Introduction

Mental disorders account for a significant disease burden for college students.¹ Approximately 1 in 3 college freshmen reports having mental health problems in the past year.² Recent reports^3-5 have highlighted concerns regarding increased frequency and severity of major depression, anxiety, and personality disorders among a broad group of college students, including graduate students. Up to 33% of college students report depression significant enough to limit function, and more than 37% report overwhelming anxiety. These data reflect a dramatic increase in the portion of the campus population affected by a mental health problem compared with a decade ago.⁶

Growing demand for on-campus mental health services has regularly surpassed the capacity of campus counseling centers, which face resource limitations. Therefore, some students who seek treatment are not able get their needs met. In addition, a significant portion of those students struggling with mental health problems do not seek treatment at all.⁶ Reasons for not seeking treatment may include perception that treatment is not needed, lack of time, preference for self-management, and stigma.^7,8 College students who do seek treatment face further challenges, including high rates of medication nonadherence. Up to 55% of students are nonadherent with their antidepressant medication.¹² Several factors contribute to medication nonadherence for people of all ages¹³; however, for college students with mental illnesses, factors such as lack of consistent day-to-day routine, lack of acceptance of mental illness, and stigma may be especially prominent.¹⁴

Innovative, engaging, and resource-efficient approaches that minimize barriers to treatment-seeking and convey messages of a supportive, nonstigmatizing campus community are needed to better address mental health needs among college students. Group therapies, in addition to individual approaches, have been purported as an approach that optimizes scarce resources while addressing many of the concerns of college students with mental illnesses, such as social isolation, interpersonal relationships, low illness acceptance, and self-esteem.¹⁵

Photovoice (PV) is a participatory-action research methodology involving photography and reflection. It involves creation of a visual reality of a person's lived experience with a particular topic. With PV, research participants are the experts on a topic and communicate their expertise via capturing photographs and engaging in deep reflection upon them, typically in individual as well as group settings. This leads to creation of meaning, collective voice about a topic, and a venue for communication to others. Despite its typical employment as a research modality, PV offers potential advantages as a therapeutic group modality.¹⁶ By drawing students in through the ability to focus on creativity and reflection, a PV approach may theoretically be more enticing and less stigmatized than traditional group therapy. It also often leads to a sense of empowerment and group support due to the shared nature of discussing experiences related to the topic, in this case, living with a mental illness.^17,18 The authors previously have undertaken PV studies^14,19,20 in partnership with college students living with mental illness and psychotropic medications, using a qualitative approach to better understand the experience of living with a mental illness and medication on a college campus. A significant takeaway from this work is that participating in a PV study was beneficial and possibly therapeutic to participants' mental health through learning, acceptance of illness and medications, and peer support. Participants have anecdotally reported that their participation in a PV study increased their medication adherence through enhanced daily structure and reduced negative perceptions about medications.²⁰

Photovoice has not previously been employed as a therapeutic intervention or adherence promotion strategy in a health care setting utilizing quantitative measurements of impact. Given the above-described benefits, we hypothesized that use of PV as an intervention for students with mental illness would result in greater improvements in mood and anxiety and would contribute to increased medication adherence compared to usual care. We describe a pilot study to explore whether using PV in this manner is more beneficial for depression symptoms, anxiety symptoms, or medication adherence among college students as compared with usual group therapy, which served as an active control.

Method

This study was conducted over an 8-week period at the campus counseling center at a midsized public academic institution in the Midwest. North Dakota State University Institutional Review Board reviewed and approved all study procedures. Participants were recruited from the student body via flyers posted throughout campus. Eligible students were those at least 18 years of age, diagnosed with depression and/or anxiety, and taking at least 1 medication to treat the same. Interested individuals who met these criteria were scheduled for an initial individual session with 1 of the 2 licensed psychologists on the counseling center staff, who confirmed study eligibility and performed a safety assessment to ensure the individual was not at risk of harm to self or others or in need of referral to a higher level of care. Each eligible participant was informed of study procedures and risks and benefits and offered the opportunity to use a pseudonym for added protection of their identity if they wished. Twenty participants who gave their informed consent to participate and have their photos and reflections published anonymously were randomized in a 1:1 fashion to PV or control (Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

All participants, regardless of group assignment, were scheduled for a 1-hour individual session every other week (n = 3 individual sessions) as well as a 1-hour group session every other week (on alternating weeks opposite of individual sessions, n = 3 group sessions). All participants were given information about the importance of protecting the privacy of other participants and encouraged not to discuss information about other participants outside of the counseling center environment.

