Abstract
Background:
Homelessness is a major problem that critically impacts the mental health and well-being of the affected individuals. This umbrella review aimed to evaluate the current evidence on the prevalence of mental disorders among homeless people from evidence-based systematic reviews and meta-analyses.
Methods:
We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and Joanna Briggs Institute (JBI) methodology for umbrella reviews. We searched 12 major databases and additional sources to identify systematically conducted reviews and meta-analyses reporting the prevalence of mental disorders among homeless populations.
Results:
We evaluated 1,277 citations and found 15 reviews meeting our criteria. Most studies were conducted among high-income countries with samples from different age groups. Studies reported high prevalence rates of depressive and anxiety disorders, schizophrenia spectrum and psychotic disorders, substance use disorders, suicidal behavior, bipolar and mood disorders, neurocognitive disorders and other mental disorders among homeless people. Moreover, studies also reported a high burden of co-occurring mental and physical health problems among the homeless experiencing mental disorders.
Conclusion:
This umbrella review synthesized the current evidence on the epidemiological burden of mental disorders in homelessness. This evidence necessitates advanced research to explore psychosocial and epidemiological correlates and adopt multipronged interventions to prevent, identify and treat mental disorders among homeless populations.
Keywords
Introduction
Homelessness is a growing population health concern worldwide (Omerov et al., 2020). In the United States, a total of 552,830 people experienced homelessness on a single night in 2018 (Henry et al., 2018). In the same year, nearly 320,000 people in the United Kingdom were recorded as homeless (Shelter England, 2018). Moreover, around 235,000 individuals were homeless during 2016 in Canada (Gaetz et al., 2016). In the twentieth century, homeless people were typically older men, whereas homelessness in recent years has been observed among females and youth of different racial and ethnic groups (Jones, 2016). This evidence provides an overall scenario of homelessness in the high-income nations; however, evidence on the severity of homelessness is scarce from low- and middle-income countries (LMICs). Socioeconomic factors associated with homelessness including rapid urbanization and industrialization, high poverty, income inequality, unemployment, maldistribution of resources between rural and urban areas, migration, and lack of access to affordable housing are prevalence in LMICs, which suggest the magnitude of homelessness is likely to be higher in those contexts (Speak, 2019). This global burden of homelessness has critical implications for health policymaking and practice. In homelessness, the rate of mortality is nearly eight times higher than the average for men and 12 times higher for women, with an average age for death at 52 years (Aldridge et al.,2019, 2018). This can be attributable to the fact that homeless individuals experience enormous health inequalities and have a higher prevalence of various medical conditions (Omerov et al., 2020). This burden increases among aging people who are homeless. They often experience multiple health problems, which become worse in the presence of a lack of access to health care services, poor social ties and continued effects of other social determinants of health (Omerov et al., 2020). In most cases, such problems are not prevented or diagnoses at earlier stages resulting in increased use of acute care services and higher cost to the health systems (Omerov et al., 2020; Rosenheck & Seibyl, 1998).
Addressing homelessness is essential for the overall health and well-being of a population. This agenda is part of the United Nations 2030 Agenda for Sustainable Development (United Nations, 2015). The action plans to alleviate homelessness include reducing poverty and ensuring health and well-being. In addition to addressing residential challenges and physical health problems, it is necessary to improve the mental health of people who are homeless (Altena et al., 2010; Dickey, 2000). Psychosocial stressors like impaired interpersonal relationships, lack of hope, loneliness and poor social capital affect the mental health and well-being during homelessness (Omerov et al., 2020). Another challenge is the deinstitutionalization of mental illness in many countries like the United States, which resulted in a reduced number of beds in indoor psychiatric facilities without strengthening community shelters and models of care (Yohanna, 2013). In addition, LMICs generally have lesser organizational capacities to provide mental health services, both in the institutional and community settings (Daund et al., 2018). Such challenges may have brought many people with severe mental illness to the temporary shelters and streets, thus increasing both the burden of mental disorders and homelessness within a given geographic region. From a population health perspective, homeless people, irrespective of a previous psychiatric diagnosis, are likely to live in disadvantaged conditions that make them vulnerable to mental disorders and remain undiagnosed as well as untreated in most cases (Patten, 2017; Yim et al., 2015).
