Increased attention continues to be placed on best practices for assisting student-athletes who show symptoms of a mental health illness. One of the most common mental health concerns for student-athletes is feeling overwhelmed or anxious. Early recognition of these symptoms can help student-athletes find early intervention.
To elucidate specific attributes associated with an increased likelihood of anxiety symptoms and treatment avoidance to help athletic trainers and other practitioners better identify and assist high-risk individuals.
Cross-sectional study.
Colleges and universities in the United States.
A total of 51 882 varsity athletes who completed the National College Health Assessment between 2011 and 2019.
Survey responses (self-report) to questions related to anxiety symptoms and disorders cross-sectioned into 9 predictors: year in school, gender, sexual orientation, race and ethnicity, international student status, school type, received information on stress reduction or depression or anxiety, anxiety about impeding academic performance, and history of mental health treatment in the last year.
More than three-fourths (82.9%, n = 43 016) of student-athletes reported feeling overwhelmed because of all that they had to do in their daily activities, whereas only half (47.5%, n = 24 645) of student-athletes described overwhelming anxiety as a symptom. Gender, sexual orientation, race and ethnicity, school type, and previous mental health treatment had large effects on these 2 symptoms. However, of the student-athletes who reported overwhelming anxiety as a symptom, only one-fourth (22.9%, n = 5649) indicated they received a diagnosis of anxiety with or without treatment.
Factors such as the year in school and availability of mental health information appeared to have little effect on the experience of anxiety-related symptoms and the propensity to seek treatment. However, other factors, including gender, sexual orientation, race and ethnicity, school type, and previous mental health treatment, did have important effects. Our study provides additional evidence that the stresses placed on collegiate student-athletes frequently led to feelings of being overwhelmed and, eventually, anxiety disorder diagnoses. However, most student-athletes experiencing anxiety-related symptoms still did not seek a diagnosis with or without treatment.
Gender, sexual orientation, race and ethnicity, school type, and previous mental health treatment had important effects on reporting anxiety symptoms.
Α history of mental health treatment had the largest effect on anxiety-related outcomes and thus can serve as an important predictor of the risk for anxiety and the need for treatment.
Among athletes reporting overwhelming anxiety in the last year (n = 24 645), 77.1% stated they received neither a diagnosis nor treatment for any anxiety disorders versus 22.9% who received a diagnosis with or without treatment.
Mental health consists of maintaining adequate, functional levels of emotional, psychological, and social well-being, posing a challenge for millions of people worldwide.1 Researchers2 have estimated that approximately 20% of US adults experience mental health disruptions, with approximately 5% experiencing serious mental illness. Individuals with mental health disruptions are plagued by feelings of hopelessness, face crucial life decisions or events as a result (eg, dropping out of school, losing a job, ending relationships), and may consider or even attempt suicide. Compounding the problem, a large proportion of people with mental health impairments seek no form of treatment (>50% in 1 study3 ) and are thus unlikely to resolve their conditions.
Given multiple unique stressors, college students are at particularly high risk for mental health problems. Mental health diagnoses are rising among college students, with an almost 50% increase over the past 10 to 15 years, doubling the need for treatment.4 The stressors unique to college students are perhaps magnified in student-athletes, who, in addition to typical academic pressures, often must meet high academic standards to maintain scholarships; encounter complex social dynamics with teammates, coaches, athletic trainers (ATs), and other athletic staff; and perceive pressure to perform athletically at a high level.
