Epidemiological assessment of mental health screening protocols for high school athlete participation in the USA, 2021

Abstract

Objectives

To examine state-level variation in mental health screening practices for high school athletes in the USA and evaluate associations between screening protocols and adolescent depression-related and suicide-related outcomes.

Design

Cross-sectional study of preparticipation examination (PPE) forms and Youth Risk Behaviour Surveillance System (YRBSS) data.

Setting

High school athletics across all 50 US states and the District of Columbia.

Participants

732 676 high school students who completed the 2021 YRBSS survey (grades 9–12).

Primary and secondary outcome measures

Prevalence of depressive symptoms and suicide-related behaviours, measured using YRBSS items on sadness/hopelessness, suicide ideation, planning, attempts and suicide attempts resulting in injury. State-level PPE forms were reviewed for presence and depth of mental health screening, including Patient Health Questionnaire-4 (PHQ-4) and eating disorder (ED) risk questions.

Results

Screening practices varied widely, ranging from no mention of mental health in some states to required evaluations with formal referral protocols in others. Of 732 676 students, 503 861 (68.7%) reported feeling sad or hopeless, 608 416 (83.0%) seriously considered suicide and 64 072 (8.8%) attempted suicide in the past year. States with required screening protocols had higher reported rates of depressive symptoms (71.0% vs 68.7%, p<0.001) but lower suicide attempt rates (8.0% vs 8.8%, p<0.001) compared with states with no screening. Use of PHQ-4 and ED risk questions was associated with lower suicide attempt prevalence (8.4% vs 8.8%, p<0.001).

Conclusions

High school athletes represent a vulnerable group in need of structured and standardised mental health screening. State-level screening policies were associated with lower suicide attempt rates, though causal inference is limited by the cross-sectional design. Standardised use of validated mental health tools may support early identification and intervention for at-risk adolescents.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Used publicly available preparticipation examination (PPE) forms from all 50 US states and D.C. to assess screening practices.

• Developed a standardised 0–3 rating system for state-level screening requirements.

• Linked PPE data with a large, nationally representative adolescent dataset (Youth Risk Behaviour Surveillance System 2021, n=732 676).

• Cross-sectional design precludes causal inference.

• Reliance on self-reported survey data and state-level aggregation may introduce reporting and measurement bias.

Introduction

The mental health of high school athletes has become an area of increasing concern as rates of adolescent depression and suicide have risen nationwide.¹,³ While participation in organised sports has been associated with protective mental health benefits,⁴ including a greater sense of school belonging, social support and reduced risk of depression and suicidality in adolescents,^{5 6} sports involvement may also introduce unique stressors such as performance pressure, balancing athletic and academic responsibilities and injury-related challenges, all of which can negatively affect psychological well-being into adulthood.⁷,⁹ The WHO defines adolescents as people aged 10–19, a period marked by rapid development and increased exposure to mental health issues, risky behaviours and victimisation, with significant health vulnerabilities including a high prevalence of mental health problems, engagement in risky behaviours and increased victimisation rates.¹⁰ Furthermore, high school athletes face unique stressors, including performance pressure, balancing academics and sports, and physical injuries due to the internal and external psychosocial stressors they may face from themselves, coaches and loved ones to perform.¹¹ Therefore, mental health issues, which are already prevalent among adolescents, can have a significant impact on athletic performance and overall well-being of these athletes.

Despite growing recognition of the significance of mental health in the context of high school athletics, the current landscape of mental health screening practices in high schools varies widely across schools and between states across the USA.¹² High school athletes present with several risk factors for predisposition to poor mental health outcomes given their vulnerable age range superimposed by performance expectations from parents, coaches and peers. Research indicates that parental pressure contributes to a threatening environment, heightened anxiety and reduced enjoyment among youth athletes.¹³,¹⁵ Thus, critical evaluation of current approaches to mental health assessment in high school athletes in the USA is warranted, as the appropriateness and feasibility of standardisation remain uncertain. While several mental health screening instruments have been developed for use in athletic populations, most have been validated in collegiate or professional athletes rather than in high school samples. Examples include the Sport Interference Checklist, developed for collegiate athletes and the Student Athlete Relationship Inventory, which was originally developed in a sample of high school and collegiate athletes and later in collegiate students exclusively.¹⁶,¹⁸ Additionally, the Athlete Psychological Strain Questionnaire and the Sports Mental Health Assessment Tool have demonstrated utility in identifying mental health concerns in elite and collegiate athletes but have not been psychometrically validated among high school athletes.^{19 20} More recently, Donohue et al developed the Mental Health Disorders Screening Instrument for Athletes, which demonstrated strong reliability and validity in a collegiate athlete sample.²¹ Despite these advances, few if any athlete-specific mental health screening instruments have undergone psychometric validation in high school athletes, representing a critical gap in the literature. General mental health instruments such as the Patient Health Questionnaire-9PHQ-9, Generalized Anxiety Disorder–7 questionnaire (GAD-7) and Eating Attitudes Test–26 (EAT-26) are commonly used in adolescents and school-based settings, but their validity in high school athlete populations remains largely untested.

