56]. It must be noted that some studies did not specify the ages at which some or all respondents initiated AAS use (See Table 3).
Similarly, participation in other sports such as athletics, cycling, hockey, and football emerged as a prominent feature of AAS users backgrounds prior to initiation of AAS use [7,11,22,25,30,32,38,41,49,52]. This is exemplified by Josh in Bardick et al.'s study [20]. Josh was a hockey player who "needed to take steroids to become the best" (p. 138). Similarly, Maycock and Howat [34] highlighted association with 'complacent' trainers or coaches as a feature of AAS users prior to the initiation of AAS use (p. 319).
Also, Gruber and Pope [27] recount the story of Ms. A. who "took all of the supplements and ergogenic drugs that her trainer recommended, including large doses of anabolic steroids". In Maycock and Howat's study [34]: Four of the interviewed sample indicated that complacency by trainers and coaches contributed to their decision to consider use. The failure of coaches and officials to investigate large increases in body mass and strength achieved by other competitors contributed to their decision to explore use (p. 319).
AAS users also showed psychological syndromes such as mood and depressive disorders as well as troubled psychosocial histories including divorce, having suffered rape, poor parental connectedness or involvement, and poor social support [14,25,27,28,44,45,[48][49][50] prior to the initiation of AAS use. In one study [27], five females initiated AAS use after the experience of rape: None used such drugs previously…Indeed, prior to experiencing rape, these five women believed that taking anabolic substances was a weakness… Subsequent to their rape, they justified the decision to start using anabolic substances as being necessary to gain muscle mass and strength, because they thought it was impossible to grow big or strong enough "naturally" (p. 275).
Also evident as a feature of AAS users prior to initiation of AAS use was eating disorders such as anorexia nervosa [25,44]. Pope et al. [44] present the cases of four persons who initiated AAS use due to anorexia nervosa and reverse anorexia nervosa. Negative body image as well as low self-esteem and low self-efficacy also emerged as features of AAS users prior to the initiation of AAS use [7,10,25,30,32,34,35,42,44]. Cases 01 and 02 of Fudala et al.'s study [25] recount the stories of a male who "stated that he was using AASs because he lacked self-esteem and was not good-looking." and another who initiated AAS use because he "felt small and [needed to] become more muscular to accomplish [his] goals" (p. 123).   Use of other drugs also emerged as a feature of AAS users prior to the initiation of AAS use [7,14,27,28,32,38]. Nøkleby and Skårderud [38] highlighted drug use networks as well as addiction clinics as major gateways for the initiation of use. In their study, Kristian commented: I have always been offered steroids at other places as well, but it never came to anything. But when I got here [addiction clinic] it (steroids) fell right in my lap. And it was the same the last place I was in treatment. It (steroids) fell right in my lap, and that made it easy to accept (p. 495).
It also emerged that many AAS users understood the debilitating consequences of AAS but nevertheless went ahead to initiate use [10,32,34,40]. In Maycock and Howat's study [34]: Prior to initiating [AAS] use all of the men interviewed undertook information searches. These included talking to friends, gym trainers and instructors, anabolic steroid users and dealers, reading magazines, underground anabolic steroid manuals and medical journals and occasionally talking to medical practitioners (p. 320).

Sources of first AAS
Studies specified several sources of users' first AAS: the illicit market (dealers, mail order, internet etc.), coaches or trainers, clinicians or health workers (doctors, pharmacists, and veterinarians), friends or teammates, gym employees, intimate partners, and relatives (See Table 5).
In addition, training associates such as coaches or trainers and gym employees emerged as a source of AAS during the initiation of AAS use [15,16,19,22,24,49,55,56]. Clinicians or health workers such as doctors, pharmacists, and veterinarians also came up as sources of AAS during the initiation of AAS use [23,24,32,39].

