Chapter 12

The "heroin overdose" mystery and other occupational hazards of addiction

Chapter 4 of this Report reviewed in detail the effects of heroin and other opiates on addicts, including deleterious physiological effects traceable to the drugs themselves. Narcotics addicts today face other physiological hazards that are traceable to the narcotics laws, to the adulteration, contamination, and exorbitant black-market prices that those laws foster, and to other legal and social (as distinct from pharmacological) factors. Dr. Jerome H. Jaffe has described some of these risks in Goodman and Gilman's textbook (1970): "Undoubtedly, the high cost and impurities of illicit drugs in the United States exact their toll. The high incidence of venereal disease reflects the occupational hazard of the many females who earn their drug money through prostitution. The average annual death rate among young, adult heroin addicts is several times higher than that for nonaddicts of similar age and ethnic backgrounds. . . . The suicide rate among adult addicts is likewise considerably higher than that of the general population." 1

Because "the preferred route of administration is intravenous," Dr. Jaffe continues, "there is sharing of implements of injection and a failure to employ hygienic technics, with a resultant high incidence of endocarditis, and hepatitis, and other infections." 2

The exorbitant price of black-market heroin, Dr. Jaffe might have added, is one of the factors that makes intravenous injection "the preferred route of administration," for "mainlining" is the cheapest way to forestall withdrawal symptoms. And the laws restricting possession of injection equipment, under penalty of imprisonment, increase the risk of needleborne infections by encouraging the sharing of implements.

There remains to be considered yet another risk of heroin addiction, the most publicized hazard of all-death from "heroin overdose." Because these deaths are a source of such widespread concern, and also because they are so widely misunderstood, even by authorities on heroin addiction and by addicts themselves, we shall examine the data in detail. Much of the discussion that follows is focused on New York City, since the deaths attributed to heroin overdose are most numerous there and since the New York City data are published in convenient form.

"Prior to 1943, there were relatively few deaths among addicts from overdosage. " 3 By the 1950s, however, nearly half of all deaths among New York City addicts were being attributed to "acute reaction to dosage or overdosage ." 4 In 1969, about 70 percent of all New York addict deaths were assigned the "overdose" labels and in 1970, the proportion was about 80 percent.- The number of deaths so designated by New York City's Office of the Chief Medical Examiner increased from very few or none at all before 1943 to about 800 in 1969 and 1970 .7

During this same twenty-eight-year span, addict deaths from all other causes-infections, violence, suicide, and so on-increased very little. The enormous increase in number of deaths among addicts shown in Figure 2 was attributed almost entirely to "overdose" deaths.

The number of deaths throughout the United States attributed to heroin overdose from 1943 to date must total many thousands. In New York City it was reported that narcotics, chiefly heroin, were the leading cause of death in 1969 and 1970 in all males aged fifteen to tbirty-five,8 including nonaddicts. This can properly be characterized as an epidemic; the general alarm over these deaths is thoroughly warranted.

There are two relatively simple ways, however, to prevent deaths from heroin overdose.

First, addicts can be warned to take only their usual dose of heroin rather than risking death by taking too much.

Second, even in cases where an addict takes a vastly excessive dose despite the warning, death usually can be readily prevented, for death from an overdose of opiates is ordinarily a slow process. "in cases of fatal poisoning with morphine, the time of death may vary roughly from one to twelve hours." 9 The first signs are lethargy and stupor, followed by prolonged coma. If, after a period of hours, death does ensue, it is usually from respiratory failure. During the minutes or hours following the injection of a potentially fatal overdose, death can be readily forestalled by administering an effective antidote: a narcotic antagonist known as nalorpbine (Nalline).10 Nalorphine brings a victim of opiate overdose out of his stupor or coma within a few minutes. Since there is plenty of time and since nalorphine is stocked in pharmacies and hospital emergency rooms throughout the country, the death of anyone due to heroin overdose is very rarely excusable.

But alas, the two standard precautions against overdose-warnings against taking too much and administration of an antidote-are in fact wholly ineffective in the current crisis, for the thousands of deaths attributed to heroin overdose are not in fact due to heroin overdose at all. The evidence falls under three major rubrics.

(1) The deaths cannot be due to overdose.

(2) There has never been any evidence that they are due to overdose.

(3) There has long been a plethora of evidence demonstrating that they are not due to overdose.

Let us review these three bodies of data in detail.