The purpose of this study is to assess the state of mental health screening practices nationwide, investigate associations between the efficacy of screening practices, if they exist, and adolescent mental health outcomes. While many studies have identified the correlation between being a youth-athlete and predisposed risk for poor mental outcomes, none have examined the adequacy and efficacy of current screening on a state-by-state level. State-level preparticipation examination (PPE) policies represent a structural factor that may influence opportunities for identification and referral of adolescents at risk for poor mental health outcomes quantified in this study as rates of depression and suicidality. These policies determine whether and how mental health questions are asked during athletic clearance, the use of validated instruments, and the presence of referral protocols. Variation across states can, therefore, be understood as variation in exposure to standardised opportunities for detection and referral. This conceptual model provides the rationale for linking state-level screening requirements to aggregated student-level outcomes while recognising that such analyses remain ecological in nature and cannot demonstrate individual-level causality.

Methods

Study design and sample

In this cross-sectional study, publicly available PPE forms,²² which students are required to submit prior to participation in high school athletics, were investigated on the websites of major high school athletic associations within each state. A structured data abstraction form was developed to systematically review each state’s publicly available PPE form as provided by the state-level association for high school athletics. Each form was obtained directly from official state high school athletic association websites, and where multiple forms existed, the most recent was selected. To standardise comparisons across jurisdictions, a Likert-type scale was applied to categorise the presence and extent of mental health screening content. In assessing the extent of mental health screening overall within each state, a score of 0 was assigned if no such screening was identified, 1 for suggested screening (eg, voluntary screenings available, but no formal assessment on PPE), 2 for superficial screening (eg, presence of a PHQ-4 or other questions pertaining to mental health) or 3 for required screening (eg, presence of a PHQ-4, eating disorder (ED) questions, and/or other questions regarding mental health wherein a positive score was followed by further evaluation). Each state was classified into one of these four distinct categories. This approach provided a consistent ordinal framework for quantifying variation in screening requirements across states. Prior to full coding, the authors conducted a pilot review of ten PPE forms to refine the scoring rubric and establish consistency across raters. Each form was screened by three different authors independently before being assigned a final score through consensus. While formal inter-rater reliability statistics (eg, Cohen’s κ) were not calculated, this independent review and consensus approach served as a safeguard to enhance consistency and reduce subjectivity in coding.

State-level data were then correlated with participant findings from the Youth Risk Behaviour Survey (YRBS), a nationally representative questionnaire conducted biennially by the Centers for Disease Control and Prevention to monitor health-related behaviours and risk factors among high school students in the USA. Data from the 2021 YRBS sample (N=738 112) were used in this study. The national YRBS used a three-stage, cluster sample design to obtain a nationally representative sample of public and private school US students in grades 9–12, allowing for a consistent assessment of behaviours and risk factors prevalent among high school students during the surveyed period. Additional information about sampling, data collection, response rates and processing is available in the YRBS overview report.²³ The full YRBS questionnaire, data sets and documentation are publicly available online.

Measures

Sociodemographic information

The student demographic characteristics gathered included age, sex, and racial and ethnic background. Students were categorised into six groups based on race and ethnicity, which included white or Caucasian, black or African American, Asian or Asian American, Hispanic or Latino, Native Hawaiian/Pacific Islander/American Indian/Alaska Native or multiple races or ethnicities. Descriptive analyses were conducted to characterise the pooled sample.

