Secondary Logo

Share this article on:

Perception of Health Status by Homeless US Veterans

Nyamathi, Adeline ANP, PhD, FAAN; Sands, Heather FNP, PhD; Pattatucci-Aragón, Angela PhD; Berg, Jill RN, PhD; Leake, Barbara PhD; Hahn, Joan Earle DNSc, RN, CDDN, CS; Morisky, Donald ScD


Perceptions of health status among 331 homeless veterans and homeless nonveterans were examined. Homeless veterans were significantly less apt to perceive their health as fair/poor (8%) compared to non-veteran homeless men (19%). Homeless veterans were also more likely to report having a regular source of care (57% versus 36%). Logistic regression analysis indicated the adjusted odds of fair/poor health were more than two times greater for persons reporting depressive symptomatology than for those without this history; veterans continue to remain less likely to report fair/poor health than nonveterans. High rates of substance abuse were observed for the entire sample. Such differences in perceived health result in important health access issues.

From the School of Nursing, University of California–Los Angeles (Nyamathi, Sands, Berg, Leake, Hahn)

School of Public Health, University of California–Los Angeles (Morisky)

From the School of Public Health, University of Puerto Rico, San Juan. (Pattatucci-Aragón)

This study was funded by the National Institute on Drug Abuse, DA11145.

Corresponding author: Adeline Nyamathi, ANP, PhD, FAAN, Professor, UCLA, School of Nursing, Room 2–250, Factor Building, Box 951720, Los Angeles, CA 90095–1702 (e-mail:

CURRENT ESTIMATES INDICATE that veterans comprise almost 40% of the homeless population and that more than 250,000 may be homeless on any given night. 1 Poverty, postmilitary mental illness, and social isolation are among the complex pattern of risk factors that predispose veterans to homelessness. Additional factors include childhood physical and/or sexual abuse and lack of support from family and friends. 2,3 Up to 70% of homeless veterans have long-standing alcohol dependence, 4 and 45% suffer from mental illness, primarily depression. 5 Homeless veterans also have high rates of posttraumatic stress disorder (PTSD), 3 comorbid psychiatric conditions, 6 and severely impaired occupational and social functioning. 3,7 A number of homeless veterans face disabilities due to serious mental illnesses. 8

As a group, post–Vietnam era veterans are more than 3 times as likely to be homeless as their nonveteran counterparts. 2 This cohort of veterans, aged 20–34, has also been found to disproportionately experience pervasive addiction problems. Moreover, rates of antisocial behavior are 5 to 6 times higher in this subgroup than in the general population. 3

Compared to other homeless adults, homeless veterans tend to be older, better educated, and are more likely to have been or currently be married. 4 They also have access to an array of health care resources through the Veterans Administration (VA). However, homeless veterans' perceptions of their health status and whether these perceptions differ from those of nonveteran homeless individuals, when other potentially important predictors of general health are considered, have not received adequate attention.

In this study, the perception of health status was compared among homeless veterans and homeless nonveterans residing in the skid row area of Los Angeles. In addition, the relationship was examined between situational, personal, cognitive, and behavioral factors and the perceptions of health among homeless adults. This information was used to control for those variables that might confound any relationship between veteran and health status. Such information can help guide the design of intervention programs to improve the health of veterans and assist them in exiting their homeless state.

Back to Top | Article Outline


Perceptions of health status have been widely used in research as a surrogate for physical health status, primarily because it is difficult to obtain objective measures of health through physician examinations. Perceived health has provided an indirect measure of an individual's sense of long-standing chronic illness, 9 has served as a useful means of gauging physical health status, 10 and has been associated with capacity for self-care. 11 As a global measure, perceptions of health status may be tapping into aspects of mastery over the environment, of coping effectiveness mediated by self-esteem, and of other personal resources. 11 Among homeless populations, Ropers and Boyer 10 found that the presence of a chronic condition was the most important correlate of perceived health, and psychological distress was a strong predictor of health perception among homeless individuals with protracted periods of depression. Taken together, however, these two variables only accounted for 20% of the variance, which suggests that perceived health status is a complex phenomenon requiring a multifactorial model explanation. Given that perceived health status is a reliable surrogate measure of general health status, Goldstein et al 9 suggests that uncovering substantial relationships with other measures of physical and psychological health, as well as those associated with health behaviors/health care access, is a realistic expectation.

