As sport, gender, and dietary habits are unique human experiences shaped by cultural and social norms, we followed an interpretive paradigm, informed by bounded relativism ontology (i.e., multiple realities exist and are rooted in personal context), and an epistemological orientation of social constructionism (i.e., reality is socially constructed) (28). To effectively explore the interplay between athletic, dietary, and gender identities, this study took an Interpretive Phenomenological Approach (IPA) (29). IPA facilitates exploration of an ‘insider look’ into adolescents’ experiences (30). As such, we focused on a first-person perspective to make sense of lived structures and experiences, uncovering the ‘essences’ of how adolescents decide what to eat at the intersection of their gender, athletic, and dietary identities (31, 32). As IPA has been used to explore gender, sex, eating disorders, and sport psychology, it is well-suited for our research aim (33). This IPA study is part of a bigger study the EATing in a GENdered world study (EatGen), a mixed-methods project focused on co-designing inclusive interventions to support athletes’ eating habits in schools. Ethical approval for the project was provided by the University of British Columbia Research Ethics board (H22-00121). This paper focuses exclusively on the qualitative interview results pertaining to experiences navigating intersections between athletic, dietary, and gender identities following best-practice qualitative methods (Supplementary Table 1).

Adolescents who attended a secondary school in British Columbia, Canada (grades 8–12) were recruited. The school is situated in a suburban community with predominately high income, two-parent households (34). This school was chosen due to its diverse athletic program and existing relationship with the research team.

Adolescents were recruited by posters and in-person presentations by two female research assistants who had no former relationships with the school or adolescents. At two time points (April–June 2022 and December 2022 – January 2023), interested adolescents scanned a QR code to provide their contact information to the research team. Eligible participants provided informed consent over the phone. Parents of adolescents under 18 years of age were also called to provide consent for their children. Adolescents were excluded if they were: pregnant, living with an eating disorder, on exchange, diagnosed with a severe dietary restriction (e.g., Crohns), and/or did not have access to technology to participate in a virtual interview. We purposively recruited adolescents to balance sex (male/female), as sports compete based on sex, and sport involvement (athlete/non-athlete) (35). To be considered an athlete, an adolescent had to have participated in a school sport and/or club sport (i.e., practices held after school with competition against other local schools/club teams) in the past year. Differing from traditional phenomenology work, we recruited a larger sample of adolescents (29) to provide insight into differences between groups and meet saturation for our broader projects objectives that interested teens were made aware of.

Aligned with an interpretive phenomenological approach, an open-ended interview guide (35) was developed by the first and last author based on concepts from behaviour change theory (36) and the Food Literacy Competencies for Young Adults (37) framework. After pilot-testing the guide with one adolescent, two questions were altered to ensure language was appropriate for the target population (35). This study focused on responses to questions about beliefs and experiences relating to dietary habits, gender, and sports. See Supplementary Table 2 for the interview guide used in the broader project.

Adolescents were asked to define their view of masculinity and femininity in their own words at the start of the interview. After this, adolescents self-identified how masculine or feminine they perceived themselves to be using a scale of 1 (masculine) to 10 (feminine). We elected to use this novel approach as more traditional measures, like the Bem Sex-Role inventory, are lengthy and restrictive (38) and do not facilitate our goal of understanding personal experience of identities. Thus, we aimed to facilitate a truer experience of what being masculine or feminine meant to each individual teen by using an open ended, self-identified approach to gender identity instead of using pre-determined definitions or categories.

Demographic information was collected through an online survey. Adolescents then participated in a semi-structured virtual interview in a private location lasting ~30 min (19–57 min). Audio recordings of the interviews were transcribed verbatim by a transcription service, and de-identified. The interviewers wrote case notes following each interview. Adolescents received a $30.00 CAD gift card for participating.

All three interviewers were trained in how to probe for adolescents’ experiences by the first author, a certified facilitator. After two interviews each, the first author listened to the audio and provided feedback on interviewers’ techniques to improve the consistency and depth of responses across interviewers. Interviewers were not previously known to participants and were all female university level students who had never competed in competitive sports.

