Subsequent review of the medical literature has revealed a plethora of inexplicable AIDS-like illnesses dating back to the late 1940s in the United States and Europe, although many of these scattered cases appear to be in people without classic HIV risk factors and could also be due to rare deficits in the immune system.

Pre-1981 cases fall into two categories: those for which stored blood samples have revealed HIV infection, and those for which no stored blood samples exist, but where the pattern of symptoms reported is highly suggestive of AIDS.

The oldest AIDS case for which an HIV diagnosis has been confirmed is that of Robert R, an American youth from St Louis who died in 1969. Robert’s case is mysterious as he appears to have had no connection with Africa, and his sexual history is murky.

The case of David Carr, a Manchester sailor who is believed to have died of AIDS in 1959 has recently been called into doubt, following the discovery that the strain of HIV isolated from stored tissue samples was virtually identical to strains prevalent 30 years later. US researchers have argued that such a similarity is implausible unless the virus was a laboratory contaminant, because HIV would normally mutate considerably during a 30 year period.

Other than Robert R, the earliest proven AIDS case involved Arvid Noe, a Norwegian sailor who had travelled in Africa and then worked as a lorry driver in Europe. Not only he but also his wife and child died in 1976.

A Portuguese man who fell ill in 1978 has been retrospectively diagnosed as one of the earliest cases of AIDS caused by HIV-2.

The identified cases in which HIV was present are important because they challenge the belief that HIV is a virus which was introduced into the West in the 1970s. Although some of these cases indicate contact with Africa, the 1969 case of an American youth shows no evidence of an African connection.

Clinical records have also been used to identify pre-1981 cases. Historians have looked for unusual cases recorded in the medical literature which appear to fit the existing definition of AIDS. A surprising number of such reports exist, dating back to at least 1940 in North America and Europe. There are numerous reports of an AIDS-like syndrome which occurred in newborn children in Germany, Poland and Scandinavia, starting in 1939 and persisting until the late 1950s. The syndrome was characterised by Pneumocystis pneumonia (PCP), cytomegalovirus disease and hypergammaglobulinaemia (very high levels of antibodies), all characteristic of AIDS in newborn children today.

The Dutch researcher Jaap Goudsmit speculates that this eruption of an AIDS-like syndrome was connected to the re-use of needles in hospitals, a common practice before the development of disposable syringes in the 1960s. He also argues that it may have originated from a strain of HIV which was less harmful to adults than the forms of HIV now in circulation, and that it was introduced to Europe by German colonists returning from Cameroon prior to the outbreak of war. However, other researchers have long argued that these cases were due to malnutrition-induced immune deficiency, a point not addressed by Goudsmit. Edward Hooper's investigations suggest that all the outbreaks occurred in regions where uranium was being mined.

There have always been rare cases of immune system collapse in humans that aren’t due to inherited immune deficiency, and cases still turn up in the literature of individuals with an AIDS-like picture of decimated T-cells but who lack risk factors and don’t test positive for HIV antibodies. This syndrome has received the name ICL (idiopathic CD4+ lymphocytopenia), meaning ‘deficit of CD4 cells of unknown origin’, and while cases are extremely rare compared with cases of AIDS, many of these early reports could be ICL, singled out for report by doctors for case report precisely because they were unusual.

Similar detailed reports do not exist for Africa, except for data concerning Kaposi's sarcoma, which became epidemic in equatorial Africa in the late 1950s. This form of KS, which took the lethal form only in about 10% of cases, was not associated with HIV or with immune suppression, and affected Africans but not Westerners resident in Africa.

A dozen or so cases of AIDS-like illness have been identified by Edward Hooper in Kinshasa, Rwanda and parts of the Congo, the earliest properly documented case having been diagnosed in 1962. African doctors tend to agree that AIDS did not appear in Africa on a growing scale before the late 1970s, and that it became epidemic only in the early 1980s. In the early years of the epidemic the syndrome was known as `Slim' in Uganda and other central African countries; it first appeared in Uganda on the north shores of Lake Victoria, in Burundi and Rwanda (states to the west of Lake Victoria), and in Kinshasa (the capital of Zaire), which lies at the crossroads of trade routes linking East Africa, West Africa, Angola and Zaire.

Isolated cases of AIDS-defining illnesses began to appear among gay men and injecting drug users from the early 1970s in New York and San Francisco. These were not linked by doctors until the beginning of 1981, when a cluster of orders for the drug pentamidine, used to treat PCP, sparked curiosity amongst officials at the Centers for Disease Control. Although the cluster of PCP cases was identified in 1981, it is possible that a low level of `pneumonia' cases were treated with standard antibiotics and hence went undetected by the Centers for Disease Control for some years before 1981.

Some researchers argue that AIDS cases could have occurred at a low level in the population before the 1940s without exciting much suspicion, because of a greater frequency of infectious diseases. It was only when infectious disease became less common that immune deficiency became more remarkable and worthy of note. Prior to the introduction of antibiotics, tuberculosis and syphilis may have masked minor clusters of HIV disease, in the view of medical historian Mirko Grmek.

However, Edward Hooper, author of The River, argues that many of these cases are misattributed, and his detailed investigations point to other causes in almost every documented case. If these explanations are correct in each case, this would place the emergence of AIDS no earlier than the late 1960s, Hooper argues.