With the range and intensity of these stressors, it is not surprising that one of the leading mental health diagnoses among college students and student-athletes is generalized anxiety disorder,2 the characteristics of which include overwhelming, persistent, and excessive worry about different aspects of life.5 In a recent survey6 of college counseling center directors, 42% stated that anxiety was the leading mental health concern in their student populations. Among student-athletes, roughly half reported feeling overwhelming anxiety in a given year, and >20% noted that anxiety was an impediment to academic achievement.6 At the same time, only 12.8% of athletes described receiving anxiety disorder diagnoses, meaning that a substantial proportion (87.2%) of athletes experiencing anxiety symptoms did not receive professional attention.6 Even more alarming, among those who did receive a diagnosis, almost one-third did not seek treatment.6 Treatment avoidance among student-athletes is perhaps not surprising given the additional stigma associated with perceived “mental weakness” in athlete culture.7–9 The rates of anxiety symptoms and diagnoses among student-athletes have also increased over the past 10 years, but, encouragingly, so have the rates of treatment-seeking behavior and openness to future treatment.10 In addition to mental health, anxiety can also affect the physical safety of athletes during practice and competition. Negative stressors can reduce intentional focus, leading to miscues,11 and anxiety during competition can generate specific physiological responses (increased muscle tension, narrowed visual field, increased distractibility) that increase the injury risk.11 Athletes with preseason anxiety symptoms have been shown to sustain more injuries than athletes without symptoms.12
The presence of an AT who is familiar with the student-athletes and their mental well-being can help reduce the risk of anxiety-related injuries as well as help students obtain needed treatment.12 In 2015, the National Athletic Trainers’ Association (NATA) adopted a consensus statement developed by ATs with mental health expertise.13 The statement provides guidelines for developing appropriate recognition and referral plans for athletes with mental health concerns.13 The National Collegiate Athletics Association (NCAA) adopted and implemented the NATA guidelines in its 2016 “Mental Health Best Practices” document,14 which provides guidelines for campus stakeholders. The NCAA specifically recognized ATs as campus “point people” for providing student-athlete health care, including identifying and referring students with signs of mental health distress.14
To increase the proportion of student-athletes seeking appropriate diagnosis and treatment for anxiety symptoms, ATs need to not only detect outward signs of anxiety but also identify athletes who may be at elevated risk. Their ability to do so is complicated by the risk for symptoms and the varying tendencies of individuals to seek diagnoses and treatment. For example, in addition to aspects of general disposition (eg, optimism versus pessimism), gender has been observed to affect both acute and generalized anxiety in response to various stimuli, with females exhibiting greater risk than males.15,16 Other researchers have shown that treatment-seeking barriers may differ by race and ethnicity.17 Factors including age, year in school, sexual orientation, and institutional traits can affect stigma and its relationships to seeking diagnoses and treatment.18–24 Thus, not every athlete has the same level of risk for impaired mental health. A clearer understanding of the factors that affect the risk or influence the reluctance to seek treatment can help ATs and other practitioners identify high-risk athletes early and aid them in obtaining needed treatment.
To further investigate factors affecting the prevalence of anxiety symptoms, diagnoses, and treatment-seeking behavior in the student-athlete population, we analyzed self-reported data from the National College Health Assessment (NCHA), a large, national health survey of US college students. The NCHA asks several mental-health–related questions and obtains extensive data on student demographics and institutional attributes, providing a powerful tool for assessing nationwide trends in factors affecting the risk for anxiety and related treatment-seeking behavior. Year in school, gender, sexual orientation, race and ethnicity, international student status, school type, availability of mental health information, academic impediments, and a history of mental health treatment were analyzed for their effects on anxiety symptoms, diagnoses, and treatment in student-athletes. The purpose of our analysis was to elucidate specific attributes associated with an increased likelihood of anxiety symptoms and treatment avoidance to help ATs and other practitioners better identify and assist high-risk individuals.
METHODS
Data came from NCHA surveys IIb (2011–2015) and IIc (2015–2019) of the American College Health Association (ACHA; data file distributed February 10, 2020, by ACHA). The NCHA is a validated national research survey providing self-reported information on US college and university student health status, habits, knowledge, behaviors, and perceptions.25 The survey helps college administrators, health service providers, health educators, counselors, and researchers understand and manage student health. We accessed and analyzed NCHA data under a data-sharing agreement with the ACHA. All available data from participating institutions during the aforementioned survey periods were included initially. Data from special-focus/miscellaneous/unclassified, 2-year, or non-US schools were then excluded. Only data on full-time, degree-seeking students self-identifying as varsity athletes aged <30 years were retained for analysis. Our university’s institutional review board determined that the study procedures were exempt.
Outcomes were anxiety-related symptoms and diagnosis with or without treatment (hereafter diagnosis/treatment) of anxiety disorders. Symptom questions included “Have you ever:” “…felt overwhelmed by all you had to do?” and “…felt overwhelming anxiety?” Responses were recoded to binary levels: yes, in the last year and no, not in the last year. Survey questions about the diagnosis and treatment of anxiety asked for information from the preceding year. Response options were no (neither diagnosed nor treated), yes, diagnosed but not treated, and 4 treatment-type options (medication, psychotherapy, medication and psychotherapy, and other). Responses were recoded into 3 levels: not diagnosed/treated, diagnosed but not treated, and treated. Rather than analyze data for individual specific diagnoses (anxiety, obsessive-compulsive disorder, panic attacks, or phobia), we collapsed all into the single category anxiety disorders.