Photovoice participants were directed to use a smartphone camera to capture photos that, for them, represented the theme of living with my mental illness. They were given a journal to record their thoughts about each photo (using the SHOWED technique²¹ as a guide; Table 1) and were encouraged to use their journals ad libitum as frequently as they liked. Photovoice participants e-mailed 1 to 5 photos to the psychologist every other week in advance of their individual photo reflection sessions. During individual sessions, the provider asked each SHOWED question about each photo and encouraged the participant to elaborate upon associated feelings or illness-related features of each photo. At the conclusion of each individual session, participants chose which of their photos they wanted to discuss during the group reflection session the following week. During group reflection sessions, participants took turns discussing their chosen photos (which were displayed on a video screen), using their journal entries as a guide. An open and interactive space was created by the psychologists, who encouraged dialogue about each photo from all participants in attendance.

TABLE 1 The SHOWED technique and explanation^a

View Fullscreen

Participants assigned to the control were given a journal and asked to reflect, weekly at a minimum, on the topic of living with my mental illness. Individual sessions lasted approximately 50 minutes and were focused on processing themes from journal entries written by participants throughout the week. The psychologists utilized person-centered approaches of reflection and processing to facilitate deeper meaning and illuminate emergent themes associated with participants' journaling. Group sessions met every other week and were process groups focusing on here-and-now interactions in addition to illuminating group themes of shared experiences.²²

All participants completed prestudy and poststudy evaluations including the Beck Depression Inventory (BDI-II),²³ Burns Anxiety Inventory (BURNS),²⁴ and Medication Adherence Rating Scale (MARS).²⁵ The BDI-II and BURNS instruments were chosen as they were already a standard part of all patients' first-visit encounter with the campus counseling center. Lower scores for each of these instruments indicate symptom improvement. The MARS was chosen due to its simplicity as well as its prior use and validation in psychiatric patient populations.^26-28 Higher scores are associated with higher likelihood of medication adherence. An exit questionnaire was also completed. Descriptive statistics and analyses of variance controlling for baseline measures were performed with quantitative data to assess for differences between groups from preassessments to postassessments based on completion analysis. Statistical significance was defined as P ≤ .05. An informal summary analysis of exit questionnaire responses was conducted with identification of concepts based on frequency.

Results

Ten participants completed the study and had prestudy and poststudy data for analysis: 4 in PV and 6 in the control. All noncompleting participants were lost to follow up as they did not indicate preference to drop out of the study but did not show for study meetings nor notify staff of reasons for absence. No study-related adverse events were reported. Figure 2 provides an example of material elicited and discussed by participants in the PV and control groups, respectively. Actual participant names were replaced with pseudonyms by the researchers.

View Figure

• Download as Powerpoint
• Download Figure

At baseline, PV participants had significantly higher BDI scores (P = .018), indicating more severe depression. Baseline BURNS and MARS scores were similar between groups (P = .29 and .84, respectively). At 8 weeks, the reduction in BURNS score in the PV group was significantly larger than the reduction in the control group (–8.5 in the PV condition vs –2.6 for controls; P = .049). There was not a difference in the mean change of the BDI scores (–6.7 in the PV condition vs 0.2 among controls; P = .26) between groups at 8 weeks despite an overall mean reduction in the PV group. In both groups, MARS scores decreased from baseline to 8 weeks with no statistical difference between groups in the change in these scores (1 in the PV condition vs 0.83 in the control; P = .16; Table 2).

TABLE 2 Depression, anxiety, and medication adherence outcome measures

View Fullscreen

All completing participants in both conditions answered affirmatively to the question Did you find the study process beneficial for your mental health?. When asked to describe which part(s) they felt were therapeutic, PV participants tended to comment about photo-taking and reflection, which helped to make my feelings more concrete, and control participants commented about the support found in the group environment, which made me feel like I am not alone. When asked to describe their least favorite aspect of the study, PV participants tended to cite preference for longer study participation, and control participants cited lack of focus/productivity of group discussions.

