Homelessness, a major social issue, is strongly linked with problematic alcohol and drug use. Alcohol and drug dependence are the most common mental health issues among homeless populations in high-income countries, with rates of alcohol use disorder ranging from 5.5 to 71.7% and drug use disorder rates varying from 0 to 72.1% (Gutwinski et al., 2021). Substance use may precipitate entry into homeless by depleting an individual’s social and economic resources (Johnson et al., 1997) or develop later as a coping mechanism to manage the stressors of homelessness (Johnson & Chamberlain, 2008).

There is an established link between social anxiety disorder (SAD) and drug and alcohol use. For example, Grant et al. (2005) found in a nationally representative American sample that 48.2% of people with lifetime SAD also had an alcohol use disorder and 22.3% a drug use disorder. Research has indicated that alcohol and drugs may be used to self-medicate and relieve anxious symptoms of social anxiety (Carrigan & Randall, 2003; Thomas et al., 2003). The self-medication hypothesis proposes that substance use is a maladaptive coping strategy which is intended to mitigate anxiety symptoms and the life difficulties that they cause but, with repeated use, leads to substance dependence instead, which can in turn exacerbate anxiety symptoms (Chutuape & de Wit, 1995; Turner et al., 2018). Additionally, social anxiety and stress may elicit cravings for substance use, thereby leading to relapse in otherwise abstinent individuals (Ghiţă et al., 2019).

Treatment of social anxiety comorbid with substance use is likely to prove challenging, especially when compounded by homelessness. Social anxiety may interfere with treatment of substance use disorders by impeding engagement and retention (Randall et al., 2001). For example, investigations of Alcoholics Anonymous (AA) have found that socially anxious members can struggle with the program’s social elements, making them less likely to speak during group therapy (Book et al., 2009). Furthermore, adults with SAD tend to report that they do not feel as good after meetings, feel less integrated into the group, and feel more shame as part of 12-step programs (Terra et al., 2006). Boddapati et al. (2014) similarly found that higher social anxiety predicted lower likelihood of remaining in a sober living house for more than 6 months, which they attributed to difficulties managing social contact required as part of residential treatment. Conversely, substance use problems may adversely impact psychological social anxiety interventions. Using substances to manage anxiety may constitute a safety behaviour which interferes with behavioural experiments and exposure tasks by depriving individuals of opportunities to experience positive social interactions without the influence of substances (Baillie & Sannibale, 2007; Rapee & Heimberg, 1997).

However, social anxiety can potentially improve engagement in substance use treatment. For example, SAD was associated with higher levels of peer-helping (e.g. guiding other patients through treatment) among adolescents in a residential 12-step program (Pagano et al., 2015). Similarly, among 128 men in residential drug and alcohol treatment, Oakland and McChargue (2014) found that alcohol users high in social anxiety stayed in treatment longer than low social anxiety users. The researchers proposed that exposure to social situations inherent to treatment improved these residents’ social anxiety symptoms alongside their alcohol recovery.

Despite the frequent co-occurrence of SAD and substance use, research on psychotherapeutic interventions for this comorbidity is scarce (Baillie et al., 2013) with only three randomised control trials (RCT) conducted to date. Stapinski et al. (2020) allocated participants with clinically diagnosed SAD and a diagnosis or sub-clinical symptoms of alcohol use disorder to receive either integrated cognitive behaviour therapy (CBT), which targeted both social anxiety and substance use, or CBT focused only on alcohol use for 10 weeks. At 6-month follow-up, both groups exhibited significant improvements in social anxiety and alcohol use symptoms and quality of life. Notably, the integrated treatment group demonstrated greater reduction of social anxiety symptoms, indicating that concurrent treatment of these disorders may yield superior results. Randall et al. (2001) compared a 12-week CBT program for comorbid alcohol dependence and SAD with a similar program for alcohol dependence only and found reductions in symptoms of both disorders across both groups at 3-month follow-up. Another RCT by Schadé et al. (2005) compared group format relapse prevention, with or without additional individual CBT for social anxiety. Additional social anxiety therapy significantly reduced anxiety and avoidance but did not impact alcohol relapse.