Back to Top | Article Outline


The organizing framework used for this study was the Comprehensive Health Seeking and Coping Paradigm (CHSCP). 12 The CHSCP, originally adopted from Lazarus and Folkman's 13 Stress and Coping Model and Schlotfeldt's 14 Health Seeking Paradigm, includes a number of components that influence health outcome. Health care providers can help individuals gain insight into their effective coping strategies, as well as assist in enhancing those strategies, by assessing factors that influence the appraisal of health status and coping behaviors used to manage the environment.

The model components examined in this study consist of personal factors, situational factors, behavioral factors and sociodemographic factors. Personal factors included current depression, duration of homelessness, and perceived seriousness of conditions (including perception of health status). Situational factors, such as environmental constraints and facilitators, also figured prominently in this study; they are represented by health care access and resources. Important behavioral factors in this study included the use of substances such as alcohol and illegal drugs, which are emblematic of emotion-focused coping. Finally, important sociodemographic factors that may affect health outcome were examined, including level of education, marital status, and employment.

Back to Top | Article Outline


Sample and setting

Data were collected between 1997 and 2002, as part of a larger study examining the effectiveness of a case managed program by a nurse versus a standard program on the compliance of homeless adults with latent tuberculosis (TB) that were offered a 6-month course of isoniazid by directly observed therapy. Subjects were determined eligible for the screening phase of this TB chemoprophylaxis study if they: (1) had spent the previous night in a homeless shelter or residential substance abuse treatment program, (2) had no self-reported history of a positive test for Protein Purified Derivative (PPD) or of completing a TB treatment program, and (3) were between the ages of 18 to 55 or over the age of 55, and reported risk activation factors for active TB. These included being an injection drug user (IDU) or experiencing certain immune compromising diseases (such as HIV/AIDS, diabetes mellitus, or severe kidney disease) or taking immuno-suppressing medications.

Interested individuals found to be PPD positive via the Mantoux method on subsequent testing (defined as having a PPD reaction of at least 10 mm of induration), and who had a negative chest X-ray and normal liver enzymes were enrolled into the study. In the larger study of TB chemoprophylaxis, a convenience sample of 415 homeless men and women that resided in 20 shelters and residential recovery programs in the skid row area of Los Angeles were consecutively enrolled. Because the recruited female subsample (84) did not contain any veterans, they were excluded from this analysis. Of the enrolled sample of homeless men (n = 331), 79 (19%) were self-reported as US veterans.

Back to Top | Article Outline


Participants received a full description of the study goals and procedures before providing written informed consent, as approved by the University of California, Human Subjects Protection Committee. Recruitment procedures consisted of posting flyers outlining the study in the targeted 20 shelters and residential recovery programs in the skid row area of Los Angeles.

These flyers directed interested persons to call or visit the research nurses located at a neighborhood clinic within walking distance of the shelters or treatment programs. Persons were excluded from the study if, during the interview, they demonstrated cognitive impairment such as active hallucinations or stupor. Those determined eligible for the study were administered a baseline questionnaire by trained nurses and outreach workers.

Back to Top | Article Outline


Several of the instruments used in this study had been tested previously with African American and Latino homeless populations 15,16 and demonstrated sound reliability and validity.

Back to Top | Article Outline

Sociodemographic factors

A structured instrument assessed age, ethnicity, education, time homeless, marital status, and history of incarceration.

Back to Top | Article Outline

Personal factors

Current depression was measured by the Center for Epidemiological Studies Depression (CES-D) scale. 17 The instrument is a 20-item scale that measures frequency of symptoms on a 4-point continuum and has well-established reliability and validity. Examples of CES-D items are “You felt depressed” and “You felt fearful.” Scores on the CES-D ranged from 3 to 53. In the current study, the Cronbach alpha coefficient for the CES-D scale was 0.85. CES-D scores were dichotomized at the customary warning level of 16 for analysis.