Survey measures

Scoring of mental health screening content was informed by two objective measures: the PHQ-4, a four-part questionnaire which allows for an ultra brief and accurate measurement of core signs of depression and anxiety, and another set of four questions pertaining to disordered eating risk. These objective measures were incorporated as anchors to enhance consistency and comparability across states, complementing our qualitative review of form content. The presence of these questionnaires provided a standard present on forms across some states that provided a foundational comparison measure to quantify efficacy of the form. The PHQ-4 asks: Over the last 2 weeks, how often have you been bothered by the following problems? Feeling nervous, anxious or on edge (0/1/2/3); Not being able to stop or control worrying (0/1/2/3); feeling down, depressed or hopeless (0/1/2/3); little interest or pleasure in doing things (0/1/2/3). Scores are rated as normal (0–2), mild (3–5), moderate (6–8) and severe (9–12). The disordered eating risk questions were as follows: Are you happy with your weight? (Yes/No); Are you trying to gain or lose weight? (Yes/No); Has anyone recommended you change your weight or eating habits? (Yes/No); Do you limit or carefully control what you eat? (Yes/No). The national landscape of screening practices was mapped based on the team’s scores, as well as based on presence of the objective components of the forms (PHQ-4 questions, ED questions or both).

Depression-related and suicide-related behaviours

This analysis included five variables about depression and attempted suicide from 2021 Youth Risk Behaviour Surveillance System (YRBSS) data. Participants were asked the following: During the past 12 months, did you ever feel so sad or hopeless almost every day for 2 weeks or more in a row that you stopped doing some usual activities? (Yes/No), During the past 12 months, did you ever seriously consider attempting suicide? (Yes/No), During the past 12 months, did you make a plan about how you would attempt suicide? (Yes/No), During the past 12 months, how many times did you actually attempt suicide? (0 times/1 time/2 or 3 times/4 or 5 times/6 or more times), If you attempted suicide during the past 12 months, did any attempt result in an injury, poisoning or overdose that had to be treated by a doctor or nurse? (I did not attempt suicide during the past 12 months/Yes/No). Each variable was dichotomised, with the presence (signified by a selected answer of Yes or 1 or more times) or absence (signified by No or 0 times) of each category of violence being recorded as 1 or 0, respectively, for each student. Students who answered at least one of our five questions of interest were included in analyses.

Analytical framework

Because PPE screening requirements are enacted at the state level, whereas outcomes are reported at the individual level in the YRBSS, an ecological approach was used. State-level policies were correlated with aggregated prevalence estimates of depression-related and suicide-related behaviours. Associations related to the inclusion of validated tools, such as the PHQ-4 and/or ED risk items, were examined separately. Potential confounding or effect modification by factors including provider availability, school-based mental health resources, socioeconomic characteristics and baseline mental health burden was acknowledged but not adjusted for, given the descriptive and cross-sectional design of the study.

Statistical analysis

Analyses were performed using IBM SPSS (V.26). Outcomes were analysed at the state level, where YRBSS binary responses were aggregated into continuous prevalence estimates. Because outcomes were aggregated into state-level prevalence estimates (continuous proportions), assumptions of analysis of variance (ANOVA) (eg, approximate normality of means, homogeneity of variances) were considered reasonable in this ecological context. ANOVA was used to compare mean prevalence estimates across screening policy categories, with the primary intent being descriptive and exploratory rather than inferential, as well as across positive responses to depressive and/or self-harm behaviours and the presence of PHQ-4 and ED questions within the PPE forms. P values reported in the results reflect one-way ANOVA tests of overall group differences, followed by pairwise post hoc comparisons between policy categories (suggested vs none; superficial vs none; required vs none). Bonferroni adjustments were applied to account for multiple comparisons. The independent variable for all analyses was state screening policy category, coded on a 0–3 scale to reflect the absence, suggested presence, inclusion or requirement of mental health screening content in state PPE forms. Dependent variables were state-level prevalence estimates for depression symptoms and suicide-related behaviours as reported in the YRBSS, with stratified analyses conducted where applicable by the presence of PHQ-4 and/or ED items on state PPE forms. Analytical expectations were framed as exploratory and descriptive. We anticipated that jurisdictions with more comprehensive screening requirements might exhibit different prevalence estimates compared with those with absent or limited requirements. These expectations were not formulated as formal inferential hypotheses but were intended to provide a framework for interpreting observed patterns.

