Online Support Groups for Depression

This mixed-methods study aimed to explore the initial process of engagement with an online support group (OSG) for depression. Fifteen British National Health Service patients experiencing depression who had not previously used an OSG for depression were offered facilitated access to an existing peer-to-peer OSG for 10 weeks. Pre- and post-measures of depression, social support, and self-stigma were taken in addition to a weekly measure of OSG usage. A follow-up qualitative interview was conducted with a subsample of nine participants. Depression and self-stigma reduced over the 10-week period, but perceived social support did not change. There was no evidence of adverse outcomes. Perceived benefits of OSG participation included connection to others, normalization of depression, and stigma reduction. However, engagement with the OSG was generally low. Barriers included concerns over causing harm to others or being harmed oneself, feeling different from others in the group, and fears of being judged by others. OSGs may potentially reduce depressive symptoms and perceived self-stigma. However, considerable barriers may hinder people with depression from engaging with OSGs. Further work is needed to determine who will benefit most from participating in OSGs for depression and how best to facilitate engagement.


OSG
The OSG selected for this study was PsychCentral.com. This was a U.S.-based peer-to-peer OSG which was "standalone," that is, not linked to another intervention. It was chosen because it was a high-traffic website that had a constructive atmosphere and was moderated by a team supervised by a clinical psychologist. Permission was granted by the website's owner for this research to take place.

Design and Procedure
This was a mixed-methods study, with quantitative measures taken before, during, and after the intervention, in addition to post-intervention qualitative interviews.
Participants were emailed an information sheet. If they consented to take part, they were emailed a link to the online questionnaire, on completion of which they were directed to the PsychCentral depression forum and given guidance on how to register. They were encouraged to write a post to introduce themselves to the group, and asked to engage with the site as they wished over the next 10 weeks. At the end of each week, they were emailed a link to the OSG Usage Report. At the end of the 10 weeks, they were emailed a link to the final questionnaire and were invited to participate in a face-to-face interview. Participants were encouraged to contact the researcher by phone or email if they had difficulties using the OSG or if they wished to discuss their experiences.
Participants whose responses to the initial questionnaires suggested that they were severely depressed or a risk to themselves or others, i.e., scores >0.3 on the risk scale of the Clinical Outcomes in Routine Evaluation [CORE] or >1 on Item 9 of the Patient Health Questionnaire  ("Thoughts that you would be better off dead, or of hurting yourself in some way") were emailed with the suggestion that they seek additional help, in line with the study risk protocol. This procedure was followed for five participants.

Measures
Pre-post measures. The following pre-post measures were used: CORE-Outcome Measure (CORE-OM; Evans et al., 2000), a 34-item questionnaire assessing general psychological distress; PHQ-9 (Kroenke, Spitzer, & Williams, 2001), a nine-item depression scale; the Medical Outcomes Study Social Support Survey (MOSSSS; Sherbourne & Stewart, 1991), a 19-item social support measure; and Depression Self-Stigma Scale (DSSS; Kanter, Rusch, & Brondino, 2008), two subscales of which were used: "General Self-Stigma" and "Treatment Stigma." Measures taken at end only. OSG Questionnaire (OSGQ; Chang, Yeh, & Krumboltz, 2001) is a nine-item scale measuring overall satisfaction with an OSG. Overall usage: Postings made to the group were identified by searching for participants' usernames. Each participant's total number of posts over the 10-week study period was recorded.
Measure taken weekly throughout the 10-week period. OSG Usage Report, designed for the current study, asked how often participants had logged on to the group during the Note. PHQ = Patient Health Questionnaire; OSG = online support group. *****severe depression. ****moderately severe depression. ***moderate depression. **mild depression. *no depression.
week, how long they had used it for, and how many times they had posted. It also had an open-ended question about experiences of the group during that week.

Interview
The semi-structured interview schedule covered three main areas: (a) depression and face-to-face social support, (b) engaging with the OSG, and (c) view of, and interactions with, others in the group. Interviews lasted about an hour and were digitally recorded. Participants were offered £10 plus expenses. The data were analyzed using thematic analysis (Braun & Clarke, 2006). For each transcript, units of meaning were identified and collated into a tentative set of themes. Themes were compared across all the transcripts and organized into general themes and sub-themes, which were checked against the original transcripts and modified accordingly. During the process of analysis, the two authors discussed the emerging thematic structure and attempted to resolve any areas of uncertainty. The interview data were also triangulated with week-by-week written feedback from the OSG usage report.

Researchers' Perspective
Qualitative research guidelines (e.g., Barker & Pistrang, 2005) recommend that researchers disclose relevant background. The first author was a female clinical psychology graduate student with a positive but limited experience of using an OSG in a personal capacity. Her initial assumptions were that OSGs can potentially be helpful, but that there might be barriers to engagement. The second author was a male clinical psychologist who had conducted research on support interventions. The researchers attempted to "bracket" their assumptions during the course of the study to be aware of how they might influence their interpretations of the data.

