Homeless and incarcerated: An epidemiological study from Canada

2

International Journal of Social Psychiatry and homeless are so enmeshed that this population has been observed to be at the greatest risk of suffering from mental health issues and substance abuse (Hawthorne et al., 2012;Hickey, 2002;McNeil, Binder, & Robinson, 2005;White et al., 2006) and therefore needs closer attention in order to identify their specific vulnerabilities and address them (Martell, 1991).

This article aims to address this gap by studying the incarcerated homeless and identifying specific vulnerabilities, which renders them different from the nonincarcerated homeless. We hypothesized that (a) there would be specific socio-demographic differences between the incarcerated and never-incarcerated and (b) there would be specific risk factors that contributed to incarceration among the homeless. In addition, we also wanted to describe the homeless population and its significant involvement with the criminal justice and enforcement system.

Method

Between May and September 2009, the BCHOHS sampled homeless populations from three cities in British Columbia, Canada: the large urban center Vancouver (n = 250), the mid-sized city and capital of province Victoria (n = 150) and the more remote, but largest rural city of Prince George (n = 100). Participants were at least 19 years of age, able to give informed consent and self-identified as being homeless. Homelessness was defined as living on the street, in a shelter, couch surfing or having no fixed address. A significant (15%+) portion of women, young people (aged 19-25 years) and Aboriginal participants were recruited through focused, purposive sampling due to the additional vulnerabilities that these groups face.

The first half of the sample was restricted to those termed 'absolutely homeless', or living off of the streets. This group was recruited via street outreach, at drop-in centers, food banks and service staff. The second half of the sample was recruited randomly from shelters. Shelter beds were randomized, and then selected shelter beds received a card with an invitation to participate in the study and to book an appointment. In order to determine the housing status of potential participants recruited from services and outreach centers, outreach staff was consulted.

Face-to-face interviews were conducted for one session by trained interviewers, who were mainly trained mental health nurses or social workers. Interviews were administered primarily in a research office. Some interviews also took place at the site of recruitment, where participants felt most comfortable. Prior to participation, participants were given a detailed description of the study and provided informed consent. Participants each received $CAD 30 whether or not the interview was completed. The Behavioural Research Ethics Board of the University of British Columbia and the Providence Health Care Research Institute provided ethics approval.

All participants completed a variety of assessments in the BCHOHS, including demographic information, the Maudsley Addiction Profile (MAP), the Childhood Trauma Questionnaire (CTQ) and the Mini International Neuropsychiatric Interview (MINI) Plus. The MINI Plus, version 5.0.0 (MINI-Plus; Sheehan et al., 1998) is a structured clinical interview based on diagnostic criteria of the Diagnostic and Statistical Manual (4 ed.; DSM-IV) and the International Classification of Diseases (ICD)-10. It was designed to assess current and lifetime Axis I substance use and mental health disorders as well as antisocial personality disorder. It has been shown to be reliable and valid in several US based and European studies (Sheehan et al., 1998). The MAP (Marsden et al., 1998) is a self-report measurement that assesses problem behavior in the past 30 days in four domains: substance use, health risk behavior, physical and psychological health and personal/social functions. For this study, we included only information on substance use. The Childhood Trauma Questionnaire -Short Form (CTQ-SF; Bernstein et al., 2003) is a retrospective self-report inventory that assesses different types of childhood maltreatment on five subscales: physical abuse, emotional abuse, sexual abuse, physical neglect and emotional neglect.

For this article, we explored two questions: what are the characteristics of the incarcerated homeless (n = 328; 65.6%), and how do they compare to the never-incarcerated participants. We defined incarceration as spending more than one night in any kind of custody involving the legal system and included jail, remand or postconviction custody. We compared these two populations (incarcerated vs never-incarcerated) and analyzed associations using t-test and chi-square tests. In order to find specific risk factors, we then ran a logistic regression analysis using a model where only significantly associated variables (p < .05) with the outcome of being incarcerated or not were included.

Results

We present results for both sets of analyses -characteristics of the incarcerated homeless population and differences between incarcerated and never-incarcerated. In the first analysis, the sample consisted of 328 (65.6%) of the total 500 that were interviewed for the study. Males (66.6%) constituted the majority of the incarcerated sample. The age of first incarceration was 20.9 years, and they had been to jail an average of 2.3 times over their lifetime.

