The Design and Implementation of a Novel Mental Health Literacy Educational Intervention Program in Gaelic Footballers

The design of the “GAA [Gaelic Athletic Association] Mental Health—Injury and a Healthy Mind” educational intervention program was informed by formative research that identified a lack of education, stigma, and negative self-attitudes as barriers to mental health help-seeking in Gaelic footballers.⁹  An MHL program was deemed most appropriate to address these factors, and a matrix of objectives (see Supplemental Material A, available online at http://dx.doi.org/10.4085/1062-6050-0463.22.S1) was developed, with the program determinants informed by the definition¹  and components of MHL.²  The intervention was underpinned by the Theory of Planned Behavior and the Help-Seeking Model. We agreed on the program content (see Supplemental Material B, available online at http://dx.doi.org/10.4085/1062-6050-0463.22.S2) after a comprehensive review of the literature and an examination of existing MHL educational programs and materials^12,15  by a multidisciplinary group of experts (a sport and exercise psychologist, athletic therapists, and a sport scientist). The content was designed to be delivered through a brief 25-minute online presentation, supplemented by a Mental Health Toolkit information pack available online.

Program design and content clarity were piloted with a convenience sample of 13 athletes from a variety of sports (age 24.6 ± 3.5 years old; men = 6; women = 7). Participants attended the presentation online via Zoom (version 5.0.1), received the Mental Health Toolkit by email, and completed a process evaluation measure after the presentation. The process evaluation measure, similar to that used by Liddle et al,¹⁵  consisted of 9 rating statements using a 4-point Likert scale ranging from 1 (not at all) to 4 (very) and 5 open-ended questions that asked participants to provide feedback on the program. Our pilot study results showed that participants rated the MHL educational program positively, with an overall average score of 3.5 ± 0.4. Detailed results, feedback, and evidence of program changes are available in Supplemental Material C, available online at http://dx.doi.org/10.4085/1062-6050-0463.22.S3.

A quasi-experimental design was used. We adopted a between-groups (intervention and control) and within-groups (time: baseline [T1] and immediately after [T2], 1 week after [T3], and 1 month after [T4]) intervention design.

Male and female Gaelic footballers (n = 160) were recruited via social media to an intervention (n = 85) or control group (n = 75). Ethical approval was granted by the Institutional Research Ethics Committee (No. 20191104), and informed consent was obtained.

The intervention group completed the baseline measures online using Google Forms. The 25-minute educational program was delivered in small groups (1 to 8 participants) via PowerPoint (Microsoft Corp) and Zoom at a time convenient to the participants. The presentation was facilitated by a qualified and experienced sports psychologist with master’s and doctoral qualifications who was not involved in program development. The facilitator underwent training, which included a detailed presentation plan and script. After the presentation, participants were given an opportunity to ask questions. We also provided participants with a Mental Health Toolkit, outlining the program content, by email to refresh their knowledge when necessary.

Follow-up measures, which were emailed to participants, were completed by the intervention group immediately after the Zoom presentation and at 1 week and 1 month afterward. Reminders were not necessary for measures completed immediately after the intervention, but reminder emails were sent after 2 days to those who failed to complete follow-up measures at 1 week and 1 month, with responses accepted up to 3 days after reminders were issued. After all repeat measures were administered, intervention group participants completed a process evaluation measure to determine their feedback. The control group supplied measures at baseline, and, as they did not receive the intervention, a 1-week period was allowed before postintervention measures were obtained; these were repeated after 1 week and 1 month. A flow chart of participants at each time point is presented in Figure 1.

The Stigma Scale for Receiving Psychological Help (SSRPH; Table 1) is a unidimensional measure used to assess perceived stigma toward seeking professional mental health treatment.²⁰  The scale demonstrated good content and construct validity²⁰  and also had good internal consistency in athletic populations.^21,22  Good internal consistency of the SSRPH was shown in the current sample at T1 (α = .78), T2 (α = .93), T3 (α = .90), and T4 (α = .92).

The Self-Stigma of Seeking Help Scale (SSOSH; Table 1) is used to measure negative self-perceptions for seeking psychological help.²³  It is a valid and reliable measure of self-stigma, with demonstrated test-retest reliability²³  and good internal consistency in athletes.^21,22  The SSOSH displayed good internal consistency in the current sample at T1 (α = .85), T2 (α = .92), T3 (α = .91), and T4 (α = .93).