Predictors were year in school (first-, second-, third-, fourth-, or fifth-year undergraduate or graduate); self-identified gender (cis male [hereafter male], cis female [hereafter female], or nonbinary), sexual orientation (heterosexual, gay/lesbian, bisexual, or other), and race and ethnicity (American Indian, Alaska Native, or Native Hawaiian; Asian or Pacific Islander; biracial, multiracial, or other; Black non-Hispanic; Hispanic of any race; White non-Hispanic)26 ; international student status (no or yes); school type (baccalaureate [liberal arts] colleges, private master’s universities, public master’s universities, private doctoral universities, or public doctoral universities); received information on stress reduction or depression or anxiety (no or yes); anxiety impeding academic performance in the last year (no or yes); and history of mental health treatment (no or yes).
Gender and racial and ethnic categories were structured using previously described methods.10 Sexual orientation response options differed between NCHA IIb (heterosexual, gay/lesbian, bisexual, or unsure) and IIc (asexual, bisexual, gay, lesbian, pansexual, queer, questioning, same gender loving, straight/heterosexual, or another identity). To merge the data, we created 4 categories with the following response options from each survey: heterosexual (IIb: heterosexual; IIc: straight/heterosexual), gay/lesbian (IIb: gay/lesbian; IIc: gay, lesbian, or same gender loving), bisexual (IIb: bisexual; IIc: bisexual), and other (IIb: unsure; IIc: asexual, pansexual, queer, questioning, or another identity). School types were derived using latent class analysis27,28 and 4 metadata variables related to school attributes: public or private, religiously or nonreligiously affiliated, enrollment (small, medium, or large),22 and Basic Carnegie Classification.29 Models with 3 to 6 classes were compared, and the model with the lowest Akaike information criterion30 was retained. This was the 5-level model closely resembling the Basic Carnegie Classification but distinguishing between public and private master’s and doctoral degree–granting institutions. Anxiety as an impediment to academic performance in the last year was dichotomized into no, not an impediment or yes, an impediment, with the latter containing responses that anxiety led to lower coursework grades; poor grades, incomplete, or dropped courses; or disruptions to theses, dissertations, research, or practicums. Responses as to past mental health treatment beyond the previous year were condensed to no/never or yes/ever.
Statistical Analysis
Given the large sample size and to correct for multiple comparisons, we set the α level at .001. Statistical analysis was performed using SAS (version 9.4; SAS Institute). Effects of predictors on outcome frequencies were analyzed using multivariable logistic regression analysis. Univariate association filtering30 was applied to first determine which predictors to include in the multivariable analyses. The univariate criteria for inclusion were P < .001 and an odds ratio indicating at least a small effect size, adjusted for exposure-event frequency.31 Multivariable models for each outcome were then run that included all predictors meeting these criteria. Post hoc relative risk (RR) was calculated for all possible level pairs within each included predictor variable, adjusted for other model variable effects.32 For ease of interpretation, RRs were structured so the numerator was always the level with the greater event frequency (ie, groups or levels were not designated a priori as exposure or control); thus, all RRs are expressed as ≥1.00. We calculated 99.9% CIs, chosen over the standard 95% CI to match the α level. Post hoc between-groups differences were designated as clinically meaningful when RR > 1.05 (ie, >5% higher risk) and the CI excluded 1.00.
Binary symptom outcomes were analyzed such that the event for RR calculations was yes, in the last year. Within the symptomatic sample, the diagnosis/treatment outcomes analysis was stratified. First, rates of no diagnosis or treatment versus any diagnosis or treatment were analyzed (event: any diagnosis/treatment). Second, the data were limited to athletes reporting diagnosis or treatment, and the rates of untreated diagnosis versus treatment were analyzed (event: untreated diagnosis).
