Introduction

Eating disorders are complex and serious mental health problems that commonly appear during adolescence. Compared with female adolescents, however, male adolescents have been relatively overlooked in research terms (Mitchison et al., 2013; Murray et al., 2017). In 2010, Murray and colleagues published a review of research on the topic of muscle dysmorphia in males in an effort to examine the appropriateness of its re-conceptualization as a feature of male disordered eating (Murray et al., 2010). Research about factors that are prospectively associated with the development of disordered eating in adolescent males has increased and there is now an understanding that phenotypic differences exist between disordered eating attitudes and behavior in adolescent males and females (Lavender et al., 2017; Mitchison & Mond, 2015). Psychometric assessment measures available to researchers and clinicians to assess disordered eating in adolescent males may be less than appropriate for this use, however, because the evidentiary basis and psychometric data they were established upon reflect female-centric disordered eating attitudes and behaviour (Murray et al., 2010). This review aims to extend current knowledge by synthesizing evidence about factors that are associated with the development of disordered eating in male adolescents and clarifying the extent to which the assessment measures used to identify those factors are appropriate for use with adolescent males..

Reviews of existing research have assisted greatly in understanding the epidemiological and demographic characteristics of disordered eating in male adolescents. Across 26 European countries, approximately 10% of adolescent males and 20% of adolescent females reported some weight reduction behavior, however, recent increases in the prevalence of weight reduction behaviors in adolescent males were compared with stable patterns for adolescent females (Dzielska et al., 2020). Approximately 5.5% of young males in the United States evidence disordered eating (Lipson & Sonneville, 2017). In Australia and Canada, the incidence of DSM-5 (American Psychiatric Association, 2013) eating disorders in male adolescents has been reported as 1.2% and 2.2%, respectively (Allen et al., 2013; Flament et al., 2015). Reviews of the contributions of both family environmental and genetic factors in the development of disordered eating indicate that substantial environmental influence is associated with genetic contributions to Anorexia Nervosa and Bulimia Nervosa (Bulik et al., 2016; Trace et al., 2013). Internationally, the prevalence of eating disorders varies with lowest rates reported in African and Latin populations followed by Japan and China (Hoek, 2016). Reviews suggest that non-Caucasian, non-western participants living in western countries are included in disordered eating research at levels lower than their proportion of population (Egbert et al., 2022).

Relational experiences with parents and carers likely contribute to disordered eating attitudes and behavior in adolescent males (Pace et al., 2018). Despite its importance, however, relatively little is known about the developmental significance of parenting factors to disordered eating in adolescent male populations. Positive and statistically significant relationships that have been demonstrated between parenting style, and parenting feeding practices have important implications for the relation between parenting and developing adolescent mental health outcomes including eating related behavior (Brown et al., 2016; Hubbs-Tait et al., 2008). In addition, parental disordered eating behavior may detrimentally influence parent–child interactions and, therefore, child development, beyond the transmission of eating behavior (Chapman et al., 2021). Variability in findings in this area may reflect a lack of consistency in operationalization and measurement of characteristics currently understood as central to disordered eating in adolescent males (Murray et al., 2017). For instance, research has suggested that adolescent males maternal but not paternal parenting style was statistically significantly associated with less extreme, but not extreme, weight control behaviors in male adolescents (Zubatsky et al., 2015). By contrast, paternal but not maternal parenting characteristics during early adolescence positively predicted negative eating attitudes and behavior in adolescent males (Pace et al., 2018). Further research that identifies parenting factors associated with the development and maintenance of disordered eating in adolescent males is, therefore, needed.

The degree of biological and social change that is evident in adolescence is unrivalled developmentally (Mendle, 2014; Paganini et al., 2021). The developmental timing of disordered eating, and its association with puberty in adolescent males, is an area of emerging importance in eating disorders research (Chapman et al., 2021; Paganini et al., 2021). Epidemiological data points to later onset of disordered eating in males compared with females, reflecting the importance of pubertal timing for onset in females (Zehr et al., 2007). Female specific factors, such as increases in ovarian hormones, have been implicated in increases in disordered eating in females but do not generalize to males (Klump et al., 2012). However, some psychosocial factors, such as anxiety, have previously been associated with increases in disordered eating in both female and male post-pubertal populations and may, therefore, generalize across the sexes (Zehr et al., 2007). Consistent with traditional conceptualizations of disordered eating, much research that has been undertaken in this area has used female participants and maintained adherence to models of disordered eating driven by attitudes and behaviors applicable to females (Neale et al., 2020; Paganini et al., 2021). Genetic, temperamental, and maturational influences on disordered eating in males appear more variable and poorly articulated than those in females (Klump, 2013), and future research directed at identifying processes unique to the relation between male disordered eating and temperamental and maturational characteristics is essential (Paganini et al., 2021).

