Katz, Mitchell H.
Section Editor(s): Feldman, Mitchell D. M.D., M.Phil.*; Feldman, Eric A. J.D., Ph.D.**; Coates, Thomas J. Ph.D.***
The HIV/AIDS epidemic struck San Francisco's gay male community harder than any community in the world. By 1992, AIDS had become the most common cause of death among men of all ages in San Francisco. Among San Francisco residents, AIDS accounted for 15% of deaths and 39% of potential years lost before age 75(1). However, the epidemic has also been more effectively contained in San Francisco than anywhere else. Despite the overall success of San Francisco's prevention efforts, particular subpopulations, especially young gay men, continue to have high rates of seroconversion. The experience of San Francisco therefore has much to teach the world about the successes and ongoing challenges of HIV prevention. I review here the nature of the epidemic in San Francisco among men who report sex with men (MSM), the prevention efforts within this community, and the lessons from San Francisco's experience.
THE AIDS EPIDEMIC IN SAN FRANCISCO
The first AIDS cases were identified in 1980 in San Francisco, Los Angeles, and New York. They occurred exclusively among MSM. Between 1980 and 1995, 22,602 AIDS cases were reported in San Francisco; 20,530 (91%) were in MSM. San Francisco ranks third among metropolitan areas in cumulative AIDS cases, after New York City and Los Angeles. However, San Francisco is markedly smaller than these other two cities, with only 724,000 residents and an estimated population of MSM of only 58,000(2).
Figure 1 shows the incidence of AIDS cases in San Francisco among MSM, MSM who report injection drug use, heterosexual injection drug users, and persons in other risk groups. The curve for MSM dwarfs the other curves, rising sharply from 1981 until 1992 and then dipping in subsequent years. Although not perfectly correlated, AIDS incidence is a relatively accurate reflection of HIV seroincidence, with a time frame shift of approximately 10 years, reflecting the mean time between HIV seroconversion and AIDS(3,4). Therefore, on the basis of the AIDS incidence curves, we can assume that the peak years of HIV seroconversion were between 1981 and 1983, with a subsequent decrease.
Although some of the decrease in AIDS cases observed between 1992 and 1993 is due to the change in the definition of AIDS to include persons with low CD4 lymphocyte counts, the magnitude of the decline is too great to be solely due to the expansion of the case definition. In support of a true decline in AIDS incidence is the corresponding decrease in AIDS mortality. Mortality rates are unaffected by changes in the case definition. Figure 2 shows the steadily increasing mortality due to AIDS in San Francisco among MSM from 1987 to 1992, with subsequent decreases in 1993 and 1994. In contrast, the number of persons dying of AIDS in other risk groups continued to increase in 1993 and 1994.
Cohort data from the San Francisco hepatitis B vaccine cohort also document a decrease of seroconversion rates in the early 1980s among MSM. The study enrolled 359 MSM who were seen at the San Francisco municipal sexually transmitted disease(STD) clinic between 1978 and 1980 for a trial of the hepatitis B vaccine(3). Because the serum specimens from this cohort were stored, the incidence of HIV infection during these early years could be determined retrospectively for the 320 men who consented to have their blood tested. Figure 3 shows the estimated incidence of HIV infection in the hepatitis B vaccine trial. The curve rises steeply between 1979 and 1984 and then levels off to 1-2% newly infected men per year.
Because the men from the hepatitis B vaccine trial were enrolled from an STD clinic, they may have been more likely to be sexually active. However, to be enrolled in the hepatitis B vaccine trial they had to be unexposed to hepatitis B, indicating lower sexual activity in the past. To overcome these biases, the San Francisco Men's Health Study was designed to obtain population-based estimates of seroprevalence. In 1984, investigators sampled 1,034 single men aged 25 to 55 years living in the 19 census tracks with the highest incidence of AIDS in San Francisco(5,6). The data from this study confirm and extend the findings from the hepatitis B vaccine trial. The investigators estimated HIV seroincidence for 1982 through 1983 by comparing HIV seroprevalence on entry into the study for men with and without recent risk behaviors. They estimated an HIV positivity rate among MSM of 22.8% in 1982, rising to 48.6% by 1984 and stabilizing thereafter. For MSM living outside of these 19 census tracks (an estimated 40,000 MSM, or 71%), seroprevalence by 1984 was estimated to be 32 and 39%(7,8). On the basis of these and other data, the Department of Public Health estimates that among MSM there were approximately 8,000 new infections per year in 1982, 6,700 in 1983, 5,600 in 1984, 1,200 in 1985, 1,000 in 1986, and about 500 per year beginning in 1987, with rates remaining stable thereafter.
