A new study by the U.S. Centers for Disease Control and Prevention (CDC) shows that new rates of HIV infection between 2006 and 2009 remained stable in the country for all but one group: young, black men who have sex with men (MSM). The CDC study relied on the most sensitive technique yet used to determine that each year about 50,000 people become infected, which is in keeping with earlier estimates. But in black gay and bisexual men between ages 13 and 29, the incidence jumped by 48%, to 6500 cases in 2009. "We're very concerned about this trend," said CDC Director Thomas Frieden at a press teleconference this afternoon.
The new CDC data, published today in PLoS ONE, show that MSM of all ethnicities accounted for 61% of the new U.S. infections in 2009. Next came heterosexual transmission (27%), the main mode of spread in most of the world, and injection drug use (9%). The use of powerful antiretroviral drugs as preventives has so reduced the number of babies becoming infected by their mothers that this route of transmission was included in the "other" category that makes up less than 1% of the total incidence.
Joseph Prejean of CDC's HIV Incidence and Case Surveillance Branch stressed that black men (which it also characterizes as African Americans) do not report higher risk-taking behavior than other MSM. "We don't have all the answers about what may be driving this trend," Prejean said. But he and his colleagues offered a long list of possibilities, including more HIV in the network of people they are having sex with, higher rates of other sexually transmitted infections that ease transmission of the AIDS virus, and deep-rooted stigma in their communities about homosexuality and the disease itself.
Overall, black men and women had 7.7 times the rate of new infections documented in whites in 2009. Latinos had 2.9 times higher incidence than whites that year. "HIV remains one of the most glaring health disparities for African Americans and Latinos," Prejean said.
Since 1994, 25 states have provided CDC with information about newly diagnosed HIV infections. But those data do not distinguish between people who were recently infected and those who had harbored the virus for years. Still, CDC used this prevalence data, combined with cases of AIDS (which had been reported since 1982), to estimate incidence. With the advent of new techniques that use HIV antibody levels to gauge how long a person has been infected, CDC in 2004 began giving funding to selected locales to do incidence surveillance.
Although the current analysis uses actual incidence data, it extrapolates information from 16 states and two cities to produce an estimated number for the entire nation. CDC researchers explain in their PLoS ONE paper how they adjusted their model to obtain the best estimates.