Outcome, Usage, and Satisfaction
Eleven of the 15 participants completed all pre-and postmeasures. Missing data were handled on an intent-to-treat basis using last-observation-carried-forward.
Depression and self-stigma reduced over the 10 weeks ( Table 2). In terms of reliable change (Jacobson & Truax, 1991), three people improved, 12 showed no change, and no one deteriorated. There was no change in perceived social support or the total CORE-OM.
Usage was generally low. The median percentage of weeks in which participants logged in was 30% (range = 0-80%). The median number of posts over the 10 weeks was 2 (range = 0-13). Eight (55%) participants completed the OSGQ; their mean score was 27.4 (SD = 10.7, range = 11-39), indicating a moderate level of satisfaction.

Qualitative Interviews
Context. Most participants interviewed described experiencing severe recurrent depression, which had affected their ability to work and had in several cases led to suicide attempts. Only one was experiencing a first episode. Several also experienced other mental health problems such as anxiety, posttraumatic stress disorder (PTSD), and trichotillomania. None had had any experience of OSGs for depression. One described a positive experience with an OSG for trichotillomania, and others had used non-mental health online communities. About half had been hopeful about what the OSG might offer them, for instance, looking forward to sharing ideas with others and finding people with similar experiences. The rest were less optimistic, reporting that they had not expected to benefit from the OSG; they appeared to be motivated by curiosity or a desire to further psychological research.
Thematic analysis. The themes generated from the qualitative interviews and weekly feedback were grouped into two domains (see Table 3): perceived benefits and barriers to engagement.

Benefits
I'm not alone in this. Participants valued the opportunity to make contact with others with experience of depression. Five said that they felt less socially isolated as a result of using the OSG and felt an emotional connection with other users, promoted by others' openness about their experiences of depression. Three felt that the site had helped to make depression  (Evans et al., 2000); PHQ = Patient Health Questionnaire (Kroenke, Spitzer, & Williams, 2001); MOSSSS = The Medical Outcomes Study Social Support Survey (Sherbourne & Stewart, 1991); DSSS = Depression Self-Stigma Scale (General Self-Stigma and Treatment subscales; Kanter et al., 2008). *p < .05. (one-tailed) seem like a common, universal experience, which they found reassuring. This was linked to the size and the international nature of the OSG.
The value of talking to strangers online. Some participants felt more comfortable talking about depression to strangers online than to their friends and family. They attributed the safe, non-judgmental nature of the OSG to the fact that users had personal experience of depression. Three felt that the anonymity enabled them to disclose more freely. Two spoke about the helpfulness of talking with people who did not have any preconceived ideas about them.
Supporting and being supported. Participants felt supported from being in contact with the group, even without any actual communication. Three said that knowing that it was there, even if they were not using it, was helpful. They appreciated being able to access the group immediately and at any time, and some contrasted this with their experience of waiting a considerable time for therapy. Two talked about wanting to help others as a result of reading their posts and the sense of reward that comes from offering help, and two others talked about the support they had got from reading about others' experiences. One felt that seeing other people going through cycles of low moods reminded her that she would once again come through her own current low patch, and another spoke about the way in which he had been inspired to try different ways of coping.
Could be useful for others or for me at a different time. Even if they had not found the site helpful, participants without exception expressed the view that the site could potentially be a useful resource for people with depression, and commented that users appeared to be supporting each other and seemed to benefit. Four felt that they had been severely depressed at the time of the research: One spoke about finding it difficult to even motivate himself to log on. Two felt that the group may have been more useful to them if they were less depressed, although others said that they would be more likely to use the site if they were having a bad time.

Barriers
Fear of negative impact on others. Participants' concern for the other people involved with the group was striking, and some cited the potential negative consequences that their involvement could have for others as one of the reasons that they did not contribute more. In some cases, the concern that they would not be able to help others held participants back from seeking support, as they felt strongly that their relationship with the site should be reciprocal. Four participants said that they sensed that others in the group were vulnerable and worried that their contributions could make people feel worse: one said that hearing about her long history with depression might be discouraging for someone who was newly diagnosed. Three worried that they might get into conflict with others in the group and had held back from posting because they were afraid that they would express anger or criticism.  worse, particularly if they were not feeling especially depressed. Three felt that others on the site could not help them, and one who posted messages did not find the replies helpful. One expressed the view that when people were depressed they were too withdrawn to interact and when they were well they were afraid of interacting with other depressed people for fear of being pulled back down into depression. Two participants described experiencing social anxiety in face-to-face situations, and worried about how they would come across in the group and whether others might be critical of them.

