King Leopold’s Pandemic

The origins of HIV offer a great example of the ways treating human impact as an afterthought—discussed in our previous post—locks the study of pathogens into limited and oftentimes downright drunken trajectories.

In 2006 Beatrice Hahn and her colleagues identified the likely source for the SIVcpz progenitor that seeded HIV-1 group M, the clade that produced the AIDS pandemic. The team identified two wild chimpanzee groups in the southeastern corner of Cameroon—50 km north of Ouesso—with SIVcpz phylogenetically closest to group M.

The team hypothesized the virus spread to humans there a hundred years or so ago before making its way south by the Sangha and Congo Rivers to Leopoldville (now Kinshasa), one of several new colonial administrative centers in the area. By a relaxed molecular clock permitting different rates of nucleotide substitution across phylogenetic branches, Worobey et al. (2008) dated the emergence of group M to 1908 (1884-1924). After circulating regionally for decades, diversifying into many of its present-day subtypes, the virus exploded in population size and made its way out onto the global travel network and to the rest of the world.

It’s an exquisite piece of work and, perhaps unbeknownst to its authors, paradigm-shifting. In adding a location to the likely dates of the host switching event, the Hahn group moves us from general speculations about HIV’s emergence into a framework in which we can begin to study the specific circumstances for the origins of the world’s deadliest pandemic.

A literature review, however, shows little effort expended on the natural follow-up: what, if anything, was going on in southeast Cameroon a hundred years ago to promote the emergence of protopandemic HIV?


There has been some speculation on the fringe. George Parris, for one, thinks AIDS arose iatrogenically in a two-step selection regime. Parris hypothesizes HIV emerged in human populations in central Africa long before 1908. He claims that over this longer history the virus came to act as a natural anti-malarial agent—a pathogen-equivalent to sickle cell anemia—by helping reverse malaria’s efforts to block apoptosis in plasmodium-infected cells.

According to Parris, HIV’s relationship with its human hosts changed once pamaquine, an early anti-malarial drug, began to be used in central Africa in the 1920s. Pamaquine blocks malaria chemically but also targets these SIVs, selecting now, in Parris’s twist, for a anti-apoptotic retrovirus that turns immune cells into virus factories.

Parris pins HIV’s transfer from Cameroon to the Congo on pamaquine only by way of a single child who participated in a 1927 trial in Leopoldville, a child whose last name might indicate a southern Cameroon origin. Parris claims the child a marker for multiple families, who likely made such a trip, bringing the new HIV into the Congo for its pamaquine makeover. Tenuous doesn’t begin to describe the assertion, especially without a single blood or paraffin sample from the trial available.

According to Parris, the newly pathogenic virus then circulated in and about Leopoldville for several decades before undergoing a second round of iatrogenic selection. This time the anti-malarial cholorquine selected for increased transcription rates, infectivity, and NF-κB activation. The drug was introduced to the region in a decades-long campaign WHO started in 1955. Parris claims the virus turned virulent in populations in which no cholorquine was subsequently administered, and so went undetected until it arrived in the cholorquine-free United States.

Parris’s theory raises some interesting ideas about ecological relationships across multiple pathogen taxa, as well as the role pharmaceuticals can play in inadvertent epidemiologies, medicine’s third rail. But his claims suffer from a conspiracy theory’s predilection to twist and turn in any and all directions to produce a logically consistent story however improbable and whatever the state of the evidence. Rube Goldberg as played by Jesse Ventura. Such scenarios are indeed nothing other than “outlandish speculation”, to use Parris’s reverse psychology. At one and the same time, they are presently untestable and closed to disproof.

Parris’s story can serve us well, however, if anything in alerting us to the dangers of explaining specific historical events with circumstantial evidence.


There are other claims of an iatrogenic introduction in Cameroon.

In 1999 Edward Hooper indicted oral polio vaccination campaigns in the Congo for accidentally introducing HIV into humans. His efforts won little but medicine’s retribution and, given the devastating implications of the accusation, repeated attempts at a stake-in-the-heart refutation, even as hospital-specific diseases are now widely acknowledged. In 2004 Worobey et al. showed the SIVs in Congolese chimp populations, kidney cells from which the vaccines were cultured in, were not related to HIV’s pandemic strain.