To address the pre-existing mental disorders and promote positive mental health and resilience, it is essential to understand the epidemiological burden of mental disorders among homeless individuals. In this regard, observational studies may provide insights about the spatial and temporal distribution of mental disorders in homelessness. Moreover, systematic reviews or meta-analysis of observational studies can address the sampling errors of individual studies and provide a less-biased estimate of the burden of health conditions of interest. Given that more than 11 systematic reviews or meta-analyses are published in a day (Bastian et al., 2010), it is essential to synthesize the evidence from existing reviews. Such reviews of the reviews, also known as umbrella reviews, have shown advantages over meta-analyses due to their bird’s eye view of evidence (Ioannidis, 2009). Furthermore, umbrella reviews are increasingly used for clinical risk prediction and evidence-based actions through synthesizing prior knowledge on several mental disorders (Fullana et al., 2019). There is a lack of an umbrella review that can provide global evidence on the prevalence of mental disorders in homelessness, which may inform future research, policymaking and practice. This umbrella review acknowledged this knowledge gap and synthesized the current evidence on the prevalence of mental disorders among people who are homeless.
Materials and methods
Search strategy of the review
This umbrella review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Liberati et al., 2009) and the Joanna Briggs Institute (JBI) methodology for umbrella reviews (Aromataris et al., 2015). We searched MEDLINE, Embase, PubMed, PsycINFO, CINAHL, Health Policy Reference Center, ERIC, Health Source (Nursing/Academic Edition), Environment Complete, Child Development & Adolescent Studies, Academic Search Ultimate and the Cochrane Library using specific keywords (please see Table 1). For each database, the titles, abstracts, subject headings and general keywords were searched with no language or time restrictions. Moreover, we searched the citations used as references of the primarily screened articles and citing articles from Google Scholar using the ‘cited by’ function. All databases and additional sources were searched from their inception to October 15, 2019, and the entire search process was repeated on December 21, 2019, for the last time. All the citations were compiled using RefWorks software and uploaded to Rayyan cloud-based citations management system for systematic evaluation.
Search strategy for this umbrella review.
Inclusion and exclusion criteria
In this umbrella review, we included reviews that systematically evaluated and reported the prevalence of mental disorders among people who are homeless. In addition, studies reporting other quantitative measures of disease burden (e.g. odds ratio (OR) or relative risk (RR) expressing the epidemiological burden of a disorder) were also considered in the absence of a prevalence value. To specify mental disorders, we adopted the classifications of mental disorders in the International Classification of Diseases (ICD) 10th revision (World Health Organization [WHO], 2016), or the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013). In addition, we acknowledged the historical evolution of the definitions and conceptual constructs within mental disorders. Therefore, we included reviews reporting mental disorders which are consistent with the earlier versions of ICD or DSM and have an equivalent diagnosis classified under the current versions of these guidelines. Moreover, there are several definitions of homelessness in different contexts, which may include some individuals while excluding others within the scope of respective definitions (Byrne & Culhane, 2015; US Health and Human Services, 2020). For example, in the United States, two major definitions are used by federal agencies to identify the homeless population and provide social care. While both definitions agree on most criteria, individuals living in motels or staying with others lacking a regular accommodation are considered homeless by the educational program definition, whereas those individuals are not eligible as homeless as per the Housing and Urban Development (HUD) definition (US Health and Human Services, 2020). Such definitions may have provided different estimations of study samples and the magnitude of different health problems in those samples. This review acknowledged such differences and considered any definition of homelessness in the eligible reviews and respective primary studies, which allowed this review to remain inclusive to diverse operational definitions that may have been used in different contexts over time.
Furthermore, articles were recruited if they fulfilled all the inclusion criteria and excluded if they did not meet any of the exclusion criteria as listed in Table 2. In this umbrella review, two authors independently evaluated the citations according to the pre-specified criteria. Any conflicts arising in the screening process were resolved at the end of the independent screening through discussion in the presence of a third author.
Eligibility criteria for this review.
ICD: International Classification of Diseases; DSM: Diagnostic and Statistical Manual of Mental Disorders.
