Key Points
Question
What is the prevalence of mental health disorders among people experiencing homelessness?
Findings
In this systematic review and meta-analysis, the prevalence of current and lifetime mental health disorders among people experiencing homelessness was high, with male individuals exhibiting a significantly higher lifetime prevalence of any mental health disorder compared to female individuals.
Meaning
These findings demonstrate that most people experiencing homelessness have mental health disorders, with current and lifetime prevalence generally much greater than that observed in general community samples.
This systematic review and meta-analysis assesses the prevalence of mental health disorders among people experiencing homelessness.
Abstract
Importance
Several factors may place people with mental health disorders, including substance use disorders, at increased risk of experiencing homelessness and experiencing homelessness may also increase the risk of developing mental health disorders. Meta-analyses examining the prevalence of mental health disorders among people experiencing homelessness globally are lacking.
Objective
To determine the current and lifetime prevalence of mental health disorders among people experiencing homelessness and identify associated factors.
Data Sources
A systematic search of electronic databases (PubMed, MEDLINE, PsycInfo, Embase, Cochrane, CINAHL, and AMED) was conducted from inception to May 1, 2021.
Study Selection
Studies investigating the prevalence of mental health disorders among people experiencing homelessness aged 18 years and older were included.
Data Extraction and Synthesis
Data extraction was completed using standardized forms in Covidence. All extracted data were reviewed for accuracy by consensus between 2 independent reviewers. Random-effects meta-analysis was used to estimate the prevalence (with 95% CIs) of mental health disorders in people experiencing homelessness. Subgroup analyses were performed by sex, study year, age group, region, risk of bias, and measurement method. Meta-regression was conducted to examine the association between mental health disorders and age, risk of bias, and study year.
Main Outcomes and Measures
Current and lifetime prevalence of mental health disorders among people experiencing homelessness.
Results
A total of 7729 citations were retrieved, with 291 undergoing full-text review and 85 included in the final review (N = 48 414 participants, 11 154 [23%] female and 37 260 [77%] male). The current prevalence of mental health disorders among people experiencing homelessness was 67% (95% CI, 55-77), and the lifetime prevalence was 77% (95% CI, 61-88). Male individuals exhibited a significantly higher lifetime prevalence of mental health disorders (86%; 95% CI, 74-92) compared to female individuals (69%; 95% CI, 48-84). The prevalence of several specific disorders were estimated, including any substance use disorder (44%), antisocial personality disorder (26%), major depression (19%), schizophrenia (7%), and bipolar disorder (8%).
Conclusions and Relevance
The findings demonstrate that most people experiencing homelessness have mental health disorders, with higher prevalences than those observed in general community samples. Specific interventions are needed to support the mental health needs of this population, including close coordination of mental health, social, and housing services and policies to support people experiencing homelessness with mental disorders.
Introduction
Mental health disorders, including substance use disorders, are among the most debilitating and costly conditions worldwide, affecting nearly 1 billion people and costing the world economy approximately $2.5 trillion annually.1 One population with particularly high rates of mental health disorders is people experiencing homelessness. While estimates vary and there are challenges to enumerating homelessness, there are more than 100 million people experiencing homelessness globally.2
Individual factors, such as substance use, relationship conflicts, and traumatic experiences, may contribute to homelessness among people with mental health disorders.3 Poverty, lack of affordable housing, transitions from foster care or institutional settings, stigma, racism, and discrimination may further perpetuate homelessness for these individuals.3 The relationship between mental health disorders and homelessness is complex and bidirectional4: mental health disorders may lead to situations that result in homelessness, or homelessness may be a stressor contributing to the development or worsening of mental health disorder symptoms. Homelessness is affected by a complex interplay of social determinants of health, including poor social and economic conditions.5 Homelessness is also associated with health inequalities, including higher morbidity, shorter life expectancy, and higher usage of health services.5
While mental health disorders are common among people experiencing homelessness, to our knowledge, there are currently no meta-analyses examining the prevalence of or factors associated with mental health disorders among people experiencing homelessness globally. The present systematic review and meta-analysis aimed to synthesize the existing literature examining the prevalence of mental health disorders among people experiencing homelessness, estimate the pooled prevalence of any mental health disorder and specific mental health disorders among people experiencing homelessness, and explore factors that may contribute to heterogeneity within these estimates.
Methods
Study Design
We registered this review with PROSPERO (CRD42021247246). The study followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline. We defined people currently experiencing homelessness as those meeting the criteria for homelessness in each included study, specifically at the time of assessment for mental health disorders in the individual component studies. Study criteria varied, but most studies specified that the individual must be living in a shelter or place not intended as a permanent dwelling. In addition, we defined individuals previously experiencing homelessness as those who had met the study criteria for homelessness in the past but were not experiencing homelessness at the time of assessment for mental health disorders.
The primary outcome was the current prevalence of any diagnosed mental health disorder. Secondary outcomes were the lifetime prevalence of any diagnosed mental health disorder, sex-specific prevalence of any mental health disorder, and the prevalence of individual mental health disorder diagnostic categories. Only mental health disorders diagnosed based on any version of the International Classification of Disease (ICD), the DSM, or validated diagnostic instruments were deemed eligible for review inclusion. We provide a list of such instruments in eTable 1 in Supplement 1.
