Introduction

Young adulthood (ages 18–25) is a critical period for identifying and treating mental health problems with the potential to change trajectories of illness and health across the life span (Wood et al., 2018). In 2020, 30.6% of young adults in the United States experienced any mental illness (10.2 million) and 9.7% experienced a serious mental illness (3.3 million; Substance Abuse and Mental Health Services Administration 2021). These high numbers continue a trend of increasing rates of mental illness in young adults over the past decade with current rates more than double the number reported in 2008 (Substance Abuse and Mental Health Services Administration 2021). These numbers have been particularly high since the COVID pandemic as isolation and disruption of daily routines took a significant toll on young adults (Czeisler et al., 2020). This is particularly troubling since serious mental illnesses often have peak periods of onset in adolescence or early adulthood (Kessler et al., 2007a, 2007b), highlighting the urgency of identification and treatment in young adulthood. Onset of mental health symptoms in adolescence or young adulthood is associated with greater risk for significant mental health problems and suicide throughout the lifespan (Kessler et al., 2007a, 2007b; Mitsui et al., 2018), making intervention and treatment during young adulthood critical. Research also shows that the delay of treatment for mental health challenges can lead to substandard treatment outcomes and more complex symptomatology (Kessler & Price, 1993; Macdonald et al., 2021). Yet, young adults consistently have the highest rates of reported unmet need for services and the lowest rates of mental health service use compared to older aged adults (Substance Abuse and Mental Health Services Administration 2021).

Young adults experiencing homelessness are a group that are particularly vulnerable to experiencing mental health problems (Edidin et al., 2012; Hodgson et al., 2013). One in ten young adults in the United States, approximately 4.2 million, experience housing instability or homelessness each year (Morton et al., 2018). And, rates of mental disorders among young adults experiencing homelessness (YAEH) are exceptionally high compared to same-aged peers without experiences of homelessness. In a systematic review, Edidin et al. (2012) found that prevalence of mental disorders in youth experiencing homelessness ranged from 48 to 98%, with many youth meeting criteria for multiple disorders. Similar rates have been found across the world such as in Australia, where youth/young adults who experience homelessness have a lifetime prevalence of 82–85% for psychiatric disorders (Black et al., 2018). These high rates are likely the result of the relationship between mental health and homelessness in which mental health symptoms contribute to homelessness, experiences while homeless contribute to the onset of new and/or exacerbation of existing mental health problems, and common root problems including family violence, abuse, inadequate social safety nets, and poverty contribute to both mental health and homelessness (Narendorf, 2017; Thompson et al., 2010). Additional factors such as high rates of co-morbid substance use and high rates of exposure to public systems such as foster care and juvenile justice highlight the need to connect YAEH to mental health supports that are sensitive to the intersectional factors that contribute to mental health problems (Narendorf et al., 2020; Whitbeck et al., 2004). The window of time when young adults are experiencing homelessness may present a critical time to intervene and connect them with mental health supports. This is particularly critical in the current context of the COVID-19 pandemic which has been documented to be particularly challenging for those experiencing homelessness due to the abrupt closure of services they relied on and the formidable challenges of navigating a pandemic without access to safe and stable housing (Rew et al., 2021; Tucker et al., 2020).

While rates of mental health problems are high among young adults, particularly those experiencing homelessness, only one third of young adults with evidence of mental health problems utilize mental health treatments (Substance Abuse and Mental Health Services Administration 2021; Solorio et al., 2006). These low rates of service use highlight the need to identify points of intervention to improve connection to mental health supports or identify alternative venues for providing supports. Formal mental health supports, including use of therapy and psychotropic medications, are the front-line treatment for mental disorders (National Alliance On Mental Illness 2022) yet utilization of these services among YAEH is low (Solorio et al., 2006). And, the developmental stage of young adulthood has been documented as a point where service-seeking for mental health problems is at its lowest point in the life course (Pottick et al., 2008). Unfortunately, YAEH experience unique mental health service access barriers that further compound the challenges of getting appropriate mental health care. These include the inability to consistently access services in one location, transportation barriers, long waitlists, complex eligibility processes, stigma around mental illness and/or service use, and providers who lack training in culturally responsive services for YAEH (Black et al., 2018; Brown et al., 2016; Gallardo et al., 2020; Macdonald et al., 2021). Understanding mental health service use among this group requires accounting for a range of barriers that may prevent successful service use.

