But the study also warns that even small increases in unsafe sex practices could allow epidemic HIV transmission to persist and be passed to new generations of gay men.
The study, published in the Aug. 1 issue of the American Journal of Epidemiology, was done by Martina Morris and Laura Dean, epidemiologists at Columbia's School of Public Health.
Homosexual men in New York City now report an average of one new unsafe partner per year, down from more than 11 a decade ago, according to the Longitudinal AIDS Impact Project (LAIP), a seven-year behavior study of New York City gay men under the direction of Dean.
To determine the long-term effect of this dramatic drop on HIV transmission in AIDS epicenters like New York and San Francisco, the Columbia researchers constructed a mathematical model based on data from the LAIP.
The model tracks the gay male population infected with HIV from 1980 to present and projects how widespread HIV will become in this group through 2030.
"We found the current sexual behavior of gay men to be right on the epidemic boundary so that HIV potentially could fade out under the right conditions or proliferate if there is even a small increase in unsafe sex habits," says Morris.
"The good news is that if safer sex is maintained and infection from other sources like injection drug use is prevented, our results indicate that the HIV epidemic could fade out among gay men," says Morris. "But this will not happen quickly. Even under the most optimistic scenario, it would take more than 50 years."
The Columbia study suggests the number of people with HIV in AIDS epicenters is gradually descending from its peak in 1989 when an estimated 40 percent of homosexual men were infected.
Under the optimistic scenario where the average number of new, unsafe sexual partners per year does not increase above one, the virus could affect less than five percent of the gay population by the year 2030.
But with an average of two new unsafe partners per year, instead of the reported one, the virus would continue to be transmitted at epidemic levels and infect about 25 percent of younger men and about 60 percent of men who reach their late 40s.
"The bad news is that it only takes an average of one more unsafe partner to keep the epidemic going strong," says Morris.
"The future of the epidemic appears to be highly sensitive to changes in current behavior patterns," says Morris. "A single individual who has one more unsafe partner puts himself at slight increased risk, but if all individuals make this choice, there is a synergistic effect and everyone's risk would be dramatically increased."
Evidence presented by Dean at last year's American Public Health Association Meeting suggests that younger gay men are already engaging in more risky sexual behavior.
Because younger gay men enter the sexually active population without infection and are more likely to select sexual partners from those of similar ages, they are slightly less at risk.
But according to the Morris/Dean model, if younger men have an average of three new, unsafe partners per year, their behavior would be sufficient to cause the disease to once again proliferate in the entire homosexual population.
"Young men are the future of this epidemic," says Morris. "It is critical to ensure that the relatively higher rates of unsafe sex already reported by younger homosexual men do not continue to rise."
According to the authors, their findings have several implications. First, it would be unwise and premature to conclude that a consistent pattern of decline in the rate of new AIDS cases means the margin of safety has increased.
Second, efforts to establish and maintain safe sex behavior are crucial to the future of the disease and third, research which tracks the sexual behaviors of gay men is still needed.
"The best way to make sure gay men maintain safer sex is to show them that their behavior really matters," says Morris. "Without the compelling evidence provided by this and other projection studies, safer sex behavior among gay men may be difficult to maintain."
This work was supported by the National Institutes of Health, the National Science Foundation, the Aaron Diamond Foundation, the New York Community Trust, the British Science and Engineering Research Council, and the Isaac Newton Institute for Mathematical Sciences and Columbia.
Morris is an assistant professor of sociology at Columbia. She received her M.A. degree in statistics (1987) and Ph.D. in sociology (1989) from the University of Chicago.
Morris' primary research interest is in modeling the role of networks in the spread of infectious disease.
She has pioneered methods for integrating egocentric network data into dynamic models for epidemiological transmission, and is currently involved in survey projects on the spread of HIV/AIDS in New York City, Uganda and Thailand.
Dean is associate research scientist at Columbia's School of Public Health. She received her M.Ed. from Columbia (1979).
She is currently the principal investigator of a landmark longitudinal study of the mental health and behavioral impact of the AIDS epidemic in New York's gay community.
She has published on the psychological effects of knowing one's HIV status, AIDS-related bereavement and antigay violence and discrimination.