Several cohort studies document that the decrease in seroconversion rates after 1984 reflect a general decrease in high-risk sexual activities, especially anal intercourse(6,9). Because changes in behavior may occur in response to the education provided to participants in these cohorts, these changes might not reflect the behaviors among MSM who did not participate in studies. As an indirect measure of broad behavior change, the San Francisco Department of Health examined rates of gonococcal proctitis in San Francisco. Rectal proctitis develops almost exclusively among men who engage in receptive anal intercourse. Rates of gonococcal proctitis dropped dramatically, from more than 5,000 cases in 1980 to 4,000 in 1982 and just over 2,000 in 1983 (Fig. 4). Therefore, substantial direct and indirect measures support a true decline in HIV seroconversions.
EARLY PREVENTION EFFORTS
A great deal of the success of San Francisco's prevention efforts can be attributed to the political power of the gay community. By the 1970s, San Francisco had become a haven for gay persons, attracting thousands of gay men and women from other areas of the country and the world. As a result of this migration, San Francisco had the highest concentration of gay persons of any major city in the United States. This high concentration of gay persons, coupled with a historic tolerance for people with alternative lifestyles, resulted in a greater tolerance for gay persons among the non-gay populace than is found in other cities in the United States. Gay cultural events, such as the annual Gay Pride Parade, attracted huge crowds of both gays and heterosexuals.
Importantly, the gay population, estimated at 20% of voting adults, voted as a block for candidates that support gay causes. Politicians could not avoid addressing the gay community's issues because they relied on them for their votes. Therefore, gay men and women were enlisted as influential advisors to non-gay politicians. The political power of gay persons was galvanized by the election of Harvey Milk, an openly gay man, to the San Francisco Board of supervisors in 1978. This political power made the gay community much more influential in San Francisco than in most cities and made the local government more responsive to the AIDS crisis.
Although until 1984 it was not determined for certain that AIDS was caused by a virus, HIV prevention efforts actually began in the early 1980s. Epidemiologic data supported the hypothesis that AIDS was sexually transmitted. Consequently, public health officials and members of the gay community began discussing ways to reduce the rates of infection. In the early 1980s, most information about AIDS was exchanged through one-on-one and group discussions. Gay men, in pairs and larger groups, discussed this alarming illness and its relationship to sexual practices. Because of other STDs that were common among gay men (e.g., hepatitis B, amebiasis, gonorrhea, genital warts), there was a context for these discussions, and many gay men were familiar with such concepts as asymptomatic carriers, incubation periods, and immunity.
From these informal discussions sprang up broader public discussions that included public health officials. These activities are well described in detail in Randy Shilts' book And The Band Played On(10). Early efforts included community meetings, often heated, with key decision-makers in the gay community. Articles and letters to the editor appeared in several newspapers. While the writers were debating as to what the policy should be, these discussions were educating the readership about AIDS. Before government funding, community activists placed posters in windows of local stores and directly mailed safe-sex brochures to single men living in San Francisco. In 1982, the first community agency dedicated to serving people with AIDS, The Kaposi's Sarcoma Education and Research Foundation, was created. The earliest prevention education messages focused on limiting the number of sexual partners and learning about the partner's health before engaging in sex. Over the years, the messages have changed and matured to focus on avoiding the exchange of body fluids, especially via unprotected anal intercourse.
These indigenous efforts received support from the government. In 1982, San Francisco approved $1 million towards AIDS education; this was the first locally funded AIDS education effort in the United States. In 1983, an additional $2.1 million was provided to AIDS programs. San Francisco's budget for AIDS funding exceeded that of the National Institute of Health's extramural AIDS research budget for the country(10).
