AIDS and Behavior

, Volume 13, Issue 6, pp 1222–1232

The Effects of HIV Stigma on Health, Disclosure of HIV Status, and Risk Behavior of Homeless and Unstably Housed Persons Living with HIV

Authors

    • Division of HIV/AIDS PreventionNational Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention
  • Sherri L. Pals
    • Division of HIV/AIDS PreventionNational Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention
  • Daniel P. Kidder
    • Division of HIV/AIDS PreventionNational Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention
  • Cari Courtenay-Quirk
    • Division of HIV/AIDS PreventionNational Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention
  • David R. Holtgrave
    • Department of HealthBehavior and Society, Johns Hopkins Bloomberg School of Public Health
Original Paper

DOI: 10.1007/s10461-008-9455-4

Cite this article as:
Wolitski, R.J., Pals, S.L., Kidder, D.P. et al. AIDS Behav (2009) 13: 1222. doi:10.1007/s10461-008-9455-4

Abstract

HIV-related stigma negatively affects the lives of persons living with HIV/AIDS (PLWHA). Homeless/unstably housed PLWHA experience myriad challenges and may be particularly vulnerable to the effects of HIV-related stigma. Homeless/unstably housed PLWHA from 3 U.S. cities (N = 637) completed computer-assisted interviews that measured demographics, self-assessed physical and mental health, medical utilization, adherence, HIV disclosure, and risk behaviors. Internal and perceived external HIV stigma were assessed and combined for a total stigma score. Higher levels of stigma were experienced by women, homeless participants, those with a high school education or less, and those more recently diagnosed with HIV. Stigma was strongly associated with poorer self-assessed physical and mental health, and perceived external stigma was associated with recent non-adherence to HIV treatment. Perceived external stigma was associated with decreased HIV disclosure to social network members, and internal stigma was associated with drug use and non-disclosure to sex partners. Interventions are needed to reduce HIV-related stigma and its effects on the health of homeless/unstably housed PLWHA.

Keywords

HIV stigmaDiscriminationPrejudiceHomelessnessAccess to medical careAdherenceSexual risk behaviorHIV disclosureSocial support

Introduction

Since the beginning of the AIDS epidemic, people living with HIV infection and AIDS (PLWHA) have experienced considerable stigma and discrimination (Herek and Capitanio 1999; Herek et al. 2002; Herek and Glunt 1988; Parker and Aggleton 2003). HIV-related external stigma reflects the negative social identity ascribed to PLWHA by some people who are not affected by HIV, which results in the social marginalization of PLWHA (Parker and Aggleton 2003). External HIV stigma is associated with blaming PLWHA for their illness, misperceptions about HIV transmission, fear of casual contagion, and negative attitudes about members of groups that have been disproportionately affected by HIV (Derlega et al. 1998; Herek and Capitanio 1993; Herek 1999; Herek et al. 2002; 2005; Parker and Aggleton 2003; Shapiro 2005). External HIV stigma has decreased since the beginning of the epidemic, but significant numbers of Americans continue to hold negative beliefs about PLWHA and are uncomfortable coming into contact with them (Centers for Disease Control Prevention 2000; Herek et al. 2002).

External HIV stigma can be internalized by PLWHA, resulting in negative self-image, feelings of shame or guilt, and other manifestations of internal (or felt) stigma (Lee et al. 2002). Stigma is not always internalized, and many PLWHA maintain positive self-images (Buseh and Stevens 2006; Siegel et al. 1998). Internal and external HIV stigma not only lead to social exclusion and discrimination, but also have the potential to affect individual and community health (Valdiserri 2002). For instance, stigma is associated with avoidance of HIV/STD testing, delayed seeking of HIV/STD medical care, reluctance to disclose seropositive HIV status to others, non-adherence to HIV treatment, and poorer self-assessed mental and physical health (Babalola 2007; Chesney and Smith 1999; Courtenay-Quirk et al. 2006; Fortenberry 1997; Fortenberry et al. 2002; Kang et al. 2006; Lee et al. 2002; Murphy et al. 2006; Rintamaki et al. 2006; Valdiserri 2002; Vanable et al. 2006).