Finally, we calculated a sensitivity analysis to compare the model results when a history of mental health treatment was excluded. We included past treatment as a predictor because it has implications for future mental health status and treatment-seeking behavior and can be used as a screening tool to improve care for incoming athletes (first-year undergraduates and transfers). However, the high correlation between past treatment and current diagnosis or treatment (and some overlap as categories) could obscure other smaller variable effects. Given that no meaningful differences between models were found in the sensitivity analysis, treatment history was retained in multivariable models when it met the univariate inclusion criteria.
RESULTS
The final sample consisted of 51 882 athletes. Sample distributions for independent variables are presented in Table 1. Most athletes were female, heterosexual, White, and US citizens. Year in school was skewed toward younger students (34.4% first-year undergraduates), and distribution by school type was 27.4% baccalaureate (liberal arts) colleges, 26.7% public doctoral universities, 21.6% private master’s universities, 13.9% private doctoral universities, and 10.4% public master’s universities. Two-thirds of students had received information on stress reduction (67.5%) and depression or anxiety (64.9%), and 34.8% reported previous mental health treatment. The univariate statistics used to determine which variables were included in each multivariable analysis are provided in Table 2.
Symptom: Overwhelmed by All You Had to Do
A large majority of athletes (82.9%, n = 43 016) reported being “overwhelmed by all you had to do.” The full multivariable model demonstrated a difference (χ2 = 3675.9, P < .001), with predictors of year in school, gender, sexual orientation, race and ethnicity, international student status, school type, and a history of mental health treatment (χ2 ≥ 25.8, P < .001). Within-group outcome frequencies by independent variable are presented in Figure 1.
The proportion of each subgroup responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelmed by all you had to do?” Only variables meeting the univariate criteria for inclusion in the multivariable models are presented.a–q Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful. A, Year in school. B, Gender. C, Sexual orientation. D, Race and ethnicity. E, International student. F, School type. G, History of mental health treatment.
The proportion of each subgroup responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelmed by all you had to do?” Only variables meeting the univariate criteria for inclusion in the multivariable models are presented.a–q Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful. A, Year in school. B, Gender. C, Sexual orientation. D, Race and ethnicity. E, International student. F, School type. G, History of mental health treatment.
Clinically meaningful pairwise comparisons are indicated below as the group with the higher outcome frequency versus the group with the lower outcome frequency (RR; 99.9% CI). When one group was meaningfully different from multiple other groups, results are shown as the group with the highest outcome frequency versus each group with a lower outcome frequency (RR; 99.9% CI) in descending order.
For year in school, second- (RR = 1.06; 99.9% CI = 1.01, 1.11) and third- (RR = 1.06; 99.9% CI = 1.01, 1.11) year undergraduates had higher outcome frequencies than graduate students. Outcome frequencies by gender were higher for female versus male (RR = 1.22; 99.9% CI = 1.20, 1.24), nonbinary versus male (RR = 1.10; 99.9% CI = 1.02, 1.18), and female versus nonbinary (RR = 1.10; 99.9% CI = 1.05, 1.15) student-athletes. By sexual orientation, outcome frequencies were higher for gay/lesbian versus other (RR = 1.08; 99.9% CI = 1.03, 1.13) and bisexual versus other (RR = 1.07; 99.9% CI = 1.03, 1.12) student-athletes. Outcome frequencies for race and ethnicity were higher for White versus Black (RR = 1.12; 99.9% CI = 1.08, 1.16), versus Hispanic (RR = 1.06; 99.9% CI = 1.02, 1.10), and versus Asian or Pacific Islander (RR = 1.09; 99.9% CI = 1.05, 1.13) student-athletes; for biracial, multiracial, or other versus Black (RR = 1.10; 99.9% CI = 1.06, 1.15) and versus Asian or Pacific Islander (RR = 1.08; 99.9% CI = 1.03, 1.12) student-athletes; and Hispanic versus Black student-athletes (RR = 1.06; 99.9% CI = 1.01, 1.11). We found higher outcome frequencies for US versus international students (RR = 1.09; 99.9% CI = 1.07, 1.11). By school type, outcome frequencies were higher for baccalaureate (liberal arts) colleges versus public master’s universities (RR = 1.07; 99.9% CI = 1.04, 1.10) and private doctoral universities versus public master’s universities (RR = 1.07; 99.9% CI = 1.04, 1.10). Higher outcome frequencies were found for students with a treatment history versus no treatment history (RR = 1.11; 99.9% CI = 1.10, 1.12).