We included English-language studies or those with English translations. We included cohort and cross-sectional designs and excluded case-control studies and reviews. We also excluded studies on suicide, self-harm, cognitive impairment (except dementia), neurological disorders, and smoking. Studies on pediatric populations (17 years and younger), specific populations (eg, veterans), small sample sizes (fewer than 50), hospital-based studies, and studies where the entire population likely had a mental health disorder (eg, studies examining the prevalence of mental health disorders among psychiatric clinic participants) were also excluded.
Information Sources and Search Strategy
We searched PubMed, MEDLINE, PsycInfo, Embase, Cochrane, CINAHL, and AMED from their inception to May 1, 2021. Free-text search terms included “mental illness,” “mood disorders,” “homelessness,” “homeless person,” “no fixed address,” “prevalence,” “incidence,” and terms that relate to specific mental health disorders; similar medical subject headings were applied to all databases permitting medical subject headings search strings. The full search strategy is available in eTable 2 in Supplement 1.
Selection Process
Search outputs from each database were imported into Covidence,6 a web-based systematic review manager. Duplicates were removed, and then 2 reviewers independently screened citations by title and abstract, followed by full-text screening using predefined criteria (R.B, L.T, A.B., and D.S). Discrepancies were resolved through consensus or consultation with a third reviewer. Interrater reliability was assessed using the κ statistic.
Data Collection
Data extraction was completed using prepopulated forms in Covidence. All extracted data were reviewed for accuracy by consensus between 2 independent reviewers (R.B, J.A, L.T., and D.S.). We extracted study details (author, journal, publication year, setting, study design, country, and follow-up duration), participant information (sample size, age, sex, ethnicity, and homelessness criteria), mental health disorder diagnosis (sources and diagnostic tools), and mental health disorder prevalence data (diagnoses, period, and sex- and diagnosis-specific estimates).
Risk of Bias Assessment
Risk of bias was appraised independently for each study by 2 reviewers (R.B, J.A, L.T., and D.S.) using the Newcastle-Ottawa Scale for cohort and cross-sectional studies.7 The risk of bias tool assesses bias related to the prevalence estimates of mental health disorder among people experiencing homelessness. We generated funnel plots and the Egger test for funnel plot asymmetry to quantify the potential for publication bias.8
Synthesis Methods
Descriptive statistics were calculated using proportions and means. In addition, data on the prevalence of mental health disorders were extracted. R version 4.2.1 in RStudio (R Foundation) was used for the meta-analyses, using the meta, metafor, and tidyverse software packages using a random-effects model and logit transformation.9 We used the generalized linear mixed effects model pooling method as recommended for meta-analysis of proportions.10,11 The primary meta-analysis synthesized the current prevalence of mental health disorders among people experiencing homelessness, generating a pooled estimate of prevalence and an accompanying 95% CI. Secondary meta-analyses examined the lifetime prevalence of mental health disorders and sex- and diagnosis-specific mental health disorder categories. Diagnosis-specific mental health disorder prevalence meta-analyses were only conducted if at least 3 representative studies were included.
Heterogeneity was quantified using the I2 statistic and its significance was determined based on the Cochran Q test P value (<.05). Heterogeneity was also quantified using τ212. We used Knapp-Hartung adjustments to calculate the confidence intervals around the pooled effects.9,13
For current prevalence of mental health disorders, we included the assessment closest to the past year if more than 1 time period was assessed in an individual study, with a maximum mental health disorder assessment time period of 3 years. Studies that reported mental health disorders among the whole population but oversampled by sex were not included in analyses examining the prevalence of mental health disorders in the general population although these studies were included in stratified sex-specific subgroup analyses.
Other subgroup analyses included region (continent), age group, risk of bias, study year, and mental health disorder diagnostic method. Meta-regression analyses explored the study-level association between the mental health disorder prevalence with the mean age of the study sample participants, study-level risk of bias assessment score, and study year.
Results
Study Selection
The initial search yielded 7729 citations, with 2655 duplicates. After title and abstract review, 291 studies underwent full-text evaluation with substantial interrater reliability (κ, 0.76) and 89% agreement.14 Eighty-five studies met all study criteria. Please refer to eFigure 1 in Supplement 1 for the PRISMA flow diagram, eTable 3 in Supplement 1 for study references, and eTable 4 in Supplement 1 for the PRISMA checklist.
Study Characteristics
Of 85 studies, 65 examined current mental health disorder prevalence, 44 examined lifetime mental health disorder prevalence, and 24 examined both. There were 48 414 people experiencing homelessness represented in the studies (11 154 [23%] female and 37 260 [77%] male). Nonresponse rates ranged from 2% to 64%, with 34 studies not reporting the response rate. Most studies were conducted in the US (n = 36), followed by Canada (n = 8) and Germany (n = 7) (eTable 5 in Supplement 1). Definitions of homelessness varied, with most including people in shelters or living in places not intended to be dwellings like streets (n = 71). Some studies limited the definition to those in shelters or seeking emergency housing (n = 9), while others used different definitions (n = 5), typically excluding those staying with friends or family. Common methods for examining mental health disorders included the Mini-International Neuropsychiatric Interview (n = 20), the Diagnostic Interview Schedule (n = 18), the Structured Clinical Interview for DSM (n = 13), the Composite International Diagnostic Interview (n = 10), and electronic medical records (n = 8), with several using multiple methods. The most common definitions were based on DSM-IV/DSM-IV-TR (n = 39), DSM-III/DSM-III-R (n = 36), and ICD-10 (n = 22).
