The Relationship Between On-Campus Service Utilization and Common Mental Health Concerns in Undergraduate College Students

Abstract

Objectives

Little research exists into the trends associated with on-campus service utilization for mental health concerns of college students. Rates of broad service utilization exist, but no published study has examined the direct relationship between a range of common mental health symptoms and on-campus service utilization. The aims of the present study are to explore: 1) which common mental health concerns are associated with specific on-campus service utilization in undergraduate students and 2) whether endorsement of more mental health concerns will predict a higher number of services utilized.

Methods

Data were utilized from 3,734 undergraduates at a large (more than 20,000 undergraduates), urban university (M_age = 19.94 years, SD = 0.55 years; female = 66%). Four on-campus services (University Counseling Services, University Health Services, The Wellness Resource Center, and Disability Support Services) were regressed onto mental health concerns associated with symptoms of three disorders (anxiety, depression, alcohol use disorder [AUD]) and two mental health risk factors (stressful life events [SLEs], antisocial behaviors [ASBs]).

Results

AUD symptoms predicted the most overall and specific service utilization, followed by depression symptoms and SLEs. Anxiety symptoms and ASBs were not significant predictors when combined with other variables.

Conclusions

This is the first study to investigate trends specific to on-campus college student service utilization. Findings will be helpful to mental health professionals on similar college campuses by providing insight into programming and outreach initiatives for these or related services.

Mental health concerns, including symptoms and risk factors for mental health disorders, are widespread on college campuses (Fletcher, Bryden, Schneider, Dawson, & Vandermeer, 2007; Saleem, Mahmood, & Naz, 2013; Storrie, Ahern, & Tuckett, 2010) yet little research has been published that investigates university-provided service utilization for these concerns among students. Because the extant literature in this area tends to focus on service utilization broadly, and rarely specifies whether students seek help on- or off-campus (Cranford, Eisenberg, & Serras, 2009; Eisenberg, Hunt, Speer, & Zivin, 2011; Herman et al., 2011; Sontag-Padilla et al., 2016; Wu, Pilowsky, Schlenger, & Hasin, 2007), there are significant gaps in knowledge about which students are most likely to use which university-provided services (Armstrong & Young, 2015; Bernhardsdottir & Vilhjalmsson, 2013; Eisenberg, Hunt, & Speer, 2012; Kanuri, Taylor, Cohen, & Newman, 2015). Although many on-campus providers and administrators internally track this data, this information is not widely available to researchers nor published in peer-reviewed journals. Further, services offered on college campuses can vary widely, even within a campus, and they all address mental health in different capacities. Therefore, the present study seeks to determine which mental health concerns predict utilization of various on-campus services. This knowledge will contribute to the service utilization knowledge base by elucidating which mental health concerns are most likely to be associated with on-campus service use. This will ultimately assist college mental health providers with decisions regarding how to properly address the needs of their students.

As many as 25% of college students suffer from common mental health disorders such as anxiety (Lipson, Zhou, Wagner III, Beck, & Eisenberg, 2016), depression (Lipson et al., 2016), and alcohol use disorder (AUD) (Blanco et al., 2008; Slutske et al., 2004; Wu et al., 2007), making these three concerns the most prevalent among college students when excluding personality disorders (Blanco et al., 2008). Even more troubling, prevalence of sub-threshold symptomatology is typically even higher for these mental health conditions (Boswell, McAleavey, Castonguay, Hayes, & Locke, 2012; Holmes & Silvestri, 2016; Kanuri et al., 2015). Further, stressful life events (SLEs) are a common risk factor for poor mental health on college campuses and other young adults of this age (Amstadter et al., 2013), including for the aforementioned conditions (Barr, 2014; Bernhardsdottir & Vilhjalmsson, 2013; Cranford et al., 2009; D’Amico, Mechling, Kemppainen, Ahern, & Lee, 2016; Fletcher et al., 2007; Holmes & Silvestri, 2016; Lipson et al., 2016). The prevalence of SLEs can be as high as 26% for specific trauma-exposed life events (Elhai et al., 2012) and 85% for lifetime endorsement of at least one aversive or stressful life event (Smyth, Hockemeyer, Heron, Wonderlich, & Pennebaker, 2008). Finally, prevalence of antisocial behaviors (ASBs) has long been known to substantially increase during the time period of adolescence (Moffitt, 1993). These behaviors have been associated with alcohol use prevalence among college-age adults (Dick et al., 2014). They are also a risk factor for later disorders and associated with various SLEs common among college-age individuals (e.g., sexual coercion and assault), as well as peer group deviance (Gámez-Guadix, Straus, & Hershberger, 2011; Moffitt, 1993). Together, these five interrelated concerns are problematic for students and should be investigated in the context of mental health service utilization.

