Political determinants of variable aetiology resonance: explaining the African AIDS epidemics

INTRODUCTION

African governments […] and international institutions have proved remarkably effective at managing the HIV/AIDS epidemic in a way that minimizes political threats. ^1,p3

This paradigm likely leads to underestimating the role of iatrogenic HIV transmission and to the omission of non-behaviour-related determinants of sexual HIV transmission. This politically biased view results in prevention policies that, notwithstanding the dramatic scope of the African epidemics, do not question existing political and economic hierarchies.

THE BEHAVIOURAL PARADIGM AND ITS POLITICAL SIDE-EFFECTS

Why has HIV spread with such unique efficiency in sub-Saharan Africa and there alone? This is, in essence, the central question to be answered to design effective prevention programmes in the region. Today, the variation of HIV prevalence between countries and world regions are primarily explained by variations in sexual behaviour (e.g. rates of unprotected sex and concurrent sexual partnerships). Sexual behaviour is the main variable in the dominant explanatory model, directly derived from studies of Western AIDS epidemics. Confronted with the extremely high prevalence in SSA, AIDS scholars thus deduced that ‘African’ sexual behaviour must fundamentally differ from that of the rest of the world. Consequently, behaviour change communication has become the core element of today's HIV prevention policies in SSA. The Joint United Nations Programme on HIV/AIDS (UNAIDS) for instance, suggests that in 2010, 73% of the funds for HIV prevention in SSA should be spent on interventions closely related to behaviour change. ² Sexual intercourse probably is the major path of HIV transmission in SSA and individual behaviour change should play an important role in preventing any further spread of the epidemic.

However, empirical evidence has made it increasingly clear that not only is there no homogenous ‘African’ sexual behaviour, but that behavioural differences fail to explain the substantial variations in HIV infection rates between countries and regions. ^3,4,Chaps1+7 Not only is there no evidence for significant variations in the frequency of sexual contacts or the average number of lifetime partners between SSA and the rest of the world, but ‘there is no empirical basis for the presumption that high rates of sexual activity correlate with high HIV prevalence’. ^{4,pp17–27,5} To be sure, the risk per exposure (i.e. unprotected sexual intercourse) varies because of the natural evolution of the virus in the body over time. ⁶ Consequently, differences in types of sexual networks might play an important role in the spread of the epidemic. At a similar number of sexual contacts, HIV is likely to spread faster in ‘nets’ (networks with regular concurrent sexual partnerships) than in ‘chains’ (people have sex with different partners over time, but only with one partner over a given time period). ⁷ The bio-medical argument and the results of mathematical modelling are compelling, ⁸ but reliable empiric data on concurrent sexual partnerships are sparse. ⁹ While some studies claim a clear correlation between concurrency and levels of HIV infection, ¹⁰ others show no correlation between the two. ^11,12 In any case, even if concurrent sexual partnerships might explain some of the variation in HIV prevalence among countries and regions, the ‘unexplained remaining variation’ remains very important. ^3,13,14 In sum, sexual behaviour is only part of the story and definitely an insufficient explanation for the dynamics of the epidemic spread of HIV in SSA.

Over the last decade, the behavioural paradigm has had to face mounting empirical evidence pointing to the unjustifiable character of its monopolistic explanatory protension. Far from attempting to settle the ongoing debate about the drivers of the HIV epidemics in SSA, this paper stresses that the translation of the available bio-medical and epidemiological evidence into defensible prevention measures has been weak. While a certain gap between scientific knowledge and public policies seems both unavoidable and necessary, the variable political resonance of different explanatory approaches (and the prevention measures that come with them) is not random: The official acknowledgement of the causal relationships underlying the spread of HIV in SSA is politically mediated. While scientific insights play a key role in the construction of aetiologies, their respective resonance is, to a great extent, the result of power relationships. What is frequently overlooked is that today's dominant explanatory paradigm of HIV/AIDS in SSA has significant political side-effects. By focusing nearly exclusively on sexual behaviour, it ascribes blame to individuals. By systematically dismissing iatrogenic transmission as a possible driver of the epidemic in SSA, it dismisses claims of government responsibility in the spread of the epidemic. While these side-effects might be altogether unintended, there are clear political incentives to favour certain explanations of the spread of HIV in SSA over others.*

The behavioural aetiology suffers from a double reductionism: (1) it concentrates on sexual transmission only (iatrogenic and, to a certain extent, mother-to-child transmission falling by the wayside) and (2) within sexual transmission, it focuses exclusively on the immediate cause of transmission (unprotected sex), neglecting structural determinants of infectivity and vulnerability to infection due to biological co-factors.