Risk of Bias in Studies
Many studies had low response rates and nonvalidated homelessness measures, but all included studies used validated methods for diagnosing mental health disorders. Most controlled for sex through restriction or stratification. Details on the risk of bias for all studies are provided in eTable 6 in Supplement 1. Publication bias assessments were insignificant, as the Egger test did not indicate funnel plot asymmetry (intercept, −0.51; 95% CI, −5.1 to 4.1; t, −0.22; P = .83) (eFigure 9 in Supplement 1).
Results of Individual Studies and Syntheses
Crude Current and Lifetime Prevalence of Mental Health Disorders Among People Experiencing Homelessness
Forty studies reported the current prevalence of mental health disorders among people experiencing homelessness, with 16 included in the meta-analysis, as many were restricted by sex or had oversampled one sex. The pooled current prevalence of mental health disorders among people experiencing homelessness was 67% (95% CI, 55-77; n = 16) (Figure15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30). In contrast, the pooled prevalence of lifetime mental health disorders was 77% (95% CI, 61-88; n = 8) (eFigure 2 in Supplement 1).
Figure. Current Prevalence of Mental Health Disorders Among People Experiencing Homelessness.
Sex-Specific Current and Lifetime Prevalence of Mental Health Disorders Among People Experiencing Homelessness
The pooled prevalence of current mental health disorders was 57% (95% CI, 42-71) among female individuals experiencing homelessness and 67% (95% CI, 58-76) among male individuals (Table 1; eFigure 3 in Supplement 1). The lifetime prevalence of mental health disorders was 69% (95% CI, 48-84) among female individuals experiencing homelessness and 86% (95% CI, 74-92) among male individuals experiencing homelessness (P = .03) (Table 1; eFigure 4 in Supplement 1).
Table 1. Summary of Subgroup Analyses of Prevalence of Mental Health Disorders .
| Grouping | Prevalence, % (95% CI) | χ2 | df | I2, % | τ2 | P value |
|---|---|---|---|---|---|---|
| Current prevalence | ||||||
| Overall | 63.5 (55.7-70.7) | 2.0 | 1 | NA | NA | NA |
| Male | 67.4 (58.1-75.5) | 168.0 | 10 | 94 | 0.32 | .16 |
| Female | 57.1 (42.3-71) | 105.2 | 8 | 93 | 0.43 | |
| Lifetime prevalence | ||||||
| Overall | 80.2 (70.4-87.3) | 4.7 | 1 | NA | ||
| Male | 84.9 (74.1-94.7) | 126.9 | 8 | 94 | 0.71 | .03 |
| Female | 68.9 (48.1-84.1) | 67.0 | 4 | 94 | 0.43 |
Abbreviation: NA, not applicable.
Diagnosis-Specific Current and Lifetime Prevalence of Mental Health Disorders Among People Experiencing Homelessness
The current and lifetime prevalence for each of the specific diagnoses are outlined in Table 2. The most prevalent mental health disorder was substance use disorder, with a current prevalence of 44% (95% CI, 30-59) and a lifetime prevalence of 56% (95% CI, 43-68) (Table 2), followed by antisocial personality disorder (26%), major depression (19%), and any mood disorder (18%).
Table 2. Individual Diagnoses (Both Sexes, Current Homelessness).
| Disordera | Current prevalence, % (95% CI) | No. of studies | Lifetime prevalence (95% CI) | No. of studies |
|---|---|---|---|---|
| Severe mental illness | ||||
| Schizophrenia | 7.1 (4.2-11.6) | 14 | 7.6 (4.6-12.3) | 9 |
| Psychotic disorder | 13.9 (9.3-20.4) | 12 | 17.4 (9.2-30.6) | 5 |
| Bipolar disorder | 8.0 (2.4-24.6) | 8 | 8.1 (4.0-15.8) | 6 |
| Mood disorders | ||||
| Mood disorder (general, including bipolar disorder) | 17.8 (11.6-26.3) | 15 | 23.8 (15.6-34.5) | 7 |
| Major depression | 18.9 (13.8-25.5) | 12 | 23.9 (22.1-25.9) | 9 |
| Dysthymia | 9 (3.4-21.6) | 5 | 12.9 (8.0-20.2) | 4 |
| Anxiety disorders | ||||
| Anxiety disorders (general) | 14.1 (9.2-20.9) | 9 | NA | NA |
| Generalized anxiety disorder | 8.2 (4.9-13.4) | 5 | 14.0 (4.9-34.2) | 4 |
| Social phobia | 10.4 (2.2-37.2) | 3 | NA | NA |
| Other disorders | ||||
| Posttraumatic stress disorder | 10.5 (4-25.2) | 7 | NA | NA |
| Obsessive-compulsive disorder | 2.8 (0.5-1.3) | 4 | NA | NA |
| Antisocial personality disorder | 26.1 (14.2-43.0) | 6 | 17.8 (6.6-39.8) | 7 |
| Substance use disorders | ||||
| Including alcohol | 43.6 (29.5-58.8) | 16 | 55.8 (42.5-68.3) | 12 |
| Excluding alcohol | 31.7 (17.1-51.1) | 20 | 29.2 (19.4-41.5) | 13 |
| Alcohol use disorder | 26.1 (19.0-34.6) | 20 | 46.0 (33.0-59.6) | 11 |
Abbreviation: NA, not applicable.