(1) Why these deaths cannot be due to overdose. The amount of morphine or heroin required to kill a human being who is not addicted to opiates remains in doubt but it is certainly many times the usual dose (10 milligrams) contained in a New York City bag. "There is little accurate information," Drs. A. J. Reynolds and Lowell 0. Randall report in Morphine and Allied Drugs (1967). "The figures that have been reported show wide variation." 12 This ignorance no doubt stems from the rarity of morphine or heroin overdose deaths. The amounts of morphine or heroin needed to kill a nonaddict have been variously estimated at 120 milligrams (oral),13 200 milligrams, 14 250 milligrams," and 350 milligrams16-though it has also been noted that nonaddicts have survived much larger doses .17 The best experimental evidence comes from Drs. Lawrence Kolb and A. G. Du Mez of the United States Public Health Service; in 1931 they demonstrated that it takes seven or eight milligrams of heroin per kilogram of body weight, injected directly into a vein, to kill unaddicted monkeys." On this basis, it would take 500 milligrams or more (50 New York City bags full, administered in a single injection) to kill an unaddicted human adult.

Virtually all of the victims whose deaths are falsely labeled as due to heroin overdose, moreover, are addicts who have already developed a tolerance for opiates-and even enormous amounts of morphine or heroin do not kill addicts. In the Philadelphia study of the 1920s, for example, sonic addicts reported using 28 grains (1,680 milligrams) of morphine or heroin per day."' This is forty times the usual New York City daily dose. In one Philadelphia experiment, 1,800 milligrams of morphine were injected into an addict over a two-and-a-half-hour period. This vast dose didn't even make him sick.20

Nor does a sudden increase in dosage produce significant side effects, much less death, among addicts. In the Philadelphia study, three addicts were given six, seven, and nine times their customary doses-"mainlined." Far from causing death, the drug "resulted in insignificant changes in the pulse and respiration rates, electrocardiogram, chemical studies of the blood, and the behavior of the addict." 21 The addicts didn't even become drowsy.22

Recent studies at the Rockefeller Hospital in New York City, under the direction of Dr. Vincent P. Dole, have confirmed the remarkable resistance of addicts to overdose. Addicts receiving daily maintenance doses of 40 milligrams to 80 milligrams of methadone, a synthetic narcotic (see Chapter 14), were given as much as 200 milligrams of unadulterated heroin in a single intravenous injection. They "bad no change in respiratory center or any other vital organs ." 23

(2) There is no evidence to show that deaths attributed to overdose are in fact so caused. Whenever someone takes a drug-whether strychnine, a barbiturate, heroin, or some other substance-and then dies without other apparent cause, the suspicion naturally arises that he may have taken too much of the drug and and died of poisoning an overdose. To confirm or refute this suspicion, an autopsy is performed, following a well-established series of procedures.

If the drug was taken by mouth, for example, the stomach contents and feces are analyzed in order to identify the drug and to determine whether an excessive amount is present. If the drug was injected, the tissues surrounding the injection site are similarly analyzed. The blood, urine, and other body fluids and tissues can also be analyzed and the quantity of drug present determined.

Circumstantial evidence, too, can in some cases establish with reasonable certainly that someone has died of overdose. If a Patient fills a prescription for a hundred barbiturate tablets, for example, and is found dead the next morning with only a few tablets left in the bottle, death from barbiturate poisoning is a reasonable hypothesis to be explored. Similarly, if an addict dies after "shooting up," and friends who were present report that he injected many times his usual dose, the possibility of death from heroin overdose deserves serious consideration.

Further, in cases where an addict has died following an injection of heroin, and the syringe he used is found nearby or still sticking in his vein, the contents of the syringe can be examined to determine whether it contained heroin of exceptional strength. And there are other ways of establishing at least a prima facie case for an overdose diagnosis.

A conscientious search of the United States medical literature throughout recent decades has failed to turn up a single scientific paper reporting that heroin overdose, as established by these or any other reasonable methods of determining overdose, is in fact a cause of death among American heroin addicts. The evidence that addicts have been dying by the hundreds of heroin overdose is simply nonexistent.

At this point the mystery deepens. If even enormous doses of heroin will not kill an addict, and if there exists no shred of evidence to indicate that addicts or nonaddicts are in fact dying of heroin overdose, why is the overdose myth almost universally accepted? The answer lies in the customs of the United States coroner-medical examiner system.