Perceived health status was measured by an individual item. Respondents were asked to rate their general health on a 5-point scale from “excellent” to “poor.” This item has been used in a number of health surveys as a valid overall indicator of physical health, 18,19 including being part of the health assessment in the RAND Medical Outcomes Study. 19

Back to Top | Article Outline

Situational factors

Health care access and resources were assessed by items inquiring about whether participants had a regular source of medical care and whether they received veteran benefits or insurance. These items were assessed using a “yes” or “no” response.

Back to Top | Article Outline

Behavioral factors

Drug and alcohol use were assessed with a revised version of the Texas Christian University (TCU) Drug History Form. 20 This questionnaire has been tested with men and women with a history of drug addiction, prostitution, and homelessness. It records the frequency of use of 16 drugs, by injection or other means, during the past 6 months and also elicits information about lifetime use. Drugs assessed are: heroin, street methadone, other opiates, cocaine, crack, methamphetamine, other amphetamines, inhalants, marijuana/hashish, hallucinogens, tranquilizers, barbiturates, other sedatives, designer drugs, alcohol, and nicotine. The CAGE questionnaire was administered to assess persons with a high likelihood of alcohol dependence or abuse. Cronbach's alpha for the CAGE instrument in this sample was 0.81. Injection drug use was defined as any use of drugs by injection regardless of frequency, specified as during the last 6 months (recent) or during lifetime. Objective measures of drug use were not obtained as they provide relatively short-term evidence of drug use. Further, our research team collected hair samples in a previous study of homeless women and we found reasonable concurrence between self-reporting and objective evidence of cocaine use. 21

Back to Top | Article Outline


Descriptive analysis consisted of frequencies and percentiles or means and standard deviations, depending on the level of the data. Differences in categorical sociodemographic and behavioral characteristics were assessed between veterans and nonveterans and between those who did or did not perceive their health to be in the category of fair/poor, with chi-square and Fisher exact tests. Differences in continuous variables between the two groups were assessed by independent sample t-tests. Logistic regression analyses were performed to estimate the independent effect of veteran status on fair/poor health when variables such as age, race, regular source of care, and history of incarceration were controlled.

Back to Top | Article Outline


Sociodemographic factors

For the overall sample, the men reported a mean age of 41.2 years (SD 8.6). The majority of the sample was African American (82.8%), 9.1% were Latino, and 6.7% were Caucasian. The majority were unemployed (88.5%), single (60.4%), and completed a mean of 12.1 (SD = 1.8) years of education. A significant proportion of the sample self-reported as US veterans.

As displayed in Table 1, the veteran subsample was older (44 versus 40 years) and better educated (13 versus 12 years) compared to their nonveteran counterparts. Homeless veterans were also less likely to report a history of incarceration than nonveteran homeless men (53% versus 66%). Within group comparison revealed that veterans born in 1954 or after were less likely to be married than those born before that time. Additionally, veterans born in 1954 or after were found to have higher rates of depression compared to veterans born before 1953 (48% versus 36%), respectively; however, these data were not significant (not tabled).

Table 1

Table 1

Back to Top | Article Outline

Drug and alcohol use

Although comparisons between veterans and nonveterans on drug and alcohol use measures showed no significant differences, substance use was notably high in both groups (Table 1). For example, almost half of the sample reported a history of alcohol dependency (44% versus 47%, for veterans and nonveterans, respectively), and 70% of both groups reported lifetime use of crack cocaine. High levels of alcohol use in the last six months were also reported by veterans (70%) and nonveterans (73%).

Back to Top | Article Outline

Health care access and resources

Table 2 shows that veterans were significantly less likely to perceive their health as fair/poor (8%) compared to non-veteran homeless men (19%). Moreover, veterans were more likely to report having a regular source of care (57%) than their counterparts (36%). Only about one fifth (19%) of veterans reported receiving veterans' benefits, while almost half (47%) reported having veterans' insurance.