CHALLENGES TO THE SAN FRANCISCO HIV PREVENTION MODEL
Although the HIV prevention effort in San Francisco has been effective in the gay community as a whole, it has not been as successful among young MSM or among MSM of color, nor has it prevented relapse among some older MSM.
Figure 5 shows the AIDS incidence among MSM by birth cohort. The curves for men born before 1940 and for those born between 1940 and 1944, 1945 and 1949, 1950 and 1954, and 1955 and 1959 show similar shapes, with a peak of cases in 1992 and then steady decreases. However, for men born between 1960 and 1964, and 1965 and 1969, the curves show a similar rise but comparatively smaller decreases after 1992, indicating that HIV infections did not level off as sharply among younger men in the mid-1980s. The number of cases among MSM who were born between 1970 and 1974 is small: However, the oldest of these men would have been only 14 years old in 1984. Therefore, this cohort of men became sexually active during the time that the safer-sex message was being well publicized in the gay male community. Similar trends by birth cohort have been found in New York City(11,12) and Los Angeles (13).
More detailed information concerning young MSM is available from a seroprevalence study conducted in 1992 by the San Francisco Department of Health(14). This survey, which used targeted sampling at public venues frequented by MSM aged 17-22 in San Francisco and Berkeley, revealed a seroprevalence rate of 12.1% for San Francisco men. A third of the young men reported that they had unprotected anal intercourse in the prior 6 months. Significant predictors of unprotected anal intercourse included negative peer norms regarding safe sex and the use of alcohol and nitrites during sex. The findings of the seroprevalence study have been substantiated by a multistage probability sample of single men from the same census tracks as the original San Francisco Young Men's Health Study(15). This study found that 17.9% of men aged 18-29 years were HIV-infected. Sixty-three percent of men reported one or more receptive anal intercourse partners in the prior year, and 41% of these young men did not use condoms consistently. In this study, the estimated HIV seroincidence, based on follow-up of those who were initially seronegative, was 2.6% per year.
Focus group discussions with young MSM reveal that these young men are knowledgeable about HIV and know how it is transmitted. Reasons they cited for having unprotected sex include the perception that they could not become infected because they were young (AIDS was an "older man's disease"), being "in love," substance use, and power dynamics in their relationship (e.g., inability to ask their partners to put on a condom). Some young gay men feel that it is inevitable that they will become infected. For others, the strong sexual urges of adolescence, coupled with the general high risk-taking that occurs during this time, result in unsafe sexual practices.
A disturbing aspect of both studies of young men was that seroprevalence rates were substantially higher among MSM of color. In the Department of Public Health study(14), the HIV seroprevalence among African American MSM was 21.2%, significantly higher than that of other racial/ethnic groups. In the San Francisco Young Men's Health study(15), the seroprevalence rates were 15.5% among whites, 25.0% among Latinos, 26.9% among Asian/Pacific Islanders, and 35.0% among African Americans. This finding supports other reports that MSM of color were not changing their risk behaviors as rapidly as other gay men. Since 1981, MSM of color represent an increasing proportion of AIDS cases among MSM. In 1981, 7.1% of AIDS cases among MSM were men of color; by 1985, 15.1% of AIDS cases among MSM were men of color, by 1990, 19.4% and by 1995, 24.3% of cases were men of color. In a population-based study of AIDS in multiethnic neighborhoods of San Francisco (the AMEN study), the seroprevalence was also higher among African American MSM (63%) than among white MSM (45%), although the sample size was relatively small and this difference did not reach statistical significance(7).