Key in differentiating between adolescent males and females is a recognition that substantive differences exist between them in disordered eating attitudes and behavior (Murray et al., 2017). Recent reviews have articulated a range of phenotypic characteristics that are now broadly understood to be uniquely associated with male disordered eating behaviors (Mitchison & Mond, 2015). While evidence accounts for body dissatisfaction as a predominant feature of disordered eating in both sexes, traditional female-centric conceptualizations of disordered eating are primarily concerned with a drive for thinness that is typically attained through processes of restriction or purging (American Psychiatric Association, 2013). By contrast, the 2010 review was instrumental in re-conceptualizing male disordered eating by placing muscularity at its heart (Murray et al., 2010). Phenotypically distinct process in males may instead reflect dual drives to increase body muscularity and, simultaneously, to increase leanness (Tylka, 2011).

Subsequent reviews have further articulated processes central to the development of male disordered eating (Mitchison & Mond, 2015; Murray et al., 2017). The first represents the pursuit of a male body ‘ideal’ characterized by increased muscularity and a concomitant increase in leanness acquired as a result of the internalization of socio-cultural processes. Second, this ideal has been articulated in association with muscularity orientated eating that serves to facilitate muscle growth and enhance body leanness (Murray et al., 2017). Mutually dependent process of bulk and cut account for the need to support muscle development through food intake while restricting food intake to attain leanness and enhance muscle appearance (Lavender et al., 2017). Muscularity orientated disordered eating has been associated with behavioral excesses in the form of increased exercise to develop muscle (Lavender et al., 2017; Murray et al., 2017). This may involve an increase in gym attendance or engagement in various forms of body building exercise, and may be associated with obsessive, compulsive or anxiety-related features where eating related targets are, or are not, met (Murray et al., 2010). Knowledge about factors associated with the development and maintenance of male-specific eating disorder phenotypes, and an informed developmental understanding of those associations, are currently needed (Jaworski et al., 2019; Lavender et al., 2017).

The internalization of a male body ideal is central to current conceptualizations of adolescent male disordered eating and male exposure to media that depicts idealized male body imagery is potentially important in influencing the development of disordered eating attitudes and behavior (Sweeting et al., 2015). Images displayed in media may influence males and females through similar processes of internalization of socio-culturally driven body ideals (Lavender et al., 2017), even if the ideals that are pursued between the sexes are dissimilar. Western media depictions of female ideals focus on themes similar to those that underscore current conceptualizations of female disordered eating presentations, namely striving to attain thinness (Mitchison & Mond, 2015). By contrast, media representations of male bodies that fit current sociocultural ideals, focus on the dual themes of a drive toward increased muscularity and increased leanness (Tylka, 2011). It is not by coincidence that efforts to attain these outcomes engender concern about current versus idealized muscularity and body fat (Lavender et al., 2017), the very themes that have recently emerged as central to muscularity orientated adolescent male disordered eating (Murray et al., 2017).

Media may advance adolescent male muscularity and leanness ideals in similar ways to those that have long pervaded media presenting idealized female body images (Bardone-Cone & Cass, 2006). Themes of diet, exercise, muscularity, substance use, and health related risk implicitly or explicitly extol the virtues of pursuing disordered eating related behavior and depiction of non-ideal bodies and marginalization further endorse socio-cultural ideals (Murray et al., 2016; Sweeting et al., 2015). The implications of the influence of media for disordered eating attitudes and behavior in adolescent males is continuously changing (Sweeting et al., 2015). The range of media currently available to adolescents is extensive, and assessment of factors of influence in driving muscularity orientated idealized body images may well be extended to include all forms of media where the presentation or marketing of idealized images occurs (Murray et al., 2016). Despite recent advances, much remains to be understood about the role of media in the development of disordered eating attitudes and behavior in adolescent males. In addition to describing associations between variables, mediational, moderational and modelling analyses hold promise to further elucidate the role of media in the establishment and maintenance of disordered eating attitudes and behaviors in adolescent males (Lavender et al., 2017).

Valid assessment is essential to developing an accurate understanding of adolescent male disordered eating attitudes and behavior. Many psychometric assessment measures that have been utilized to assess female populations have been utilized with males with reliability; however concern remains about their validity for use with males (Smith et al., 2017). Psychometric assessment measures developed consistent with criteria which typically underlie characteristics of female disordered eating commonly exclude items that capture the desire to increase muscle mass (Reas et al., 2012; Smith et al., 2017). For instance, the Eating Disorders Inventory-3 (EDI-3; Garner, 2004) assesses people’s perceptions of specific body parts such as “I think my hips are too big” or “I think my buttocks are too large”. These concerns may not generalize well across the sexes, potentially rendering them inappropriate for assessing male body image (Smith et al., 2017). Furthermore, a comparison of the Eating Disorder Examination Questionnaire (EDE-Q; Fairburn & Beglin, 1994) across a matched sample of male and female adolescents diagnosed with Anorexia Nervosa found that males scored significantly lower on the EDE-Q global score and the shape concern and weight concern subscales than females (Darcy et al., 2012). Disparities such as these may be addressed by developing normative male data on existing psychometric assessment measures that have been developed consistent with female-centric assumptions (Mitchison & Mond, 2015). This solution may, however, not demonstrate sufficient reach in addressing male specific attitudes and behavior.