Patient and public involvement

Patients and the public were not involved in the design, conduct, reporting or dissemination plans of this research. This study was based entirely on publicly available data and did not involve direct interaction with patients, student-athletes or members of the public. As such, the development of research questions, outcome measures and recruitment strategies did not include patient or public input. However, the implications of this work aim to inform policy and screening practices that ultimately affect student-athletes, with the potential for future research to include stakeholder perspectives from students, parents, educators and healthcare providers.

Results

Mental health screening practices by state

Figure 1 illustrates the mental health examination ratings across the USA based on high school athletics preparticipation forms from each state. The map is colour-coded to represent different levels of screening to denote where mental health examination is required (red), where superficial screening is in place (pink), where screening is suggested (yellow) and where high school athletic associations do not perform screening (grey). Figure 2 provides a detailed map showing the presence of PHQ-4 and/or ED risk questions within high school athletics preparticipation forms for each state. Screening tools in several states include only ED risk questions (blue), some use only PHQ-4 (light green), and some states incorporate both ED risk and PHQ-4 questionnaires (striped blue/green), while others include neither the ED risk questions nor PHQ-4 on their forms (grey). Consistency across reviewers was maintained through independent screening of each state by three authors and subsequent consensus discussions. Although formal reliability coefficients were not computed, this process minimised variability in coding outcomes.

Figure 1. Map detailing mental health examination ratings based on high school athletics preparticipation form of each state. Note: This figure was created by the authors from publicly available data and is licensed under CC BY (Creative Commons Attribution licence).

Figure 2. Map detailing the presence of PHQ-4 and/or ED risk questions within high school athletics preparticipation form of each state. Note: This figure was created by the authors from publicly available data and is licensed under CC BY (Creative Commons Attribution licence). ED, eating disorder; PHQ-4, Patient Health Questionnaire-4.

Demographic characteristics

Sociodemographic characteristics of the sample from the YRBSS data were assessed. The study included 732 676 participants, with a relatively even distribution across ages 14–18. The sample was slightly 52.2% female and predominantly white or Caucasian (410 335, 56.9%), with representation from black or African American (82 371, 11.4%), Asian or Asian American (28 204, 3.9%), Hispanic or Latino (134 591, 18.7%), Native Hawaiian/Pacific Islander/American Indian/Alaska Native (30 965, 4.3%), and multiracial groups (34 642, 4.8%).

Analyses

Table 1 describes the prevalence of depression- and suicide-related behaviours among high school students, categorised by the extent of mental health screening undergone, which ranges from no evaluation (rating 0) to full evaluation required (rating 3). The percentage of students who felt sad or hopeless, considered attempting suicide or made a plan about how they would attempt suicide was highest for students in states which required mental health screening compared with other groups. Meanwhile, the percentage of students who actually attempted suicide and the percentage of suicide attempts that resulted in an injury were lowest in states which required mental health screening compared with other groups.

Table 1. Prevalence of depression-related and suicide-related behaviours in high school students who underwent required, superficial, suggested or no mental health screening.

Variable	Mental health screening rating
	None	Suggested		Superficial		Required
			P value*		P value*		P value*
Felt sad or hopeless	68.7%	67.7%	0.005	67.9%	<0.001	71.0%	<0.001
Seriously considered attempting suicide	82.8%	81.2%	<0.001	82.4%	<0.001	83.7%	<0.001
Made a plan about how they would attempt suicide	85.9%	84.3%	<0.001	85.4%	<0.001	87.0%	<0.001
Actually attempted suicide	8.8%	9.8%	<0.001	9.1%	<0.001	8.0%	<0.001
Suicide attempt resulted in an injury, poisoning or overdose that had to be treated by a doctor or nurse	2.9%	3.3%	<0.001	3.0%	0.024	2.6%	<0.001

P values reflect one-way ANOVA tests with Bonferroni-adjusted pairwise comparisons (suggested vs none; superficial vs none; required vs none). Participants who completed the YRBS survey were asked the following: During the past 12 months, did you ever feel so sad or hopeless almost every day for 2 weeks or more in a row that you stopped doing some usual activities? (Yes/No); During the past 12 months, did you ever seriously consider attempting suicide? (Yes/No); During the past 12 months, did you make a plan about how you would attempt suicide? (Yes/No); During the past 12 months, how many times did you actually attempt suicide? (0 times/1 time/2 or 3 times/4 or 5 times/6 or more times); If you attempted suicide during the past 12 months, did any attempt result in an injury, poisoning or overdose that had to be treated by a doctor or nurse? (I did not attempt suicide during the part 12 months/Yes/No). Each variable was dichotomised, and the presence (signified by a selected answer of Yes or 1 or more times) or absence (signified by No or 0 times) of each category was recorded as 1 or 0, respectively, for each student.

