Neither God nor others: a qualitative study of strategies for avoiding suicide among homeless people

Introduction: Drug addiction and alcoholism characterize the existential condition of most homeless people, while the risk of suicide runs in parallel. Objectives: Following the Provisional Model (PM), this study aimed to explore the relationships between addiction, suicidal ideation, and religiosity among 13 homeless people, and the roles of bonding ties (within the group) and bridging ties (intergroup). Method: The study is rooted in the field of qualitative psychology research. A survey was conducted, analyzing the personal accounts of participants and applying the PM psychological interpretation integrated with theory from literature on drug addiction, religiosity, and suicidal ideation in the field of homelessness. Results: Outcomes show that suicidal ideation appears in the first phase of homelessness and is opposed by alcoholism and drug addiction. Religiosity does not help to counteract suicidal ideation or to create bonding relationships. Conversely, drugs and alcohol seem to be more useful for preventing suicide, but also do not help in bonding relationships. Conclusion: Our survey only partially confirmed the PM, because the main result was the importance of relationships between suicidal ideation and alcohol/drug abuse during the initial phase of becoming homeless, while the importance of bonding ties deriving from addiction behavior did not emerge.


Introduction
Homeless people are frequently unable to acquire and maintain regular, safe, and adequate housing, 1 often because of alcoholism, drug addiction, and migration problems. In fact, over 70% of homeless people abuse alcohol and/or illicit drugs, 2-4 while refugees tend to become homeless because of loss of their social identity and support networks. 5,6 Some surveys that have focused on integration processes and group relationships have found that they orient relationships, especially towards bonding (ties within a group), whilst avoiding bridging (intergroup ties). 7 In particular, it seems that there is a strong sense of unity within groups of homeless alcoholics, more so than among the homeless in general. 8 These studies arose in the area of the Provisional Model (PM), proposed by Farrington and Robinson, 9 and aimed to analyze aggregation among homeless people and the construction of their social identity, suggesting that their strategies change in relation to the amount of time (years) a person has been homeless. 10 The PM describes different phases of homelessness. The first, "Aspirant exiters," develops over the initial ten-totwelve months, when people identify themselves as homeless and make favorable intragroup comparison "on the basis of their past accomplishments and the perceived likelihood of escaping from homelessness," 9 since at this point escape is quite likely. However, as time passes, escape from homelessness seems less and less likely and it does not appear as useful anymore for positive intragroup comparisons.
In the second phase, "Deniers," during the second year, individuals no longer identify themselves as homeless and, unable to find new coping strategies, may deny their reality in order to escape from suffering.
Although this could be an adaptive mechanism, if the isolation from reality becomes too pervasive, they can shift toward mental illness.
During the third phase, "Subgroupers" (2-4 years), individuals associate themselves with a particular group of homeless people and can make favorable intergroup comparisons against other homeless groups.
Elements that can enhance a positive identification among homeless include religion, and even drug or alcohol abuse. However, Farrington and Robinson suggest that even this strategy, over time, might become impractical, since individuals may "lose control over their associations and lifestyle" 9 (p. 186).
The fourth and last phase is divided into 3 sub-phases.
In sub-phase 4a, "Carers and sharers" (5-12 years), they might develop roles for specific identities, making positive intragroup comparisons. During sub-phase 4b "Family" (12-15 years), any comparison ceases. Lastly, sub-phase 4c "Typicals" is different from the others, since it can start after 3 years and a half and last until the end of the process, and may be caused by a failure in the previous mechanisms, leading the person to identify himself/herself as a prototypical homeless person.
From Farrington and Robinson's perspective, the progression from the first to the second phase is the most critical moment, because of the severe stress and the lack of solidarity ties.
The PM also considers the possibility that some changes might be influenced by other factors, for example increasing dependence on alcohol. 9 As highlighted by the social identity theory 11 and by the self-categorization theory, 12,13 religion is a cultural dimension that constitutes an important factor for social and personal identity. 14 Furthermore, evidence confirms that it supports psychological health, 15,16 improving emotional stability against depression, death anxiety, self-harm and suicide. [17][18][19] Since religion is also

Method
This survey is founded on the narrative method and was conducted using in-depth interviews 25

Ethical approval
The study followed the consolidated criteria for reporting qualitative research (CORE-Q) check-list, 40 and obtained approval from the ethics commission at the psychology department of the Università degli Studi di Padova (University of Padova).