Table 2

Table 2

Back to Top | Article Outline

Relationships with perceived fair/poor health status

A number of significant relationships emerged when homeless men who reported fair/poor health status were compared to those who perceived better health status (Table 3). Homeless men who perceived their health to be worse were more likely to report depressive symptomatology (67% versus 44%) compared to those experiencing better health. Perceptions of fair/poor health also were associated with injection drug use. Moreover, men who reported fair/poor health were more likely to have been incarcerated than those who reported better health. Significant differences were not found between perceived health status and race or history of substance abuse other than injection drug use.

Table 3

Table 3

Within the group of the veterans reporting receipt of veterans' benefits (19%), none reported fair/poor health. Moreover, among the 47% that reported having veterans' insurance, none reported fair/poor health. Health differences were not found between veterans and nonveterans that indicated no regular source of care.

Back to Top | Article Outline

Predictors of fair/poor health status

Logistic regression analysis indicated that nonveterans are more likely to report a perception of fair/poor health status than veterans, and incarceration was positively associated with fair/poor health status (Table 4). Moreover, the adjusted odds of fair/poor health were more than two times greater for persons with depressive symptomatology than for those without this history. In addition, those less educated were more likely to report fair/poor health than those not reporting these factors. Significant differences were not found with respect to age.

Table 4

Table 4

Back to Top | Article Outline


In 1996, 9%, or 2.2 million, veterans received some level of disability compensation, 22 with a growing trend toward seeking service-connected disability payments. 23 Enrollees of the Veterans Administration may have other coverage, including Medicare, private insurance, or Medicaid, with approximately 73% having alternative forms of coverage and 38% using exclusively the Veterans Administration for health care. 24

Nationwide, homelessness continues to affect a significant percentage of veterans in the United States, despite the availability of federal benefits for those meeting eligibility requirements. 3,25 Similar results were found in this study of homeless adults. Although nearly 25% of the sample reported veteran status, only about one fifth of them acknowledged receipt of VA health benefits. This poor coverage may reflect the convoluted process of establishing eligibility and subsequently registering for VA health benefits. This can involve appealing for benefits when denied coverage or undergoing review to reestablish eligibility for disability payments, usually every other year. 26 This is highlighted in Applewhite's 27 analysis, indicating that homeless veterans report high levels of stress and frustration related to the VA service delivery system. Prior to 1996, a complicated method of eligibility determination rendered many veterans with either very limited benefits or none at all. Although the situation is somewhat improved today, in order to receive VA benefits, veterans must still complete a lengthy application form and provide an address where “official notification” can be mailed. Further, even with resources available to them, homeless veterans may face insurmountable barriers in seeking health care services, as they must still procure transportation, which may be difficult or too problematic to coordinate. If the VA is to reach homeless veterans, then the process will need to be facilitated through diligent case management and a waiver of co-payments without elaborate means testing requirements.

Although few veterans in our sample were receiving VA benefits, more than half reported having a regular source of care. Wenzel et al 28 found that the reported need for services resulting from chronic medical problems, mental health needs, and substance abuse faced by a cohort of homeless veterans in Los Angeles was positively related to use of health services, a finding that mirrors a description of veterans using the VA health care system. 29 In particular this study reported that veterans, who accessed the VA health care system, were found to be more disabled, report more complex health problems, and be less economically stable than the general US population of veterans who do not use the VA health care system. It is possible that the use of health services promotes a greater sense of awareness, which in turn amplifies awareness/perceptions of need for health services.

Although the data revealed no significant differences between homeless veterans and nonveterans on history of depression, alcohol dependency, crack cocaine use, and ongoing use of alcohol or drugs, fewer veterans rated their health status as fair or poor compared to non-veteran homeless. Furthermore, homeless veterans were significantly less likely to report incarceration histories compared to their non-veteran counterparts. They were also more likely to be older and better educated than non-veteran homeless men. Whether greater availability of health care combined with greater maturity due to age and education relates to better perception of health and improved social behavior is of interest to consider in further investigations of homeless populations.