Diagnostic categories and subcategories are partially overlapping and individual studies reported on some or all possible diagnostic categories.
Investigation of Possible Causes of Heterogeneity
The prevalence of mental health disorders across various subgroups is reported in Table 3. Studies with a higher risk of bias, as measured by the Newcastle-Ottawa Scale, also reported a higher prevalence of mental health disorders among people experiencing homelessness (75.4% vs 54.2%; P = .02). The prevalence of mental health disorders among people experiencing homelessness in studies published before 2010 was lower than in studies published more recently (48.0% vs 76.2%; P < .001). Regional differences in the prevalence of mental health disorders were also observed, with North America having the highest prevalence at 77% (eFigures 5-8 in Supplement 1) compared to other regions. The year of publication was a significant predictor of prevalence in meta-regression (β, 0.06; 95% CI, 0.02-0.09; P < .001) (Table 4). Subgroup analysis by income level of countries could not be undertaken due to limited studies in low- and middle-income countries.
Table 3. Subgroup Analyses.
| Prevalence, % (95% CI) | df | χ2 | I2, % | τ2 | P value | |
|---|---|---|---|---|---|---|
| Population age group | 1 | |||||
| Young adult population | 73.1 (29.7-94.6) | 2 | 21.6 | 90.8 | 0.48 | .41 |
| General adult population | 64.2 (50.4-76.0) | 13 | 641.7 | 98.0 | 0.93 | |
| Region | 3 | 14.5 | ||||
| North America | 76.6 (58.2-88.5) | 6 | 145.5 | 95.9 | 0.79 | .002 |
| Europe | 60.3 (30.5-84.0) | 4 | 186.8 | 97.3 | 0.97 | |
| Asia | 58.7 (23.5-86.9) | 2 | 25.9 | 33.7 | 0.34 | |
| Australia | 46.8 (41.3-52.4) | 0 | NA | NA | NA | |
| Risk of bias | 1 | |||||
| High (1-4 stars) | 75.4 (58.6-67.7) | 8 | 205.8 | 96.1 | 0.96 | .02 |
| Low (5-8 stars) | 54.2 (39.9-67.7) | 6 | 147.7 | 95.9 | 0.36 | |
| Study year | 1 | 12.3 | ||||
| Before 2010 | 48.0 (35.0-61.4) | 5 | 84.7 | 94.1 | 0.23 | <.001 |
| After 2010 | 76.2 (62.7-86.0) | 9 | 212.5 | 95.8 | 0.76 | |
| Method of measuring mental health disorder | 1 | |||||
| Administered testing to full study population | 64.5 (51.0-76.0) | 13 | 454.9 | 97.1 | 0.88 | .16 |
| Administrative data diagnoses | 80.0 (7.3-100.0) | 1 | 28.8 | 96.5 | 0.46 |
Abbreviation: NA, not applicable.
Table 4. Sensitivity Analyses for Current Prevalence of Mental Health Disorders Using Meta-Regression.
| Variable | Estimate (95% CI) | SE | df | P value (meta-regression) |
|---|---|---|---|---|
| Mean age (continuous) | −0.02 (−0.07 to 0.03) | 0.02 | 13 | .37 |
| Risk of bias (continuous) | −0.38 (−0.81 to 0.04) | 0.20 | 14 | .07 |
| Study year (continuous) | 0.06 (0.02 to 0.09) | 0.02 | 14 | .009 |
Discussion
The findings in this systematic review and meta-analysis indicate that 67% of people experiencing homelessness had a current mental health disorder while the lifetime prevalence of mental health disorders among people experiencing homelessness was 75%. The prevalence of lifetime mental health disorders was higher among male individuals experiencing homelessness compared to female individuals. The most common mental health disorders included substance use disorders, antisocial personality disorders, major depression, and general mood disorders, and estimates of mental health disorders were higher than that reported in general community samples.31,32,33,34,35,36,37,38,39
Our results are similar to those of a previous review,40 which estimated that 76% of people experiencing homelessness living in high-income countries experience mental health disorders. However, our review included more studies overall (85 vs 39) and in our meta-analysis (16 vs 8). Most population-based studies estimate the current prevalence of mental health disorders to be approximately 13% to 15% and the lifetime prevalence to be between 12% and 47%.31,32 Additionally, specific mental health disorder prevalence among people experiencing homelessness far exceeds that in the general population, including major depression (19% vs 5%), psychotic disorder (14% vs 0.4%), general mood disorders (18% vs %1-10), bipolar disorder (8% vs 0.7%), posttraumatic stress disorder (10.5% vs 1.1%), and substance use disorders (44% vs 12%-15%).33,34,35,36,37,38,39 The prevalence of antisocial personality disorder was reported as being particularly high among people experiencing homelessness in the studies included our review when compared to the general population (26% vs 0.6%-4%).38 This finding may be due to common risk factors for homelessness and antisocial personality disorder, such as poverty or early life trauma.41,42 In addition, the diagnostic criteria for antisocial personality disorder can include a failure to maintain work and disregard for personal safety, which may be challenging to disentangle from the circumstances in which people experiencing homelessness live unintentionally.43
This review identifies gaps in current research related to mental health disorder research among people experiencing homelessness. In contrast to previous reviews, ours included studies from low- and middle-income countries, although there were only a few studies from these countries (n = 5). Also, validated or standardized definitions of homelessness are lacking, likely contributing to the heterogeneity in the prevalence estimates. Response rates are often low or unreported in primary studies, and there is a shortage of research on individuals with past experiences of homelessness. Future studies should address these limitations to build upon the current research.
