Why is that? The connection with HIV transmission and male homosexual acts have a well documented relationship. While it may not explain all of it, I don't see how you can argue on one hand that there is no relationship but then reference transmission for those in prison as a prevailing factor. So how can you have it both ways? If they're having sex only with men in prison and are not "on the down low" (meaning they won't latter go out and have unprotected hetersexual sex) why is there such a disproportion increase in AA female HIV contraction rates between other ethnicities? But you're right, prison does seem to be a breeding ground for this, even Primetime referenced a study which states up to 60% of prisoners have sex while incarcerated.
While not in the strickest sense is this on the down low I guess, it's still the same behaviour pattern which is causing an increase in infection rates.
While there is not data which clearly connects the facts of HIV and the down low phenomenon, there are enough data to make this a logical conclusion.
1) HIV risk differences between African-American and white men who have sex with men. Journal of the National Medical Association. 91(2):92-100, 1999 Feb
African-American men who have sex with men remain at disproportionately greater risk for contracting human immunodeficiency virus (HIV) infection. While high HIV seroincidence has been documented among homosexual African-American men, behavioral research has rarely studied the HIV risk issues confronting these men. This study assessed a sample of 253 men who have sex with men to determine if African-American (n = 79) and white (n = 174) men report different rates of HIV risk behaviors and differ in characteristics indicative of risk. African-American men who have sex with men were more likely to be HIV-seropositive, to report past treatment for gonorrhea and syphilis, and to have a recent unprotected sex partner known or believed to be HIV-seropositive. Multivariate analyses of covariance, controlling for group differences in age, education, and income, revealed that African-American men who have sex with men were less open about their sexual orientation, scored lower in HIV risk behavior knowledge, had more female sexual partners, and more frequently used cocaine in association with sex relative to white men who have sex with men. Human immunodeficiency virus prevention programs tailored to the needs and risk issues of African-American men who have sex with men are needed.
2)Seroprevalence of human immunodeficiency virus type 1 and sexual behaviors in bisexual African-American and Hispanic men visiting a sexually transmitted disease clinic in New York City. Source American Journal of Epidemiology. 147(3):269-72, 1998 Feb 1.
A cross-sectional human immunodeficiency virus type 1 (HIV-1) serosurvey was conducted between January 1988 and January 1993 at a New York City sexually transmitted disease clinic serving predominantly African-American and Hispanic patients. Overall, 14% (415/3,069) of participating men reported having sex with men; among these, only 52 (13%) were classified as "homosexual" (having had sex with men exclusively since 1978). Most men (87%) who reported having sex with other men also reported having sex with women. These included 147 (35%) "bisexuals" (sex with more than one man and at least one woman since 1978) and 216 (52%) "heterosexuals" (sex with women since 1978 and sexual contact with men before 1978 or only once thereafter). Although HIV-1 seroprevalence was highest among "homosexual" men (70%, 95% confidence interval (CI) 55-81), it was also high among "bisexual" men (35%, 95% CI 27-43) and "heterosexual" men (17%, 95% CI 13-23), and was lowest in men who reported having no male sex partners in their lifetime (9%, 95% CI 7-10). It is possible that transmission of HIV-1 from bisexual men to female sexual partners plays a greater role in heterosexual transmission than was previously recognized.
There was also a more recent article which quantified somewhat the bisexual behavior AA men, I believe the number was 2% of all AA males, which was much higher than any other ethnicity, I'll see if I can find that article.