Despite limited empirical research on psychological treatments for comorbid social anxiety and substance use, some key treatment considerations and recommendations for this population recur throughout the literature. Turner et al. (2018) proposed that concurrent or integrated treatment is the gold standard for comorbid social anxiety and substance use given the frequent co-occurrence and mutually reinforcing nature of these difficulties. Bruce et al. (2005) indicated that SAD typically does not remit without addiction intervention, and Kushner et al. (2005) found that addiction patients are more likely to relapse if their SAD remained untreated during addiction intervention. Despite the apparent need for integrated treatment, patients with this comorbidity report struggling to find clinicians who can address both their addiction and anxiety-related needs (Melchior et al., 2014). Consequently, greater clinician awareness of and capability with evidence-based integrated interventions is required (Turner et al., 2018). Nonetheless, the literature remains primarily dependent on expert opinions to inform treatment of this comorbidity (Smith & Book, 2008), with a striking lack of research voicing the experiences of the clients themselves and their treatment needs.

The extant literature also largely fails to consider how homeless individuals may respond to treatment for substance use and social anxiety. While a recent meta-analysis suggested that 12-step programs can produce higher levels of abstinence than CBT (Kelly et al., 2020), qualitative research suggests that people who have experienced homelessness and substance use often reject abstinence-based programs. This may be because these programs (e.g. AA) do not meet the needs of homeless individuals (Timko, 2008). While 12-step programs emphasise that substance use is maladaptive and must be ceased, it may be perceived as an adaptive coping mechanism by some homeless individuals. Collins et al.’s (2018) qualitative investigation of homeless individuals revealed that drinking can be a social ritual which fosters community among homeless individuals and assists in managing difficult social situations.

This mismatch between abstinence-based programs’ values and homeless individuals’ expectations (Collins et al., 2012) may manifest in negative views of these programs. For example, Grazioli et al. (2015) found that perceptions of 12-step programs among homeless individuals were polarised but predominantly negative, with 54.8% of participants criticising the discourse, members, therapeutic value, and/or structure of 12-step programs. Homeless individuals have reported repeatedly engaging in abstinence-based treatment for various reasons; rather than seeking to cease consuming alcohol, they may wish to meet their basic needs (e.g. stable accommodation), engage socially with other homeless individuals, and take a break from (rather than completely stop) drinking (Collins et al., 2016). Furthermore, homeless individuals have proposed that recovery from alcohol use would be better facilitated through alternative pathways such as provision of stable housing, harm reduction, and opportunities for meaningful activities and social connections (Collins et al., 2016).

The current study primarily aimed to develop a deeper understanding of the treatment experiences and needs of homeless individuals with social anxiety and substance use. Given the dearth of current research into treatment for comorbid social anxiety and substance use, and the literature being guided primarily by expert opinions (Smith & Book, 2008), this study sought to give voice to the lived experience of homeless individuals with this comorbidity.

Method

Design

Due to the lack of existing qualitative research among individuals with social anxiety and substance use issues, especially among homeless populations, a qualitative exploratory design was chosen. This study was informed by a critical realist epistemological framework (Willig, 1999), acknowledging that the experiences of participants may not be a direct representation of reality, with the researchers interpreting the meaning of these experiences. Homeless individuals are recognised as a difficult to reach research population (Booth, 1999). Accordingly, all participants were recruited from the Mission Australia Centre using convenience sampling. The Mission Australia Centre is a specialist service based in New South Wales, Australia, that provides supported housing units for adult men experiencing homelessness. Focus groups were chosen to facilitate participants comparing and contrasting their views (Kitzinger, 1995), given the variability in the experiences of co-occurring mental health difficulties and homelessness (Gutwinski et al., 2021; Neale, 2001; Teesson et al., 2003). Additionally, focus groups were pragmatic given the recruitment context, as the service setting offered residents regular group meetings. Thematic analysis was deemed most appropriate for our study aims, a data analysis method which involves identifying common ideas and themes within discussions (Braun & Clarke, 2006). This study was approved by the Human Research Ethics Committee at the Australian College of Applied Psychology.