The incarcerated sample also indulged in unsafe sex 11.3 times till date in their lifetime and used infected syringes about 0.25 times during the same period. As for the crimes for which details were recorded, 25.8% sold drugs, 11.9% indulged in survival sex or prostitution and 2.7% shoplifted. Most health-care visits were to the emergency rooms (3.6 times).

In the second analysis, comparing the incarcerated with the never-incarcerated, several significant differences were observed (see Table 1). The incarcerated group had more males (66.6%) than the never-incarcerated (52.3%). More incarcerated homeless had also been in foster care (56.4%) than the never-incarcerated (35.5%) and had also used more substances (2.7). Among the substances used, there were significant differences observed for reported use of crack cocaine (69.9% vs 30.1%) and crystal methamphetamine (78.7% vs 21.3%). The incarcerated also reported a significant and higher degree of childhood emotional abuse (p = .03), emotional neglect (p = .003) and sexual abuse (p = .008). Adult sexual abuse was also higher in the incarcerated (31.6%) compared to the never-incarcerated (p = .05). Regression analysis yielded a significant influence of variables with R 2 = .084 (p < .001). Risk factors identified which had a positive predictor value were male gender (p < .001; odds ratio (OR) = 2.8; 95% confidence interval (CI): 1.7-4.4), a diagnosis of depression (p = .02; OR = 2.2; 95% CI: 1.1-4.4) and a severe emotional neglect (p = .02; OR = 1.8; 95% CI: 1.1-3.2) in the childhood.

Discussion

The participants in our survey are from three different regions of British Columbia. It focuses on and addresses the problems of an absolutely underprivileged population -that of the homeless incarcerated. We found an unusually high prevalence of incarceration among the homeless (65%), which is higher than the 50%-55% prevalence that has been reported in earlier studies from similar populations (Burt et al., 1999;Wilder Research, 2007). This number assumes significance considering the fact that the sample was interviewed in three different regions, which respectively represent an urban, a semi-urban and a rural region of Canada, thereby suggesting that these findings may be generalizable to the entire country.

Our study sample was also predominantly male, had higher public arrests than the general population and included those who were often arrested for offences related to drug use such as selling drugs, theft and shoplifting or sex work (Snow, Baker, & Anderson, 1989). This is in line with other studies, which reported that the criminal activities of homeless individuals are predominantly minor crimes that directly result from their efforts to survive on limited resources -for example, shoplifting or stealing (Fischer, 1988(Fischer, , 1992Snow et al., 1989). We also observed an early age of first incarceration, that of 20 years, which can continue to lead to repeated incarcerations over time. The concerning fact observed in this study was that nearly one in eight persons was found indulging in prostitution, which is similar to some other studies from Canada that have estimated (De Matteo et al., 1999;Roy et al., 2000) that the proportion of runaway and homeless who have been involved in prostitution range from 12% to 32%. This study observed a similar figure of nearly 12% who were involved in survival sex in order to maintain themselves and their drug habits.

More concerning is the fact that there is a high degree of unprotected sex among this population (11.3 times as observed in our study), which renders them vulnerable to a host of other chronic illnesses. The homeless incarcerated also had greater emergency and acute admissions (Arfken et al., 2004;Pasic, Russo, & Roy-Bryne, 2005;Saarento, Hakko, & Jaukamaa, 1998), which translates to greater medical cost than usual nonhomeless patients (Hwang, Waver, Aubry, & Hoch, 2011). This is in addition to the fact that, in comparison to the general population, homeless people are at increased risk for all-cause mortality (Hwang, 2000;Hwang, Wilkins, Tjepkema, O'Campo, & Dunn, 2009;Kasprow & Rosenheck, 2000) and often experience higher rates of acute and chronic health conditions (Goering, Tolomiczenko, Sheldon, Boydell, & Wasylenki, 2002). All this adds to a serious burden to the society in terms of both economic costs as well as health burden.

