Mental Health Literacy in Emerging Adults in a University Setting: Distinctions Between Symptom Awareness and Appraisal

Procedure

Prior to beginning the study, approval was obtained from the local institutional review board. Students from a mid-sized Canadian university were recruited. In compensation for their time, students were given the option of receiving 1% credit toward their final grade through the university’s participant pool or entering into a draw for one of two CAD$50 gift certificates to a large box-store. Efforts were made to recruit students across disciplines through presentations in classes across campus; however, participation was higher for those receiving credit (approximately two thirds of responses). The study was completed online using SurveyMonkey.

Participants

A total of 154 individuals completed the survey. Prior to providing consent, participants were informed that incomplete surveys would be considered a withdrawal from the study should more than 20% of questions be left unanswered. Nine participants left more than 20% of questions unanswered and were subsequently removed from the sample. Individuals above the age of 25 were excluded from the sample in order to more accurately represent emerging adults in the university setting. This resulted in the removal of an additional 23 participants. The final sample for this study consisted of 122 undergraduate students ranging in age from 17 to 25 years. Demographic characteristics are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Demographic Characteristics of Sample (n = 122).

Characteristic	Statistics
Age, M ± SD	20.67 ± 2.03
Sex
Male	19.7%
Female	80.3%
Year of study
First year	29.5%
Second year	25.4%
Third year	20.5%
Fourth year or higher	24.6%
Marital status
Single/never married	93.4%
Married/common-law	6.6%
Race
Caucasian	72.1%
Asian/Pacific Islander	14.8%
Aboriginal, First Nations, or Metis	7.4%
African American	0.8%
Middle Eastern	0.8%
Other (not specified) a	4.1%

a

Individuals who did not identify with one of the cultural groups listed above.

Demographic factors are integral to mental health concerns. Socioeconomic variables were not directly queried in the current study; however, the university population from which the sample was sought has previously been examined and compared with national rates. A brief review of these factors allows for a better contextualization of our study’s findings. Demographics were drawn from the Office of Resource Planning, University of Regina (2011a, 2011b). Demographic information from the current sample is described further in the “Results” section.

The average age of undergraduate students in the sampled population is 24 years, slightly older than the national average (22 years). The number of female students is on par with the national average (66% in sampled population, 67% nationally). Although a smaller percentage of students identify as a visible minority (19%) in the studied population compared with the national average (25%), the population of aboriginal students is higher (10%) than the national average (5%). These rates are similar to the demographic of the current sample (see Table 1).

A similar proportion of students report living with their parents during their studies in the sampled population than nationally (43% and 40%, respectively); however, more students report being employed during their studies (62%), compared with national rates (56%). The most common source of financing for undergraduate students at the current university is from parents, family, or spouse (60%), which is similar to the national average (58%). Thus, although socioeconomic status was not directly examined in the current sample, the population’s characteristics are comparable with that of the national average, with only cultural distinctions.

Measures

Study Design

Upon following the link to the study, participants were provided with a description of the procedure and purposes of the study. Once participants provided informed consent to participate, they completed demographic questionnaire, the KUQ, and the ATSPPH-SF. Subsequently, participants responded to the open-ended Perception of Mental Health and Help-Seeking Questionnaire. The Warning and Action Signs Checklist was the final portion of the study completed by participants. The open-ended questions were asked prior to the presentation of the Warning and Action Signs Checklist to ensure that responses were not influenced by being presented a list of symptoms. Quantitative analyses were undertaken with questionnaire results to better understand the relationship between help-seeking and various factors (e.g., beliefs).