Participants and Recruitment

To be eligible to participate in this study, participants needed to be currently experiencing homelessness, substance use, and social anxiety. Homelessness was indicated by meeting established admissions criteria of Mission Australia Centre. Social anxiety was indicated by a score of 30 or more on the Liebowitz Social Anxiety Scale-Self Report (LSAS-SR; Rytwinski et al., 2009), while substance use difficulties were evidenced by self-reported current or historical substance use on the Addiction Severity Index (ASI; McLellan et al., 1992). In the interest of ecological validity, participants were not required to meet full diagnostic criteria of SAD or substance use disorder. Focus group participants were recruited with the assistance of Mission Australia between March and December 2020. Mission Australia Centre staff informed clients of the impending study a week prior to recruitment, and interested clients were invited to attend an information session run by the researchers, advising them of the study requirements and that participation was completely voluntary.

A total of 14 men participated in two focus groups (n = 7 in each group). Participants were all male and aged between 27 and 63 years (M = 45.21, standard deviation, SD = 9.90). The social anxiety profile of the sample, as indicated by the LSAS-SR, ranged from mild social anxiety (score of 40) to very severe social anxiety (score of 120). The average score was 75.36 (marked social anxiety). The substance use profile of the sample, as indicated by the ASI, varied considerably. Problematic substances included alcohol (any use at all), alcohol (to intoxication), cannabis, cocaine, amphetamines, other opiates/analgesics, barbiturates, sedatives/hypnotics/tranquilisers, hallucinogens, and inhalants. Participants reported difficulties with substance use ranging from 1 to 47 years, with periods of abstinence ranging from 0 to 2 years. Previous treatment for alcohol abuse ranged from 0 to 6 occasions (M = 1.0), and previous treatment for drug abuse ranged from 0 to 10 occasions (M = 2.21). All participants provided written consent to participate and for the focus group to be audio-recorded. Two focus groups were conducted, both of which were facilitated by the first author and ran for 2-h duration. The discussion was audio-recorded and transcribed verbatim, with transcripts de-identified to protect participants’ confidentiality.

Materials

The focus group was semi-structured, with six questions (see Table 1). Given that social anxiety and substance use appear to be mutually reinforcing (Bruce et al., 2005), and that substance use may precipitate or follow homelessness (Johnson et al., 1997), these questions were intended to elicit participants’ beliefs about the relationship between social anxiety and substance use based on their own experience, rather than presupposing causality. The following questions sought to elicit participants’ experiences with treatment for social anxiety and/or substance use. This was important as clients often report that their treatment needs for these disorders have not been met (Melchior et al., 2014). Finally, the focus group sought to identify participants’ perceived needs for treatment of comorbid social anxiety and substance use and anticipated barriers to engagement in therapy. As substance use may interfere with social anxiety treatment and vice versa (Randall et al., 2001; Rapee & Heimberg, 1997), it was important to understand such barriers.

Table 1 Focus group questions
Full size table

Analysis

Dedoose 8.3.17 (SocioCultural Research Consultants, 2020) was used to analyse the data, following Braun and Clarke’s (2006) six-step process for thematic analysis. The first author transcribed the focus groups, gaining familiarity with the data by reading and re-reading the transcripts several times while documenting initial thoughts and observations. Secondly, initial codes were generated using an inductive approach, by highlighting points of interest such as concepts or linguistic features. The third step involved collating these codes into emergent themes. Fourth, these themes were reviewed to form a thematic map highlighting connections between the identified themes. Next, the themes were further refined, updated, and named during several face-to-face meetings with the second and third authors and organised in line with a consensus achieved between the authors. Finally, the analysis of these themes was compiled, accompanied by selected extracts of the discussion from the focus groups.