Jaques Pépin and Annie-Claude Labbé have since shifted the locus north to Cameroon, placing the blame now on efforts there to vaccinate against yaws,

We reviewed, for 1921–1959, records of health services in Cameroun, Oubangui-Chari, Gabon and Moyen-Congo. We calculated the incidence of diseases whose treatment required the administration of IV drugs, and compared these with previously published data on HCV prevalence…

Several IV drugs were used against African trypanosomiasis, leprosy, yaws and syphilis. However, yaws was the only disease whose incidence was high enough so that up to half of some birth cohorts could have acquired HCV. Yaws incidence varied dramatically between regions, and was often >200 per 1000 per year in southern Cameroon, where extremely high HCV prevalence was found. Yaws incidence peaked between 1935 and 1955, a period which coincided with the emergence of HCV and HIV.

It’s the zombie hypothesis—it just won’t die even without proof. But given colonial medicine’s fast and loose (and oft-unethical) practice, particularly about parenteral injections, its haunting may be just desserts, an opportunity to again and again raise the roles—good, bad, and very bad—doctors played under genocidal regimes.

A second class of explanation cuts the Cameroon problem a different way. De Sousa et al., cross-listing an ethnographic atlas and the colonial medical literature, hypothesize Cameroon one of three African colonies (along with Belgian Congo and Cote d’Ivoire) with concurrently high levels of genital ulcer disease and low levels of male circumcision, synergistic co-factors for increasing transmission of a nascent HIV.

The team estimate ulcerative syphilis 100-300 times more in “promiscuous” cities there over 1920-1945 than later mid-20th century (and, in the other direction, in pre-colonial villages). These new cities were defined by a male bias in the sex ratio, promoting commercial sex industries at a time when antibiotics were also as of yet unavailable,

The earliest sharp growth periods of Kinshasa, Douala, and Abidjan, when these cities exceed 20,000 inhabitants, roughly coincide with [the most recent common ancestor] estimates of HIV-1-M, HIV-1-O, and HIV-2, respectively…and these viruses spread earlier in their respective countries.

De Sousa et al. refuse following up the implications of their own work, if true. We’ll try here—albeit in broad strokes—if only because the scope of what one is willing to investigate is critical to the resulting chain of causality acted on.

Europe’s colonial efforts annihilated much of traditional African societies and, à la Chinua Achebe, radically rearranged both the population’s physical distribution across the region’s landscape and its social order. As broadly described by Walter Rodney, the mass of African labor was redirected by both force and economic compulsion to producing for European export.

HIV/AIDS, then, emerged into—or by way of—a novel African geography: a network of growing urban cores and hinterland plantations and deposits connected into a regional production line.

The new Africa melded and juxtaposed pre-colonial and provincial social behaviors. The resulting colonial regimen—individual psyches socialized into the bottom line, population fragmentation and reorganization, relocation and cycle migration, sex and age biases urban and rural, with sex commodified professionally and less formally—spurred an increase in the virus’s transmission rates and reach. One finds multiple generations of such juxtapositions across colonial, neo-colonial, and neoliberal orders of development, widespread across the region but also organized by country-specific dynamics.

Colonial Africa was also now found at the forest’s taxonomic junctures. Clear-cutting associated with accelerated exploitation of the forest broadened sub-Sahara’s wildlife-human interface. Animals, and their pathogens, until then at one and the same time more tightly integrated at the level of the local village and more marginalized at the regional scale, became more exposed to the new subjugation. In short, as the forest’s edge grew in extent, so did the epizoological traffic.

Deforestation concomitantly turned bushmeat from a subsistence food item into a commodity that supported logging camps in the thousands and, later, farming towns growing on the edges of the contracting forest. Associated logging roads and rail better integrated the deepest forest with regional cities.

The argument against the hunting hypothesis is that the origins of the various HIV clades (HIV-1: M, N, O; HIV-2: A and B) occurred before peak exploitation. But assuming something more than chance host switching, spillover need only supersede a deforestation threshold as defined over both effort and space. Just need enough clearing over enough space to reach the requisite spillover. Indeed, spillover may be best with enough monkeys and apes about. Populations depressed to extirpation by hunting or more indirectly by widespread loss of primary forest are less likely sources for new infections.


Under this explanatory umbrella, events and circumstances stemming directly out of the colonial program permitted ancestral HIV to cross into their new hosts via repeated bouts of spillover, with a few eventually making their way out onto the new regional travel network and one, with just the right epidemiological momentum, infecting 60 million people worldwide.

