The Heritage Foundation
and the "Risk-Aids Hypothesis"

The third crucial arena for dissenting views during this time period was a lengthy (8,900-word) cover story by Duesberg and Bryan J. Ellison published in the summer 1990 issue of Policy Review, a publication of the Heritage Foundation, the well-known, right-wing think tank.^[99] The essay was actually written by Ellison, a politically conservative graduate student in Duesberg's department and self-appointed popularizer of Duesberg's views. "Scientists weren't going to listen to him. They couldn't afford to," Ellison explained. "So I realized he had to take his case to the general public."^[100] The article became, in the editor's words, "one of the three or four most-talked-about articles in the history of the magazine, … [eliciting] more letters to the editor than any in Policy Review 's history."^[101] The article also incorporated the first formal presentation by Duesberg of an alternative explanation for the etiology of AIDS, which Duesberg and Ellison dubbed the "risk-AIDS hypothesis."

The biographical note explained that Duesberg had published critiques of the accepted "virus-AIDS hypothesis" in a number of scientific journals, such as Cancer Research and the Proceedings of the National Academy of Sciences . In this way, Duesberg's accumulated scientific credibility was now converted into credibility in a different, more public forum. However, a chief strategy of the Policy Review article was to present the critique not as Duesberg's personal crusade, but as the clamor of a growing chorus, within which Duesberg was just one voice. The article therefore attributed dissenting views whenever possible to people like Walter Gilbert and Harvey Bialy. More generally, it described "an increasing number of medical scientists and physicians [who] have been questioning whether HIV actually does cause AIDS"; the article linked together those who said HIV could not play a role, those who said HIV had not conclusively been proven to play a role, and those who argued for cofactors. The reader might never have heard of this expanding group, the authors explained, because "most of these doubters prefer not to be quoted, out of fear of losing research funding or of disapproval by peers." Skepticism therefore remained a minority position "due largely to inadequate attention provided by media sources."^[102]

In the article, Duesberg and Ellison reiterated their standard arguments but also presented in expanded form a criticism that Duesberg

had not previously discussed in print in great detail: they argued that the notion that HIV caused AIDS was based fundamentally on a tautology. According to the CDC's 1987 update of its surveillance definition, AIDS was (usually) diagnosed by a positive HIV antibody test, in the presence of one or more diseases from a list: "The disease-list includes not only Kaposi's sarcoma and P. carinii pneumonia, but also tuberculosis, cytomegalovirus, herpes, diarrhea, candidiasis, lymphoma, dementia, and many other diseases. If any of these very different diseases is found alone, it is likely to be diagnosed under its classical name. If the same condition is found alongside antibodies against HIV, it is called AIDS. The correlation between AIDS and HIV is thus an artifact of the definition itself."^[103] Perfectly ordinary illnesses got stuck with the label "AIDS" if the ill person happened to be HIV positive; then researchers would turn around and say that, since everyone with AIDS was HIV positive, HIV must be the cause. This was an interesting argument, one which threw into question not only the logic of the causal claim but also the very status of "AIDS" as a legitimate disease category. It was an argument that Duesberg would often repeat in subsequent years; but it was somewhat disingenuous as posed.

Although the CDC's 1987 definition listed a number of diseases that, in an HIV-infected person, would result in an AIDS diagnosis,^[104] many of them—like Pneumocystis carinii pneumonia, toxoplasmosis, and cryptosporidiosis, diseases typical of AIDS patients—were relatively rare in general. Others, like CMV, herpes, and candidiasis, were indeed common, but in these cases the CDC's specifications went further, requiring that the conditions be present in parts of the body where these infections normally did not take root. Similarly, tuberculosis was on the list—but only if it involved at least one site other than the lungs. Diarrhea, of course, was not on the list; Duesberg and Ellison were referring casually to what the CDC called the "HIV wasting syndrome," defined as "profound involuntary weight loss > 10% of baseline body weight plus either chronic diarrhea (at least two loose stools per day for > 30 days) or chronic weakness and documented fever (for > 30 days, intermittent or constant) in the absence of a concurrent illness or condition other than HIV infection that could explain the findings. …"^[105] The diagnostic definition of AIDS-related dementia was similarly restrictive. Overall, clinical markers of AIDS were rare diseases and conditions generally not seen in people who were not HIV positive. By failing to explain these details of the CDC's diagnostic algorithm and by suggesting that ordinary diarrhea and tuberculosis

were being taken as markers of AIDS, Duesberg and Ellison were misleading their lay audience.