Even among older white MSM, several reports document the frequency of relapse from safer sex(7,16). These studies show that relapse from safer sex is a more common pattern than ongoing high-risk behavior. This is an important observation because interventions to maintain behaviors differ from those aimed at initiating behavioral change. Among men who relapse to unsafe sex, the reasons cited include being "turned on," having been requested by partner, concomitant use of alcohol and drugs, and being "in love." The problem of relapsing to unsafe sexual practices is not surprising, given experience with efforts to change other harmful behaviors, such as substance abuse and cigarette smoking. It is more difficult to maintain behaviors over a series of years than to change them initially. This is especially true of behaviors that are strongly motivating, such as sexual relations, and that are influenced by lust, love, and substance use. Relapse is particularly dangerous among MSM in San Francisco because the seroprevalence of HIV is so high. Therefore, when MSM relapse into unprotected sex, the risk for seroconversion is very high.
A case-control study undertaken in San Francisco examined factors associated with HIV seroconversion in MSM after 1984(8). The cases comprised 83 MSM who were enrolled in one of the three San Francisco cohort studies of MSM and who seroconverted between 1984 and 1989. The controls were seronegative men from these cohorts. As with the earlier results from these cohorts, the major risk factors for seroconversion were the total number of intercourse partners and receptive anal intercourse. Although HIV prevention efforts had made major inroads by 1984, a smaller number of men still seroconverted for the same reasons as before.
Most of the data on sexual behaviors, risk reduction, and relapse among MSM come from these three San Francisco cohorts assembled in the early 1980s(17,18). With the age of the cohort members increasing(by 1996 the youngest men were 30) and the attrition that has occurred, many wonder if the information gathered from these cohorts is applicable to the remainder of the MSM population. In addition, these early cohorts had few non-white participants.
To address this concern, new cohorts of MSM in San Francisco, Denver, and Chicago were assembled for a study beginning in February of 1993. The purpose was to study the feasibility of conducting preventive HIV vaccine trials among high-risk MSM. The sample for San Francisco is ethnically diverse, with 32% men of color(19). The HIV seroincidence was 2.7/100 person years (95% confidence interval 1.8-4.2%). The high rate of seroincidence is not surprising because these men were recruited on the basis of reporting recent anal sex (protected or unprotected) or of having been seen at an STD clinic. The independent predictors of seroconversion in the combined cohorts in San Francisco, Denver, and Chicago were similar to the predictors of seroconversion identified in studies conducted in the early 1980s: unprotected receptive anal intercourse, injection drug use, having a known HIV-seropositive partner, and gonococcal/nongonococcal urethritis. Condom failure was also an independent risk factor for HIV seroconversion.
The same investigators are studying predictors of condom breakage (brand of condom, technique, being inebriated while having sex) and methods of educating men on how to use condoms properly to decrease the risk of breakage. The data, presented at the National Conference for Vaccine Development Groups(20), reveal that frequent amphetamine use is an independent predictor of condom failure during insertive anal intercourse, whereas alcohol bingeing and frequent "popper" use predict condom failure during receptive anal intercourse.
LESSONS LEARNED FROM SAN FRANCISCO'S PREVENTION EFFORTS
San Francisco's prevention efforts reveal much about the successes and challenges in HIV prevention. Among them are the following. (a) Prevention works: Small studies show that prevention interventions such as small group discussions and counseling can decrease the rate of HIV infection(21). However, evidence documenting the effectiveness of HIV prevention among broad populations is limited. San Francisco's experience should be viewed as proof of the efficacy of population-based prevention. The success of this HIV prevention effort is objectively confirmed by the decreasing incidence of AIDS and the reduction in the rate of gonococcal proctitis, and in the reported rates of anal intercourse.
(b) Community mobilization is a cost-effective method of dramatically decreasing infection: San Francisco's gay community has been a driving force for HIV prevention. Although the effort is supported financially and politically by the San Francisco government, especially the Public Health Department, it is led by members of the gay community. The major question raised by this lesson is, can the government spark this type of movement? San Francisco has sought to do this by collaborating and funding community-based organizations to carry out community mobilization efforts that seek to change group norms. These efforts typically involve training members of the community to be peer advocates of change within their community. The San Francisco experience is that this method of prevention is harder to perform effectively in communities that are not tightly organized(such as substance users). It is not clear how effective this strategy has been among young MSM. After the release of the results of the Department of Public Health's Young Men's Study in 1992, several campaigns were launched that were aimed at young men. These campaigns appear to have had limited success. Preliminary results from a Department of Public Health Study launched in 1994, using methods similar to those in the 1992 study, showed a slightly lower HIV seroprevalence rate of 8.1% for San Francisco men. Despite the lower seroprevalence, the proportion of young men surveyed who had engaged in unprotected anal intercourse in the prior 6 months was 31%, almost identical to the rate (33%) found in the earlier study. Therefore, although the epidemic among young men does not appear to have the explosive nature of the earlier epidemic among MSM, the success of community mobilization in this group is unproved.