The exact details of service provision for mental health needs naturally vary by campus but most include general health care, women’s centers, mental health care, disability services, wellness resource centers, career counseling, and alcohol and other drug education programs. Many factors contribute to seeking on-campus help for mental health concerns, including perceived risk, stigma, insurance, living on- or off-campus, and knowledge of how to use services (D’Amico et al., 2016). Although it has been reported that up to 93% of students are aware of services related to mental health (Fletcher et al., 2007), on-campus utilization of counseling services is fairly low (13%) for those with any mental health problem, including anxiety (18%) and depression (16%) (Eisenberg et al., 2011). Most other estimates of utilization rates among college students combine on- and of-campus services, making it difficult to gauge the impact of individual on-campus services. This broader service utilization for mental health concerns appears to fluctuate by disorder, gender, and race/ethnicity. Specifically, utilization varies greatly depending on whether a student has anxiety (5–35%), depression (3–39%), or alcohol-related concerns (3–39%) (Caldeira et al., 2009; Cellucci, Krogh, & Vik, 2006; Cranford et al., 2009; Eisenberg et al., 2011; Lipson et al., 2016; Wu et al., 2007). Females are widely known to utilize more services than males for all health concerns, including mental health, as well as ultimately receive more treatment (Eisenberg et al., 2011; Eisenberg et al., 2012; Sontag-Padilla et al., 2016). Comparatively, college students who are Black, Asian, or Hispanic have consistently been found to seek fewer services, receive less treatment, and report more barriers to treatment compared to white students (Eisenberg et al., 2012; Herman et al., 2011; Miranda et al., 2015). Thus, mental health concerns go untreated more often than not among college students, with much unknown about how and where they seek services and whether previously noted trends in who is more likely to utilize resources are also evident when restricted to examining on-campus services. To help students get the help they need, it is important to track on-campus service utilization in the context of specific mental health concerns and risk factors.

Purpose & Research Aims

There is a dearth of published evidence as to which on-campus services are more likely to be used by college students struggling with various mental health concerns. The current study seeks to address this gap by examining whether common and often co-presenting mental health concerns, anxiety symptoms, depression symptoms, AUD symptoms, SLEs, and ASBs, predict on-campus utilization of four distinct services that all address mental health needs. The specific aims are to determine 1) which mental health concerns predict specific service utilization and 2) which concerns are associated with greater number of services utilized.

Methods

Participants

Data was analyzed from the Spit for Science study, which is an ongoing, longitudinal study at a large (more than 20,000 undergraduates), urban university with the purpose of exploring how genes and environment impact alcohol/substance use and mental health (Dick et al., 2014). As of Spring 2016, there have been four waves of data collection with a total of 9,892 participants. The demographics are representative of the larger university population. Data were collected and managed using REDCap (Research Electronic Data Capture) (Harris et al., 2009). Students 18 years and older were invited to voluntarily complete an initial survey in REDCap and to provide their genomic data via in-person saliva collection their freshman year. They received $10 and a t-shirt for their participation. Follow-up surveys were completed every spring thereafter. A complete overview of the Spit for Science project can be found elsewhere (Dick et al., 2014), including an overview of institutional board approval, participant consent, and methodology.

Analyses in the current study were limited to data from college sophomores across all waves. Sophomore cohorts were chosen in order to maximize sample size (freshman and sophomores are the best-represented cohorts due to survey attrition) and cohort opportunity to use on-campus health services (freshmen have had only a few weeks’ opportunity, whereas sophomores have had about 18 months’ opportunity). Predictors of attrition between freshman and sophomore years can be found in the supplement (see Supplemental Text and Supplemental Table 1). The total analytic sample size was 3,734 (M_age = 19.94 years, SD = 0.55 years; female = 66%; Caucasian = 50.06%; Black/African American = 19.21%; Asian = 16.56%; another race = 8.07%; Hispanic/Latino = 6.09%). In addition, only participants who completed 50% or more of a measure were used in analyses of that measure to reduce bias introduced by potentially non-compliant responders.

