The polio vaccine AIDS hypothesis
The claim that HIV entered humans through a polio vaccine trialled in the Congo in the late 1950s. The vaccine was tested and was clean, and the virus is older than the trial.
Why this verdict
Debunked by testing and by dating: surviving samples of the vaccine contained no HIV, no SIV and no chimpanzee DNA, and phylogenetic analysis places the origin of the pandemic HIV lineage decades before the trials took place.
- Period
- Region
- Reading time
- Sources
01The claim
That the CHAT oral polio vaccine, administered to around a million people in the Belgian Congo, Rwanda and Burundi between 1957 and 1960, was prepared in chimpanzee kidney cells contaminated with simian immunodeficiency virus, and that this transmitted the virus that became HIV-1 — with the medical establishment resisting the hypothesis to avoid liability and damage to vaccination programmes.
02Where it came from
The idea was raised by Louis Pascal and then given wide circulation by Tom Curtis in Rolling Stone in 1992. Edward Hooper developed it at length in The River (1999), a long and seriously researched book that made the geographic coincidence its central argument: the early AIDS cases and the vaccination sites appeared to overlap. That coincidence was the hypothesis's strength and, in the end, its only substance.
03What the record shows
✕ DEBUNKED — Investigated and refuted. Here is what actually happened.
The hypothesis was taken seriously enough to be tested properly, which is the opposite of suppression. The Royal Society convened a meeting on it in September 2000, at which Hooper presented alongside his critics.
Three independent lines then closed it.
Surviving CHAT vaccine samples held at the Wistar Institute were tested by laboratories in three countries. They contained no HIV and no SIV. They also contained no chimpanzee DNA — the cells were macaque, which is what Hilary Koprowski and colleagues had said all along, and which removes the proposed transmission route entirely. The results were published in Nature and Science in 2001.
The molecular clock closed it independently. Phylogenetic analysis of HIV-1 group M, including work by Michael Worobey using a 1960 Kinshasa lymph node sample alongside a 1959 plasma sample, dates the common ancestor to roughly 1908–1933. The virus was already diversifying in humans in central Africa decades before the vaccine trials began. No vaccine administered in 1958 can be the origin of a lineage that existed in 1920.
The geographic correlation that motivated the hypothesis is explained by the same work: both the vaccine campaign and the early spread of HIV followed the population centres and transport routes of colonial central Africa, because that is where the people were.
The natural-transfer account — repeated cross-species transmission through the hunting and butchering of primates, of which SIV lineages show many independent instances — is supported by the same phylogenetics.
04What it cost
The hypothesis is the clearest case in this archive of a claim advanced in good faith, investigated properly, and refuted — and of the damage that can follow anyway. It entered anti-vaccination material and, in the 2000s, contributed to resistance to polio vaccination in parts of Africa and South Asia, where campaigns already faced claims that the vaccine caused sterility or HIV. Polio vaccinators have been murdered in several countries. Hooper has continued to defend the hypothesis; the vaccine testing and the molecular dating have not been answered.