One foundational aspect to any type of help seeking, however, is first recognizing symptoms as a mental health problem (Cauce et al., 2002). The recognition that psychological symptoms of distress may represent a mental health problem, a phenomenon we call self-identification, may be a critical point of intervention to accompany interventions that address the structural access barriers present in the environment. Prior work in young adults has found that the perception of need for mental health services, a distinct but related concept to self-identification, is a barrier to accessing treatment with unique predictors (Narendorf & Palmer, 2016). Cauce et al, (2002), described help seeking pathways in adolescence as beginning with problem identification, then proceeding to a decision to seek help, which then culminated in selection of formal mental health services or other alternatives such as seeking informal supports from family and friends. And, they noted that for ethnic minority youth, these pathways are often shaped by cultural and contextual influences (Cauce et al., 2002). Differences by race and gender in perceptions of need for service use have been found in prior studies using national data such as the National Survey of Drug Use and Health where Narendorf and Palmer (2016) found that non-white racial-ethnic populations and those that identified as men had lower perception of need for services. Understanding factors such as these which are associated with self-identification may present opportunities to target interventions to adapt structures so that self-identification is less critical to accessing supports and also assist young people in understanding problems that may benefit from mental health treatment. Given the barriers present in the environment, including stigma about mental illness, YAEH may resist identifying symptoms of depression, anxiety, and traumatic stress as evidence of a mental health problem or connecting those symptoms to a need for services (Brown et al., 2016). YAEH also may simply not recognize their symptoms as mental health problems due to different perceptions about the meaning of mental distress symptoms.

Currently, there are gaps in our knowledge related to self-identification of mental health problems in YAEH. While we know that YAEH have significant mental health symptoms, it is not clear whether they are identifying symptoms as problems in need of treatment. This study focused on self-identification of mental health problems among a large sample of YAEH across seven cities and the relationship of self-identification to perception of unmet need for services and use of formal mental health services including therapy and medications. Specifically, the study aimed to 1) examine the rates and correlates of identification of mental health problems among those that screened positive for mental health symptoms and 2) examine whether problem identification was associated with past year use of therapy and psychiatric medications and with unmet need. We hypothesized that self-identification would be related to white race/ethnicity and female gender and that it would be positively associated with service use.

Methods

Data came from the Homeless Youth Risk and Resilience Survey (HYRRS) conducted by study authors in seven U.S. cities (New York, Los Angeles, San Jose, Phoenix, St. Louis, Houston, and Denver). Data were collected in 2016–2017. We were intentional about city selection to ensure there was representation from cities in each of the US census regions (Northeast, Midwest, South, and West). The team created a standardized instrument and study protocol for recruiting and screening that was utilized consistently across all locations. To be eligible for the study, participants had to be experiencing housing insecurity the prior night (being in a shelter, streets, transitional housing, not being able to stay with family or acquaintances for more than 30 days) and be between the ages of 18 to 26. All study procedures were approved by the respective institutional review boards at each investigator’s University. Each site had a PI who individually funded the project data collection in their city including incentives for participants in the form of a gift card to a local store ($20-$25 across sites). Participants self-administered the survey after initial screening by study staff. It took approximately 45–60 min to complete.

Study Procedures

Approximately 200 YAEH were recruited in each of the seven locations using purposive sampling from settings that provided different types of services specifically for young adults experiencing housing instability or homelessness. Researchers intentionally sampled from different settings including shelters, drop-in centers, and transitional housing programs in order to include experiences of young adults accessing different types of homeless services. Research assistants approached and screened all youth at a given location for eligibility in person during scheduled data collection days. Potential participants were asked questions to establish whether they were experiencing housing instability and were between ages 18 and 26. Once eligibility was confirmed and informed consent obtained, participants were assessed for literacy using a modified version of the REALM-SF (Murphy et al., 1993) that asked them to read 10 of the more difficult words used in the survey and then assigned one point for each word read accurately. The REALM-SF has been validated in adolescent populations and a score of 3 or higher indicates a reading level above 4th grade (Manganello et al., 2017). For our study, those who scored 3 or lower on the screener were offered assistance with reading the survey. These procedures ensured that all young adults could fully participate and provide valid responses and allowed us to identify when those with lower literacy might need assistance. Across all seven sites, only about 1% required assistance. Those who passed the literacy screener, self-administered the survey on a tablet or computer. Young people were assigned a unique identifier based on a series of questions that created a code unique to them. This was used to ensure YAEH only took the survey one time.