Measures

For mental health concerns, full surveys were reduced in order to minimize participant burden. Survey items were selected using item response theory so that only items that provided good discrimination at various locations along the range of the latent scale were retained.

Anxiety

Anxiety symptoms were assessed via four items from the Symptom Checklist-90 (SCL-90) (Derogatis & Cleary, 1977), including: 1) nervousness or shakiness inside; 2) suddenly scared for no reason; 3) feeling fearful; 4) spells of terror or panic. For each question, participants were asked to rank their endorsement of each symptom for the last 30 days on a scale of 1 (“Not at All”) to 5 (“Extremely”). A sum score for these questions was created with a range of 4–20, with higher scores representing more anxiety (Cronbach’s α = .84).

Depression

Depression symptoms were assessed via four items from the Symptom Checklist-90 (SCL-90) (Derogatis & Cleary, 1977), including: 1) feeling blue; 2) worrying too much about things; 3) feeling no interest in things; 4) feeling hopeless about the future. For each question, participants were asked to rank their endorsement of each symptom for the last 30 days on a scale of 1 (“Not at All”) to 5 (“Extremely”). A sum score for these questions was created with a range of 4–20, with higher scores representing more depression (Cronbach’s α = .85).

AUD

Sixteen questions from the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA) (Bucholz et al., 1994) were used to assess symptoms of DSM-5 (American Psychological Association, 2013) AUD. Each item assessed symptoms experienced since starting college. There were 16 questions assessing drinking behaviors in the past 12 months. These questions included items such as interference in daily activities, taking drugs while drinking, being arrested for drunken behavior, and trembling or sweating when not drinking. For each question, participants were either asked to rank their endorsement of each symptom on a scale of 1 (“Never”) to 5 (“3 or more times”) or on a scale of 0 (“No”) to 1 (“Yes”), depending on the question. An AUD sum score with a range from 0–11 was created, representing the 11 DSM-5 AUD criteria, with endorsement of more symptoms indicating increasing AUD severity (Chronbach’s α = .87).

SLEs

Participants were asked to report if they had experienced any of the following SLEs since starting college in a binary fashion, 0 (“No”) or 1 (“Yes”): 1) a broken engagement or steady relationship; 2) separation from other loved one or close friend; 3) serious illness or injury; 4) burglarized or robbed; 5) trouble with police; 6) laid off or fired from a job; 7) major financial problems; 8) serious housing problems; 9) serious difficulties at school; 10) someone close to you passed away; 11) a parent had a serious illness or injury; 12) someone else close to you had a serious illness or injury. A sum score of total SLEs since starting college was created with a range of 0–12 (Chronbach’s α = .61).

ASBs

Antisocial behaviors were assessed via items from the SSAGA (Bucholz et al., 1994), including the following three questions regarding participants’ behavior since starting college: 1) have you deliberately damaged or destroyed property that did not belong to you; 2) have you broken into a car or van to try to steal something out of it; 3) have you carried a knife or other weapon with you for protection or in case you needed it in a fight. For each question, participants were asked to rank their endorsement of each symptom on a scale of 1 (“Not at all”) to 4 (“6 or more times”). A sum score for these questions was created with a total range of 3–12, with higher scores representing more antisocial behavior (Chronbach’s α = .21).

College Services

Information was collected on utilization of four college health services that are fairly common on universities nationwide: 1) University Counseling Services, 2) University Health Services, 3) The Wellness Resource Center, and 4) Disability Support Services. All of these services are free to students who pay their student health fee, which is required for full-time students. Participants were asked if they had used each health service since starting college in a binary fashion. In addition to utilizing these extant variables for aim 1, a sum score representing total number of services used was created for use in aim 2.

University Counseling Services (2016) provides therapy services (individual, group, couple, crisis), professional training, consultation and outreach, and specific services for international students. While the remaining services do not address mental health as comprehensively as University Counseling Services, they each address such concerns in their own manner, detailed below.