(c) Prevention efforts must be ethnically, culturally, linguistically, and age appropriate: The early prevention efforts worked, in part, because gay men were talking to other gay men. However, until the late 1980s, many persons of color perceived HIV as a disease that affected white gay men. Perhaps infections among MSM of color could have been prevented but were not. Similarly, until 1992 very few compaigns targeted young MSM because youth were not recognized as a distinct group that required special prevention efforts.
(d) Prevention efforts must go beyond knowledge to include skills building, treatment of substance abuse, and assessment and management of mental health issues: Early prevention efforts appropriately focused on increasing knowledge of how HIV is transmitted. For many at-risk persons, this was enough. If they knew that they had to use condoms every time they had anal intercourse, they always would. For others, making or maintaining such changes over the course of their life may not be possible. Therefore, knowledge is a necessary but not a sufficient condition for HIV prevention. For this reason, San Francisco's Department of Public Health funds interventions to increase skills to talk to one's partners about safe sex, to take control of substance use, and to build self-esteem.
(e) Data must be continually collected and used to change prevention strategies: Prevention needs are not static. Studies monitoring the epidemic are essential to document trends in the epidemic. For example, were it not for the findings of the San Francisco Young Men's Health Study(15), the high rates of new infection among this group would not be known. Without this information, the Department of Public Health would not have made a major investment in prevention programs for young MSM.
(f) The government, working collaboratively with the community, can forge a powerful alliance for HIV prevention: Much of what has been accomplished could never have occurred without the strong gay activism in San Francisco and the support and leadership of the gay community. The partnership has held strong despite what could have been divisive issues, such as the closing of bath houses in the early 1980s. Although some in the gay community felt that the bath houses should remain open on civil libertarian grounds, the gay community in general supported the closures. Similarly, as sex clubs (establishments without baths or saunas that provide an environment for people to meet and have sex) have emerged in the 1990s, the strongest opposition has been from members of the gay community, who have insisted that the Public Health Department allow to operate only those clubs that ensure that unprotected sex does not occur on their premises. Furthermore, the gay community has been actively involved in establishing guidelines for the operating and monitoring of these clubs.
Based in part on the success of working with the community in San Francisco, the Centers for Disease Control and Prevention requires all localities to create an HIV prevention planning group, consisting of members of the affected communities, their advocates, behavioral scientists, and public health experts. These planning groups have been set up across the country and are reshaping prevention in a collaborative fashion with the local government.
San Francisco's AIDS epidemic has devastated a small community, with HIV seroprevalence rates that are higher than anywhere else in the United States. However, this devastating experience helped to spur prevention efforts that have taught us much about how partnerships can be formed between communities and government to develop methods of promoting behavior change and preventing the spread of HIV.
Acknowledgment: I am grateful to Jeff Newman, M.D., for noting the mortality trends, Ling Hsu, M.P.H., for data analysis, and Valerie Kegebein, M.P.H., Giuliano Nieri, Paul O'Malley, and Sandra Schwarcz, M.D., for their insightful comments.
1. San Francisco Department of Public Health. Mortality among San Francisco residents; 1991-92. Epidemiol Bull 1994;10:9-11.
2. Communications Technologies. HIV-related knowledge, attitudes, and behaviors among San Francisco gay and bisexual men: results from the fifth population-based survey. San Francisco: Communication Technologies, 1990.
3. Hessol NA, Lifson AR, O'Malley PM, Doll LS, Jaffe HW, Rutherford GW. Prevalence, incidence, and progression of human immunodeficiency virus infection in homosexual and bisexual men in hepatitis B vaccine trials, 1978-1988. Am J Epidemiol 1989;130:1167-75.
