REVIEWS OF INFECTIOUS DISEASES (Chicago), Volume 9 Number 6: Pages 1109-19,
November-December 1987.

Cultural practices contributing to the transmission of human immunodeficiency virus in Africa.

Daniel B. Hrdy

From theDivision of Infectious and Immunologic Diseases,
University of California Davis Medical Center, Sacramento;
and the Department of Anthropology, University of
California Davis, Davis, California

Differences between the epidemiology of AIDS cases in Africa and that in Western societies have prompted speculation regarding risk factors that may be unique to Africa. Because of the age and sex distribution of AIDS cases in Africa, emphasis has been placed on sexual transmission of human immunodeficiency virus (HIV). Factors thought to influence this sexual transmission include (1) promiscuity, with a high prevalence of sexually transmitted disease; (2) sexual practices that have been associated with increased risk of transmission of AIDS virus (homosexuality and anal intercourse); and (3) cultural practices that are possibly connected with increased virus transmission (female "circumcision" and infibulation). Other nonsexual cultural practices that do not fit the age distribution pattern of AIDS but may expose individuals to HIV include (1) practices resulting in exposure to blood (medicinal bloodletting, rituals establishing "blood brotherhood," and possibly ritual and medicinal enemas); (2) practices involving the use of shared instruments (injection of medicines, ritual scarification, group circumcision, genital tatooing, and shaving of body hair); and (3) contact with nonhuman primates. At the current time promiscuity seems to be the most important cultural factor contributing to the transmission of HIV in Africa.

The recent spread of AIDS throughout Africa raises the question of whether the mode of transmission of human immunodeficiency virus (HIV) in Africa is different from that in the United States and other Western countries. Although there is widespread seropositivity to HIV in Central and East Africa [1,2] and to the related T cell-lymphyocytic virus Type III (STLV-IIIAGM) in West Africa [3], the absence of risk factors of intravenous drug abuse and homosexuality points to patterns of transmission that are different from those in Western society. The difference is especially apparent because the male-to-female ratio of affected individuals is ~1:1 in Africa vs. 19:1 in the United States and Europe [4]. This report briefly examines cultural practices that may contribute to the spread of AIDS in Africa and highlights areas that require further research.

Any hypothesis that attempts to account for the equal sex distribution of AIDS cases in Africa must take into account the apparent age distribution of the disease. Cases are found in infants (who presumably acquire the disease from their mothers) and in sexually active adults. Although data for young children are still incomplete, AIDS cases have been reported only infrequently among those age groups, except in cases of blood transfusions. (Earlier reports of HIV seropositivity in children [5] may have been the result of nonspecific reactions [6].) Hence emphasis has been placed on sexual transmission of HIV. Factors thought to influence sexual transmission in Africa include (1) sexual promiscuity, with a high prevalence of sexually transmitted disease (STD); (2) sexual practices that have been associated with a high degree of transmission of HIV (homosexuality and anal intercourse); and (3) cultural practices that are possibly connected with increased virus transmission (female "circumcision" and infibulation). Other nonsexual cultural practices that do not fit the age distribution of AIDS but may expose individuals to HIV include (1) practices resulting in exposure to blood (medicinal bloodletting, rituals establishing "blood brotherhood," and possibly ritual and medicinal enemas); (2) practices involving the use of shared instruments (injection of medicines, ritual scarification, group circumcision, genital tattooing, and shaving of body hair); and contact with nonhuman primates.

Female circumcision and Infibulation

It seems to be relatively difficult to pass HIV during normal vaginal intercourse. Only 10% of habitual heterosexual contacts of hemophiliacs with AIDS become seropositive [7]. Thus, it has been proposed that heterosexual transmission is somehow enhanced in Africa. Of course, even with a low rate of transmissibility, large numbers of sexual contacts will place a promiscuous individual at high risk for acquiring the infection. Enhanced heterosexual transmissibility may not required for explanation of the equal sex ration among AIDS cases in Africa if it is assumed that the virus originated and was spread in the promiscuous heterosexual population. The same type of "epidemiologic accident" may account for HIV transmission in the promiscuous homosexual and drug addict populations in the West. However, it has been proposed that heterosexual transmission is, in fact, enhanced in Africa because of the widespread practice of female circumcision [8,9].

Map of female excision areas

Figure 1. Distribution of female excision (hatched area) and infibulation (cross hatched area) in Africa (see [1] [4]).

Female circumcision is a euphemism for female genital mutilation. Although it is usually performed at or shortly before puberty in Africa, female circumcision has little relation to the practice of male circumcision and is not usually an initiation rite per se. Three types of female circumcision occur in Africa. The most extreme, termed infibulation or pharaonic circumcision, involves partial closure of the vaginal orifice after excision of varying amount of tissue from the vulva, In its extreme form, all of the mons veneris, labia majora and minora, and clitoris are removed and the involved areas closed by means of sutures or thorns. After the operation the thighs are strapped together for 4—8 weeks, with complete occlusion of the introitus being prevented by the insertion of a matchstick or other wooden object. A more moderate form of female circumcision is excision, which involves removal of the clitorus and part of the labia minora. The mildest form, Sunna circumcision is circumferential excision of the clitoral prepuce [10]. Another practice that involves female genital mutilation is making "gishiri cuts," which are incisions on the vaginal wall and presumably serve the same purpose as female circumcision [11]. In may cases of infibulation and occasional cases of excision, the vaginal opening must be cut open by the husband (defibulation) in order for childbirth (or in severe cases, sexual relations) to occur. After childbirth the woman is often sewn up again.