University Health Services (2016) provides outpatient primary care in the areas of allergy shots, blood/body fluid exposures, immunizations, mental health, nutrition consultations, travel health, and women’s health. They specifically employ psychiatrists to help with students’ mental health concerns, including evaluations and pharmacologic treatment. A student is referred to University Counseling Services if they request formal counseling from University Health practitioners.

The Wellness Resource Center (2016) disseminates resources about healthy and safe lifestyles, including general wellness education, confidential advocacy services for sexual violence, alcohol and other drug education, early intervention programs for adjudicated students, free HIV testing, and a collegiate recovery program for substance use disorders. The director of the Wellness Resource Center is a licensed psychologist and most staff hold clinical licensures. Mental health is a key component of wellness and permeates what the center offers (i.e., advocacy naturally includes emotional and behavioral support).

Finally, Disability Support Services (2016) assists students with gaining access to all university programs, services, and events by providing information on disability services, making recommendations, helping the university comply with the Americans with Disabilities Act, providing technical assistance, and consulting with necessary personnel. It is not uncommon for students to require disability services assistance for common mental health concerns such as anxiety and depression, nor for such concerns to be the product of having a different type of disability.

Statistical Analyses

All statistical analyses were performed in the R software package (R Core Team, 2014). Univariate logistic regressions were used to examine the relationship between mental health concerns and utilization of each service. For aim 1, the binary score for each service was regressed separately onto sum score for each group of items (cluster) and demographic covariates to determine which concern predicts specific service utilization. Holm’s Family-Wise Error Rate (FWER) method was used to correct for the fact that multiple univariate tests were run per dependent variable. All significant variables were then entered simultaneously into a multiple logistic regression model. For aim 2, univariate linear regressions were used to examine the relationship between concerns and total number of services utilized. Specifically, the total number of services utilized was regressed separately onto each mental health sum score and demographic covariates. After correcting for Holm’s FWER, all significant variables were then entered simultaneously into a multiple linear regression model. Self-reported sex, race/ethnicity, and age were treated as covariates in all analyses.

Results

Descriptive Statistics

See Table 1 for means and standard deviations for each of the assessed mental health concerns and services utilized, stratified by gender and race/ethnicity.

Table 1.

Means and standard deviations for mental health concerns and endorsement of services assessed in the current study.

	Anxiety Symptoms	Depression Symptoms	Stressful Life Events	Antisocial Behaviors	Alcohol Use Disorder Symptoms	University Counseling Services	University Health Services	The Wellness Resource Center	Disability Support Services	Total Service Utilization
Range Mean (Standard Deviation)	4–20	4–20	0–12	3–12	0–11	0–1	0–1	0–1	0–1	0–4
Total Sample	6.51 (3.04)	9.50 (4.01)	1.88 (1.82)	3.45 (1.04)	1.03 (1.52)	.19 (.39)	.66 (.47)	.23 (.42)	.02 (.15)	1.11 (.86)
Male	5.92 (2.54)	8.67 (3.82)	1.65 (1.81)	3.71 (1.28)	1.21 (1.66)	.18 (.38)	.54 (.50)	.26 (.44)	.03 (.16)	1.00 (.88)
Female	6.79 (3.21)	9.90 (4.03)	2.00 (1.81)	3.31 (.86)	.94 (1.45)	.20 (.40)	.73 (.45)	.21 (.41)	.02 (.15)	1.16 (.84)
White	6.85 (3.13)	9.81 (4.02)	1.94 (1.77)	3.59 (1.15)	1.11 (1.69)	.22 (.42)	.68 (.47)	.21 (.41)	.03 (.17)	1.15 (.87)
Black	5.76 (2.48)	8.91 (3.84)	1.93 (1.87)	3.31 (.90)	.90 (1.30)	.15 (.35)	.70 (.46)	.27 (.44)	.01 (.09)	1.12 (.84)
Hispanic	6.67 (3.22)	9.58 (3.85)	2.06 (1.94)	3.36 (.95)	1.16 (1.43)	.24 (.43)	.67 (.47)	.34 (.48)	.03 (.17)	1.29 (.88)
Asian	6.31 (2.98)	9.19 (4.11)	1.55 (1.78)	3.31 (.93)	.92 (1.37)	.16 (.36)	.59 (.49)	.19 (.40)	.02 (.13)	.95 (.82)
Another race / ethnicity	6.86 (3.38)	9.85 (4.15)	2.04 (1.85)	3.40 (.92)	.96 (1.33)	.22 (.42)	.70 (.46)	.20 (.40)	.05 (.21)	1.17 (.91)

Specific Service Utilization

See Table 2 for the full parameter estimates showing the relationship between mental health concerns and utilization of each service.