An important finding from our review is the higher prevalence of mental health disorders among people experiencing homelessness among studies published more recently. This could be due to changing study types or better mental health care access leading to increased diagnoses. Alternatively, it may reflect overall temporal trends in prevalence of mental health disorders in the general population, but global mental disorder prevalence remained appears to have been relatively stable recently, except in the US among younger adults.44,45 Another possibility is that people with mental health disorders are increasingly likely to experience homelessness due to factors like housing affordability, limited mental health care access, low income, or challenges reintegrating after mental health treatment.46,47 Conversely, the increase in prevalence could be the result of people experiencing homelessness being more likely to develop a mental health disorder due to stressors related to being unhoused, given the bidirectional relationship.4 North America has the highest mental health disorder prevalence among people experiencing homelessness, likely influenced by various factors like mental health care access, cultural differences, social support, housing affordability, and working conditions.46,47 Additionally, higher-income countries may report higher rates of mental health disorders, possibly due to increased stigma in lower- and middle-income nations, leading to underreporting.45 Regardless, our finding that there is increasing prevalence of mental health disorders among people experiencing homelessness is concerning and effective strategies are needed to address the significant mental health needs of this population.
Given the relationships between mental health disorders and homelessness, it is crucial to have health and social strategies to both address the mental health needs of people experiencing homelessness and to effectively support people with mental health disorders to prevent them from experiencing homelessness. Various interventions have been developed to support mental health disorders among people experiencing homelessness. The most extensively studied are referred to as housing first strategies, which provide housing supports without requiring any specific treatments for mental health disorders.3,4 Housing first strategies are effective at improving housing retention but demonstrate mixed results for outcomes such as reduced substance use and improved mental health.48,49 Critical time interventions are another approach that connects individuals to mental health supports during specific times, such as transitions from shelters or inpatient psychiatric care.4,10 While both housing first and critical time interventions improve housing retention, their impact on mental health disorders varies.10,48 Assertive community treatment, involving a multidisciplinary team providing intensive community-based mental health services, has demonstrated effectiveness in reducing mental health disorder severity and homelessness among people with persistent and difficult to treat mental health disorders.50 However, more evidence-based methods are needed to lower mental health disorder risk and severity among those experiencing or at risk of homelessness and existing models of care need to be more widely implemented to reduce the prevalence of mental health disorders among people experiencing homelessness.51
Implications of Findings
Our findings reveal a high and potentially increasing prevalence of mental health disorders among people experiencing homelessness. Integrated interventions are vital, focusing on substance use disorders, antisocial personality disorder, major depression, and psychotic disorders—conditions more common in people experiencing homelessness than in the general population. Sex-specific approaches are needed, as male individuals experiencing homelessness had a higher prevalence of mental health disorders than female individuals, but addressing mental health disorders for both sexes in shelters and support services is crucial. Complex, multicomponent interventions are necessary due to the multifaceted nature of homelessness and its health disparities. While housing first and critical time interventions and assertive community treatment programs show promise in improving housing retention, additional research is needed to identify interventions that most effectively address mental health outcomes among people experiencing homelessness.
Strengths and Limitations
This systematic review and meta-analysis has several strengths. First, it only includes studies that use a validated measure of mental health disorders to ascertain the outcomes, reducing measurement bias.7 Only diagnosed mental health disorders (rather than general measures of mental wellness) were considered. Second, the large number of studies included in this review allowed us to examine specific subgroups. We also examined a broader range of mental health disorders than previous reviews and examined both current and lifetime prevalence. Third, this review did not restrict by country, allowing us to examine the prevalence of mental health disorders among people experiencing homelessness worldwide and regionally. Our review also examined the prevalence of mental health disorders among people experiencing homelessness by sex. Moreover, the review included studies that were conducted during a broad range of dates to help determine if prevalence rates have changed over time.
This review also has limitations. Obtaining a representative sample of people experiencing homelessness is challenging, and many studies potentially miss the so-called hidden homeless who do not access homeless services and may be living in cars or staying with friends or family.52 Future studies should aim to include these populations by sampling from locations identified by people experiencing homelessness as locations that are frequented by this population. Heterogeneity in estimates exists due to geographical, cultural, measurement, and sampling differences. Some studies rely on health care records, likely underestimating mental health disorder prevalence because some people with mental health disorders may not seek care or mental health disorders may not be coded as the primary reasons for health visits. Several studies had high nonresponse rates. We included these studies, as nonresponse rates only lead to bias when missing data differs systematically from measured data, which is difficult to assess in the individual studies.53 Duration and frequency of homelessness were not examined in many studies, and exclusion of individuals with severe mental illness who could not provide consent may underestimate mental health disorder prevalence.