Whenever anyone dies without a physician in attendance to certify the cause of death, it is the duty of the local coroner or medical examiner to investigate, to have an autopsy performed if indicated, and then formally to determine and record the cause of death. The parents, spouse, or children of the dead person can then ask the coroner for his findings. Newspaper reporters similarly rely on the coroner or medical examiner to explain a newsworthy death. No coroner, of course, wants to be in a position of having to answer "I don't know" to such queries. A coroner is supposed to know-and if he doesn't know, he is supposed to find out.

At some point in the history of heroin addiction, probably in the early 1940s, the custom arose among coroners and medical examiners of labeling as "heroin overdose" all deaths among heroin addicts the true cause of which could not be determined. These "overdose" determinations rested on only two findings: (1) that the victim was a heroin addict who "shot up" prior to his death; and (2) that there was no evidence of suicide, violence, infection, or other natural cause.24 No evidence that the victim had taken a large dose was required to warrant a finding of death from overdose. This curious custom continues today. Thus, in common coroner and medical examiner parlance, "death from heroin overdose" is synonymous with "death from unknown causes after injecting heroin."

During the 1940s, this custom of convenience did little apparent harm. Most deaths among heroin addicts were due to tetanus, bacterial endocarditis, tuberculosis, and other infections, to violence, or to suicide, and they were properly labeled as such by coroners and medical examiners. It was only an occasional death which baffled the medical examiner, and which was therefore signed out as due to "overdose." But, beginning about 1943, a strange new kind of death began to make its appearance among heroin addicts .25 The cause of this new kind of death was not known, and remains unknown today-though it is now quite common.

A striking feature of this mysterious new mode of death is its suddenness. Instead of occurring after one or more hours of lethargy, stupor, and coma, as in true overdose cases, death occurs within a few minutes or less-perhaps only a few seconds after the drug is injected. Indeed,

11 collapse and death are so rapid," one authority reports, "that the syringe was found in the vein of the victim or on the floor after having dropped out of the vein, and the tourniquet was still in place on the arm ." 26 This explains in part why nalorphine and other narcotic antagonists, highly effective antidotes in true opiate overdose cases, are useless in the cases falsely labeled overdose.

An even more striking feature of these mysterious deaths is a sudden and massive flooding of the lungs with fluid: pulmonary edema. In many cases it is not even necessary to open the lungs or X-ray them to find the edema; "an abundance of partly dried frothy white edema fluid [is seen] oozing from the nostrils or mouth " 27 when the body is first found. Neither of these features suggests overdose-but since "overdose" has come to be a synonym for "cause unknown," and since the cause of these sudden deaths characterized by lung edema is unknown, they are lumped under the "overdose" rubric.

Not all of the deaths attributed to heroin overdose are necessarily characterized by suddenness and by massive pulmonary edema, but several studies have shown that a high proportion of all "overdose" deaths share these two characteristics.28

(3) Evidence demonstrating that these deaths are not due to overdose is plentiful. This evidence has been summarized in a series of scientific papers, beginning in 1966, by New York City's Chief Medical Examiner, Dr. Milton Helpern, and his associate, Deputy Chief Medical Examiner, Dr. Michael M. Baden. At a meeting of the Society for the Study of Addiction held in London in 1966, Dr. Helpern explained that the most conspicuous feature of so-called "overdose" deaths is the massive pulmonary edema. When asked the cause of the edema, he cautiously responded:

This is a very interesting question. To my knowledge it is not known why the pulmonary edema develops in these cases.... This reaction sometimes occurs with the intravenous injections of mixtures, which as far as is known, do not contain any heroin, but possibly some other substance. The reaction does not appear to be specific. It does not seem to be peculiar to one substance, but it is most commonly seen with mixtures in which heroin is the smallest component.

