Research Article

High GUD Incidence in the Early 20th Century Created a Particularly Permissive Time Window for the Origin and Initial Spread of Epidemic HIV Strains

  • João Dinis de Sousa mail,

    Affiliation: Laboratory for Clinical and Evolutionary Virology, Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leuven, Belgium

  • Viktor Müller,

    Affiliation: Institute of Biology, Eötvös Loránd University, Budapest, Hungary

  • Philippe Lemey,

    Affiliation: Laboratory for Clinical and Evolutionary Virology, Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leuven, Belgium

  • Anne-Mieke Vandamme

    Affiliations: Laboratory for Clinical and Evolutionary Virology, Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leuven, Belgium, Centro de Malária e Outras Doenças Tropicais, Instituto de Higiene e Medicina Tropical, Universidade Nova de Lisboa, Lisboa, Portugal

  • Published: April 01, 2010
  • DOI: 10.1371/journal.pone.0009936

Reader Comments (6)

Post a new comment on this article

Important contribution but treatment of alternatives is misleading

Posted by jmoore on 14 Apr 2010 at 17:32 GMT

De Sousa et al.'s article adds valuable data relevant to our understanding the origins of HIV, and may be correct in its conclusions. However it is seriously misleading in its representation of history and of alternative theories regarding the adaptation of SIV to humans.

Much of the force of their argument comes from the fact that the authors' "hypothesis satisfies both temporal and spatial coincidence between the factors which we invoke and the emergence of a pathogen" (p. 12). Of course; researchers generally situate their theories in space and time as those are understood when the theory is proposed.

de Sousa et al. leave the reader with the impression that theirs is the only theory consistent with a rise in risk factors c. 1900 – 1940 followed by a decline, the pattern necessary to account for multiple adaptations of SIV to humans during a narrow historical window. This is not the case. Colonial medical practices in conjunction with brutal slave labor and campaigns to control sleeping sickness and smallpox (in addition to STDs) created a constellation of risk factors that peaked roughly between 1910 – 1930 in French Equatorial Africa and slightly later in French West Africa (consistent with timing of HIV-1 and HIV-2 origins).

After (rightly) rejecting the polio vaccine hypothesis, the authors introduce alternative theories with "[a]nother hypothesis proposes that unsterile injections serially transmitted SIV from a bushmeat handler to other humans in a chain of acutely infected people, improving its adaptation to the new host [4,23,24]. Also hunting intensification [5], social changes, urbanization, and/or increased human mobility [3] have been invoked as explanations for HIV emergence" (p. 2). This is problematic. Their ref. 5, Chitnis et al. 2000, presents the case for colonial medical practices and mentions hunting in passing. Their references 4 and 23 situate unsterile needles firmly in the context of post-WWII plastic disposables, failing to distinguish Pépin & Labbé's (2008, their ref. 24) thorough description of pre-WWII medical practices – a critical distinction given de Sousa et al.'s recognition of the importance of satisfying temporal coincidence in such disease origin theories.

Doctors in French Equatorial Africa were faced with numerous diseases. Their response included aggressive campaigns against smallpox and sleeping sickness as well as forced inoculations within labor camps. This involved processing of up to 800 or more people a day. Literally millions of injections and diagnostic samplings (also with needles) were performed under logistically difficult field conditions.

The urbanization of Kinshasa and Brazzaville led to a combined population of about 35,000 by the mid-1920s. By comparison, the first mobile sleeping sickness team examined and treated some 89,000 people starting in 1917; they left Brazzaville with only six syringes. Prior to the development of dried vaccine in 1914, smallpox eradication teams employed arm-to-arm serial inoculations to carry the vaccine to the interior (no data are available on exactly how common this was); some 30,000 people were inoculated in Haute Sangha in 1907. The upper Sangha River passes less than 10 km from the hypothesized origin of HIV-1 group M, and Ouesso, on the Sangha bordering southeast Cameroon, had a major colonial medical facility

The upper Sangha was a center of ivory and rubber extraction between 1890 and 1914. Tactics used to force collection were brutal and resulted in mass displacements as people fled. In the 1920s, forced conscription for railroad construction camps further destroyed social networks and involved assembly-line medical treatment with multiple injections on arrival in camp. The estimated mortality rate in French Equatorial Africa and western Belgian Congo between 1880 and 1920 was about 15 million: 50 percent of the population. Physical and mental stress can depress immune function, making the adaptation of a new virus to human hosts all the easier.