Data extraction and synthesis
A data extraction tool was developed adopting the JBI data extraction tool for systematic reviews and research synthesis (Munn et al., 2014). Two authors used this tool and independently extracted data on the following domains: the objectives and types of each review, year of publication, names of databases searched in respective reviews, the timeframe of searching databases, sample size, location of the primary studies, demographic characteristics of the participants, recruitment strategy and key findings on the prevalence of mental disorders among homeless people. Furthermore, a narrative synthesis of the research findings was conducted considering high heterogeneity in terms of operational definitions of homelessness as well as mental disorders, methodological approaches and instruments within the primary studies and reviews. The synthesized findings on the prevalence rates (percentage, proportion, OR, RR, or other quantitative measures) with specific or range estimations within 95% confidence interval (CI) were reported from the respective reviews.
Evaluation of the methodological quality
To evaluate the methodological quality of the reviews, we used the JBI critical appraisal checklist for systematic reviews and research synthesis checklist (Aromataris et al., 2015). Two reviewers independently evaluated each of the included reviews. At the end of the primary evaluation, two reviewers discussed the evaluation findings, reached a consensus for all the items, and finalized the overall quality ratings. The checklist consists of 10 items, and each item could receive 1 point. Therefore, the overall quality score of a review could range from 0 to 10. In this umbrella review, articles receiving 0–4, 5–7 and 8–10 were categorized as the low, medium and high-quality studies, respectively.
Results
We found 723 citations from searching 12 databases and 554 citations from additional sources, totaling 1,277 citations (please see Figure 1). After eliminating 429 duplicates, we evaluated the titles and abstracts of the remaining 848 citations as per the pre-specified criteria for this review. At the end of this stage, we removed 826 citations due to non-compliance with our criteria and evaluated full texts of the remaining 22 citations. Seven articles were excluded at this stage, and we retained 15 articles in this review (Ayano, Tesfaw, & Shumet, 2019; Ayano, Tsegay, et al., 2019; Bassuk et al., 2015; Burra et al., 2009; Depp et al., 2015; Duke & Searby, 2019; Embleton et al., 2013; Ennis et al., 2015; Fazel et al., 2008; Folsom & Jeste, 2002; Hodgson et al., 2013; Parks et al., 2007; Schreiter et al., 2017; Smartt et al., 2019; Spence et al., 2004). The included articles consisted of seven meta-analytic reviews (Ayano, Tesfaw, & Shumet, 2019; Ayano, Tsegay, et al., 2019; Bassuk et al., 2015; Depp et al., 2015; Embleton et al., 2013; Fazel et al., 2008; Schreiter et al., 2017) and eight non-quantitative reviews (Burra et al., 2009; Duke & Searby, 2019; Ennis et al., 2015; Folsom & Jeste, 2002; Hodgson et al., 2013; Parks et al., 2007; Smartt et al., 2019; Spence et al., 2004). The summary findings of the included reviews are provided in Table 3.

Flow diagram of the umbrella review.
Characteristics and the key findings of the included systematic reviews and meta-analyses.
Characteristics of the included reviews
Most reviews (n = 10) included primary studies from multiple countries without specific geographic or economic focus, whereas three reviews emphasized on high-income countries (Bassuk et al., 2015; Fazel et al., 2008; Schreiter et al., 2017) and two reviews included studies from LMICs or resource-constrained contexts (Embleton et al., 2013; Smartt et al., 2019). Among reviews without a contextual focus, most primary studies were in the United States, Germany, Canada, the United Kingdom, Australia and other developed countries with fewer studies from LMICs (Ayano, Tesfaw, & Shumet, 2019; Ayano, Tsegay, et al., 2019; Burra et al., 2009; Depp et al., 2015; Folsom & Jeste, 2002). Moreover, the earliest review was published in 2002, and a total of five reviews were published before 2010 (Burra et al., 2009; Fazel et al., 2008; Folsom & Jeste, 2002; Parks et al., 2007; Spence et al., 2004), whereas most (n = 10) reviews were published after 2010. Furthermore, the median number of databases searched in the respective reviews was three, with a range from two to 16. The number of primary studies included in the respective reviews ranged from 10 to 50. In addition, the evaluation of the methodological quality (please see Supplemental file) found seven reviews with medium quality (Embleton et al., 2013; Ennis et al., 2015; Folsom & Jeste, 2002; Hodgson et al., 2013; Parks et al., 2007; Schreiter et al., 2017; Spence et al., 2004) and eight reviews with high methodological quality (Ayano, Tesfaw, & Shumet, 2019; Ayano, Tsegay, et al., 2019; Bassuk et al., 2015; Burra et al., 2009; Depp et al., 2015; Duke & Searby, 2019; Fazel et al., 2008; Smartt et al., 2019).