Motives/drives for initiating AAS use
Motives for initiating AAS use were for: aggression, enhanced appearance, securing sports scholarships, enhanced muscle or strength, occupational (non-sporting) activities, personal security, psychological well-being or satisfaction, physiological recovery or injury prevention, sexual attraction, and for sporting or competitive activities. Other drives were trainers' approval, curiosity, family influence, use by famous athletes portrayed in the media, peer influence, and use of AAS as a sport or social norm (See Table 6). Of the above motives and drives, initiation of AAS use for enhanced appearance or body image, muscle or strength, and sports or athletic performance were most prominent in the literature. Indeed, in a study of Australian AAS users [24], the most paramount motives for the initiation of AAS use were improved appearance (46%), increase in size (33%), increase in strength (7%), and improved sporting performance (6%). Case 04 of Fudala et al.'s study [25] also tells the story of a 22-year-old male who initiated AAS use at the age of 17 "in order to increase his size and power for football" and consecutively increased his AAS consumption "in order to compete in bodybuilding events". Paula, a 39-year-old affirms the relationship between her AAS use and sports participation in Grogan et al.'s study [26] with the confession "I will stop [using steroids] when I stop competing yeah" (p. 853). Similarly, others initiated AAS use for physiological recovery or injury prevention [9,11,12,20,21,23,24,28,32,33,36,37].
Related to enhanced sports performance, enhanced occupational functioning also emerged as motive for the initiation of AAS use [6,26,34,35,40,46,47]. In support of this motive, Matt, a 33-year-old male stripper commented in Scull's study [47]: "All the guys [male strippers] take steroids, you know?…See, you won't last long in this industry if you don't use steroids. They all do steroids" (p. 567). Improved occupational functioning was again highlighted in Maycock and Howat's study [35]: For the doormen and security workers, it was about projecting physical competence; for the power lifters, it was about projecting the image of brute strength; for the sex workers or gay men using for body image reasons, it was about the presentation of a natural healthy look. For bodybuilders, it was about projecting their muscles, size and shape (p. 861).
Sexual attraction or attractiveness also emerged as an important motive for the initiation of AAS use [11,32,40,42]. This is highlighted by Kusserow's [32] finding that 18% of AAS users initiated use in order to "be more successful with the opposite sex" (p. 7). In addition, Petrocelli et al. [42] indicated that AAS use: increased and enhanced [users'] confidence and love life, as they claimed having a defined, muscular physique allowed them to meet and have sexual relations with more partners (p. 1194).
Social pressure in the form of media influence, peer influence, and sport or social norms also emerged as an important drive for the initiation of AAS use. Related to this, Petrocelli et al. [42] found long-term exposure to muscle magazines as a feature of AAS users prior to initiation of AAS use. In addition, Joe a 29-year-old male commented: "I came from a solid family that stressed competition and giving it 110%. So when I didn't see the results in the gym, I went to steroids" [22, p. 10]. In Grogan et al.'s study [26], John, a 25-year-old indicated: The more I trained, the more magazines I looked at, the bigger I wanted to be. …and there was an ITV programme [about body builders] and when I watched these people it made me feel really depressed. I didn't look as good as them. And it had a massive effect on my decision to take steroids. In fact it was probably one of the biggest reasons why I did take them seeing other people bigger than me (p. 853).
There is however contrary evidence of the influence of media on AAS use. In Walker and Joubert's study [10], 66% of respondents stated that the media had no influence on their desire to use AAS although these respondents believed that most muscular men portrayed in the media use AAS.
It is important to note however that motives for AAS use may change with time. For instance, in an Australian study [24], 46% of users indicated that they initiated use in order to improve their appearance. However, only 35% of these respondents mentioned improved appearance as motive for their most recent use indicating motive change in some users after initiation. Disparities were also discovered for other motives (p. 37). A security worker also elucidated motive change in a recent study by Kimergård [6]: At this moment in time, I'm not looking to get any bigger as a bodybuilder for example. I like to increase my strength, and now it's more for conditioning… My next cycle, I'll be doing a 'cutting' cycle, I'll be dieting and getting down to a reasonable healthy weight (p. 3).

Implications for policy and practice
Arguably, our findings represent an important basis for policymaking and planning. First, with evidence from the present study that most AAS users initiate use under 30 years, AAS use interventions should focus primarily on adolescents and young adults. Thus preventive interventions should be tailored mainly for these age cohorts. In addition, with evidence from our study that negative body image, psychological disorders, and sports participation (particularly in power sports) precede initiation of AAS use for most persons, AAS use interventions must target persons demonstrating these characteristics as well as focus on relevant environments. Moreover, AAS use interventions must be targeted at individuals with: eating disorders, low self-esteem and self-efficacy, 'doping-complacent' trainers, long-term exposure to media images of muscular persons, troubled backgrounds, drug use histories and milieus, and psychological disorders. AAS use interventions should also be aimed at athletes especially power sportspeople, doormen and security workers, male sex workers, and gay men as these groups emerged as popular AAS users in this qualitative metasynthesis.
Again, it is worrying that although some AAS users appreciated the debilitating consequences of AAS, they nevertheless went ahead to initiate use [6,10,32,34,40]. We also found that sources of first AAS were mainly users' immediate social networks and the illicit market. Furthermore, it is worthy of note that in the most recent qualitative studies presenting sources of AAS [6,10,21,38,46], the only sources were the illicit market, relatives, and friends. This is perhaps attributable to the increasing illegalization of AAS use since the 1990s [1]. Nevertheless, with the proliferation of both legal and illegal substances on the illicit market and the internet, as well as the expectedly 'drug-clean' environments of addiction clinics [38], better ways of dealing with the global public health problem of AAS use will need to be found.

Strengths and weaknesses
The present study has several strengths. To our knowledge, it is the first-ever systematic review and synthesis of qualitative studies on AAS use initiation. The systematic and advanced strategy for identifying, reporting, and synthesizing qualitative studies, the 'global' and comprehensive nature of the present study, and the inclusion of a large number of both peer-reviewed and grey literature are also notable assets.
Despite the aforementioned strengths of the present study, some limitations ought to be noted when interpreting our results. First, we restricted our analysis to English language literature. Though this is not an uncommon practice for systematic reviews [57], it is possible that the exclusion of non-English language literature influenced our results. However, it must be noted that Moher et al. [57] found no evidence of biased results with the exclusion of non-English studies. Nevertheless, it is worth pointing out again that our themes and results were robust in the sensitivity analysis. Furthermore, it is plausible that the case studies included in the present study were reported due to their 'unusual' or 'exceptional' nature. Thus, these cases may not be representative of the typical AAS user.

Conclusions
Arguably, our findings represent an important basis for AAS use interventions. Findings from the present study denote the importance of psychological and social factors in the initiation of AAS use. Our findings also complement available evidence from quantitative studies on the initiation of AAS use. There is the need for improved ways of dealing with the global problem of AAS use with the increased availability of both legal and illegal substances on the illicit market and the internet.