Though adolescent males typically return lower scores on psychometric assessment measures developed consistent with female disordered eating criteria, they nevertheless do endorse items on those measures (Carey et al., 2019; Smith et al., 2017). Importantly, therefore, males may be susceptible to internalizing a broader range of socio-cultural body ideals than adolescent females, suggesting a degree of applicability to males in both drive for thinness and bulk and cut models of disordered eating (Carey et al., 2019; Smith et al., 2017). Research has reported that male as well as female participants endorse psychometric assessment items that reflect a drive for thinness (Thurfjell et al., 2003), and for some minority male populations, similar levels of endorsement between male and female participants have been noted (Cella et al., 2010; Klimek et al., 2021). One of the implications of using psychometric assessment measures developed consistent with models of disordered eating in females is an underestimation of rates of prevalence and patterns of presentation of disordered eating in males (Mitchison & Mond, 2015). Some measures have been developed in consideration of male features of disordered eating (for instance the Eating Disorder Assessment fo Men; Stanford & Lemberg, 2012), however, they require further evaluation in clinical settings (Murray et al., 2017; Smith et al., 2017).

Current Study

Much research has been undertaken in adolescent male disordered eating since 2010; however, reviews of prospective longitudinal literature, and systematic reviews, are limited. A systematic review of prospective longitudinal research that examines factors associated with the development of disordered eating attitudes and behavior in male adolescents is, therefore, timely. Specifically, a review of research that extends current understanding of the influence of familial, developmental, social, emotional, behavioral and socio-cultural processes, as well as the psychometric assessment of those processes, on trajectories towards disordered eating, is warranted. This review aimes to address this gap by examining recent prospective research that identifies psychosocial risk factors relevant to the development of disordered eating in adolescent males. The primary aim of the review is to identify psychometrically assessed psychosocial risk factors that are prospectively associated with the development of disordered eating attitudes and behavior in adolescent males. Its secondary aim is to assess the appropriateness of the psychometric assessment measures used to identify those factors in the included research.

Methods

Protocol and Registration

The systematic review was conducted in accordance with the PRISMA guideline for reporting systematic reviews (Page et al., 2021). The protocol was originally registered with PROSPERO (CRD 42018085385) and refined to focus solely on male adolescents. It was developed in line with the procedures outlined in the Cochrane Handbook for systematic reviews (Higgins & Green, 2011).

Inclusion/Eligibility Criteria

Participants

Studies that included participants aged 11–19 years (i.e.,: adolescents) were eligible for inclusion in this review. Studies that included participants who met current DSM or ICD diagnostic criteria for an Eating Disorder at baseline, and those that included participants with neurodevelopmental disorders, diabetes, or who had undergone bariatric surgery, were excluded. Studies that included participants who were involved with child protective services or who were incarcerated were also excluded.

Outcomes

Studies were included in this review if they examined psychosocial risk factors identified to influence the development of disordered eating behavior in male adolescents.

Study Design

Prospective longitudinal studies that captured psychometrically assessed data from at least two points in time and had previously published psychometric properties were included in our review. Psychometric assessment measures targeted in the review assessed a variety of factors including generic disordered eating attitudes and behavior (n = 5), diagnosable categories (Anorexia Nervosa, Bulimia Nervosa and Binge-Eating Disorder, n = 2), characteristics included in well-established assessment measures (EDI, EDE-Q, n = 3), as well as weight reduction behaviors (n = 1) and aspirational items (n = 1).

Search Strategy

An electronic search of the databases Scopus (Elsevier), PsycINFO (EBSCO), PsycARTICLES (APA), Medline (Ovid), Web of Science Core collection and PubMed was completed on May 13, 2022. The search was limited to studies published in the English language in peer-reviewed journals between 2010 and 2022, to coincide with the year the Murray et al. (2010) review was published. The databases were searched through a multi-field format using the terms listed in Table 1. A manual search of reference lists from relevant articles and chapters failed to identify any additional research that met the inclusion criteria.