Homeless persons are at increased risk of HIV infection and homelessness negatively affects the health of PLWHA (Aidala and Sumartojo 2007; Kidder et al. in press; 2007a; Shubert and Bernstine 2007; Wolitski et al. 2007). Homeless and unstably housed PLWHA may be more likely to experience HIV-related stigma, but little is known about how different subgroups of PLWHA experience HIV stigma. Some PLWHA, such as homeless PLWHA, may be particularly vulnerable to HIV stigma because they are viewed as “responsible” for their infection or experience multiple layers of stigma associated with other stigmatized personal characteristics, such as race/ethnicity, gender, age, sexual orientation, or housing status (Dodds 2006; Emlet 2006; Fullilove and Fullilove 1999; Herek et al. 2005; Herek and Glunt 1988; Herek and Capitanio 1999; Kang et al. 2006; Reidpath and Chan 2005). In particular, homeless PLWHA may experience greater stigma than housed PLWHA because of the negative social status ascribed to those who are homeless, blame that is attributed to homeless persons for their lack of stable housing, reduced familial and other sources of social support to cope with stigma, and opposition to housing for homeless PLWHA in some communities (Link et al. 1995; Takahasi 1999). Despite the likely prevalence and importance of HIV stigma among homeless and unstably housed PLWHA, there is a relative lack of research on this issue.

The amount of stigma experienced by some PLWHA may also differ because negative attitudes about PLWHA may be more or less prevalent in their communities and social networks (Centers for Disease Control Prevention 2000; Rao et al. 2008; Swendeman et al. 2006). Highly educated persons tend to have less stigmatizing beliefs about PLWHA (Bos et al. 2001; Centers for Disease Control Prevention 2000). Whites may have more strongly negative feelings toward PLWHA than do African Americans, but African Americans may express a stronger desire to avoid contact with PLWHA (Herek et al. 1993). HIV stigma also exists within the gay community (Courtenay-Quirk et al. 2006). Men who have sex with men (MSM) living with HIV/AIDS, however, may experience less stigma than other PLWHA because of the gay community’s collective experience with HIV/AIDS (Revenson and Schiaffino 2000) and more positive attitudes toward homosexuality, which have been associated with decreased HIV-related stigma in general population studies (Herek 1999; Herek and Capitanio 1999).

The present study describes levels of HIV-related stigma observed in a 3-city study of homeless and unstably housed PLWHA. Homeless and unstably housed PLWHA represent a vulnerable population that has been shown to be at increased risk for multiple health threats including substance abuse, mental illness, violence, poor access and adherence to HIV medical care, and high-risk sexual practices (Aidala et al. 2007; Aidala and Sumartojo 2007; Eastwood and Birnbaum 2007; Henny et al. 2007; Kidder et al. 2007a; Leaver et al. 2007; Shubert and Bernstine 2007; Wolitski et al. 2007). As such, this population may be particularly vulnerable to HIV-related stigma and its negative effects. Stigma may act as a stressor that affects stigmatized individuals’ ability to cope with HIV and take appropriate action to protect their own and others’ health (Meyer 1995) or may adversely affect health due to discrimination in access to or quality of health care (Sohler et al. 2007). We hypothesized that higher levels of internal, perceived external, and total HIV-related stigma would be associated with poorer physical and mental health and associated health behaviors. In addition, we hypothesized that HIV-related stigma would be associated with less frequent disclosure of HIV serostatus and with increased sexual risk behavior.

Methods

Participants and Design

Data for these analyses were taken from the Housing and Health Study, a randomized controlled trial that evaluated the effects of housing rental assistance on the health and risk behavior of homeless and unstably housed PLWHA. The study rationale and methods are described in detail elsewhere (Kidder et al. 2007b). All research activities were approved by institutional review boards at the CDC and the collaborating institutions conducting the research. As an additional safeguard, the project received a Federal Certificate of Confidentiality.

Participants were recruited through a variety of marketing and community outreach efforts conducted in non-clinical settings by local housing service agencies in Baltimore, MD, Chicago, IL, and Los Angeles, CA. Potential participants were screened by these agencies to determine eligibility for rental assistance supported by the Department of Housing and Urban Development’s Housing Opportunities for Persons with AIDS (HOPWA) program. HOPWA-eligible persons were referred to the local research site, which was in a separate location, where interested persons were screened for research eligibility.

Potential participants were determined to be eligible if they were: (1) 18 years of age or older, (2) HIV-seropositive, (3) homeless or at severe risk of homelessness, (4) had income less than 50% of median area income, (5) spoke English or Spanish, and (6) were willing and able to provide informed consent. Over 3 baseline sessions, eligible participants completed a computer-based interview consisting of interviewer- and self-administered sections, provided a blood sample for CD4 and HIV RNA viral load testing, and received an HIV intervention. These interviews were conducted in 2004 and 2005 and took approximately 90 minutes to complete. Participants received a total of $40 for the baseline sessions.

Measures

The baseline interview included measures of housing history, physical and mental health, access to medical care, adherence to HIV medication, substance use, sexual practices, disclosure of HIV status, and sociodemographic characteristics. Key measures are described below.