In premodern times there were various "ritual" explanations for the practice of female circumcision. However, the practice continues in modern Christian and Muslim Africa. When it can be afforded, infibulation now is often performed in hospitals (primarily in northwestern Africa [12]), so a "ritual" or "traditional" explanation for the practice seems less likely. One recent theory proposes that the practice is an effort by males (and lineages) to curtail female sexual pleasure and hence illicit sexual liasons, thereby increasing certainty regarding paternity [13]. In rural tribes where female circumcision occurs, it is nearly universal in the female population; however, its prevalence is decreasing in urban areas [11]

Female circumcision has been postulated to increase the likelihood of AIDS transmission via increased exposure to blood in the vaginal canal [8]. The presumed explanation is that the small introitus, the presence of scar tissue (which may cause tissue friability), and the abnormal anatomy of a mutilated vagina would predispose to numerous small (or large) tears in the mucosa during intercourse. These tears would tend to make the squamous vaginal epithelium similar in permeability to the columnar mucosa of the rectum, with increased absorption of secretions (and virus). A less likely explanation involves sexual intercourse shortly at or shortly after the time of female circumcision, when open wounds are present.

Map of high HIV seropositity areas

Figure 2. Tentative distributions of areas with a high level of seropositivity for human immunodeficiency virus.

There are several reasons why female circumcision may not be an adequate explanation for enhanced heterosexual AIDS transmission. Although the presence of lesion in the vagina may increase male-to-female transmission, it is unclear how female-to-male transmission would be enhanced in this situation. A possible cofactor is untreated STD, which could result in the breakdown of the mucosal integrity of the male sex organs. [4]

Another major problem is the lack of correspondence of the distribution of AIDS and that of female circumcision (figure 1). Infibulation occurs mainly in the Arabian peninsula, isolated areas of West Africa, the Horn of Africa, the Sudan and northen Kenya—areas that are not known at present as "AIDS areas." Thus, infibulation is not correlated with AIDS transmission. Excision is performed in sub-Saharan Africa (both in West Africa and in East Africa) but not in most areas of Central Africa. The only area in Zaire that is affected is in the north.

Although data are sketchy [10, 11, 14], female circumcision is not practiced in areas with the highest level of HIV seropositivity. From evidence presently available, these areas of high seropositivity are eastern Zaire, Rwanda, Burundi, western Uganda, northwestern Tanzania, and northern Zambia [1, 3, 5, 15-32] (figure 2). There is some overlap of areas in which excision is performed and areas with a lower degree of seropositivity, including parts of Zaire, Kenya, Central African Republic, and Tanzania. Likewise, areas of STLV-IIIAGM seropositivity overlap with areas where excision is performed in West Africa. However, if female circumcision is an important determinant of AIDS transmission, it is difficult to understand why parts of Central Africa where levels of AIDS are highest and where AIDS was first described do not practice female circumcision.

It must be stressed that data are incomplete for these areas. Traditional anthropologists tend to pursue details of sexual practices in their studies, and various political upheavals have made work in the regions involved difficult in recent times. There is also a definite problem in data collection by foreigners (especially male foreigners) on this topic.

It is possible that population movements have introduced the practice of female circumcision into urban areas where it was previously not found. In fact, it is not clear that increasing westernization and urbanization have reduced the practice of female circumcision. For example, except for the Luo, the practice is still widespread in urban areas of Kenya. This pattern may begin to change now that President Arap Moi has spoken against the practice; in contrast former President Jomo Kenyatta felt that excision was a traditional part of Kikuyu life [33].

Questions relating to female circumcision that require further research include the following: (1) Is there a correlation of female AIDS cases with circumcision? (2) Is viral antigen present more often and at higher levels in vaginal secretions of circumcised than noncircumcised women? (3) Is the degree of trauma, bleeding, and mucosal tearing associated with intercourse greater in women who are circumcised than in those who are not? (4) What is the degree of female circumcision in Central Africa? (5) Does the distribution of population groups (i.e., tribes) that have high rates of seropositivity correlate with the distribution of population groups that practice female circumcision? Needless to say, the study of many of those issues would be extremely difficult from both technical and political standpoints.


Promiscuity, especially the total number of sexual partners, is correlated with AIDS in both the United States and Africa [34,35]. Although generalizations are difficult, most traditional African societies are promiscuous by Western standards. Promiscuity occurs both premaritally and postmaritally. For instance in the Lese of Zaire, there is a period following puberty and before marriage when sexual relations between young men and a number of eligible women are virtually sanctioned by society. The father of a woman may judge the suitability of the man on the basis of the perceived willingness to invest in his daughter [36]. In the so-called "matrilineal belt" centered in south-central Africa, there is an especially high degree of adolescent promiscuity and uncertainty about paternity. This situation has probably contributed to the prominent family role of the mother's brother. For example family wealth is inherited by offspringe of the maternal uncle rather than by patrilineal descendents from the husband. That is, wealth is passed on to a known biologic relative, rather than to the offspring of a wife who may or may not be biologic kin. Matrilineal inheritance thus may reduce societal pressure to prevent promiscuity; matrilineal societies are often promiscuous societies [37]. However, promiscuity is correlated not only with matrilineal societies. Many patrilineal African societies are promiscuous as well.

There is a high prevalence of female sterility (30%-50%) in may areas of sub-Saharan Africa, especially in central Africa [38]. The distribution of infertility is patchy in affected areas. Regions of low fertility border on areas of high fertility. For example in two neighboring districts in the Sudan, the local infertility rates vary from 3.2% to 42.5% [38]. This primary sterility is thought to be due to high levels of STDs that result in pelvic inflammatory disease in young women [39]; transmission of STDs is presumably enhanced by promiscuity. It is of interest that the "infertility belt" is in areas with a high prevalence of antibody to AIDS virus, which also may be related to promiscuity.