Given that one would expect older persons to report poorer health in general, it is possible that homeless adults may perceive their normal aches and pains of growing older differently than the general population, since they have major survival challenges that are not experienced by nonhomeless populations. Thus, these health problems may be relegated to a lower priority or ignored entirely. Alternatively, another explanation, called a disability paradox, 30 may be the adjusting of standards, values, or expectations given a particular level of impaired functioning as a means of accommodating to the day-to-day realities of living with chronic conditions or illnesses. 31,32 Veterans, like others who are aging, may face increasing frailty and disability. Moreover, whether growing old and facing debilitating disease is of great concern for homeless veterans with mood disorders and/or substance abuse problems is largely yet unexplored. 33 How context mediates the interpretation of symptoms/afflictions/nuisances and informs one's overall view of perceived health remains an important unanswered question.

The data also reveal that homeless men reporting current depression are more likely to perceive their health as fair or poor. This finding is consistent with Ropers and Boyer's 10 investigation, revealing psychological distress as a strong predictor of health perceptions among homeless individuals with protracted periods of depression. Sullivan et al 34 reported greater dissatisfaction with health status for homeless individuals with mental illness and poorer physical functioning compared with other homeless individuals without mental illness. Although the conventional view of health perception as a measure of one's sense of chronic illness was not examined, the findings clearly showed that veterans were more likely to identify a regular source of care compared to nonveterans. They likewise were less likely to report fair/poor perception of health status. Although it is unclear if a proxy relationship exists between a regular source of care and chronic illness, a recent study has found a higher proportion of chronic diseases among veteran populations in the United States and their unmet needs by traditional interventions. 35 However, use of health care resources is related to the existence of an acute or chronic illness. 28

In disentangling the differences between homeless veterans and nonveterans, particularly with regard to perception of health status, having current depressive symptoms was seen as a predictor for perceived health status among veterans and nonveterans. Similarly, other studies have found associations between posttraumatic stress disorder and combat-related military experience. 25,36 This study did not find a significant association between veteran status and problematic use of alcohol compared to nonveterans, but alcohol use in the overall sample was still high. Although other studies 3,5 have reported a higher prevalence of alcohol problems among veterans than nonveterans, it may be that a leveling effect is occurring, where differences observed earlier are not as apparent as the population gets older. Hard drugs were also more expensive when those studies 3,5 were conducted. Furthermore, as high rates of hard drug use were found, particularly crack, this could have a mediating effect on alcohol use, if prices are low and hard drugs are widely available. While addressing such factors may not improve one's perception of health status, it may decrease the likelihood of remaining homeless. A concerted attempt to improve upon delivery of health care services to homeless veterans and to change the culture in the military that encourages alcohol use would go a long way toward decreasing the number of young men that adopt substance-abusing behavior.

Incarceration was found to be a significant overall predictor of perceived health status, and veterans were significantly more likely to have a history of incarceration compared to nonveterans. Looking at depression as a predictor of health perception, a significant number of veterans develop post-military mental illness and frequently face social isolation upon return from active duty. 37 Lack of family support has been reported as a primary factor in adjustment problems of young men making the transition from a structured military setting (and, frequently, active duty) to civilian life. 38–40 Gamache and others 2 suggest that veterans who come from disadvantaged backgrounds not only confront the same challenges as veterans from stable family backgrounds, but also must adjust and make the transition out of the military with an absence of economic or social support from their families.

If perceived health status is conceptualized as a sense of mastery over one's environment or self-esteem and a belief that one is able to take care of oneself as McDonnell and others 11 suggest, then health educators implementing any program will need to build on existing strengths of the veteran while addressing their substance use, particularly alcohol, as well as the environmental factors, such as the ability to find work or housing. Further, as research findings have revealed that depression, drug use, and PTSD are acute problems among homeless veterans, a pressing need is presented for a broad array of medical, psychiatric, social, and support services targeted toward homeless veterans. In particular, veteran services that are easily accessible and associated with a sense of entitlement may provide a powerful incentive for homeless veterans to utilize such services. 41 This is particularly true as employment is a problem and military job skills frequently do not translate to the civilian market.

Limitations of this study comparing perception of health status among veterans and nonveterans include the use of a relatively small convenience sample limited to a specific geographic location, which may or may not accurately reflect the experience of the general population of homeless veterans and nonveterans. Finally, further exploration of the driving forces leading to depression, as well as ways of coping, may provide clarity and direction with regard to treatment strategies.