Work more specific to Cameroon’s colonial history is, however, still required. A cursory review shows some good places to start. Against the colonial ideology the lands Europeans coveted were virgin territory or that the few millions who lived there primitives stuck in time, the area that would comprise Cameroon hosted the rise and fall of civilizations long before Europe itself emerged.

As Mariteuw Chimère Diaw describes southern Cameroon’s shifting landscape,

This early Bantu expansion marks the beginning, in the forest, of the enclosure movement that was to later undergo a tremendous acceleration in the 19th century, a period of much significance with regard to the present configuration of village territories and land tenure. It triggered a major recomposition of the ethno-cultural landscape until it was abruptly curtailed by the Germans who, in the years 1903 [sic], forced the resettlement of Bantu villages near to trading roads and trails. This policy was later enforced by the French colonial government. It was implemented in Gabon in the 1930’s with the systematic resettlement of villages along the newly created roads network and the C.F.C.O. railway line…

Germany colonized Cameroon in 1884, the lower limit of Worobey’s group M dating. German merchants had taken advantage of British reluctance and expanded their operations into Cameroon. In response Brits in the Gold Coast imposed import duties. German Chancellor Otto von Bismarck sent troops to establish a Cameroonian protectorate and keep clear the merchants’ regional access. The Germans would use a series of uprisings in the 1890s and early 20th century as a pretext to invade inland. In 1911 France ceded the Germans parts of French Equatorial Africa, including the southeastern corner where Hahn’s two chimp populations are presently found.

By forced labor, expropriation, and bribery, the German colonial authorities developed substantial infrastructure, including a chain of rubber, cocoa, and palm plantations for export. An extensive network of roads, bridges, and rail was built deep into the interior, including across the Sanaga South branch. Commercial logging was begun. Indigenous resistance and reform efforts back in Germany cut into these efforts. In 1916 Cameroon was invaded by an allied force and divided between Britain and France.

Ironically, Parris provides a good summation of colonial intrigue about another river, the Sangha, this one closer to our chimp populations, even if only in the context of setting up his pamaqunine theory,

Germany had colonized Cameroon beginning in 1884, and, through a series of agreements with the French, the Sangha River corridor from Cameroon to the Congo River was ceded to Germany in 1911 to provide commercial access for eastern Cameroon to the Cong trade routes. Ouesso, however, the northern-most navigable port on the west bank of the Sangha River, was still held by the French; and in 1922 (after WWI) all this territory was returned to French control.

Regardless of the national claimant, private concessions controlled most of the commercial activity and trade in central Africa until after WWI. In particular, through much of this time, a private concession company (or its derivatives) headed by Anton Greshoff controlled most development and trade on the upper Sangha River. By 1890, steamers facilitated active trade among the colonial (concession) posts. It is clear from the correspondence between Greshoff and Stanley (available in Stanley’s collected papers) that trade between his private concessions on the upper Sangha and Leopoldville was routine.

The key to this trade was the fact that downstream for Leopoldville and Brazzaville, the Congo River is a series of impassable rapids, which were bypassed with a railroad. The railroad was constructed by the Belgians between 1890 and 1898 south of the Congo River from Leopoldville to Matadi, where oceangoing vessels could dock. Thus, all rubber, tropical hardwoods and other products of the Congo basin (regardless of who produced it) was typically transported via steamer to Leopoldville and on to Matadi by rail. [From Matadi] a lot of the colonial trade went into Europe via Antwerp, Belgium.

Tamara Giles-Vernick writes of the effects of the colonists’ ‘Great Game’ on local populations and their traditional patterns of migration,

Movement continued to be a crucial dynamic of the twentieth century, as the middle and upper Sangha was increasingly incorporated into wide political economies. During [the 20th] century, however, French and German colonial administrations, concessionary companies, and other, mostly European-owned, commercial enterprises provided part of the impetus for movement. Commercial enterprises sought to exploit rubber, ivory and other natural resources of the equatorial forest but faced the formidable task of extracting labor from small, scattered populations. They tackled this challenge by tempting African workers with promises of access to new consumer goods such as cloth, salt, and alcohol.

Colonial administrations found themselves in the contradictory position of encouraging and discouraging movement across colonial boundaries. On one hand, they competed with one another and with commercial enterprises to gain control of African labor. From the 1890s to 1916, French and German administrations competed intensively to attract populations to their own territories on either side of the Sangha River…But after 1915 the French administrations sought to prevent movement, rooting people in their districts so that they could be compelled to pay taxes and labor on public works projects and to prevent the spread of sleeping sickness to other populations of the Sangha basin.