Measures

We included measures of demographic characteristics, current and historical measures of mental health diagnosis and symptoms, a measure of self-identification of mental health problems, and measures of service use and perceived unmet need. We provide further details about each of these measures below.

Demographic Characteristics

Characteristics assessed included age, gender identity (including those identifying as transgender or non-binary), sexual orientation, race/ethnicity, education level, and foster care history. Gender was assessed with six categories which were collapsed into three categories for analysis– cisgender male, cisgender female, and gender expansive, which included YAEH who identified as transgender male, transgender female, gender queer, or something else. Sexual orientation was assessed with six options and re-coded dichotomously as heterosexual or LGBQ. The separation of gender identity and sexual orientation means that those identifying as transgender are represented in our gender expansive category and also as the sexual orientation with which they identify rather than be included as T in an LGBTQ category. Eight categories of race/ethnicity were also assessed then collapsed into five categories for analysis (Black, Latinx, non-Hispanic White, Multiracial, Other). Participants were also asked about history of being in foster care (yes/no) and whether they had a high school diploma/GED (yes/no).

History of Mental Health Diagnosis

Six questions created by the investigators asked separately about whether the participant had been diagnosed by a doctor or mental health professional with ADHD, Oppositional or Conduct Disorder, Bipolar Disorder, Depression, PTSD or Psychosis. These were combined into a single indicator for analysis of whether a participant had history of any mental illness (yes/no).

Mental Health Symptoms

Three scales were administered to screen for current and past-year symptoms of mental health problems. The Kessler-6 was used to assess psychological distress in the past 12 months (Kessler et al., 2003). The recommended cut-off of 13 and higher was used to identify individuals who were positive for psychological distress. Previous studies have looked at the psychometric properties of the Kessler-6 with youth (15–19) and adults (20–64) finding internal consistency was high (α = 0.86) (Ferro, 2019). The PHQ-9 was used to screen for current depression over the past two weeks (Kroenke et al., 2001). Scores of 10 and above are considered moderately or severely depressed and this was used as the cut point for considering participants positive for depression. The reliability of the PHQ-9 has also been researched by many for use across multiple age groups and studies have found good to excellent internal consistency ( α = 0.87–0.92; Bentley et al., 2021; Villarreal-Zegarra et al. 2019). The four-item primary care-PTSD (PC-PTSD) screening was administered to screen for current Post Traumatic Stress Symptoms (PTSS; Prins et al., 2003). In a study conducted by Soliemanpour et al. (2020) the PC-PTSD showed high sensitivity in identifying PTSD in adolescents. For this study, if participants responded “yes” to 3 of the 4 symptom questions they were considered to have screened positive for PTSS. We then used the three screeners described above to create a dichotomous variable for whether there was indication of screening positive for mental health symptoms, coding participants as positive who met criteria on any of the three scales as yes for mental health symptoms.

Severity of Mental Health Symptoms

After limiting the sample based on being positive on one of the three screeners, we also used the continuous version of the K6 scale in analyses to control for greater severity of symptoms among those who screened positive.

Problem Identification

To assess self-identification of mental health problems, participants were asked one question created for this study: “Are you currently experiencing problems with your mental health?” Response options included yes, no, or not sure.

Past Year Mental Health Service Use

Standard questions from the National Survey on Drug Use and Health (NSDUH, Center for Behavioral Health Statistics and Quality, 2016) were used to identify if participants had used mental health services. Separate questions asked if the participant had ever received psychiatric medications or ever received outpatient counseling. If participants selected yes, they were then asked whether they had received that service in the past year.

Unmet Need for Treatment

A standard dichotomous item from the NSDUH (Center for Behavioral Health Statistics and Quality, 2016) asked whether participants had ever thought they needed mental health services but hadn’t received them. If yes, a follow up question asked whether this was in the past year.

Creation of Study Sample

The sample for the current study was limited to those who screened positive for a mental health problem on one of the three screeners described above. Of the total original sample of 1426 participants across cities, 1369 responded to the three mental health screeners and 892 of those (65.2%), screened positive on one or more of the three instruments. Of the original sample, nearly half (48.1%, n = 656) screened positive for moderate to severe depression in the past two weeks, 41.6% (n = 567) screened positive for psychological distress in the prior year, and 39.6% (n = 541) screened positive for current post traumatic stress symptoms. Those who screened positive on one or more of the mental health screeners were significantly more likely to be female or gender expansive (X2 = 17.8, df = 2, p < 0.001), LGBQ (X2 = 31.3, df = 1, p < 0.001), and White or mixed race (X2 = 10.5, df = 3, p < 0.02) compared to those who did not indicate mental health symptoms.