In a paper published in the New York State Journal of Medicine for September 15, 1966, Dr. Helpern again cast doubt on the myth that these deaths are due to overdose. "Formerly such acute deaths were attributed to overdose of the heroin contained in the sample injected," Dr. Helpern reported-but he went on to cite several lines of evidence arguing against the overdose theory:

... Unexpected acute deaths may occur in some addicts who inject themselves with heroin mixtures even though others who take the same usual . . . dose from the same sample at the same time may suffer no dangerous effect. In some fatal acute cases, the rapidity and type of reaction do not suggest overdose alone but rather an overwhelming shocklike process due to sensitivity to the injected material. The toxicologic examination of the tissues in such fatalities, where the reaction was so rapid that the syringe and needle were still in the vein of the victim when the body was found, demonstrated only the presence of alkaloid, not overdosage. In other acute deaths, in which the circumstances and autopsy findings were positive, the toxicologist could not even find any evidence of alkaloid in the tissues or body fluids. Thus, there does not appear to be any quantitative correlation between the acute fulminating lethal effect and the amount of heroin taken. . . .30

Dr. Helpern's associate, Deputy Chief Medical Examiner Baden, went on to further discredit the already implausible overdose theory at a joint meeting of two American Medical Association drug-dependency committees held in Palo Alto, California, in February 1969.

"The majority of deaths," Dr. Baden told the AMA physicians, "are due to an acute reaction to the intravenous injection of the heroin-quinine-sugar mixture. This type of death is often referred to as an 'overdose,' which is a misnomer. Death is not due to a pharmacological overdose in the vast majority of cases." 31

At the same AMA committee meeting and at a meeting of the Medical Society of the County of New York, Dr. Baden cited six separate lines of evidence overturning the "heroin overdose" theory.

First, when the packets of heroin found near the bodies of dead addicts are examined, they do not differ from ordinary packets. "No qualitative or quantitative differences" are found.32 This rules out the possibility that some incredibly stupid processor may have filled a bag with pure heroin instead of the usual adulterated mix.

Second, when the syringes used by addicts immediately before dying are examined, the mixture found in them does not contain more heroin than usual.

Third, when the urine of addicts allegedly dead of overdose is analyzed, there is no evidence of overdose.

Fourth, the tissues surrounding the site of the fatal injection show no signs of high heroin concentration.

Fifth, neophytes unaccustomed to heroin rather than addicts tolerant to opiates would be expected to be susceptible to death from overdose. But "almost all of those dying" of alleged overdose, Deputy Chief Medical Examiner Baden reported, "are long-term users."

Sixth, again according to Dr. Baden, "addicts often 'shoot' in a group, all using the same heroin supply, and rarely does more than one addict die at such a time." 33

These definitive refutations of the heroin overdose theory should, of course, have led to two prompt steps: a warning to addicts that something other than overdose is causing these hundreds of addict deaths annually -and an intensive search for the true cause of the deaths. But neither of these steps has been taken. Hence the news media go right on talking about "heroin overdose" deaths. "Death from acute reaction to heroin overdose" and other complicated phrases are also used; these phrases similarly conceal the fact that these deaths are not due to overdose.

How can the "heroin overdose" myth not only survive but flourish even after these repeated scientific debunkings? Two stenographic transcripts provide an answer.

The first is the transcript of a press conference held at the Rockefeller University on October 27, 1969, in connection with the Second National Conference on Methadone Treatment. In the course of his remarks to the assembled reporters, Deputy Chief Medical Examiner Baden there discussed at some length a case of what he described as an "addict who died of an overdose of heroin." 34 The reporters present naturally referred thereafter to this death as a "heroin overdose" case.

At the scientific meeting held in the same room on the same day, however, Dr. Baden described the same death in quite different terms. To the scientists he stated that the addict in question "died of acute reaction to injection of heroin, a so-called overdose." When even this description was challenged by a fellow physician, who pointed out that addicts don't die following even enormous doses,35 Dr. Baden went on to explain that whenever I say 'overdose,' it is in quotation marks." 36

The reporters, of course, could not see those invisible quotation marks when they listened to Dr. Baden at his press conferences and interviews. They quite naturally took him literally-and continued to inform the public that addicts were dying of overdose.*

Even Chief Medical Examiner Helpern eventually became convinced that the "heroin overdose" publicity emanating from his office was "dangerously wrong." In testimony before the Select Committee on Crime of the United States House of Representatives on June 27, 1970, Dr. Helpern stated:

A difficulty has been that people have considered these fatal reactions the result of overdose. Now, to some people the designation overdose means [taking] more than usual with the implication that if you are careful of how much is taken there is no danger of anything other than the usual effect. This impression which many addicts have is dangerously wrong. 38

Yet a full year after Dr. Helpern testified, neither he nor Dr. Baden nor anyone else had yet ventured to correct the "dangerously wrong" view that was being foisted on the New York and national news media. Almost everyone who did not read Dr. Helpern's and Dr. Baden's papers in the medical journals still believed that heroin addicts by the hundreds were dying of overdose. Worse yet, nobody had as yet even begun to investigate seriously the crucial question: If these hundreds of addicts a year aren't dying of overdose, what are they dying of?