Housing Status

Persons who reported sleeping in shelters or other facilities for homeless persons or places not suitable for human habitation (e.g., cars, parks, sidewalks) at least one night in the prior 90 days were considered to have been recently homeless. Persons who were frequently relocated, or moved between temporary housing situations, or had received notice that they were to be evicted from their current residence were considered to be unstably housed or at severe risk of homelessness (Kidder et al. 2007b).

Mental Health

Three standardized mental health assessments were administered. Depression was assessed using the 10-item CES-D (Andresen et al. 1994). Items were scored on a scale of 1 = rarely (<1 day) to 4 = most days (5–7 days). Item scores were recoded and summed so that higher scores represent greater depressed mood (Andresen et al. 1994). Perceived stress was assessed using the 10-item Perceived Stress Scale (Cohen and Williamson 1988). Perceived stress was scored on a 1 = never to 5 = very often scale. Items were recoded and summed so that higher total scores indicate higher perceived stress. Three mental health scales from the Medical Outcomes Study Short Form-36 v.2 (SF-36; Ware et al. 2000) were used to assess general mental health, social functioning, and role/emotional functioning. The SF-36 has been used with a wide variety of populations and has been shown to have advantages over HIV-specific measures such as the MOS-HIV (Shahriar et al. 2003). The 3 SF-36 scales were also weighted and combined to yield a Mental Component Summary (MCS) score, which is norm-based (population M = 50, SD = 10). Higher scores indicate better mental health or functioning. Internal consistency for all of the mental health scales was very good, with alphas ranging from 0.76 to 0.85.

Physical Health

Self-assessed physical health was measured using the SF-36 (Ware et al. 2000). Five subscales that assess different aspects of physical health (general health, physical functioning, role/physical limitations, bodily pain, and vitality) were administered (α = 0.76 to 0.91). The 5 SF-36 scales were also weighted and combined to yield a Physical Component Summary score, which is norm-based (population M = 50, SD = 10). Higher scores on all SF-36 measures indicate better self-perceived physical health. Blood specimens were obtained for CD4 lymphocyte testing and HIV-1 viral load (Roche Amplicor HIV-1 Monitor Test, Version 1.5) at a central laboratory. The limit of detection for viral load was 400 copies of HIV-1 RNA per ml.

Utilization of Medical Care

Medical care utilization was assessed by asking about the receipt of care in the prior 6 months from 8 different sources (e.g., private physician, medical clinic, hospital, emergency room). A dichotomous variable was computed, representing receipt of care from any source. HIV medication utilization was assessed by a single item assessing whether participants reported current use of any HIV medications. Participants who reported using HIV medication were provided pictures and names of approved HIV medications and asked to indicate which medications they were currently using. For each medication they were on, participants were asked how many pills they were supposed to take each day. The computer program summed the total number of pills they were supposed to take over two days and the interviewer asked how many pills they missed taking over the previous two, and then seven, days. Persons who missed taking any pills during the time period were considered non-adherent. In addition, because homeless persons are less likely to be taking optimal HIV treatment regimens (Kidder et al. 2007a), a variable representing whether participants were taking a recommended highly active antiretroviral therapy (HAART) regimen was also computed (DHHS Panel on Antiretroviral Guidelines for Adults and Adolescents 2006).

Disclosure of HIV Status

The percentages of close friends, family members, and other support network members who were aware of the respondent’s HIV status were assessed separately by first asking about the number of persons in a given network and then the number of network members who knew the respondent’s HIV status. The proportion of the entire social network disclosed to was calculated by dividing the numbers of persons who knew the respondent’s HIV status by the total numbers of friends, family, and other persons in the respondent’s networks. These variables were not normally distributed and were categorized as follows: <50% disclosed to, ≥50% and <100%, and 100%. For main and non-main partners, participants were asked a series of questions to assess the HIV status of these partners and whether respondents had disclosed their own HIV status to these partners before the first sexual encounter. A variable was computed to represent whether respondents had failed to disclose their HIV status to any partner before the first sexual encounter with that partner.

Drug Use and Sexual Behavior

Participants were asked about injected and non-injected drug use during their lifetime and the prior 90 days. Two dichotomous variables indicating any lifetime injected drug use and any drug use in the prior 90 days were computed. Participants were asked separate sets of questions about their sexual practices with their main partner and HIV-seropositive and HIV-seronegative/unknown serostatus non-main partners. Participants first answered the complete set of questions for main partners, which included the partner’s HIV status. These were followed by questions about non-main partners, which were asked separately for HIV-seropositive and HIV-seronegative/unknown status partners. For all partner types, participants were asked whether they had engaged in a given sex behavior and if they had, the number of times they had done so with and without a condom. The following behaviors were assessed: (1) receptive oral sex, (2) insertive oral sex, (3) vaginal intercourse, (4) receptive anal intercourse, and (5) insertive anal intercourse. Dichotomous variables were computed to indicate whether respondents reported any unprotected vaginal or anal sex with any partner or with HIV-seropositive or seronegative/unknown status partners.

