Advancing Sexual and Reproductive Health and Rights
 
Perspectives on Sexual and Reproductive Health
Volume 36, Number 5, September/October 2004
DIGEST

Subgroups of Homeless In San Francisco Have Different HIV Risk Factors

Members of San Francisco's indigent population have many sexual and other risk factors that put them at particularly high risk of HIV infection.1 Among a sample of the city's homeless and marginally housed population, the overall HIV prevalence was 11%; for the subgroup of men who have sex with men, the prevalence was 30%. Sexual risk factors including being a man who has sex with men, having ever had syphilis, having ever traded sex for drugs or money, and having had receptive anal sex within the previous year were significantly associated with HIV infection among the overall sample (odds ratios, 1.6–4.6). Certain sexual risk factors were also significant among men who have sex with men and injection drug users.

To examine the prevalence of and risk factors associated with HIV infection among homeless and marginally housed adults in the San Francisco area, researchers interviewed a sample of clients at overnight shelters, midday free-meal programs and low-cost residential hotels throughout the city and county. Between April 1996 and December 1997, participants answered questions about their social and demographic backgrounds; their sexual behavior and substance use histories; and their current, chronic and lifetime homelessness status. In addition, participants provided a blood sample for HIV testing. The final sample consisted of 2,508 individuals. The researchers used bivariable and logistic regression analyses to examine the independent associations between respondents' characteristics and HIV status for the overall sample and for three subgroups: men who have sex with men, injection drug users (excluding men who have sex with men) and those not included in either of the other subgroups (referred to as the residual subgroup).

Three-quarters of the sample were male, and six in 10 were nonwhite. The vast majority were aged 30 or older and had been living in the San Francisco area for a year or more; most were heterosexual and had at least a high school education. Twenty-four percent had ever been in prison, and 14% had had a blood transfusion between 1978 and 1985. Seventy-eight percent of the sample had been homeless as adults, 43% had been homeless the night before the survey and 48% had spent the previous night in a single-room occupancy hotel. The median monthly income was $585.

A history of drug use was common among the sample; for example, 45% of respondents reported having ever injected drugs, and 63% had used crack. Current drug use was also common: Forty-four percent reported having used crack, other cocaine, heroin or stimulants within the last 30 days, and 35% reported having injected drugs during that time. Twenty-seven percent had ever shared a needle or syringe, 26% had ever taken part in a needle exchange program and 13% had ever used a shooting gallery.

Although less prevalent than drug-related risk factors, sexual risk factors were somewhat common among the homeless and marginally housed. Nineteen percent were men who have sex with men, 18% had had five or more sexual partners in the past year, 11% had had receptive anal sex in the past year and 9% had ever had syphilis. Nearly one-third had ever traded sex for money or drugs.

Overall, the prevalence of HIV infection among the sample was 11%. In bivariable analysis, HIV infection was significantly associated with being male, white, 18–29, bisexual, a gay male and having lived in San Francisco for at least one year. In addition, HIV infection was associated with all sexual and drug use risk factors, except injection of cocaine. Of the three subgroups, men who have sex with men had the highest prevalence of HIV (30%); the prevalence of infection was 8% among injection drug users and 5% in the residual subgroup. The bivariable results suggest that each subgroup has different risk factors for HIV infection.

As in the bivariable analyses, the factors found to be significant in the multivariable analyses differed by group; however, for all groups, sexual risk factors seemed to be more important than other factors. In the overall sample, men who have sex with men were more likely than others to be infected with HIV (odds ratio, 4.6). Other factors associated with HIV infection included ever having had syphilis, being a white injection drug user, having had a blood transfusion, being nonwhite, having ever traded sex for money or drugs, and having had receptive anal sex in the last year (1.6–2.2). Among men who have sex with men, those who were white and traded sex were more likely than others to be infected with HIV (5.9); other significant risk factors for this subgroup were being nonwhite, having had receptive anal sex in the last year and ever having had syphilis (2.0–3.4). Having ever had syphilis was also significantly associated with increased risk of HIV infection among the injection drug use group (3.3), along with having less than a 12th-grade education (2.6), ever having been in prison and having had a blood transfusion (2.1 for each). Finally, among the residual subgroup, HIV infection was associated with having had five or more sexual partners in the last year and being a female crack cocaine user who traded sex (2.9 and 6.1, respectively).

The researchers comment that the high HIV prevalence among the indigent of San Francisco, which is five times that of the general public, is "striking." They conclude that although the homeless and marginally housed have "numerous complex problems such as extreme poverty, social marginalization and drug abuse," sex is the most important factor in regard to HIV infection among this population. The researchers suggest that HIV intervention programs for men who have sex with men—the subgroup with the highest prevalence of infection—should "reinforce the focus on sexual risk."—J. Rosenberg

REFERENCE

1. Robertson MJ et al., HIV seroprevalence among homeless and marginally housed adults in San Francisco, American Journal of Public Health, 2004, 94(7):1207–1217.