As the resources of the area were stripped, the populations crashed, diseases were brought under control, and international human rights campaigns brought pressure to bear on colonial abuses, these risk factors were reduced. Like the incidence of GUD described by de Sousa et al., these risk factors for SIV transmission and adaptation rose in the late 1800s, peaked around the estimated times of HIV-1 and HIV-2 origins, and subsided soon after.

We may never truly know how HIV originated, and most or all of the mechanisms hypothesized to account for origins are likely to have played roles in spread of the diseases. Any of the scenarios COULD have happened. If the function of studying HIV origins is to see what lessons we can draw so we can minimize risk of the next such plague, it is important to openly discuss all viable theories and not try to sweep inconvenient ones under the rug.


Chitnis A, Rawls D, Moore J (2000) Origin of HIV-1 in colonial French Equatorial Africa? AIDS Res Hum Retroviruses 16: 5–8. doi:10.1089/088922200309548

Moore, J. (2004) The puzzling origins of AIDS. Amer. Sci. 92: 540-547. Available at

Pe´pin J, Labbe´ AC (2008) Noble goals, unforeseen consequences: control of tropical diseases in colonial Central Africa and the iatrogenic transmission of blood-borne viruses. Trop Med Int Health 13: 1–10. doi: 10.1111/j.1365-3156.2008.02060.x

No competing interests declared.

RE: Important contribution but treatment of alternatives is misleading

jdsousa replied to jmoore on 20 Apr 2010 at 05:46 GMT

You write:

“Colonial medical practices in conjunction with brutal slave labor and campaigns to control sleeping sickness and smallpox (in addition to STDs) created a constellation of risk factors that peaked roughly between 1910 – 1930 in French Equatorial Africa…”.

Forced (slave??) labor indeed peaked in early 20th century, and even before 1910; particularly its brutality was higher in the period you mention. We agree that it could contribute to several conditions potentially favoring SIV acquisition (by prompting flight from villages), and transmission (by promoting sexual promiscuity and commercial sex work (CSW)). But our model already accounts forced labor indirectly: high genital ulcer disease (GUD) incidences were caused by the behavior of CSWs and “femmes libres”, which in turn was promoted by highly male biased sex ratios in cities and other environments, which in turn resulted from organized recruitment (often forced) of men by colonial enterprises.

As for parenteral risks, at least measured by the intensity of injections and smallpox inoculations, they did not peak in early 20th century. They generally continued to increase well into the 1940s and 1950s. For example, let us consider the Jennerian inoculations performed annually in each relevant colony. In the Belgian Congo, they were 300,000–570,000 in the late 1920s, and then increased consistently to become 4.2–5.1 million in the late 1950s; in Cameroon they increased from 260,000–460,000 in the 1920s to 440,000–800,000 in the 1950s. In Côte d’Ivoire they increased from 170,000–405,000 in the 1920s to 0.6–1.1 million in the 1950s. And the figures from before 1920 are usually in the range of thousands or tens of thousands.

Annual yaws treatments. In Belgian Congo they increased from 33,000–89,000 in the 1920s to 210,000–290,000, then declining to 105,000–160,000 in the late 1950s; in Cameroon they increased from 14,000–30,000 in the late 1920s to 126,000 in 1958 (having peaked in the period 1945–55). Generally, yaws treatments peaked in the period 1935–55 in Central and West African colonies, but even after 1955, they remained higher than in the period 1900–1930.

Trypanosomiasis treatments peaked in the late 1920s in Cameroon, in the period 1932–38 in AEF (Pépin and Labbé 2008), and in the 1940s in AOF (our unpublished survey).