Bold values denote statistically significant differences (P < 0.05).

^*Significant if p<0.05.

ANOVA, analysis of variance; YRBS, Youth Risk Behaviour Survey.

Table 2 highlights the prevalence of depression-related and suicide-related behaviours among high school students whose PPE forms included a PHQ-4 questionnaire, ED risk questions or both. For states where forms included the PHQ-4, the percentages of students who felt sad or hopeless, students who considered attempting suicide and students who made a plan about how they would attempt suicide were all higher compared with those that did not use the PHQ-4. The actual suicide attempt rate was lower in states that used the PHQ-4 than those without it. For students screened with ED risk questions, a higher proportion of students felt sad or hopeless vs states that did not use ED risk questions. Students with ED risk screening showed minimal differences in suicidal behaviour outcomes. When both PHQ-4 and ED risk questions were included, depressive symptoms and considered attempting suicide were slightly more prevalent, but suicide attempt and injury rates remained comparable to students without the presence of both screenings.

Table 2. Prevalence of depression-related and suicide-related behaviours in high school students whose preparticipation forms contained a PHQ-4 questionnaire, disordered eating risk questions or both.

Variable	PHQ-4			Disordered eating			Both
	Yes	No	P value*	Yes	No	P value*	Yes	No	P value*
Felt sad or hopeless	69.8%	68.6%	<0.001	69.5%	68.0%	<0.001	69.8%	67.8%	<0.001
Seriously considered attempting suicide	83.3%	82.5%	<0.001	82.9%	82.9%	0.880	83.3%	82.8%	<0.001
Made a plan about how they would attempt suicide	86.6%	85.6%	<0.001	86.0%	86.2%	0.014	86.6%	86.1%	<0.001
Actually attempted suicide	8.4%	8.9%	<0.001	8.6%	8.8%	0.060	8.4%	8.8%	<0.001
Suicide attempt resulted in an injury, poisoning or overdose that had to be treated by a doctor or nurse	2.9%	2.8%	0.002	2.9%	2.8%	0.186	2.7%	2.8%	0.580

Participants who completed the YRBS survey were asked the following: During the past 12 months, did you ever feel so sad or hopeless almost every day for 2 weeks or more in a row that you stopped doing some usual activities? (Yes/No); During the past 12 months, did you ever seriously consider attempting suicide? (Yes/No); During the past 12 months, did you make a plan about how you would attempt suicide? (Yes/No); During the past 12 months, how many times did you actually attempt suicide? (0 times/1 time/2 or 3 times/4 or 5 times/6 or more times); If you attempted suicide during the past 12 months, did any attempt result in an injury, poisoning or overdose that had to be treated by a doctor or nurse? (I did not attempt suicide during the part 12 months/Yes/No). Each variable was dichotomised, and the presence (signified by a selected answer of Yes or 1 or more times) or absence (signified by No or 0 times) of each category was recorded as 1 or 0, respectively, for each student.

Bold values denote statistically significant differences (P < 0.05).

^*Significant if p<0.05.

PHQ-4, Patient Health Questionnaire-4; YRBS, Youth Risk Behaviour Survey.

Discussion

In the present study, we evaluated the landscape of mental health screening practices among high school athletes on a nationwide basis and investigate associations between current state screening practices with rates of adolescent depression and suicide. We found that there is a spectrum of practice of screening in this population, ranging from no screening in place to required protocols in certain states. The clinical tools leveraged also vary within the preparticipation forms analysed in this study; two objective measures examined here were a set of 4 ED risk questions and the PHQ-4, both of which were implemented in varied capacities across the USA. A lack of standardised practice was observed alongside variation in reported outcomes; however, causality cannot be inferred.

