July 14, 2003
Studies from around the world have reported recent increases in sexual risk behavior among men who have sex with men (MSM), some with accompanying rises in STDs and a few observing resurgences in HIV incidence. The propagation of HIV and other STDs requires unprotected sex with, on average, two or more partners of opposite infection status. Few recent studies have measured levels and trends in unprotected sex between MSM who knowingly or unknowingly have sex with persons of different HIV serostatus. In the current study, researchers characterized potentially HIV-serodiscordant sex in a large community-recruited sample (n=10,579) of MSM from 1999-2001.
Volunteers and staff members from the STOP AIDS Project, a San Francisco community-based organization providing HIV prevention services, conducted brief interviews with MSM at diverse gay-oriented venues or events (e.g., dance clubs, bars, gay pride parades, street fairs) and on sidewalks in neighborhoods with a high volume of MSM pedestrian traffic. The number of respondents declined from 1999 (n=4,940) to 2000 (n=2,674) because of fewer volunteers and decreasing participation among persons approached. Seventeen percent of subjects (n=1,811) reported they had completed a STOP AIDS Project questionnaire at some time in the past (any time since 1994) and therefore may have been included more than once in the data.
Subjects were classified as having potentially serodiscordant unprotected anal intercourse (UAI) if they reported UAI with at least two persons whose HIV status was not known to them in the preceding six months. Unprotected was defined as "not always" using a condom with anal intercourse. Anal intercourse was insertive, receptive, or both.
Most MSM (90.5 percent) had previously tested for HIV, 64.2 percent were white, and 59.5 percent were over age 30. Overall, 1,343 (12.7 percent) reported UAI in the preceding six months with at least two partners whose HIV serostatus was not known. Potentially serodiscordant UAI was higher among HIV-positive respondents (20.8 percent) compared with HIV-negative respondents (12.1 percent). For MSM who did not know or report their HIV serostatus, any UAI with two or more anal sex partners (reported by 13.4 percent) was considered potentially serodiscordant.
In multiple logistic regression analysis, the associations between potentially serodiscordant UAI with age and race/ethnicity differed according to respondents' HIV serostatus. HIV-positive MSM who engaged in potentially serodiscordant UAI were more likely to be over 30 years old and white. In contrast, younger age was associated with potentially serodiscordant UAI among HIV-negative MSM along with white race/ethnicity. Among MSM who did not know or report their HIV serostatus, nonwhite ethnicity was associated with UAI.
Potentially serodiscordant UAI increased overall from 11.7 percent in 1999 to 16.4 percent in 2001. In multiple logistic regression analysis, the increased temporal trends were significant after controlling for age and race/ethnicity among HIV-positive respondents and HIV-negative respondents but not among MSM with unknown serostatus. Upward trends were parallel for all age and ethnic groups.
"Recent increases in UAI previously reported among MSM in San Francisco are not only the result of increases in UAI between MSM who are known to be of the same HIV status," the authors reported. "Our data point to the urgent need for prevention messages promoting condom use and disclosure of HIV serostatus to sex partners. ... Our data suggest targeting particularly older, white, HIV-positive MSM. At the same time, our data also highlight the need for HIV counseling and testing among younger MSM and MSM of color to increase awareness of their own HIV serostatus and to encourage them to avoid potential exposure to HIV," researchers concluded.