Symptom: Overwhelming Anxiety
Approximately half of athletes (47.5%) reported having “overwhelming anxiety.” The full multivariable model demonstrated a difference (χ2 = 5151.2, P < .001), with predictors of gender, sexual orientation, race and ethnicity, international student status, and a history of mental health treatment (χ2 ≥ 46.2, P < .001). Within-group outcome frequencies by independent variable are presented in Figure 2.
Bars represent the proportion of each subgroup responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelming anxiety?” Only variables meeting the univariate criteria for inclusion in the multivariable models are presented.a–k Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful. A, Gender. B, Sexual orientation. C, Race and ethnicity. D, International student. E, History of mental health treatment.
Bars represent the proportion of each subgroup responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelming anxiety?” Only variables meeting the univariate criteria for inclusion in the multivariable models are presented.a–k Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful. A, Gender. B, Sexual orientation. C, Race and ethnicity. D, International student. E, History of mental health treatment.
By gender, outcome frequencies were higher for female versus male (RR = 1.42; 99.9% CI = 1.37, 1.47) and nonbinary versus male (RR = 1.43; 99.9% CI = 1.27, 1.60) student-athletes. By sexual orientation, we found higher outcome frequencies for gay/lesbian versus heterosexual (RR = 1.16; 99.9% CI = 1.07, 1.25), bisexual versus heterosexual (RR = 1.23; 99.9% CI = 1.16, 1.29), other versus heterosexual (RR = 1.09; 99.9% CI = 1.02, 1.16), and bisexual versus other (RR = 1.12; 99.9% CI = 1.03, 1.21) student-athletes. Outcome frequencies by race and ethnicity were higher for White versus Black (RR = 1.12; 99.9% CI = 1.04, 1.21), Asian or Pacific Islander versus Black (RR = 1.18; 99.9% CI = 1.07, 1.30), and biracial, multiracial, or other versus Black (RR = 1.17; 99.9% CI = 1.11, 1.27) student-athletes. By international student status, outcome frequencies were higher for US versus international students (RR = 1.18; 99.9% CI = 1.10, 1.25). By history of mental health treatment, outcome frequencies were higher for students with a treatment history versus no treatment history (RR = 1.68; 99.9% CI = 1.65, 1.71).
Diagnosis and Treatment
Among athletes reporting overwhelming anxiety in the last year (n = 24 645), 77.1% (n = 18 996) reported receiving neither a diagnosis nor treatment versus 22.9% (n = 5649) who reported receiving a diagnosis or treatment. Within the latter group, 72.3% (n = 4083) received a diagnosis and treatment, whereas 27.7% (n = 1566) had an untreated diagnosis. The full multivariable model displayed a difference (χ2 = 3894.7, P < .001), with predictors of year in school, gender, sexual orientation, race and ethnicity, anxiety impeding academic performance, and a history of mental health treatment (χ2 ≥ 46.7, P < .001). Within-group outcome frequencies by independent variable are provided in Figure 3.
Results of the analysis of diagnosis and treatment outcomes within athletes responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelming anxiety?” Frequencies of athletes reporting that they received neither a diagnosis nor treatment (gray bar segments), had an untreated diagnosis (white bar segments), or were diagnosed and treated for an anxiety disorder (black bar segments).a–k Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses comparing frequencies of athletes reporting no diagnosis or treatment versus those reporting diagnosis and/or treatment. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful.l When limited to only those athletes reporting either a diagnosis or treatment, only a history of mental health treatment had a significant and clinically meaningful effect on the RR of untreated diagnosis versus diagnosis and treatment. A, Year in school. B, Gender. C, Sexual orientation. D, Race and ethnicity. E, Academic impediment. F, History of mental health treatment.
Results of the analysis of diagnosis and treatment outcomes within athletes responding “Yes, in the last year” to the survey question: “Have you ever:” “…felt overwhelming anxiety?” Frequencies of athletes reporting that they received neither a diagnosis nor treatment (gray bar segments), had an untreated diagnosis (white bar segments), or were diagnosed and treated for an anxiety disorder (black bar segments).a–k Letters at the ends of the bars indicate the results of pairwise relative risk (RR) analyses comparing frequencies of athletes reporting no diagnosis or treatment versus those reporting diagnosis and/or treatment. Matching letters at the ends of 2 bars indicate that their pairwise RR was statistically significant and met the criteria to be considered clinically meaningful.l When limited to only those athletes reporting either a diagnosis or treatment, only a history of mental health treatment had a significant and clinically meaningful effect on the RR of untreated diagnosis versus diagnosis and treatment. A, Year in school. B, Gender. C, Sexual orientation. D, Race and ethnicity. E, Academic impediment. F, History of mental health treatment.