Participants, data collection, and qualitative analysis
The participants were 13 homeless people able to understand and speak Italian, who were recruited at a volunteer association in Northern Italy. The interviewer gave potential participants detailed information regarding the goals of the study and how the interview would be conducted. Consent was requested verbally.
Most of the participants were immigrants and only three of them were Italian. Table 1 lists their sociodemographic data.
The interviews covered a series of issues that were dealt with in as much depth as possible. 5 all further themes were introduced as participants began to tell their stories. Further questions followed,  Six main phases characterized the process: preparatory organization, generation of categories or themes, coding data, testing emerging understanding, searching for alternative explanations, and writing up the report. 43

Results
Four main areas of thematic prevalence emerged from the analysis: "Suicide as a solution in the initial phase of homelessness"; "Addiction as the real remedy"; "Religion, but with drugs and music"; and "Atheism, but with drugs and gambling."   were "music," and "love for wife and children." The most frequent words were "I" and "wife." These two Once back in Africa, he divorced and left the home and his savings to his ex-wife and children. He was an animist and felt he was "special" but also "followed by some spirits that have been haunting me." He had been living on the streets for the last six years, without a job and taking illicit drugs. Music and drugs gave him the strength to survive and brought him to "a different world" in a tension towards transcendence. "You have to keep searching and never give up. Still, I haven't found anything at the present, but I am not disheartened."

Suicide as a solution in the initial phase of homelessness
Although his religion did not provide support for him to cope with street life, his main topics were religion and music, and the most frequent words were "worlds," "pain", and "life," to indicate his perseverant existential enquiries into the meaning of the world and into the suffering that the human (and his) condition implies.

Atheism, but with drugs and gambling
In the final narratives, religiosity disappeared completely while addictions remained as a support.  As it turned out, use of substances seemed to offer an alternative. Drugs and alcohol, although harmful to health, contributed to addressing daily suffering.
Using Emile Durkheim's concept of "suicidal erosion," 53 we could interpret this outcome by seeing addiction as the fundamental adaptive behavior. 3 However, we cannot confirm that addiction and alcoholism offer the opportunity to create stronger ties and identity within the group of homeless people.

Conclusion
This study with 13 homeless people adopted the Provisional Model and analyzed relationships between their suicidal ideation and alcohol and drug abuse, in order to investigate a possible role of religion as a protective factor against suicide and as a factor promoting bridging/bonding ties.
All participants showed a tendency to indulge in vices due to the failure of their condition, and more than a third of them had seriously considered dying by suicide. Issues relating to religion and representation of death did not arouse any particular interest and the responses were sometimes laconic. Even though almost all of them were 'believers' and they saw death as a passage, religion appeared neither to be a factor of protection against suicide and addiction nor to improve social ties. Finally, substance abuse constituted a protective factor against suicide and increased with progression through the stages described by the PM, while religiosity progressively disappeared and no positive in-group relationships were developed thanks to either alcohol or religion.
The limitations of this study consist of the small number of participants and the difficulties related to analysis of representations of religion and God. In future research, we could help participants by providing them some inputs that may facilitate narratives, in particular when linked to biographical aspects.
Moreover, in order to better grasp the complex condition of homeless people who are also immigrants (as were the majority of our participants), it would be important to investigate in greater depth their specific points of view, exploring how they perceive their life in the new country, especially in terms of social relationships, cultural differences, stigmatization, and support received. Another significant topic to explore could also be comparison between the lives, perceptions of the self, and coping mechanisms of homeless immigrants and homeless people who were born in the country in which they currently live, in order to assess similarities and potential differences, with the objective of offering both groups more adequate support, based on their specific needs.

Disclosure
No conflicts of interest declared concerning the publication of this article.