Results

The analysis produced a set of four key themes and 15 subthemes (summarised in Table 2) describing participants’ perceptions of social anxiety and substance use and their experiences of and needs for treatment of these conditions.

Table 2 Summary of main themes and subthemes
Full size table

Longstanding Difficulties with Substance Use and Social Anxiety

This theme captures participants reflecting on their long histories with substance use and social anxiety, which had both typically persisted for decades. Some participants reported that they did not initially realise that their substance use was problematic. Others were aware that they were experiencing substance use difficulties but acknowledged active denial and continuation of these behaviours:

I never realised I was an alcoholic, I was always denying. I drink every day but I say I’m not an alcoholic… [I was drinking for] maybe more than 40 years… I started when I was 15 years old so I never considered myself an alcoholic or a drug addict or anything. (Participant F)

I’m coming into my 14th month of being sober… up until 14 months ago, I was living in denial. (Participant E)

Similarly, participants’ experiences with social anxiety were primarily characterised by a lack of awareness that their psychological symptoms constituted social anxiety. Furthermore, they indicated that lack of awareness of social anxiety was perpetuated by broader societal unawareness of the condition. For example:

In the old days, I guess people just would have said you’re shy, or you’re not good with people… and that’s what I always assumed. I didn’t think I had some kind of disorder… I just thought there was something wrong with me that meant that in a social situation, I couldn’t cut it and… there was just some limitation or failure in me to connect with people. It took me a long time to realise that… I had social anxiety. A real thing. (Participant A)

Participants also reflected that, with hindsight, they wish they had sought or had the opportunity to access substance use and social anxiety treatment earlier. They suggested that early intervention would likely have alleviated significant, ongoing distress.

For me, I should have done [a social anxiety] program probably 20, 30 years ago. But, it took me a long time to work it out… I didn’t want to ask for help. (Participant E)

“I wish I had known about it [social anxiety and substance use treatment] 30 years ago. God, I could have saved a lot of heartache.” (Participant A)

Mechanisms of Substance Use, Social Anxiety, and Their Interaction

Participants explained their perception of how substance use and social anxiety function and interact in their lives. Most participants reported that their social anxiety was present prior to substance use.

I definitely think that the anxiety came first in my perspective. In high school, and then the substance use came after (Participant K)

Participants reported that substance use and social anxiety go hand in hand and are mutually reinforcing. Consequently, both conditions gradually escalated to problematic levels over time:

I think they would have existed in parallel, simultaneously. But they were potent together and just reinforced each other… alcohol to deal with the social anxiety and also I think probably some social anxiety as a result of the use… It became a crutch and I couldn’t sort of function without it. So if you took away the alcohol… socially I was just useless. (Participant C)

While participants initially believed that substance use, especially alcohol, assisted social functioning, they acknowledged that this was an unsustainable short-term solution. Over time, substance use failed to alleviate, or even exacerbated, anxiety. For example:

As my addiction went on, [alcohol] worked less and less well to the point where it actually made almost no difference because my level of anxiety was so high it would just punch right through the… effects of the alcohol… The alcohol addiction had trumped the anxiety in a sense that I was only drinking just to stay level and physically comfortable, so the anxiety was as bad as ever, if not worse. (Participant A)

Disenfranchisement with Existing Treatment Options

This theme highlights how, despite participants’ openness to treatment, they have largely grown disenchanted with the treatment options available. A number of participants discussed in negative terms prior attempts to treat their social anxiety and/or substance use with medication.