By year in school, no adjusted RRs were clinically meaningful. Outcome frequencies by gender were higher for nonbinary versus male (RR = 1.11; 99.9% CI = 1.04, 1.18) and versus female (RR = 1.07; 99.9% CI = 1.01, 1.13) student-athletes. By sexual orientation, outcome frequencies were higher for bisexual versus heterosexual student-athletes (RR = 1.06; 99.9% CI = 1.02, 1.09). We observed that, by race and ethnicity, outcome frequencies were higher for American Indian, Alaska Native, or Native Hawaiian versus White (RR = 1.09; 99.9% CI = 1.01, 1.18), versus Black (RR = 1.15; 99.9% CI = 1.04, 1.26), versus Hispanic (RR = 1.11; 99.9% CI = 1.01, 1.22), versus Asian or Pacific Islander (RR = 1.15; 99.9% CI = 1.05, 1.26), and versus biracial, multiracial, or other (RR = 1.12; 99.9% CI = 1.02, 1.23) student-athletes. They were also higher for White versus Asian or Pacific Islander student-athletes (RR = 1.06; 99.9% CI = 1.03, 1.08). By academic impediment, we noted higher outcome frequencies for reporting anxiety as an academic impediment versus not reporting anxiety as an academic impediment (RR = 1.15; 99.9% CI = 1.13, 1.17). Outcome frequencies by a history of mental health treatment were higher for students with a treatment history versus no treatment history (RR = 1.30; 99.9% CI = 1.28, 1.31).
Symptoms, Diagnosis, and Treatment
The final analysis was limited to athletes who stated they experienced overwhelming anxiety in the last year and who either had an untreated diagnosis or were diagnosed and treated. With the exception of a history of mental health treatment, none of the adjusted RRs for any pairwise comparison met the criteria for being clinically meaningful. For past treatment, athletes with no history of treatment were more likely than those with a history of treatment to report a current untreated diagnosis (RR = 1.15; 99.9% CI = 1.12, 1.18; Figure 3).
DISCUSSION
We analyzed survey data on factors influencing anxiety symptoms, diagnosis, and treatment in collegiate student-athletes. Anxiety symptoms were common among student-athletes, but relatively few sought a corresponding diagnosis or treatment. More specifically, factors including year in school, gender, sexual orientation, race and ethnicity, international student status, school type, availability of mental health information, academic impediment, and history of mental health treatment affected anxiety-related outcomes. Whereas some factors, such as year in school and availability of mental health information, had minimal effects, others, including gender, sexual orientation, race and ethnicity, school type, and history of mental health treatment, had important effects on the experience of anxiety-related symptoms and the propensity to seek treatment.
Regarding gender, nonbinary and female athletes more frequently described anxiety-related symptoms than male athletes. Correspondingly, nonbinary and female athletes more often received a diagnosis or treatment for anxiety disorders than male athletes, with nonbinary athletes receiving a diagnosis or treatment more often than female athletes. Among athletes indicating overwhelming anxiety symptoms, nonbinary athletes were more likely to have received a diagnosis or treatment versus male and female athletes. However, among those acknowledging a diagnosis or treatment, gender had no effect on the probability of an untreated versus treated diagnosis. In other words, although symptomatic athletes of different genders may seek diagnosis or treatment at different rates, no gender difference existed in the probability of pursuing treatment after a diagnosis had been made. Among symptomatic athletes, about 25% (n = 1566) of all genders who received a diagnosis reported it was untreated. These results are consistent with those of previous researchers15,16,33 examining self-reported depression and anxiety rates among student-athletes and nonathletes who demonstrated that female athletes indicated higher levels of depression and anxiety compared with male athletes. Similarly, earlier investigators34 showed higher rates of anxiety in transgender and nonbinary adolescents and young adults, with additional variation related to specific self-identified genders relative to sex assigned at birth. Carrying over of this finding to collegiate athletes is not surprising. Given that we collapsed gender into relatively broad categories, future authors should focus on finer-grained analyses of gender expression in student-athletes and its effects on mental health.
