Ten myths and one truth about generalised HIV epidemics

James D Shelton a

The Lancet 2007; 370:1809-1811. DOI:10.1016/S0140-6736(07)61755-3

Despite substantial progress against AIDS worldwide, we are still
losing ground. The number of new infections continues to dwarf the
numbers who start antiretroviral therapy in developing countries.1,2
Most infections occur in widespread or generalised epidemics in
heterosexuals in just a few countries in southern and eastern Africa.
Although HIV incidence has fallen in Uganda, Kenya, and Zimbabwe, the
generalised epidemic rages on. Something is not working. Ten
misconceptions impede prevention.

HIV spreads like wildfire—Typically it does not. HIV is very
infectious in the first weeks when virus levels are high,3 but not in
the subsequent many-year quiescent phase. Only about 8% of people
whose primary heterosexual partners have the virus become infected
each year.4 Thus Kenya has more couples in which only one person is
infected than couples in which both are (figure).5 This low
infectiousness in heterosexual relationships partly explains why HIV
has spared most of the world's populations. However, the exceptional
generalised epidemics in Africa seem largely driven by concurrent
partnerships, in which some people have more than one regular
partner. This pattern allows rapid dissemination when a new infection
is introduced6 and probably involves more frequent risky sex than in
sporadic or exclusive relationships.

Sex work is the problem—Formal sex work is uncommon in these
generalised epidemics. In Lesotho, fewer than 2% of men reported
paying for sex in the previous year, although 29% reported multiple
partners.7 Nuanced economic support is an important enabler of
regular concurrent partnerships and transactional sex, but the
targeting of sex work in prevention campaigns has limited usefulness.

Men are the problem—The behaviour of men, including cross-
generational and coercive sex, contributes substantially to the
establishment of generalised epidemics. But a heterosexual epidemic
requires some women to have multiple partners.3 The importance of
women in generalised epidemics is evidenced by the high proportion
(sometimes the majority) of discordant couples in which the woman,
not the man, is HIV positive (figure).5

Adolescents are the problem—Generalised epidemics span all
reproductive ages. Although adolescent women are affected through sex
with older men, HIV incidence increases in women in their 20s and
later in life.8 Men are infected at even older ages. Thus
interventions in young people, including abstinence, although
important, have limited usefulness.

Poverty and discrimination are the problem—These factors can surely
engender risky sex. But HIV is paradoxically more common in wealthier
people than in poorer people, perhaps because wealth and mobility
support concurrent sexual partnerships.9 Moreover, HIV has declined
without major improvements in poverty and discrimination, notably in
Zimbabwe (notwithstanding substantial economic and social distress).
Condoms are the answer—Condom use, especially by sex workers, is
crucial to the containment of concentrated epidemics, and condoms
help to protect some individuals. But condoms alone have limited
impact in generalised epidemics. Many people dislike using them
(especially in regular relationships), protection is imperfect, use
is often irregular, and condoms seem to foster disinhibition, in
which people engage in risky sex either with condoms or with the
intention of using condoms.8

HIV testing is the answer—That learning one's HIV status (hopefully
with counselling) should lead to behavioural change and reduced risk
seems intuitive. However, real-world evidence of such change is
discouraging, especially for the large majority who test negative.3
Moreover any changes must be sustained for years. And very newly
infected people, who are highly infectious, do not yet test HIV-
positive.

Treatment is the answer—Theoretically, treatment and counselling
might aid prevention by lowering viral levels (and infectiousness) in
those treated, reducing denial about HIV, and promoting behavioural
change. However, no clear effect has emerged. Indeed these salutary
effects might be outweighed by negative effects, such as resumption
of sexual activity once those on antiretrovirals feel well, and
disinhibition when people realise that HIV might no longer be a death
sentence.

New technology is the answer—Many resources are devoted to vaccines,
microbicides, and prophylactic antiretrovirals. Unfortunately any
success appears to be far off. Moreover, such innovations might be
mainly targeted only at very high-risk populations, rely on
behavioural compliance, and engender disinhibition.10 Similarly,
treatment of sexually transmitted infections to prevent HIV has been
disappointing.11 Even male circumcision, an already available,
unmistakably effective, and compelling priority will take years to
have additional substantial effect.