As people leave rural villages and migrate to urban areas, the general level of promiscuity usually increases. This increase may be attributable in part to the relaxation of traditional village values but appears to be due primarily to the destitution of poor migrant women, who may become prostitutes, and to the greater mobility and rootlessness of young male migrants and soldiers. Unlike some Asian societies, traditional African societies have no apparent pattern of ritual prostitution, and it is unlikely that women who become prostitutes for purely monetary reasons would be tolerated in traditional surroundings. Increased prosmiscuity is especially common among upper- and middle-class urban men, who can afford the services of prostitutes.

As has been noted previously, levels of STDs are generally high in Africa [4]; this fact may reflect both casual attitudes toward sex and high levels of promiscuity as well as the lack of easily available treatment. Prostitutes are important in the transmission of STDs in Africa [40] and may be important in the transmission of AIDS both in the West and in Africa [41]. Schuster [42] provides an in-depth treatment of the lives of Zambian career women who preceive themselves as better off trading sex for favors and expensive gifts than marrying; contemporary urban life provides wives with little of the traditional support systems of the village, and the lives of married women are isolated, bleak, and improvished [41] It is perhaps significant that the first cases of AIDS in Central Africa were reported by in upper class Zaireans seeking medical treatment in Europe [14]. Except for a few rural areas (e.g., Ruwanda and western Uganda), AIDS in Africa may be commoner in urban centers [1, 43], and the mass of increased promiscuity in urban areas is in accord with this possible trend. However, not enough data are available on either the presence of AIDS in rural areas or sexual patterns of urban and rural areas for the establishment of definite correlations.

Population movements in Africa contribute to the "sexual mixing" of various African groups and may be related to the spread of AIDS. The entire Central African area (and indeed the whole of sub-Saharan Africa) is experiencing large shifts in population. Some patterns have existed for long periods, such as the movement of Arabic and Nilotic peoples into the northern part of Central Africa [44]. The long-term movement of rural population into urban areas is also continuing. Other more recent trends include the movement of migrant workers from Zaire and Rwanda to neighboring countries (e.g., to the copper belt in Zambia), the movement of armies on the Uganda-Tanzania border, and the presence of large numbers of refugees, especially from Uganda. It is probably significant that AIDS cases seem to have been present in Africa only since the 1970s [15, 45, 46]--a time frame that correlates with the intensification of urbanization and population shifts.

The relative efficiencies of HIV transmission from male to female and from female to male are still unclear. If these efficiencies are equal and prostitutes represented the major reservoir of HIV infection, a higher male-to-female ratio of cases would be expected since each prostitute has many sexual contacts. However, it is unlikely that female promiscuity is confined to "professional" prostitutes, especially in urban areas [42]. It is also possible that the male-to-female transmissibility of HIV is higher than female-to-male transmissibility, presumably because of the higher concentration of HIV in semen than in cervical secretions [47]. Both of these factors would tend to produce a more equal sex ratio among cases of HIV infection.

Data from a high-infertility area of Uganda indicate that the rate of carriage of gonorrhea is 8.9% among men and 18.3% among women. (These values are in contrast to the corresponding carriage rates of 4.2% and 2.4% found in a low-infertility area [48].) Hence gonorrhea rates in this high-infertility region tend to be at least as high in women as in men. Of course, data on gonorrhea and other STDs are not strictly relevant to HIV infection since classical STDs can be successfully treated and are more often asymptomatic in women than in men.

Although the link between the risk of acquiring AIDS and promiscuity seems to be clear at this time, there are some unanswered questions: (1) How does viral transfer occur from male to female and from female to male? (2) Is there a close correlation between AIDS and sexual promiscuity in rural areas as well as in urban areas, or is the distribution of AIDS in rural Africa due to some other factor--e.g., transmission from a nonhuman species or via a nonsexual cultural practice? (3) Are the few cases of AIDS reported in prepubertal children (e.g., cases of "slim disease" in patients aged 9, 11, and 13 years [17] related to sexual activity? (4) Does the spread of HIV seropositivity in Africa track movments of migrant laborers, armies, and prostitutes out of Central Africa?

Homosexuality and Anal Intercourse

Homosexuality is not a part of traditional societies in Sub-Saharan Africa [44, 49]. The few instances of homosexuality noted are related to societal institutions where an older man has authority over younger males. In the Bwamba of Central Africa, a male teacher of some young boys was reported to have exposed his penis and then asked the boys to "blow it like a whistle" [50]. Homosexuality probably also exists to some extent in migrant labor camps, where few women are present. These anecdotal accounts do not indicate widespread homosexuality like that which seems to occur in some societies. Nowhere is traditional African society is there the kind of sequential homosexual activity between men that is found in urban Western societies.

This pattern seems to hold in urban areas of Africa. Most Africanists uniformly deny the presence of significant homosexual activity, as do Africans themselves. It was reportedly difficult to obtain African labor for railroad work in East Africa because Africans were revolted by the homosexual practices of Indian laborers that they refused to work with them [51]. It is likely that, as elsewhere in the world, there are pockets of homosexuality in Africa, but homosexuality does not seem to be practiced as overtly and commonly as in other parts of the world.