The Attitudes Toward Seeking Professional Psychological Help-Short Form (ATSPPH-SF; Table 1) is used to assess professional help-seeking attitudes²⁴  and demonstrated good internal consistency^12,21  and test-retest reliability¹²  among athletes. Good internal consistency of the ATSPPH-SF was present in the current sample at T1 (α = .78), T2 (α = .92), T3 (α = .91), and T4 (α = .92).

The Mental Health Literacy Scale (MHLS; Table 1) can be divided into 3 subscales measuring recognition (MHLS-R; 8 items), knowledge (MHLS-K; 11 items), and attitudes (MHLS-A; 16 items).²⁵  The MHLS can be used to examine all 3 major components of MHL with due regard to cultural context,³  and good internal consistency and test-retest reliability have been shown.²⁵  The MHLS displayed good internal consistency in the current sample at T1 (α = .91), T2 (α = .97), T3 (α = .96), and T4 (α = .96).

The Brunel Mood Scale (BRUMS; Table 1) assesses mood states and is composed of 6 subscales: tension, depression, anger, fatigue, confusion, and vigor.²⁶  Total mood disturbance is calculated by subtracting the positive vigor subscale from the sum of the 5 negative mood state subscales (tension, depression, anger, fatigue, and confusion) and adding a constant of 100 to eliminate negative values. Good internal consistency was evident for each subscale, and content and criterion validity were demonstrated in athletic populations.²⁶  In the current sample, the BRUMS exhibited good internal consistency at T1 (α = .89), T2 (α = .92), T3 (α = .92), and T4 (α = .89).

Data were analyzed using SPSS (version 25; IBM Corp). A significance level of .05 was set for all statistical tests. Descriptive statistics were conducted to examine participant characteristics, and separate independent-samples t tests, Pearson χ² tests, or Fisher exact tests were calculated to examine differences between the intervention and control groups. Linear mixed modeling (LMM) was applied to evaluate differences between groups over time for each of the dependent variables (SSRPH, SSOSH, ATSPPH-SF, and MHLS score). The LMM was deemed appropriate as it accounts for missing data, unequal variances, and correlated data that are common with repeated measurements on individual participants.²⁷  We adopted the intention-to-treat principle, which allows for an unbiased estimate of treatment effect and minimizes type I error.²⁸  The LMM was conducted with group, time, and group-by-time interaction treated as fixed effects and participant as a random effect. The BRUMS total mood disturbance score was treated as a covariate in all models tested to control for mood. The participants in each group were assessed as a repeated effect using the unstructured covariance matrix. Bonferroni adjustments for multiple comparisons were applied to limit type I error. Effect sizes, determined using the Cohen d, were classified as small (d = 0.2), medium (d = 0.5), and large (d = 0.8). Descriptive statistics were determined for the process evaluation responses, and open-ended responses were coded and summarized.

Of the 85 participants recruited, 70 (age = 25.1 ± 4.5 years) completed the intervention (Figure 1). No differences were noted in demographics or baseline measures between those who completed the intervention (n = 70) and those who dropped out (n = 15), and the latter were subsequently removed from the analyses. A total of 75 participants (24.4 ± 6.0 years) were recruited to the control group. No differences between groups were present for sex, age, education level, current playing level, years playing Gaelic football, or injury history (P > .05; Table 2). Average scores in outcome measures at baseline and after the intervention are shown in Table 3.

A significant group-by-time interaction for SSRPH scores was seen (F_3,124.1 = 22.9; P < .001). In the intervention group, the SSRPH score decreased from T1 to T2 (P < .001; d = 1.7), T1 to T3 (P < .001; d = 1.3), and T1 to T4 (P < .001; d = 1.5). The scores were not different between T2, T3, and T4. No differences occurred in the control group between T1, T2, T3, and T4. Differences in SSRPH scores were evident between the intervention and control groups at T1 (P = .04; d = 0.3), T2 (P < .001; d = 1.8), T3 (P < .001; d = 1.6), and T4 (P < .001; d = 1.5; Figure 2A).

We observed a significant group-by-time interaction effect for the SSOSH score (F_3,124.9 = 9.6; P < .001). In the intervention group, scores decreased from T1 to T2 (P < .001; d = 0.8), T1 to T3 (P < .001; d = 0.7), and T1 to T4 (P < .001; d = 0.7). The scores were not different between T2, T3, and T4. No differences were found in the control group between T1, T2, T3, and T4. Differences were demonstrated in the SSOSH scores between the intervention and control groups at T1 (P = .002; d = 0.5), T2 (P < .001; d = 1.2), T3 (P < .001; d = 1.1), and T4 (P < .001; d = 0.9; Figure 2B).