Table 2.

Logistic regression parameter estimates showing relationship between mental health concerns and service utilization

Model and Variables	Univariate Regression OR (95% CIs)	Multiple Regression OR (95% CIs)
Model 1 (DV = University Counseling Services)
Anxiety	1.136 (1.102–1.170)***	1.034 (.986–1.060)
Depression	1.130 (1.102–1.159)***	1.099 (1.057–1.141)***
Stressful Life Events	1.157 (1.096–1.220)***	1.062 (.994–1.135)
Antisocial Behaviors	1.126 (1.103–1.234)	--
Alcohol Use Disorder	1.108 (1.022–1.202)	--
Age	1.175 (.957–1.443)	--
Female	1.170 (.938–1.459)	--
Black	.595 (.446 – .794)**	.648 (.476–.881)**
Hispanic	1.124 (.744–1.698)	--
Asian	.645 (.482–.862)*	.757 (.557–1.030)
Another race / ethnicity	1.012 (.689–1.487)	--
Model 2 (DV = University Health Services)
Anxiety	1.030 (1.001–1.060)	--
Depression	1.026 (1.004–1.049)	--
Stressful Life Events	1.140 (1.082–1.201)***	1.079 (1.001–1.152)*
Antisocial Behaviors	.974 (.897–1.058)	--
Alcohol Use Disorder	1.064 (.981–1.154)	--
Age	.898 (.750–1.076)	--
Female	2.269 (1.898–2.714)***	1.979 (1.578–2.483)***
Black	1.085 (.861–1.367)	--
Hispanic	.961 (.660–1.399)	--
Asian	.664 (.530–.832)**	.642 (.504 – .818)***
Another race / ethnicity	1.103 (.778–1.564)	--
Model 3 (DV = The Wellness Resource Center)
Anxiety	1.014 (.983–1.045)	--
Depression	1.010 (.986–1.034)	--
Stressful Life Events	1.132 (1.075–1.192)***	1.092 (.996–1.197)
Antisocial Behaviors	1.072 (.980–1.173)	--
Alcohol Use Disorder	1.137 (1.052–1.228)*	1.164 (1.053–1.287)**
Age	.984 (.799–1.212)	--
Female	.785 (.642–.959)	--
Black	1.342 (1.048–1.717)	--
Hispanic	1.911 (1.309–2.790)**	1.163 (.673–2.010)
Asian	.884 (.673–1.161)	--
Another race / ethnicity	.889 (.595–1.330)	--
Model 4 (DV = Disability Support Services)
Anxiety	1.145 (1.073–1.222)***	1.042 (.906–1.197)
Depression	1.130 (1.062–1.201)**	1.076 (.950–1.218)
Stressful Life Events	1.184 (1.039–1.349)	--
Antisocial Behaviors	1.390 (1.161–1.663)**	.889 (.610–1.297)
Alcohol Use Disorder	1.270 (1.072–1.503) *	1.228 (1.027–1.469)*
Age	1.541 (1.074–2.209)	--
Female	.826 (.478–1.429)	--
Black	.271 (.095–.771)	--
Hispanic	.979 (.340–2.816)	--
Asian	.585 (.267–1.283)	--
Another race / ethnicity	1.604 (.725–3.550)	--

Note. Only significant univariate parameters were moved into the multiple models; Reference sex = male; Reference race = white;

^*p < .05;

^**p < .01,

^***p < .001 (asterisks reflect corrected p-values)

University Counseling Services

Anxiety symptoms, depression symptoms and SLEs significantly predicted utilization of University Counseling Services, while being Black and being Asian significantly predicted not utilizing this service. When these variables were combined in a multiple logistic regression, only depression and being Black were significantly associated with use of University Counseling Services. Each unit increase in the depression score was associated with a 9.9% (p < .001) increase in likelihood of using University Counseling Services while being Black was associated with a 35.2% (p < .05) decreased likelihood of utilizing this service.