Fear
Not my kind of people. Participants mentioned various aspects of how others in the group communicated which discouraged them from becoming involved. The site being based in the United States made it hard for some participants to relate to the other users: they mentioned differences in language, social contexts, and medication names as barriers to understanding. Three felt that it was hard to gauge whether others were being genuine, making it difficult to feel comfortable sharing thoughts and feelings. Three said that the level of discussion on the site was too shallow and people's posts could be experienced as patronizing.
It was hard to talk (to strangers) online. Participants said that although they wanted to express themselves, it was difficult to do so in the OSG, because they felt embarrassed and vulnerable. Four spoke about the difficulty in establishing rapport online (sometimes attributed to factors such as the asynchronous nature of responses and the absence of body language and eye contact) which would have made them feel comfortable enough to reveal personal information. Two said that both having time to think about what they were about to say, and not having the pressure of having to produce something for a listener in front of them, were the inhibiting factors. Four felt that people who really knew them were in a much better position to help them through their depression than strangers online. In interactions with people who knew them well, these participants valued that they did not have to explain themselves and their history, that people could tell how they were feeling, and that they could remind them what they were like when they were not depressed, which instilled a sense of hope. Comments from strangers, although supportive, were experienced by some as "meaningless" whereas real friends' support had more impact.

Discussion
This mixed-methods study introduced a small sample of 15 people with depression, recruited through the British National Health Service, to a depression OSG and tracked their engagement with it. Most had moderate to severe levels of depression at baseline. Over the 10-week period of the study there were modest reductions in depression and perceived stigma, with no change in perceived social support. Overall, there were low levels of engagement with the OSG. The reduction in depression is consistent with the results of many other studies ). However, in the absence of a control group, it is not possible to attribute a causal role to OSG participation, particularly because engagement was low.
The low levels of activity may reflect the way in which people naturally engage with OSGs, that is initially spending a long time viewing the site without posting ("lurking") before joining the discussion. Lieberman (2007) found that 74% of participants in breast cancer OSGs lurked before they posted, with half of these doing so for 2 to 8 weeks. To observe the engagement process as it naturally occurs, a longer study period would be needed.
Participants not only described several benefits of taking part in the OSG but also emphasized factors that had hindered their engagement. In contrast to previous studies that have recruited participants from among existing users of OSGs (who are likely to be at least reasonably satisfied with such groups), the present study recruited people who had never before used a depression OSG, tracking their engagement and allowing the voices of those who find such groups less helpful to be heard.
Participants showed a striking concern for the well-being of others in the group and wanted to help them, but several felt unable to do so, often because their own struggles left them without resources. Sometimes this inhibited participants from using the group, as they felt that their relationship with it should be reciprocal. This raises the question of what could be done to help users access support without feeling a pressure to reciprocate immediately.
The main barriers to participation were fears of having a negative impact on others in the group, or the group's having a negative impact on oneself. Participants also worried about their depression being made worse by their interaction with the group and reported avoiding reading posts with negative content. This may reflect a more general tendency toward emotional avoidance among participants which may have affected their low engagement with the group (Cameron et al., 2005). In addition, negative beliefs about seeking help (e.g., that it is a sign of weakness) may also have inhibited participation.
These concerns are understandable in the light of the well-documented "cognitive triad" of negative beliefs about the self, the world, and the future that characterizes depressive thinking (Beck, Rush, Shaw, & Emery, 1979). Levine and Moreland (2006) suggested that in general people try to join groups that are maximally rewarding and minimally costly. Depressive thinking may have led participants to underestimate the rewards and overestimate the costs of joining the OSG. As they did not engage, there was no opportunity for them to gain evidence to disconfirm their negative preconceptions.

Limitations
This was a small-sample, mixed-methods study. The quantitative arm had low statistical power; the measures were included to give an indication of participants' overall outcomes. The sample was a convenience one, but it did have the advantages of having a range of ages and ethnic backgrounds, and of giving people access to an OSG for the first time.
Limiting participants' choice to a single OSG was problematic, as there is evidence that when people choose a group that "fits" them best, they are more likely to engage with it (Lieberman, 2004(Lieberman, , 2007. Moreover, the choice of a predominantly North American group was with hindsight unfortunate, as cultural differences were often mentioned as a barrier to participation. Thus, it is unclear whether the low levels of engagement and barriers to participation found in this study reflect engagement with OSGs for depression in general or with this OSG in particular. When the study started, there was no comparable U.K.-based OSG in terms of moderation, positive group climate, and high volume of traffic.

Implications
There was no evidence of negative effects from the OSG, and some participants reported benefits, which suggested that clinicians need not be concerned by their clients' involvement with OSGs. It appears that prospective members who anticipate negative effects simply do not engage with such groups. Many of the barriers to participation relate to people's fears of what might happen were they to take part, rather than to actual experiences of participation. It is unclear to what extent these fears are related to cognitive aspects of depression. Depressed clients may benefit from some therapeutic work to help them test the accuracy of their negative predictions in the initial stages of engaging with such a group.
Some of the barriers identified in this study are related to the negative thinking style, withdrawal and lack of motivation characteristic of depression. It would be valuable to design an intervention to help participants overcome these initial barriers and evaluate if this had any impact on engagement. If barriers to active participation can be minimized, people with depression may then be able to derive more benefit from online support.