Three participants were excluded from the final sample, and their interviews excluded from analysis, as they no longer met eligibility criteria after implying they had previously lived with an eating disorder during their interview.

Aligned with IPA, analysis started with open coding of transcripts line by line, where each line of data was assessed for ‘codes’ to indicating meaning in adolescents’ experiences. All coding was done inductively (i.e., from the data itself) to allow for true representation of participants experiences to guide the process (29, 39). Similar codes were then grouped together and considered for overarching commonalities within a transcript. Next, tentative overarching themes where formed (39). This process was repeated on the next transcript and so on in an idiographic nature (29). Once all transcripts were coded, differences in experiences were compared to confirm themes across adolescents experiences (39). The first author, a female researcher who previously competed in competitive sport and is now a coach of competitive high school athletes, led the analysis. The process of ‘bracketing’, where one intentionally tries to set aside their own beliefs and assumptions, (29, 39) was done in two ways. First, the lead author, who did not interact with participants and therefore had a degree of separation, led the analysis and discussed developing codes and themes with an individual of a different sex who had previous competed in sports and was not involved in the study. Second, an additional coder who did not play competitive sports reviewed all transcripts (authors two or three) during the open coding and theme consolidation stages.

Thirty-three adolescents participated (Table 1). Of these, 18 were athletes (n = 9 girls; n = 9 boys) and 15 were non-athletes (n = 4 non-binary; n = 5 girls; n = 6 boys). Self-identity ratings (1 = masculine and 10 = feminine) in each group were: 1.5 among athlete boys, 7.5 among non-athlete boys, 6.5 among athlete girls and 8.0 among non-athlete girls. No non-binary non-athletes gave the same self-identity rating. Eleven athletes played a club and a school sport and seven played only a club sport.

The IPA analysis generated three themes: (1) De-gendering of protein and muscles; (2) Food displays what I value; and (3) Being [too] masculine is bad. Differences in experiences between athletes and non-athletes from diverse genders are highlighted below. A summary of all themes and sub-themes can be found in Table 2 with additional quotes presented.

In our sample, athletes expressed being more fixated on protein intake and achieving muscular body shapes compared to non-athletes. As literature suggests that gaining muscle and protein are traditionally tied with masculine identities, (7, 24, 25) our findings paint a new picture. Athletes in this sample view muscles and protein as an inherent part of being an athlete and not an inherent part of being masculine. Though promising in terms of challenging gender norms around dietary identities, some athlete girls in this sample still felt inner conflicted for consuming protein sources like red meat that carry traditional ties to masculinity (26, 40). Research conducted in collegiate athletes suggests that women balanced their muscular athletic build with feminine norms by choosing larger, more-masculine-appearing romantic partners (42). Taken in the context of our own findings, it is possible that athlete girls and women still experience underlying gender-based pressures that create tension between their own gender identities and athletic identities that manifest through dietary identities and idealized body shapes. However, the pressure athletes’ boys may experience in the context of their bodies should not be overlooked.

Our sample considered of a wide array of sports that are often associated with traditional masculinity like basketball and soccer (16). Recent literature exploring associations between dietary habits, mental health and injury risk in young adult dancers did not find significant differences in athletes’ body acceptance or dietary habits (43). However, female athletes did report higher self-esteem (43). This may suggest that male athletes experience greater stigmatization of their bodies compared to male athletes in a traditionally male sport like hockey (16), playing an unfavorable role in a male athletes’ perceptions of their bodies. Previous research supports this notion with adolescent male tootball players reported less body dissatisfaction compared to male runners (44). As we did not have the sample size to explore themes across individual sports, future research should explore how the intersection between sport, dietary and gender identities within different sports may vary to ensure the unique pressures boys (and girls) experience in sport are not overlooked.

