Blind spots in the epidemiology of HIV in black Americans

Introduction

In the United States there is currently much concern about the large disparity in risk of human immunodeficiency virus (HIV) infection between black Americans and other ethnic groups, ¹ a concern that was explicitly articulated at least two decades ago. ² This reality has generated extensive discussion and soul searching. For example, the US Centers for Disease Control and Prevention (CDC) sponsored a special supplement about the disparity in Sexually Transmitted Diseases (July 2006,33:S1–77) and the National Minority AIDS Council (NMAC) produced a report on this topic in late 2006. ³ The data are sobering: although black Americans comprised only 13% of the US population, they accounted for half of the new HIV/AIDS diagnoses between 2001 and 2005 and ‘in 2005, the annual rates of HIV/AIDS diagnosis among black men and women were seven and 21 times higher than rates among white men and women, respectively.'^1(p.190) Specifically, in the 33 states that report HIV cases to CDC, there were 93,447 HIV/AIDS cases in blacks, including 35,160 women, between 2001 and 2005. ¹

In investigating the uneven distribution of HIV across the ethnic groups in the United States, considerable energy has been focused on contextual factors. The epidemio-logical horizon has been broadened by evidence and speculation that epidemic trajectories are shaped by demographic, social, economic and network configurations. ⁴ Regrettably, this emphasis on community susceptibility has seemingly eclipsed attention to actual vectors for HIV transmission.

Analyses Focus Chiefly on Contextual Factors

A measure of how far HIV epidemiology has strayed from assessing variables directly connected to transmission is the fact that in none of the dozen articles in the Sexually Transmitted Diseases supplement, nor in the NAMC report, nor in a recent CDC summary of the disparities, ¹ is there mention of heterosexual anal intercourse – the practice most strongly associated with sexual transmission of HIV in heterosexuals. ⁵ Neither is this key variable properly measured in the national HIV surveillance system coordinated by the CDC. For example, the extensive CDC questionnaire ⁶ used to interview selected newly reported cases of HIV or AIDS from sentinel states comprises nearly 65 pages of demographic, social, economic, behavioural and medical questions. Virtually all sections concerning risk behaviours include an ‘ever’ question (e.g., ‘Have you ever used a needle to inject [shoot] drugs …’) and assess ‘when’ or ‘how often’ (e.g., ‘… how many times in your life have you injected …'?), but the exception is anal intercourse. For this risk factor, respondents are asked only whether they had anal sex the last time they had sex. Despite this weak assessment, Lansky and her CDC colleagues concluded, based on their analysis of these surveillance data, that: ‘… our findings suggest the importance of providing specific information on risk of HIV transmission through anal sex among heterosexuals….'^7(p.488) Unfortunately, this conclusion did not affect any of the subsequent recommendations published by the CDC and NMAC.

Considering Direct Transmission Vectors

Associations between poverty, illicit drug use, incarceration and risk of HIV infection are noted in both the Sexually Transmitted Diseases supplement ⁸ and the NMAC report. ³ Johnson and Raphael suggest that ‘… the higher incarceration rates among black men… explain a substantial share of the racial disparity in AIDS infection between black women and women of other racial and ethnic groups.'^9(p.3) Especially interesting is the recently reported strong association between receiving a tattoo and HIV seroconversion while in prison, a finding which the authors highlight as worthy of further investigation. ¹⁰ It might be that poverty is also more closely connected to iatrogenic transmission events than conventionally imagined. David Gisselquist (personal written communication, 23 March 2007) speculates that ‘in the US, many poor people who do not have health insurance buy an unknown and largely unstudied proportion of their health care out of pocket from unlicensed and illegal practitioners in situations where standard precautions might not be followed.’ It is noteworthy that while the formal CDC HIV surveillance instrument ⁶ assesses many indirect, contextual variables, it does not collect information on direct puncturing exposures other than those for injection drug use and (obliquely) blood transfusion. That unhygienic medical, dental or cosmetic procedures may occur in the USA (as they do in many poor countries) should not be dismissed a priori. Indeed, results of a recent investigation of nosocomial hepatitis B transmission in a USA dental surgery, where strict infection control practices were reportedly followed, suggest that blood exposures ought to be routinely and comprehensively assessed when investigating the transmission of blood-borne viruses, including HIV.^11,12

Very little research has examined anal intercourse as a risk factor for HIV transmission in heterosexual black Americans. This may be due to what Voeller terms ‘anorectophobia.' ¹³ Perhaps another reason, acknowledged by CDC researchers, is fear of embarrassing study participants.^14,15 Nonetheless, Chirgwin et al. found self-reported anal intercourse to be the best predictor of incident HIV infection in a sample of inner-city black women who did not inject illicit drugs. ¹⁶ To our knowledge, no other cohort studies of HIV incidence have been conducted in similar populations of urban or rural black American women. The Chirgwin et al. finding is consistent with studies showing that anal sex is associated with incident HIV infection in other heterosexual populations^17,18 and with in vitro studies showing that rectal tissue could be readily infected with HIV while reasonably healthy vaginal tissue could not.^19,20 Although the prevalence of anal sex in women in the American general population does not differ substantially across ethnic groups ²¹ , such ecological comparisons cannot reveal factors involved with HIV transmission at the individual level.

Reducing HIV Transmission in Black Heterosexual Populations

On 9 March 2007, the CDC announced its intention to do much more to reduce HIV propagation via its ‘Heightened National Response to the HIV/AIDS Crisis’ in black American communities, recommending that we ‘scale up known, effective HIV-prevention interventions and implement new, improved, culturally appropriate HIV/AIDS strategies.'^1(p.190) We concur and propose that the CDC should also consider implementing its own six-year-old recommendation ⁷ that heterosexuals be specifically warned about the high risks of anal intercourse. Encouragingly, the Foundation for AIDS Research (amfAR) recently funded several studies ‘aimed at increasing understanding and prevention of rectal HIV transmission'; ²² Dr Rowena Johnson notes that ‘25 years after the first identification of AIDS, the taboos that surround an open discussion of sexual behaviour are still haunting us in our efforts to contain this epidemic. It is time for us to take an honest and unflinching look at how HIV is spread and how to minimize the risks. This new research should help us further untangle this riddle.' ²² We recommend that such efforts be augmented by assessing unconventional and unsuspected exposures to potentially contaminated sharps and employing research designs that allow routes of transmission to be determined with confidence.^{11,12,23–25} These designs involve tracing sexual and non-sexual contacts of recently infected persons and uninfected controls, coupled with DNA sequencing of infected persons’ HIV isolates. In fact, these investigations could be quickly and widely conducted with minor modifications to routine HIV contact tracing procedures. ²⁶

We do not know the extent to which unsuspected blood exposures, unprotected anal intercourse or any other un(der)assessed yet plausible HIV transmission vector can account for ethnic disparities in HIV transmission in the United States. Neither do public health authorities. The point is that to determine validly what is driving the HIV epidemic in high-risk communities, researchers must return to epidemiologic fundamentals by looking at the full spectrum of transmission modes and using the most illuminating investigative tools. Appropriate prevention strategies critically depend on such fundamental evidence.