In the first round of investigation, none of the 1,041 specimens taken in 1981 had HIV. Four infected specimens were taken in 1982, one in 1983, and six in 1984, making a total of eleven in 12,979 specimens, or less than 0.1 per cent. Four were men and seven women. Eight were recent arrivals in the district: four from other parts of Malawi (including the main city, Blantyre), two from Tanzania, and two from Zambia. Not only was the disease brought from several outside sources almost simultaneously, but several different subtypes were introduced. The two arrivals from Tanzania brought subtypes A and D, the two forms dominant in East Africa. Of the other nine specimens from this period, six were later identified as subtype C, while the other three could not be positively identified but were closest to subtype C and possibly an extinct variety of it.
Map 3 Southern Africa
Of the six individuals definitely identified with subtype C, one came from Zambia, two had been born in Zambia but had lived in Blantyre, two had come from elsewhere in Malawi, and one was a long-term resident of Karonga. Subtype C was to dominate the southern African epidemic, causing some 94 per cent of infections there in 2001.2 It may have originated in the southern DR Congo,3 which had many links with neighbouring Zambia, especially through the mining towns of Katanga and the Zambian Copperbelt. A possible reconstruction, compatible with evidence of early infection elsewhere in Malawi and Zambia that will be quoted later, is that elements of the East African epidemic (subtypes A and D) spread across the border into rural Karonga, but that the bulk of infection (subtype C) was carried from the southern DR Congo into Zambia, probably first to the Copperbelt, spread to other urban centres (including Blantyre) by 1983, and was carried from these centres into Karonga. Something can even be known of the process of infection. Of the six specimens with subtype C, four were so closely related genetically as to form a single cluster (cluster 1) with a single origin. One of the four was the long-term resident of Karonga. The other three had come from other parts of Malawi. The most likely scenario is that one person introduced the strain from elsewhere in Malawi and infected the other three after arriving in Karonga, although this cannot be certain.
Cluster 1 becomes central when attention shifts to the second round of blood collection in 1987–9. This revealed not 11 but 189 HIV-positive specimens, a prevalence of 2 per cent. Of the 168 specimens that could be analysed by subtype, 152 (90 per cent) belonged to subtype C, 6 to D, 3 to A, 3 were unclassified, and 4 were recombinants. Not only had subtype C established itself as the dominant form, but so had cluster 1: 40 per cent of those with subtype C (61 people) were infected with variants of that strain, probably introduced no more than five or ten years earlier by a single individual. Nothing could illustrate more vividly the explosive potential of a virus whose existence in their bodies was almost certainly unknown to most of those harbouring it.
The data collected in 1987–9 reveal much more about HIV epidemiology in Karonga. A majority of those infected were women, with an especially rapid increase in the late 1980s among women aged between 15 and 24, whereas men with HIV were generally older. Some 87 of the 189 infected people had not been present in the district in 1981–4, divided between 48 returning absentees and 39 new immigrants. Clearly the epidemic was still driven chiefly by mobility beyond the district. Prevalence increased with years of schooling and was most common among traders, salaried employees, casual labourers, and generally those who were not peasant farmers. Those with the best and the worst housing had higher prevalence than those with houses of intermediate quality. Of eighteen couples in which both partners were infected, only twelve were infected with closely related viral strains. Most intriguing was the dominance of subtype C, for one unanswered question about the epidemic is whether this subtype, which by the 2000s was responsible for more than half the world’s HIV infections, had greater evolutionary fitness than other subtypes. Despite much research and several detailed differences in its mode of operation, no conclusive evidence of this had emerged by 2005, although one study had shown that viral concentrations were more than three times as high in the blood and semen of Malawian men, over 90 per cent of them with subtype C, than in Americans with subtype B.4
Although the data from Karonga are uniquely detailed, it was clearly not the first part of Malawi to experience HIV infection. Study of stored blood taken in southern Africa before 1974 has revealed no evidence of HIV, but the first 17 Aids cases were reported from Malawi’s health facilities in 1985, some with aggressive Kaposi’s sarcoma, and a year later nearly 4 per cent of Malawian mineworkers in South Africa were HIV-positive, the only national group from Central Africa significantly infected. Given the long incubation period before the appearance of symptomatic Aids, and given the wide extent of HIV infection evident by the mid 1980s, Malawi’s silent epidemic probably began before 1980, or only slightly after HIV can be discerned around Lake Victoria. Census data show that mortality in Malawi increased significantly between 1977 and 1987, but chiefly among children, who commonly died of Aids more quickly than adults.5
The virus may have reached Zambia slightly earlier than Malawi, although the evidence is indirect. In 1983 Anne Bayley, a surgeon in Lusaka, found herself treating unprecedented numbers of young adults afflicted with aggressive Kaposi’s sarcoma. When tested in 1984, 91 per cent of these were found HIV-positive. Bayley later thought that the first case might date back to 1980 and that HIV had probably reached Zambia in the mid 1970s, initially spreading slowly. She added – a conclusion presumably reached by retrospective testing – that in 1981 fewer than 1 per cent of women at Lusaka’s antenatal clinics were HIV-infected.6 Many early patients in Lusaka with Kaposi’s sarcoma had associations with the Copperbelt, where tuberculosis cases suddenly multiplied from 1984 and the first small HIV tests in the general population of mining communities in 1985 showed 13.5 per cent prevalence in males and 21 per cent in females. Of deaths from Aids reported from Zambia between March and July 1987, 46 per cent were from the Copperbelt and 18 per cent from Lusaka.7 Yet the situation in the capital was alarming enough, for tests there in 1985 showed that 8 per cent of pregnant women were infected. In February 1986 Aids patients were also dying in Livingstone on Zambia’s southern border with Zimbabwe.8
In reality, the silent epidemic had penetrated Zimbabwe some time before, although perhaps three or four years later than Zambia and Malawi as the virus was carried southwards. The first cases of Aids and aggressive Kaposi’s sarcoma were diagnosed in 1983. Alarm arose only when blood was first screened in 1985 and it was revealed that over 2 per cent of donors had HIV. Infection then concentrated in the northern city of Harare, with only 0.05 per cent of donors testing positive in Bulawayo, further south. Thereafter, however, expansion became general and rapid. At the district hospital at Hurungwe in Mashonaland West, the annual number of patients diagnosed with HIV rose between 1986 and 1988 from 16 to 292. In Manicaland province, on Zimbabwe’s eastern border, all districts recorded increased mortality from the late 1980s. By 1990 national antenatal prevalence was 12.9 per cent.9
Botswana was invaded next, slightly later than Zimbabwe and just as stealthily. Over 200 blood specimens collected in the north during 1984 showed no HIV. ‘It’s not a problem in Botswana,’ an official declared, ‘AIDS is primarily a disease of homosexuals and there is no homosexual in Botswana.’ The first case reported at that time was indeed a white homosexual.10 When the first Tswana tested positive a year later, the Minister of Health ‘allayed fears by mentioning that the modes of transmission of the disease means that it could not become a big epidemic’. Within another year, however, he was speaking of ‘a scourge that could decimate a large portion of the human race’ and recommending ‘a stable, faithful relationship with another uninfected person’. By then Botswana had 30 known HIV cases and feared that the real number might be more like 3,000. In 1990 tests showed that 5–7 per cent of blood donors in towns and 1–2 per cent in the countryside were infected.11
During the 1990s