Sexual behaviour will not change—Actually, facing the prospect of
deadly illness, many people will change. Homosexual men in the USA
radically changed behaviour in the 1980s. And the reductions in HIV
incidence in Kenya and eastern Zimbabwe were accompanied by large
drops in multiple partners,8,12 probably largely as a spontaneous
reaction to fear.

Truthfully, our priority must be on the key driver of generalised
epidemics—concurrent partnerships. Although many people sense that
multiple partners are risky, they do not realise the particular risk
of concurrent partnerships. Indeed, technical appreciation of their
role is recent.6 But partner limitation has also been neglected
because of the culture wars between advocates of condoms and
advocates of abstinence, because it smacks of moralising, because
mass behavioural change is alien to most medical professionals, and
because of the competing priorities of HIV programmes.

Fortunately we can enhance partner-limitation behaviour, akin to the
behaviour change that many people have adopted spontaneously. State-
of-the-art behaviour-change techniques, including explicit messages,
that are sensitive to local cultures, can raise perception of
personalised risk. Even modest reductions in concurrent partnerships
could substantially dampen the epidemic dynamic. Other prevention
approaches also have merit, but they can be much more effective in
conjunction with partner-limitation. Now, more than 20 years into HIV
prevention, we have to get it right.

I thank Daniel Halperin and Willard Cates for helpful ideas on this
Comment. My views here are not necessarily those of USAID. I declare
that I have no conflict of interest.

References

1. UNAIDS, WHO. AIDS epidemic update. December, 2007:
http://data.unaids.org/pub/EPISlides/2007/2007_epiupdat...
(accessed Nov 21, 2007)..

2. World Health Organization, UNAIDS, UNICEF. Towards universal
access: scaling up priority HIV/AIDS interventions in the health
sector. Progress report, April 2007. April 17, 2007:
http://www.who.int/hiv/mediacentre/univeral_access_prog...
(accessed Nov 21, 2007)..

3. Cassell MM, Surdo A. Testing the limits of case finding for HIV
prevention. Lancet Infect Dis 2007; 7: 491-495.

4. Wawer MJ, Gray RH, Sewankambo NK, et al. Rates of HIV-1
transmission per coital act by stage of HIV-1 infection, in Rakai,
Uganda. J Infect Dis 2005; 191: 1403-1409.

5. Central Bureau of Statistics, Ministry of Health Kenya, Kenya
Medical Research Institute, Centers for Disease Control and
Prevention Kenya, ORC Macro. Kenya demographic and health survey
2003. 2004:
http://www.measuredhs.com/pubs/pub_details.cfm?ID=462&c...
(accessed Nov 21, 2007)..

6. Halperin D, Epstein H. Concurrent sexual partnerships help to
explain Africa's high level of HIV prevalence: implications for
Pevention. Lancet 2004; 364: 4-6.

7. Ministry of Health and Social Welfare Lesotho, Bureau of
Statistics Lesotho, ORC Macro. Lesotho demographic and health survey
2004. 2005:
http://www.measuredhs.com/aboutsurveys/search/metadata....
(accessed Nov 21, 2007).

8. Shelton JD. Confessions of a condom lover. Lancet 2006; 368: 1947-
1949.

9. Shelton JD, Cassell MM, Adetunji J. Is poverty or wealth at the
root of HIV?. Lancet 2005; 366: 1057-1058.

10. Imrie J, Elford J, Kippax S, Hart G. Biomedical HIV prevention—
and social science. Lancet 2007; 370: 10-11.

11. Gray RH, Wawer MJ. Randomized trials of HIV prevention. Lancet
2007; 370: 200-201.

12. Gregson S, Garnett GP, Nyamukapa CA, et al. HIV decline
associated with behavior change in eastern Zimbabwe. Science 2006;
311: 664-666.

Affiliations

a. Bureau for Global Health, US Agency for International Development,
Washington, DC 20523, USA