But at the same time, Duesberg and Ellison presented arguments to counter the assumption that the rare AIDS diseases, like Kaposi's sarcoma and PCP, were in fact so rare. They maintained that "not only have all 25 of these AIDS conditions existed for decades at a low level in the population, but HIV-free instances of the same diseases are still being diagnosed."^[106] They also described a recent letter to Lancet by Robert Root-Bernstein, an associate professor of physiology at Michigan State University and recipient of a MacArthur fellowship—one of the so-called "genius grants" provided, no-strings-attached, to individuals in a variety of fields who have been deemed unusually promising. Root-Bernstein's review of the medical literature had led him to conclude that perhaps 15 to 20 percent of all Kaposi's sarcoma cases before 1979 fit the pattern generally believed to have arisen only with the AIDS epidemic: young victims with a short survival time. Citing Sonnabend, Root-Bernstein had written: "Several hypotheses must be entertained—that AIDS is not new; that HIV is only one of several possible causes of AIDS; or that HIV is itself a new, opportunistic infection that takes advantage of previously immunosuppressed individuals."^[107]

The existence of "AIDS" diseases in people who are not antibody positive, in Duesberg and Ellison's view, was evidence for their alternative hypothesis, the risk-AIDS hypothesis. They proposed "that the AIDS diseases are entirely separate conditions caused by a variety of factors, most of which have in common only that they involve risk behavior." But like Sonnabend and others who had trod this path before them, the authors recognized that "a risk hypothesis must explain the recent increases in the various AIDS diseases, and why these have all been concentrated in particular risk groups." So Duesberg and Ellison put forward a potpourri of potential causes of the AIDS marker illnesses, linking Kaposi's sarcoma with popper use by gay men; AIDS dementia with psychoactive drugs and syphilis; and the wasting syndrome, "found most heavily in African AIDS patients," with "the extremes of malnutrition and the lack of sanitation on most of that continent," compounded in recent years by "wars and totalitarian regimes."^[108]

Many of these arguments were widely familiar from debates early in the epidemic: the claims about African health conditions, for example, mirror Sonnabend's speculations about Haitians in 1983. And indeed, to explain the systemic failure of immune response that is

characteristic of AIDS, Duesberg and Ellison's article explicitly endorsed the immune overload hypothesis,^[109] incorporating it within their risk-AIDS hypothesis: "Joseph Sonnabend, a New York physician who founded the journal AIDS Research in 1983, has pointed out that repeated, constant infections may eventually overload the immune system, causing its failure; still worse are simultaneous infections by two or more diseases."^[110] Duesberg and Ellison also pointed to heavy drug use as a major cause of immunosuppression. They claimed that abuse of alcohol, heroin, cocaine, marijuana, Valium, and amphetamines "can all be found as part of the life histories of many AIDS patients"; "when combined with regular and prolonged malnutrition, as is done with many active homosexuals^[111] and with heroin addicts, this can lead to complete immune collapse." To round out the picture, the authors noted the long-term immunosuppressive effects of antibiotics and claim that "active homosexuals … often [take] large amounts of tetracycline and other antibiotics each evening before entering the bath houses."^[112]

Duesberg and Ellison didn't provide any sources for their ethnographic data, and in interviews both of them acknowledged having little direct knowledge of gay life despite its vibrant expression in San Francisco, only miles from the Berkeley campus. In part the authors were drawing on early medical claims about "how the gay lifestyle" was related to the epidemic of immune suppression, which in turn borrowed from earlier and contemporary medical literature on gay men who attended clinics for treatment of sexually transmitted diseases (see chapter 1). Communication with John Lauritsen may also have played its part in shaping their biased understandings of gay male behavior. In a letter to Duesberg written just a few months earlier, Lauritsen had characterized the Mineshaft, the Saint, and St. Mark's Baths—the most prominent New York City venues for uninhibited gay male sex in the years before the epidemic—as "hell-holes which were the arenas for truly psychopathic drug abuse as obligatory tribal ritual."^[113]

Gay men and injection drug users had always been the focus of immune overload theories. But no one promoting such a perspective in 1990 could avoid discussion of the other "risk groups," and Duesberg and Ellison understood this. They explained (again echoing Sonnabend's claims from seven years earlier) that blood transfusion recipients were at risk of developing immunodeficiency because of pathogens present in transfused blood. Moreover, people receiving blood

transfusions typically did so because they were already quite ill or had undergone surgery, and both the trauma of the surgical procedure and the anesthesia could have immunosuppressive effects. In fact, Duesberg and Ellison claimed, "with or without HIV infection, half of all [transfusion] recipients do not survive their first year after transfusion." Similarly they noted that "hemophilia has always been a fatal condition," and that the blood products received by its sufferers were immunosuppressive. Finally, cases of AIDS in infants could be traced to "combinations of most of the above risk factors"; 95 percent of these babies were born to mothers who either used drugs or were sex partners of drug users, or had received transfusions, or had hemophilia. According to Duesberg and Ellison: "The risk behavior of many of their mothers has reached these victims, but their conditions are renamed AIDS when in the presence of antibodies against HIV."^[114]

In the conclusion to the article, Duesberg and Ellison turned to the policy implications of their argument. "The most urgent of these," they said, concerned the widespread administration of AZT. This powerful drug worked by inhibiting the replication of the virus, but "by doing this the drug also kills all actively growing cells in the patient," including immune system cells. If the virus was harmless, as the authors maintained, then "inhibiting HIV would accomplish nothing, while AZT actually produces the very immune suppression it is supposed to prevent." AZT, by this view, was just another harmful drug—like heroin, cocaine, and poppers—that contributed to immune overload. Second, the risk-AIDS hypothesis called into question the existing AIDS education strategies. Condoms and sterile needles were fine if the goal was to prevent hepatitis and other infectious diseases. But the hazard of these programs, Duesberg and Ellison maintained, was that they lulled the practitioners of risk behaviors into a false sense of security. By failing to "[emphasize] the danger of the risk behavior itself—particularly drug-taking—[these programs] may inadvertently encourage spread of the disease."^[115]