These findings should be interpreted within the ecological framework of the study. Jurisdictions with required screening demonstrated higher prevalence of reported sadness, hopelessness and suicide planning, yet lower prevalence of suicide attempts and attempt-related injuries. This pattern is consistent with the possibility that stronger screening policies facilitate recognition of symptoms while also creating pathways for evaluation and referral; however, the associations observed cannot be assumed to reflect direct causal effects. Alternative explanations, including differences in baseline mental health burden, reporting practices and contextual factors such as provider supply, school-based resources or cultural attitudes towards disclosure, may contribute to the observed variation. Given the cross-sectional design and reliance on aggregated state-level data, the results should be understood as descriptive associations between policy categories and reported outcomes rather than evidence of policy effectiveness at the individual level.

Our study found that after analysis of the nationwide protocols for screening mental health in high school athletes, a substantial number of states, particularly in the Midwest and Northeast regions, have mandatory mental health screenings for high school athletes, highlighting a regional prioritisation of mental health given by state-level interventions and protocols. For example, the state of Delaware, where mental health screening prior to high school athletics participation is required and whose preparticipation form contains both a PHQ-4 and ED risk questions, requires a referral and clearance by the athlete’s primary care provider if he or she scores 3 or higher on the PHQ-4. Of note, associations between such requirements and outcomes should be interpreted cautiously, as other contextual factors may also contribute. In contrast, several states, notably in the western and southern regions, lack mental health screening requirements of any kind; preparticipation evaluation from the Alabama High School Athletic Association includes a physical exam and concussion assessment, while the California Interscholastic Federation places limitations on practice time, a maximum of 18 hours of practice time per week and no more than 4 hours per day but does not screen formally. Existing data suggest that only one-fifth of public schools in the USA engage in mental health screenings,²⁴ with the need for practical guidance at the state level and a universal standardisation of expectations and measures noted as major limitations.^{25 26}

While there is limited literature that has examined the disparity of mental health screening at a national level, several geographical factors may correlate with these findings. First, a shortage of primary care providers and trained mental health specialists in certain regions, particularly in rural and underserved areas, significantly impacts the availability and implementation of mental health screenings for high school students and athletes.²⁷ In areas with limited access to primary care, fewer opportunities for early detection and intervention can be detrimental to the adolescent and young adult population in those areas.²⁸ Provider shortages can also contribute to overburdened healthcare providers, reducing the likelihood of implementation and consistent utilisation of mental health protocols. A lack of provider availability may also discourage school systems from mandating mental health screenings due to an inability to address any identified concerns with follow-up care. Additionally, while the prevalence of mental health concerns in rural areas has also been shown to parallel that of urban environments,²⁹ prior research has suggested that residents in rural areas are less likely than residents of urban areas to seek professional help for psychological concerns.³⁰ In particular, student-athletes who identify as male or are earlier on in their educational careers may have a stigma toward seeking sport psychology consulting,³¹ which may be exacerbated by the difficulty faced in accessing such resources in certain areas. Further research may explore solutions to mitigate the impact of primary care shortages amidst mental health screening implementation, such as integrating telehealth services or promoting school-based health initiatives.

Our study did result in some paradoxical outcomes. The findings displayed in table 1 show that states with required mental health screenings reported higher rates of self-reported symptoms and planning behaviours but lower rates of attempts. These associations are consistent with possible detection and referral effects, but alternative explanations such as differences in baseline mental health burden or reporting practices cannot be excluded. Despite these heightened rates, the prevalence of true suicide attempts and injury-related suicide attempts was lowest in states with required screenings. While this may seem counterintuitive, this may suggest that states with required mental health screenings as part of their physical clearance for high school athletes may identify and record higher rates of sadness, hopelessness and suicide planning behaviour across their general student populations. An explanation could be that these states with screening practices were motivated to implement these practices due to a pre-existing higher baseline risk for poor mental health outcomes. The lower rates of suicide attempts in these states may then reflect the success of their policies in addressing what could have been worse outcomes without the screenings. These indicators of poor mental health would potentially go undetected in states not mandating these screenings and thus potentiate artificially low rates of these behaviours. However, while this may suggest that screening requirements contribute to greater detection of symptoms and potentially facilitate referral, alternative explanations must also be considered. States with higher baseline rates of depressive symptoms may have been more likely to implement stronger screening requirements, raising the possibility of reverse causality. In addition, variation in reporting practices, cultural norms around disclosure or differences in survey administration may influence prevalence estimates independent of screening policy. These factors underscore the importance of interpreting the observed paradox as a descriptive association within an ecological framework, rather than as evidence of direct causal effects.