Total number of people treated annually (any disease) in each colony may be roughly proportional to number of injections (since these were used to treat many different diseases). This number was always on the increase, as health systems developed. Very roughly speaking, and for the colonies for which we collected data spanning many decades, the numbers multiply 10-fold between the period 1905–15 and the period 1925–35, and another 10-fold up to the late 1950s (our unpublished survey). Thus, if HIV-1 group M started to spread around 1908, as suggested by Worobey et al (2008), it did so in an epoch when health care in Central Africa was about 100 times less intense than in the 1950s.

You rightly point out that, in early 20th century, particularly risky procedures such as arm-to-arm smallpox inoculations peaked. The French campaigns you mention in Haute Sangha, however, were made at a time when the neighboring Cameroon was a German colony, and so hardly these particular campaigns could have serially transmitted the SIVcpz that gave rise to HIV-1-M. The trypanosomiasis and other injection campaigns of early 20th century could have had lower standards of needle hygiene than at other times, but this is hard to test in colonial medical papers that almost never mention errors, and are biased towards narratives of success.

You are right in pointing out that our reference to Chitnis et al (2000):

“Also, hunting intensification [5]…have been invoked…”

…may mislead the reader into thinking that Chitnis et al (2000) proposed “hunting intensification” as the only factor responsible for HIV origins. Therefore, we will introduce an edited Note, acknowledging that Chitnis et al (2000) call attention to a plethora of risk factors that peaked in early 20th century in French Equatorial Africa (AEF) and French West Africa (AOF) and may have contributed to HIV emergence: hunting intensification possibly associated with rubber collection, fleeing from forced labour, and gun proliferation; parenteral risks, including smallpox inoculations and use of unsterile needles; increase in sexual promiscuity, including commercial sex work.

Chitnis et al (2000) is the only main paper about the origins proposing that hunting intensification may have been a factor in HIV emergence, hence our phrasing. Also, our reference to Marx et al (2001), Gisselquist (2003), and Pépin and Labbé (2008) (refs. 4, 23, and 24 of the paper) in the previous sentence is justified because these authors are the only ones proposing only parenteral serial transmission as the key factor. Of course, other authors, including Chitnis et al (2000) also mention parenteral risks as potentially important.

Finally, we could not agree more that we should “openly discuss all viable theories and not try to sweep inconvenient ones under the rug”. We had no intention in doing that with our paper, and we explicitly state in the Discussion that we do not dismiss parenteral risks as having played a role, and do not consider our paper a refutation of alternative theories. The GUD theory however, offers a more parsimonious and consistent explanation of the known facts, and has the conceptual advantage of explaining HIV emergence by invoking no other factors than those most promoting HIV spread nowadays: GUD, CSWs, sexual promiscuity.

The references cited are listed in the paper.

No competing interests declared.

RE: RE: Important contribution but treatment of alternatives is misleading

jmoore replied to jdsousa on 21 Apr 2010 at 18:40 GMT

Thanks very much for the clarifications, Joao.
Very briefly: though number of injections rose during the 20th century, so did the standards of medical practice (e.g. sterilization of needles & syringes). While it is hard to determine exactly how careful or careless early medical campaigns were, inferences can be made from e.g. a mobile team diagnosing/treating nearly 90,000 people in 1917-1919 with only 6 syringes (McKelvey 1973). Pepín & Labbé describe at length the evidence that risks of parenteral infection were greater in this colonial period than in the post-WWII medicine that Gisselquist and Marx address. The two periods are simply discontinuous with respect to risk factors.

French, Germans and Haute Sangha. There are two issues: first, how significant was the border pre-1915, and second, how significant is the date?

The border: The French medical center at Ouesso (3rd largest in FEA; Headrick 1994) was on the Sangha just across the Ngoko River from Cameroon. Africans moved back and forth across "international borders" frequently (here I am generalizing from comments in Hochschild [1998] about the Congo and Oubangi as permeable borders between FEA and Belgian Congo; the Ngoko is a small river).