The apparent lack of AIDS among homosexuals in Africa also supports the absence of significant homosexual activity. However, a caveat must be injected here: many field workers have noted that it is difficult to obtain accurate information on sexual practices that Africans perceive to be offensive to Westerners. In addition, homosexuality is illegal in many African countries. Moreover, there is a common desire for informants to answer any question with the answer that the questioner is perceived to want. An accurate survey of homosexuality in urban areas has probably never been conducted.

Although the spread of AIDS in Africa seems to be primarily a heterosexual process, studies of the frequency and distribution of homosexuality in Africa need to be done. Needless to say, such studies would be difficult to carry out because of the sociologic and political obstacles that have already been mentioned.

Another practice that is correlated with the acquisition of AIDS in Western societies is anal intercourse. There is less information available on anal intercourse than on homosexuality. Again, this as a practice whose existence is denied by Africanists who specialize in Central African societies (e.g., K. Obbo and I. Schuster, personal communications) as well as Africans themselves. The only significant reports of significant rates of anal intercourse are claims made by students of genital mutilation [8, 9], who state that anal intercourse is used of necessity in cases of infibulation. However this practice is mostly limited to Arabic-influenced areas rather than Central Africa. It must be mentioned that societal disapproval of the practices of homosexuality and anal intercourse also exists as a result of the prevalence of Christianity in Central and East Africa. Catholicism, which is one of the strongest forces opposing homosexuality, is common in Central Africa, encompassing over 50% of the population in Rwanda and eastern Zaire. Any study of the practice of anal intercourse would be limited by all of the factors mentioned above.

Practices Resulting in Exposure to Blood

Bloodletting for medicinal purposes is common in sub-Saharan Africa, especially in rural areas. The practice involves a practitioner's using an instrument such as a razor blade or knife to make cuts deep enough to allow blood to flow freely. The practitioner then takes a mixture of ashes from leaves and rubs it into the wound, potentially exposing himself to the patient's blood. In eastern Zaire the cuts are made over the affected organ--e.g., the forehead for headache, the abdomen for stomach pains, and the joints for arthritis. A conservative estimate is that most people among the Lese have bloodletting (chanja) performed at least five times per year from infancy onward. Many people have more than 100 chanjas performed yearly. (R. Bailey, personal communication). The practice is primarily rural but continues to some extent in urban areas.

Traditional rural healers and midwives also are occasionally exposed to blood during their duties. African birthing practices can involve a minimum of interference (such as the in the !Kung San, where the woman gives birth in an isolated area) or active interference (e.g., pulling the baby from the uterous by force). Among the Efe of Zaire, a woman is assisted by other women, one of who inserts her fingers into the vagina periodically to monitor the baby's position. After birth, the umbilical cord is cut with a hunting arrow [52].

Blood brotherhood was formerly a widespread practice in much of East and Central Africa, especially along pastoral groups such as the Masai of Kenya and Tanzania and the Bokongo of Zaire [53]. Louis Leakey became a blood brother in the Masai tribe in Kenya [54]. This practice has been decreasing in recent years. The custom involves exchanging a small amount of venous blood via the binding of excisions. This practice is typically prevalent among males only, although some females have reportedly become "blood brothers" [50]. A male may become blood brothers with several people, but the custom is usually limited to members of a person's own tribe. One report identifies a case of AIDS possibly acquired from blood brotherhood, but other risk factors may have been present [55].

Enemas, usually of local herbs and water, are commonly given for both ritual and medicinal purposes in Central and Southern Africa. To date AIDS virus has not been found in stool, so it is difficult to ascertain a possible relation between AIDS transmission and the giving of enemas. Piot et al. [45] found no correlation of the practices with AIDS in Zaire.

The above-mentioned practices seem to present little opportunity for the transmission of AIDS virus on a large scale. However, occasional cases may be related to these practices.

Practices involving the Use of Shared Instruments

Many researchers have raised the possibility that unsterilized or shared needles may be responsible for AIDS transmission in Africa. "Injection doctors" are common in many third-world countries. These are usually untrained traveling practitioners who administer over-the-counter parenteral antibiotics. Their practice involves intramuscular injections, usually with equipment that is less than optimally sterilized, for a wide variety of indications. Many in not most of the indications would not be treated with antibiotics (oral or parenteral) in developed countries. However, clinics and other more formal medical providers in developing countries may also reuse improperly sterilized equipment (L. Altman, personal communication). A high incidence of abscesses related to injections has been reported in many countries [56].

A major difference between drug use by addicts in the West and drug administration by injection doctors is that iv-drug users inject substances directly into the bloodstream and often draw blood back into the syringe to ascertain whether the needle is in the vein, whereas injection doctors utilize im injection, which involves less exposure to blood. (Intravenous-drug abuse in reportedly rare in Africa, but insufficient hard data exist on this point.) Needle stick injuries can transmit HIV [57] but this mode of transmission is relatively uncommon [58,59]. The injections administered by injection doctors from contaminated syringes are often fairly widely separated in time, which makes transmission even less likely. HIV is relatively unstable in the environment and is inactivated by a number of substances [60]. However, high-tier virus was found to be infective in a dried state at room temperature for 3 days [61]. It is unknown what relation this finding has to the survival of low-titer virus on a syringe.

Previous studies have reported no relation between AIDS cases and needle exposure [2, 45, 62, 63], but exposure to needles is common among both patients and controls. Mann et al. [64, 65] recently reported a possible correlation between a history of injections and AIDS in Zaire. However, as patients with AIDS became sicker, they may have sought out injections as a possible cure, with a resulting bias in the findings of these authors. Other evidence against needles as a major cause of HIV transmission includes the relatively low number of childhood AIDS cases and the presence of AIDS in upper-class Central Africans who do not use unsterilized needles.

