Intended for healthcare professionals

Education And Debate

Sailors and star-bursts, and the arrival of HIV

BMJ 1997; 315 doi: (Published 20 December 1997) Cite this as: BMJ 1997;315:1689
  1. Edward Hooper, writer and medical researchera
  1. a PO Box 4087, Worthing BN14 7LQ


Tracking the origins and early history of a newly recognised disease is more than just an academic exercise. To appreciate how a disease began can help medical science to combat it. The classic example is John Snow's investigation of the cholera epidemic in Golden Square, London, in 1854: his removal of the handle of the Broad Street pump contained the outbreak.1 An appreciation of causation may also help to prevent similar events occurring in the future. The recent evidence, for example, about the origins of new variant Creutzfeldt-Jakob disease2 3 will, hopefully, sensitise those research scientists who are transplanting baboon livers in humans to the potentially catastrophic impact of zoonoses—human diseases acquired from animals.4

Three outbreaks of AIDS…

In the case of AIDS, three related but distinct causes have been recognised in the past 16 years—namely the three human immunodeficiency viruses (HIV-2 and HIV-1 groups M and O). It is now widely accepted that HIV-2 is the result of a zoonotic transfer of a simian immunodeficiency virus from the sooty mangabey (a species of African monkey). HIV-1 groups M (for “main”) and O (“outlier”) seem to result from two separate zoonotic transfers of different variants of simian immunodeficiency virus in chimpanzees.

Summary points

Learning about the origins of a disease may help us to control it and also to prevent similar diseases arising in the future

The earliest confirmed case of AIDS in the world was in a young Norwegian sailor who was infected with HIV-1 group O—probably in Cameroon in 1961-2

The earliest evidence of HIV-1 group M is from 1959 and of HIV-2 from 1965. It seems, therefore, that all three HIVs may have emerged around the same time

Phylogenetic evidence shows that HIV-1 groups M and O show a “star-burst” phylogeny, with different subtypes suddenly emerging around 1959

Opinion is divided about whether this star-burst arose from the natural transfer of simian immunodeficiency virus to humans or from iatrogenic introduction—for example, through a vaccine

HIV-1 group M has probably caused over 99% of the world's 12.9 million cumulative AIDS cases to date5; by contrast, group O has probably caused less than 0.1%, perhaps because the virus (like HIV-2) is less transmissible. None the less, the rarer HIV-1 may also have lessons to teach us.

Two mariners…

Earlier this year, characterisation by polymerase chain reaction sequencing of an archival HIV-1 isolate from a 29 year old Norwegian former merchant seaman showed that he had been infected with a group O virus.6 With this announcement, another piece of the jigsaw of the early history of the HIVs has slotted into place.

The Norwegian sailor died of AIDS in 1976, at the age of 29, as did his wife and youngest daughter, born in 1967. Since the debunking of the case of the sailor from Manchester who died in 1959 with symptoms of immunosuppression (but not, it would appear, HIV infection),7 8 the members of this Norwegian family now represent the earliest confirmed cases of AIDS. The first symptoms appeared in 1966 in the sailor, in 1967 in his wife, and in 1969 in their daughter.

The great majority of group O isolates come from people originating from west central Africa, and in particular Cameroon and Gabon.9 The central and coastal provinces of Cameroon (containing, respectively, the capital, Yaoundé, and the main port and commercial centre, Douala), have the highest current prevalence of group O, which causes just over 5% of all HIV infections in these two regions.10

The Norwegian sailor's maritime history is interesting. Between 1961 and 1965 he travelled the world's oceans, calling at ports in all six inhabited continents. On his first voyage, which began in August 1961 just after his 15th birthday, he worked as a kitchen hand on a Norwegian vessel that sailed down the west African coastline, calling at ports in Senegal, Guinea, Liberia, Côte d'Ivoire, Ghana, Nigeria, and Cameroon (almost certainly Douala). A gonorrhoeal infection during this trip shows that he was already sexually active. He returned home in May 1962, and apart from a two day stopover in Mombasa, Kenya, in 1964, he never returned to Africa (K F Wefring, personal communications, 1993, 1994, 1997). The sailor was, however, most unlikely to have been infected in Kenya, for only one group O isolate has been identified from that country—and that in 1995-6.11 No evidence exists to suggest that the sailor was bisexual, which means that sexual contact with a woman in Douala is the most straightforward explanation for his infection. This would suggest that HIV-1 group O has been circulating in that part of Africa for at least 35 years.