Compared with heterosexual athletes, athletes self-identifying as gay/lesbian, bisexual, or other more frequently described overwhelming anxiety as well as receiving a diagnosis or treatment for anxiety disorders. However, among athletes reporting a diagnosis or treatment, sexual orientation had no effect on the likelihood of undergoing treatment versus having an untreated diagnosis. Among symptomatic athletes, about 25% to 30% of athletes of all sexual orientations who stated they obtained a diagnosis or treatment acknowledged an untreated diagnosis. Improving access to treatment may require increased sensitivity to the anxiety-related mental health needs of athletes identifying as gay/lesbian, bisexual, and other sexual orientations. Findings in this area are encouraging, with researchers35 observing that most ATs participating in a survey and especially female ATs had positive attitudes toward lesbian, gay, bisexual, transgender, and queer student-athletes. This positivity and open-mindedness toward lesbian, gay, bisexual, transgender, and queer athletes is a strong foundation on which to build additional awareness regarding mental health. However, increasing this open-mindedness and building empathy among ATs toward anxiety-related risk factors in these athletes is important to improve screening, prevention, and treatment.
Similar to sexual orientation, social factors related to race and ethnicity can elevate stress levels among individuals from groups identified as racial and ethnic minorities. College campuses, where racial and ethnic minority students and student-athletes are frequently underrepresented, have the potential to further exacerbate societal influences on anxiety. Still, we determined that White athletes were most likely to feel overwhelmed, followed by biracial, multiracial, or other athletes, whereas Black athletes were least likely to feel overwhelmed. White, Asian or Pacific Islander, and biracial, multiracial, or other athletes were also all more likely to describe overwhelming anxiety compared with Black athletes. At the same time, White athletes who noted anxiety-related symptoms also commented that they received a diagnosis or treatment for anxiety disorders more frequently than athletes identifying as Black, Asian or Pacific Islander, and biracial, multiracial, or other. Despite not reporting symptoms at meaningfully higher rates than other groups, American Indian, Alaska Native, or Native Hawaiian athletes received a diagnosis or treatment more frequently than athletes of any other race and ethnicity, with effect sizes that were considerably larger than for other pairwise racial and ethnic comparisons. The results of our study are consistent with those in the literature. Tran36 found that White student-athletes had a higher rate of anxiety compared with student-athletes who were Black, Hispanic/Latinx, Asian/Pacific Islander, Native American, and multiracial, whereas, conversely, student-athletes who were Black, Hispanic/Latinx, Asian/Pacific Islander, Native American, and multiracial had higher rates of depression and suicide compared with White student-athletes. Among those athletes reporting a diagnosis or treatment, however, race and ethnicity had no effect on the likelihood of undergoing treatment versus having an untreated diagnosis. Among symptomatic athletes, about 25% to 40% of all races and ethnicities who obtained a diagnosis stated that it was untreated. In general, then, the results of Tran36 suggest that racial and ethnic minority student-athletes do not experience, or at least do not report experiencing, anxiety-related outcomes more frequently than their White counterparts.
Often, the school environment can affect the stressors that cause anxiety among collegiate student-athletes. For example, some institutions will create “soft landing” courses to allow athletes an opportunity to maintain the required grade point average to participate in sports. In our study, students attending more expensive, and generally more academically rigorous, schools (baccalaureate [liberal arts] colleges and private doctoral universities) were more likely to feel overwhelmed than those attending more affordable, regional public schools (eg, public master’s universities). One interpretation is that the pressure on athletes to maintain high grades to keep an athletic scholarship at an expensive and academically challenging school may elevate anxiety relative to athletes at more affordable and less selective schools. Nonetheless, the interpretation may not be this simple. In recent work, Stokowski et al37 observed low levels of anxiety and depression among athletes at NCAA Division III schools, which tend to be selective and expensive baccalaureate (liberal arts) colleges, many of which are also private. Most survey respondents did not identify closely with the athlete role, despite participating in varsity athletics, and therefore were likely able to focus more on academics in preparation for postcollegiate life as a nonathlete. Moreover, Division III schools do not offer athletic scholarships, perhaps removing the pressure and anxiety tied specifically to maintaining this route to educational access. More exploration is needed to understand how different institutional qualities affect student-athlete anxiety so that preventive and treatment strategies can be appropriately tailored.