In conclusion, the findings of this study underscore the need for focused interventions for each subgroup of homeless persons. Homeless veterans seemingly have better access to health care benefits, yet in this study, only one fifth utilized these benefits. Health care professionals, including health educators and community health workers can address this need by being knowledgeable about the dynamics of health care access for homeless veterans and facilitating their use of these benefits.

Back to Top | Article Outline


1. O'Hare W. A new look at poverty in America. Population Bulletin. 1996;51:41–48.
2. Gamache G, Rosenheck R, Tessler R. The proportion of veterans among homeless men: A decade later. Soc Psychiatry Psychiatr Epidemiol. 2001;36:481–485.
3. Rosenheck R, Frisman L, Chung A. The proportion of veterans among homeless men. Am J Public Health. 1994;84:466–469.
4. Tessler R. Comparison of homeless veterans with other homeless men in a large clinical outreach program. Psychiatr Q. 2002;73:109–119.
5. Roth D, Bean J. Alcohol problems and homelessness: Findings from the Ohio Study. Alcohol Health & Research World. 1985;10:14–15.
6. Kulka RA, Schlenger WE, Fairbank JA, Hough RL, Jordan BK, Marmar CR, Weiss DS. Trauma and the Vietnam War Generation: Report of Findings from the National Vietnam Veterans Readjustment Study. New York: Brunner/Mazel; 1990.
7. Frueh BC, Turner SM, Beidel DC, Cahill SP. Assessment of social functioning in combat veterans with PTSD. Aggression and violent behavior: A Review Journal. 2001;6:79–80.
8. Department of Veterans Affairs, Office of Congressional Affairs, Testimony & Legislation. Statement of Thomas L. Garthwhite, MD, Deputy Under Secretary for Health, Department of Veterans Affairs, on maintaining capacity to provide for the specialized treatment and rehabilitation needs of disabled veterans before the Subcommittee on Health, Committee on Veterans Affairs, US House of Representatives. July 23, 1998 [VA testimony at hearings before the House Veterans Affairs Subcommittee on Health]. Available at Accessed on September 6, 2003.
9. Goldstein M, Siegel J, Boyer R. Predicting health changes in perceived health status. American Journal of Public Health. 1984;74:611–614.
10. Ropers R, Boyer R. Perceived health status among the new urban homeless. Soc Sci Med. 1987;24:669–678.
11. McDonnell M, Turner J, Weaver M. Antecedents of adherence to antituberculosis therapy. Public Health Nurs. 2001;18:392–400.
12. Nyamathi, A. Comprehensive health seeking and coping paradigm. J Adv Nurs. 1989;14:281–290.
13. Lazarus R, Folkman S. Stress, Appraisal, and Coping. New York: Springer Publishers; 1984.
14. Schlotfeldt R. Nursing in the future. Nurs Outlook. 1981;29:295–301.
15. Nyamathi A, Flaskerud J, Leake B, Lewis C, Bennett C. Outcomes of specialized and traditional AIDS counseling programs for impoverished women of color. Res Nurs Health. 1993;16:11–21.
16. Nyamathi A, Leake B, Gelberg, L. Sheltered versus nonsheltered homeless women differences in health, behavior, victimization, and utilization of care. J Gen Intern Med. 2000;15:565–572.
17. Radloff S. The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385–401.
18. Aday L, Anderson R. Equity of access to medical care: A conceptual and empirical overview. Med Care. 1981;19:4–27.
19. Sherbourne CD, Stewart AL. The MOS social support survey. Social Science & Medicine. 1991;32:705–714.
20. Simpson D, Chatham L. TCU/DATAR Forms Manual. Ft. Worth, TX: Institute of Behavioral Research, Texas Christian University; 1995.
21. Nyamathi A, Leake B, Longshore D, Gelberg L. Reliability of homeless women's reports: Concordance between hair assay and self-report of cocaine use. Nurs Res. 2001;50:165–171.