Table 1 Search terms for electronic database search
Full size table

Assessment of Methodological Quality

A modified version of the Downs and Black (1998) checklist was used to assess the methodological quality and risk of bias of each included study. Of the original 27 checklist items, 11 items in the areas of reporting (questions 1–3, 6, 7, 10) external validity (questions 11 and 12) and internal validity (questions 17–20) were considered appropriate to assessment in systematic reviews of longitudinal studies. Previous authors have utilized the same items in assessing methodological quality and risk of bias in systematic reviews of longitudinal studies (Duch et al., 2013; Vanderloo, 2014).

Data Extraction

Relevant data from the included studies were recorded in a data extraction table designed for this review. Extraction was undertaken by authors SB and JM independently of each other and checked for consistency. The following information was extracted: key study details (author, year, and country), study characteristics (design and duration), setting (population), sample characteristics (baseline sample size, final sample size, sex, mean age, attrition rate), disordered eating outcomes (restriction, purging, binge eating), Eating Disorder outcome measure and predictors/factors associated with disordered eating development.

Data Synthesis

The review process extracted data and synthesized it in narrative and tabular form. The potential to use meta-analytic approaches was explored, however, as effect, data was not extracted or assessed, meta-analytic approaches were judged as being of limited use.

Results

Study Selection

A total of 347 records were identified through the database searches. Of these, 72 studies were removed as duplicates and a further 103 studies were removed upon title and abstract screening. The remaining 77 full text studies were reviewed by two authors SB and JM and inter-rater reliability was estimated at (K = 0.91) using Cohen’s kappa (Landis & Koch, 1977). Any disagreements in final study selection were resolved through discussion and a total of 21 studies were identified as meeting inclusion criteria. Figure 1 presents a PRISMA flow chart of the study selection process.

Fig. 1
figure 1

PRISMA flow chart depicting study selection process

Full size image

Study Characteristics

Of the k = 21 studies that were eligible for inclusion in the review, k = 4 studies (Juarascio et al., 2016; Perkins & Brausch, 2019; Petersen & Hyde, 2013; Stojek et al., 2017) were based in USA, k = 3 studies (Van Durme et al., 2018; Verschueren et al., 2020; Verstuyf et al., 2014) were based in Belgium, k = 3 studies (Espinoza et al., 2019; Ferreiro et al., 2011, 2014) were based in Spain, k = 3 studies (Hoffmann & Warschburger, 2017, 2019; Sehm & Warschburger, 2018) were based in Germany, k = 2 studies (Jackson & Chen, 2011, 2014) were based in China, k = 2 studies (Mougharbel et al., 2020; Zaitsoff et al., 2020) were based in Canada, k = 1 study (Rosenrauch et al., 2017) was based in Australia, k = 1 study (Dakanalis et al., 2015) was based in Italy, k = 1 study (Rotenberg & Sangha, 2015) was based in the UK and k = 1 study (Dzielska et al., 2020) was based in Europe and in North America. Table 2 presents the characteristics of included studies.

Table 2 Characteristics of included studies
Full size table

Research Design

All eligible studies utilized a prospective longitudinal design, however the duration of the studies varied from 1 to 16 years. A median study length of two years was reported in k = 7 studies.

Sample Characteristics

The sample sizes of studies included in this review ranged from n = 116 to n = 639,194 participants. Across eligible studies, data was drawn from a total of n = 658,917 participants. Females were over-represented in k = 17 studies. Female participation in the remaining studies ranged from 46–49.6%. The age of participants at the commencement of each study ranged from 11.52 years (SD = 0.31) to 16.54 years (SD = 1.19). The studies demonstrated considerable sample homogeneity with all participants drawn from secondary school based community samples.

Quality Assessment

The Downs and Black checklist recommends that all studies that receive a score above 5 out of a possible 11 indicate low risk of bias (Downs & Black, 1998). The risk of bias was assessed by two of the review authors (SB and JM). No studies were rated below 5 and thus no studies indicated a high level of bias. This may, in part, be due to the prospective nature of all the included studies. Rate of agreement for overall quality of each study using the Downs and Black (1998) tool was considered excellent (Cicchetti, 1994) as reflected by an intra-class correlation coefficient of 0.91. The results of the quality assessment are summarized in Table 3.

Table 3 Risk of bias within studies
Full size table

Outcomes

All studies included in this review reported finding at least one male-specific predictor of disordered eating development in adolescents. Results are summarized in Table 4 and narratively synthesized below. Factors were classified under five factor groupings as follows: psychological factors (driven primarily by an individual’s cognitive, affective, somatic or behavioral processes); body appearance factors (driven primarily by an individual’s perception of, or attributions about, their physical self); sociocultural factors (driven primarily by implicit or explicit social or cultural expectations, assumptions or values); family factors (driven primarily as a result of parental and sibling influences within family systems) and peer factors (driven primarily as a result of peer based influences including assumptions about, and reports from, peers). In total, 35 male-specific predictors of disordered eating and behavior in males including psychological factors (n = 14), body appearance factors n = 10), sociocultural factors (n = 4), family factors (n = 3) and peer factors (n = 4).