A subset of practitioners that can spread HIV may exists. Further studies on the use of injections in AIDS patients are required. It may be possible to culture used needles for HIV or to test for transmission by needles in a primate model. The whole area of viral transmission by inanimate objects is currently poorly understood; fortunately, the topic seems amenable to laboratory investigation.

Other African cultural practices involve the use of shared instruments. These practices – scarification, group circumcision, and genital tatooing – are associated with traditional African societies [44]. As westernization and urbanization continue, such practices are becoming increasingly less common. Scarification in particular is highly disapproved of by missionaries. Colonial governments banned scarification at one time, and this prohibition left a negative attitude even in postcolonial times. However, it is difficult to generalize about any cultural trait in Africa since cultural differences differ widely from tribe to tribe.

Scarification is probably the commonest of all practices involving shared instruments. It is found among many groups in West, Central, and East Africa (e.g., Nuba, Dinka, Tiv [66]). Typically, cuts are made for beautification as well as ritual purposes. Linear scars are produced by deep cuts with a sharp instrument. For round cuts a hooked horn is typically used to life the skin and pull it up, and a small razor blade or knife is used to slice the raised skin and produce a prominent scar [66]. Keloid formation is often a desired result. Scarification is carried out during multiple procedures at different ages throughout childhood, with puberty being especially important. The ritual explanation is that these scars enforce group affiliation and promote tribal integration. The marks vary from tribe to tribe and can include both facial markings and body markings as earlobe stretching. Possible exposure to virus results from the sharing of an instrument to perform the operations. The rarity of AIDS in children argues against scarification as a significant factor in the transmission of HIV.

Another form of scarification involves a type of witchcraft in which the subject (often a young batchelor) puts a mixture believed to have magic properties on a razor blade or knife, unobstrusively approaches an unwilling victim (typically) a young woman), and makes an incision in the victim's back or arm [50]. In rural Africa razor blades are often difficult to come by; thus they are frequently used by more than one individual for a number of functions, such as shaving facial or body hair. The shaving of body hair has been noted to occur in Rwanda and other parts of Central and North Africa for purposes of beautification [66]. Again, despite the constraints of an unstable virus, AIDS transmission may be at least theoretically possible, if unlikely.

Another variant of scarification includes genital tattooing, which is practiced in parts of Central and North Africa (e.g., Ashanti; K. Paige, personal communication). In females the practice involves repetitive insertion of a nonsterilized needle with pigment directly into the labia and vaginal wall. Similarly, traditional circumcision or subincision practices involve the cutting of the foreskin of young males around puberty. These practices are often performed at long intervals (10-20 years) in each village; thus a relatively large number of individuals of varying ages will be operated on at one time. Since one instrument is typically used, it is conceivable that viral transmission could occur, although the physical constraints on viral survival would apply. It should be mentioned that "ritual activities are usually performed separately for males and females, so that we might expect patterns of transfer to be female-to-female and male to male. Further research in an area where AIDS antibodies and ritual scarification are prevalent may determine whether clustering of cases occurs by sex or by age--patterns that might point to nonsexual transmission of HIV.

Contact with Nonhuman Primates

Isolation of HIV-like viruses from African green monkeys and macaques has raised the possibility that AIDS has somehow been passed to humans from non-human primates. Such transmission could have been an isolated, rare event, with HIV evolving independently from a single point of origin. It is also possible, though less likely, that such transmission occurs relatively frequently and some amount of AIDS in Africa is the result of recurrent transmission from monkeys. Desrosiers [67] has pointed out that the low degree (<75%) of sequence homology between human T-cell-lymphotropic virus type III/lymphadenopathy-associated virus (HTLV-III/LAV) and STLV-IIIAGM, argues against recent tranmission of viruses from monkeys to humans.

The presence in West Africa of STLV-IIIAGM, which is different from HIV in Central Africa and the West, points to several sources of origin of HIV in Africa. However, the extremely high rate of mutation of HIV may also be responsible for this viral variability.

Map of green monkey habitat

Figure 3. Distribution of Cercopithecus aethiops (see [69]).

To date, African green monkeys (Cercopithecus aethiops, or vervets) are the only species reported to have an AIDS-type virus in the wild [68]. Hence vervets are the principal suspect in the transmission of disease. Although the distribution of the vervet's forest-fringe and savannah habitats does not overlap perfectly with the currently known distribution of AIDS (particularly central Zaire), the degree of overlap is generally fairly good (figure 3) [69]. This opportunistic and ecologically adaptable primate species is spread widely, if discontinuously, throughout areas of high HIV seropositivity.

The existence of STLV-IIIAGM is wild vervets may be relevant to the situation in humans. There is a striking analogy between promiscuity as a risk factor in humans and the "promisuous" behavior of vervets. Typically, female vervets, unlike baboons, are sexually receptive for long periods (many weeks [70]) and during that time mate with multiple male partners, sometimes engaging in dozens of copulations on a single day -- activity that may lead to traumatic lesions of the vaginal or perineal area. Although vervets do not exhibit the large and fragile "sexual swellings common to other cercopithecine monkeys (like baboons), vervet perineal skin is slightly edematous during the breeding season. Exposure to multiple sexual partners may be a factor in the spread of STLV-IIIAGM throughout vervet populations. However, other factors, such as between-group transfer, population structure, and ecologic variables, must be studied in order to determine how the virus spreads in primates.