I think they just give you the medication before you’re even back to your base self, before they can even see what the original root of it was. How can you diagnose someone when they’re not quite themselves because they’re on drugs or something like that? I don’t even know what they [medications] do, there’s such a variety. It’s an experiment. (Participant N)

The amount of times I’ve had a 10-minute session with a doctor or a psychiatrist… where they’re virtually finishing your sentences for you and prescribing SSRIs, SNRIs or double… And that was it, over and over again… And that was the kind of frontline approach to anxiety… Incredibly unhelpful, because they obviously weren’t really thinking it through. (Participant C)

A number of participants noted prior experiences attempting to address their substance use and social anxiety through rehabilitation centres. These were predominately described as a temporary fix.

Rehab is fine while you’re there, you’re in, like, a protective bubble… The real work, for me, starts when you get out because that’s when you’ve got to deal with the real world. I always looked for quick fixes going into short-term rehabs and they didn’t work (Participant E)

[Rehab] just gives you tools, it’s up to you if you follow that… You’re the one who makes the decision (Participant M)

Participants noted that for the homeless individuals experiencing social anxiety and substance use, rehab often served their housing needs more than their substance use needs.

You go there for sleep if you’re homeless. Somewhere you’ve got a roof for a little while. That’s my thinking about rehab. (Participant D)

Everyone I know who’s gone to rehab has come out… with a habit that’s fucking worse… But it gives you a roof if you didn’t have one. (Participant N)

Participant also criticised the overall structure of many residential rehabs, suggesting that they can cause harm and that the therapeutic community model can yield harmful outcomes. For example:

A lot of rehabs are based on this therapeutic community model which is very, very peer-to-peer. I think that can be really dangerous… It’s the inmates running the asylum. (Participant C)

They’re treated like animals. You go [to rehab] for six months and you are clean for six months. When you get out of there, all of the people, they relapse… because they learn nothing. They’re just locked there, really. (Participant D)

Participants reported that existing treatment options, primarily rehab and community peer support programs such as 12-step, were particularly difficult for substance abusers who also had social anxiety.

I get anxious when I go to fellowship meetings, because of the big crowd of people… When I get asked to share, the anxiety flares back up and I just shut down completely. (Participant M)

I’d be anxious about [group therapy]. Not wanting to go when you see how many people are there. (Participant K)

Needs for Future Treatment

This theme captures participants’ thoughts and suggestions for improving treatment of comorbid substance use and social anxiety. Participants highlighted a lack of understanding of social anxiety and the consequent importance of basic psychoeducation about social anxiety. As one participant said:

I’d just be curious maybe to get an insight into what the current thinking is around [social anxiety], even a definition. I mean, I know how uncomfortable I might feel, scared I might feel in social situations or how I might want to avoid it but… I’m not really conversant with what social anxiety actually is. (Participant C)

Participants reported the importance of practical skills to manage anxiety as an adaptive substitute for substance use:

The practical side of CBT, techniques on how to get through it (Participant H)

I think it’s about some practical advice, tools, skills, things that can be used in the moment… whether that be a mindful exercise or just sort of recasting in your mind what’s at stake. For example, you know, does it matter if people think I’m boring right now? Or… practical little mind tricks that I could do at the time. That’s what I’m looking for. (Participant A)

Participants also discussed crucial aspects of the therapeutic space. They highlighted the importance of being understood by both participants and therapists in group treatment. Some participant suggested that they are more amenable to treatment that incorporated peer support or treatment led by individuals with lived experience of substance use:

Sometimes you’ll come across people that really haven’t experienced it themselves… it’s all good and well if people can read everything out of a book… but I’ve found that the best people that help you, they actually experienced it themselves… they can be more understanding as to how you’re feeling and what you’re going through (Participant E)

Peer support is amazing… There’s nothing quite like someone who can look you in the eye and know exactly what you’re talking about. To get that camaraderie and that shared experience.” (Participant A)

However, participants also recognised the importance of appropriately trained therapists and the necessary balance between lived experience and clinical expertise. For example, some participants raised concerns about individuals with histories of substance use working as therapists.

