Men who have sex with men (MSM) have been the most severely affected risk group since the beginning of the AIDS epidemic, accounting for more than half of all AIDS cases reported to the Centers for Disease Control and Prevention (CDC) through December, 1995 . Largely because of AIDS in MSM, HIV infection has been the leading cause of death in men aged 25–44 years since 1991 , and was responsible for 23% of all deaths of men in this age group in 1994 . However, from 1984 to 1990, the annual increases in AIDS incidence for MSM became progressively smaller, and from 1987 to 1990 AIDS incidence for MSM seemed to plateau in some large metropolitan statistical areas . In this report, we describe the trends in annual AIDS incidence among MSM from 1990 to 1995. Because national trends are a composite of trends in population subgroups, we also describe trends among geographic, racial/ethnic, and age subgroups of MSM.
Material and methods
AIDS surveillance data
Health care providers are required to report confidentially AIDS cases to the state or local health department. In all states, case reports of AIDS, without personally identifying information, are forwarded by state health departments to the CDC. Case reports contain information on age, sex, race/ethnicity, date of diagnosis, residence at the time of diagnosis, AIDS-defining condition [i.e., an AIDS-defining opportunistic illness (‘clinical AIDS’), or severe immunosuppression (CD4 T-lymphocyte count of fewer than 200 × 106 cells/l, or CD4 percentage of less than 14% of total lymphocytes, or ‘immunologic AIDS’), and risks for HIV transmission, which allow classification of cases into probable exposure categories.
We analysed data for all MSM aged at least 13 years, who received diagnoses of AIDS-defining conditions  between January 1990 and December 1995, and whose cases were reported to the CDC up to September 1996. For AIDS surveillance purposes, the MSM risk category includes men who report having sex with another man after 1977, but before their first positive HIV-antibody test. For this analysis, MSM who also reported injecting drug use were excluded.
Method of estimating clinical AIDS diagnosis dates
The 1993 expansion of the AIDS surveillance case definition [5,6], allowed for the diagnosis of AIDS based not only on clinical criteria (AIDS-defining opportunistic illnesses), but also on new immunologic criteria (immunologic AIDS). Once patients with immunologic AIDS are reported to CDC, the later occurrence of clinical AIDS is often not reported. To compare AIDS incidence before and after 1993, we estimated the time of occurrence of clinical AIDS in patients reported with immunologic AIDS on the basis of the distribution of time from immunologic AIDS to clinical AIDS .
These distributions were estimated using data from the Adult and Adolescent Spectrum of HIV Disease Surveillance (ASD) project, an ongoing survey of medical records of HIV-infected patients receiving medical care from more than 100 clinics, hospitals, and private medical practices in 10 cities in the United States . The estimates were based on 9261 patients receiving care from January 1990 to August 1996 and having at least one CD4 determination before the diagnosis of clinical AIDS. Separate distributions were applied, based on the level of CD4 count at the diagnosis of immunologic AIDS (CD4 strata 0–9, 10–29, 30–69, 70–129, 130–149, 150–189, 190–279, and ≥ 280 cells × 106/l). The median time from immunologic AIDS to clinical AIDS diagnosis for all persons in ASD was 20 months .
AIDS incidence rates were also adjusted for the anticipated redistribution of those cases initially reported with no identified risk , and for reporting delays . The result of these adjustments was the adjusted estimated incidence of clinical AIDS. Estimates were rounded as described in the HIV/AIDS Surveillance Report of 1995 . In this report, this estimated incidence measure will be referred to as AIDS incidence.
Rate calculation and presentation of data
AIDS incidence rates are reported as estimated cases of clinical AIDS per 100 000 men aged at least 13 years per year (AIDS rate). Denominators were derived from 1990 census data for 1990 rates, and 1991–1995 intercensal estimates for 1991–1995 rates. For reasons of confidentiality and stability of estimates, data are not presented for any analysis group in which the rate was based on fewer than 40 AIDS cases. For regional analyses, only the 50 states plus the District of Columbia were included; national analyses also included Puerto Rico and other United States territories.
Annual AIDS rates in MSM increased from 25.5 cases per 100000 men (25 100 cases in 1990) to 28.5 cases per 100000 men (29 600 cases in 1995), a 12% increase (Fig. 1). The increase differed substantially by race/ ethnicity (Fig. 1). Black men, for whom the rates were highest throughout the period, had a proportional rate increase of 45% over the 5-year period (7700 cases in 1995). In 1995, incidence rates rose 53% among American-Indian/Alaskan native men (100 cases in 1995), 23% for Hispanic men (4400 cases in 1995), and 19% for Asian/Pacific Islander men (320 cases in 1995). AIDS rates for white men decreased by 2% (17 000 cases in 1995).