Although a possible source of exposure of humans to monkey virus is bites from monkeys kept as pets, the commonest type of exposure to monkeys, including vervets, is the hunting of monkeys for meat. In all rural areas and in many cities (e.g., Kigali, the capital of Rwanda), smoked monkeys are a common food item for sale in markets [71]. The monkeys are either trapped or shot, eviscerated, and prepared for eating by singeing of the hair and then smoking of the carcass over a fire. During the evisceration process, humans are exposed to blood and internal secretions that presumably contain virus. Although smoked monkeys are sold throughout sub-Saharan Africa, monkey hunting is mainly limited to forested areas. These forested areas occur most prominently in Central Africa. Actual transmission of virus to humans through this process would require a break in the skin for systemic viral absorption. Although monkeys are eaten throughout sub-Saharan Africa, it is extremely unlikely that the eating of monkeys per se is a source of virus exposure. As has been mentioned, it is far more likely that any transmission of HIV-like viruses from monkeys to humans has occurred as a rare event rather than as a recurring transmission pattern.

Another source of interspecific transmission of AIDS virus could be arthropod vectors. A recent report of survival of HIV in bedbugs supports this hypothesis [72]. Since adequate testing of insect species in the wild has not been carried out, it is not clear whether vector-borne virus transmission plays an important part in the epidemiology of AIDS. Even a low-titer presence in insects could be a source of rare transmission between humans and nonhuman primates. However, more data are needed on this point.

Questions that require further study include the following: (1) What is the distribution of HIV-like viruses in wild primate populations? (2) Do monkeys other than vervets (many of them also exhibiting promiscuous mating habits [73]) also have HIV-like viruses in the wild? (3) What is the genetic interrelation of primate HIV-like viruses? (4) Do areas where vervets are hunted overlap with areas of AIDS seropositivity (especially in rural regions)? (5) In what monkey tissues is the HIV-like virus present, and are humans exposed to these tissues? (6) How easily can different HIV-like viruses be transferred between species by various routes?

Received for publication 20 November 1986 and in revised form 3 April 1987.
The author is a reciprient of a Burroughs-Welcome Young Investigator Award.
A preliminary version of this paper was prepared for a meeting of the Committee on a National Strategy for AIDS of the Institute of Medicine, National Academy of Sciences.
The helpful comments of Drs. Robert Bailey, Elizabeth Colson, William Davis, Peter Ellison, Elliot Goldstein, Sarah Hrdy, P. E. Palmer, and Joseph Skorupa and the secretarial assistance of Ms. Judy Blankenship and Ms. Brenda Burris are gratefully acknowledged.
Please address requests for reprints to Dr. Daniel B. Hrdy, Division of Infectious and Immunologic Diseases, University of California Medical Center, 4301 X Street, Sacremento, California 95817.