Characteristics of the study populations
The study populations in this review were homeless individuals with diverse characteristics, as identified in different study samples. The sample size among studies ranged from eight to 326,073 (Duke & Searby, 2019; Hodgson et al., 2013). Three reviews focused on adult participants who were homeless (Burra et al., 2009; Depp et al., 2015; Spence et al., 2004). Three reviews focused on homeless children and adolescents (Bassuk et al., 2015; Embleton et al., 2013; Parks et al., 2007). One review included primary studies with youth participants aged from 15 to 24 years (Hodgson et al., 2013). Moreover, one review focused on studies with women participants only (Duke & Searby, 2019). In many samples, racial and ethnic minorities were over-represented within the homeless groups compared to the general population (Spence et al., 2004). The study participants were recruited from multiple sources, including homeless shelters, social services, temporary residential facilities, downtown locations, suburban areas, health care facilities, community locations and streets (Bassuk et al., 2015; Burra et al., 2009; Duke & Searby, 2019; Embleton et al., 2013; Ennis et al., 2015; Fazel et al., 2008; Folsom & Jeste, 2002; Parks et al., 2007; Schreiter et al., 2017; Smartt et al., 2019; Spence et al., 2004).
Prevalence of mental disorders among homeless people
The prevalence of mental disorders varied across homeless samples. For example, Hodgson and colleagues (2013) reported the overall prevalence of mental disorders ranged from 48.4% to 98% among primary studies included in that review. In a meta-analysis by Schreiter and colleagues (2017), the pooled prevalence of mental disorders was found as 77.5% (95% CI: 72.4–82.3). The prevalence rates varied across geographic regions of corresponding primary studies among the reviews. For example, Fazel and colleagues (2008) found that primary studies conducted in the United States had lower pooled prevalence of mental disorders (9%, 95% CI: 6–12) compared to samples from Mainland Europe (12%, 95% CI: 7–16), the United Kingdom (19%, 95% CI: 9–29) and Australia (16%, 95% CI: 10–22). Such variations were also noted between studies from high-income countries and LMICs, studies published in the last decade and earlier years, and for different screening measures used across primary studies (Ayano, Tesfaw, & Shumet, 2019). Moreover, reviews reported different prevalence rates of specific mental disorders among homeless populations, which are presented in the subsequent sections.
Depressive disorders and anxiety disorders
Six reviews reported the prevalence of depressive disorders among homeless people, which ranged from 11.4% to 57.9% (Bassuk et al., 2015; Duke & Searby, 2019; Fazel et al., 2008; Hodgson et al., 2013; Schreiter et al., 2017; Spence et al., 2004). For example, Bassuk and colleagues (2015) found 13.8%–46.3% of the participants in the primary studies had depressive disorders. Another review reported 17.6%–28.1% of the study samples were suffering from depression (Hodgson et al., 2013). Moreover, three reviews reported the prevalence of anxiety disorders, which ranged from 10% to 32% across samples (Bassuk et al., 2015; Hodgson et al., 2013; Schreiter et al., 2017). For example, a meta-analysis by Schreiter and colleagues (2017) found the pooled prevalence of anxiety disorders was 17.6% (95% CI: 12.9–22.8) among homeless participants.
Schizophrenia spectrum and other psychotic disorders
Seven reviews reported the prevalence of schizophrenia spectrum and other psychotic disorders, which ranged from 1% to 45% (Ayano, Tesfaw, & Shumet, 2019; Depp et al., 2015; Duke & Searby, 2019; Fazel et al., 2008; Folsom & Jeste, 2002; Schreiter et al., 2017; Spence et al., 2004). For example, a systematic review found that young participants had higher rates of schizophrenia (13%–21%) than older participants (8%–14%), women participants had higher rates up to 35% compared to men (8%–12%), and chronically homeless individuals had higher prevalence (18%–27%) compared to newly homeless individuals (2%–14%) (Folsom & Jeste, 2002). Another meta-analytic review reported the pooled prevalence of psychotic disorder in homeless people was 21.21% (95% CI: 13.73–31.29), whereas the pooled prevalence rates of schizophrenia, schizophreniform disorder, schizoaffective disorder and psychotic disorder not otherwise specified (NOS) were 10.29% (95% CI: 6.44–16.02), 2.48% (95% CI: 0.16–28.11), 3.53% (95% CI: 1.33–9.05) and 9% (95% CI: 6.92–11.62), respectively (Ayano, Tesfaw, & Shumet, 2019).