Table 4 Summary of factors found to predict adolescent male disordered eating
Full size table

Psychological Factors

Fourteen psychological factors that predicted the development of disordered eating in male adolescents were identified in k = 12 of the included studies. Disordered eating attitudes were significantly and positively predicted by high affect reactivity and high internalizing symptoms in combination with high distress intolerance at baseline but not over time (Juarascio et al., 2016). In addition, depressive symptoms mediated (Ferreiro et al., 2011, 2014), and self-esteem and suicidal ideation prospectively predicted (Ferreiro et al., 2011; Perkins & Brausch, 2019) disordered eating behavior. In addition, internalizing symptoms, poor effortful control and identity confusion all decreased over time and were therefore identified as significant negative predictors of social emotional wellbeing (Verschueren et al., 2020), and increased emotional eating, in conjunction with eating characterized by loss of control, were significant predictors of subsequent disordered eating attitudes (Stojek et al., 2017). Affective (depressive) symptoms and low self-esteem predicted subsequent dietary restraint and binge eating and, in turn, binge eating predicted subsequent affective symptoms (Dakanalis et al., 2015; Sehm & Warschburger, 2018). By contrast, increased lack of trust in close others and loneliness were both identified as predictors of increased bulimic symptoms (Rotenberg & Sangha, 2015). One theme evident in the relation between psychological factors and disordered eating behavior was role of affect. Specifically, increased negative affect, reduced positive affect, and increased affect reactivity were all positively associated with the development of disordered eating attitudes and/or behavior over time. In addition, perception of self, either with respect to identity or esteem, emerged as central among psychological factors.

Appearance Factors

Ten appearance-related predictors of disordered eating amongst male adolescents were identified in k = 10 studies. In particular, BMI and overestimation of one’s body weight were significant positive predictors of weight reduction behaviors (Dzielska et al., 2020; Jackson & Chen, 2014). Increased muscularity concerns, increased weight and shape concerns (Hoffmann & Warschburger, 2017, 2019), and overweight or underweight status (Hoffmann & Warschburger, 2017) were identified as significant positive predictors of eating attitudes and behaviors, dietary restraint and muscularity-oriented behavior over time (Hoffmann & Warschburger, 2017, 2019). Zaitsoff et al. (2020) identified increased body checking as a significant positive predictor of eating disordered attitudes and behaviors amongst male adolescents, whilst studies by Ferreiro et al. (2011), Jackson and Chen (2011, 2014), identified body dissatisfaction as predictive of eating attitudes and behaviors in mid-adolescence. Self-objectification via appearance anxiety, and body shame were identified as significant prospective predictors of dietary restraint and binge eating amongst male adolescents (Dakanalis et al., 2015) Appearance related themes were predominantly related to the over- or under-estimation of weight and/or shape in self or others, regardless of whether it was related to estimating muscularity or leanness.

Family Factors

Three family related factors were identified in k = 2 of the retained studies. The activating strategy of rumination in attachment anxiety, and the deactivating strategy of emotional control in attachment avoidance with their caregivers statistically significantly predicted bulimic symptoms in male adolescents (Van Durme et al., 2018). Additionally, perceived appearance pressure from parents and dating partners predicted adolescent male eating attitudes and behavior (Jackson & Chen, 2014). Importantly, no studies that met criteria for inclusion in this review assessed sibling influences in disordered eating.

Sociocultural Factors

Four individual socio-cultural factors related to adolescent male disordered eating were identified in k = 8 studies. Eating attitudes and behavior in early adolescence were predicted by perceived appearance pressure from mass media to lose weight or to be thin (Jackson & Chen, 2011, 2014). Significant predictors reported to apply to middle adolescence included perceived pressure from peers, friends, media and fear of negative evaluation (Jackson & Chen, 2011; Mougharbel et al., 2020). The internalization of appearance ideals significantly and positively predicted a range of disordered eating outcomes including binge eating and dietary restraint in two studies (Dakanalis et al., 2015; Sehm & Warschburger, 2018), disordered eating attitudes and behavior (Espinoza et al., 2019), eating regulation (Verstuyf et al., 2014) and disordered eating attitudes and behavior, dietary restraint and muscularity-oriented behavior (Hoffmann & Warschburger, 2019). Sociocultural factors all related on some level to the delivery and internalization of sociocultural ideals through media.