Fortunately, enough is already known to suggest some promising directions for immediate research.

Most deaths from so-called overdose, as noted above, are characterized by suddenness and by pulmonary edema. No other cause of death-such as tetanus, bacterial endocarditis, hepatitis, or a knife or gunshot wound -is found. In approximately 60 percent of autopsies, a 1970 study indicates, there is also cerebral edema (accumulation of fluid in the brain) along with widespread fragmentation of the astrocytes (star-shaped cells) in the brain.39 A death with these characteristics, occurring in a heroin addict, constitutes a dramatic and readily identifiable syndrome which Dr. Helpern has called "acute fatal reaction to the intravenous injection of crude mixtures of heroin and other substances." We shall here apply a less cumbersome label: "Syndrome X."

1 Thus in the New York Times for December 16, 1969, a reporter was led to state without qualification: "About 800 addicts of all ages died this year from overdoses, according to Dr. Baden." 37

One clue to the true cause of Syndrome X is its initial appearance about 1943, its relative rarity for the next few years, and its recent rapid increase in frequency. The time sequence obviously suggests that the cause of Syndrome X must be some factor introduced about 1943 and affecting a vastly increased number of addicts during 1969 and 1970. Heroin clearly does not qualify; it was widely used long before 1943. Indeed, a highly significant fact about Syndrome X is that it has become more and more frequent as the amount of heroin in the New York City bag has gone down and down. These deaths are, if anything, associated with "underdose" rather than overdose.

One theory sometimes advanced is that Syndrome X deaths are caused by the quinine in the bag. Quinine was introduced as an adulterant of heroin sometime after 1939, when an epidemic of malaria spread by contaminated injection needles hit New York City addicts 40 thus the time of introduction fits the Syndrome X timetable. Some addicts discovered that the quinine contributed to the sensation known as a "rush" immediately after injection. Heroin traffickers also discovered that the bitter taste of the quinine makes it impossible for addicts to gauge the concentration of heroin in the bag by tasting the mixture. For these and possibly other reasons, quinine has remained a standard adulterant of New York City heroin ever since.

Perhaps the first suggestion that quinine might be causing New York City's Syndrome X deaths came from Dr. F. E. Camps, the United Kingdom Home Office pathologist in charge of investigating opiate deaths in England. At a conference of the Society for the Study of Addiction held in London in September 1966 (which Chief Medical Examiner Helpern attended), Dr. Camps stated: "The only comparable drug to heroin which causes rapid death with pulmonary oedema is quinine. In this case patients start off with discomfort in their chest, and then rapidly die. It is conceivable that this could have some relation to [New York City] heroin deaths ." 41

At the same conference an American pathologist, the late Dr. Rudolph J. Muelling of the University of Kentucky Medical School, added that a type of lung lesion similar to that found in Syndrome X deaths "is found to occur when one studies pure quinine cases. In the United States this kind of lesion has been found in several nurses attempting to induce abortions on themselves. They take the quinine orally and the condition comes on quite rapidly. The patients die of quinine alone." 42

A second possible cause of Syndrome X deaths can best be illustrated by two examples.

One is the case of "C. G.," a heroin addict long accustomed to mainlining his drug, who one day got drunk, took his "customary injection of heroin and collapsed shortly thereafter." Subsequent X-rays showed lung edema.43

Another is the case of a heroin addict whose death was recently reported by Dr. George R. Gay and his associates at the Haight-Ashbury Medical Clinic, San Francisco. This addict first "shot some reds" (that is, barbiturates) and then "fixed" with heroin following the barbiturates. He died of what was diagnosed as "overdose of heroin ." 44

Cases such as these have given rise to the question whether Syndrome X deaths may result from injecting heroin (with or without quinine) into a body already laden with a central-nervous-system depressant such as alcohol or a barbiturate.

Addicts themselves would seem to deserve credit for first suspecting that so-called "heroin overdose" deaths might in fact result from the combined action of alcohol and heroin. Back in 1958, a team headed by Dr. Ray E. Trussell and Mr. Harold Alksne interviewed more than 200 New York City addicts-alumni of the Riverside Hospital addiction treatment program (see Chapter 10). In this as in other pre-1960 studies, few addicts drank alcohol while on heroin, and they did not drink much. When asked why, the addicts commonly gave two reasons.