Analysis

After limiting the sample only to those with indication of mental health need, we used IBM SPSS statistics (Version 27) (IBM Corp., 2020) to examine correlates of the three categories of identification of mental health problems (yes, no, unsure). ANOVA and chi-square tests were used to assess bivariate relationships then multinomial logistic regression was used to examine predictors of the relative risk of being in each self-identification group compared to each of the other groups. Multinomial logistic regression was run twice with different categories selected as the reference group to enable us to report on all possible comparisons. We conducted three separate logistic regression analyses focused on the contribution of problem identification to predicting past year use of medication, therapy, and unmet need, while controlling for other variables. Missing data was less than 5% across all variables.

Study Results

The sample for the current study (n = 892) was majority cisgender male (55%), 34% LGBQ, and racially/ethnically diverse with 21% identifying as White, 35% as African American, 17% as Hispanic and 28% as Multiracial or other races with an average age of 21 (see Table 1). Over a third of the sample reported a history of being in foster care (39.1%) and only 70% had completed high school or received a GED. Most reported having received a mental health diagnosis at some point in their lives (71.4%).

Table 1 Description of sample overall and by identification category
Full size table

Problem Identification

Just over one-third of the sample (35.2%, n = 309) identified as ‘yes’ to having a mental health problem. Another 22.2% (n = 198) endorsed the option ‘not sure’ about having a mental health problem. Notably, 41.5% (n = 370) responded ‘no’ that they did not consider themselves to have a mental health problem, even though the sample only included those that screened positive on mental health symptom screeners. Frequencies and bivariate associations between key variables and self-identification are presented in Table 1. Cisgender males had lower rates of self-identification (28%) compared to cisgender females (42%) and to gender expansive individuals (44%; X2 = 29.87, df = 4, p < 0.001). White youth had higher rates of self-identification (43%) and Hispanic (22%) and Black (33%) youth had lower rates (X2 = 26.71, df = 6, p < 0.001). Those who indicated that they had previously been diagnosed by a doctor with a mental disorder (X2 = 78.65, df 2, p < 0.001) and those who reported having been in the foster care system (X2 = 7.44, df = 2, p < 0.05) were also more likely to identify as having a mental health problem.

Multinomial logistic regression was used to examine relative risk ratios (RRR) of being in each category compared to each of the other two categories (Table 2). Older age (RRR = 1.15; CI = 1.06, 1.24), being female (RRR = 1.65; CI = 1.14, 2.40) or gender expansive (RRR = 2.33; CI = 1.15, 4.72), and being LGBQ (RRR = 1.74; CI = 1.19, 2.53) were each related to greater likelihood of stating “yes” to having a mental health problem compared to those that stated “no” they did not. Having been given a prior diagnosis (RRR = 4.62; CI = 2.96, 7.23) and having a higher psychological distress score (RRR = 1.06; CI = 1.03, 1.09) were also associated with being in the “yes” category compared to the “no” category. Those that identified as Hispanic were less likely than White youth to be in the “yes” group compared to the “no” group (RRR = 0.36; CI = 0.20, 0.65).

Table 2 Multinomial logistic regression model comparing relative risk of each identification category
Full size table

Prior diagnosis and higher psychological distress score were also significant predictors of likelihood of being in the “yes” category compared to the “not sure” category. Those who had been given a diagnosis in the past had relative risk of 3.32 (CI = 2.03, 5.45) greater of being in the “yes” category compared to the “not sure” category. Those with higher psychological distress scores also had greater relative risk of being in the “yes” category compared to the “not sure” category (RRR = 1.04, CI = 1.01, 1.07). Being cisgender female was also associated with greater likelihood of being in the “yes” category compared to the “not sure” category (RRR = 1.54, CI = 1.01, 2.35). The differences between being in the “not sure” category compared to the “no” category were not significantly related to the mental health indicators (prior diagnosis or psychological distress score). Instead, likelihood of being in the “not sure” category compared to the “no” category was related to older age (RRR = 1.12, 1.03, 1.23) and identifying as gender expansive compared to male (RRR = 2.58, CI = 1.26, 5.27).