In addition to quantitative analyses, qualitative analysis of each open-ended question was conducted. Analysis of the open-ended questions employed a thematic content analysis approach (Dixon-Woods, Agarwal, Jones, Young, & Sutton, 2005). Analysis occurred within each question using a data-driven approach in which no a priori categories were established; rather, emerging thematic categories were examined. Through an inductive process of becoming familiar with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, each researcher generated a list of themes for each question through an independent analysis of the total data set (Braun & Clarke, 2006). Pulling from Braun and Clarke’s (2006) definition of a theme, in which “a theme captures something important about the data . . . and represents some level of patterned response or meaning within the data set” (p. 82), potential themes were evaluated based on their relevance to the research questions and the degree to which they were emphasized both across and within participant responses.

Subsequently, through collaborative discussion between both researchers, themes were combined, collapsed, expanded, and renamed until a final list of themes was completely agreed upon. All data were re-examined and applied to the themes to ensure that identified themes appropriately captured the data (Bradley, Curry, & Devers, 2007). Themes were categorized as “primary” or “secondary” consistent with the degree of emphasis placed on the theme by participants. In addition, responses were further categorized and examined based on prior mental health service use; however, no differences in themes emerged when comparing those who had sought services in the past with those who had not sought services. There was high completion rate of the open-ended questionnaire, as 121 of 122 participants provided responses.

Treatment of Missing Data

Prior to beginning the analyses, an examination of missing data was performed. Nine participants had completed less than 80% of the survey. These participants were considered to have withdrawn from the study, as outlined in the consent form, and were automatically removed without further analyses of responses. Patterns of missing data across remaining participants were examined. Missing data were minimal; however, differences in means were examined for each variable between individuals with and without missing data (Tabachnick & Fidell, 2007). No significant differences were found across variables as a result of missing data.

Mental Health Service Use

Among participants, 82% were reportedly aware of the mental health services offered on campus. Frequency of on- and off-campus mental health service use is summarized in Table 2.

OPEN IN VIEWER

Table 2.

Frequency of Mental Health or Counseling Service Access On and Off Campus.

Service access	On campus (%)	Off campus (%)
0 times	82.8	72.0
1-2 times	9.1	9.9
3-4 times	3.3	6.6
5+ times	4.9	11.5

Identification and Comparison of Pertinent Warning and Action Signs

Quantitative results

To address the first two purposes of this study, the frequency of the sample that endorsed each of the symptoms as a warning sign or as an action sign was calculated (see Table 3). All the symptoms, with the exception of “confused thinking” were identified by 50% or more of participants as being a warning sign of mental health concerns. Although several symptoms were endorsed by a large portion of the sample, “prolonged sadness,” “suicidal thoughts,” and “self-harming behaviors” were symptoms that were most frequently identified as a warning sign, with 82%, 81%, and 80% of participants endorsing the symptoms, respectively.

OPEN IN VIEWER

Table 3.

Frequency of Sample Identifying Symptoms as Either a Warning Sign or Action Sign.

Item	% identified as warning sign	% identified as action sign
Confused thinking*	28.7	12.3
Prolonged sadness**	82.0	53.3
Prolonged Irritability**	59.0	32.0
Feeling of extreme highs and lows**	77.0	42.6
Excessive fears**	51.6	32.0
Excessive worries or anxieties**	70.5	45.9
Social withdrawal**	61.5	36.9
Dramatic changes in eating habits**	62.3	29.5
Dramatic changes in sleeping habits**	61.5	31.1
Strong feelings of anger**	64.8	37.7
Delusions	71.3	59.8
Hallucinations	77.9	73.8
Growing inability to cope with daily problems**	73.8	45.9
Growing inability to cope with daily activities **	64.8	41.0
Suicidal thoughts	81.1	75.4
Self-harming behaviors	80.3	80.3
Numerous unexplained physical ailments**	51.6	32.0
Substance abuse	68.0	62.3
None of the above	1.6	2.5

Note. * and ** indicate significant difference in identification as warning and action signs.

*

p < .002. **p < .001.

Contrary to the high endorsement rates for warning signs, only 6 of the 18 symptoms were identified by more than 50% of participants as symptoms warranting action. In particular, only “prolonged sadness,” “delusions,” “hallucinations,” “suicidal thoughts,” “self-harming behaviors,” and “substance abuse” were symptoms identified as most likely to lead to action.