A SINGLE-MINDED FIXATION ON SEXUAL TRANSMISSION

While most official figures estimate that about 90% of adult HIV infections in SSA are due to sexual transmission, some authors have voiced concern about a possible and potentially dramatic underestimation of iatrogenic transmission of HIV. ¹⁵ Gisselquist et al. even claim that ‘[e]vidence [concerning the early years of the epidemic] permits the interpretation that health care exposures caused more HIV than sexual transmission’. ^16,p148 While such an extremely high rate of iatrogenic transmission seems questionable in today's transmission networks, the role of transmission in health-care settings most probably remains underestimated in SSA. ^17,18 Indeed, even a brief stay in a public hospital or health centre in most sub-Saharan countries suffices to cast reasonable doubt on the official numbers concerning the percentages of sterilized health equipment and blood transfusions effectively tested for HIV. More often than not, these numbers are at 100% or conspicuously close to it. This comes as no surprise since anything else would be considered an inexcusable failure of public health authorities. The ensuing estimates of iatrogenic transmission of HIV are conspicuously low, if mentioned at all. Notwithstanding the official, impressive figures, the improvement of blood screening is frequently raised as an urgent priority in several sub-Saharan countries. The mere existence of this claim puts the official figures into question.

The epidemiological debate on this complex issue continues. This, however, in no way detracts from the fact that the political implications of (and potential reasons for) the possible underestimation of iatrogenic transmission of HIV/AIDS in SSA have not received adequate attention. Different authors put forward potential explanations for the omission of evidence for iatrogenic transmission. ^† None of them, however, mentions that there are clear political incentives to inflate the estimates of sexual HIV transmission. Since modes of transmission other than sex are more amenable to control by public authorities, these can potentially be associated with failures of African governments and of donor agencies. Acknowledging iatrogenic transmission would thus be politically threatening. At the same time, not adequately addressing it could become politically explosive in the future if health-care-related transmission turns out to be a significant driver of the epidemic in certain countries. This combination of the impossibility to acknowledge the role of iatrogenic transmission and the necessity to act against it might help explain some governments' recent paradoxical stand. They post up reassuring official numbers on safety in health-care settings while, at the same time, collaborating with donors to improve blood screening and other prevention measures of iatrogenic HIV transmission. Having long been neglected, the contribution of iatrogenic transmission to blood-borne epidemics in SSA might be gaining some ground for while policy-makers continue to reassure the public, some have started to take action behind the scenes.

OMITTING THE NON-BEHAVIOURAL DRIVERS OF SEXUAL TRANSMISSION

While appropriately acknowledging that the risk of sexual transmission is particularly high during the first weeks of infection, the studies based on the behavioural paradigm implicitly assume the average risk of transmission per sexual exposure to be identical all over the world. It is likely, however, that this risk varies substantially from one region to another. The risk per exposure is determined by two factors: (1) the ‘vulnerability’ to infection of the HIV-negative person and (2) the ‘infectivity’ of the HIV-positive partner, which greatly depends on his or her viral load. ¹⁹ Apart from the use of antiretrovirals, a person's viral load depends to a great extent on his or her general health and nutritional status. The healthier a person, the better his or her immune system can keep the virus from entering and replicating in the body. ^‡ In the same way, an HIV-negative person's ability to avoid infection in case of exposure depends, among other things, on his or her general health status. Another important factor in determining the risk of transmission is the health of both partners' genital mucous membranes. Any infection, even asymptomatic, increases their permeability to the virus and thus probably significantly adds to the risk of transmission in either direction.

A growing range of bio-medical and epidemiological studies analyses the often complex effects of different health variables on the risk of infection per exposure. ^§ Beyond the now well-known facilitating role of sexually transmitted diseases, these findings underline that the vulnerability to HIV infection also varies with conditions such as malnutrition (which is frequently exacerbated by worms), ^21–24 malaria, ^25,26 genital schistosomiasis ^{27–30,31Chap3} and other parasitic diseases bringing about often chronic immune dysregulation. ²⁹ Most of these conditions are highly co-endemic in SSA, which frequently accounts for the overwhelming share of their global disease burden, and their distribution is very similar to that of HIV/AIDS. ³¹ The prevalence of these co-factors greatly depends on the populations' access to sufficient and balanced nutrition, safe water, sanitation, adequate housing and quality primary health care. These co-factors are thus, to a great extent, poverty related. All of them impact people's immune responses. Some additionally lead to an increased viral burden in the genital tract or to an alteration of the genital mucous membranes.

The ensuing combination of increased infectivity and vulnerability certainly contributes to explaining Africa's generalized epidemics in its heterosexual populations and the substantial inter-regional variations in HIV prevalence. Indeed, in the absence of transmission-facilitating co-factors, penile–vaginal intercourse is known to be a relatively inefficient way to transmit HIV. Among otherwise healthy partners, the average risk of transmission per unprotected discordant penile–vaginal intercourse is estimated to be between 0.1% and 0.2%. ³² However, the risk per exposure probably varies dramatically according to differences in people's health status: in SSA, HIV fell on ‘fertile terrain’. ³³

Notwithstanding the claim's somewhat obscurantist and paternalistic undertone, one could argue that there should be a difference between what we know about AIDS and what we say about it. Fear of iatrogenic transmission could keep people away from health facilities and an erroneous interpretation of the ‘protective’ effects of good nutrition, general health and intact genital mucous membranes might mislead people into engaging in unprotected sex. The public taboo about some aspects of the epidemic, however, should not keep us from using our knowledge as a basis of policy. Nevertheless, the debate about different HIV prevention measures remains heavily influenced by these discursive taboos. These taboos are not politically neutral.