AIDS rates over the 5-year period increased for men aged 30 years and older, but decreased for men aged 13–29 years (Table 1). The largest increases in rates over the 5-year period was for men older than 59 years (32% increase), although the rates in this age group were low.
Changes in AIDS rates for MSM also differed by geographic region in the USA (Table 1): rates in the west decreased slightly (−2%), but those in all other regions increased. Although the differences in national AIDS rates by race/ethnicity were fairly consistent within regions (data not shown), the highest race/region-specific increases in AIDS rates occurred among black MSM in the south (59%, 3700 cases in 1995) and the mid-west (59%, 1200 cases in 1995). The only region/race-specific decrease was among white MSM in the west (−8%, 5700 cases in 1995); in other regions of the country, rates among white MSM were stable between 1990 and 1995.
Larger increases in AIDS rates occurred in metropolitan statistical areas of 50 000 to 249 999 people and rural areas (non-metropolitan statistical areas; Table 1), although these smaller areas had lower AIDS rates than the large metropolitan statistical areas in 1990. The largest metropolitan statistical area size/race-specific increases in rates occurred among black (132%) and Hispanic (100%) MSM in rural areas, black (83%) MSM in metropolitan statistical areas with 50 000 to 249 999 residents, and black (100%) and American-Indian/Alaskan native (100%) MSM in metropolitan statistical areas of 250 000 to 999 999 people. The only metropolitan statistical areas size/race-specific decrease in rates occurred among white MSM in metropolitan statistical areas of more than 2.5 million (9% decrease).
The AIDS rates and changes in rates have differed in the 10 metropolitan statistical areas that have reported the largest cumulative number of AIDS cases in MSM (Table 2). Rates for black MSM increased in all nine metropolitan statistical areas evaluated, and rates for Hispanic MSM increased in six of eight metropolitan statistical areas with sufficient AIDS diagnoses for analysis; in contrast, rates for white MSM decreased in seven of 10 metropolitan statistical areas. The largest decrease in the rate in these cities occurred in Houston, Texas (−21%); the largest rate increase occurred in Miami, Florida (34%), and was associated with a large increase in the incidence rate for black MSM (110%).
The rate of AIDS for MSM remained high in the early 1990s, and MSM have continued to represent the largest number of new AIDS cases reported to CDC in the past 5 years. Although the increases in AIDS incidence rates for MSM are not as dramatic as earlier in the epidemic, large numbers of AIDS cases in MSM continue to be reported each year; a total of 190 227 new cases were reported from 1990 to 1995.
The reasons that annual increases in AIDS incidence have slowed may be a balance of the reductions in high-risk behaviors by some MSM and continuing high levels of risk behaviors by others. HIV prevention efforts may have reached MSM communities in differing degrees. In the late 1980s, a slowing in the growth of AIDS was reported in some groups of MSM  and may have been due, in part, to a decrease in high-risk sexual behaviors and in the HIV incidence among MSM in the mid- to late 1980s . The reduction of high-risk behaviors in the 1980s was associated with decreases in the rates of gonorrhea, amebiasis, and syphilis among MSM during that period [12–14]; these outcomes may serve as surrogate measures of changes in high-risk sexual behavior by MSM. However, new infections continue to occur, and some MSM either fail to adopt safer sex practices or relapse to unsafe behavior [15,16]. Studies of young MSM indicate that the prevalence of high-risk sexual behavior [17–20] and the HIV seroprevalence [17–19] remain high; young age, having an HIV-seropositive sex partner, injecting drug use, condom failure, and unprotected receptive anal sex may be important predictors of seroconversion in MSM .
Because improved antiretroviral therapies may delay the occurrence of opportunistic illnesses, it is possible that trends in clinical AIDS may be influenced by the changing availability of therapies over time . The data used for this analysis included AIDS diagnoses up to December 1995 and few of the most promising therapies were in use during that time. Therefore, the results of this analysis should only be minimally influenced by the prolonged intervals between HIV infection and AIDS diagnosis that are associated with improved therapies.
Regional differences in AIDS incidence have been recognized [4,23]. In this analysis, region-specific increases in rates occurred in the north-east, the mid-west, and the south, but rates decreased in the west. Differences in regional trends could reflect temporal differences; e.g., the epidemic was introduced later among MSM in some regions, and thus a moderation of incidence related to infection of a large proportion of men at risk may have occurred later in these regions. There also may be differences in the prevalence of high-risk sexual behaviors among MSM in different areas of the country or in metropolitan statistical areas of different sizes. For example, MSM in small southern cities have been reported to be several times more likely than men in San Francisco, California to engage in unprotected anal intercourse. .