In other words, HIV may have emerged during a period in which colonial administrations battled for populations across the Dja, Ngoko and Sangha Rivers separating German Cameroon and French Equatorial Africa. Rebecca Hardin quotes one French communiqué,

Along almost the entire border the neighboring villages were French. The relations I had with them were excellent, they assured me of their Francophile sentiments, but in the end they allowed themselves to be tempted by the insinuations of the German authorities, insinuations about which the indigenous chiefs had briefed me and which were the following: The Germans are richer and stronger than the French, who won’t be able to protect you against us; if you don’t come voluntarily we will bring you by force. The French collect taxes from you and we, on the contrary, we give you gifts. The French don’t bring commerce to your homeland.

“Were French” is a touch presumptuous, oui? The locals instead played the powers off each other, a dangerous game in a watershed Adam Hochshild describes hosts one of the most brutal colonial histories in the world: wholesale extermination, murder, rape, chopping heads and hands off (including those of children), burning villages to the ground, raids by paramilitaries and company militia, kidnapping, forced labor, and out-and out slavery.


German and French expansion into and competition in the Congo Basin, and specifically in southeastern Cameroon, begs whether colonial maps housed in Berlin, Paris, Yaoundé, Brazzaville, Bertoua, Ouesso, or in company archives show specific population relocations or tracts cleared away on specific dates, including near or through what were the chimpanzees’ extant home ranges. We should be sure, however, to aim less for identifying Human Zero, epidemiology’s red herring, than for an understanding how a specific colonialism ratcheted xenospecific exposure.

Other scenarios also require attention. Worobey et al.’s estimates date HIV-1’s origins only within a range, 1884-1924. Cameroon was ruled or infiltrated by a series of colonial powers and companies, any one of whose exploitation may have sprung HIV. Of course, blaming any one power may take away from the accumulative damage Cameroon’s colonization caused to the country’s epizoological resilience.

But cataloging the efforts—by a single power or in the cumulative— sets us down a road in which we can finally begin to more finely map shifts in the country’s political fortunes onto its evolutionary epidemiology, including, if more historical samples from chimps and humans alike become available, onto HIV’s genetic code.

Are there more mature attempts at tying specific acts of exploitation to the emergence or spread of a xenospecific pathogen? In the next post we will explore an example other than HIV or influenza or anything else we’ve yet covered this past year, Farming Pathogens’s first now complete. Another epidemiology is indeed possible. We can better see infectious disease when we bother to see ourselves.


2 Responses to “King Leopold’s Pandemic”

  1. George Parris Says:

    Dear Mr. Wallace:

    I encourage your skepticism and share your hope that Dr. Worobey will obtain older tissue samples and continue his work. I would also encourage you to provide citations of my work so that your readers can read my work, understand the source of some of your comments and decide for themselves how far on the fringe I am. Readers may find my papers by searching “parris ge” on You can also find papers by the other authors cited in the blog with the author’s name.

    I have asked these questions of others and will pose them here for you to consider: Is it reasonable that humans would be the only primates in Central Africa that have never had a benign retrovirus strain of the SIV/HIV family especially since there are about a dozen examples of SIV- to-human transfers in the last 100 years and the virus has been infecting primates for much longer than that. As a matter of fact, Wertheim and Worobey (PLoS, 2009) indicate that the transfer of SIVcpz to humans resulting in HIV-1 group M was 1853 (1799-1904, 95% CI). Although Worobey et al. favor their calculation of 1908 (1884–1924, 95% CI) for the divergence of HIV-1 group M strains (cited in PLoS, 2009 from Nature 2008) a careful reading of the Nature (2008) paper (see Table 1 in that paper) reveals the following: The model that actually best fit the data was a “constant population size demographic model” prior to generation of the samples (1959-1960) and it yields a date of 1921 (1908-1933, 95% CI). This model was rejected by Worobey et al. in favor of a more poorly fitting Bayesian skyline plot apparently because they viewed the constant population size as not “biologically plausible.” The reason for this decision is actually two-fold: (i) Conventional wisdom (i.e., Hahn et al.) insists that transfer of SIV to humans per se was enough to produce HIV-1 group M as a zoonosis (i.e., the disease immediately followed the transfer) although this idea is not consistent with the numerous well-known ape-to-human transfers that have not produced pandemic strains. I sense that the 2008 (Nature) and 2009 (PLoS) papers are trying to close the gap between infection of humans with HIV-1 (earlier than 1904) and the emergence of group M (later than 1908) to prove the bushmeat hypothesis. (ii) One of the preferred rationales for escape of the pandemic strain of HIV (without any evolution) was increase in human population in Central Africa (after 1900). Of course, all the other human infections with SIV have experienced similar population growth and density without effect. I have proposed, based on published reports from the laboratory in Leopoldville (1927), that a pre-pubescent child (age 5 in 1927) from Cameroon carrying the benign HIV strain was subjected to a clinical trial of a powerful anti-malarial (pamaquine) that is know to effect the blood and that this exposure may have selected for a group of mutations that enhanced the resistance of the infected cells to apoptosis and reduced transcription of the provirus. That hypothesis accounts for the constant demographic into the late 1950s very nicely, without any ad hoc assumptions regarding the growth of population (i.e., the population growth was irrelevant as the child did not have sex for est. 10 years and the virus was not very active and did not cause AIDS at that time. By the way, if Hahn et al. are correct why didn’t anyone notice AIDS between 1904 and 1980? Why wasn’t AIDS spreading world-wide as a result of foreign soldiers and workers in Central Africa during WWI and WWII. There were major health studies in the 1920s and 1930 involving sampling blood from most of the people in Cameroon, but no one noticed AIDS, why?

