The "Drug-Aids Hypothesis"
As the implications of Maddox's intervention were sorted out in various arenas, Duesberg scored another partial victory:
the publication in a professional journal of a formal statement of his own hypothesis as to the causes of AIDS. The article was published in Biomedicine & Pharmacotherapy after a yearlong, ultimately unsuccessful, campaign by Duesberg to publish once again in the far more prestigious Proceedings of the National Academy of Sciences . The Proceedings had published Duesberg's views on AIDS twice before but each time had sent the submissions out for peer review, breaking with its ordinarily relaxed procedures for academy members. Both times the reviews had been critical; both times editor Igor Dawid ultimately had relented. This time Dawid put his foot down, writing Duesberg in February 1991 that the article had been rejected. One referee had called the paper a "flight of ideas" and "grossly incomplete," while the second referee favored publication. The third referee acknowledged, "I am no expert in the fields concerned" and made a political argument: "In all likelihood the publication of this article in PNAS would be harmful to the reputation of the journal, and has a potential for being harmful to the HIV infected segment of the population."
What was all the fuss about? In "The Role of Drugs in the Origin of AIDS," Duesberg presented what could be considered the fourth version in a sequence of publicly expressed views about the etiology of AIDS. Back in 1987, Duesberg had started out saying he didn't know what caused AIDS; all he was certain of was that it wasn't HIV. "The charge was then leveled that I was destructive, I was only negative, I was not contributing anything," Duesberg later recalled. But very soon afterward, he began making statements consistent with the immune overload hypothesis. In 1990 his student Bryan Ellison's Policy Review article had expanded this hypothesis (as reworked by Root-Bernstein) and extended it to the other risk groups, formalizing it as the risk-AIDS hypothesis. But in the very course of discussions with Ellison as the article was being written, Duesberg had become increasingly uncomfortable with the generalized focus on any and all forms of risk behavior.
In place of the risk-AIDS hypothesis, Duesberg began to formulate a more parsimonious explanation, the "drug-AIDS hypothesis," which emphasized toxicological causes of AIDS. In a move that signaled his disagreement with some of the other HIV dissidents, such as Sonnabend and Callen, Duesberg dismissed the significance of repeated infections. AIDS was not caused by an infectious agent—not HIV, not CMV, not Epstein-Barr virus, not any combination of the above. Drug consumption—not promiscuous sex—was the lifestyle practice associated with AIDS, Duesberg increasingly became convinced.
Leaving aside the hemophilia and transfusion cases, which had their own explanations, nearly every case of AIDS in the United States and Europe could be attributed to drug abuse.
In his 1992 article in Biomedicine & Pharmacotherapy , Duesberg presented his case in more formal terms. The article included 132 references, mostly to scientific publications but also to popular works by Jad Adams and John Lauritsen. Ever since the turn of the century, Duesberg wrote, "evidence has accumulated that addiction to psychoactive drugs leads to immune suppression and clinical abnormalities similar to AIDS, including lymphopenia, lymphadenopathy, fever, weight loss, septicemia, and increased susceptibility to infections and neurological disorders." These clinical abnormalities became epidemic in the early 1980s as a result of "a massive escalation in the consumption of psychoactive drugs," Duesberg explained, with reference to Justice Department statistics. "Thus the American AIDS epidemic is a subset of the AIDS epidemic." Indeed, only half of the drug-induced immunodeficiency cases receive public notice, due to the hegemony of the HIV hypothesis: "Only the pneumonias, tuberculoses, and dementias of the 50% of American intravenous drug users with HIV are recorded as AIDS, while those of their HIV-negative counterparts are diagnosed by their old names."
How would Duesberg prove a claim that AIDS is caused by drugs? Koch's postulates were not relevant in this case, since they applied only to infectious agents. Duesberg was arguing that the relation between drugs and AIDS was analogous to that between smoking and lung cancer: prolonged and repeated exposure to the toxic substance or substances eventually resulted in disease in some percentage of cases. The problem is that causal relationships of this sort are notoriously difficult to establish, and epidemiologists devote considerable energy to teasing out the various lifestyle risk factors that might confound the relationship (Is the smoker also overweight? Does the smoker drink alcohol?). But Duesberg was not an epidemiologist and had conducted no controlled studies. Nor did his proven expertise in molecular biology or retrovirology have much bearing in this case. Duesberg simply set out to construct a persuasive argument, relying on his background in chemistry and facts at his disposal from his scouring of the published literature.