To address the third purpose of this study, McNemar’s test, a non-parametric test for paired dichotomous data, was used for each symptom. These analyses allowed for identification of significant difference between the number of participants endorsing the symptom as a warning sign and the number of participants endorsing the symptom as an action sign. As multiple comparisons were made, Bonferroni’s correction was used to reduce the risk of Type I error. As a result, comparisons were considered statistically significant when p ≤ .0026 (corrected to allow for McNemar testing on the 19 items, that is, p = .05/19). Significant differences between awareness and appraisal for each sign are highlighted in Table 3.

All symptoms, with the exception of “delusions,” “hallucinations,” “suicidal thoughts,” “self-harming behaviors,” and “substance abuse,” were significantly more likely to be identified as a warning sign than as an action sign. In all cases where significant differences were found between warning signs and action signs, participants were less likely to act on the symptom, despite identifying it as a warning sign. When no significant differences were found, endorsement rates were typically high for both warning and action signs, suggesting that a subset of symptoms are concerning enough to lead to both awareness and appraisal of need for action.

Qualitative findings

To ensure that no signs or symptoms of mental health decline were overlooked with the checklist, participants were asked to identify warning signs of mental health decline and signs that would motivate action through open-ended questions. Emerging themes are summarized in Table 4.

OPEN IN VIEWER

Table 4.

Summary of Themes From Narrative Responses.

	Primary themes	Secondary themes
Warning signs	Depression/sadnessSocial isolation/withdrawalSelf-harm/suicidal thoughts	Inability to cope with daily tasks
		Anxiety/stress
		Physical symptoms
		Anger/irritability
		Changes in behaviors/routine
Action signs	Suicidal thoughts“Severe” or “extreme” depression“Severe” or “extreme” anxiety	Violence toward others/aggressionImpairment that cannot be managed on one’s own“Breakdowns”
Barriers to help-seeking	Stigma	Availability/wait times
	Finances	Lack of time
	Insufficient knowledge to access	Concerns about confidentiality
Facilitators to help-seeking	Support from family or friendsUrging and insistence from family or friendsKnowledge of services	Assured confidentialityEase of access

Warning signs

Analysis of responses revealed that participants were able to independently identify several symptoms that might be indicative of a problem. In particular, three primary themes emerged. First, participants shared that depression and sadness warned them of a need for help regarding their mental health. As one participant reported,

Feelings of anger, sadness, emptiness, extending for long periods of time. These kinds of feelings may ultimately lead to depression and if I was concerned about that then professional help may be very useful to me at such a time.

Second, participants reported that thoughts or attempts of self-harm or suicide would also be significant warning signs that assistance with their mental health was necessary. Third, participants identified social isolation and withdrawal from their family and friends to be warning signs of mental health concerns, such as identified by one participant who articulated that “disconnect from family, friends and loved ones” would be cause for concern.

Overarching the specific warning signs identified was the notion that change in routines or behaviors was a central indicator of mental health decline. This is exemplified by one participant who stated, “If myself, or others in my life noticed significant changes to my life regarding moods and actions such as depressed mood or grades falling.”

Action signs

Participants were similarly asked to identify behaviors, problems, or feelings that would have to be occurring for them to actually seek help for their mental health. Although specific symptoms were identified, an overarching theme across responses was the level of severity and duration of symptoms. Participants frequently used terminology such as “severe” or “extreme.” For instance, consistent with this idea of severity, one participant stated,

If I was feeling extremely depressed or anxious for an extended period of time and it was caused by things that usually do not bother me/ or if I did not know the origin of my feelings I would want to seek mental health.

Suicidal ideation was a common theme across participants as a symptom that would cause participants to take action. For instance, one participant wrote “having hateful/suicidal thoughts” would cause him to seek out help. Moreover, participants reported that violence or aggression toward others, impairment that cannot be managed on one’s own (i.e., no social supports, interference with daily life, can no longer cope with symptoms alone), and “breakdowns,” would be alarming enough for them to utilize mental health services.

