Sullivan et al.  explored pathways to homelessness for mentally ill persons by examining mental illness as a risk factor for homelessness as distinct from other personal vulnerabilities (such as histories of poverty, abuse, or family instability) that are likely to increase the risk of homeless-ness when affordable housing is in short supply. Since a longitudinal study of a community sample over many years, with repeated assessments of all potential risk factors including mental illness, would be prohibitively expensive, they used data from two existing data sets: the COH project (described in Sullivan et al. ) and the National Epidemiological Catchment Area (ECA) Survey . To examine pathways to homelessness, they conducted three analyses. First, they compared and contrasted three groups: the mentally ill homeless, the non-mentally-ill homeless (obtained from the COH study), and the mentally ill housed (obtained from the ECA study). The ECA survey, conducted in five sites across the USA between 1980 and 1984, was designed to estimate the prevalence of mental disorders in both treated and non-treated community populations. Data from the Los Angeles ECA site of the non-institutionalized (n = 2901) were used.
The comparisons revealed that the mentally ill homeless are more demo-graphically similar to the non-mentally-ill homeless than they are to mentally ill housed persons. Current alcohol and drug dependence follow a similar pattern. Like the non-mentally-ill homeless, the mentally ill homeless are at very high risk of substance abuse. Homeless subjects have almost twice the prevalence of alcohol dependence and six times the prevalence of drug abuse of housed subjects. These comparisons show that homeless persons, whether or not they are mentally ill, are more likely to be socially disadvantaged (less educated, ethnic minorities) and to have a high likelihood to be currently dependent on alcohol or drugs.
Homeless persons appear to have experienced considerable poverty in childhood. About one in five stated that their family was on welfare and that their primary caregiver was never or rarely employed. The mentally ill homeless did not differ significantly from the non-mentally-ill homeless in terms of childhood poverty. However, the mentally ill homeless did experience significantly more family and home instability. Of the mentally ill homeless sample, 60% had a primary caregiver who was either mentally ill or physically disabled, and more than one out of four were placed at least once in an institution or foster care. Furthermore, mentally ill homeless persons were also more likely to come from backgrounds marked by physical or sexual abuse. Compared with the non-mentally-ill homeless, twice as many mentally ill homeless (almost 40% of the sample) reported having lived in a household where violence or abuse took place regularly. One-third had actually been physically abused, while 5% reported having been sexually abused. Both physical abuse (19% vs. 13%) and sexual abuse (12% vs. 1%) were more frequent in women. By logistic regression, five factors uniquely associated with being mentally ill were identified: having been physically abused (OR = 2.88; P < 0.0001), being white (OR = 1.78; P < 0.0001), residential instability in childhood (OR = 1.60; P = 0.005), caregiver illness (OR = 1.39; P = 0.02), and having some college education (OR = 1.38; P = 0.02).
The authors stated that the relationship between homelessness and mental illness is rather complex. In some ways the mentally ill homeless appear to be more privileged (better educated, less likely to be of minority ethnicity) than other homeless persons. On the other hand, they share with other homeless people backgrounds marked by poverty: dependency on welfare, childhood hunger and family unemployment. The mentally ill homeless appear to have more in common with other homeless people than they do with the mentally ill housed population. Furthermore, home-lessness appears to be a phenomenon rooted in the impoverished and disadvantaged backgrounds of homeless people regardless of their subsequent mental health status.
However, the mentally ill homeless are distinct in terms of childhood risk factors. They have significantly higher scores on every indicator of childhood family instability and violence or abuse. About one-fourth of the mentally ill homeless experienced residential instability with their family as a child, about one-fourth were placed out of their homes, and more than one-third either witnessed violence within the household or personally experienced abuse. The authors conclude that the mentally ill homeless have received a "double dose" of disadvantage— poverty with the addition of childhood family instability and violence.
Sullivan et al.'s analyses  do not support the notion that mental illness represents a distinctive pathway to homelessness, but rather that the relationship between mental illness and homelessness is both complex and dynamic. While programmes that attempt to improve the symptoms and functioning of homeless adults and to alleviate the chronic stresses of homelessness certainly help some individuals, they fail to address the deeper origins of homelessness, arising from both the structural and personal vulnerabilities that exist for all homeless people. For the subpopulation of seriously mentally ill adults, effective interventions to prevent or treat substance abuse appear to be important in reducing the risk of home-lessness. Consequently, programmes designed to help the adult mentally ill homeless should be coupled with programmes that address childhood risk factors for homelessness and readdress the structural changes that underlie contemporary homelessness.
Was this article helpful?