HIV Stigma

Two 6-item scales adapted from prior research (Berger et al. 2001) were used to measure internal and perceived external HIV stigma (α = 0.88 and 0.91, respectively). The scales were shortened for this study and the external stigma items reworded to more explicitly focus on the respondents’ social networks. Prior adaptations of these scales have been shown to be valid and reliable in other populations. (Bunn et al. 2007; Wright et al. 2007) Responses were reported on a 4-point scale (strongly disagree, disagree, agree, strongly agree), with larger values representing higher levels of stigma. Scale scores were computed by averaging the responses to items on each scale, and a total HIV stigma score (α = 0.92), was computed by averaging responses to both scales.

Statistical Analyses

Principal components factor analysis with varimax rotation was performed to confirm the factor structure. The analysis confirmed the presence of two orthogonal factors with factor loadings ranging from 0.83 to 0.64 and 0.84 to 0.74 for the internal and perceived external stigma scales, respectively. The remaining data analyses focused on assessing the association of the 3 HIV stigma measures with participant characteristics, health status, risk behavior, and other outcomes. Linear regression analyses in which each HIV stigma measure served as a dependent variable were used to identify participant characteristics associated with stigma. Participant characteristics that were associated with any of the stigma measures (P < .10) were controlled for in logistic and linear regression analyses that examined the association between the health and risk behavior outcomes of interest and the HIV stigma measures.

Results

Characteristics of study participants are shown in Table 1. Scores on the HIV-stigma scales and responses to individual items indicated that the majority of respondents experienced at least some manifestations of HIV-related stigma (Table 2). Factor analysis confirmed that the internal and perceived external HIV-stigma scales represented distinct constructs (Table 2). Multivariable models showed that participants with no education beyond high school, and females (differences between MSM and non-MSM men were not statistically significant) had higher scores on all 3 HIV-stigma measures (Table 1). Total HIV stigma was higher among those with a more recent HIV diagnosis. Higher internal and total stigma scores were also observed for those who were recently homeless. Recruitment site was associated with perceived external stigma, with higher scores observed among participants recruited from Chicago.
Table 1

Independent associations of demographic characteristics of homeless and unstably housed persons living with HIV/AIDS with internal, perceived external, and total stigma—Baltimore, Chicago, and Los Angeles, 2004–2005

Characteristic

No.

Percent

Internal stigma

External stigma

Total stigma

Meana

Fb

Meana

Fb

Meana

Fb

Recruitment site

1.60

 

3.16*

 

2.08

    Baltimore

213/637

33.4

2.10

2.52

2.31

    Chicago

213/637

33.4

2.21

2.68

2.45

    Los Angeles

211/637

33.1

2.24

2.52

2.38

Housing status

4.26*

 

2.79

 

4.44*

    Unstably housed/severe risk of homelessness

463/637

72.7

2.11

2.51

2.31

    Recently homeless

174/637

27.3

2.25

2.63

2.44

Race/ethnicityc

N/A

 

N/A

 

N/A

    Hispanic

59/635

9.3

2.31

2.59

2.45

    Black/African-American

499/635

78.6

2.14

2.54

2.34

    White

51/635

8.0

2.13

2.27

2.20

    Other

26/635

4.1

2.04

2.44

2.24

Age

0.84

 

1.15

 

1.14

    18–29

66/637

10.4

2.29

2.70

2.50

    30–39

171/637

26.8

2.17

2.50

2.33

    40–49

307/637

48.2

2.19

2.57

2.38

    50 and older

93/637

14.6

2.09

2.52

2.30

Gender and HIV exposure category

 

3.36*

 

3.87*

 

4.60*

    Male—men who have sex with men (MSM)

262/637

41.1

2.08

2.45

2.27

    Male—non-MSM

182/637

28.6

2.18

2.58

2.38

    Female

193/637

30.3

2.29

2.68

2.48

Ever injected drugsc

N/A

 

N/A

 

N/A

    Yes

181/635

28.5

2.10

2.45

2.27

    No

454/635

71.5

2.17

2.55

2.36

Household income (past 30 days)c

N/A

 

N/A

 

N/A

    Less than $600

266/584

45.5

2.19

2.57

2.38

    $600 or greater

318/584

54.5

2.12

2.48

2.30

Education

4.69**

 