University Health Services

SLEs and being female significantly predicted utilization of University Health Services, and being Asian significantly predicted not utilizing this service. Each of these variables remained significant after they were combined in a multiple logistic regression. For each SLE experienced, individuals were 7.9% (p < .05) more likely to use University Health Services. Further, females were 97.9% (p < .001) more likely than males and Asians were 35.8% (p < .001) less likely than whites to use this service.

The Wellness Resource Center

AUD symptoms, SLEs, and being Hispanic significantly predicted utilization of The Wellness Resource Center. Only AUD symptoms were a significant predictor after each of these were combined in a multiple logistic regression. Each additional AUD symptom was associated with a 16.4% (p < .05) increase in likelihood of using the Wellness Resource Center.

Disability Support Services

Anxiety, depression, AUD symptoms, ASBs, and age significantly predicted utilization of Disability Support Services, while being Black significantly predicted not utilizing this service. When these variables were combined in a multiple logistic regression, only AUD symptoms were significant. Each additional AUD symptom predicted a 22.8% (p < .05) increase in likelihood of using Disability Support Services.

Overall Service Utilization

See Table 3 for the full parameter estimates showing the relationship between mental health symptoms and risk factors and total number services utilized. In the univariate linear regression, there was a significant effect of anxiety, depression, AUD symptoms, SLEs, ASBs, being female, and being Asian on service utilization. When these variables were entered into a multiple linear regression, only AUD symptoms were significantly associated with overall service utilization, with each additional AUD symptom score predicting use of 7.5% more services (p < .001).

Table 3.

Linear regression parameter estimates showing relationship between mental health concerns and total number of services utilized

Variable	Univariate Regression β (95% CIs)	Multiple Regression β (95% CIs)
Anxiety	.036 (.025–.047)***	.009 (−.016–.033)
Depression	.020 (.022–.039)***	.018 (−.001–.037)
Stressful Life Events	.079 (.060–.099)***	.029 (−.005–.062)
Antisocial Behaviors	.039 (.005–.073)*	−.028 (−.080–.024)
Alcohol Use Disorder	.063 (.033–.092)***	.075 (.037–.112)***
Age	.016 (−.059–.091)	--
Female	.163 (1.066–1.675)***	.044 (−.073–.160)
Black	−.023 (−.118–.065)	--
Hispanic	.140 (−.011–.291)	--
Asian	−.193 (−.287–−.010)***	−.109 (−.239–.022)
Another race / ethnicity	.021 (−.117–.158)	--

Note. Only significant univariate parameters were moved into the multiple models; Reference sex = male; Reference race = white;

^*p < .05;

^**p < .01,

^***p < .001 (asterisks reflect corrected p-values)

Discussion

The current study was the first to explore the relationship between key mental health concerns/risk factors (anxiety, depression, and AUD symptoms, SLEs, ASBs) and on-campus service utilization (University Counseling Services, University Health Services, The Wellness Resource Center, and Disability Support Services) outside of potential internal university reports. We sought to determine which concerns predicted specific service utilization and greater number of services utilized. Results suggest that each service may be sought for unique yet sometimes overlapping mental health concerns. This implies that while on-campus health systems can be difficult to navigate, students have some understanding of the unique purpose of various services.

Specifically, endorsement of internalizing concerns (anxiety, depression, SLEs) predicted utilization of University Counseling Services, of which only depression remained significant when taking other variables into consideration. This indicates that despite being aimed at general mental health services, not all mental health concerns drive students to seek help here. SLEs were notably the only significant mental health predictor of University Health Services, perhaps speaking to the more physical nature that some SLEs can have (i.e., traumatic illness), including those assessed in this study (note that stressful events related to sexual assault or natural disaster were not included in the current measure for SLEs). SLEs along with AUD symptoms predicted utilization of the Wellness Resource Center, only the latter of which remained significant after controlling for other variables. As with similar health and wellness-focused organizations around the country, two of this center’s three primary foci are advocacy and recovery services, which align with SLE and AUD concerns, respectively. Finally, all concerns except SLEs predicted use of Disability Support Services, highlighting the often overlooked role that this service plays in supporting students’ mental health needs. However, only AUD remained significant in the full model. Mental health concerns can both lead to and be a result of disability. In regard to the relationship between mental health concerns/risk factors and the total number of services utilized, endorsement of each mental health condition was associated with an increase in utilization. Again, only AUD symptoms significantly predicted overall service utilization when all conditions were considered jointly.