4. Rutherford GW, Lifson AR, Hessol NA, et al. Course of HIV-1 infection in a cohort of homosexual and bisexual men: an 11 year follow up study. BMJ 1990;301:1183-8.
5. Winkelstein W, Lyman DM, Padian N, et al. Sexual practices and risk of infection by the human immunodeficiency virus: the San Francisco Men's Health Study. JAMA 1987;257:321-5.
6. Winkelstein W, Samuel M, Padian N, et al. The San Francisco Men's Health Study: III. Reduction in human immunodeficiency virus transmission among homosexual/bisexual men 1982-86. Am J Public Health 1987;77:685-9.
7. Fullilove MT, Wiley J, Fullilove RE, et al. Risk for AIDS in multiethnic neighborhoods in San Francisco, California: the population-based AMEN study. West J Med 1992;157:32-40.
8. Moss AR, Bacchetti P, Osmond D, et al. Seropositivity for HIV and the development of AIDS or AIDS related condition: three year follow-up of the San Francisco General Hospital Cohort. BMJ 1988;296:745-50.
9. McKusick L, Wiley JA, Coates TJ, et al. Reported changes in the sexual behavior of men at risk for AIDS. San Francisco, 1982-84. The AIDS Behavior Research Project. Public Health Rep 1985;100:622-8.
10. Shilts R. And the band played on. New York: Penguin, 1987.
11. Fordyce EJ, Williams RD, Surick IW, Shum RT, Quintyne RA, Thomas PA. Trends in the AIDS epidemic among men who reported sex with men in New York City: 1981-1993. AIDS Educ Prev 1995;7:3-12.
12. Stoneburner R, Lessner L, Fordyce EJ, Bevier P, Chaisson MA. Insight into the infection dynamics of the AIDS epidemic: a birth cohort analysis of the New York City AIDS mortality. Am J Epidemiol 1993;138:1093-104.
13. Greenland S, Lieb L, Simon P, Ford W, Kerndt P. Evidence for recent growth of the HIV epidemic among African-American men and young male cohorts in Los Angeles County. J Acquir Immune Defic Syndr 1996;11:401-9.
14. Lemp GF, Hirozawa AM, Givertz D, et al. Seroprevalence of HIV and risk behaviors among young gay and bisexual men: the San Francisco/Berkeley Young Men's Survey. JAMA 1994;272:449-54.
15. Osmond DH, Page K, Wiley J, et al. HIV infection in homosexual and bisexual men 18 to 29 years of age: the San Francisco Young Men's Health Study. Am J Public Health 1994;84:1933-7.
16. Stall R, Ekstrand M, Pollack L, McKusick L, Coates TJ. Relapse from safer sex: the next challenge for AIDS prevention efforts. J Acquir Immune Defic Syndr 1990;3:1181-7.
17. Ekstrand ML, Coates TJ. Maintenance of safer sexual behaviors and predictors of risky sex: the San Francisco Men's Health Study. Am J Public Health 1990;80:973-7.
18. Samuel MC, Hessol N, Shiboski S, Engel RR, Speed TP, Winkelstein W. Factors associated with human immunodeficiency virus seroconversion in homosexual men in three San Francisco cohort studies, 1984-1989. J Acquir Immune Defic Syndr 1993;6:303-12.
19. Buchbinder SP, Douglas JM, McKirnan DJ, Judson FN, Katz MH, MacQueen KM. The feasibility of conducting preventive HIV vaccine trials in homosexual men in the United States. J Infect Dis (in press).
20. Stone E, Vittinghoff E, O'Malley P, Buchbinder SP. Should non-injection gay/bisexual men be targeted for HIV prevention trials? [Abstract]. Presented at the National Conference for Vaccine Development Groups, 1996.
21. Choi K-H, Coates TJ. Prevention of HIV infection. AIDS 1994;8:1371-89.
Publication of this supplement has been supported by the Japan Foundation, Center for Global Partnership
© Lippincott-Raven Publishers.