22. Oboler S. Disability evaluations under the Department of Veterans Affairs. In: Rondinelli RD, Katz RT, eds, Impairment Rating and Disability Evaluation. Philadelphia: Saunders; 2000.
23. Frueh BC, Elhai JD, Gold PB, Monnier J, Magruder KM, Keane TM, Arana GW. Disability compensation seeking among veterans evaluated for posttraumatic stress disorder. Psychiatr Serv. 2003;54(1):84–91.
24. Shen Y, Hendricks A, Kazis L. Health insurance and use of services by veterans; 1999 large health survey of veteran enrollee [Executive report]. Washington, DC: Department of Veteran Affairs, Veterans Health Administration, Office of Quality and Performance; 2000.
25. Winkleby M, Fleshin D. Physical, addictive, and psychiatric disorders among homeless veterans and nonveterans. Public Health Rep. 1993;108:30–36.
26. Frueh BC, Elhai JD, Gold PB, Monnier J, Magruder KM, Keane TM, Arana GW. Disability compensation seeking among veterans evaluated for posttraumatic stress disorder. Psychiatr Serv. 2003;54(1):84–91.
27. Applewhite S. Homeless veterans: Perspectives on social services use. Soc Work. 1997;42:19–30.
28. Wenzel S, Bakhtiar L, Caskey N, Hardie E, Redford C, Sadler N, Gelberg L. Homeless veterans' utilization of medical, psychiatric, and substance abuse services. Med Care. 1995;33:1132–1144.
29. Kizer KW. Re-engineering the veterans health care system. In: Ramsaroop P, Ball MJ, Beaulieuu D, et al, eds. Advancing Federal Sector Health Care. New York: Springer Verlag; 2000, 79–96.
30. Albrecht Gl, Devlieger PJ. The disability paradox: High quality of life against the odds. Soc Sci Med. 1999;48:977–988.
31. Daltroy L, Larson M, Eaton H, et al. Discrepancies between self-reported and observed physical function in the elderly: the influence of response shift and other factors. Soc Sci Med. 1999;48:1549–1561.
32. Nordeson A, Engstrom B, Norberg A. Self-reported quality of life for patients with progressive neurological diseases. Qual Life Res. 1998;7:257–266.
33. Gardner I, Kendig H, Brooke L. Social isolation in the Australian veteran community: needs and interventions. Lincoln Gerontology Centre School of Public Health, La Trobe University, Victoria, Australia; 1997, unpublished manuscript.
34. Sullivan G, Burnam A, Koegel P, Hollenberg J. Quality of life of homeless persons with mental illness: Results from the course-of-homelessness study. Psychiatric Services. 2000;51:1135–1141.
35. Nodhturft V, Schneider JM, Hebert P, Bradham DD, Bryant M, Phillips M, Russo K, Goettelman D, Aldahondo A, Clark V, Wagener S. Chronic disease self-management improving health outcomes. Nursing Clinics of North America. 2000;35(2):507–510.
36. Rosenheck R, Leda C, Gallup P. Combat stress, psychosocial adjustment, and service use among homeless Vietnam veterans. Hosp Community Psychiatry. 1992;43:145–149.
37. Rosenheck R, Fontana A. A model of homelessness among male veterans of the Vietnam War generation. Am J Psychiatry. 1994;151:421–427.
38. Fontana A, Rosenheck R. Traumatic war stressors and psychiatric symptoms among World War II, Korean, and Vietnam War veterans. Psychol Aging. 1994;9:27–33.
39. Irving L, Telfer L, Blake D. Hope, coping, and social support in combat-related posttraumatic stress disorder. J Trauma Stress. 1997;10:465–479.
40. Benotsch E, Brailey K, Vasterling J, Uddo M, Constans J, Sutker P. War zone stress, personal and environmental resources, and PTSD symptoms in Gulf War veterans: A longitudinal perspective. J Abnorm Psychol. 2000;109:205–213.
41. Gamache G, Rosenheck R, Tessler R. Factors predicting choice of provider among homeless veterans with mental illness. Psychiatr Serv. 2000;51:1024–1028.

disability; homeless veterans; perceived health status; tuberculosis

© 2004 Lippincott Williams & Wilkins, Inc.