The ATSPPH-SF scores displayed a significant group-by-time interaction effect (F_3,127.2 = 17.4; P < .001). In the intervention group, the ATSPPH-SF scores increased from T1 to T2 (P < .001; d = 0.8), T1 to T3 (P < .001; d = 0.9), and T1 to T4 (P < .001; d = 0.7). The scores were not different between T2, T3, and T4. For the control group, the ATSPPH-SF scores decreased from T1 to T2 (P = .004; d = 0.3) and increased between T2 and T3 (P = .05; d = 0.2). Differences were present in the ATSPPH-SF scores between groups at T2 (P < .001; d = 1.2), T3 (P < .001; d = 1.0), and T4 (P < .001; d = 0.9; Figure 3).

We saw a significant group-by-time interaction effect for the MHLS scores (F_3,118.4 = 38.8; P < .001). Similarly, significant group-by-time interaction effects were observed for the recognition (F_3,125.9 = 14.3; P < .001), knowledge (F_3,125.4 = 23.3; P < .001), and attitude (F_3,118.5 = 29.2; P < .001) subscales.

In the intervention group, the MHLS scores increased from T1 to T2 (P < .001; d = 1.1), T1 to T3 (P < .001; d = 1.0), and T1 to T4 (P < .001; d = 1.1), but it was not different between T2, T3, and T4. In the control group, the MHLS scores decreased from T1 to T2 (P < .001; d = 0.4), T1 to T3 (P = .001; d = 0.3), and T1 to T4 (P < .001; d = 0.3). Differences were found in the MHLS scores between the intervention and control groups at T2 (P < .001; d = 1.3), T3 (P < .001; d = 1.2), and T4 (P < .001; d = 1.3; Figure 4A).

The MHLS recognition score in the intervention group increased from T1 to T2 (P < .001; d = 1.0), T1 to T3 (P < .001; d = 1.0), and T1 to T4 (P < .001; d = 1.0), but it was not different between T2, T3, and T4. In the control group, the recognition score increased from T2 to T3 (P = .007; d = 0.3) and T2 to T4 (P = .001; d = 0.3). Differences were demonstrated in the recognition scores between the intervention and control groups at T2 (P < .001; d = 1.2), T3 (P < .001; d = 1.0), and T4 (P < .001; d = 0.9; Figure 4B).

The MHLS knowledge score in the intervention group increased from T1 to T2 (P < .001; d = 1.1), T1 to T3 (P < .001; d = 1.1), and T1 to T4 (P < .001; d = 1.1), but it was not different between T2, T3, and T4. No differences were shown in the control group between T1, T2, T3, and T4. Differences were evident in the knowledge scores between the intervention and control groups at T2 (P < .001; d = 1.3), T3 (P < .001; d = 1.3), and T4 (P < .001; d = 1.4; Figure 4C).

In the intervention group, the MHLS attitudes score increased from T1 to T2 (P < .001; d = 0.7), T1 to T3 (P < .001; d = 0.6), and T1 to T4 (P < .001; d = 0.8); however, it was not different between T2, T3, and T4. In the control group, the attitudes score decreased from T1 to T2 (P < .001; d = 0.5), T1 to T3 (P < .001; d = 0.4), and T1 to T4 (P < .001; d = 0.4). We found differences in the attitudes scores between the intervention and control groups at T2 (P < .001; d = 1.1), T3 (P < .001; d = 1.0), and T4 (P < .001; d = 1.1; Figure 4D).

Participants rated the program positively, with an overall score of 3.6 ± 0.5 (see Supplemental Material D, available online at http://dx.doi.org/10.4085/1062-6050-0463.22.S4). They liked that the program was informative, specific to Gaelic football, well communicated, relevant, interesting, clear, and concise; addressed mental health as an important topic; was available online; and provided additional online resources (see Supplemental Material D). Of the intervention group participants who offered feedback (n = 55), the majority (n = 38, 69%) stated that there were no elements of the program that they disliked or would change, but some participants (n = 12, 22%) disliked the repetitiveness of the measures and would have liked more concise surveys. In addition, 4 participants (7%) reported that the MHL educational program was lacking detail and was too simple, only delivered on 1 occasion, supplied no new information, and required more in-depth psychology detail. A small number of participants acknowledged the need for greater interaction (n = 3, 5%), access to a recorded version of the presentation (n = 2, 4%), and more detailed information (n = 4, 7%). Participants indicated that the additional information pack received after the presentation was informative (n = 25, 45%), a useful reminder and reference tool (n = 17, 31%), and clear and concise (n = 3, 5%). Detailed results of the feedback received are available in Supplemental Material D.