Conclusions
This systematic review and meta-analysis highlights the high prevalence of mental health disorders among people experiencing homelessness among all subgroups and regions that have reported on this. Future research should explore factors increasing mental health disorder risk in people experiencing homelessness and homelessness risk in those with mental health disorders. Additionally, evaluating interventions to reduce mental health disorders among people experiencing homelessness is essential. Efforts to improve access to evidence-based programs to prevent homelessness among people with mental health disorders or improve access to mental health care for people who have mental health disorders and who are currently experiencing homelessness is critical to reduce the burden of mental health disorders among this vulnerable population.
eTable 1. Valid AMH Indices
eTable 2. Search Strategy
eTable 3. Systematic Review Study References
eTable 4. PRISMA Checklist
eTable 5. Countries Represented
eTable 6. Risk of Bias Analysis
eFigure 1. PRISMA Flow Diagram for Study Eligibility
eFigure 2. Lifetime Prevalence of MHDs among People Experiencing Homelessness
eFigure 3. Current Prevalence of MHDs by Sex
eFigure 4. Lifetime Prevalence of MHDs by Sex
eFigure 5. Current Prevalence of MHDs by Continent
eFigure 6. Current Prevalence of MHDs by Youth vs. Non-Youth Populations
eFigure 7. Current Prevalence of MHDs among Studies that Examine MHDs Using Standardized Tests Administered to the Entire Study Population vs. Studies that use Healthcare Administrative Data
eFigure 8. Prevalence of MHDs Pre- and Post-2010
eFigure 9. Funnel Plot Examining Publication Bias Among Studies Examining Current Prevalence of MHDs Among Both Sexes
Data sharing statement
References
- 1.The Lancet Global Health . Mental health matters. Lancet Glob Health. 2020;8(11):e1352. doi: 10.1016/S2214-109X(20)30432-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.United Nations . SDH indicator metadata 2021. Accessed July 15, 2023. https://unstats.un.org/sdgs/metadata/files/Metadata-11-01-01.pdf
- 3.Piat M, Polvere L, Kirst M, et al. Pathways into homelessness: understanding how both individual and structural factors contribute to and sustain homelessness in Canada. Urban Stud. 2014;52(13):2366-2382. doi: 10.1177/0042098014548138 [DOI] [Google Scholar]
- 4.Padgett DK. Homelessness, housing instability and mental health: making the connections. BJPsych Bull. 2020;44(5):197-201. doi: 10.1192/bjb.2020.49 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stafford A, Wood L. Tackling health disparities for people who are homeless? start with social determinants. Int J Environ Res Public Health. 2017;14(12):1535. doi: 10.3390/ijerph14121535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Veritas Health Innovation . Covidence systematic review software. Accessed December 20, 2023. https://www.covidence.org.
- 7.Wells G, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analysis. Ottawa Hospital Research Institute. Accessed May 20, 2023. https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
- 8.Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. doi: 10.1136/bmj.315.7109.629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Harrer M, Cuijpers P, Furukawa T, Ebert D. Doing Meta-Analysis with R: A Hands-On Guide. CRC Press; 2022. [Google Scholar]
- 10.Manuel JI, Nizza M, Herman DB, et al. Supporting vulnerable people during challenging transitions: a systematic review of critical time intervention. Adm Policy Ment Health. 2023;50(1):100-113. doi: 10.1007/s10488-022-01224-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schwarzer G, Chemaitelly H, Abu-Raddad LJ, Rücker G. Seriously misleading results using inverse of Freeman-Tukey double arcsine transformation in meta-analysis of single proportions. Res Synth Methods. 2019;10(3):476-483. doi: 10.1002/jrsm.1348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Viechtbauer W. Bias and efficiency of meta-analytic variance estimators in the random-effects model. J Educ Behav Stat. 2005;30(3):261-293. doi: 10.3102/10769986030003261 [DOI] [Google Scholar]
- 13.Knapp G, Hartung J. Improved tests for a random effects meta-regression with a single covariate. Stat Med. 2003;22(17):2693-2710. doi: 10.1002/sim.1482 [DOI] [PubMed] [Google Scholar]
- 14.McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb). 2012;22(3):276-282. doi: 10.11613/BM.2012.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Awirutworakul T, Pitukthanin A, Chiangchaisakulthai K, Anukul C, Vallibhakara SAO. Prevalence of psychiatric disorders in homeless population in Bangkok, Thailand. J Med Assoc Thai. 2018;101(8):1065-1071. [Google Scholar]