Unsurprisingly, a history of mental health treatment had the largest effect on anxiety-related outcomes in effect sizes (eg, almost twice as likely to report feeling overwhelming anxiety and the largest RR [1.30] for recent diagnosis or treatment) and thus can serve as an important predictor of risk for anxiety and need for treatment. One stark finding was that 77.1% of all athletes experiencing overwhelming anxiety did not receive a subsequent diagnosis or treatment, meaning that substantial barriers to student-athlete mental health treatment-seeking behavior (eg, stigma, lack of available resources or insufficient information about available resources, insurance and financial concerns) still exist. Yet the main factor that appeared to best predict who would seek treatment in conjunction with symptoms was a history of mental health treatment. Athletes with any history of mental health treatment were more likely to describe symptoms, especially feeling overwhelming anxiety. Among symptomatic athletes also indicating a recent diagnosis or treatment, those who had past mental health treatment were recently treated at a rate of 77%, with just 23% acknowledging a recent untreated diagnosis. In contrast, only 39% of symptomatic athletes without a history of treatment had been treated recently, whereas 61% reported a recent untreated diagnosis.
It therefore seems clear that knowledge of a history of mental health treatment can serve as an important screening tool to identify student-athletes who are at higher risk for anxiety disorders and who may also be more open to receiving treatment with appropriate guidance. This prescreening process could be easily incorporated into annual preparticipation physical examinations, but few researchers have investigated trends of colleges and universities implementing this type of practice. Our study combined with the NATA and NCAA consensus statements13,14 can provide the necessary guidance for ATs to help recognize and develop referral policies and procedures for student-athletes with anxiety. Our work also presents opportunities for future evaluations of whether such screening tools increase the rates of treatment-seeking behavior among symptomatic student-athletes.
This study had limitations, including the self-reported nature of the data set, which introduced the potential for error via dishonest or inaccurate responses. However, the anonymous survey and large sample size likely minimized this effect. Honest anonymous answers likely “wash out” an anticipated small number of dishonest ones, resulting in broadly unbiased results. Bias may have also existed if students who chose not to respond to the survey were more likely to be athletes or those with mental health concerns such that these groups were underrepresented in the sample. We have no way to determine if this occurred, and the results may have underestimated the athlete mental health burden. Finally, a small number of students may have taken the survey in multiple semesters, appearing twice in the data set. Although we implemented no procedures to prevent this, based on an analysis of institutional identifications and years of survey administration, it is unlikely that this accounts for many observations. We also had no information about when in a given semester (ie, beginning, middle, finals week) each institution administered the surveys. To the extent that the data set included uncontrolled repeated observations, the effect would be artificial narrowing of the CIs. The reader is cautioned to interpret the results while considering the possible effects of these limitations. Although we focused on anxiety among student-athletes, future researchers should address other specific mental health diagnoses (ie, mood disorders, eating disorders, and substance abuse). Anecdotally, these are common mental health concerns seen among the student-athlete population, and ATs could benefit from a better understanding of factors influencing the risk as well as the propensity for and barriers to treatment-seeking behavior. Such findings would broaden the capacity of ATs to identify athletes at elevated risk for a set of mental health conditions and guide them toward treatment as needed.
Our study provided additional evidence that the stresses placed on collegiate student-athletes frequently led to feelings of being overwhelmed and, eventually, a diagnosis of anxiety disorder. Nonetheless, most student-athletes experiencing anxiety-related symptoms did not seek a diagnosis or treatment. Demographic and institutional factors affected the prevalence of anxiety-related symptoms and diagnoses but generally not of treatment-seeking behavior after a diagnosis. The exception was a history of mental health treatment, which not only predicted symptom reporting but also was a key factor in treatment-seeking behavior corresponding to the diagnosis. This information can be used by schools, athletic departments, and ATs to better identify student-athletes who are at risk for anxiety disorders and may be more likely to seek treatment versus those who need additional encouragement to do so.
ACKNOWLEDGMENTS
The opinions, findings, and conclusions reported in this article are those of the authors and are in no way meant to represent the corporate opinions, views, or policies of the ACHA. The ACHA does not warrant nor assume any liability or responsibility for the accuracy, completeness, or usefulness or any information presented in this article.