Between 1969 and 1973 or 1974, the Norwegian sailor experienced a remission of symptoms and was employed as a long distance lorry driver, ferrying goods to various destinations in Europe, including Germany, France, Belgium, Holland, Switzerland, Austria, and Italy. Seventy per cent of these journeys were to Germany, and his major pickup point for return cargoes was at Wesseling, 16 km south of Cologne. Former work colleagues believe that during the course of his travels he had sex with various women, including prostitutes. In this context it would be interesting to know whether the German musician who died in Cologne of clinically defined AIDS in 1979 (and whose first symptoms appeared in December 1976)12 was infected with HIV-1 group O. The musician was bisexual (not homosexual, as previously reported), and apparently used to hire female prostitutes to participate in orgies. Tissues from the postmortem examination are still in existence.


Route of the Norwegian sailor's first voyage, between Oslofjord (Norway) and Douala (Cameroon), August 1961 to May 1962

The next likely case of group O infection to feature in the literature is the second child of a French barmaid from Reims, who died in 1981 “following a clinical history highly suggestive of neonatal AIDS”; in 1992 a group O virus was isolated from the mother, who by then had AIDS.13 14 It may be significant that a major lorry route between Liège and Lyons—two of the Norwegian sailor's delivery stops—runs past Reims. Alternatively, since Reims is a garrison town, the woman's sexual partners might have included members of the French military who had served in Cameroon.15

The first group O isolate to be partially characterised and reported in the literature was the Cameroonian ANT-70, in 1990.16 Given the Norwegian sailor's sexual history, this long gap between earliest known infection and scientific recognition may seem surprising. There are, however, various possible explanations. One is that in 1961–2 the group O virus might have been new to Homo sapiens and not yet well adapted to transmission among humans. Another is that group O may be intrinsically less transmissible than group M, as suggested by the fact that the Norwegian sailor's first two daughters, born in 1964 and 1966, are both HIV negative. What this also shows, however, is that early, sporadic cases of any new disease tend to be missed.

Especially for a lentivirus like HIV, a considerable lag time may occur between the earliest known appearance in humans and its recognition as a cause of illness, which generally occurs when sufficient cases exist to establish a pattern. In the case of HIV-1 group M, we have reliable evidence that the seed was present in humans as early as 1959 in what is now Kinshasa, Congo,17 and yet the dreadful first flowering in American homosexual men only came to the notice of the medical profession in 1981.18 We now know that other cases of AIDS occurred in the Congo during the 1970s19 20 and possibly as early as 1962 (J Sonnet, personal communication, 1992),21 but the significance of such cases was recognised only retrospectively, once the syndrome and its viral cause had been identified. This highlights the worrying possibility that other HIV variants may already be spreading, unrecognised, in humans.

In 1994 Gerry Myers of the HIV sequence database reported that HIV-1 groups M and O both exhibit star-like phylogenetic trees, and proposed that the divergence of different subtypes within both these groups might have occurred around the year 1959.22 23 This is consistent with the molecular clocks proposed by many geneticists24 25 26 and with the fact that no HIV isolate has yet been discovered from before 1959. Others have referred to this divergence as a “big bang,”27 though perhaps the best image is that of a silent star-burst,28 viewed years later across space and time, perhaps by a sailor on night watch in the midst of a dark ocean.

…and a monkey puzzle tree

Speculation abounds about why the two explosive HIV-1 divergences should have suddenly occurred around the end of the 1950s. In fact, there might even have been three roughly contemporaneous star-bursts, as the earliest epidemiological evidence of all three HIVs pertains to the same time period—1959 for HIV-1 group M, 1961–2 for group O, and 1965–6 (in different parts of west Africa) for HIV-2.29 30

Proponents of the “natural transfer” school believe that simian viruses may have been transferred to humans during the skinning and butchery of chimpanzees and sooty mangabeys or the keeping of these primates as pets. They seek to explain the synchronicity of divergence and spread of the HIVs by proposing that urbanisation and new sexual freedoms around the time of decolonisation brought these rare human viruses in from the bush.

Others, members of the iatrogenic school, believe that the hand of medical science may have played an unintended role. They propose that the capture of monkeys and apes for scientific purposes, or the administration in Africa of vaccines made in substrates of primate kidney, may have been the initial means whereby the precursor simian viruses were transferred to humans.31 Many of them believe that the star-burst phenomenon is suggestive of several simultaneous iatrogenic transfers.

Perhaps when further isolates of simian immunodeficiency viruses from chimpanzees and sooty magabeys and archival HIV isolates have been sequenced, we shall have greater insight into the question of where, when, and how the HIVs came into being, and how best to minimise the risk of further zoonotic disasters in the new millennium.


Much of the information in this article is based on tape recordings and notes of interviews between EH and various scientists or on personal letters from those scientists. The map is based on an original drawing by Sally Griffin.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
View Abstract