The HIV hypothesis "has not yet saved a single life, despite federal spending of $3 billion per year," Duesberg and Ellison reminded their readers in closing.^[116] Instead of sinking more money down the same hole, the government should begin funding "studies on the causes of the separate AIDS-diseases and their appropriate therapies." The rest of the $3 billion "might then be saved and returned to the taxpayers," wrote the authors in a suggestion that presumably did not clash with the conservative agenda of Policy Review .

The next issue of the magazine was devoted to letters in response to Duesberg and Ellison—the total length of the letter section was over 13,000 words, one and a half times the length of the original article. Both the establishment and the dissenters were well represented. Howard Temin stressed the "tragic" pediatric evidence: in one study, fifteen of sixteen HIV-infected children of infected mothers had AIDS or pre-AIDS symptoms, while none of thirty-nine uninfected children of infected mothers showed signs of illness. Wrote Temin: "Duesberg and Ellison state that 'the risk behavior of many of their mothers has reached these victims.' It is clear that what reached the children was HIV."^[117] Warren Winkelstein, the Berkeley epidemiologist, wrote in with the most recent results from the ongoing San Francisco Men's Health Study. Out of 386 homosexual men who had been HIV positive when entering the study six years before, 140 (36 percent) had developed AIDS, and the majority of them had died. Forty homosexual men had become infected since entering the study, and 2 (5 percent) had developed AIDS. But of 370 homosexual men who had remained uninfected, none developed AIDS.^[118]

An interesting letter came from Michael Fumento, who had written a popular book called The Myth of Heterosexual AIDS . Each of them a controversial figure, Duesberg and Fumento shared the belief that AIDS was a "risk group disease" and not a threat to the general population. But they were on opposite sides when it came to the etiological debate. Noting that his "initial reaction to anyone challenging the AIDS industry in any way is favorable," Fumento continued: "but in the case of Peter Duesberg and his co-author Bryan Ellison, I really must demur." After raising objections to Duesberg and Ellison's arguments, Fumento threw down the gauntlet: "What I would suggest, in perfect seriousness, is that before the authors write another article suggesting that it is perfectly okay for HIV-infected persons to have unprotected sex with uninfected persons or vice-versa, that they, in a public forum, inject themselves with HIV. Apparently Duesberg has hinted he may do it; I think he should go beyond that. Readers have a right to know just how much faith the authors have in their own theory."^[119]

Duesberg and Ellison were given the last word, and they had plenty to say.^[120] They began by expressing their pleasure that "the debate that should have occurred … years ago" was finally taking place. They then launched into a critique of the cohort studies that Winkelstein

and other letter writers had cited as definitive. The existing studies proved nothing, in Duesberg and Ellison's view, because they failed to demonstrate that illness was the consequence of HIV, not risk behavior. A controlled study actually designed to distinguish between the two causal hypotheses would be set up quite differently. It would compare two large groups of people, HIV positives and HIV negatives. But the two groups would be carefully matched for "every health risk that might possibly be involved in the various AIDS diseases."

Duesberg and Ellison also responded to various arguments that letter writers had raised—about babies with AIDS, about wives of hemophiliacs, about needle-stick injuries. Only "media sensationalism," they argued, could convince people that wives of hemophiliacs were at great risk of AIDS if they had no other risk factors. Those cases that had occurred were quite explainable: "Since AIDS is merely, by definition, a list of old diseases that are renamed when in the presence of antibodies against HIV, one should not be surprised to find an occasional such wife who happens to contract HIV and, coincidentally, one of the many diseases on the AIDS list." A controlled study, they believed, would show that HIV-positive wives developed AIDS indicator diseases at the same rate as HIV-negative wives.

The real problem, in Duesberg and Ellison's view, was that the established AIDS researchers abandoned scientific principles when it suited their interests. Instead of controlled studies, these researchers invoked anecdotal evidence. When Koch's postulates failed them, they "casually try to abandon those timetested, commonsensical" rules of scientific method. And "when all else fails," they started "changing the rules," "rather than bringing the hypothesis into question," as real scientists were supposed to do. To explain why so few antibody-positive people had AIDS, "a latent period first had to be invented, then extended to its present, and still growing, total of 10 to 11 years." Duesberg and Ellison concluded by declaring themselves "quite willing to carry out the Fumento test." But their degree of interest in doing so depended on the attention it could attract to their cause: "If he will arrange for sufficient national publicity, if he would be convinced by our action, and if he will thereafter help us bring exposure to our viewpoint, we will indeed be quite happy to have ourselves publicly injected with HIV. Perhaps Fumento will also be willing to check on our health status in the year 2000, or after whatever additional time is eventually added to the virus' latent period."^[121]