Brief Reports: Psychiatric Illness and Substance Abuse Among Homeless Asian-American Veterans

A small but growing body of research suggests that Asian Americans have significant psychiatric and substance abuse problems ( 1 , 2 ). However, few, if any, nationally based studies have directly compared rates of mental illness and related problems between Asian Americans and other racial or ethnic groups. Homelessness and related risk factors, including mental illness, have also yet to be examined among Asian Americans. A significant body of research has demonstrated the association of psychiatric illness and substance abuse with homelessness among U.S. veterans ( 3 , 4 ). To our knowledge, there have been no published studies of homeless Asian Americans, either civilian or veteran.

Since 1987 the Department of Veterans Affairs (VA) Health Care for Homeless Veterans (HCHV) program, now operating at almost 150 locations across the country, has conducted community outreach to contact homeless veterans in community settings and link them with VA and non-VA health services, including time-limited residential treatment. Basic sociodemographic and clinical data are gathered for all clients through a structured intake assessment at the time of program entry, which provides a unique basis for identifying the relative risk of homelessness among Asian-American veterans and the presence of various risk factors. In this study we examined the proportion of Asian Americans among homeless veterans and among veterans in the general population and compared Asian-American and other homeless veterans on measures of psychiatric illness and substance abuse.

Methods

Data were gathered from all intake assessments conducted at all 141 sites of the HCHV program between 1997 and 2001 (N=157,628). The entire sample was used to determine the proportion of Asian-American participants. However, to minimize the diluting impact of sites with only minimal numbers of Asian-American veterans in the examination of the relative prevalence of clinical problems, only sites that reported at least five Asian-American veterans were included in the comparative analysis of clinical characteristics (22 sites, 67,441 veterans).

Data were obtained from administrative intake data used to monitor program performance at the national level. Intake assessment involved completion of a 60-item structured assessment interview administered by program clinicians, primarily social workers and nurses. The interviews were predominantly conducted during outreach visits to locations such as soup kitchens and shelters. Data collected include patients' self-reports of basic demographic information, past and current drug and alcohol use, and past and current psychiatric status, as well as clinicians' diagnostic assessments of substance abuse and psychiatric illness. Waiver of informed consent was obtained from the institutional review board of the VA Connecticut Healthcare System.

To determine the relative risk of homelessness among Asian-American veterans, the proportion of Asian Americans within each of 13 age strata in the homeless population was compared with the proportion of Asian Americans in the general veteran population in the 2000 census ( 5 , 6 , 7 ). The relative risk of homelessness among Asian Americans was then calculated with its 95 percent confidence interval ( 8 ).

To compare characteristics of homeless Asian-American veterans with those of other racial or ethnic groups, bivariate associations between the dependent diagnostic variables and demographic characteristics were reviewed to identify variables for inclusion as covariables in the analysis of clinical differences between ethnic groups. Because all dependent variables were dichotomous, logistic regression was used to examine the association of ethnic group membership (dummy variables coded for black, Hispanic, and Asian Americans, with white as the reference group) and mental health status indicators. The following control variables were included in each model: sex (dummy variable for female gender), age at time of interview (continuous), marital status (coded 1 if married), Vietnam-era service (coded 1 if yes), self-reported experience of combat fire (coded 1 if yes), number of days homeless in the past 30 days (continuous), literal homelessness (coded 1 if residence at the time of the interview was a shelter or the street), and receipt of public support (coded 1 if the veteran was currently receiving service-connected disability payments, non-VA disability payments, a pension that was not connected to service, or other public support), and number of days worked for pay in the past 30 days (continuous).

Results

Of the 157,628 veterans who were homeless in our sample, 504 (.32 percent) were Asian American, which is about one-third of the 1.1 percent of the general veteran population in fiscal year 2000 that was Asian American ( 7 ). The estimate of the proportion of Asian Americans in the general population of veterans increased slightly, to 1.19 percent when standardized to the specific age distribution in the homeless sample. The age-standardized risk of homelessness among Asian Americans compared with other veterans was thus .266 (Z=32.0, p<.001).

Of the 67,411 veterans from sites with at least five homeless Asian-American veterans, 402 were Asian American (.6 percent). Five tables presenting detailed demographic and clinical data for Asian-American, white, black, and Hispanic veterans can be found in the online version of this brief report at ps.psychiatryonline.org. Briefly, the mean±SD age of the Asian-American veterans was 45.6±11.9. Ninety-five percent were men, 13 percent were married, and 62 percent were literally homeless. Thirty-nine percent had served during the Vietnam era. Twenty-seven percent reported combat exposure. Fifty percent received some form of public support or benefits. Clinicians' ratings indicated that 33 percent had alcohol abuse or dependence, 31 percent had current drug abuse or dependence, 40 percent had serious psychiatric problems, and 21 percent had co-occurring disorders. Thirty-two percent reported at least one previous psychiatric hospitalization.