Substance-related and addictive disorders
In this umbrella review, seven reviews were identified that reported the prevalence of substance-related and addictive disorders, which ranged from 4.5% to 60.9% across homeless samples (Depp et al., 2015; Duke & Searby, 2019; Embleton et al., 2013; Fazel et al., 2008; Hodgson et al., 2013; Schreiter et al., 2017; Spence et al., 2004). For example, a meta-analytic review found the pooled prevalence of alcohol dependence was 37.9% (95% CI: 27.8–48) (Fazel et al., 2008). Another review found that 11%–43.7% had alcohol or other substance use disorder (Hodgson et al., 2013).
Neurocognitive disorders
Six reviews reported the prevalence of neurocognitive disorders, which ranged from 4% to 80% (Burra et al., 2009; Depp et al., 2015; Ennis et al., 2015; Parks et al., 2007; Schreiter et al., 2017; Spence et al., 2004). The most commonly reported problems were cognitive impairments among homeless people. For example, Ennis and colleagues (2015) found 18%–55.4% homeless participants had general cognitive deficits. A meta-analytic review found 25.4% of the homeless adults had cognitive impairment (Depp et al., 2015). Another review by Parks and colleagues (2007) found 11%–35% of homeless children had impaired cognitive functions and associated disabilities.
Bipolar disorders and mood disorders
Four reviews reported the prevalence of bipolar and mood disorders among homeless people, which ranged from 5.1% to 41.3% (Bassuk et al., 2015; Depp et al., 2015; Hodgson et al., 2013; Schreiter et al., 2017). A review by Hodgson and colleagues (2013) found 26.9% of homeless participants had bipolar disorders. Moreover, 12.2%–41.3% of homeless samples had mood disorders in this review. Furthermore, a review by Depp and colleagues (2015) reported the prevalence of affective or mood disorders was 27.6% (SD = 18.8, median = 24%) in 10 studies.
Suicidal behavior disorder
Two reviews identified prevalence rates for suicidal ideation, attempt and self-injury among homeless people (Ayano, Tsegay, et al., 2019; Hodgson et al., 2013). Hodgson and colleagues (2013) reported the prevalence rates of self-harm (69%), suicidal ideation (22%–36.8%) and suicidal attempts (8.8%–46%). Another review by Ayano and colleagues reported the pooled prevalence of current suicidal ideation was 17.83% (95% CI: 10.73–28.14), whereas the prevalence of lifetime suicidal ideation was 41.6% (95% CI: 28.55–55.95) (Ayano, Tsegay, et al., 2019). Moreover, the pooled prevalence rates of current and lifetime suicidal attempts were 9.16% (95% CI: 4.1–19.2) and 28.8% (95% CI: 21.66–37.18), respectively.
Other mental disorders among homeless people
Several other mental disorders were reported across reviews. Two reviews reported the prevalence of attention-deficit/hyperactivity disorder and conduct disorder ranging from 4.4% to 34% and 36% to 76.7%, respectively (Bassuk et al., 2015; Hodgson et al., 2013). Moreover, two reviews reported that 23.1%–29.1% of homeless people had personality disorders (Fazel et al., 2008; Schreiter et al., 2017). Furthermore, Bassuk and colleagues found 19.7% of children and adolescents had disruptive behavioral disorders. A review by Smartt and colleagues (2019) found seven studies from LMICs, which reported 8%–47.4% of the homeless samples had severe mental disorders. Another review by Duke and Searby reported that 29.1%–41.4% of homeless women had posttraumatic stress disorder, whereas Hodgson and colleagues found the prevalence of psychiatric comorbidity in posttraumatic stress disorder ranged from 48% to 80.9%. Moreover, 40%–67.3% of homeless individuals had co-existing substance use disorder and posttraumatic stress disorder. Such co-existence of multiple mental disorders and other clinical conditions was reported in several reviews (Depp et al., 2015; Duke & Searby, 2019; Hodgson et al., 2013; Smartt et al., 2019; Spence et al., 2004).