You see a lot of people, they’ve done their rehab and then all of a sudden, they’re getting in the job of helping other people, being professionally paid with minimal training to help people who are in rehabs (Participant E)

Discussion

The current investigation deepens insight into homeless individuals’ experience of comorbid social anxiety and substance use difficulties. Participants described substance use and social anxiety as longstanding conditions which can endure over several decades. They reported that they were parallel processes, wherein drugs (primarily alcohol) were used to manage anxiety in social situations and developing dependence upon substances in turn enhanced social anxiety. Participants noted that substance use was indeed initially helpful to alleviate social anxiety in the short term but unsustainable in the long term as it did not address the root cause of the anxiety. Therefore, anxiety intensified and the anxiolytic effects of substance use diminished.

Strikingly, despite the distress elicited by this comorbidity, this cycle continued for many years, and participants reported regretting not seeking treatment much earlier. This desire for treatment underscores the currently unmet need for accessible integrated treatments for this population (Melchior et al., 2014). This apparent willingness to engage in treatment may signal a window of opportunity for integrated programs to intervene earlier in the cycle of social anxiety and substance use. Psychological services providing psychoeducation about the relationship between social anxiety and substance use and outreach-based treatment, especially to at-risk groups such as homeless people, may encourage individuals with this comorbidity to seek help sooner, thereby interrupting this cycle before it can progress and cause years of distress. Early intervention may be particularly important for this comorbidity given that young people with SAD and substance use may more actively engage in treatment and peer-helping than those without SAD (Pagano et al., 2015).

Despite this window of opportunity, barriers to treatment for social anxiety and substance use were raised during the focus group. For example, participants reported denying or not recognising their problematic substance use, and that broader society did not understand that their social anxiety constituted a legitimate psychological diagnosis. While these disorders are ostensibly better understood and recognised today, the limited research on their integrated treatment (Smith & Book, 2008) may suggest that understanding of these disorders still requires improvement, both among the general population and researchers. Ongoing empirical and qualitative investigation is required to improve common and scientific knowledge of these conditions and assist in development of integrated comorbidity interventions.

Melchior et al. (2014) suggested that individuals with anxiety and substance use disorders perceive their needs to be unmet due to a scarcity of available services, and the current study highlighted various other factors which may contribute to this dissatisfaction. For example, participants indicated that they were alienated by the peer-to-peer therapeutic model of 12-step programs. These qualms with existing treatments underscore the importance of developing care options congruent with the values of individuals with histories of substance use and social anxiety. Aligning interventions with treatment-seekers’ values may be even more important when their personal context is further complicated by homelessness. Homeless individuals are a special population whose needs are often not considered in the development of 12-step programs (Collins et al., 2012; Timko, 2008), which can lead to negative appraisals of these services (Grazioli et al., 2015). Furthermore, alcohol use in this population may be perceived as adaptive as it forms part of important bonding rituals (Collins et al., 2018). Given that socially anxious adults have been found to benefit less from 12-step programs than those without social anxiety (Terra et al., 2006), substance use may be an even more salient safety behaviour among homeless individuals, where it has a clear social advantage. Consequently, homeless individuals have proposed that they would derive more therapeutic benefit from being provided with stable housing or engaging in harm reduction (rather than abstinence-based treatments; Collins et al., 2016). Therefore, while there may be a window of opportunity to address comorbid social anxiety and substance use, interventions must recognise and be responsive to the potential for chaos in homeless individuals’ lives (Collins et al., 2018; Neale, 2001; Teesson, et al., 2003) and how their substance use and anxiety interact in this context.