- 16.Castro AL, Gustafson EL, Ford AE, et al. Psychiatric disorders, high-risk behaviors, and chronicity of episodes among predominantly African American homeless Chicago youth. J Health Care Poor Underserved. 2014;25(3):1201-1216. doi: 10.1353/hpu.2014.0124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ferguson A. Examining risk factors of chronic homelessness: mental illness, substance abuse, trauma, and criminal history. Diss Abstr Int B Sci Eng. 2014;74(11-B(E)). [Google Scholar]
- 18.Fischer PJ, Shapiro S, Breakey WR, Anthony JC, Kramer M. Mental health and social characteristics of the homeless: a survey of mission users. Am J Public Health. 1986;76(5):519-524. doi: 10.2105/AJPH.76.5.519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Herrman HE, McGorry PD, Bennett PA, Singh BS. Age and severe mental disorders in homeless and disaffiliated people in inner Melbourne. Med J Aust. 1990;153(4):197-200, 204-205. doi: 10.5694/j.1326-5377.1990.tb136859.x [DOI] [PubMed] [Google Scholar]
- 20.Kovess V, Lazarus CM. Use of mental health services by homeless people in Paris. Int J Ment Health. 2001;30(3):26-39. doi: 10.1080/00207411.2001.11449523 [DOI] [Google Scholar]
- 21.Lin WC, Bharel M, Zhang J, O’Connell E, Clark RE. Frequent emergency department visits and hospitalizations among homeless people with medicaid: implications for Medicaid expansion. Am J Public Health. 2015;105(Suppl 5):S716-S722. doi: 10.2105/AJPH.2015.302693 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Madianos MG, Chondraki P, Papadimitriou GN. Prevalence of psychiatric disorders among homeless people in Athens area: a cross-sectional study. Soc Psychiatry Psychiatr Epidemiol. 2013;48(8):1225-1234. doi: 10.1007/s00127-013-0674-2 [DOI] [PubMed] [Google Scholar]
- 23.Nishio A, Yamamoto M, Horita R, et al. Prevalence of mental illness, cognitive disability, and their overlap among the homeless in Nagoya, Japan. PLoS One. 2015;10(9):e0138052. doi: 10.1371/journal.pone.0138052 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Salize HJ, Horst A, Dillmann-Lange C, et al. Needs for mental health care and service provision in single homeless people. Soc Psychiatry Psychiatr Epidemiol. 2001;36(4):207-216. doi: 10.1007/s001270170065 [DOI] [PubMed] [Google Scholar]
- 25.Saperstein AM, Lee S, Ronan EJ, Seeman RS, Medalia A. Cognitive deficit and mental health in homeless transition-age youth. Pediatrics. 2014;134(1):e138-e145. doi: 10.1542/peds.2013-4302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Schütz C, Choi F, Jae Song M, Wesarg C, Li K, Krausz M. Living with dual diagnosis and homelessness: marginalized within a marginalized group. J Dual Diagn. 2019;15(2):88-94. doi: 10.1080/15504263.2019.1579948 [DOI] [PubMed] [Google Scholar]
- 27.Stenius-Ayoade A, Haaramo P, Erkkilä E, et al. Mental disorders and the use of primary health care services among homeless shelter users in the Helsinki metropolitan area, Finland. BMC Health Serv Res. 2017;17(1):428. doi: 10.1186/s12913-017-2372-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Toro PA, Wall DD. Research on homeless persons: diagnostic comparisons and practice implications. Prof Psychol Res Pr. 1991;22(6):479-488. doi: 10.1037/0735-7028.22.6.479 [DOI] [Google Scholar]
- 29.Vázquez C, Muñoz M, Sanz J. Lifetime and 12-month prevalence of DSM-III-R mental disorders among the homeless in Madrid: a European study using the CIDI. Acta Psychiatr Scand. 1997;95(6):523-530. doi: 10.1111/j.1600-0447.1997.tb10141.x [DOI] [PubMed] [Google Scholar]
- 30.Yim LCL, Leung HCM, Chan WC, Lam MHB, Lim VWM. Prevalence of mental illness among homeless people in Hong Kong. PLoS One. 2015;10(10):e0140940. doi: 10.1371/journal.pone.0140940 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Castaldelli-Maia JM, Bhugra D. Analysis of global prevalence of mental and substance use disorders within countries: focus on sociodemographic characteristics and income levels. Int Rev Psychiatry. 2022;34(1):6-15. doi: 10.1080/09540261.2022.2040450 [DOI] [PubMed] [Google Scholar]
- 32.Kessler RC, Angermeyer M, Anthony JC, et al. Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative. World Psychiatry. 2007;6(3):168-176. [PMC free article] [PubMed] [Google Scholar]
- 33.The World Health Organization . Depressive disorder (depression). Accessed August 25, 2023. https://www.who.int/news-room/fact-sheets/detail/depression.