5.95**

 

6.32**

    Less than high school

226/637

35.5

2.28

2.62

2.45

    High school graduate or equivalent

186/637

29.2

2.22

2.69

2.45

    More than high school

225/637

35.3

2.05

2.42

2.23

Years since HIV diagnosis

 

2.53

 

2.57

 

3.16*

    <5 years

198/619

32.0

2.31

2.69

2.50

    5 to <10 years

173/619

27.9

2.20

2.63

2.42

    10 to <15 years

146/619

23.6

2.11

2.46

2.29

    ≥15 years

102/619

16.5

2.10

2.51

2.31

Ever diagnosed with AIDS

.32

 

.21

 

.00

    Yes

246/636

61.3

2.16

2.59

2.38

    No

390/636

38.7

2.20

2.56

2.38

aAll means except those otherwise noted are adjusted for all variables having a significant bivariate association (data not shown) with one or more of these HIV stigma scale scores. Variables included as covariates were: recruitment site, housing status, age, gender and HIV exposure category, education, years since HIV diagnosis, and ever diagnosed with HIV

bF tests were based on multiple linear regression models (one for each stigma scale) that included predictor variables representing the participant characteristics that were associated (P < .10) with one or more of the stigma scales in bivariate analyses (data not shown)

cVariable was not associated with stigma scales in bivariate analyses (P ≥ .10) and was not included in multiple regression models. Means reported for these variables are unadjusted and N/A indicates that F-tests from multiple regression analyses are not applicable

P < .05; ** P < .01

Table 2

Internal, perceived external, and total stigma among homeless and unstably housed persons living with HIV/AIDS—Baltimore, Chicago, and Los Angeles, 2004–2005

 

N

Mean

SD

Agree or strongly agree (%)

Internal stigma scale

637

2.15

0.75

    I feel I am not as good a person as others because I have HIV

637

2.01

0.97

27.6

    Having HIV makes me feel unclean

635

2.15

0.96

33.1

    Having HIV makes me feel I am a bad person

636

1.86

0.80

14.9

    People’s attitudes about HIV make me feel worse about myself

635

2.33

0.97

39.7

    I feel guilty because I have HIV

637

2.25

1.04

37.7

    Having HIV in my body is disgusting to me

634

2.31

0.98

40.2

Perceived external stigma scale

637

2.52

0.78

    People I know would treat someone with HIV as an outcast

637

2.55

0.94

53.5

    People I know would be uncomfortable around someone with HIV

637

2.59

0.94

57.1

    People I know believe that a person with HIV is dirty

635

2.40

0.91

45.4

    People I know would reject someone with HIV

631

2.71

0.96

60.5

    People I know would not want someone with HIV around their children

630

2.54

0.94

52.4

    People I know think that a person with HIV is disgusting

637

2.34

0.93

41.1

Total stigma

637

2.34

0.68

Note: Scores for all individual stigma scale items and for the internal, external and total stigma scales ranged from 1 to 4

Linear regression analyses revealed that HIV stigma was independently associated with poorer self-assessed mental and physical health (Table 3). Increased levels of internal, perceived external, and total HIV stigma were consistently associated with poorer scores on all self-assessed measures of mental and physical health. HIV stigma was not associated with receipt of medical care, current use of any HIV medication, or current use of HAART (data not shown). Perceived external HIV stigma was associated with missing HIV medication doses (P < .05) in the past 2 days (Table 3).
Table 3

Association of internal, perceived external, and total stigma with mental health, physical health, and health care utilization—Baltimore, Chicago, and Los Angeles, 2004–2005

 

N

Internal stigma

External stigma

Total stigma

Β

Fa

β

Fa

β

Fa

Mental health

Depression

617

3.94

124.81**

3.40

95.77**

4.68

147.23**

Perceived stress

616

3.97

114.00**

3.39

85.85**

4.69

132.48**

SF-36: general mental health

618

−6.90

116.06**

−5.59

77.46**

−7.96

127.09**

SF-36: social functioning

616

−5.17

68.69**

−4.65

58.89**

−6.27

83.77**

SF-36: role/emotion

617

−4.71

52.42**

−3.99

39.71**

−5.55

59.68**

SF-36: mental component summary score

614

−6.54

100.36**

−5.21

65.38**

−7.48

108.01**

Self-assessed physical health

SF-36: general health

618

−5.31

91.12**

−4.82

79.17**

−6.47

112.96**

SF-36: physical functioning

618

−2.90

24.90**

−2.81

25.13**

−3.65

32.35**

SF-36: role/physical

617

−3.03

35.80**

−3.17

42.69**

−4.00

51.50**

SF-36: bodily pain

618

−2.36

13.75*

−3.30

29.62**

−3.64

27.11**

SF-36: vitality

618

−4.67

74.34**

−4.00

57.10**

−5.53

85.94**

SF-36: physical component summary score

614

−2.09

16.10**

−2.71

29.86**

−3.08

29.10**

 