Factors That Influence Help-Seeking Behaviors

Quantitative findings

Consistent with the fourth purpose of this study, specific factors that predict actual help-seeking were examined. In particular, the influences of attitudes toward psychological help and knowledge of counseling services were examined. To better understand the role of these variables on help-seeking, a hierarchical multiple regression was performed. The predictors of interest were attitudes toward seeking psychological help, as measured by the total score on the ATSPPH-SF, and knowledge of university counseling services, as measured by the total score on the KUQ. Help-seeking behaviors were assessed by using the participants’ reported frequency of access of university counseling services as the dependent variable. This dependent variable was selected for several reasons. First, given the university population, the free services available on campus are likely the most accessible services for individuals seeking help. Second, the specific role of knowledge of this service on use of the service was of interest. Using frequency scores for broader access of mental health services would overlook the specific relationship between knowledge of counseling services and use of counseling services.

Given findings from previous research suggesting that demographic characteristics influence use among university students (Nam et al., 2010; Yorgason et al., 2008), the influences of age and sex were controlled for by entering these variables first (Step 1). Next, ATSPPH-SF and KUQ total scores were entered in Step 2. Finally, frequency of access of counseling services was entered as the dependent variable. Age and sex explained 5% of the variance in service use. After entering ATSPPH-SF and KUQ scores in Step 2, the model explained 28% of the variance in campus counseling service use, F(4, 113) = 10.98, p < .001. The addition of these measures in Step 2 explained an additional 23% of the variance in service use, after controlling for age and sex, R² change = .23, F change (2, 113) = 18.22, p < .001. In the final model, only knowledge of counseling services was statistically significant (β = .48, p < .001). These results suggest that service use is impacted by knowledge of counseling services and attitudes toward health care providers; however, when examined more closely, knowledge of counseling services has the greatest effect on services use, and is the only factor that makes a unique statistically significant contribution to services use. The regression results and standardized regression coefficients (β) are summarized in Table 5.

OPEN IN VIEWER

Table 5.

Results of Regression Examining Influence of Attitudes and Knowledge of Access of Counseling Services on Frequency of Counseling Services Access.

Predictor variable	B	β	t
Model 1
Age	.09	.14	1.56
Sex	.56	.18	1.93
Model 2
Age	.06	.10	1.22
Sex	.45	.14	1.73
ATSPPH-SF	.01	.03	.32
KUQ	.22	.48	5.66*

Note. ATSPPH-SF = Total score on the Attitudes Toward Seeking Professional Psychological Help–Short Form Questionnaire; KUQ = Total score on the Knowledge and Use Questionnaire; R² = .05 for Model 1, p = .06; R²Δ = .23 for Model 2, p < .001; Total R² = .28, p < .001.

*

p < .001.

Relevant Warning and Action Signs

It could be argued that any symptom which indicates a mental health problem should automatically warrant action. Based on this logic, warning signs and action signs should be appraised in a near identical fashion. On the contrary, students perceived warning signs to be quite distinct from action signs. The majority of signs and symptoms from the list with which students were presented were appraised to be a warning sign, but not something that would prompt help-seeking. Similarly, when asked to independently generate warning signs of mental health problems, students were able to identify the more subtle symptoms of mental health concerns, such as impairment in daily activities, social withdrawal, and changes to patterns of behavior. Despite having a good sense of which behaviors/problems are signs of mental health decline, only the most severe of symptoms (i.e., suicidal thoughts, hallucinations) were equally appraised to be both a warning and action sign. When asked to generate action signs that would prompt them to seek help, students placed more emphasis on the severity of symptoms. Although “depression/sadness” was one of the primary warning signs identified in students’ narrative responses, “severe or extreme depression” was one of the primary action signs. In addition, the responses unique to action signs all include some element of blatancy. When symptoms become apparent to others or difficult to hide, students seem more likely to seek help.