One was that the effect of alcohol is "offensive" to a man on heroin. "The narcotic alone has an analgesic effect which tends to quiet the individual. Alcohol, on the other hand ... has the capacity to agitate the individual in his relationships with other people. This generally is offensive to the addict." 45

The other reason given by addicts in 1958 for not drinking while on heroin is the first extant clue to the possible relationship between alcohol and death from "heroin overdose." Addicts, the Trussell-Alksne team noted, "believe that the use of narcotics and alcohol in combination is dangerous and might possibly lead to the death of an individual." 46 By the 1960s, this awareness of the hazard of shooting heroin while drunk had disappeared from the addict scene. Addicts, like others, were evidently convinced by the official announcements that those deaths were indeed due to heroin overdose.

If the theory is sound that even an ordinary dose of an opiate injected while drunk can produce death, then death could occur when an ordinary drunk who is not addicted is brought into a hospital emergency room with a painful injury and is given a routine (10 milligram) injection of morphine to ease his pain. Drs. William B. Deichmann and Horace W. Gerarde report in their Toxicology of Drugs and Chemicals (1969 edition) that death may in fact occur under such conditions.

"The ordinary safe therapeutic dose of morphine," they warn, in italics, in their textbook, "may be fatal to persons who have been drinking alcoholic beverages. Morphine in therapeutic doses [similar to the doses commonly injected by addicts] resulted in fatalities in individuals whose blood alcohol levels ranged from 0.22 to 0.27%. Morphine is also synergistic with barbiturates and related drugs." 47 Thus the hazard of death from shooting an opiate while drunk on alcohol or a barbiturate is familiar to some toxicologists even though it has been ignored by authorities on drug addiction-and by coroners and medical examiners-through the years.

If this alcohol-heroin and barbiturate-heroin explanation is correct, the fact is of the utmost practical importance-for hundreds of deaths a year might be prevented by warning addicts not to shoot heroin while drunk on alcohol or barbiturates.

The alcohol-barbiturate hypothesis fits the Syndrome X time schedule. Throughout the nineteenth century and well into the twentieth, opiate addicts were known for their dislike of alcohol while on opiates. As noted earlier, they turned to alcohol only when deprived of their opiate supply or when trying to "kick the habit." This remains generally true today; an addict rarely drinks while on heroin. He often drinks, however, when his heroin supply runs out and withdrawal symptoms set in. During World War 11, many heroin addicts were abruptly deprived of their heroin supply for longer or shorter periods. If some of them turned to alcohol, then connected with a fresh heroin supply and "shot up" while still drunk, the first few identified Syndrome X deaths might have occurred. The recent sharp increase in Syndrome X deaths might similarly be explained by an increased tendency to alternate alcohol or barbiturates with heroin as a result of high heroin prices. As the amount of heroin in the New York City bag went down and down, according to this theory, more and more addicts got drunk-and died of Syndrome X following their next "fix."

Evidence in recent years for the use of alcohol by addicts shortly before their death has been assembled from the New York City files by Drs. Jane McCusker and Charles E. Cherubin. They reviewed 588 city toxicology reports found in the files on addicts who died in 1967. In 549 of these cases, tests for alcohol had been run-and in 43 percent of the cases tested, alcohol was in fact found .48 (Barbiturates were not reported on.) Their findings led Drs. McCusker and Cherubin to suggest that further research be promptly launched into the possible role of alcohol and the barbiturates in so-called "heroin overdose" cases.

The same suggestion has been tentatively made by Dr. Gay of the Haight-Ashbury Medical Clinic. Thirty-seven percent of the addicts attending the clinic, Dr. Gay states, report using barbiturates "for sedation and sleep" when heroin withdrawal symptoms set in; and 24 percent report using alcohol similarly.49 Thus the stage is set for shooting heroin while drunk on one or the other-and, perhaps, for sudden death from overdose."