Peer Factors

Four individual peer factors related to adolescent male disordered eating were identified in k = 3 studies. Eating behavior was positively and significantly predicted by sexual harassment via self-surveillance (Petersen & Hyde, 2013) and perceived appearance pressure from desired/current dating partners significantly increased disordered eating at particular points in adolescence (Jackson & Chen, 2014). For male adolescents, the inverse relationship between friends healthy eating habits and meal skipping reduced across adolescence when skipping lunch but not breakfast meals (Rosenrauch et al., 2017). Peer factors, therefore, appeared to have potential to mitigate risk that resulted from peer interactions.

Psychometric Assessment of Predictors

Across the k = 21 eligible studies, n = 14 different psychometric assessment measures were used to gather information about adolescent male disordered eating. Three studies utilized novel surveys that were designed for their study, with some form of psychometric assessment reported within or subsequent to the retained study (Dakanalis et al., 2015; Rosenrauch et al., 2017; Verstuyf et al., 2014). The most commonly used outcome measures were the EDE-Q (Fairburn & Beglin, 1994) and the Child Eating Disorder Examination Questionnaire (Kliem et al., 2017), each of which was used to generate data in k = 6 (28.6%) of studies. The psychometric assessment measures used in each study are included Table 2. Measure characteristics including the sex and age of the normative population used for measure validation are included in Table 5.

Table 5 Summary of psychometric assessment measures used in studies retained for inclusion in the review
Full size table

Discussion

Research regarding disordered eating in adolescence has historically focused on females, however, research into the development of disordered eating attitudes and behaviours in adolescent males is emerging (Mitchison et al., 2013; Murray et al., 2017). A systematic review of prospective longitudinal research was undertaken to synthesize current knowledge about factors that are prospectively associated with the development of disordered eating in adolescent males. The primary aim of the review was to identify psychometrically assessed, psychosocial risk factors that are prospectively associated with the development of disordered eating attitudes and behaviors in adolescent males. Its secondary aim was to assess the appropriateness of the psychometric assessment measures used to identify those risk factors in the reviewed research. Across the k = 21 studies that were eligible for inclusion in the review, we identified 35 risk factors and 12 psychometric assessment measures. The risk factors ranged across psychological, body appearance, sociocultural, family, and peer concerns and influenced a range of disordered eating outcomes identified in the reviewed studies and synthesized in this review. The review indicated that risk for disordered eating in male adolescents is established developmentally however, given limitations in the validity of the psychometric assessment measures utilized in the reviewed studies, caution may be warranted with respect to findings.

Across studies, adolescent male disordered eating attitudes and behaviors were predicted by all five factor groupings. In general terms, psychological factors predicted characteristics including those related to cognitive and affective processes, regulation and suicidal ideation. They also predicted perfectionism and self-esteem. Appearance concerns such as muscularity, weight and shape concerns, and weight status influenced eating attitudes and behavior, dietary restraint and muscularity orientated behavior and appearance. Sociocultural factors such as engagement of mass media and engagement of peers via screen use appeared to play a role, possibly in influencing the extent to which individual male adolescents internalize socially or culturally determined appearance standards and the perception of self that result. Sociocultural factors were, therefore, instrumental in determining, and/or depicting, the appearance standards to which male adolescents held themselves and the eating attitudes and behavior that resulted. These last factors included eating regulation, binge eating, and muscularity orientated eating behavior outcomes.

One pattern that emerged in our review indicated that developmentally important relationships may be predictive of subsequent eating attitudes and behavior, however this association appeared to be relatively complex. For instance, affect regulation is established in the context of early developmental relationships (Feldman, 2012). Within adolescence, well developed affect regulation is protective against future risk of mental ill-health. For instance, processes of attachment were predictive of adolescent affect regulation (Van Durme et al., 2018). The importance of family to the development of disordered eating was further demonstrated in findings that attachment anxiety was related to Bulimia Nervosa and that attachment avoidance was related to emotional control (Van Durme et al., 2018). A second factor, fundamental to early relationships, is trust (Rotenberg & Sangha, 2015). Trust is built within the context of reciprocal or synchronous attachment relationships (Stolle & Nishikawa, 2011), and trust beliefs were found to negatively predict loneliness and Bulimia Nervosa (Rotenberg & Sangha, 2015). The development of strong, regulated, early attachment relationships is, perhaps, one of the most significant predictors of affect regulation and, therefore, reduced disordered eating attitudes and behaviors in adolescent males.

Two additional factors of risk identified as occurring early in trajectories toward disordered eating were an internalizing disposition in general (Verschueren et al., 2020) and appearance related anxiety in particular (Dakanalis et al., 2015; Juarascio et al., 2016). Both, arguably, also develop in the context of early developmentally important relationships (Gar et al., 2005). Appearance-related anxiety was found to positively influence disordered eating behavior via its influence on negative affect (Dakanalis et al., 2015). This influence was exerted via limitations in food intake and increased food regulation (Dakanalis et al., 2015; Jackson & Chen, 2014), In addition, an internalizing disposition and high distress intolerance were found to be strong and positively associated with disordered eating attitudes (Espinoza et al., 2019; Stojek et al., 2017).