After quickly reviewing his case for the implausibility of the reigning hypothesis, Duesberg began by noting the "chronological coincidences" between the AIDS and drug epidemics in the United States. Moreover, "drugs and AIDS appear to claim their victims from the
same risk groups." Intravenous drug users comprised about a third of all AIDS patients, Duesberg explained. And about 60 percent of AIDS patients in the United States were male homosexuals, who, according to Duesberg, were disproportionate consumers of drugs. Duesberg reported on a number of studies, including a 1990 survey of "3,916 self-identified American homosexual men, the largest of its kind," which found that 83 percent had used one or more drugs—including poppers, cocaine, amphetamines, and LSD—during the previous six months. Finally, Duesberg turned to another "risk group," healthy antibody positives who had taken "cytocidal DNA chain terminators" such as AZT. "Thus an unknown, but possibly a high percentage of the 30,000 Americans that currently develop AIDS per year have used AZT prior to or after the onset of AIDS."
Even if accepted at face value, these arguments about prevalence of drug use among AIDS risk groups were not particularly weighty in establishing the role of drugs in the causation of AIDS. Indeed, in his own attacks on the HIV hypothesis, Duesberg had frequently invoked the maxim that "correlation is not causation": just because HIV, or drug use, or anything else had been correlated with AIDS, researchers could not necessarily conclude that they had identified a cause. However, AZT presented Duesberg with a particularly convenient target, because by 1992 no one liked the drug very much despite its widespread administration. AZT did not cure AIDS, and it had substantial and potentially dangerous side effects. The initial study that showed it prolonged life in AIDS patients had been ended early when, for ethical reasons, the drug was supplied to study participants getting only a placebo, and some argued that, as a result, there was no clear evidence of the drug's long-term effects. Since 1990 the drug had also been prescribed to asymptomatic HIV positives in hopes of preventing progression to AIDS. But recent studies had been equivocal, suggesting that while the drug might indeed delay the onset of opportunistic infections, it might have no ultimate effect on longevity. By this reading, HIV positives faced a Hobson's choice in the short term—refuse AZT and suffer minor opportunistic infections or take AZT and endure its adverse effects—but arrived at the same place in the end. Another often-criticized but much publicized study had suggested that AZT might be less effective in African-Americans and Latinos than in whites.
To be sure, most doctors continued to prescribe AZT (or chemically related drugs), and public health authorities continued to promote it
as the indicated treatment for HIV positives with abnormally low T-cell counts. But enthusiasm for the drug had waned appreciably. Over the years, particularly in New York, some in the AIDS movement had come out against AZT—including dissenters in the causation controversy, like Ortleb, Lauritsen, Sonnabend, and Callen, as well as some who accepted the HIV hypothesis. Callen—who was well known for the accomplishment of being alive a decade after his AIDS diagnosis—attributed his survival, in large part, to his refusal to take AZT.
In his critique, Duesberg emphasized the drug's side effects: anemia, nausea, muscle atrophy, hepatitis, insomnia, headaches, seizures, and vomiting, among others. Yet although none of these conditions would justify an AIDS diagnosis, Duesberg did not explain his claim (perhaps borrowed from the title of Lauritsen's book, AZT: Poison by Prescription ) that AZT is "AIDS by prescription." Certainly if Duesberg were held to the same rigorous standards of proof that he proposed for the HIV orthodoxy, his argument would have to be found wanting. He had provided no conclusive evidence isolating long-term drug use as the cause of AIDS; he could point to no controlled longitudinal studies of the kind he insisted that the AIDS establishment must perform. And along the way he presented a number of arguments that can only be characterized as specious: "Within 48 weeks on AZT, 172 (56%) out of 308 Australian AIDS patients developed one or more new AIDS diseases, including pneumonia and candidiasis. This indicates that AZT induces AIDS disease within less than 1 year and thus much faster than the 10 years HIV is said to need to cause AIDS." This was like arguing that if a flu sufferer took aspirin, and four hours later her fever returned, then aspirin must cause fever even more rapidly than the influenza virus. Perhaps comments such as these were intended only to goad his critics and were not meant to be taken too seriously. Or perhaps by this point, Duesberg was so embittered by the behavior of his scientific colleagues—who, he believed, had black-balled him, tried to silence him, and succeeded in cutting his funding—that he was willing to employ any rhetorical device at his disposal to cast doubt on the worth of their accomplishments.
By a different calculus, Duesberg might be said to have achieved his objectives with the article in Biomedicine & Pharmacotherapy: he had supplemented what, after all, was his main point—that HIV could not be the cause of AIDS—by proposing an alternative explanation in a legitimate scientific publication. In the past, experts had taunted Duesberg: "Perhaps he would be willing to tell us what, in his view, is
the cause of AIDS and what he would do about it. … It would seem only fair for Professor Duesberg either to come up with an equally strong candidate or to lend his support to eradicating HIV and thus AIDS." Now Duesberg could claim to have met that challenge, and he could return to exerting public pressure on the proponents of the AIDS orthodoxy to prove the official story of AIDS causation.