The date: Prior to the development of dried vaccine in 1914, smallpox eradication teams employed arm-to-arm serial inoculations to carry the vaccine to the interior. Records are spotty, but some 30,000 people were inoculated in Haute Sangha in 1907 (Headrick 1994). In August 1914, French forces under Gen. Aymerich invaded Cameroon by coming up the Sangha through Ouesso; Eugene Jamot was the chief medical officer of this force. The several thousand French (i.e., Congolese) plus about 600 from the Belgian Congo spent all of 1915 moving from SE Cameroon through Lomie to Yaounde; this is precisely the region where modern SIVcpz closest to HIV-1 group M is found. From the fall of Yaounde in January 1916, of course, SW Cameroon was under French control and a focus of mobile sleeping sickness eradication work in the 1920s (see Strachan 2004, Jamot 1930; Suret-Canale 1971). It is worth noting that after he was expelled from Cameroon (over medical malpractice by a subordinate) in 1931, Jamot was sent to French West Africa where he continued his fight against sleeping sickness (Suret-Canale 1971).

The period 1900 – 1925 had at least (1) forced labor transport and camps, resulting in concentrations of immunocompromised people; (2) a dramatic increase in sexual networks via CSW and social disruption (cf. 'femmes libres'); (3) the use of syringes in involuntary medical diagnoses and treatments of hundreds of thousands of people under unsanitary conditions; (4) urbanization; and (5) (presumed) changes in the use of bushmeat. Urbanization, CSW and the use of medical injections all continued to increase in the years following, but medical practices improved and (as you show) GUD prevalence decreased, so that the risks of those two factors vis a vis establishing a zoonotic disease went down even as the practices become more prevalent. It is theoretically possible that any ONE of those 5 risk factors could have been responsible for the adaptation of SIVcpz to humans; synergies among two or more seem probable.

I personally don't think it likely that we will be able to prove which set of factors was in fact responsible for adaptation of SIVcpz; at least #s 1-4 all would have contributed to the spread of the virus once it became HIV. There is a great deal to gain from better understanding specific aspects of the risks, such as the role of GUD in elevating transmission probabilities, but a great deal to be lost if the search for HIV origins is pursued as a zero-sum game in which primacy of one factor is advocated without full, explicit analysis of other proposed mechanisms. I've suggested elsewhere that the main value of these analyses is as "simulation exercises" for how zoonoses CAN become epidemics; on that view, there is something to learn from each (including the disproved disposable needles and polio vaccine theories) (Moore 2004).

Headrick R. 1994. Colonialism, Health and Illness in French Equatorial Africa, 1885–1935. African Studies Association Press, Atlanta, Georgia.
Hochschild A. 1998. King Leopold’ s Ghost: A Story of Greed, Terror, and Heroism in Colonial Africa. Houghton Mifflin, Boston.
Jamot, E. 1930. La Maladie du Sommeil au Cameroun. Africa 3: 161-177.
McKelvey JJ. 1973. Man against Tsetse: Struggle for Africa. Cornell University Press, Ithaca, New York.
Moore, J. 2004. The puzzling origins of AIDS. Amer. Sci. 92: 540-547.
Strachan, H. 2004. The First World War in Africa. Oxford University Press, Oxford.
Suret-Canale, J. 1971. French Colonialism in Tropical Africa 1900-1945. Pica Press, New York.

No competing interests declared.

RE: RE: RE: Important contribution but treatment of alternatives is misleading

jdsousa replied to jmoore on 05 May 2010 at 14:08 GMT

We agree that all factors you describe could have favoured SIV zoonoses. We referred that Cameroon was German-administrated before 1916, just to stress that the *particular* French health campaigns you referred initially would be unlikely to treat Cameroonians (although they could, since, as you say, borders were loosely controlled and people could cross them).

In our paper, we calculate that HIV-1 group M split from its closest SIV lineage around 1876 (1847–1907) (Table 1). And the most recent datings of the time when the human strain starts to spread have confidence intervals ranging from 1873 to 1933 (Worobey et al 2008) (Table 1). The cross-species transfer necessarily happened between the two events, more likely closer to the second.

So, factors intervening in HIV-1 group M adaptation to humans, if they took place in south Cameroon, could have happened either during German or during French control of the area. If parenteral risks had a role (and you can see in our Discussion that we do not dismiss this hypothesis), they could have been caused by the health teams of either colonial power.

No competing interests declared.