There were significantly more women among Asian-American veterans (p<.001), and both Hispanic and Asian-American veterans were more likely to be married than white veterans (p<.001). White veterans were more likely than Asian-American veterans to have served in the Vietnam era (p<.001), as only 39.4 percent of Asian Americans reported such service. However, Asian-American veterans were slightly more likely than veterans of other ethnic groups to report exposure to enemy fire during combat (27 percent; p<.001). Asian-American veterans were significantly less likely than the other racial or ethnic groups (except Hispanics) to be literally homeless (sleeping in an emergency shelter or on the streets) at intake (62.2 percent; p<.001).

As shown in Table 1 , after the analyses controlled for other factors, compared with the other three racial or ethnic groups, Asian-American veterans were less likely to report alcohol abuse or related problems and less likely to be given a diagnosis of such disorders by VA clinicians. They were less likely than blacks or Hispanics to report drug abuse problems or being given a diagnosis of such problems, but they were not significantly different from whites in reporting or being given a diagnosis of such problems.

Enlarge table

Both Asian-American and white veterans were more likely than Hispanic or black veterans to report current psychiatric illness or history of a previous psychiatric hospitalization (p<.001), but no significant difference was found between whites and Asian-Americans on these self-report measures. Asian-American and black veterans were more likely than whites to receive a clinical diagnosis of schizophrenia (p<.01 and p<.001, respectively), and Asian Americans were more likely than all other racial or ethnic groups to receive a clinical diagnosis of other psychotic disorder (p<.001). Asian-American veterans were also more likely than the other three groups to receive a diagnosis of any psychotic disorder (p<.01). Clinicians' assessment of the presence of serious psychiatric illness at any time was more frequent among white veterans than among Asian-American veterans (p<.01). Clinicians' assessment of dual diagnosis was highest among white veterans, and Asian Americans were the least likely to receive such a diagnosis (p<.001).

Discussion

These analyses of a large administrative data set suggest that Asian-American veterans are about one-fourth as likely as age-matched veterans in the general population to become homeless. Analyses of the homeless veteran population suggest that by both self-report and clinicians' ratings, Asian Americans are less likely to have problems related to alcohol abuse. Findings for drug abuse were more mixed. Asian Americans were less likely than blacks and Hispanics to report drug use or to be given a diagnosis of drug use disorder by VA clinicians, but they were no different on this measure from whites. Although Asian Americans were no more likely to report psychiatric illness than whites, they were more likely than whites to be given a diagnosis of schizophrenia or other psychotic disorders by VA clinicians. Mental illness and substance abuse are both strong risk factors for homelessness in the general population ( 9 ), and in our sample, the lower rate of alcohol abuse observed among Asian-American veterans may account for both the lower risk of homelessness overall and the higher rate of schizophrenia among those who become homeless.

However, the high incidence of diagnosed schizophrenia and psychosis, in the presence of no differences in self-reported psychiatric illness, suggests a possible discrepancy between self-report data and clinician diagnostic judgments that may reflect cross-cultural miscommunication or even clinician bias, with VA clinicians diagnosing psychosis more readily among Asian Americans than in other racial or ethnic groups.

The lower risk of homelessness among Asian-American veterans may also reflect the way some Asian-American families have been described as coping with mental illness ( 10 ). Mental illness carries an intense stigma in most Asian cultures, and Asian families may at first try to shelter and hide their mentally ill family members from the larger society. However, as the illness progresses, families may find the family member with mental illness increasingly unmanageable, resulting ultimately in rejection ( 11 ). Studies have shown that Asian-American patients consistently present later in the course of illness for psychiatric care and often present with more severe illness, most likely as a consequence of this delay in seeking treatment ( 2 ).

This study has several limitations. The sample of homeless Asian-American veterans was small and based on a clinical convenience sample. Our results thus may not be representative of all homeless Asian-American veterans or of specific Asian ethnic subgroups. Another limitation is possible clinician bias in assessing mental illness in racial or ethnic minority groups. The higher rate of schizophrenia and other severe psychiatric illnesses among Asian Americans may reflect unfamiliarity with cultural expressions.

Conclusions

Asian-American veterans appear to experience a lower risk of homelessness than veterans from other racial or ethnic groups, and Asian-American veterans who become homeless appear to experience less alcohol abuse but possibly higher relative rates of psychotic illnesses. The lower rates of alcohol abuse among Asian-American veterans may help explain their reduced risk of homelessness.