  1. Van de Perre P, Rouvroy D, Lepage P, Bogaerts J, Kestelyn P, Kayhigi J, Hekker AC, Butzler JP, Clumeck N. Acquired inmmodeficiency syndrome in Rwanda. Lancet 1984;2:62-5
  2. Kreiss JK, Koech D, Plummer FA, Holmes KK, Lightfoote M, Piot P, Ronald AR, Ndima-Achola MB, D'Costa LJ, Roberts P, Ngugi EN, Quinn TC. AIDS virus infection in Nairobi prostitutes; spread of the epidemic in East Africa. N Engl J Med 1986;314:414-8
  3. Barin F, M'Boup S, Denis F, Kanki P, Allan JS, Lee TH, Essex M. Serological evidence for virus related to simian T-lymphotropic retrovirus III in residents of West Africa. Lancet 1985;2:1387-9
  4. Biggar RJ. The AIDS problem in Africa. Lancet 1986;1:79-82
  5. Biggar RJ, Melbye M, Kestens L, de Feyter M, Saxinger C, Bodner AJ, Paluko L, Blattner WA, Gigase PL. Seroepidemiology of HTLV-III antibodies in a remote population of eastern Zaire. Br Med J 1985;290:808-10
  6. Biggar RJ, Giagase PL, Melbye M, Kestens L, Sarin PS, Bodner AJ, Demedts P, Steven WJ, Paluke L, Delacollete C, Blattner WA, ELISA HTML retrovirus antibody reactivity associated with maleria and immune complexes in healthy Africans. Lancet 1985;2;520-3
  7. Kreiss JK. Kitchen LW, Prince HE, Kasper CK, Essex M, Antibody to human T-lymphocytic virus type III in wives of hemophiliacs: evidence for heterosexual transmission. Ann Int Med 1985;102:623-6
  8. Mannes RA. African practices may offer insights on AIDS. New York Times, 22 November 1985
  9. Link U. AIDS in Africa [letter]. Science 1986;231:203
  10. Verzin JA. Sequelae of female circumcision. Trep Doct 1975;5:163-9
  11. Hosken F. The Hosken report: genital-sexual mutilation of females. Lexington, Mass; Women's International Network, 1983
  12. Futuh-Shandall AA. Circumcision and infibulation of females. Sudan Medical Journal 1967;5:178-211
  13. Hrdy S. The woman that never evolved. Cambridge, Mass: Harvard University Press, 1983
  14. Pieters G. Gynecologie au pays des femmes cousues. Acta Chir Belg 1972;71:173-93
  15. Chumek N, Sonner J, Taelman H, Mascart-Lenore F, De Bruyere M, Vandeperre P, Dasnow J, Marcellis L, Lamy M, Jonas C, Eychmans L, Noel H, Vanhaeverbeek M, Butler J-P. Acquired immunedeficiency syndrome in African patients. N Engl J Med 1984;310:492-7
  16. Buchanan DJ, Downing RG, Tedder RS. HTLV-III antibody positivity in Zambian copper belt [letter]. Lancet 1986:1;155
  17. Serwadda D, Mugerwa RD, Sewankambo RK, Lwegaba A, Carswell JW, Kirya GB, Bayley AC, Downing RG, Tedder RS, Clayden SA, Weiss RA, Dalgleish AG. Slim disease: a new disease in Uganda and its association with HTLV-III infection. Lancet 1985;2:849-52
  18. Galli M, Orlando G, Crocchiolo P, Ciantia F, Zanetti A, Lazzarin A. Low prevalence of LAV/HTLV-III infection in a malaria-highly endemic area of Central Africa as af May 1984 [poster 368]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  19. Froment A, Larouze B, Dazza MC, Rey M, Abondo A, Brun-Vezinet F. Lack of evidence for LAV/HTLV-III infection in remote areas of Cameroon [poster 369]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  20. Forthal DN, Getchell JP, Mann JM. Antibody to human T-lymphotrophic virus type III/lymphadenopathy-associated virus (HTLV-III/LAV in sera collected in 1976, Equateur region, Zaire [poster 370]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  21. Kalamba K, Mann JM, Francis H, Lurhuma Z, Muyembe T, Piot P. LAV/HTLV-III seroprevalence among patients without AIDS or AIDS-related complex hospitalized at the University Hospital, Kinshasa, Zaire [poster 371. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  22. Lloyd G, Downing RG, Serwadda D, Sewankambo N, Mugerwa R, Carswell W. Seroepidemiological and virological studies on slim disease in Uganda [poster 372]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  23. Nasidi A, Harry TO, Ajose-Coker OO, Ademuluyi SA, Akinyanju OO. Evidence of the LAV/HTLV III infection and AIDS-related complex in Lagos, Nigeria [poster 373]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  24. Merlin M, Gonzalez JP, Josse R, Josserand R, Ivanoff B, Georges AJ. Evaluation of the prevalence of anti LAV-HTLV III antibodies in the Central African population: about 12 sample surveys [poster 374] In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  25. Delaporte E, Ivanoff B, Brun-Vezinet F, Frost E, Salmot AG, Larouze B, IgG antibodies to LAV/HTLV-III in a rural population in Gabon [poster 375]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  26. Georges AJ, Gonzalez JP, McCormick JB, Georges MO, Merlin M, Meunier DMY, Lesbordes JL. Antibodies to LAV/HTLV III in various population groups of the Central African Republic [poster 376]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  27. Durand JP, Garrique GP, Bouloumie J, Jeandel P, Sicard J Fabresse F. AIDS in Cameroon [poster 377]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  28. Titti F, Aceti A, Verani P, Rossi GB, Bile K, Sebastiani A. Seroepidemiologic screening for antibodies to HTLV-III/LAV in Somalia [poster 380]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  29. Piot P, Plummer FA, Rey M, Ngugi EN, Rouzioux C, Ndinya-Achola JO, D'Costa LJ, Vercauteren G. Retrospective seroepidemiology of AIDS virus infection in Nairobi populations [communication 51. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  30. Bila K, Mann JM, Ngange K, Vgaly B, Miatudila M, Colebunders R. Natural history of Lav/HTML-III infection in Zaire [communication 51]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  31. Hira SK, Perine PL, Redfield RR, Chikamara DM, Wadhawan D, Myendafilumba D. The epidemiology and clinical manifestations of the acquired immune deficiency syndrome (AIDS) and its related complex (ARC) in Zambia [communcation 52]. In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  32. Nzilambi N, Mann JM, Francis H, Masaki MN, Dikilu K, Malonga M, LAV/HTLV-III seroprevalence among tuberculosis patients in Zaire [communication 105] In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  33. Kenyatta J. Facing Mount Kenya. New York: Vintage, 1962