Adolescents highlighted social pressures related to femininity, but not masculinity, to directly impacting their dietary identities. This matches other literature where taking smaller bites, eating smaller portions and actively choosing ‘healthy’ foods are associated with a more feminine dietary identity (26). However, our study is one of the first to show that these pressure to eat ‘femininely’ are tied to personal identity (i.e., self-perception of how masculine or feminine an individual feels) and not gender identity (i.e., boy, girl or non-binary youth) or sex (i.e., male or female). This finding is important as current measurement of sex and gender in research exploring athletes’ experiences with food may not be sufficient to capture underlying norms. It is especially important as the push for more equal and equitable opportunities in sports sees greater inclusion of athletes of diverse sexes and genders training together (45). Team culture and social norms are key influencers in athletes’ food choices, (46) making it imperative that robust strategies to capture societal norms in dietary identities of athletes are developed. This will better ensure that athletes who need the most support in adopting a dietary identity focused on their performance needs instead of social norms can be identified.

Our findings on the intersection between dietary identities and gender identities through idealized bodies differ from previous literature, which suggests that weight is primarily a feminine issue (41). In this study, all non-athletes, including masculine self-identifying boys and non-binary adolescents, experienced pressure to avoid weight gain or lose weight. Further, athlete boys and girls experienced pressure to be muscular. Athletes’ focus on being muscular regardless of gender or self-identity likely comes from their lived experiences having their bodies constantly evaluated for performance (10). As such, embodied expectations to adopt a dietary identity consistent with promoting muscularity as part of athletic identity may place high school athletes at risk for disordered eating compared to non-athletes. Such observations have been corroborated in elite sport settings, (47) but not in high school sport settings until now.

Many adolescents in our study described masculinity associated with negative personality and behavioral traits. In some cases, it resulted in adolescents in this sample feeling less inclined to participate in sports. This contrasts views from other adolescents who were mostly athlete boys in this study who identified masculinity as an entrenched, and positive part of their personal identities. Previous literature suggests that sport creates environments where historically masculine traits, like being competitive or aggressive, are viewed more favorably among boys (13) but not among other gender groups (42). Recent literature suggests that gender beliefs in sport are still present across a host of sports among collegiate athletes, with most sports associated as masculine or neutral, but not feminine (48). Further, the authors suggest that ‘gender typing sports’ (i.e., how gendered a sport is) may prevent athletes from participating or staying involved due to misalignment between their own gender identity and that of a particular sport, (48) a notion that may explain our findings here. As such, greater research is needed to understand how views of masculinity may hinder or support involvement in sports and for whom.

Our findings highlight the notion that de-gendering dietary and sport identities is important, but that there are caveats to be aware of. First, removing gender pressures around protein and muscularity as masculine are important, but have created a situation where all athletes, regardless of gender in our sample, fixated on protein as a part of athletic identity. This manifested as a use of proteins supplements (e.g., powders) instead of wholefoods. As the current position from sport experts and dietitians is to promote a ‘food first but not food only’ approach, athletes’ focus on protein supplements could be problematic for overall health (8). Second, the de-gendering of protein and muscles as a part of athletic identity might increase the risk of overlooking disordering eating among female athletes. For example, over-exercising under the label of athletic identity and increasing muscle might not be obviously interpreted by coaches, trainers or parents as a form of disorder eating. Coaches, trainers and clinicians in sport spaces need to be aware of gender sensitive tools to assess athletes’ eating habits in the context of sport such as the Athletes’ Relationship with Training scale (ART) (49). Finally, programs where athletes and non-athletes learn about food together (e.g., home economic classes) need to account for the impact of athletic identities to avoid inadvertently harming an athlete’s dietary identity. For example, prioritizing food as fuel to keep esthetic and weight goals out of the picture.

Our findings may not be generalizable to all athletes as we had a relatively homogenous sample in terms of socio-economic status, geography, and ethnicity. However, our sample cuts across different grades, sports, and genders. This suggests that our work may be transferable to athletes in similar socioeconomic conditions and school systems. The number of non-binary adolescents in our sample was also small and not sufficient to fully understand this groups’ experiences. Despite this, our study is one of the first to consider the experiences of non-binary adolescents. Our analysis further did not explore the impact of culture or social influences on adolescents’ dietary habits, perceptions of sport or gender norms. As such, we can not speculate on who or what ultiamtely contributed to adolescents’ perceptions of their identities.