- 34.Ferrari AJ, Stockings E, Khoo JP, et al. The prevalence and burden of bipolar disorder: findings from the Global Burden of Disease Study 2013. Bipolar Disord. 2016;18(5):440-450. doi: 10.1111/bdi.12423 [DOI] [PubMed] [Google Scholar]
- 35.Moreno-Küstner B, Martín C, Pastor L. Prevalence of psychotic disorders and its association with methodological issues. a systematic review and meta-analyses. PLoS One. 2018;13(4):e0195687. doi: 10.1371/journal.pone.0195687 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Merikangas KR, McClair VL. Epidemiology of substance use disorders. Hum Genet. 2012;131(6):779-789. doi: 10.1007/s00439-012-1168-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Demyttenaere K, Bruffaerts R, Posada-Villa J, et al. ; WHO World Mental Health Survey Consortium . Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA. 2004;291(21):2581-2590. doi: 10.1001/jama.291.21.2581 [DOI] [PubMed] [Google Scholar]
- 38.Lenzenweger MF, Lane MC, Loranger AW, Kessler RC. DSM-IV personality disorders in the National Comorbidity Survey Replication. Biol Psychiatry. 2007;62(6):553-564. doi: 10.1016/j.biopsych.2006.09.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Karam EG, Friedman MJ, Hill ED, et al. Cumulative traumas and risk thresholds: 12-month PTSD in the World Mental Health (WMH) surveys. Depress Anxiety. 2014;31(2):130-142. doi: 10.1002/da.22169 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Gutwinski S, Schreiter S, Deutscher K, Fazel S. The prevalence of mental disorders among homeless people in high-income countries: an updated systematic review and meta-regression analysis. PLoS Med. 2021;18(8):e1003750. doi: 10.1371/journal.pmed.1003750 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Dell NA, Vaughn MG, Huang J, Mancini M, Maynard BR. Correlates of homelessness among adults with personality disorder. Psychiatr Q. 2023;94(2):281-295. doi: 10.1007/s11126-023-10027-w [DOI] [PubMed] [Google Scholar]
- 42.Davies BR, Allen NB. Trauma and homelessness in youth: psychopathology and intervention. Clin Psychol Rev. 2017;54:17-28. doi: 10.1016/j.cpr.2017.03.005 [DOI] [PubMed] [Google Scholar]
- 43.Fisher KA, Hany M. Antisocial Personality Disorder. In: StatPearls. StatPearls Publishing; May 8, 2023. [PubMed] [Google Scholar]
- 44.Twenge JM, Cooper AB, Joiner TE, Duffy ME, Binau SG. Age, period, and cohort trends in mood disorder indicators and suicide-related outcomes in a nationally representative dataset, 2005-2017. J Abnorm Psychol. 2019;128(3):185-199. doi: 10.1037/abn0000410 [DOI] [PubMed] [Google Scholar]
- 45.Freeman M. The World Mental Health Report: transforming mental health for all. World Psychiatry. 2022;21(3):391-392. doi: 10.1002/wps.21018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Elliott M, Krivo L. Structural determinants of homelessness in the United States. Soc Probl. 1991;38(1):113-131. doi: 10.2307/800641 [DOI] [Google Scholar]
- 47.Mago VK, Morden HK, Fritz C, et al. Analyzing the impact of social factors on homelessness: a fuzzy cognitive map approach. BMC Med Inform Decis Mak. 2013;13:94. doi: 10.1186/1472-6947-13-94 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Woodhall-Melnik J, Dunn J. A systematic review of outcomes associated with participation in Housing First programs. Housing Stud. 2016;31(3):287-304. doi: 10.1080/02673037.2015.1080816 [DOI] [Google Scholar]
- 49.Busch-Geertsema V. Housing First Europe: final report. Housing First Europe . Accessed December 23, 2023. http://www.habitat.hu/files/FinalReportHousingFirstEurope.pdf
- 50.Coldwell CM, Bender WS. The effectiveness of assertive community treatment for homeless populations with severe mental illness: a meta-analysis. Am J Psychiatry. 2007;164(3):393-399. doi: 10.1176/ajp.2007.164.3.393 [DOI] [PubMed] [Google Scholar]
- 51.Stergiopoulos V, Schuler A, Nisenbaum R, et al. The effectiveness of an integrated collaborative care model vs. a shifted outpatient collaborative care model on community functioning, residential stability, and health service use among homeless adults with mental illness: a quasi-experimental study. BMC Health Serv Res. 2015;15:348. doi: 10.1186/s12913-015-1014-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Deleu H, Schrooten M, Hermans K. Hidden homelessness: a scoping review and avenues for further inquiry. Soc Policy Soc. 2023;22:282-298. doi: 10.1017/S1474746421000476 [DOI] [Google Scholar]
- 53.Higgins JPT, Savović J, Page MJ, Elbers RG, Sterne JAC. Chapter 8: Assessing risk of bias in a randomized trial. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 6.4. Cochrane, 2023. http://www.training.cochrane.org/handbook [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Valid AMH Indices
eTable 2. Search Strategy
eTable 3. Systematic Review Study References
eTable 4. PRISMA Checklist
eTable 5. Countries Represented
eTable 6. Risk of Bias Analysis
eFigure 1. PRISMA Flow Diagram for Study Eligibility
eFigure 2. Lifetime Prevalence of MHDs among People Experiencing Homelessness
eFigure 3. Current Prevalence of MHDs by Sex
eFigure 4. Lifetime Prevalence of MHDs by Sex
eFigure 5. Current Prevalence of MHDs by Continent
eFigure 6. Current Prevalence of MHDs by Youth vs. Non-Youth Populations
eFigure 7. Current Prevalence of MHDs among Studies that Examine MHDs Using Standardized Tests Administered to the Entire Study Population vs. Studies that use Healthcare Administrative Data
eFigure 8. Prevalence of MHDs Pre- and Post-2010
eFigure 9. Funnel Plot Examining Publication Bias Among Studies Examining Current Prevalence of MHDs Among Both Sexes
Data sharing statement