Across studies, a clear association was evident between affect and disordered eating attitudes and behavior in adolescent males. A considerable range of affective processes including affective reactivity, positive and negative affect, emotional eating and distress intolerance were evaluated in the reviewed studies. High negative affect, in particular, was found to positively predict dietary restraint (Dakanalis et al., 2015), binge eating (Sehm & Warschburger, 2018) and disordered eating (Jackson & Chen, 2011, 2014). In addition, affective reactivity was found to be a positive predictor of disordered eating attitudes (Juarascio et al., 2016) and depressive symptoms were positively predicted by binge eating (Sehm & Warschburger, 2018). Importantly, binge eating and dietary restraint were reported to have a reciprocal relation (Dakanalis et al., 2015).

A further clear pattern of association across studies was evident between the internalization of socio-cultural ideals and disordered eating behavior in adolescent males. The internalization of socio-cultural ideals was positively predicted by high levels of negative affect (Dakanalis et al., 2015), an association that underscored its importance as a predictor of disordered eating behavior (Hoffmann & Warschburger, 2019). The internalization of socio-cultural ideals positively predicted self-objectification, which in turn predicted appearance anxiety together with dietary restraint and binge eating (Dakanalis et al., 2015). Further, the internalization of socio-cultural ideals interacted positively with distress intolerance to predict increases in eating disordered attitudes in adolescence (Juarascio et al., 2016). Disordered eating was influenced by media in young adolescent males (Jackson & Chen, 2011; Sehm & Warschburger, 2018). It also predicted binge eating in research by Sehm and Warschburger (2018) and weight/shape concerns and muscularity concerns were, generally, positive predictors of disordered eating (Hoffmann & Warschburger, 2017).

Appearance related factors identified in the review were somewhat more varied than those previously mentioned, however, they often appeared to work in conjunction with psychological and socio-cultural factors that predicted disordered eating and behavior in adolescent males. In addition to self-objectification, body checking and body dissatisfaction both positively predicted disordered eating behavior in mid-adolescence (Ferreiro et al., 2011; Zaitsoff et al., 2020). Sexual harassment, identified as potentially preceding self-objectification, may influence the commencement of self-surveillance and, therefore, add risk for disordered eating (Dakanalis et al., 2015; Petersen & Hyde, 2013). Body dissatisfaction, in addition to Bulimia Nervosa, was also identified as a risk factor for suicidal ideation (Perkins & Brausch, 2019). One study in particular addressed the dual themes of a drive toward increased muscularity and a drive toward increased leanness (Hoffmann & Warschburger, 2019). Weight and shape concerns were found to be prospective predictors of restrained eating and the internalization of an athletic ideal was found to prospectively predict muscularity orientated behavior (Hoffmann & Warschburger, 2019). Similarly, muscularity concerns were broadly associated with eating concerns (Hoffmann & Warschburger, 2017). Time spent watching television positively predicted subsequent BMI scores (Mougharbel et al., 2020). In turn, BMI scores positively predicted disordered eating attitudes and high loss of control eating in conjunction with high emotional eating positively predicted subsequent disordered eating (Jackson & Chen, 2014; Stojek et al., 2017).

Psychometric Assessment Measures

The studies included in this review identified factors of risk for the development of disordered eating in adolescent males; however, these factors were identified using psychometric assessment measures that may not adequately capture male adolescent eating attitudes and behaviour (Darcy & Lin, 2012; Murray et al., 2018). In the validation studies published in support of the psychometric assessment measures used in the reviewed studies, between 47 and 100% of participants were female. Of those psychometric assessment measures, only one (The Drive for Muscularity Scale; McCreary, 2007) specifically assessed features consistent with current conceptualizations of disordered eating in adolescent males. This is not to say that other assessment measures do not reflect male specific characteristics. For instance, the desire to attain leanness, currently understood as a feature of male disordered eating, may be informed by female-centric assessment measures that seek responses to items about thinness. This suggests that the attitudes and behaviors reported in response to assessment items in the identified studies may more accurately reflect the degree to which adolescent males endorse attitudes and behaviors consistent with adolescent female disordered eating. For example, the drive for thinness subscale of the EDI-3 (Garner, 2004) and the restraint subscale of the EDE-Q (Fairburn & Beglin, 1994) include items which may appropriately assess ideals traditionally associated with female disordered eating. By contrast, males may experience a desire for increased muscularity and increased leanness, and therefore experience fear of weight or muscle loss, as opposed to gain (Dakanalis et al., 2015; Mitchison & Mond, 2015). Thus, on an item-by-item comparison of the EDE-Q, and despite being diagnosed with Anorexia Nervosa, male adolescents were less likely than females to affirm wanting a flat or empty abdomen, being uncomfortable while eating in front of others, eating in private, or holding a desire to lose weight (Darcy & Lin, 2012).

