The findings imply that doctors should not rely on a man's self-described sexual orientation in assessing his risk of HIV and other sexually transmitted diseases, researchers report in the Annals of Internal Medicine.
Instead, they should ask patients specific questions about their sexual behavior, according to the researchers, led by Dr. Preeti Pathela of the New York City health department.
The findings are based on a 2003 health department survey that included 4,193 men age 18 and up. Respondents were asked about their sexual behavior and their sexual orientation.
Almost 4 percent said they were homosexual, while 91 percent described themselves as "straight." The rest said they were bisexual, "unsure," or declined to answer.
But of men who considered themselves heterosexual, nearly 10 percent had had sex with a man, but no woman, in the past year, Pathela's team found. And of the 337 survey respondents who'd had sex with another man, almost 73 percent identified themselves as straight.
Cultural norms may have played a significant role in the discrepancy, according to the researchers. Foreign-born men, who make up a large proportion of New York City men, were more likely than their U.S.-born counterparts to call themselves heterosexual despite having sex with other men.Men raised in cultures less accepting of homosexuality may be "reluctant" to identify themselves as such, Pathela's team notes, or they may have a narrow definition of what constitutes homosexuality.
In general, self-described heterosexuals, whether they had sex with men or not, had fewer sexual partners than men who said they were homosexual. However, self-described straight men who had sex with other men were less likely than gay men to have had an HIV test recently or to use condoms.
This pattern is "troubling," according to the researchers, and it highlights the need to target STD prevention messages beyond men who call themselves homosexual.
"It is of utmost importance for providers to take a sexual history that ascertains the sex of (the) partner or partners," Pathela's team writes. "Given our data, asking about a patient's sexual identity will not adequately assess his risk."
SOURCE: Annals of Internal Medicine, September 19, 2006.