The rates of clinical AIDS and increases in incidence rates were higher for black and Hispanic MSM than for white MSM, regardless of size of metropolitan statistical area or geographic region. Several studies with large proportions of non-white participants have suggested that the prevalence of unsafe sexual behavior may be higher for non-white MSM than for white MSM [25–27]. For racial/ethnic minorities, inaccurate appraisal of HIV risk, decreased access to HIV prevention services, and high rates of sexually transmitted diseases may also contribute to high AIDS incidence rates [28,29]. These AIDS surveillance data demonstrate that racial/ethnic disparity in AIDS rates has grown even larger in the past 5 years, which suggests that the current prevention strategies have not been as effective for minority populations as for white men. These data speak vividly for the need to strengthen prevention efforts among black and Hispanic MSM. These strategies must address the prevention and support system needs of these specific communities, e.g., cultural populations may differ on openness to discussions of homosexuality, sexual behavior, or HIV infection [30, 31].
Among MSM of various ages, AIDS rates were smallest for the youngest MSM. Because the interval between HIV infection and AIDS diagnosis is variable and may be long, men who are infected at a young age may not receive an AIDS diagnosis until they are in an older age category. Thus, AIDS incidence data may not fully reflect the levels of HIV infection in young MSM, and small rates of AIDS diagnoses should be interpreted with caution in these younger age groups.
Despite the fact that AIDS rates decreased modestly for MSM aged 13–29 years, seroprevalence and behavioral studies indicate that HIV is still an important health problem for young men. HIV seroprevalence data demonstrate that many young MSM are infected with HIV; recent studies indicate a high prevalence of HIV infection in young MSM in New York City, New York (9%) , San Francisco, California (17.9%) , and Pittsburgh, Pennsylvania (7.1%) . These rates are 12–30-fold higher than HIV seroprevalence rates among all men in the United States . The annual rate of new HIV infections in these groups of young MSM was also high; 2% in New York City and 2.6% in San Francisco. In addition, in all three cohorts, high percentages of young MSM reported risky sexual behaviors [17–19].
The use of rates (number of cases per 100 000 men) to evaluate trends in AIDS incidence allows comparison of the impact of the epidemic on persons of different races and ethnicities, from different areas of the country, and of different ages. However, rates calculated in this way underestimate the impact of the epidemic on MSM, because the true number of MSM at risk is smaller than census counts of all men older than 13 years. Variations in rates of AIDS attributed to male–male sexual contact may reflect variations in the prevalence of male homosexual behavior as well as differences in the prevalence of HIV infection in different communities. For example, according to a recent study , the prevalence of men who self-identified as homosexual was ninefold higher for men in the 12 largest central cities than for men living in rural areas. Therefore, rates based on census counts of all men may underestimate the rates for MSM in smaller metropolitan statistical areas compared to larger ones. There is no evidence that the prevalence of male–male sexual activity differs for racial/ethnic groups ; however, of men with male–male sexual risk, minority men are more likely to also have reported heterosexual or injecting drug use risks . For this reason, rates of AIDS among racial/minority MSM may be underestimated in this analysis to a greater extent than for white MSM, because men with multiple risk behaviors (for example, male–male sex and injecting drug use) would be not be counted as cases in this analysis.
The 12% increase in clinical AIDS incidence rates for MSM during the 5-year period reflects the net effect of a declining rate for white MSM, and increasing rates for black and Hispanic men, older men, and men in rural areas and small metropolitan statistical areas (less than one million residents). The AIDS epidemic among MSM is not homogenous but is a composite of multiple epidemics with different times of onset and patterns of spread. Each community needs to plan HIV prevention activities in response to the characteristics of its local epidemic and to the characteristics of the populations at risk .
1. Centers for Disease Control and Prevention: HIV/AIDS Surveillance Report;
2. Kochanek KD, Hudson BL: Advance report of final mortality statistics, 1992
. Monthly Vital Stat Rep
, 6 (suppl):24.
3. Singh GK, Kochanek KD, MacDorman MF: Advance report of final mortality statistics, 1994
. Monthly Vital Stat Rep
, 3 (suppl):24.
4. Karon JM, Berkelman RL: The geographic and ethnic diversity of AIDS incidence in homosexual/bisexual men in the United States
. J Acquir Immune Defic Syndr
5. Centers for Disease Control and Prevention: 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults
(no RR-17): 1–19.
6. Centers for Disease Control and Prevention: Update: Trends in AIDS diagnosis and reporting under the expanded surveillance definition for adolescents and adults – United States, 1993
7. Karon JM, Green TA, Hanson DL, Ward JW: Estimating the number of AIDS-defining opportunistic illness diagnoses from data collected under the 1993 AIDS surveillance case definition
. J Acquir Immune Defic Syndr
1997 (in press).