  34. Landesman SH, Ginzburg HM, Weiss SH. The AIDS epidemic. N Engl J Med 1985;312:521-5
  35. Clumeck N, Vandeperr P, Carael M, Rouvroy D, Nzaramba D. Heterosexual promiscuity amoung African patients with AIDS [letter] N Engl J Med 1985;313;182
  36. Bailey RC. The sociology of Efe Pygmy men in the Ituri Forest, Zaire. Doctoral [PhD] thesis. Cambridge, Mass: Harvard University, 1985
  37. Hartung J. Matrilineal inheritance: new theory and analysis. Behavioral and Brain Sciences 1985;8;661-70
  38. Belsey MA. The epidemiology of infertility: a review with particular reference to sub-Saharan Africa. Bull WHO 1976;54:319-41
  39. Westrom L. Effect of acute pelvic inflammatory disease on fertility. Am J Obstet Gynecol 1975;121:707-13
  40. D'Costa LJ, Plummer FA, Bowner I, Fransen L, Piot P, Ronald AR, Nsanze H. Prostitutes are a major reservoir of sexually transmitted disease in Nairobi, Kenya. Sex Trans Dis 1985;12;64-7
  41. Vandeperre P, Clumeck N, Carael M, Nzabihirmann E, Robert-Guroff M, De Moi P, Freyens P, Butzler J-B. Female prostitutes: a risk factor for infection with human T-cell lymphotrophic virus type III. Lancet 1985;2:524-6
  42. Schuster I. The new women of Lusaka. Palo Alto, calif: Mayfield, 1979
  43. Norman C. Politics and science clash on African AIDS. Science 1985;230:1140-2
  44. Murdock GP. Africa: its peoples and their cultural history. New York: McGraw-Hill, 1959
  45. Piot P, Quinn TC, Taelman H, Feinsod FM, Minlangu FB, Mbendi N, Wobin O, Mazebo P, Ndangi K, Stevens W, Kalambayi K, Mitchell S, Bridts C, McCormack JB. Acquired immunodeficiency syndrome in a heterosexual population in Zaire. Lancet 1984;2:527-9
  46. Carswell JW, Sewankambo N, Lloyd G, Downing RG. How long has the AIDS virus been in Uganda? [letter] Lancet 1986;1:1217
  47. Wofsy CB, Cohen JB, Hauer LB, Padian NS, Michaelis BA, Evans LA, Levy JA. Isolation of AIDS-associated retrovirus from genital secretions of women with antibodies to the virus. Lancet 1986;1:527-9
  48. Arya OP, Taber SR. Correlates of venereal diseases and fertility in rurnal Uganda. WHO/VDT/RES/75.339, WHO/VDT/RES/Gon/75.96 (unpublished documents cited in [38]). Geneva: World Health Organization, 1975
  49. Texton RB. A cross-cultural summary. New Haven, Conn: HRAF Press, 1967
  50. Winter EH. Beyond the mountains of the moon. Urbana, Ill: University of Illinois Press, 1965
  51. Miller C. The lunatic express. New York: Macmillian, 1971
  52. Morrelli G, Will S, Zwahlen D. Perinatal practices among the Efe (Pygmies) of Zaire. In: Proceedings of the International Conference on Infancy Studies, New York, 1984
  53. Bernardi B. The age-system of the Nilo-Hamitic peoples. Africa 1952;22:316-32
  54. Leakey L. Some notes on the Masai. Jeournal of the Royal Anthropological Institute of Great Britain and Ireland 1930;60:185-209
  55. Morfeldt-Manson L, Lindquist L. Blood brotherhood: a risk factor for AIDS? [letter] Lancet 1984;2:1346
  56. Carswell JW. Injection absesses. Ugandan Medical Journal 1983;5:16-20
  57. Koenig RE, Gautier T, Levy JA. Unusual introfamilial transmission of human immunodeficiency virus [letter]. Lancet 1986;2:627
  58. Hirsh MS, Wormser GP, Schooley RT, Ho DD, Felserstein D, Hopkins CC, Joline RN, Duncanson F, Sarngadharan MG, Saxinger C, Gallo RC. Risk of nosocomial infection with human T-cell lymphotrophic virus III (HTML-III). N Engl J Med 1985;312:1-4
  59. Weiss SH, Saxinger CW, Rechtman D, Grieco MD, Nadler J, Holman S, Ginzburg HM, Groopman JE, Goedert JJ, Markham PD, Gallo RC, Blattner WA, Landesman S. HTLV-III infection among health care workers: association with needle stick injuries. JAMA 1985;254:2089-93
  60. Martin LS, McDougal JS, Loskoski SL. Dinfection and inactivation of the human T lymphotrophic virus type III/lymphadenopathy-associated virus. J Infect Dis 1985;152:400-3
  61. Resnick L, Veren K, Salahuddin SZ, Tondreau S, Markham PD. Stability and inactivation of HTLV-III/lav under clinical and laboratory environments. JAMA 1986;255:1887-91
  62. Lepage P, Vandeperre P, Carael M, Butzler JP. Are medical injections a risk factor for HIV infection in children? [letter] Lancet 1986;2:1103-4
  63. Melbye M, Njelesani EK, Bayley A, Mukelabai K, Manuwele JK, Bowa FJ, Clayden SA, Levin A, Blattner WA, Weiss RA, Tedder R, Biggar RJ. Evidence for heterosexual transmission and clinical manifestations of human immunodeficiency virus infection and related conditions in Lusaka, Zambia. Lancet 1986;2:1113-5
  64. Mann JM. Epidemiology of LAV/HTLV-III in Africa [communication 49] In: Program and abstracts of the 2nd International Conference on AIDS, Paris, 1986
  65. Mann JM, Francis H, Davachi F, Baudoux P, Quinn TC, Nzilambi N, Bosenge N, Colebunders RL, Piot P, Kabote N, Asila PK, Malonga M, Curran JW. Risk factors for human immunodeficency virus seropositivity among children 1-24 months old in Kinshasha, Zaire. Lancet 1986;2:654-6
  66. Desrosiers RC. Origin of the human AIDS virus [letter]. Nature 1986;319:728
  67. Kanki PJ, Alroy J, Essex M. [Isolation of T-lymphotrophic retrovirus related to HTLV-III/LAV from wild-caught African green monkeys. Science 1985;230:951-4
  68. Wolfheim JH. Primates of the world: distribution, abundance and conservation. Seattle, Wash: University of Washington Press, 1983
  69. Andelman S. Concealed ovulation and continuous receptivity among vervet monkeys. American Naturalist 1987;129:785-99
  70. Haddow AJ. Field and laboratory studies on an African monkey. Cercopithecus ascanius schmidti matchie. Proceedings of the Zoological Society of London 1952;122:297-394
  71. Lyons SF, Jupp PB, Schoub BD. Survival of HIV in the common bedbug [letter]. Lancet 1986;2:45
  72. Hrdy S, Whitten P. The patterning of sexual activity among primates. In: Smuts B, Cheney D, Seyfarth R, Wranghan R, Struhsaker D, eds. Primate societies. Chicago: University of Chicago Press, 1987

(File revised 7 March 2006)

Return to CIRP library