Relations Between Problem Identification and Mental Health Services

The relations between problem identification and past year psychiatric medication, past year use of therapy, and past year unmet need for treatment were examined in separate multivariable logistic regression models with identification included as an independent variable with three categories using no as the reference group (see Table 3). Those who affirmatively stated that they had a mental health problem had significantly higher odds of using medications (OR = 2.53; CI = 1.73, 3.70) and higher odds of using outpatient therapy (OR = 2.26; CI = 1.52, 3.35) compared to those that stated no, they did not have a mental health problem. Both those that affirmatively stated they had a mental health problem (OR = 2.81; CI = 1.95, 4.05) and those that stated they were not sure whether they had a problem (OR = 1.91; CI = 1.28, 2.84) had significantly higher odds of reporting unmet need compared to those that stated they did not have a mental health problem, even when controlling for other variables. There were multiple other variables associated with mental health services such as prior diagnosis, prior foster care history, gender, age, and psychological distress. These relations are reported in Table 3 but are not discussed here as the focus of this paper is self-identification.

Table 3 Logistic regression models examining relations between problem identification and past year service use, unmet need for services
Full size table

Discussion

This study examined a concept that is foundational to understanding mental health service use -the young person’s own identification of mental health challenges. We found that self-identification was relatively uncommon (35.2%) in our sample of YAEH that were identified through standardized screening measures with evidence of mental health symptoms. We also noted that there was a significant group (22.2%) who reported uncertainty about whether they had mental health problems. The largest group, however, stated that they did not have a mental health problem (41.5%). Problem identification was positively related to use of therapy and medication services, highlighting it as a potential target for intervention to increase service utilization. Several findings are notable and have implications for practitioners and researchers.

First, findings from this study are in line with prior work on the importance of illness identity and illness narratives in shaping service use behaviors during young adulthood (Floersch et al., 2009; Munson, Floersch & Townsend; Munson et al., 2018). For example, Munson et al (2018) identified three distinct types of illness narratives in a population of young adults with mental health histories who had been involved in public systems of care (n = 40) – overwhelmed, integrated, and distanced. Individuals with an overwhelmed narrative spoke about or mentioned their illness and symptoms as being all consuming (Munson et al., 2018). Individuals with the integrated narrative would describe illness and symptoms as part of who they are (Munson et al., 2018). The distanced group, who were largely not engaged with mental health services, did not identify as having mental health problems and narrated stories of struggles with mental health as normal developmental struggles rather than an illness (Munson et al., 2018). This distanced group may be similar to the YAEH in this study who did not identify as having a mental health problem. While critical aspects of access to services need to be addressed to ensure YAEH who have more severe symptoms receive them, it appears that there is a mismatch between requiring YEAH to identify as having a mental health problem to access services and the current narratives of many YAEH that do not see themselves as having a problem.

Illness perceptions in young adults have also been found to vary by race and gender with cisgender men of color being least likely to identify with certain aspects of illness identity (Narendorf et al., 2018). This is congruent with the findings in our study where cisgender men were less likely to identify as having a mental illness compared to cisgender women or gender expansive young adults. And, young adults that identified as Hispanic were less likely to identify as having a mental health problem. This supports the idea that there are cultural dimensions to how young people come to understand symptoms and frame their problems as a mental illness or not, a potential contributor to differential rates of mental health service use (Cauce et al., 2002).

We also saw evidence that problem identification may be part of a developmental process. Identifying as having a mental health problem was related to older age, potentially indicating that young adults learn to frame their problems as mental health challenges as they have experiences of mental health symptoms over time. This is similar to prior work where youth taking psychiatric medications came to understand their effects over time through periods of use and non-use (Floersch et al., 2009; Narendorf et al., 2015). YAEH that are early into the transition to adulthood may require a different approach that provides more basic psychoeducation or assists in understanding symptoms compared to older young adults that may have developed some of these insights through their experience in transitioning to adulthood with mental health challenges. YAEH that are further into adulthood might benefit from more targeted psychoeducation matched to their conceptions of their illness as their illness narratives may be more engrained and require more specific supports to challenge those perspectives.