N

AOR

95% CI

AOR

95% CI

AOR

95% CI

Laboratory measures of physical health

Viral load detectable

615

1.08

0.85, 1.38

0.89

0.70, 1.12

0.97

0.74, 1.27

CD4 count below 200

615

1.01

0.76, 1.33

0.99

0.76, 1.30

1.00

0.73, 1.36

Self-reported health care and medication utilization

Any medical care in past 90 days

617

1.11

0.83, 1.49

1.05

0.79, 1.39

1.10

0.80, 1.52

Received care from a clinic or private MD in past 6 months

617

1.21

0.95, 1.54

1.18

0.94, 1.49

1.26

0.96, 1.64

Currently taking HAART

618

0.98

0.78, 1.24

1.07

0.85, 1.33

1.03

0.80, 1.34

Missed any HIV medication doses in past 2 daysb

413

1.30

0.92, 1.85

1.40

1.00, 1.95

1.50

1.01, 2.22

Missed any HIV medication doses in past 7 daysb

412

1.05

0.77, 1.44

1.41

1.05, 1.89

1.31

0.93, 1.85

aF tests were based a series of multiple linear regression models (with mental and physical health measures as outcomes) that included as covariates all participant characteristics that were associated (P < .10) with one or more of the stigma scales in bivariate analyses. These were: recruitment site, housing status, age, gender and HIV exposure category, education, years since HIV diagnosis, and ever diagnosed with HIV

bAsked only of persons taking HIV antiretroviral therapy

* P < .001; ** P < .0001

Ordinal logistic regression analyses controlling for participant characteristics associated with HIV stigma revealed that perceived external stigma was associated with disclosure to fewer social network members (Table 4). Internal HIV stigma was not associated with disclosure to friends, family, and other social network members but was significantly associated with non-disclosure of HIV status to sex partners. Higher levels of internal HIV stigma, but not perceived external HIV stigma, were associated with failure to disclose HIV seropositive status to one or more recent sex partners prior to the first sexual encounter.
Table 4

Association of internal, perceived external, and total stigma with disclosure of HIV status, sexual behavior, and substance use—Baltimore, Chicago, and Los Angeles, 2004–2005

Outcome variable

Na

Internal stigma

External stigma

Total stigma

AORb

95% CI

AORb

95% CI

AORb

95% CI

Disclosure of HIV status

  % of social networkc

605

0.99

0.80, 1.23

0.73

0.59, 0.90

0.81

0.64, 1.03

    % of friendsc

502

0.95

0.73, 1.24

0.69

0.53, 0.90

0.77

0.57, 1.03

    % of familyc

541

1.02

0.79, 1.33

0.75

0.59, 0.96

0.84

0.64, 1.11

    % of other personsc

479

0.82

0.60, 1.13

0.63

0.46, 0.88

0.66

0.46, 0.95

  Did not disclose to 1+ sex partnersd

368

1.47

1.09, 2.00

1.11

0.83, 1.50

1.37

0.97, 1.94

Sexual behavior in past 90 days

  Any sex

611

0.89

0.69, 1.14

1.00

0.78, 1.28

0.93

0.70, 1.23

  Unprotected vaginal or anal sex—any partner

599

1.02

0.79, 1.31

0.86

0.67, 1.10

0.92

0.69, 1.22

  Unprotected vaginal or anal sex with HIV seropositive partner

610

1.10

0.79, 1.53

0.97

0.70, 1.34

1.04

0.72, 1.51

  Unprotected vaginal or anal sex with HIV seronegative or unknown serostatus partner

603

1.04

0.77, 1.40

0.88

0.66, 1.18

0.94

0.67, 1.32

Any drug use in past 90 days

614

1.38

1.09, 1.74

1.08

0.86, 1.34

1.28

0.99, 1.65

aDenominator for analysis

bOdds ratios were adjusted for recruitment site, housing status, age, gender/HIV exposure category, sexual identity, education, years since HIV diagnosis, and ever diagnosed with AIDS in multiple logistic regression models

cOdds ratios and 95% confidence intervals are estimated using ordinal logistic regression models with disclosure category (0 to <50%, 50 to <100%, 100%) as the dependent variable. Odds ratios represent the association of stigma with higher levels of disclosure

dAmong persons with 1+ sex partners in the prior 90 days, failure to disclose HIV status before first sex to 1 or more of these partners

None of the HIV stigma scales was associated with being sexually active or with risky sexual behavior, but internal HIV stigma was independently associated with increased drug use when other participant characteristics were controlled.