POLITICAL BIAS IN THE SELECTION OF LEGITIMATE AETIOLOGIES

A certain type of bio-medical and epidemiological knowledge has been consistently overlooked in the process of policy formulation. Notwithstanding solid scientific evidence, HIV prevention policies in SSA disregard the potentially central issues of non-sexual transmission (e.g. via puncturing exposures) and populations' vulnerability to infection due to biological co-factors of sexual transmission. As a result, the battle against HIV/AIDS continues to focus nearly exclusively on the ‘immediate’ causes of individual HIV infection. Today's individualist explanation of the epidemic leads to the formulation of last-minute prevention interventions that put all efforts into avoiding transmission at the last possible moment, i.e., the instant of sexual intercourse. ^{4,pp179–180}

Why have certain elements of epidemiological knowledge received so little attention? As previously stated, the dominant explanation for the African HIV epidemics has profound political implications, for it places the blame on some and ‘exonerates’ others. To be sure, the ascription of blame to the individual is nothing new in the domain of public health and Paul Farmer's analysis in AIDS and Accusation illustrates that HIV/AIDS is by no means an exception in this respect. ³⁴ This remains true. Of course, even in SSA an individual has very good chances to avoid HIV infection through ‘responsible’ behaviour. At the population level, however, behavioural differences are insufficient to explain the epidemic spread of the virus in SSA and its rather limited spread in generalized, heterosexual populations in the rest of the world.

The dominant epidemiological interpretation is accompanied by a largely silent political context. If, as the dominant explanation suggests, the spread of HIV in SSA is principally due to individual behaviour, any protest levelled at the authorities related to AIDS would be viewed as superfluous, if not absurd. While it seems somewhat understandable that African governments subscribe to this explanation, it is more troubling that most international agencies and many social scientists also sing the same tune. Indeed, many AIDS researchers have become party to this nearly complete sexualization of the HIV epidemics in SSA.

A look at the potential political and economic implications of structural interventions based on the co-factor theory of vulnerability may help explain why these interventions have not been promulgated more consistently by policy-makers in SSA. To be sure, there are some ‘win–win’ interventions that address the biological vulnerability to HIV infection (e.g. massive drug administration against parasites ³⁵ ) and whose absence is chiefly due to governmental negligence. However, the policy responses necessary to effectively address issues such as nutrition, sanitation, adequate housing, access to safe water and quality primary health care are a lot more structural in nature. Indeed, satisfying these basic needs as a national priority would imply a major shift away from current national and international policies. It would entail fundamental changes in domains such as land tenure and agricultural policies, ^36,37 international trade policies, ^38,39 social and tax policies, and international agencies' requirements (e.g. IMF budget ceilings on health expenditure), to name a few.

In sum, addressing the root causes of the epidemic spread of HIV in SSA would entail profound changes in the living conditions of the poor and thus of the vast majority of the population in SSA. Economic inequality, for instance, is clearly and positively correlated with HIV prevalence. ³² Policies aiming at seriously reducing inequality are, however, fundamentally at odds with today's dominant economic paradigm. And structural HIV prevention is politically charged.

As de Waal notes, ‘AIDS has been politically domesticated’. ^1,p119 He forgets to mention, however, that the dominant explanatory paradigm of the spread of HIV in SSA substantially contributed to this political domestication. The focus on individual behaviour as the alleged main cause of African epidemics draws attention away from socioeconomic inequality and exclusion from access to essential goods and services, both of which play major roles in the spread of HIV. This depoliticized framing decouples the problem of HIV/AIDS from matters of distributional justice and prevents existing patterns of (re)distribution from being called into question.

CONCLUSION

The intricate complexity of the underlying causalities of African AIDS epidemics excludes any mono-causal explanation. It would thus be absurd to give an exclusively ‘environmental’ explanation and to disqualify any individual-based prevention effort. However, strong paradigm effects ⁴⁰ exist in the research and policy arenas of HIV/AIDS. By fuelling conformist analysis and inhibiting reasonable debate, these tendencies have done a lot of harm to both science and prevention policies.

Finally, claiming that there are incentives to keep ‘politically uncomfortable’ aspects of the epidemic off the public agenda and that the depoliticizing ‘side-effects’ of today's dominant paradigm might be more than coincidental does not require resorting to conspiracy theory. To be sure, the spread of HIV/AIDS in SSA is unintended. No doubt, African leaders don't want their people to die from AIDS. But to what extent are those currently in power willing to accept fundamental changes in the allocation of political and economic resources in order to effectively address the epidemic's structural drivers? The question seems worth asking and it has been largely ignored in most of the scientific literature and public debate on HIV/AIDS in SSA.