Two of the most publicized "overdose deaths" of 1970, Dr. Gay informed the National Heroin Conference in June 1971, fit precisely this pattern. These were the deaths of the rock musician Jimi Hendrix and the singer Janis Joplin. Hendrix was known to use both alcohol and barbiturates-and possibly also heroin. Janis Joplin "drank [alcohol] like an F. Scott Fitzgerald legend," Dr. Gay adds-and also used narcotics.50 The magazine Time reported on October 19, 1970, shortly after Janis Joplin's death:

The quart bottle of Southern Comfort [whiskey] that she held aloft onstage was at once a symbol of her load and a way of lightening it. As she emptied the bottle, she grew happier, more radiant, and more freaked out....

Last week, on a day that superficially at least seemed to be less lonely than most, Janis Joplin died on the lowest and saddest of notes. Returning to her Hollywood motel room after a late-night recording session and some hard drinking with friends at a nearby bar, she apparently filled a hypodermic needle with heroin and shot it into her left arm. The injection killed her.51

Janis Joplin's death, of course, was popularly attributed to "heroin overdose." If the alcohol-barbiturate-heroin theory is correct, her fatal injection of heroin while drunk on alcohol was the prototype of many other deaths similarly mislabeled "overdose."

The British experience with deaths attributed to heroin overdose is consistent with the alcohol-barbiturate hypothesis. Dr. Ramon Gardner of the Bethlem Royal Hospital and Maudsley Hospital in London studied the records of 170 deaths known to have occurred among addicts in Britain during the five-year period 1965 through 1969. Twenty of these deaths were deemed suicides, 24 were traceable to infections, 12 were from natural causes, 11 were drownings, falls, murders, or other accidents, and 6 occurred during treatment (of which two followed abrupt withdrawal of narcotics when the addicts were imprisoned). Eight more were due to overdose of barbiturates or other nonopiate drugs. This left a maximum of 89 mysterious deaths out of 170 which might have been caused by accidental opiate overdose-or by something else.52

Dr. Gardner then went on to study in more detail 47 of these deaths possibly due to heroin overdose. In a number of cases, he found that the addicts bad been confined in a hospital, prison, or detention center or had for other reasons been abstinent from opiates for a week or longer, and had thus lost at least a portion of their tolerance for opiates. They had then injected an opiate-some of them on the day of discharge, others within the next day or two. Thus these deaths might have been due to overdose-though evidence was lacking that the victims had in fact taken fatally large doses. (Merely doubling or quadrupling the dose, it will be recalled, will not kill even nonaddicts.)

But in at least 21 of the 47 cases, there had been no withdrawal from opiates prior to death, so that tolerance had not been lost. And in some cases, the dose preceding death was so small-as little as 20 or 30 milligrams of heroin, for example-as to establish beyond question that overdose was not the cause .53

These British deaths, accordingly, remain mysterious, like deaths from Syndrome X in the United States. Among several likely explanations, Dr. Gardner himself noted, is the possibility that these addicts may have taken some other drug, perhaps a central-nervous-system depressant, at the same time. Since there is no quinine in British opiates, that drug must be exonerated in the British deaths.

Another British drug authority adds that in Britain as in the United States, "many of those who die, in fact, have taken barbiturates as well [as opiates] at the same time." 54

It might prove absurdly easy to confirm the alcohol-barbiturate hypothesis. All that might be necessary would be to addict a few monkeys or other primates to heroin, intoxicate them on alcohol or barbiturates, and then inject modest doses of heroin. If the monkeys drop dead of Syndrome X, a warning against shooting heroin while drunk on alcohol or barbiturates might save many hundreds of lives a year throughout the world.

Several other possible explanations of Syndrome X deaths have been offered. No theory has yet been proved. Worse yet, no theory has ever been experimentally tested. The time has surely come to determine the cause (or causes) of Syndrome X and bring to a close this tragic series of deaths. If 800 respectable citizens instead of heroin addicts had dropped dead in New York City of a mysterious syndrome in 1970, a gargantuan research program would no doubt have been promptly launched.

If the Syndrome X deaths are due to quinine or to any other adulterant or contaminant in the bag, the responsibility clearly rests with the American heroin black market for selling unsafe mixes. If the cause of these deaths is the shooting of heroin while drunk on alcohol or barbiturates, the black-market distribution system remains at least indirectly responsible, for it is largely the high cost of black-market heroin that makes heroin users turn to alcohol and barbiturates on occasion-and thus, perhaps, to risk death from Syndrome X.

The two steps which must now be taken are (1) to stop sweeping these mysterious deaths under the carpet by falsely labeling them "overdose" and (2) to launch an intensive clinical and experimental search for what is in fact killing these addicts.