To this end, participants in the current study offered insight into their treatment needs for social anxiety and substance use. They reported hoping to learn practical strategies to manage anxiety and replace substance use. Participants acknowledging the need to learn more adaptive ways to manage anxiety than resorting to substance use underscore the inextricable link between these conditions and the need for integrated treatment (Turner et al., 2018). Furthermore, it reflects the first time that voice is being given to the people with lived experience of social anxiety and substance use on how this comorbidity can be addressed, in a literature characterised by expert academic opinions and minimal empirical research (Smith & Book, 2008). Participants also highlighted a desire for psychoeducation about social anxiety. Their admitted lack of understanding of social anxiety reflects the historical misunderstanding of social anxiety among the general population, which perpetuated their difficulties over multiple decades, and once again emphasises the potential value of outreach to homeless populations at higher risk of developing these difficulties (Gutwinski et al., 2021). Furthermore, the reported desire for a better fundamental understanding of social anxiety and the relatively generic descriptions of what participants desire from treatment (e.g. “practical advice, tools, skills”) emphasises the current dearth treatment options for this comorbidity.

Reflecting on their needs for a therapeutic space, participants emphasised the importance of feeling able to express themselves and be understood. Furthermore, they suggested that it could be advantageous for the facilitator to also have lived experience with substance use; however, this was counterbalanced by recognition that therapists should be trained professionally. These findings reinforce the importance of designing interventions for this comorbidity that account for the target population’s values. For example, group-based interventions may be particularly valuable to offer the sense of connection and unity that may be imparted by communal drinking among homeless individuals (Collins et al., 2016), while offering more adaptive strategies to manage social anxiety and curb substance use. Given that one RCT on group-based therapy for social anxiety and substance use found an increase in drinking behaviour, which may have been attributable to anxiety elicited by exposure tasks (Randall et al., 2001), ensuring a safe, supportive environment in therapy appears essential to facilitate successful treatment of this population.

This study had a number of limitations. Participants were all residents at the Mission Australia Centre and were therefore being provided with stable accommodation. Consequently, the perspectives in this study may be those of individuals who, while homeless, have less chaotic daily lives (Collins et al., 2018) than those without steady accommodation. Participants may therefore have been more contemplative of their treatment needs for social anxiety and substance use, compared to homeless individuals without housing, who likely place much higher value on receiving basic needs such as shelter or food. Future research may investigate how perceived treatment needs may differ between individuals at different stages of homelessness, which would assist with more effectively targeting higher risk individuals earlier in their homelessness trajectories. Furthermore, due to the recruitment strategy which relied on convenience sampling to access a niche population, all participants were male and varied in age considerably. Further research could focus on how the experiences of homeless men and homeless women with comorbid substance use and social anxiety differ and any differences in attitudes towards social anxiety and substance use according to participant age. Similarly, future research could examine whether the interaction between substance use and social anxiety differs as a function of substance type, as it was not possible to do so given the reliance on convenience sampling in this study.

The use of focus group methodology in a population with social challenges may have precluded participation from individuals with more severe social anxiety. Thus, the current study represents the views of those individuals with some capacity to engage in a social group format. It is also noteworthy that the discussion in the focus group primarily centred on substance use and social anxiety, with minimal references to homelessness. As all participants had stable accommodation, homelessness may have been a less salient concern compared to their difficulties with substance use and ongoing social anxiety (which require daily attention to manage). While substance use has been reported to contribute to homelessness (Johnson et al., 1997; Mabhala et al., 2017), and socially stressful situations which induce coping-based substance use may frequently occur in homeless individuals’ daily lives (Collins et al., 2018), it remains unclear if and how social anxiety may contribute to entry into and maintenance of homelessness. It may be valuable in future studies to explicitly explore the perceived interaction, if any, of social anxiety with homelessness trajectories.

In summary, the current study explored the experiences of homeless men with histories of social anxiety and substance use. Participants noted that these conditions fed into each other and went untreated for extended periods of time. Furthermore, they noted that the treatments available for substance use and social anxiety did not fully align with their values. Participants made key recommendations that future treatments should focus on practical anxiety and substance use management skills and ensure a therapeutic space which facilitates understanding between group members and therapists.