This distinction between warning and action signs suggests that university students in the emerging adult age group are reactive rather than proactive in regard to mental health concerns. Students’ appraisal that only the most severe symptoms, such as hallucinations and suicidal thoughts, should prompt action is particularly worrisome. Delaying treatment until symptoms are extreme makes intervention more difficult and recovery less likely (Kisely, Scott, Denney, & Simon, 2006; Rosenkranz & Muller, 2011). Furthermore, as symptoms increase, intentions to seek help tend to decrease, making delayed help-seeking all the more problematic (Rickwood et al., 2005). Students’ tendency to wait until symptoms are extreme before seeking formal help may contribute to the increasingly high prevalence and severity of mental health problems among this population.

Barriers and Facilitators to Help-Seeking

Several factors were identified that might influence responding to one’s awareness and appraisal of symptoms. The importance of knowledge to help-seeking was identified in both the regression results and qualitative findings. Students reported insufficient logistical information (e.g., location, fees) as well as procedural information (e.g., who is eligible, what to expect in treatment) as barriers to taking action.

Although the quantitative results indicated that beliefs/attitudes about help-seeking are not predictive of help-seeking, the qualitative responses suggested that certain beliefs/attitudes do play a role. This discrepancy is likely due to the way in which beliefs and attitudes were queried in the two methods. The specific attitudes/beliefs presented in the questionnaire were not necessarily the same attitudes/beliefs that emerged in the open-ended qualitative responses.

Stigma is one such attitude/belief that students indicated as a particularly important barrier to seeking services. In fact, students indicated that increasing the secrecy of help-seeking would help limit stigma, and thus, would facilitate seeking services. Stigma as a barrier to help-seeking has been widely discussed in past research (Eisenberg, Downs, Golberstein, & Zivin, 2009; Livingston & Boyd, 2010), and widely targeted in mental health initiatives and awareness campaigns (Stuart, 2008). The current results suggest that the many forms of stigma (e.g., shame, fear of judgment from others, self-stigmatizing attitudes) remain a major impediment to the receipt of mental health services. Students recommended several logistical changes, such as online booking of appointments and housing campus services in an allied health center rather than in a separate distinct area, that could be implemented to delimit anticipated stigma.

Students also identified support and insistence from family or friends as major facilitators to accessing mental health services on campus. This finding is consistent with past research, which has identified social support or encouragement from others as a primary facilitator among young people seeking help (Gulliver et al., 2010).

Conceptual Model of MHL and Help-Seeking

Although models of help-seeking have been previously proposed (e.g., Rickwood et al., 2005), they do not sufficiently capture the distinction between awareness and appraisal identified in the current study. Results suggest that awareness and appraisal are distinct processes, and that they should be recognized as such when considering emerging adults’ understanding of mental health symptoms. To better capture this distinction, a conceptual model, based on integration of the qualitative and quantitative findings, is proposed in Figure 1. This model highlights the distinction between awareness and appraisal, as well as the progression toward help-seeking that is impacted by barriers and facilitators identified through our results. Following the same process from awareness to action, the quantitative and qualitative findings allow for proposed recommendations to improve MHL and access to services.

OPEN IN VIEWER

Figure 1.

Conceptual model of mental health literacy and help-seeking in university students.

At the stage of awareness, quantitative and qualitative results suggest that students have a good sense of which behaviors or symptoms are “warning signs” of a mental health concern. A subset of more subtle signs of mental health decline, such as irritability and changes to sleep and eating, were less frequently identified as warning signs. Increased education around the less obvious, yet equally pertinent signs could improve awareness across students.