Discussion

Photovoice participants had significantly greater reductions in anxiety scores than those in the control group. This suggests the possibility that PV may be an effective approach for reducing anxiety symptoms when it is employed with college students for several possible reasons. First, it is possible that visual representation of concepts and feelings during the PV process allowed for greater emotional processing, leading to reductions in anxiety. Because of its incorporation of visual images, PV has been touted to lead to deeper connections with emotions and to “tap into a different part of human consciousness than do words-alone” approaches.^29(p23) This individual reflection process allows for significant self-examination and deep processing of experience.³⁰ Other PV work has shown enhanced emotional expressions, a greater healing impact,³¹ and safer space for sharing³² than traditional approaches that rely on nonvisual means of relaying information. An extension of this possibility, which was suggested to us by a PV participant, is that use of images allows the focus of group discussion to be about a picture rather than on the individual, reducing anxiety and vulnerability associated with talking about symptoms and/or being in a group setting. This is consistent with research showing that reduced self-focused attention by increasing attention to things external to oneself can result in improvements in social anxiety symptoms.^33,34 Hence, perhaps the PV participants experienced greater reductions in anxiety than usual care participants due to the ability to transfer their attention to their photographs and less on themselves. Unfortunately, we did not measure self-focused attention nor collect disease-specific diagnostic information about the presence or absence of social anxiety disorder, so these ideas can only be hypothesis generating. Additionally, the PV process often results in feedback from other participants, rhetorical questions on issues and experiences, and the formation of a collective stance on issues significant to participants based on discussion of how they view themselves and others.^35,36 A sense of community forged by common experience and “opportunities to positively reframe meanings of their mental illness, medications, and lives” may also flow from group reflection sessions,^18(p121) which could further lend itself to reduced anxiety. It also is possible that use of an image as the focus point during group sessions allowed for more organization and productivity. This is supported by the qualitative finding that control participants felt group discussions were nonproductive, and PV participants did not. Finally, it is possible that PV participants were more likely to benefit regardless of the intervention employed because of their greater severity of baseline depression. However, lack of greater improvement in depression symptoms in the PV group compared with controls seems to contradict this notion. Further study is needed with larger samples to explore these ideas.

Completing participants reported poorer medication adherence at 8 weeks than at baseline with no difference between the PV and control groups. Participants in our sample may have been overreporting their adherence at baseline, which has been previously reported among adolescents and young adults with psychiatric illness.³⁷ Perhaps increased trust in therapists ensued over time during the 8-week study, leading to greater participant honesty in adherence reporting at study end. It is also possible that, as participants learned more about their illness and aspects of their treatment through both the PV and usual care programming, they gained greater insight into their medication-taking behaviors, realizing they were more nonadherent than they had previously identified. It is also possible that, as participants engaged in group and individual nonpharmacologic therapies and as their symptoms improved, they felt their medications were of lower importance, leading to poorer adherence with them. This would be contradictory to our previous anecdotal findings associated with PV participation,²⁰ but it is not out of the question. Unfortunately, we did not collect information about reasons for nonadherence or employ an objective, nonself-report measure of medication adherence to further explore or validate these findings. Future studies should seek to examine these potential contributing factors.

The high participant attrition rate and low sample size are significant limitations of this work. Initially, participants were enthused to enroll and share their experience. However, nearly half of the participants did not return after their initial enrollment session. In our experience, attrition of this magnitude is common regardless of the type of mental health service being provided as a semester progresses and students become more stressed after the midterm period, often leading to lack of follow-through. In addition, attrition rates for group mental health services are often higher than individual sessions.¹⁵ Although we had hoped that the PV activity might be an approach that would overcome this barrier to group mental health services, this did not seem to be the case. We did not collect information about individual sessions that were also a part of both PV and usual care. Therefore, we cannot comment on any differences between individual sessions nor how they may have impacted the overall outcomes.

Conclusion

Photovoice may hold promise as a resource-efficient and effective intervention to reduce anxiety among college students with a mental illness, but it did not seem to promote greater retention of students in the campus counseling center, and it may have led to greater medication nonadherence. Further study with larger sample sizes is needed to replicate these findings.