There were also significant effects of sex and race at both the individual and combined levels of analysis. Females were more likely to utilize University Health Services. This effect did not diminish when combined with other significant variables and is consistent with past literature (Eisenberg et al., 2011; Sontag-Padilla et al., 2016; Wang et al., 2005). Differences in mental health service utilization among minority individuals are also in line with past research and epidemiological studies (Eisenberg et al, 2011; Jackson et al., 2004; Miranda et al., 2015; Sontag-Padilla et al., 2016). Specifically, we found that Black individuals were less likely to utilize University Counseling Services even after other variables were taken into consideration. Asian individuals were less likely to seek out University Counseling or University Health Services, the latter trend holding when combined with other significant variables. Asian individuals were also less likely to seek services overall. There are several possibilities for the lower utilization of services by minority college students (Eisenberg et al., 2012). A recent study explicitly tested many of these, finding that lack of lack of time and stigma were barriers in minority but not white students. (Miranda et al., 2015). Finally, Hispanic individuals were more likely to utilize The Wellness Resource Center, a trend that is contrary to past findings of health seeking behaviors broadly; however, there is not a direct comparison to this service in extant literature.

The current findings expand on past literature by going beyond examination of the prevalence of mental health disorders (Barr, 2014; Bernhardsdottir & Vilhjalmsson, 2013; Cranford et al., 2009; D’Amico et al., 2016; Fletcher et al., 2007; Holmes & Silvestri, 2016; Lipson et al., 2016; Regehr, Glancy, & Pitts, 2013; Sontag-Padilla et al., 2016) and the broad service utilization rates of college students (Buscemi et al., 2010; Caldeira et al., 2009; Cellucci et al., 2006; Cranford et al., 2009; Eisenberg et al., 2011; Herman et al., 2011; Wu et al., 2007) to explore how various mental health concerns predict specific on-campus service use. While there are no studies that are directly comparable to the current findings, continuities and trends can be noted. Specifically, on-campus prevalence rates of alcohol use and related disorders are notable (Blanco et al., 2008; Dawson, Grant, Stinson, & Chou, 2004; Slutske, 2005; Wu et al., 2007), with service utilization in the current study somewhat in line with such findings, showing that students with AUD symptoms utilize the most services. Somewhat surprisingly, the current study did not find an association between service utilization and anxiety, depression, or SLEs in the larger multiple regressions. It has been previously noted that on-campus use of counseling services among those with anxiety and depression are notable and that anxiety, depression, and SLEs are moderately prevalent on college campuses (Eisenberg et al., 2011; Elhai et al., 2012; Lipson et al., 2016; Smyth et al., 2008). These conditions are often comorbid with each other and this shared variance could explain a lack of power to predict significant associations when combined in a multiple regression. Finally, this is the first study to examine ASBs in relation to service utilization among a college sample and adds to that body of research. While significance of ASBs in predicting utilization was not found when combined with other significant factors, it was still found to significantly predict use of University Counseling Services and Disability Support Services in univariate regressions.

Limitations

The present results should be considered in the context of several limitations. First, the sample was limited to college sophomores and a single college campus. This may limit generalizability to other populations. Specifically, the current sample had more racial/ethnic minorities, as well as slightly different proportions of each race/ethnicity, compared to the composition of average US universities (American College Health Assessment, 2016). Our population was also somewhat self-selecting, in that students could choose whether or not to participate. However, there was a ~70% response rate and the demographic characteristics of these individuals was in line with the larger university population (Dick et al., 2014). In addition, due to the nature of the current sample, results will likely differ from populations more clinical in nature (e.g., treatment seeking).