An additional consideration is the possibility of reverse causality and differential reporting. States with higher baseline rates of depressive symptoms or suicide-related concerns may have been more likely to adopt stronger screening policies, which could contribute to the observed associations. Likewise, differences in cultural norms around disclosure, survey implementation or willingness to self-report sensitive information may influence prevalence estimates independently of screening practices. These alternative explanations highlight the need to interpret the paradox as an ecological association rather than evidence of direct policy effects.

In addition to detection and referral effects, alternative explanations should be considered. It is possible that states with a higher baseline burden of depressive symptoms or suicide-related concerns were more likely to adopt stronger screening requirements, creating the potential for reverse causality in the observed associations. Differential reporting practices may also contribute; for example, variation in cultural norms around disclosure, survey administration or willingness to self-report sensitive information could influence prevalence estimates independently of screening policy. These factors highlight the importance of interpreting the findings as descriptive associations within an ecological framework rather than as evidence of direct policy effects.

There is a considerable overlap between states that require mental health screening in high school athletics and those using both the PHQ-4 and ED risk questionnaires in preparticipation forms (figures1 2). Both criteria independently show a statistically significant decrease in the rate of high school students who attempted suicide compared with states with superficial or suggested screening protocols (8.0% vs 9.1% and 9.8%, p<0.001) and those that did not use both tools (8.8% vs 8.4%, p<0.001). Adolescents engaged in sports face heightened pressures that may exacerbate mental health conditions such as depression, anxiety and EDs,³²,³⁴ with some findings suggesting athletes are particularly vulnerable to these disorders.³⁵ While this study cannot determine whether these tools directly improve outcomes, associations with use of validated tools suggest potential differences in detection and reporting.

This study has several important strengths. By using a large, nationally representative sample from the YRBSS, the findings describe patterns in reported outcomes across the USA. The use of validated screening tools, such as the PHQ-4 and ED risk questions, in combination with electronic pre-participation forms bolstered the assessment, though further work is needed to establish whether standardised screening contributes to improvements. Finally, focusing on state-level practices helped highlight geographic disparities, providing valuable insights that could inform targeted policies and interventions. However, this study is not without limitations. First, the analysis relied on general adolescent YRBSS data rather than athlete-specific outcomes, which may not fully capture the unique experiences of student-athletes despite the study’s focus on pre-participation screening policies. Additionally, the cross-sectional design and absence of longitudinal data prevent any conclusions about the durability or temporal sequence of associations between screening requirements and mental health outcomes. While the focus on state-level data provides a broad overview of mental health screening practices and their implementation across the USA, it may fail to capture nuances and variability in policies, practices and resource allocation at the local or school district level. The use of self-reported survey data introduces potential biases, including reporting and recall biases, which may limit the precision and generalisability of the findings. Finally, while ANOVA was retained for this descriptive work, more advanced models specifically tailored to proportion data and potential clustering (such as logistic regression, generalised linear mixed models or spatial analyses) would be more appropriate for future confirmatory studies. Future studies which use longitudinal methods and focus specifically on high school athletes may help address the gaps identified in this study and provide a clearer understanding of how mental health screening practices can best support this population.

Clinical implications

The present findings invite critical dialogue around how mental health screening practices can most effectively support the well-being of high school athletes. Recommendations regarding the inclusion of validated tools such as the PHQ-4 and ED risk questions in PPE should be regarded as tentative and subject to confirmation through future research. In particular, studies that incorporate athlete-specific behaviour and symptom tracking, longitudinal designs and follow-up of intervention outcomes will be essential to further determine the extent to which standardised screening practices translate into meaningful improvements in mental health and safety for high school athletes. Relatedly, there is a need to consider how screening results are communicated, addressed and integrated into athlete-centred care. Ensuring that follow-up processes are sensitive to the unique psychosocial challenges faced by student-athletes—including stigma, performance pressure and access to care – —may enhance the impact of these screenings. Educational and athletic institutions should also consider how to tailor screening protocols to reflect the developmental stage, sport-specific demands and cultural context of this population. Moving forward, thoughtful engagement between clinicians, educators, athletic staff and families will be essential to optimising the utility of screening tools and ensuring they lead to meaningful, supportive interventions.

Data availability statement

Data are available in a public, open access repository.