Our aims were to design a novel educational intervention program for improving MHL in Gaelic footballers and examine its effectiveness. As hypothesized, the “GAA Mental Health—Injury and a Healthy Mind” educational program was effective in decreasing mental health stigma, improving attitudes toward help-seeking, and increasing the recognition and knowledge of mental health conditions compared with a control group. The effect sizes were large and the changes were sustained at 1-week and 1-month follow-ups.

The MHL educational program decreased the stigma associated with seeking and receiving psychological help from baseline to after the intervention, improvements that were retained at follow-up. Previous researchers similarly identified that a brief internet-based mental health and destigmatizing intervention¹²  and an educational mental health program^10,11  reduced stigma in elite and collegiate athletes, respectively. The reduction of stigma in Gaelic footballers was retained at the 1-month follow-up, which was not evident¹⁰  or reported in earlier studies.¹¹ 

We found differences in stigma at baseline between the intervention and control groups. With the quasi-experimental design, intervention group participants were recruited first, followed by control group participants. Despite no differences in participant characteristics at baseline, those who initially signed up may have been more willing to be involved possibly due to prior experience with mental health issues or exposure to those with mental health issues, which may explain the lower stigma scores in the intervention group. In addition, self-stigma in the intervention group at baseline (mean = 23.9 ± 5.5) was lower than levels observed in collegiate athletes (mean = 25.1–25.4),^10,22  while the control group had a greater SSOSH score (mean = 27.2 ± 7.3). Along with the possibly greater willingness of the intervention group to engage with mental health research, as outlined previously, this may be attributed to the additional pressures collegiate athletes face, such as pressure to maintain scholarships or achieve professional contracts,²⁹  resulting in greater help-seeking self-stigma than evident in amateur Gaelic footballers. Self-stigma was also lower in Gaelic footballers at the 1-month follow-up (mean = 19.9 ± 6.5) than observed at a similar follow-up in collegiate athletes who completed an MHL program (mean = 23.1 ± 5.6).¹⁰  Perceived stigma for receiving help at baseline (mean = 6.2 ± 3.3) was in line with the level observed in collegiate athletes (mean = 6.1),²²  but the control group had higher SSRPH scores (mean = 7.4 ± 4.1).

Athlete mental health is often framed using language consistent with mental illness,³⁰  so help-seeking stigma can be heightened in an athletic setting.³¹  Athletes need to be comfortable in their environment, where they are free to ask for help without judgement or negative consequences and will receive the necessary assistance from expert mental health professionals.³²  Our findings provide evidence for the implementation of a brief MHL educational program in community sport to reduce the stigma associated with help-seeking. The current intervention presented here aimed to normalize mental health and help-seeking by debunking common myths and misconceptions about mental health conditions and offering the perspectives of elite players. The use of athlete role models to convey destigmatizing information and address stereotypes¹³  can supply vicarious reinforcement to players³³  and proved to be effective in eliciting changes in stigma perceived by Gaelic footballers. By reducing stigma and creating a healthy mental health culture advocating for mental well-being and mental health help-seeking, Gaelic footballers may be better equipped to manage stressors.

Attitudes toward help-seeking improved in the intervention group, with the differences sustained at 1 week and 1 month. Previous researchers¹⁰  identified similar increases in help-seeking attitudes after a 4-week educational program in collegiate athletes (age = 19.2 ± 1.5 years) from a variety of sports. Extensive MHL programs may not be necessary, as our brief educational intervention elicited comparable effects.

Baseline help-seeking attitudes were lower in both the intervention (mean = 20.0 ± 4.6) and control (mean = 18.5 ± 5.7) groups than preintervention attitude scores in collegiate athletes (mean = 27.5 ± 4.6),¹⁰  indicating that Gaelic footballers may have worse attitudes toward help-seeking. Irish elite athletes previously reported positive attitudes and a willingness to seek sport psychology consulting when dealing with injury and rehabilitation.³⁴  The conflicting findings suggest differences in attitudes between mental health help-seeking and sport psychology consulting for physical performance benefits, with a preference for seeking help in the latter situation.³⁴  Professional athletes may also have greater access to mental health professionals and, therefore, better attitudes toward help-seeking than Gaelic football players.

As with stigma, normalizing the conversation and increasing visibility can improve overall attitudes toward mental health issues.³⁵  Exposing Gaelic footballers to an MHL educational program may provide that visibility, and using specific examples of the mental health experiences of Gaelic footballers can help players relate to the content, supporting better help-seeking attitudes. Thus, a sport-specific educational intervention program may improve attitudes toward help-seeking and is an essential step in preparing Gaelic footballers to effectively manage mental health conditions.