Second, it is important to note that other campuses will likely have services that vary from those offered at the university under study (e.g., The Wellness Resource Center) and that utilization of services may change as students progress through their course of study. We were unable to directly study the effects of diagnosed mental health disorders (i.e., we relied on endorsement of symptoms/indicators), which would likely reflect greater severity and thus greater utilization. However, we chose to rely on measures that reliably and validly assess symptom and risk factor endorsement to reduce administration burden, maximize item-level endorsement, and therefore increase power.

Finally, the current study did not assess the amount of use and type of treatment sought (i.e., each service offers many treatment options). Therefore, any service utilization reported was not necessarily for mental health concerns, and the frequency of service utilization was not available. Relatedly, we were not able to assess whether students sought treatment for alcohol due to mandatory disciplinary action. Despite these limitations, this study captures the unique service-seeking behaviors of a large university population, and results indicate that different mental health concerns and demographic characteristics influence the use of different services.

Future Directions and Implications

The goal of providing on-campus services is to increase students’ health at multiple levels of care, including individual level treatment and population level prevention. Thus, the current study has implications for both research and programming, which support one another and aid in the endeavor of improved student heath. Future research efforts should focus on conducting similar studies at different types of universities, as services vary by school and the mental health concerns associated with services in this study may not generalize to different populations. More research should also be published on which types of therapy, programming, and other assistance received at each on-campus service are most effective for certain mental health conditions. This may lend insight into why certain mental health concerns and demographics influence service utilization. A few studies have examined the effect of therapy type broadly in college samples (Bernhardsdottir, Champion, & Skärsäter, 2014; Stewart, Dispenza, Parker, Chang, & Cunnien, 2014), but no comprehensive studies of this topic exist. Further, the issues surrounding why students do not seek out services should be explored to ensure that they are not due to preventable barriers such as mental health stigma or health insurance. For example, at least one study (Nunes et al., 2014) has highlighted the barrier of health insurance when seeking mental health treatment among college students.

Finally, there is always more to be done to study the nuances of mental health disorders on college campuses among racial/ethnic minorities, LGBTQ students, and international students. Again, some studies have examined these nuances broadly (Kerr, Santurri, & Peters, 2013; Miranda et al., 2015; Nilsson, Berkel, Flores, & Lucas, 2004) but not as many exist that comprehensively examine this in relation to specific disorders and/or services utilized. For example, at least one previous study has found that racial minorities cite more treatment barriers to on-campus mental health counseling services compared to white students (Miranda et al., 2015). This knowledge combined with the results of the current study indicate that one potential future direction could be to examine demographic characteristics as mediators of the relationship between mental health and on-campus service utilization. Such information could provide strong evidence for adjusting on-campus mental health care and awareness to be more inclusive.

Despite the gaps in knowledge surrounding on-campus service utilization, the current findings can inform programming at similar universities. Our results suggest that service advertising is being received and internalized by students, at least to a certain degree. However, outreach from each service can expand toward students who struggle with the mental health concerns not currently associated with treatment-seeking at the specific provision. For example, if University Counseling Services can treat internalizing and externalizing disorders, but students only seek these services for internalizing symptoms, then awareness of the scope of this service should be expanded. Relatedly, students may not seek help when or where their concerns would be best addressed, as evidenced by the fact that the most concerns predicted utilization of Disability Support Services. If students are potentially waiting until their concerns are at a disabled level (one reason students seek help at this service), then the other on-campus services need to work at reducing stigma and barriers keeping students from seeking treatment sooner. Finally, efforts such as an inter-service task force that pool resources in order better inform prevent and intervention efforts for students should be undertaken.

Conclusions

The current study found that AUD symptoms predicted the most overall and specific service utilization, followed by depression symptoms and SLEs. While mental health challenges are widespread on college campuses and broad trends are well studied (Hoeppner, Hoeppner, & Campbell, 2009; Storrie et al., 2010), it is important to examine how specific mental conditions are managed by college students. The current findings indicate which types of services appeal to students the most (i.e., those with AUD symptoms utilize overall wellness services) and provides information about where specific types of services and practitioners (psychiatrists, addiction counselors, etc.) may be best utilized. This may potentially inform universities with similar on-campus services as to whether students with specific disorders are being appropriately targeted in programming across campus.