Both the quantitative and qualitative results support the notion that emerging adults appraise symptoms as problematic when they are severe or obvious markers of mental health problems (e.g., suicidal ideation, hallucinations). Similarly, when they perceive themselves as at risk for harming themselves or others, symptoms are appraised as warranting action. Paralleling these findings, we propose that education of the consequences of symptoms, and the benefits of early intervention, even for less obvious symptoms may result in increased help-seeking. Professionals and educators should consider promoting awareness about the implications of failing to seek help in response to warning signs. Moreover, instruction on accurately appraising mental health symptoms should be incorporated into the education system curriculum. If instruction was provided during school-aged years, people may be better prepared to take health-related action in emerging adulthood.

Participants identified several barriers and facilitators that might intervene between appraisal and action. As discussed, stigma, lack of a knowledge, and financial/time constraints were detrimental to access, while support, as well as a guarantee of privacy supported help-seeking. Several recommendations emerged from these findings, as well as directly from the participants’ recommendations. Continued work to decrease stigma on campuses remains needed. In addition, mental health professionals and educators should also consider promoting logistical and procedural knowledge of formal help-seeking. Given that participants suggested that parents and peers played a big role in their help-seeking behaviors, education geared toward providing skills in assisting a loved one in seeking services could be useful. Finally, implementing changes that allow for increased privacy of clients accessing service may encourage students to seek help.

Reflection on Cultural/Institutional Context

The participants in the current study consisted of university students in a mid-sized western Canadian university and the results reflect on the general university culture in several ways. First, participants gave the impression that the university atmosphere is not conducive to, or supportive of, formal help-seeking for mental health concerns. This is in line with past research indicating that components of campus culture (e.g., lack of marketing of mental health services, over-valuation of physical health, and under-valuation of mental health) can dissuade students from help-seeking (Chen, 2013). Second, participants strongly asserted that mental health stigma is still very much alive within the Canadian university culture. These findings are consistent with past research proposing that mental health stigma is particularly widespread within university settings (Golberstein, Eisenberg, & Gollust, 2008). As peer influence is highly important to identity formation and decision making in emerging adulthood (Seiffge-Krenke & Beyers, 2007), perceptions of peer beliefs and the experience of social pressures are integral to help-seeking within this culture (Goodman, Peterson-Badali, & Henderson, 2011).

Third, participants indicated that mental health knowledge is lacking in this institutional context. As those emerging adults who are attending university are often assumed to be more educated and advantaged (i.e., higher socioeconomic status) than their non-university student peers, this finding is particularly disconcerting. Nevertheless, MHL would have to be examined in a non-university population of emerging adults in order to extrapolate the current findings to diverse cultural, socioeconomic, and institutional contexts.

Limitations

Several limitations to this study should be acknowledged. To begin, the majority of the sample was female and Caucasian. As noted, the sampled population is similar on demographic variables to institutions across Canada; however, the population sample includes a higher rate of aboriginals and lower overall rate of visible minorities. This demographic distribution may have uniquely influenced responses in the current study. The small number of participants within minority groups did not allow comparisons between cultural group. Consequently, certain cultural influences on help-seeking (e.g., underutilization of formal supports among minority students; Kuo, Kwantes, Towson, & Nanson, 2006) may not be represented in the current results.

Similarly, our sample is most representative of a middle-class socioeconomic status and may not generalize to emerging adults in alternate socioeconomic strata. Moreover, data collection was limited to one university, and the majority of students in the sample were enrolled in kinesiology, psychology, and nursing. As a result of the sample limitations, our findings may not be generalizable across different universities and among all university students. The purpose of our study was to gain an enhanced understanding of the MHL of emerging adults in the university setting; therefore, further research is needed to examine whether there is variation in MHL across different universities and students in diverse disciplines.

In addition, the checklist used to determine students’ perspectives on warning signs versus action signs was simplistic, in that it did not provide contextual details about their decision making. For example, the checklist did not allow for participants to indicate the degree of symptoms that would warrant action, or certain situations in which symptoms would warrant action. Another limitation of the checklist method is the potential for participants to interpret symptoms differently. For example, one student’s definition of “substance abuse” may not be the same as the next student’s definition.