The educational intervention program improved the recognition and knowledge of mental health issues and help-seeking resources, with increases sustained at follow-up. Improved recognition or knowledge was similarly facilitated in collegiate^10,11  and youth athletes,¹⁵  coaches,³⁶  elite sport staff,³³  parents of youth sports participants,^37,38  and sports club leaders.^39,40  Baseline MHL was greater in the intervention (mean = 129.5 ± 10.9) and control (mean = 125.8 ± 17.8) groups than in student-athletes (mean = 123.4 ± 11.1).¹⁰  Mental health literacy decreased in the control group from baseline to immediately after the intervention and follow-up, whereas the scores on the attitudes subscale decreased across all time points. This suggests that MHL can fluctuate over time and may be affected by experiences of daily living. Better MHL reflects a greater intention to seek help⁴¹ ; combined with a lack of knowledge of mental disorders and their symptoms and poor awareness of help-seeking services previously acknowledged as key barriers to help-seeking among Gaelic footballers,⁹  the findings indicate the importance of an MHL education program for players.

Improving MHL can not only benefit the development and use of knowledge to improve overall mental well-being but also change attitudes, help overcome stigma, and create opportunities to seek help.³  In particular, those with greater mental health knowledge have more intentions to engage with those who have a mental health condition,⁴²  which may be due to less stigma and more positive attitudes toward mental health. Interventions can also support intervention participants in transferring newly acquired knowledge or helping behaviors with family, friends, and social networks.³³  Organized sport, particularly Gaelic football (the most popular participatory sport in Ireland),⁴³  is a promising avenue for mental health education and promotion, whereby teammates look out for and support one another on the field (a concept that can be extended off the field), to facilitate and support mental well-being among players.¹⁵  Therefore, mental health and well-being MHL educational programs in community sports should be seen as an essential component of athlete care and well-being, and stakeholders must consider providing mental health programs for all players across the organizations.

Participants positively rated the MHL educational program and acknowledged that it addressed mental health as an important topic needing more attention and discussion in Gaelic football environments. Gaelic football is a popular community sport in Ireland, and the current findings can be translated to similar sports environments internationally. In addition, problems identified with previous interventions^2,13  were addressed: the program was underpinned by an appropriate theoretical framework, addressed all domains of MHL, and used valid and reliable psychometric measures to assess changes in MHL, and we recruited a large sample size and control group.^2,13 

We adopted a quasi-experimental design, as a randomized controlled trial was not feasible, which may have resulted in unobserved and potentially confounding variables between groups attributed to the lack of random assignment. However, this design is more robust than a single-group pre-post design and was deemed most appropriate. Participants’ previous mental health experiences or education was not measured, and self-selection bias may have had an effect, with some players being potentially more willing to engage. Future researchers may minimize these factors by accounting for prior mental health education and using multiple recruitment strategies. Intentions to seek help were not measured, and follow-up was conducted over only a 4-week period, which did not allow us to examine the long-term implications and sustained benefits of the program. Future investigators should consider longer follow-up periods or use longitudinal studies to evaluate actual help-seeking behaviors after the intervention.

The “GAA and Mental Health—Injury and a Healthy Mind” educational program, a brief, accessible, structured, and targeted program designed specifically for Gaelic games athletes, may be a practical and viable approach for eliciting effective changes in MHL. Our findings demonstrated that the program was beneficial in improving Gaelic footballers’ recognition, knowledge, and attitudes and reducing help-seeking stigma, with large effect sizes evident. This pioneering and novel educational program sparks the conversation around mental health and paves the way for the support of mental health and well-being. The program also essentially showed utility in transferring to other community sports beyond Gaelic football, extending the benefits for improving MHL and help-seeking across Ireland and internationally, which is key to further normalizing the conversation around mental health in the community.

This research was funded by the Technological University of the Shannon: Midlands-Midwest Seed Funding.

Supplemental Material A. Matrix of objectives.

Found at DOI: http://dx.doi.org/10.4085/1062-6050-0463.22.S1

Supplemental Material B. Program content.

Found at DOI: http://dx.doi.org/10.4085/1062-6050-0463.22.S2

Supplemental Material C. Pilot study results and feedback.

Found at DOI: http://dx.doi.org/10.4085/1062-6050-0463.22.S3

Supplemental Material D. Participant feedback.

Found at DOI: http://dx.doi.org/10.4085/1062-6050-0463.22.S4