As expected, use of mental health services, both therapy and medications, was related to problem identification compared to those who did not identify any mental health problems. For some YAEH, there may be benefit of providing nonjudgmental and destigmatizing psychoeducation about common mental health symptoms and explaining that therapy and medications can provide relief from many symptoms. Those who indicated that they were uncertain about having a mental health problem may be a particular target for increased understanding of their symptoms and options for intervention, as they reported higher rates of unmet need for treatment compared to those that did not see themselves as having mental health problems. This indicates that this group is potentially seeing a need for treatment but has not yet secured services. General psychoeducation provided through youth service venues (shelters, drop-in centers, outreach) could be paired with clear paths for connection to services such as co-located mental health services so that YAEH can easily access services if they have increased awareness that their symptoms may merit mental health intervention and they decide they want services. Co-located mental health services, can also reduce the structural barriers to care as YAEH would not need to travel to multiple sites to get support.

It should be noted, however, that identifying as having a mental illness may not be the best path for all YAEH to find mental health support. Literature on stigma resistance has found that mental illness identity deflection can be an adaptive strategy for those that view mental health treatment as stigmatizing (Thoits, 2016). Given the large number of YAEH who clearly stated they did not have a mental health problem, homeless service providers should consider providing universal supports to enhance mental wellness that do not require identification of a mental health problem. Programming such as mindfulness and yoga can enhance emotion regulation, reduce reactivity, and reduce stress for all young people without requiring YAEH to identify as having a problem (Goyal et al., 2014; Sibinga et al., 2011). Framing therapeutic services as wellness supports rather than specifically tying them to mental health “problems”, may reduce stigma and the barrier of self-identification to increase use of mental health services and improve mental health among YAEH.

It is also important that we conduct a critical analysis of the current housing and mental health service system available to YEAH, as systemic problems may make young people less comfortable identifying mental health problems. Furthermore, our finding that YAEH who did identify as having a mental health problem, and were more likely to get therapy and medication, but were also more likely to report that they had unmet mental health needs, suggests need for improvements in the service system to better meet mental health and support needs. Previous literature has found our existing service systems are often poorly equipped to address the multitude of intersecting and systemic adversities young people experience and thus have difficulty engaging YAEH (Altena et al., 2017). YAEH have often developed an adaptive distrust of formal support systems (Auerswald & Eyre, 2002) and report experiencing stigma and discrimination, particularly around mental health needs, which deters them from seeking formal support from service providers (Hudson et al., 2010; Naert et al., 2020). Peer support, provided by those with shared lived experiences, has been shown to provide mental health support in a uniquely flexible, youth-led, and relational way (Erangey et al., 2020), supporting self-directed growth rather than conventional treatment goals (Erangey et al., 2021). Such adaptations to the mental health service model would not only offer more affirming options to YEAH seeking services but possibly also shift young peoples’ willingness to explore and gain greater understanding of their own needs for mental health support.

This study includes new information about how YAEH view their mental health symptoms, however, it does have some limitations. Data are cross-sectional and hence it is not possible to disentangle causal relationships. We hypothesized that problem identification would be associated with service use and although the hypothesis was supported, the relationship is likely bidirectional with individuals that utilize mental health services framing their problems as mental health problems. In addition, only one question was used to assess problem identification and this was a measure created for this study. Future research should include more detailed measures of stigma that more fully assesses young people’s understanding of their own symptoms and of mental health problems in general. Finally, all YAEH included in this study were purposively recruited from service settings. This means that these findings will not generalize to the experiences of all YAEH as many are not connected to formal services. In addition, agencies and services varied across cities which then influenced which YAEH might be included in the sample in each city. Results should be viewed as representing a large purposive sample of YAEH from across the United States rather than representative of all YAEH.

Conclusion

YAEH have high rates of mental health symptoms, but only one third of those who experienced significant symptoms identified themselves as having a mental health problem. Our study highlights problem identification as a potential modifiable factor, for some, that could increase rates of mental health service use. Approximately one in five YAEH screening positive for mental health symptoms expressed uncertainty about whether they had a problem, and psychoeducation may be particularly important in increasing use of services for this group. For young people who stated they did not have a mental health problem, mental health supports could be provided in ways that do not require identification as having a mental health problem. Strategies may include creating universal interventions and programming that utilize peers and have wellness, relational support, and resilience embedded into them without the need for individuals to name their symptoms as mental health problems. Pairing supports with education as a dual strategy can assist in reaching YAEH who are conceptualizing symptoms in a variety of ways to both provide low barrier support and open the door for formal mental health supports.