Discussion

More than 25 years since the first cases of AIDS were recognized, stigma associated with the disease continues to negatively affect the lives of PLWHA. In this study, external HIV stigma was widely perceived to exist within the social networks of homeless/unstably housed PLWHA. External stigma had been internalized by some participants who acknowledged negative feelings (including guilt and disgust) about their own HIV status. The continued presence of perceived external and internal HIV stigma demonstrates a need for increased effort to combat HIV stigma at a societal level and to strengthen supportive services for homeless (and perhaps other) PLWHA that improve their ability to cope with stigma and the myriad stressors of life with HIV/AIDS.

Participants’ experience of HIV stigma varied and was sometimes greater among those who most likely experienced multiple sources of stigma. This may be due to the layering of stigma associated with other stigmatized characteristics (e.g., homelessness, limited educational attainment), the amount of HIV-related stigma within various social networks, and variations in individual ability to cope with stigma. Evidence for the layering of stigma was not always observed, however. For example, racial and ethnic minority participants reported comparable levels of HIV-related stigma and MSM in this study reported less HIV-related stigma. Shared experiences within these subgroups and coping mechanisms developed to deal with racism and homophobia may in part account for these findings. In the case of MSM, the lower rates of HIV-related stigma may exist because of the history of the gay community’s positive response to the HIV/AIDS crisis and the likely presence of other PLWHA in the networks of HIV-seropositive MSM. Other research, however, has found higher levels of HIV stigma among gay-identified MSM and youth (Swendeman et al. 2006). Given the evolving profile of the HIV epidemic in the United States, additional research is needed to better understand HIV stigma, the ability of PLWHA to cope with it, and the benefits of improving social support in various subgroups of PLWHA.

Perceived external, internal, and total HIV stigma were negatively associated with physical and mental health. These associations were strong and observed for all self-assessed measures employed in this study, confirming earlier findings (Courtenay-Quirk et al. 2006; Lee et al. 2002; Murphy et al. 2006). Given this pattern of results, it was surprising that laboratory-based measures of physical health (CD4 and HIV viral load) were not associated with any HIV-stigma measure.

The reasons for these incongruous findings are not clear. It is possible that persons who view the world and their own situation negatively perceived higher levels of stigma and rated themselves less healthy than they actually were. Alternatively, HIV stigma may have had real effects on mental health, which in turn negatively affected self-assessed physical health. Negative affect has been associated with increased symptom reporting and poorer self-assessed physical health in prior research (Koller et al. 1996; Pettit et al. 2001; Salovey and Birnbaum 1989; Salovey et al. 2000). It is also possible that HIV stigma has indirect effects on physical health that were not adequately captured by CD4 and HIV viral load but were better measured by the more comprehensive SF-36 health assessment. Additional research on the effects of HIV stigma on health, including studies employing longitudinal designs and multiple objective assessments of physical health, is warranted.

Perceived external HIV stigma was associated with disclosure to fewer network members, suggesting that PLWHA in this study evaluated their social environment and chose not to disclose to those that would not be supportive. In an unsupportive social environment, concealment of HIV status may be an appropriate response. In general, however, concealment of HIV status acts as a barrier to social support (Hays et al. 1993; Mansergh et al. 1995; Serovich et al. 2000; Smith and Rapkin 1996; Zea et al. 2005) and may negatively affect the health of PLWHA (Strachan et al. 2007). Interestingly, disclosure to sex partners was not associated with perceived external stigma, but it was associated with internal stigma. Given the cross-sectional nature of this study it is not known whether higher levels of internal stigma led to non-disclosure or if failure to disclose increased internal stigma because respondents later felt shame or guilt about having sex without disclosing their HIV status.

The present results suggest that for homeless and unstably housed PLWHA, HIV stigma may play only a limited role in the sexual transmission of HIV. None of the HIV stigma scales were associated with sexual risk behavior. However, internal HIV stigma was associated with drug use and non-disclosure of HIV status to sex partners. These findings, coupled with the lack of association between perceived external stigma and drug use or disclosure to sex partners, suggests that efforts to help PLWHA overcome feelings of guilt or shame about their HIV status might affect drug use, which may be used as a coping strategy by some PLWHA, and increase disclosure to sex partners.