8. Farizo KM, Buehler JW, Chamberland ME, et al.
: Spectrum of disease in persons with human immunodeficiency virus infection in the United States
9. Green TA, Karon JM, Nwanyanwu OC: Changes in AIDS incidence trends in the United States
. J Acquir Immune Defic Syndr
10. Karon JM, Devine OJ, Morgan WM: Predicting AIDS incidence by extrapolating from recent trends
. In Mathematical and Statistical Approaches to AIDS Epidemiology
. Lecture Notes in Biomathematics, Vol 83
. Edited by Castillo-Chavez C. Berlin: Springer-Verlag; 1989:58–88.
11. Winkelstein W, Wiley JA, Padian NS, et al.
: The San Francisco Men's Health Study: continued decline in HIV seroconversion rates among homosexual/bisexual men
. Am J Public Health
12. Centers for Disease Control and Prevention: Declining rates of rectal and pharyngeal gonorrhea in males – New York City
13. Sorvillo FJ, Lieb L, Mascola L, Waterman SH: Declining rates of amebiasis in Los Angeles County: a sentinel for decreasing immunodeficiency syndrome
? Am J Public Health
14. Aral SO, Holmes KK: Sexually transmitted diseases in the AIDS era
. Sci Am
15. Stall R, Ekstrand M, Pollack L, et al.
: Relapse from safer sex: the next challenge for AIDS prevention efforts
. J Acquir Immune Defic Syndr Hum Retrovirol
16. Hospers HJ, Kok G: Determinants of safe and risk-taking sexual behavior among gay men: a review
. AIDS Educ Prev
17. 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
18. Dean L, Meyer I: HIV prevalence and sexual behavior in a cohort of New York City gay men (aged 18–24)
. J Acquir Immune Defic Syndr Hum Retrovirol
19. Silvestre AJ, Kingsley LA, Wehman P, Dappen R, Ho M, Rinaldo CR: Changes in HIV rates and behavior among homosexual men, 1984 to 1988/92
. Am J Public Health
20. Hays RB, Kegeles SM, Coates TJ: High HIV risk-taking among young gay men
21. Buchbinder SP, Douglas JM, McKirnan DJ, Judson FN, Katz MH, MacQueen KM: Feasibility of human immunodeficiency virus vaccine trials in homosexual men in the United States: risk behavior, seroincidence, and willingness to participate
. J Infect Dis
22. Centers for Disease Control and Prevention: Update: trends in AIDS incidence, deaths, and prevalence – United States, 1996
23. Hu DJ, Fleming PL, Mays MA, Ward JW: The expanding regional diversity of the acquired immunodeficiency syndrome epidemic in the United States
. Arch Intern Med
24. Kelley JA, St Lawrence JS, Brasfield TL, Stevenson Y, Diaz YE, Hauth AC: AIDS risk behavior patterns among gay men in small southern cities
. Am J Public Health
25. Linn LS, Spiegel JS, Matthews WC, Leake B, Lien R, Brooks S: Recent sexual behaviors among homosexual men seeking primary care
. Arch Intern Med
26. Doll LS, Byers RH, Bolan G, et al.
: Homosexual men who engage in high-risk sexual behavior: a multicenter comparison
. Sex Transm Dis
27. Lemp GF, Hirozawa AM, Givertz D, et al.
: Seroprevalence of HIV and risk behaviors among young homosexual and bisexual men. The San Francisco/Berkeley Young Men's Survey
28. National Commission on AIDS: The Challenge of HIV/AIDS In Communities of Color
. Washington, DC: National Commission on AIDS; 1992.
29. United States Conference of Mayors: Assessing the HIV-prevention Needs of Gay and Bisexual Men of Color
. Laurel: Health Consultants International; 1993.
30. Peterson JL, Coates TJ, Catania JA, Hilliard B, Middleton L, Hearst N: Help-seeking behavior for AIDS high-risk sexual behavior among gay and bisexual African-American men
. AIDS Educ Prevent
31. Szapocznik J: Research on disclosure of HIV status: cultural evolution finds an ally in science
. Health Psychol
32. Karon JM, Rosenbuerg PS, McQuillan G, Khare M, Gwinn M, Petersen LR: Prevalence of HIV infection in the United States, 1984–1992
33. Laumann EO, Gagnon JH, Michaels S: The Social Organization of Sexuality: Sexual Practices in the United States
. Chicago: University of Chicago Press; 1994.
34. Rogers SM, Turner CF: Male-male sexual contact in the USA: findings from five sample surveys, 1970–1990
. J Sex Research
35. Fleming PL, Ward JW, Dean HD, Ciesielski CA, Drotman DP: AIDS in minority men who report male-male sex
. United States Abstracts of the IX International Conference on AIDS
. Berlin, June 1993 [abstract PO-C12-2879].
36. Valdiserri RO, Aultman TV, Curran JW: Community planning: a national strategy to improve HIV prevention programs
. J Commun Health