Discussion
Overview of synthesized findings of this umbrella review
To our knowledge, this is the first umbrella review reporting the overall prevalence of different mental disorders among people who are homeless. The synthesized findings from existing evidence-based reviews inform a high burden of depressive disorders, anxiety disorders, schizophrenia spectrum and psychotic disorders, bipolar and mood disorders, substance use disorders, suicidal behavior and self-injury, posttraumatic stress disorders, neurocognitive disorders and other psychiatric conditions. In addition to the adults, high prevalence rates were found among children and adolescents who are homeless. Fewer studies reported high prevalence rates of mental disorders among homeless women (Duke & Searby, 2019). Most reviews had a higher number of primary studies from high-income countries, and the proportion of racial and ethnic minorities was higher in many reviews (Bassuk et al., 2015; Fazel et al., 2008; Schreiter et al., 2017; Spence et al., 2004). The prevalence rates were different across study samples, which necessitates an in-depth evaluation of the potential reasons contributing to such heterogeneity to better understand the findings of this review.
Psychiatric research on the homeless population is often constrained by several factors, which may have affected the existing evidence base in this domain. The working definition of homelessness may differ across contexts, which may affect the estimation of homeless individuals and evaluate any health problems among this mobile and vulnerable population (Fazel et al., 2014; Williams, 2017). For example, a study in the United States evaluated how changed definitions impacted estimations of homeless populations and found that a change in the definitions excluded nearly half of the chronically homeless individuals (Byrne & Culhane, 2015). Moreover, definitions and measurements of homelessness may also result in inaccurate estimations of mental health conditions. A study found that objectively defined homelessness was associated with higher rates of alcohol use and substance use disorders compared to subjectively reported homelessness (Eyrich-Garg et al., 2008). These differences are critical as the psychosocial epidemiology can be uniquely different among the included or excluded individuals based on such changes. It is essential to consider these issues while using the findings of this study as well as conducting future research in this domain.
Another issue is the changing definitions of mental disorders and instruments measuring the same, which may have resulted in different prevalence estimations across studies (Ayano, Tesfaw, & Shumet, 2019). Also, the low number of reviews and the median number of databases suggest a need to synthesize robust evidence from more data sources focusing on diverse mental health outcomes among homeless people.
The geographic and contextual focus of many studies provided evidence for those areas, whereas many nations with a high number of homeless populations may remain under-examined among the existing reviews, which include countries in South America, Sub-Saharan Africa and South Asia. This can be a result of a gap in evidence synthesis or a critical lack of primary studies conducted in those regions.
Furthermore, the co-existence of multiple mental disorders as well as physical comorbidity among the homeless with mental disorders highlights the severity of disability across samples (Depp et al., 2015; Duke & Searby, 2019; Hodgson et al., 2013; Spence et al., 2004). This may inform inadequate evidence if the primary studies measured only one or a few mental health conditions rather than a thorough evaluation of multiple health problems among the study samples.
Finally, limited evidence is found on how the mental health status changed among people before and after experiencing homelessness or how different the prevalence rates are between the homeless samples and the general population within the same geographic and sociocultural contexts. These issues should be considered to contextualize the findings of the current review and to inform future knowledge synthesis.
Implications for future research, policy development and practice
The findings of this review provide valuable insights to conduct future research, adopt appropriate policies and improve psychosocial care through better practice. First, longitudinal studies using standardized research instruments should be used to evaluate the mental health conditions and associated factors among homeless people, which may further improve the quantity and quality of knowledge in this area and inform evidence-based practice.
Second, psychopathological processes among socioeconomically marginalized populations who are vulnerable to homelessness should be examined. For example, people who experienced forced migration and associated psychosocial trauma may experience higher burden of mental disorders while living in temporary shelters or unstable social settings (Hossain & Purohit, 2018). It is essential to identify various psychosocial factors associated with mental health resilience and outcomes during homelessness, which may help in preventing mental disorders among homeless people through early psychosocial interventions (Hughes et al., 2010; Lee et al., 2011).