Overall, evidence regarding a direct association between HIV stigma and unprotected sex is limited and mixed (Courtenay-Quirk et al. 2006; Preston et al. 2004, 2007; Swendeman et al. 2006). The finding that HIV stigma was associated with disclosure and not sexual risk behavior is consistent with prior research indicating that these behaviors are distinct, determined by different factors, and often not causally linked (i.e., disclosure of HIV status does not consistently lead to reduced sexual risk) (Crepaz and Marks 2003; Hart et al. 2005; Marks and Crepaz 2001; Parsons et al. 2005; Simoni and Pantalone 2005; Wolitski et al. 1998). It is premature to draw firm conclusions about the association between HIV stigma and sexual risk, particularly given the association between internal stigma and recent substance use, as it remains understudied.

Caution in interpreting these findings is warranted due to the cross-sectional nature of the study and other limitations. The causal nature of the relationships observed between HIV stigma and various outcomes cannot be determined from these data—longitudinal analyses are needed to establish causal relationships. With the exception of CD4 and HIV viral load, the data were based on self-report, which may be subject to recall and other biases (Catania 1999; Catania et al. 1990; Ostrow and Kalichman 2000). However, the study used a number of well-established measures that have been shown to be valid and highly reliable in varied samples. Participants were not representatively sampled and some subgroups, such as African Americans and homeless persons living on the street, may have been under- or over-sampled. Thus, even though a range of recruitment strategies were used, the sample may not adequately represent the characteristics or experiences of homeless and unstably housed PLWHA in these or other American cities.

This study adds to the existing evidence showing the persistence and negative impact of HIV stigma in vulnerable subgroups of PLWHA in the United States. There are encouraging signs that HIV stigma has declined over time (Herek et al. 2002), but it remains a troubling social and public health problem. HIV stigma affects not only people who have been diagnosed with HIV, but also has negative effects on those with undiagnosed HIV infection, those who are at-risk for HIV or are members of groups at increased risk, and the friends and family members of PLWHA (Bogart et al. 2008; Murphy et al. 2006). Exposure to information and positive portrayals of PLWHA in the mass media and other intervention strategies have been shown to reduce HIV stigma (Brown et al. 2003; O’Leary et al. 2007; Hutchinson et al. 2007), but too few have been implemented widely in the United States. There is a critical need to mobilize additional resources to counter the pernicious presence of HIV stigma and its deleterious effects.

Acknowledgments

This study was funded by the Centers for Disease Control and Prevention of the Department of Health and Human Services through a contract (200-2001-0123, Task 9) with Research Triangle Institute, Inc. Funding for tenant-based rental assistance was provided by the Department of Housing and Urban Development. The questionnaire was approved by the US Office of Management and Budget (OMB No. 0920–0628). We would like to thank the many people who made this study a success. The following individuals played a lead role in the design or implementation of the Housing and Health Study (in alphabetical order) include Angela Aidala (Columbia University), Arturo Bendixen (AIDS Foundation of Chicago), Kate Briddell (City of Baltimore, Department of Housing and Community Development), Shahry Deyhimy (City of Los Angeles Housing Department), Paul Dornan (HUD), David Harre (Independent consultant, formerly with HUD), Myrna Hooper (Housing Authority of the City of Los Angeles), Jennafer Kwait (RTI), Fred Licari (RTI), Shirley Nash (City of Chicago Department of Public Health), Scott Royal (RTI), William Rudy (HUD), Ron Stall (University of Pittsburgh), Esther Sumartojo (CDC), and David Vos (HUD). In addition, the following persons were instrumental to the success of the study: Rusty Bennett, Maria Caban, Sylvia Cohn, Lynne Cooper, Jay Cross, Maria DiGregorio, Clyde Hart, Kirk Henny, Kelly Kent, Lee Lam, Eugene Little, Ellen McCarty and Jerusalem House, Joyce Moon Howard, Noelle Richa, Danny Ringer, Randy Russell, Ruth Schwartz, and Tom Spira. This study would not have been possible without the many contributions and support of the collaborating HUD grantee agencies in the participating cities: City of Baltimore Department of Housing and Community Development, City of Chicago Department of Public Health, AIDS Foundation of Chicago, City of Los Angeles Housing Department, Housing Authority of the City of Los Angeles, Shelter Partnership, Housing Authority of the County of Los Angeles, County of Los Angeles Office of AIDS Programs and Policies, Los Angeles Homeless Services Authority. Most importantly, we would like to acknowledge and thank the individuals who volunteered to participate in this study. Their strength and resilience in dealing with the multiple challenges they faced in their lives, in particular those associated with living with HIV, housing instability, and stigma, should be an inspiration to us all.

Disclaimer

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.

Copyright information

© Springer Science+Business Media, LLC 2008