Items were also identified on psychometric assessment measures that may have relevance to the assessment of disordered eating in adolescent males, but which may benefit from refinement to account for the variability in characteristics of presentation across the sexes. For example, dieting may be characterized by calorie counting to change (increase or decrease) weight (Lavender et al., 2017) depending on the motivation. Similarly, food preoccupation or restriction may occur to facilitate or suppress the consumption of specific foods or food groups appropriate to the desired outcome (Griffiths et al., 2013; Murray & Touyz, 2013). By contrast, oral control may involve restricting the intake of certain foods, depending on whether increased muscle or leanness is the desired outcome (Badenes-Ribera et al., 2019). Finally, items referring to dissatisfaction with body parts, such as the hips and buttocks, may be broadened in an effort to reflect the range of potentially desired outcomes as a function of disordered eating symptomatology across the sexes (Garner, 2004; Lavender et al., 2017). The factors identified in this review as prospectively associated with the development of disordered eating in adolescent males are primarily concerned with increasing muscle and increasing leanness (Tylka, 2011). The development of psychometric assessment measures that demonstrate validity in accounting for processes that facilitate attainment of these ideals in adolescent males is essential in advancing the area (Hoffmann & Warschburger, 2019).

Limitations and Future Directions

This review has some noteworthy limitations. Inclusion and exclusion criteria were used to determine the studies that would be included in the review. Those that were reviewed were prospective longitudinal research studies and, as a result, a range of cross-sectional characteristics that are potentially important in building a comprehensive picture of adolescent male disordered eating were not reviewed. Rather, and where possible, the implications of development over time were reported. A majority of the reviewed studies spanned the timeframe where puberty likely occurred, and it was unfortunate that changes in functioning as a result of puberty could not be reviewed because the onset of puberty was reported to a limited extent in only one study (Espinoza et al., 2019). The implications of puberty for the developmental course of disordered eating in adolescent males, therefore, remain unclear. Similar concerns are present with respect to factors such as culture and ethnicity. Where it has been possible to report in an informed way on the implications of culture or ethnicity that has been included in the review. However, less than half the reviewed studies included information about participant’s cultural or ethnic status limiting the generalizability of findings to non-Western populations. Thus, the implications of a range of factors for the development of disordered eating attitudes and behaviors in adolescent males remain less than clear.

A comprehensive understanding of disordered eating in adolescent males remains absent from the existing evidence-base, in large part due to limitations in psychometric assessment for this cohort. Efforts to delineate male factors require the accurate identification of those factors prior to the development of comprehensive models of male disordered eating and, subsequently, the development of comprehensive and integrated models of adolescent risk across sexes. Achieving this goal will provide an account of both male- and female-specific factors of risk, and also account for the intersection of those factors such that risk is represented at the level of a syndrome of features applicable across adolescence.

This review also has implications for the development of strategies of preventative intervention, as well as treatment. Research suggests that strategies of preventative intervention that function at a family systems level early in childhood hold promise to address subsequent social, emotional and behavioral functioning in adolescence and, therefore, adulthood. Whether indicated or selected, relationally based strategies of preventative intervention appear to have potential to mitigate both genetic and relational aspects of disordered eating. In identifying and clustering factors associated with the development of disordered eating in adolescent males, this review also informs components of treatment which are responsive to change as it occurs in adolescence. The development of psychometric assessment measures that can capture symptoms as they commence early in adolescence will be of the upmost assistance in interrupting trajectories toward diagnosable eating disorders in adolescent males. Aspects of those diagnoses, as they are developed, are also informed by this review.

Conclusion

The evidence-base regarding disordered eating in adolescence has traditionally focused on females, though literature considering the perspective of adolescent males is emerging. This review was undertaken to synthesize research findings about psychosocial risk factors that are prospectively associated with the development of disordered eating attitudes and behavior in male adolescents. The risk factors identified in the review were psychological, body appearance, sociocultural, family and peer factors, and highlight a range of characteristics that are unique to adolescent male disordered eating. The review also critically appraised the psychometric assessment measures used in the reviewed studies to identify those factors of risk. Accurate assessment is imperative in generating reliable and valid research and informing clinical practice and the review concluded that existing female-centric psychometric assessments normed predominantly on female participants may be inappropriate for use with adolescent males. Informing strategies for preventative intervention to avert trajectories toward disordered eating and, ultimately, eating disorders in adolescent males remains a priority for research in this field.