Third, it is necessary to examine how different mental disorders evolve over time and change their courses among homeless populations. Multiple psychiatric diagnoses among people who are already experiencing one or more neuropsychiatric conditions is a major global mental health concern (Hossain, Khan, et al., 2020; Hossain, Purohit, et al., 2020). In addition, homeless individuals generally experience instability in terms of residential, occupational, cultural, social and environmental aspects, which is likely to exert a compounding effect of psychosocial burden among those individuals. Therefore, the prognosis of mental disorders among homeless may not be similar to the general population living in comparatively stable conditions (Dawson & Jackson, 2013; Hodgson et al., 2013; Stubbs et al., 2019). Such variability among mental health prognosis may inform the psychiatrists, psychologists, social workers and other caregivers to provide adequate support based on the individual mental health needs.
Fourth, little is known about mental health-seeking behavior among homeless individuals. Identifying the barriers and facilitators of individuals and groups in terms of seeking mental care can be a potential research agenda for future studies, which may inform effective policy development in mental health. In this discourse, homelessness and associated mental health outcomes should be evaluated in respective contexts, especially how these constructs are defined and conceptualized in those places. It is necessary to assess and compare how such problems are different than the general population and what are the determinants of mental health in those contexts. Such contextualization of evidence may inform better decision-making to address mental health gaps across homeless populations.
Fifth, health services research should be conducted to examine how health systems and existing modalities of mental health care are equipped to address the burden of mental disorders among homeless people. This may require policy analyses and evaluations of mental health as well as social care programs to assess the effectiveness of the existing approaches, identify the gaps through rigorous research and address the same through evidence-based multipronged pharmacological and non-pharmacological interventions (Baxter et al., 2019; Fitzpatrick-Lewis et al., 2011).
Last but not least, improving the mental health outcomes among homeless people would necessitate addressing the underlying causes of homelessness and persistent psychosocial stressors (Embleton et al., 2016; Nooe & Patterson, 2010), which should be prioritized across communities and organizations. A meaningful partnership among major stakeholders can mobilize resources alleviating homelessness and improve mental health among homeless individuals.
Limitations
This umbrella review has several limitations. One such limitation is a potential selection bias as we did not search all the databases and excluded unpublished studies and reports. Another limitation is the publication bias within the scientific literature as less significant findings are less likely to get published, thus may not contribute to the evidence base. Moreover, we did not conduct a quantitative evaluation of the patient-level data, which could have eliminated between-study and within-study variations and provided uniform evidence of prevalence estimations across samples. However, this umbrella review was conducted using systematic guidelines involving more than two reviewers to ensure scientific rigor at each stage of review. Future evidence synthesis should address the existing limitations and advance the knowledge base in this domain.
Conclusion
Homelessness is a prevalent problem across societies with enormous psychosocial impacts on population health. This umbrella review systematically evaluated the current evidence on the prevalence of mental disorders among homeless people. The findings of this review inform a high epidemiological burden of mental disorders in homelessness, which requires multi-level interventions to address the same. Moreover, future research should be conducted to improve the evidence base on psychopathological correlates, processes and outcomes associated with homelessness across populations. The current evidence on mental health problems warrants a better understanding of the underlying socioeconomic challenges that impact overall health and well-being. Nonetheless, the definitions of homelessness and mental health conditions continue to evolve across places and over time, which is a continued challenge for synthesizing knowledge and translating the same in practice. It is essential to acknowledge and address the methodological and contextual issues that may inform a better understanding of mental health among homeless populations. Finally, evidence-based insights should be translated to future policies, programs and services envisaging effective prevention, diagnosis and treatment of mental disorders among people who are homeless.
Supplemental Material
Supplemental_file – Supplemental material for Prevalence of mental disorders among people who are homeless: An umbrella review
Supplemental material, Supplemental_file for Prevalence of mental disorders among people who are homeless: An umbrella review by Md Mahbub Hossain, Abida Sultana, Samia Tasnim, Qiping Fan, Ping Ma, E Lisako J McKyer and Neetu Purohit in International Journal of Social Psychiatry
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