Comparing the incarcerated with the never-incarcerated within our sample population, significant differences were observed in both gender and substance abuse, with males outnumbering females (67% vs 33%) and the incarcerated using more substances than the never-incarcerated (p = .003), suggesting that substance abuse is especially highly prevalent in the homeless population and is additionally associated with involvement in the criminal justice system (Benda, 1993;Desai, Lam, & Rosenheck, 2000;Fischer, 1988Fischer, , 1992Kushel, Hahn, Evans, Bangsberg, & Moss, 2005;Lindelius & Salum, 1976).

This study observed an overall greater prevalence of substance abuse, which is similar to another study from Toronto where current drug use was reported by 40% of the sample (Smart & Adlaf, 1991) and which is at least three times that of the general population (Canadian Centre on Substance Abuse, 2006). Our finding that homeless people with mental disorders who have been incarcerated have higher rates of co-occurring substance-related disorders than those who are not homeless is consistent with research in other settings who have defined incarceration as spending more than one night in any kind of custody during their lifetime (Drake & Wallach, 1999;McNeil et al., 2005;Salkow & Fichter, 2003). In addition, inmates with dual diagnoses were more likely than those without co-occurring disorders to be homeless, to spend more time being incarcerated and to be charged with violent crimes than other inmates (McNeil et al., 2005).

The substances used that were significantly different between both groups were crack cocaine and crystal meth. Both these drugs have been associated with depression and other psychiatric illnesses (Duailibi, Ribeiro, & Laranjeira, 2008;Grant et al., 2012;Marshall & Werb, 2010;Smart, 1991). We also observed a similar higher prevalence of depressive disorder and psychotic disorders among the incarcerated homeless in our sample. Combined together, substance use and mental illness can increase the risk of homelessness by undermining their ability to maintain social ties and economic stability (Vangeest & Johnson, 5 2002). In addition, it also renders the homeless vulnerable to numerous adverse health effects, including overdoses and infectious diseases (Chyvette & Latkin, 2007;Little et al., 2005).

Having been in foster care and having a history of childhood emotional and sexual abuse were both significantly higher among our incarcerated sample. Similar findings of high emotional and sexual abuse have been observed in earlier studies (Chyvette & Latkin, 2007;Thompson & Hasin, 2012;Vangeest & Johnson, 2002). It may be hypothesized that having been in foster care increased the risk of emotional and sexual abuse, but we were unable to find such an association. Homeless individuals may be traumatized at an early age, put into foster care, rendered homeless, initiated into substance use and re-traumatized on repeated occasions in adult life, rendering them vulnerable to incarceration and mental illness (Kerr et al., 2009).

We identified male gender (OR = 2.8), being emotionally neglected (OR = 1.8) and having depression (OR = 2.2) as major risk factors for being incarcerated. It appears that the most vulnerable population that runs the risk of being incarcerated repeatedly are males, with a history of childhood emotional neglect and a history of major depressive disorder. Possible reasons for this may be the difficulty that this group faces in making bail (either due to financial or social conditions), being less likely to be released on their own recognizance and often being arrested multiple times for a similar crime. This pattern of repeated traumatization and incarceration represents a failure of the current health-care and prison systems (McNeil et al., 2005). Crimes committed by these individuals are not recognized as substantially related to their untreated mental or substance-related disorders. These individuals are pushed through the criminal justice system that has been designed more for incapacitation and punishment than for treatment. A disheartening lesson that emerges is that it appears that many individuals with mental disorders and substance use are not being stabilized under the current system but, rather, are cycling between the street and jail.

Instead of cycling in this vicious circle, the authors recommend that homeless people should have minor charges dismissed, and, where appropriate, may have more serious misdemeanor charges before the court reduced or dismissed. In addition, traditional sanctions such as fines and custody should be replaced by community-based treatment and services, with the absolute goal of rehabilitation rather than punishment (American Bar Association Commission on Homelessness and Poverty, n.d.).

To conclude, the incarcerated homeless represent a population that has been severely affected on several domains, in addition to suffering from increased substance use and mental disorders. Identifying the most vulnerable population from the risk factors observed in this study could help to strengthen our current health and judicial systems so as to better serve the needs of this population. We, however, recognize that our findings need to be accepted in view of several limitations, the most important being that our analyses were largely exploratory. We proceeded in light of current, albeit limited, knowledge about both homelessness and incarceration. Despite this limitation, we have been able to identify what appear to be likely risk factors for increased incarceration among the homeless population.