    Although you have specifically been critical of my paper regarding a hypothesis for how HIV got from Cameroon to Leopoldville, none of the people you cite (Worobey, Hahn et al.) have offered an explanation for why the epicenter of HIV is not SE Cameroon. My hypothesis actually supports their hypothesis and my hypothesis is based on the eye-witness accounts of an American doctor working in Leopoldville in 1921. There are several circumstantial pieces of information connecting the girl to SE Cameroon: Her name (Ntongo) is common in SE Cameroon and there is even a town named Ntongo-Atani on the Sangha River, she had sleeping sickness that was common in Cameroon at the time, steamers were known to be bringing patients from Cameroon to Leopoldville for treatment of sleeping sickness in the 1920s.

    Contrary to your claim, there are many ways to prove my hypothesis wrong. For example, show that infection of a human with SIVcpz causes AIDS. Show that the 1927 date for the origin of HIV-1 M is implausible. Show that many people had HIV/AIDS in 1930 (at least 22 years after Worobey said that SIVcpz was transferred to humans). I suspect that even if we had hundreds of tissue samples from before 1950, we would be very unlikely to find HIV in any of them. Of course, some one could take SIVcpz in human blood and treat it with pamaquine and see if there is any selection for specific phenotypes. It would also be interesting to see what happens if chimps infected with SIV were dosed with chloroquine. There are a number of papers supporting the idea that chloroquine inhibits HIV, there is little doubt that virtually every one in Central Africa was dosed over a long period of time with chloroquine starting in the late 1950s; we know that HIV evolves; why hasn’t this been studies in detail (see Naarding et al. 2007 in Retrovirology)?


    George E. Parris
    3 11 2010

  2. rgwallace Says:

    First, George, blogs allow readers to directly access the sources discussed–including yours–by clicking on the links provided here in orange.

    Second, all your publications save one since 2004 have been published in a single obscure journal. ‘Medical Hypotheses’ isn’t peer-reviewed and has been mired in accusations of pseudo-science, including publishing claims people with Down Syndrome are ‘mongoloid’ and Peter Duesberg’s declarations AIDS isn’t responsible for African deaths. Your publications acknowledge funding only from a foundation with your family name. And your work is little *but* ad hoc supposition. Dude, I think ‘fringe’ is the nice way of describing your scholarship.

    Now I mentioned your work not merely as an example of the dangers of organizing circumstantial evidence, whatever the hypothesis, but because you 1) were one of the few who bothered with following up the Hahn report, however much I disagree with your conclusions, and 2) did some informative leg work in summarizing some of the Congo Basin’s colonial history.

    One of the markers of a mature mind is to be able to separate the wheat from the chaff in all work, including that on the fringe. That said, the failures of the dominant paradigm do not magically embody support for any other, yours included. None of the tests you propose, and refuse to conduct, would refute other hypotheses, or, frankly, lend support to your own. I’m intrigued–but not enough to spend more time on this–by your inability to see anything other than an iatrogenic source for the evolution of HIV’s virulence.

    I wish you the best of luck in your efforts to ‘suspect’ the truth.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: