Sign the Resolution
Contents | Feedback | Search
DRCNet Home | Join DRCNet
DRCNet Library | Schaffer Library | Cocaine
This chapter introduces the pharmacology of crack cocaine to help readers evaluate the claims that have been made regarding its danger to individual users and society. Because, materially and pharmacologically, crack is cocaine, most of what is known about cocaine applies to crack as well. The fact that crack is smokedrather than sniffed, swallowed, or injectedis significant. Our review of the evidence indicates, however, that its importance has been exaggerated. Clearly, using either cocaine powder or crack entails risks, but both can also be used in less or more risky ways. In fact, among the things we will show is that the amount of harm resulting from the use of powder cocaine and crack has less to do with their pharmacological properties than with the social circumstances of their use.
FROM COCA LEAVES TO CRACKFor centuries, people have consumed cocaine to enhance work performance, forestall drowsiness, lift mood, and produce feelings of elation and euphoria. In the South American Andes, for over a thousand years people have ingested cocaine by chewing coca leaves or brewing them into a tea. This form of consumption seems not to be associated with significant biological harm or social dysfunction (Aldrich and Barker, 1976; Antonil, 1978; Forno et al., 1981; Weil, 1986) and has not, by and large, been subjected to repressive government control (Henman, l990; Morales, 1989).
There was little use of coca in the United States or Europe until the mid-nineteenth century, when the plant's principal active ingredient was extracted and made available as a water-soluble powdercocaine hydrochloride Western physicians soon discovered that cocaine was an effective local anesthetic; they also used it, although less effectively, as an antidepressant, asthma remedy, and a treatment for opiate addiction About the same time, cocaine was added to numerous patent medicines and tonics that people purchased without prescription to combat a variety of common ailments, including chronic fatigue.
During the late nineteenth century, Americans also consumed cocaine recreationally, often in beverage form. Vin-Mariani wine and Coca-Cola for example, were popular cocaine-based drinksand the latter was even marketed as a "temperance beverage" to people wishing to avoid alcohol (Pendergrast, 1993). There is less information available about the recreational use of cocaine powder during the nineteenth century, but it seems to have been most common among members of the "criminal underworld" (Grinspoon and Bakalar, 1985; Inciardi, 1992; Musto, 1987)a fact that helped fuel public support for increased government controls. Also precipitating anticocaine legislation around the turn of the twentieth century were growing concern about cocaine's potentially harmful physical effects (Alexander, 1990; Courtwright, 1982; Grinspoon and Bakalar, 1985) and fear, especially in the South, that the drug caused blacks to behave violently (Morgan, 1981; Musto, 1987; Pendergrast, 1993). However, because this was an era of increasing government control over most available intoxicantsincluding alcoholthe laws regulating cocaine may have had little to do with this drug's particular characteristics and effects.
As early as 1887, states began passing anticocaine laws (Ashley, 1976); and in 1906, with enactment of the first Pure Food and Drug Act, the federal government began requiring that products with cocaine (and some other drugs) be labeled as to content. Then, in 1914, Congress passed the Harrison Act, which originally only imposed tax and registration requirements on the legitimate providers of certain drugs, including cocaine. However, courts soon interpreted this law as giving federal drug enforcement officials the power to decide what constituted "legitimate" use of these drugs; and, through this power, they quickly transformed the Harrison Act into a law prohibiting all recreational use of cocaine.
One immediate consequence of cocaine prohibition was the elimination of cocaine tonics and beverages. Another was the emergence of an organized black market in cocaine hydrochloride, which was smuggled into the country from South America. Almost certainly, the purity of the product available to users declined, the price rose far above the $2 an ounce that had been common during the previous century (Courtwright, 199l), and use became even more concentrated in deviant subcultures (Ashley, 1976; Grinspoon and Bakalar, 1985; Inciardi, 1992).
After 1930, when a number of synthetic stimulants (particularly amphetamine) became available, cocaine use may have decreased further, although it continued to be used by some artists and entertainers, who generally sniffed it, and by intravenous heroin users, who employed it either as an occasional alternative to heroin or mixed with heroin to form a "speed-ball" (Grinspoon and Bakalar, 1985). During the 1960s, as part of the more general increase in the use of illegal drugs among more "mainstream" Americans, the use of cocaine probably increased as well, although in 1972, still less than 3% of the population (aged twelve and over) said they had tried it (Johnson and Muffler, 1992).
During the remainder of the 1970s and into the early 1980s, cocaine use increased steadily, especially among young adults aged eighteen to twenty-five (NIDA, 1991a). Probably contributing to cocaine's appeal was the government's success, first, in curtailing diverted medicinal amphetamine (Brecher, 1972; Inciardi, 1987; Morgan and Kagan, 1978) and, second, in interdicting enough marijuana substantially to decrease its availability and increase its price (Cowan, 1986; Hamid, 1992; Lazare, 1990). As the demand for cocaine increased, supplies increased as well, and by 1982, approximately 28% of eighteen-to twenty-five-year-olds had at least tried it (NIDA, 1991a). However, because of cocaine's relatively high priceup to $100 for a gram of powder in the early 1980suse was most prevalent among the middle and upper classes (Grinspoon and Bakalar, 1985). The typical mode of ingestion was to sniff cocaine hydrochloride powder into the nose, which permits absorption through the nasal mucosa.
Today, essentially all cocaine enters the U.S. in the form of hydrochloride powder. This powder is extracted in a process that begins by mixing pulverized coca leaves with a solvent (such as ether or gasoline) and partially drying it. Then, to make the product water-soluble, this "coca paste" is treated with hydrochloric acid and dried to a white powder. In this form, cocaine can be sniffed, swallowed, or dissolved in water for injection, but it cannot be smoked because igniting it degrades the cocaine before it will volatilize. However, through a series of fairly simple chemical procedures, cocaine can be turned into "freebase"a product that resembles the smokeable coca paste. To produce freebase, cocaine hydrochloride is mixed in water with a liquid base (such as ammonia, baking soda, or sodium hydroxide) to remove the hydrochloric acid. The resulting alkaloidal cocaine is then dissolved in a solvent (such as ether) and gently heated, causing most of the liquid to evaporate. The product created, when placed in a glass pipe and ignited, produces vapors of relatively pure cocaine.
Inhaling cocaine vapor into the lungs delivers the drug more rapidly to the bloodstreamand therefore to the brainthan does sniffing the powder; as a consequence, it produces quicker, more intense effects. Most cocaine users do not want this more dramatic experienceespecially because it means, as well, a more rapid diminishing of the drug's effects. Also reducing freebase's attractiveness is the somewhat complicated conversion processwhich occasionally can be dangerous because some of the solvent used in the preparation may remain in the product being ignited. Nonetheless, freebasing did increase in popularity in the early 1980s (Hamid, 1992; Inciardi, 1987; Siegel, 1984), mainly attracting people who were already fairly heavy users of powder cocaine (Siegel, 1984; Waldorf et al., 1991; Washton et al., 1986).
Around 1985, another form of smokeable cocainecalled "rock" or "crack"became available. Its production resembles that of freebase, but without the final purification process: cocaine hydrochloride is dissolved in water, sodium bicarbonate (baking soda) is added, and the mixture is heated and then dried into hard, smokeable pellets. These pellets contain not only alkaloidal cocaine, but sodium bicarbonate and whatever other fillers and adulterants had been added earlier to the powder; thus, crack is not as highly purified as freebase, and street samples tend to range from 10 to 40% cocaine by weight (Inciardi, 1987). Still, igniting crack produces a vapor that is largely pure cocaine (Snyder et al., 1988), making the experience of smoking crack quite similar to that of smoking freebase. However, unlike freebase, which users generally produced themselves from the powder, crack was usually cooked (or "cracked up") by drug dealers who then sold it in ready-to-smoke form (Hamid, 1990).
Crack quickly gained in popularity, although it never became as popular as cocaine powder. For example, in 1991, nearly twenty-four million Americans (aged twelve and over) said they had tried cocaine, compared to less than four million for crack (NIDA, 1991b). Although the price of cocaine had been decreasing and its quality increasing during the early 1980s, it was still, in 1985, too expensive to be used very much by the poor. What crack did was to lower dramatically the cost of the "cocaine high." Simply because smoking delivers a drug more efficiently to the brain than does snorting, an amount of cocaine too small to produce an effect in powder form becomes an effective dose when converted to crack. In 1986, a single dose of crack could be purchased for as little as $5 or $10; and, over the next few years as the price of cocaine powder fell even further, the price of a pellet of crack fell as low as $2 in some parts of the country (Cohn, 1986). Thus, by the late 1980s, what had once been called "the champagne of drugs" had become available to the poorand its use spread especially quickly in impoverished urban areas where enterprising youth turned powder cocaine into crack and sold it on the streets (Fagan and Chin, 1989; Hamid, 1990; Williams, 1992).
As Chapter 2 showed, once crack had been introduced to the inner-city poor, the "crack epidemic" became a major media eventwith literally thousands of articles appearing in newspapers and magazines in 1986 alone. At the time, no scientific studies of the drug had been conducted, but journalists found and quoted a handful of "experts"mostly law enforcement officials and drug treatment providerswho had decided that crack was "the most dangerous drug known to man." They claimed that crack was highly potent and highly toxic, causing record numbers of heart attacks, seizures, and strokes. They blamed crack for recent increases in crime, family violence, and child abandonment. They claimed that crack was "instantly addicting," making moderate and controlled use impossible. And when used by pregnant women, crack was said to produce babies so severely damaged that they would never fully recover.
Before long, articles supporting these claims appeared in the drug abuse and medical literatures. Although clothed in scientific garb, most of these "studies" were simply "case reports" of crack and cocaine users enrolled in drug abuse treatment programsa self-selected and nonrepresentative sample (Gold et al., 1986; Honer et al., 1987; Isaacs et al., 1987; Miller et al., 1989; Mody et al., 1988; Spitz and Rosecan, 1987; Washton et al., 1986; Weiss and Mirin, 1987). Even today, few of the "facts" that are "well known" about cocaine and crack come from careful scientific studies. Nonetheless, they have made their way into government documents, drug education materials, and anti-drug public service announcements particularly those of the Partnership for a Drug-Free America. In addition, although a number of journalists have been critical of the media's handling of the crack story (Gladwell, 1986; Martz, 1990; Morley, 1989; Weisman, 1986), exaggerated tales of cocaine- and crack-caused horror still appear regularly in the popular press.
The faulty assumption on which such drug horror stories are based is that a drug's pharmacology holds the key to understanding the patterns of its use and the behavior of its users. One of our goals in this chapter is to demonstrate that this "pharmacocentrism" is misleading. In doing so, we are not suggesting that a drug's pharmacology is unimportant. After all, for a drug to be used recreationally, people have to like how it makes them feel, and how a drug makes people feel is a product of its "pharmacological fit" with the human organism. However, a description of this "fit" cannot explain why only some people use a particular drug, why only some of them become regular users, or why fewer still use it in a volume and frequency that disrupts their lives. In short, to explain how a drug works in the brain reveals no more about why and how people use it than explaining how a specific food is processed by the body reveals why and how people eat it. Like food consumption, drug consumption must be understood, primarily, as a social-psychological phenomenon. In fact, one of the things we hope to show in the following overview of crack cocaine's pharmacology is how little it reveals about the drug's popularity or the social consequences of its use.
PHARMACODYNAMICS:Like all stimulant drugs, those prescribed by physicians as well as those taken recreationally, cocaine produces a psychoactive effect by interacting with the central nervous system, stimulating it to perform its ordinary functions more intensely. This system operates through the release of various neurochemical transmitters (from the nerve cells in which they are produced) and their binding to receptor sites on neighboring cells. The constant release and binding of these neurotransmitters forms a pathway of "messages" that travel throughout the body, sustaining life and making possible the organism's response to environmental stimuli.
COCAINE'S INTERACTION WITH THE HUMAN ORGANISM
Cocaine also has an impact in the autonomic (or involuntary) division of the central nervous system, which helps regulate a variety of bodily functions that are generally free of volitional impact, including respiration, circulation, digestion, and body temperature. Ordinarily, these functions are maintained at relatively stable levels throughout the day. But they are slowed down during periods of rest through diminished production, release, and binding of neurotransmitters and can be speeded up, as needed, through increased neurotransmitter activity. Cocaine operates in this system by increasing the concentration and binding activity of the body's own neurotransmittersparticularly dopamine. Thus, what people experience as cocaine's stimulant effect is an intensification of the body's normal stimulatory mechanisms.
Cocaine is both a quick-acting and a short-acting drug. When cocaine enters the bloodstream directly, via injection, it reaches the brain quickly, and users feel its effects within minutes. Inhalation also delivers cocaine quickly to the brain because air passages in the lungs are positioned close to capillary accesses to the bloodstream. When cocaine is sniffed, the onset of effect is slower because the drug must pass through the nasal mucosa before entering the bloodstream. Swallowing cocaine delays delivery to the brain even more because most of the drug is passed through the gastrointestinal tract before it crosses through cell membranes into the bloodstream.
Controlling for dose, sniffing and swallowing also produce less intense effects. This is not only because these routes of administration cause active cocaine molecules to reach the brain more gradually, and therefore in lower concentrations, but also because the additional passage of time allows more of the cocaine molecules to be transformed into inactive byproducts (or metabolites) before they reach the brain. Both injection and inhalation deliver a greater number of active cocaine molecules per dose to the brain than snorting.
Whatever the route of administration, within thirty to sixty minutes, the processes of biotransformation and excretion cut in half cocaine's concentration in the bloodwhich is one reason its effects are of relatively short duration. However, even before this decline, cocaine's effects are diminished through other "protective mechanisms." The most important is the rapid distribution of the drug from the bloodstream to the rest of the body, including to sites with no cocaine receptors. Thus, the same activity that delivers cocaine rapidly to active sites in the brain and heart also removes it from these sites, thereby reducing the drug's effects. At the same time, cocaine's effects are also diminished through homeostatic mechanisms that reduce neurotransmitter activity at receptor sitesthe same mechanisms that operate when factors other than drugs cause an increase in neurotransmitter activity. It is this diminishing of effectsprior even to the decline in the drug's concentration in the bloodthat cocaine users experience as "acute tolerance." That is, to maintain a stable effect over time, users must follow each dose of the drug with a larger subsequent dose. However, acute tolerance also can be viewed as a preexisting protective mechanism because it diminishes cocaine's potentially harmful effects on the cardiovascular and central nervous systems.
Cocaine's Psychostimulant EffectsThe intensity of cocaine's impact on the central nervous system depends largely on dose. At low doses, cocaine's effects are fairly similar to those of caffeine: it combats drowsiness and fatigue, increases energy and alertness, and enhances mental acuity. With increasing doses, most users begin to experience negative effectssuch as nervousness, jitteriness, sleeplessness, and agitationand at very high doses, feelings of suspicion, hypervigilance, and paranoia are common (Cohen, 1989; Erickson et al., 1987; Spotts and Shontz, 1980; Waldorf et al., 1991).
Extremely high does of cocainelike extremely high doses of many stimulant drugscan produce a toxic psychosis, with symptoms similar to the delirium of high fever. However, toxic psychoses appear to be rare among cocaine users, probably because of the body's protective mechanisms referred to previously. In addition, because cocaine is relatively short acting, when psychosis does occur, it tends to be short-lived (Weil, 1986). Permanent psychosis is found occasionally among cocaine users (Washton, 1989; Weiss and Mirin, 1987), but there is no evidence of a causal link, and, for most people, even heavy and prolonged use appears to have no permanent impact on mental health, personality, mood, cognition, memory, or perception.
Among people predisposed to behave violently, cocaine may increase the likelihood of their involvement in violent episodes, but there is no evidence that cocaine causes generally nonviolent people to behave violently. Some researchers have identified crack as more violence producing than cocaine powder (Peterson, 199l; Washton, 1989), and journalists have been prone to attribute increases in violent crime to the pharmacological properties of crack. However, a growing number of social scientists refute these claims.
Crack use by women has also been blamed for rising rates of child abuse (Peterson, 199l). However, to the extent that crack users seem to "lose their mothering instinct" and begin abusing or neglecting their children, it is probably due less to the pharmacology of the drug than to the lifestyle that accompanies heavy involvement in the street drug sceneregardless of the drug (Rosenbaum et al., 1990). In fact, research on a variety of drugs shows that the same drug is associated with very different behaviors in different cultures, which indicates that there is no direct link between any specific drug and any specific behavior (see, e.g., MacAndrew and Edgerton, 1969; Zinberg, 1984). In this culture, cracklike alcoholis associated with violence primarily because it is often used by people already at high risk for behaving violently and because it is often used in social settings in which violence is already common (Williams, 1992). No drug directly causes violence simply through its pharmacological action.
Cocaine's Physiological EffectsBecause it constricts blood vessels and speeds up the heart, cocaine has the potential to produce cardiovascular disease. However, at low doses, the increases in blood pressure and heart rate caused by cocaine are fairly similar to those associated with over-the-counter appetite suppressantsand are less dramatic than those experienced by most people during aerobic exercise, urban driving, or sex. With larger doses, cocaine's cardiovascular effects become more pronounced, and users face an increased risk of harm to the heart through coronary artery constriction or arrhythmia. High-dose users also face an increased risk of adverse stimulant effects in the central nervous system, including seizures, convulsions, and strokes (Cregler and Mark, 1986).
Because intravenous injection allows the rapid delivery of a large dose of cocaine, injectors are more likely to experience adverse physiological effects. Rapid consumption of multiple doses increases the risk associated with other routes of administration, but the body's capacity quickly to diminish cocaine's effects protects even most high-dose smokers, sniffers, and swallowers from serious harm. Oral ingestion clearly has the greatest safety margin, although an extremely large dose swallowed can be dangerous, as indicated by the death of "body packers"people who swallow balloons or condoms filled with cocaine to smuggle it across borders (Amon et al., 1986; Suarez et al., 1977).
Although most people can consume a fairly high dose of cocaine with out serious harm, even a low dose can be dangerous for people with preexisting central nervous system or cardiac abnormalities (Isner et al., 1986; Mittleman and Wetli, 1987). People with enzyme deficiencies that interfere with cocaine's biotransformation may also be at higher risk (Devenyi, 1989). There is recent evidence that consuming alcohol with cocaine may be risky, especially for persons with heightened sensitivity to cocaine's effects (Karch, 1992).
Cocaine Toxicity as a Cause of DeathCocaine's lethal potential has been demonstrated through the administration of high doses to animals (Finkle and McCloskey, 1978; Smart and Anglin, 1987), but until fairly recently, death was regarded as a rare occurrence among human cocaine users (Lundberg et al. 1977). Then, in 1979, in the Journal of the American Medical Association, the coroner from Dade County Florida reported that during the previous decade there had been sixty-eight deaths from "recreational use of cocaine" (Wetli and Wright, 1979). This article is still widely cited as evidence of cocaine's dangers, even though the "study" actually involved little more than attributing to cocaine all deaths in which evidence of cocaine had been detected by postmortem examination.
Examining the sixty-eight cases used by Wetli and Wright, Bruce Alexander (1990) found that fifteen of the deaths had actually been caused by trauma (automobile accidents, drownings, gunshot wounds) and that another had been a suicide. Drugs other than cocaine were found in the bodies of all sixty-eight victims; and, in fact, at the time of the autopsy, twenty-nine deaths had been officially attributed to "multiple drug intoxication." In none of these cases was there an attempt to determine each drug's contribution to the death or to identify the presence of any physical abnormalities that might have played a contributory role. Alexander estimates that in only seventeen of the sixty-eight cases was it likely that cocaine had made a pharmacological contribution to the deathand five of these seventeen deaths were of "body packers" who had smuggled extremely large quantities internally in balloons or condoms that broke. The remaining twelve victims may have died after using cocaine "recreationally," although because the records contained no information regarding dose or frequency of use, there is no way of knowing how much cocaine these people had used or by what mode of ingestion.
Also widely cited as evidence of cocaine's deadly potential are data from a national sample of hospitals and coroners' offices compiled in association with the federally funded Drug Abuse Warning Network (DAWN) project. These data, which show steadily increasing numbers of cocaine-related deaths during the second half of the 1980s when crack use was spreading, are no better for determining cause of death than those used by Wetli and Wright. According to DAWN guidelines, any death that "involves" drug abusedefined as any use of a controlled substance for its psychic effects, without medical approvalcan be counted as drug related, without proof that drugs actually caused the death. In fact, in some cases, deaths are attributed to drugs solely on the basis of circumstantial evidence of drug abuse, without even toxicological verification of their presence at the time of death (Benowitz, 1992). In 1990, nearly twenty-five hundred deaths nationwide were estimated by DAWN to be cocaine related, but in about three-quarters of those, one or more additional drugs were mentionedmost commonly alcohol.
Casting further doubt on the validity of DAWN's fatality data is Tardiff et al.'s (1989) review of the coroner's reports from 935 New York City deaths that had been officially labeled cocaine related. They found that about half these deaths had been due to trauma (caused by accidents homicides, and suicides) and that less than 12% were even possibly related to the pharmacological effects of cocaine. If these 935 cases are at all representative of the coroners' reports used by DAWN, the number of deaths caused by cocaine nationwide in 1990 may have been more in the neighborhood of 250 than the 2500 estimated by DAWN (NIDA, 1991c) . In fact, a finding from this NIDA report that has not been widely publicized is that in only 172 cases (less than 7% of the 2483 officially identified as cocaine related) was cocaine identified as the single, direct cause of death. Cocaine-Related Medical Emergencies
DAWN also compiles data on drug-related hospital emergency room visits and, since the mid-1980s, has reported steadily increasing numbers of emergencies related to cocainewith increases in some years as much as 100%. However, like the data for drug-related deaths, the data for drug-related emergency room visits are problematic because DAWN compilers count as a "drug-related episode" any emergency room visit that "involves" drug abuse. And, if more than one drug is identified in an episode, each is reported as a separate drug "mention," whether or not it contributed substantially to the condition prompting the visit. In 1990, for example, cocaine was mentioned in an estimated eighty thousand emergency episodes, but in only about one-quarter of those was it the only drug mentioned; the drug most often mentioned in combination with cocaine was alcohol (NIDA, 1991d).
The DAWN data cannot be used to estimate the incidence of medical complications associated with the use of cocaine or other drugs because there is not, among the six possible "reasons" for emergency room contacts recognized by DAWN, one for physical symptoms related to drug use. Surely some of the visits included in the category "unexpected drug reactions" (e.g., 22.9% of the cocaine-related episodes reported in 1990) involved physical symptoms, but even many of those may not have required medical attention. In fact, given the recent publicity regarding cocaine's dangers, it is possible that some users became frightened by fairly mild cardiovascular symptoms and were prompted to seek medical attention they did not actually need.
Despite these (and other) problems with the research methodology most of which inflate the incidence of cocaine-related toxicitythe DAWN data are routinely offered as proof that cocaine users face grave risks of physiological harm. At the same time, the fact that cocaine-related emergency room mentions have continued to rise, even as overall rates of cocaine use have declined, is used as evidence that hard-core cocaine abuse has risen (Millman, 1991; White House, 1989). Increases in both cocaine abuse and cocaine-related emergency room visits are, in turn, often attributed to the increased use of crack (Kandel, 199l; Schuster, 1990).
There is little direct evidence that crack users suffer more physiological harm than do those who use comparable amounts of powder cocaine. It is difficult to know if crack users are over-represented among DAWN's cocaine mentions, but if they are, it may be because people who use crack are more likely than cocaine powder users to go to emergency rooms. After crack first appeared in the mid-1980s, politicians, clinicians, journalists, and drug czars all predicted that its use would quickly spread to all communities and all socioeconomic groups. However, as earlier chapters have shown, crack has remained a drug used primarily by the urban poor, who are most likely to seek treatment for their medical problems, drug related or not, at hospital emergency rooms (Wishner et al., 199l ).
Such a socioeconomic explanation for the overrepresentation of crack users in the DAWN reports is supported by Waldorf et al.'s (1991) study of heavy cocaine users. Within their sample of predominantly middle-class cocaine snorters and smokers, they found that smokers (who used freebase and/or crack) developed no more drug-related health problems than did cocaine sniffersalthough the smokers did, as a group, develop health problems earlier (see Chapter 4). In addition, although most of the predominantly middle-class cocaine users in their sample did attribute some health problems to cocaine, there is no evidence that they visited hospital emergency rooms. Indeed, like most of the twenty-five million Americans who have used cocaine, very few ever sought medical help for the physiological problems they believed were related to their use of the drug. Clearly, both cocaine and crack have the potential to produce serious harm, even death; but the evidence isdespite government reports showing ever-increasing harmthat most people consume these drugs in a way that does not cause them lasting or even temporary harm.
CRACK AS AN ADDICTIVE DRUGEvaluating the addictive potential of cocaineor any drugis complicated by the fact that addiction is, as Bakalar and Grinspoon (1984) put it, "an essentially contested concept." In fact, the term "drug addiction" seldom appears in the substance abuse literature anymore, having been replaced by "drug dependence" or "substance abuse disorder"conditions diagnosed largely on the basis of behavior. Definitions vary, but most include as a core criterion the continued use of a drug despite the appearance of negative consequences for the user's health, work, financial stability, relationships, and the like. The key element is "compulsion" or "loss of control," which suggest an inability to alter drug-taking behavior (see, e.g., Jaffe, 1990).
There is no question that some people use crack or cocaine in other forms despite its negative impact on their lives. Furthermore, as the crack users and freebasers in Chapter 4 suggest, in virtually every sample of cocaine users, some people identify themselves as "addicts" or describe their relationship to cocaine as "obsessive," "compulsive," or "out of control" (Cohen, 1989; Erickson et al., 1987; Siegel, 1980; Waldorf et al., 1991). Thus, among drug users as well as drug experts, cocaine, especially in the form of crack, is generally accepted as having substantial addictive potential. Because cocaine does not produce physical dependence and withdrawal of the sort associated with opiates (Gawin and Kleber, 1986), cocaine addiction was once thought to be primarily psychological in nature. Increasingly, however, it is being described in physiological terms. In fact, some drug treatment entrepreneurs maintain that cocaine addiction must be physical because "no drug can become psychologically compelling without there being physical (indeed cellular) changes in brain activity that both result from and contribute to its continued use" (Washton, 1989:36-37).
A Biochemical Theory of Cocaine AddictionCocaine produces psychoactive effects by increasing the neural activity of dopamine and other neurotransmittersin effect, stimulating the "pleasure system" that is activated when humans have pleasurable real-life experiences. This "chemical reward" is the pharmacological basis for cocaine's use as a recreational drugand for cocaine users' descriptions of its effects as "intensely pleasurable," "euphoric," or, in the case of crack, even "orgasmic." Because people tend to use frequently only drugs that produce pleasurable effects, cocaine's stimulation of the brain's "pleasure system" is also what gives it addictive potential.
However, this is far from a sufficient explanation of cocaine addiction. After all, the government's own data show that most people who try cocaine do not even become regular users, much less "addicts" (NIDA, 1991a). What Washton (1989) and others who have attempted to prove cocaine's inherent addictiveness suggest is that, with continued use, cocaine alters the chemical structure and functioning of cells in the brain's "pleasure system," to the point where the cells themselves begin to "crave cocaine." In fact, Washton claims that it is this new understanding of cocaine's ability physically to alter brain cells that has made the old distinction between psychological and physical addiction meaningless.
The physiological mechanisms of cocaine addiction are presumed to operate more intensely when the drug is smoked, and many early media reports went so far as to identify crack as "the most addictive drug known to man." Before long, similar claims also began to appear in the drug abuse literature, most of them attributing crack's unique addictiveness to the intensity of its high, the rapid onset and short duration of its effects, and the severe "crash" that accompanies its decline (Miller et al., 1989; Spitz and Rosecan, 1987; Washton, 1989; Washton et al., 1986). In fact, Washton maintained that crack so quickly altered the brain's functioning that it was often "instantaneously addicting." Today, many doubt that crack causes instant addiction, but it continues to be widely acceptedeven among scholars who challenge most other unproven "drug truths"that crack is much more addictive than powder cocaine (Inciardi, 1987; Kleiman, 1992; Musto, 1987; Trebach, 1987).
Comparing the Addictiveness of Crack and CocaineThe hypothesis that smoked cocaine is more likely to lead to addiction than is an equal dose used intranasally has never been tested on either animals or humans. Nor is it likely to be because animals cannot be easily trained to use either of these drug administration techniques and fortunately ethical reasons prevent the random assignment of human subjects to an experimental condition (smoking) that is believed to be more dangerous. However, NIDA data can be used to calculate and compare the "continuation rates" for crack and cocaine: the proportion of people who, after trying each drug, continue to use it regularly. Of course, even regular users of crack and cocaine are not necessarily addicts, but if crack is, indeed, more addictive than cocaine, the continuation rates for crack should be markedly higher.
The best data available for this comparison are the population estimates from the National Household Survey on Drug Abuse sponsored by the National Institute on Drug Abuse (NIDA, 1991b). Readers reared on the frightening claims of clinicians, politicians, and the media may be surprised to learn from the NIDA survey that only about one in twelve (8%) of Americans aged twelve and over who have ever tried cocaine had used it at all in the month prior to the survey. This figure was somewhat higher for crack, but still only about one in eight (12.3%) of those who have ever tried crack had used it in the month prior to the survey. The fraction of these "past-month" users who go on to daily use and therefore, arguably, to "addiction" is far smaller. In interpreting these data, it is also important to recognize that precisely because smoking is a more direct mode of ingestion offering a much more intense high, the fraction of cocaine users who are drawn to crack is very likely to be among the heaviest users to begin with. Further, crack was introduced and systematically marketed in impoverished inner-city communities where powder cocaine was less affordable and less available (Hamid, 1992; Inciardi, 1987), which means that crack has been disproportionately available to just those parts of the population who are most vulnerable to the abuse of any drug (Anthony, 1991; Kandel, 1991). Thus, the different continuation rates for crack and powder cocaine may be explained in part by differences in the social circumstances of users themselves.
Data from NIDA's High School Senior Survey make much the same point. For example, in 1991, among students who reported having ever tried crack, only one in thirty-five reported daily or near daily userates virtually identical to those for powder cocaine. In fact, among high school seniors, the continuation rates for alcohol, marijuana, cigarettes, and LSD were all higher than for either powder cocaine or crack (Johnston et al., 1991). Regular use of any drug, licit or illicit, is not something anyone wants to see among high school students. But when the best available evidence shows that the vast majority of young people who try crack do not go on to use it regularly, and when only a small fraction of even these go on to daily use, it is clear that the claim that crack is "instantaneously addicting" is false.
These data indicate not only that relatively few cocaine users become "dependent"whatever their route of administrationbut that smoking cocaine by itself does not increase markedly the likelihood of dependence. This latter finding is important because it means that the claim that cocaine is much more addictive when smoked (Gold, 1984; Inciardi, 1987; Jekeletal., 1986; Jerietal., 1978; Siegel, 1982,1984; Washtonetal., 1986) must be reexamined. We think that a more accurate interpretation of existing evidence is that already abuse-prone cocaine users are most likely to move toward a more efficient mode of ingestion as they escalate their use. The claims of Washton, Gold, and others about crack's extreme dependence liability are based on treatment populations and those who call help hotlinespeople who are, by definition, among the most problematic users. Thus, claims made on the basis of their reports cannot be safely generalized to all who have experimented with crack or freebase.
The Pharmacology of Cocaine BingeingA cocaine binge is an episode of continuous drug taking, lasting several hours or more, in which additional doses of the drug are consumed in an effort to forestall diminution of the effects. People binge on drugs other than cocaine, and, in fact, occasional alcohol bingeing seems almost a "rite of passage" for American youth. Still, bingeing seems to be particularly prevalent among cocaine usersa fact that may be related to cocaine's specific pharmacological action. As discussed earlier, soon after cocaine produces its effects, the body's mechanisms of homeostasis respond to diminish them. This is why, for example, cocaine's cardiovascular effects diminish more quickly than does its concentration in the blood (Fischman et al., 1985). By and large, the more dramatic the cocaine effect, the more dramatic the homeostatic response; and, following a large dose, blood pressure, heart rate, and the like may actually go below normal before returning to normal.
Similar mechanisms operate to diminish cocaine's psychoactive effects, and the bigger the dose, the more dramatic the neural system's homeostatic response. Thus, a dose large enough to produce feelings of euphoria may, a short time later, produce feelings of dysphoria as neurotransmitter activity declines to below normal levels (Waldorf et al., 1991:223-226). Not all cocaine users experience a dramatic shift from euphoria to dysphoria (Van Dyke et al., 1976), but it is common among cocaine users who engage in binges. In fact, it is during this "crash phase" that they report an intense craving for the drugespecially once they have learned that consuming an additional dose restores the euphoria, if only temporarily. Of course, each restoration of effect is followed by another "crash" in which users will have to decide, again, whether to continue or stop.
Cocaine bingeing is reported with all routes of administration (Cohen, 1989; Erickson et al., 1987; Siegel, 1982; Waldorf et al., 199l), but appears to be more common among cocaine smokers than sniffers. This makes sense because, even if the dose consumed through smoking is smaller than the dose sniffedand it often issmoking delivers the drug to the brain in a more concentrated form, producing first a more dramatic high and then a more dramatic "low" as the neural system responds to cocaine's presence. In a sense, smoking almost "tricks" the organism into responding to a relatively small dose of cocaine as if it were a large dose; as a consequence, the "crash" following the smoking of crack is likely to be more intense. Thus, as the accounts in Chapter 4 suggest, when using the drug, crack smokers may indeed experience more intense "craving" to continue bingeing than do cocaine sniffers and may, as a result, find it harder to resist the urge to binge, especially if the drug is readily available. In addition, because the duration of effect is shorter with smoking than sniffing, crack users are likely to consume more doses during a similar time period. This doesn't mean necessarily that they will consume more cocaine than sniffers do during a typical binge. What makes the crack binge more dramatic is that the transitions from euphoria to dysphoria are more frequent and more intense.
Whatever the association between bingeing and route of administration, bingeing per se is not evidence of drug dependence. Most people who meet the diagnostic criteria for cocaine dependence probably do engage m episodes of bingeing. However, both cocaine and crack users may binge occasionallyand experience "craving" and "compulsion" during the bingewithout becoming dependent (see Cohen, 1989; Waldorf et al., 1991). Of course, during periods of temporary abstinence, many regular cocaine users also report "craving" the drug. But these feelings probably have little to do with cocaine's pharmacological properties because they occur even long after cocaine's impact on the central nervous system has disappeared and are similar to what people describe when they give up other drugs (and even other activities) they enjoy. Indeed, we suspect that the craving linked to cocaine's pharmacological activity is short-lived, making it harder for cocaine users to resist bingeing, but not harder for themafter a drug-taking episode is overto resist using the drug again. The desire to use cocaine again is probably not pharmacologically linked to whether or not the preceding episode of drug taking was a binge or whether the drug was smoked or sniffed. To become "cocaine dependent," users must repeatedly decideduring periods of diminished or absent pharmacological effectto use the drug again.
Pharmacology Is Not DestinyAll the data gathered by NIDA since the 1970S show lower continuation rates for cocaine than for most other drugs. Among high school seniors who have tried cocaine, only 5.2% report having tried unsuccessfully to stop using ita lower percentage than for most other drugs (Bachman et al., 1991a). Most cocaine users take the drug occasionally and recreationallywithout experiencing compulsion, without bingeing, and without developing symptoms of drug dependence.
The likelihood that cocaine will be used in a dysfunctional way seems to be greater when the drug is smoked, sniffed, or injected than when it is swallowed. In this culture, the most common routes of administration are sniffing and smoking; and, controlling for other variables, smoking appears to be marginally riskierprobably increasing the incidence of bingeing, but not dramatically affecting whether current users decide to continue or cease taking the drug. Some people who use cocaine do become "dependent" on it, but many also, at some point, stop or reduce their use, often without obtaining drug treatment (Cohen, 1989; Erickson et al., 1987; Kandel et al., 1985; Shaffer and Jones, 1989; Siegel, 1980; Waldorf et al., 1991). No route of administration makes it easier (or harder) for "addicts" to overcome their "addiction," although there is evidence that crack smokers begin the process sooner (Millman, 1991; Washton et al., 1986)probably because their greater propensity to binge creates more of the problems that motivate drug users to change their behavior.
There is no evidence to support the claim made by Washton and others that continuous use of cocaine permanently alters brain cells in a way that "compels" people to keep using itor that smoking crack cocaine is markedly more addictive than sniffing cocaine powder. Instead, the literature shows that, even with direct modes of cocaine ingestion like crack smoking, use patterns and consequences vary widely. This evidence supports the theoretical perspective outlined in the beginning of this book: that abusive use patterns and addiction are more a function of the characteristics of certain users and certain social circumstances of use than of the drugs themselves (see Peele, 1985; Szasz, 1974; Zinberg, 1984). This is why all drugs, including cocaine and crack, show such enormous variation in patterns of use. In fact, as the following section discusses in more detail, research with animals shows that the more the conditions under which drugs are administered resemble the conditions under which humans take drugs, the more variation in their drug-taking behavior.
Animal Self-Administration of CocaineLaboratory scientists sometimes joke that the definition of a drug is any substance that, when injected into a rat, produces a journal article. Hundreds of studies have proven that laboratory animals can be taught to self-administer cocaine, even to the point of causing their own death. The earliest such studies, conducted in the late 1960s (Deneau et al., 1969; Pickens and Thompson, 1968) are important because they show that even drugs that do not produce physical dependence and withdrawal can be highly "reinforcing"; that is, after being administered the drug, lab animals can be made to self-administer more of it. Deneau et al., for example, demonstrated that monkeys would push a lever for cocaine over twelve thousand timesnearly as many times as physically dependent monkeys push it for heroin. By the late 1980s, over five hundred articles describing the reinforcing properties of cocaine had been published (Johanson and Fischman, 1989).
The assumption on which animal research is justifiedand repeatedly fundedis that much can be learned about human cocaine use from studying cocaine self-administration in caged animals (Bozarth, 1988; Brady and Griffiths, 1976; Fischman, 1988; Washton, 1989). However, to provoke animals to self-administer cocaine (and most other drugs), they must be "trained" to do so. In order to maximize the dose and frequency of use, researchers tether animals to the cage and surgically implant a permanent injection apparatus in their backs. This unreachable catheter injects cocaine intravenously following operant behavior (such as depressing a lever). Many researchers starve the rats before training begins because this increases the likelihood that animals will repeatedly inject cocaine.
But just as humans are typically distracted from drug use by other pleasures and life commitments, so are animals. Simply giving a cocaine-injecting rat a solution of water sweetened with glucose and saccharin decreases the injection rate (Carroll et al., 1989); so does maintaining rats on an adequate diet (Carroll et al., 1979). In addition, if instead of unlimited access, animals are given cocaine (or heroin) under conditions of limited access, they tend to arrive at a controlled daily dose and do not "choose drugs over life." In fact, in these settings, if the concentration of the drug is increased, the animals tend to administer fewer injections, holding constant their total daily doses (e.g., Wilson et al., 1971).
When animals are allowed to interact socially, their drug consumption also tends to decrease. In one series of studies, rats were trained to drink a morphine solution but then permitted access to an open area populated with other rats and scattered with objects for inspection and play. These opportunities for exercise, play, and socializing markedly decreased their consumption of morphine (Alexander et al., 1981). Environmental factors also affect the trainability of rats for cocaine injection-for example Schenk et al. (1987) found that rats reared in groups were less likely to self-administer cocaine than were rats reared in isolation
Studies of drug self-administration by rodents, dogs, and even primates have garnered much attention but have not contributed much to understanding cocaine use in humans. This is true because the conditions used in most animal studies are so extreme, so unlike the conditions of ordinary human life. In fact, experimental conditions are expressly designed to maximize animals' self-injection of cocaine. For example, test animals are raised in isolation or removed from social interaction with others of their kind. They are outfitted for solitary life and implanted with an IV injection apparatus. They are often starved to prepare them for their lives as cocaine "addicts" and almost always denied all opposing reinforcerseven sweetened water. And experimenters make unlimited supplies of cocaine constantly available. Thus, it is not surprising that researchers can train "nine out of ten laboratory rats" to inject themselves with lethal doses of cocaine (Bozarth and Wise, 1985). Such studies are then cited as scientific "proof" of cocaine's extreme addictivenessimplying that what is true for rats is also true for humans. This is the clear message in the Partnership for a Drug-Free America's "Dead Rat" video, which has been shown frequently on television.
The National Institute on Drug Abuse has paid for much of this animal research and continues to do sonow defining as a prime objective the discovery of a cocaine "antagonist" that will block or counter cocaine's effects and be useful for "treating" cocaine and crack addiction in humans (Leary, 1993; McNeil, 1992). This effort is premised on the idea that current "cocaine addicts" cannot stop using the drugan idea that is continuously reinforced by the animal self-administration studies. However, the accumulated data on human cocaine use show that most users do not become addicted to the drug, and, of those who do, most eventually stop or greatly reduce their use.
CRACK, COCAINE, AND PREGNANCYAnother core claim in the most recent War on Drugs concerns so-called crack babies. Among the earliest reports of possible fetal damage associated with cocaine was a study by a group of clinical investigators that appeared in the New England Journal of Medicine in 1985. Chasnoff and his colleagues found a higher than normal incidence of certain abnormalities among babies who had been exposed prenatally to cocaine and suggested that cocaine might be more harmful to fetuses than previously believed. Prior to this, cocaine had been largely ignored by researchers interested in drugs and pregnancy and, in fact, was not even discussed in either of the research monographs on fetal drug effects published by NIDA in 1985 (Chiang and Lee, 1985; Pinkert, 1985). However, interest in the topic grew quickly; and within a few years, dozens of articles on fetal exposure to cocaine had appeared, most of them reporting evidence of harm. However, few of the studies using human subjects used rigorous standards of scientific investigation, and those using animals (as we explain later) provide little insight into cocaine's effects in humans.
When pregnant laboratory animals (usually rats) are given cocaine, among the abnormalities noted in their offspring are low birth weight, eye and skeletal defects, cardiovascular malformation, delayed social development, impaired reflexes, increased shock sensitivity, and heightened reaction to painful stimuli (Fantel and Macphail, 1982; Finnell et al., 1990; Mahalik et al., 1980, 1984; Smith et al., 1989; Webster and Brown-Woodman, 1990). However, some of these effects have been found in only a single study; and, in some cases, researchers following similar experimental protocols have been unable to replicate earlier findings (Mayes, 1992; Neuspiel and Hamel, 1991). Thus, although taken as a whole, these studies indicate some vulnerability to cocaine among fetal rats, they do not constitute a "settled" body of research. The value of this research is limited by the fact that the animals are generally given extremely large doses of cocainesometimes twenty-five or more times (per kilogram of body weight) those typically consumed by humans. This is a serious methodological flaw because the effects of high-dose drug use are often not only quantitatively different from low-dose effects, but qualitatively different. In fact, if these animals had been given doses comparable to those consumed by humans, there might have been no adverse effects at all.
Even if adverse effects occur in fetal rats following doses of cocaine comparable to those consumed by humans, it does not necessarily mean human fetuses will be similarly affected. As shown in the literature on prenatal exposure to other drugs, fetal structure, function, and development are quite different in rats and humans (Juchau, 1976, 1985; Miller and Kellogg, 1985; Rudolph, 1985; Wang et al., 1985). For one thing, the human placentaunlike that of the rathas some capacity to metabolize drugs thus, although some active cocaine molecules are transferred from the pregnant woman to the fetus, they tend to be a lower proportion of those consumed than occurs in rats (Spear et al., 1989). In addition, human fetuses themselves have more drug-metabolizing enzymes than ratsgiving human fetuses a shorter period of exposure to whatever active cocaine they receive. Finally, because fetal development is more rapid in rats than humans, the impact of a single drug episode is likely to be more pronounced in rats than in humans. Because of these differences, effects found in rats exposed prenatally to cocaine might never occur in exposed human offspring.
To study cocaine's fetal effects in humans, researchers compare the babies of women who used cocaine during pregnancy to the babies of women who did not. However, only if the women in the two groups are otherwise similar can adverse pregnancy outcomesprematurity, low birth weight physical deformities, and the likebe attributed to prenatal cocaine exposure. In most comparative studies, women in the drug-exposed and control groups differ substantially. In fact, although cocaine use is well distributed across class and racial groups, the cocaine users selected for fetal impact studies are overwhelmingly poor and minoritywhich means they are less likely to have had adequate nutrition and medical care during their pregnancies and less likely to have healthy babies, whether they use cocaine or not. Further complicating the results of these studies is the fact that poor pregnant women who use cocaine are more likely than pregnant women generally to have an infectious disease and to use other drugs, particularly alcohol and tobaccoconditions known to contribute to fetal harm (Graham and Koren, 1991; Koren et al., 1990). Even the best designed of the cocaine and pregnancy studies control for only a few of these confounding variables, and many studies control for none (Neuspiel and Hamel, 1991). As a consequence, researchers have been unable to determine the magnitude of cocaine's impact on pregnancyor, indeed, whether cocaine has an independent impact at all.
Had this research been published at any other time, it might have gone unnoticed outside the scientific community. However, its appearance in the late 1980sat the height of the crack scarepractically guaranteed the attention of the popular press. In fact, although the studies themselves generally made no mention of the route through which pregnant women had consumed cocaine, journalists almost uniformly identified crack as the drug causing extensive fetal harm. By ignoring the methodological limitations in the scientific research, they presented preliminary data as fact.
THE CRACK BABY: A MEDIA-CREATED CRISISVirtually every adverse outcome found in every fetal study involving cocainewhether the subjects were humans or ratswas reported in the mass media as evidence that crack causes damage in babies. Journalists described "crack babies" as permanently impairedphysically, intellectually, and emotionally. Some of these babies, it was claimed, so lacked "normal human feelings" and "impulse control" that, as they matured, they were certain to pose a danger to others. Continually, Americans were told about the financial cost to taxpayers of the growing number of crack babiesmany of whom would need extensive medical treatment, special education, and long-term institutional care. Estimates of the magnitude of the "crisis" varied, but the media often quoted a Department of Health and Human Services report predicting one hundred thousand crack-damaged babies per year, at an annual cost to society of about $20 billion (Kusserow, 1990).
Journalists also continually portrayed crack babies as having been born "addicted" to cocaine. For example, one television news broadcast depicted a tiny African-American baby in an incubator waving his arms in apparently futile gestures as a voice-over described the horror of watching such babies "craving cocaine." In numerous magazine and newspaper articles as well, journalists described "tiny addicts" who were "poisoned in the womb" and then forced, at birth, into a "world of nightmarish withdrawal."
Of all the drug horror stories ever told, perhaps none has provoked as much public concern as that of the crack baby. In response, various remedial programs were implemented, particularly in the public schools, with the goal of helping crack babies compensate for their handicaps (Chira, 1990; Toufexis, 1991), but more commonly, a punitive approach has been taken. For example, hospitals now regularly test the urine of babies whose mothers they suspect of having used drugs, and babies are often taken away on the basis of a positive drug test alone (Siegel, 1991). In some parts of the country, women are prosecuted and imprisoned for using drugs during pregnancy (see Chapter 12 of this volume; Paltrow, 1992; Siegel, 1991), and state legislatures are searching for new ways to control pregnant drug usersfor example, laws that would force them, once detected, to choose between drug treatment and sterilization (Berrien, 1990; Chavkin, 1991). A recent survey of college students found widespread support for such policiesparticularly when the drug being used by pregnant women was cocaine (Vener et al., 1992)and probably most Americans would agree. Indeed, among defenders of drug prohibition, the goal of "saving crack babies" is now often offered as the primary justification for escalating the entire War on Drugs.
The "crack baby" on which drug policy is increasingly based does not exist. Crack babies are like Max Headroom and reincarnations of Elvisa media creation. Cocaine does not produce physical dependence, and babies exposed to it prenatally do not exhibit symptoms of drug withdrawal. Other symptoms of drug dependencesuch as "craving" and "compulsion"cannot be detected in babies. In fact, without knowing that cocaine was used by their mothers, clinicians cannot distinguish so-called crack-addicted babies from babies born to comparable mothers who had never used cocaine or crack (Hadeed and Siegel, 1989; Parker et al., 1990).
In the scientific literature itself, the issue of fetal damage related to cocaine is more complicated, but journalists have blatantly misrepresented that literature by reporting only studies that found evidence of harm  and then minimizing, if not ignoring, the limitations in their research design. The mass media have consistently portrayed crack as a direct cause of adverse pregnancy outcomes even though no study has convincingly shown that to be so. In fact, there is now evidence that cocaine actually contributes little to the abnormalities detected in the babies of women who use cocaine during pregnancy.
A number of people have criticized the cocaine and pregnancy studies, pointing out how biased sample selection and the lack of control over other variables prevent their being used as evidence that cocaine causes fetal harm (Alexander, 1990; Kandall, 1991; Mayes, 1992; Mayes et al., 1992; Neuspiel and Hamel, 1991). In addition, the few studies that have monitored cocaine-exposed babies during the first few years of life have found that the differences detected at birth almost disappear by age two (Chasnoff et al., 1992; Graham et al., 1992). However, the study we find most persuasive was done by a group of Canadian researchers who combined data from the twenty best-designed studies published prior to 1989 and performed a "meta-analysis" that challenges most of their findings (Lutiger et al., 1991). A meta-analysis is particularly useful when the results of similarly designed studies are inconsistent, as they are in this case. It also reduces the impact of selection bias, increases control over potentially confounding variables, and eliminates some of the problems of small sample sizethus permitting the use of more sophisticated statistical measures.
After combining the data from both drug users and controls, Lutiger et al. compared the reproductive risks associated with (1) polydrug use, including cocaine; (2) polydrug use, excluding cocaine; (3) cocaine use only; and (4) no drug use. Analyzing the data as a whole, they discovered that most of the fetal effects associated with cocaine disappeared. They did find significant differences between the offspring of women who had used drugs during pregnancy and those who had notbut both the type and rate of fetal abnormalities were similar regardless of the drugs consumed.
This latter finding is important because it calls into question the alleged harmful consequences of cocaine's vasoconstrictive impact on the umbilical cord and placenta. In sufficient doses, cocaine probably does restrict the flow of blood from mother to fetus, but because infants exposed prenatally to cocaine tend to be indistinguishable from those exposed to drugs that do not cause vasoconstriction, we cannot conclude that cocaine's slowing of the blood flow compromises fetal development. In fact, there is evidence that, in response to cocaine's presence, receptors in the placenta "down-regulate" fairly quickly, reducing vasoconstriction even before serum levels decline substantiallythus shortening the period of time in which blood from the mother is restricted (Wang and Schnoll, 1987).
We still know almost nothing about cocaine's interaction with the fetal brain, although the incidence of cardiovascular and central nervous system damage seems to be quite low (Neuspiel and Hamel, 1991). It has been suggested that the fetal neural system is more sensitive than that of adults and therefore more easily damaged by cocaine. But it is just as possible that the opposite is true. We know that fetal anatomy and function differ from those of adultsso much so that inferences about a drug's fetal effects can never be made on the basis of detected effects in adults (Miller and Kellogg, 1985; Rudolph, 1985; Wang et al., 1985). Some drugs are less harmful to fetuses than adults and some are more harmful; however, overall, human fetuses have proven to be remarkably resistant to the drugs consumed by their mothers (Alexander et al., 1985).
Given the recent increases in cocaine use and our failure to persuade some pregnant women not to take it, it is fortunate that the evidence to date does not suggest that cocaine is among the drugs that are particularly damaging to the fetus. This does not mean that cocaine use by pregnant women poses no risk. However, it is now clear that the high rate of abnormalities found in babies exposed prenatally to cocaine has less to do with the pharmacological effects of the drug than with other factors of high-risk pregnancy that "cluster" in drug usersparticularly impoverished drug users who more often have poor diets and no prenatal care and who are more frequent victims of violence against women and other crimes.
The route through which cocaine is administered probably makes little difference, although the greater use of crack by the inner-city poor means that crack users are more likely than powder cocaine users to have unhealthy babies. In addition, impoverished drug users are more likely than their wealthier counterparts to be enmeshed in a deviant lifestyle that carries with it many additional pregnancy risks. This association between crack use and adverse pregnancy outcomes will continue to exist as long as poor women are over-represented among crack users and as long as socioeconomic status remains a critical determinant of many non-drug-related pregnancy risks. Again, there is no evidence that crack is a direct cause of fetal harm, so reductions in crack use will not lead automatically to a reduction in the number of unhealthy babies being born.
CONCLUSIONPopular beliefs and attitudes about cocaine and crack have been shaped by journalists. Because the media are businesses seeking ever-larger markets of readers and audiences, they generally frame stories in ways that resonate with the sympathies and antipathies that make up conventional wisdom regarding drugs. In this sense, the crack story is simply the most recent installment in a series of morality tales that simultaneously construct and confirm Americans' belief in the power of drugs to disinhibit and harm users. However, there is something newor at least refinedin crack journalism: the emergence of a group of "drug experts" who use pharmacological language and concepts to support existing drug myths while ignoring pharmacological principles and evidence that challenge those myths. Some of the articles published in drug abuse and medical journals appear scientific but are not because the taken-for-granted premise of their authorslike that of most journalistsis simply that any crack use is highly destructive.
Our review of the available literature indicates that most of the claims that have been made about crack's hazards are either exaggerated or unfounded. In both powder and crack form, cocaine can be toxic, especially when consumed in large doses, and even small doses may produce harm in some users. However, most users experience no serious adverse health consequences related to their use. Cocaine also appears to be weak as a fetal toxin, and no physical or developmental abnormalities in infants can be attributed causally and specifically to maternal use of cocaine or crack. In both fetuses and adults, the relatively large safety margin associated with cocaine is probably linked to humans' extensive homeostatic responses to stimulant drugsprotective mechanisms confirmed by pharmacological science but rarely even mentioned by those interested in publicizing cocaine's harms.
Cocaine does not produce physical dependence, and babies are not born addicted to this drug. Numerous studies have shown that laboratory animals can be manipulated to self-administer cocaine repeatedly, but such studies provide very little insight into cocaine's addictive potential in humans. Among humans, cocaine addiction is relatively rare as a proportion of the total number of people who have tried it, regardless of the form in which the drug is employed. Early claims that smokeable cocaine caused instant addiction were clearly wrong. In fact, there is no evidence that the rapid onset/rapid decline of effect associated with smoking makes addiction or even escalated use inevitable. As Reinarman et al. suggest in Chapter 4, smoking may increase the likelihood that cocaine users will engage in bingeing. But it may also turn out that the problems associated with such bingeing may move crack users"drug dependent" or notmore quickly toward quitting or curtailing their use. Because the excessive use of a drug over a short period of time is likely to cause more individual and social dysfunction than moderate use over a long period of time, the tendency of crack users to binge means that crack can be viewed as more risky than powder cocaine. However, it is important keep in mind that many crack users take the drug occasionally, do not engage in prolonged binges, and do not become dysfunctional.
We have argued that the route of cocaine administration matters less than the public has been led to believea conclusion based on comparing smoking and sniffing, the two modes of ingestion most prevalent in American society. The practice of swallowing cocaine, although not free from abuse potential, almost certainly provides users with a substantially wider safety margin. Of course, swallowing is also a more "inefficient" way to consume a drug, and under a system of drug prohibition, such milder (and more "expensive") modes of ingestion tend to disappear. In this sense, the emergence of crack is part of a general trend that has been operating since cocaine prohibition was put into place early in the twentieth century. Fortunately, this more efficient mode of ingesting cocaine has not dramatically increased the risks associated with its use. Although there are risks involved in using crack, they have been consistently exaggerated. As the other chapters in Part I of this book demonstrate, most of the problems associated with crack are products of the social context in which it arose and is used, not its pharmacological powers or "efficient" route of administration.
NOTESThe authors acknowledge the valuable help of Lester Grinspoon, M.D., and Michael R. Aldrich, Ph.D.
1. For a more complete history of cocaine in the U.S. prior to the introduction of crack, see Grinspoon and Bakalar (1985). [back]
2. Some artists included cocaine in their work. For example, in the 1930s film classic Modern Times, Charlie Chaplin's "little tramp" accidentally consumed cocaine and became a hero when, energized by the substance, he single-handedly stopped a jailbreak. [back]
3. Inciardi (1992) notes that without the building of new highways in Peru making transport across the Andes Mountains easier, cocaine supplies to the U.S. could not have grown as they did. See also Morales (1989). [back]
4. For a more detailed description of freebase production, see Raye (1980). [back]
5. Injecting cocaine also produces "more effect for the money," but generally only highly committed users inject. Because crack is "smokeable," it appealed to users who are reluctant to inject any drug. [back]
6. For example, in July 1981, Time magazine ran a story identifying cocaine as a drug of the "rich and famous." [back]
7. Only recently have medical scientists become more interested in cocaine. In 1984, only 89 articles on cocaine were listed in the Cumulated Index Medicus. In 1989, the number was 426 and in 1992, 842. "Crack" was added as a separate category in 1992 and 45 articles were listed. [back]
8. See, for example, documents published by the National Institute on Drug Abuse (NIDA, 1990, 1991e, 1991f, 1991g, 1991h). All focus on the hazards of cocaine, none reports the drug's margin of safety, and none discusses the possibility of controlled use. [back]
9. Drug tales take on the character of folkloric horror narratives that can be told either with or without the drug theme. They usually focus on the drug user as crazed, dangerous, possessed of superhuman strength, sexually rapacious, and from a different social class or race than the teller of the tale (Brecher et al., 1988; Morgan and Kagan, 1980). [back]
10. We would make the same argument for nonpsychoactive drugs. For example, pharmacology can describe how aspirin works to reduce pain, but it cannot explain why some people choose to endure pain and avoid its use, why some people take aspirin at the first sign of pain, while others wait for pain to escalate; or why others take aspirin when they are not experiencing pain at all. [back]
11. An extreme example of the autonomic nervous system's capacity quickly to increase the release of neurotransmitters is when the organism is faced with danger and survival requires it to "fight" or "flee." Increased neurotransmitter activity stimulates the heart to beat faster and more forcefully and heightens the nervous system's responses to stimuli. [back]
12. Two other neurotransmitters that are known to play a role in operation of the autonomic nervous system are serotonin and norepinephrine. Less is known about their activity, but cocaine probably interacts with them, too, to produce or modulate its stimulant effects. [back]
13. All stimulant drugs work by interacting with this system, although in somewhat different ways. For example, the drug bromocriptine, which is used to treat Parkinson's disease, produces an effect by binding directly to and activating the nerve cell's receptors for dopamine. Other stimulants (e.g., amphetamine) work by entering the nerve cell and "displacing" the body's own neurotransmittersforcing their release at a faster pace and making more available for binding to receptor sites on neighboring cells. Cocaine works more indirectly, blocking the nerve cell's ordinary "reuptake" of neurotransmitters once they have performed their function and are released from the receptor site; these neurotransmitters thus remain in the space between cells (the synapse) and are available for additional activation of receptors. [back]
14. Drug molecules diffuse into cells based, in part, on their lipid solubility. The cell membrane is largely lipid (fatty) in character, and drugs behave as if they were dissolving in the membrane to pass through it. Thus, the likelihood that a drug molecule will enter cells is reflected in its likelihood to dissolve in lipid, nonhydrous solvents (ether, toluene, carbon tetrachloride). Plant alkaloids (compounds containing a nitrogen, including cocaine, mescaline, morphine, ibogaine, ephedrine, and atropine) generally have rapid cell penetration, making possible rapid psychopharmacological activity. [back]
15. Both time to onset of effect and peak concentration of effect are quite similar for inhalation and injection (Jones, 1984). Some cocaine users report a slightly quicker onset of effect with inhalation (Inciardi, 1992; Miller et al., 1989; Weil, 1986), but if this is the case, it is a difference of only a few seconds. [back]
16. Laboratory detection of the by-product benzoylecognine allows identification of cocaine users, through urinalysis, for up to seventy-two hours following useeven longer for heavy users (Weiss, 1988). [back]
17. Pharmacologists identify the time it takes for blood concentrations to decline to half a previous concentration as a drug's "half-life." Cocaine's half-life is thirty to sixty minutes, which also approximates its duration of effect. [back]
18. Among the known mechanisms of "acute tolerance" to cocainein the neural systemare "down-regulation" of dopamine receptors (leaving fewer available for binding) and "autoinhibition" of the nerve cell's dopamine excretion process. Outside the brain, other homeostatic adjustments occur; for example, a "baroreceptor" in the neck senses the rise in blood pressure caused by cocaine, as it does for other reasons, and relays a message to the brain to diminish cardiovascular activity. These mechanisms are effective enough so that, in experiments in which cocaine is continuously injected to maintain a constant blood concentration, both mood elevation and increased heart rate disappear within four hours (Ambre et al., 1988). [back]
19. Acute tolerance to a drug may develop during a single episode of use and occurs in the presence of bodily mechanisms that counter or compensate for the drug's effects. Cocaine produces an acute rather than a chronic tolerance. Therefore, when people stop using it, for even a few hours, responsiveness begins to return. However, the impact of additional doses of cocaine, consumed while the body is actively countering the effects of the earlier dose, will be diminished. By consuming continuously larger subsequent doses of cocaine, users might come close to re-creating the effects of the original dose, but at some point, acute tolerance may be so nearly complete that even extremely high doses produce little effect. [back]
20. Under the influence of cocaine, people may perform various cognitive and motor tasks more quickly and effectively. There is no research on cocaine's impacts on performance, and it is unlikely that it would be funded by NIDA or other government bodies. There is evidence that amphetamine and caffeine can enhance performance in athletics and other endeavors (Laties and Weiss, 1982; Weiss and Laties, 1962). [back]
21. See, for example, "Users of Crack Cocaine Link Violence to Drug's Influence," Washington Post, March 24, 1989, p. A11; "Capital Offers a Rare Market to Drug Dealers," New York Times, March 28, 1989, p. A1; "Crack Murder: A Detective Story," New York Times Magazine, February 15, 1987, p. 29; "Crack and Crime," Newsweek, June 16, 1986, pp. 16-22. [back]
22. For example, Fagan and Chin (1990) found few differences in the criminal histories of crack and cocaine powder users. Corman et al. (1991) found no evidence of an increase in homicide rates specifically related to the introduction of crack. And Goldstein and his colleagues in Chapter 6 of this volume show that almost none of the homicides identified by the police as crack related were due to its pharmacological effects; in fact, most were due to the unregulated (thus often violent) illicit crack market. [back]
23. Taken together, alcohol and cocaine generate production of cocaethylene, a toxic condensation product that, like cocaine, prevents the reuptake of dopamine and norepinephrine, thereby increasing their synaptic concentrations. That cocaethylene biodegrades more slowly than cocaine makes it potentially more dangerous than cocaine (Bailey, 1993; Jatlowetal., 1991). [back]
24. In a study of 137 cocaine users seeking admittance to an emergency room Derlet and Albertson (1989) found that the most common complaint (29%) was an "altered mental state." Most remaining patients reported at least one physical symptom commonly associated with cocaine, but this does not necessarily mean that they suffered physical harm or really needed medical treatment. [back]
25. This may be why nearly 60% of the cocaine users who entered emergency rooms in 1990 left without being admitted to the hospital (NIDA, 1991d). Because cocaine's effects wear off quickly, the symptoms that bring users to emergency rooms may disappear before they can be officially admitted. [back]
26. The additional problems include frequent changes in the panel of reporting hospitals; only recently has NIDA decided that the sampling adequately reflects national trends. In addition, we believe that the training of those who collect the data bias the process toward inflating drug mentions, as does the failure to confirm drug mentions toxicologically. We know, for example, that the availability of amphetamine look-alikes makes self-reports of amphetamine use unreliable (Morgan et al., 1987). [back]
27. For a more in-depth examination of possible reasons for the incongruence between the DAWN data and the national drug use data, see Harrison (1992). [back]
28. Using the DAWN data, Adams et al. (1990) and Gampel (1992) found that crack users were overrepresented, but this finding is questionable because, for a large majority of the emergency room visits attributed to cocaine, there was no evidence of route of administration in the record (NIDA, 1991d). [back]
29. This race/class difference can also be seen in the DAWN statistics: in 1988, although accounting for only 37% of past-month crack users and 15% of past-month cocaine users, blacks made up 48% of cocaine-related emergency room episodes (Gampel, 1992). [back]
30. Olds and Milner (1954), who first identified a structural substrate for "reward," showed that animals will repeatedly self-administer electrical shocks if electrodes are planted in certain areas of the brain. Because giving animals some stimulant drugs (including cocaine) causes a reduction in the voltage required to maintain self-administered shocks, it is assumed these drugs operate within the same system. For more recent research in this area, see Gardner (1992). [back]
31. Drugs that people do not experience as pleasurable have little "abuse potential." For example, antipsychotic drugs, such as chlorpromazine, are almost never used recreationally, and people who take them under medical supervision do not "crave" them when use is discontinued, even in the face of a withdrawal syndrome (Jaffe, 1990). In fact, it is often patients' unwillingness to sustain use of nonpleasurable psychoactive drugs, rather than patients' overuse, that is defined as a problem (Weintraub, 1975). [back]
32. For variations on this biochemical determinism theme, see Dackis and Gold (1985), Gawin and Ellinwood (1988), Gold et al. (1985), Nahas (1989), Spitz and Rosecan (1987), Washton and Gold (1984). [back]
33. For example, an article from Newsweek (June 16, 1986, p. 17) announces that "when smoked, cocaine molecules reach the brain in less than 10 seconds; the resulting euphoric high is followed by a crushing depression. The cycle of ups and downs reinforces the craving and, according to many experts, can produce a chemical dependency within two weeks." [back]
34. Other studies have confirmed these findings for cocaine and other stimulants (Johanson et al., 1976; Yanagita et al., 1973). To our knowledge, no later study has reproduced the twelve thousand lever pushes, and some studies of "extinction" have reported many fewer pushes prior to cessation (Griffith et al., 1979). [back]
35. This interaction experiment highlights the need for isolation in IV injection studies. The apparatus customarily employed will be inspected, bitten, and disrupted by another animal placed in the cage. [back]
36. The Partnership ad corrupts the animal study even further by depicting a rat just prior to death drinking water "laced with cocaine." Again, to cause rats to die from cocaine, researchers must limit their food, tether them to an injection apparatus, provide unlimited access to cocaine, and eliminate all alternative stimuli and activitiesconditions virtually never present in human life. [back]
37. There had, however, been some research on cocaine's fetal impacts in rats and mice (e.g., Fantel and Macphail 1982; Mahalik et al., 1980, 1984). [back]
38. Giving animals a dose comparable to that consumed by humans decreases substantially the probability that researchers will find a drug effect. Thus, they escalate doses to whatever level is necessary to achieve an effectwhich is one reason why they so often find effects. [back]
39. This quote comes from the Readers' Digest. See Yeager (1991). [back]
40. See, for example, the importance of the crack baby story in the remarks of antireform participants at the Hoover Institution s Conference on U.S. Drug Policy held at Stanford University in 1990 (see Hay, 1991; Peterson, 1991; Rosenthal, 1991). [back]
41. This may be typical of journalists. For example, Koren and Klein (1991) searched major newspapers for coverage of two radiation damage studies published in the same issue of the Journal of the American Medical Associationonewith positive and one with negative findings. The study finding harm was given considerably more attention. [back]
42. This result is all the more remarkable given that studies failing to find any adverse cocaine effects had been systematically excluded from the medical science literature (Koren et al., 1989). Thus, if Lutiger et al.'s findings contain any bias, it is probably in the direction of exaggeratingnot minimizingcocaine's effects. [back]
43. Although vasoconstriction is generally thought to be detrimental, it may be cocaine's initial vasoconstrictive action in the umbilical cord and placenta that protects the fetus from receiving an even larger dose of active cocaine from the mother. Thus, even if cocaine is potentially teratogenic, the actual occurrence of cellular damage, resulting in malformation, may be quite rare. Clearly, the incidence of congenital abnormality is much lower in humans than in ratswhich suggests that the capacity of the human placenta to metabolize drugs also plays a role in protecting human fetuses from harm. In fact, the only physical defect consistently found in cocaine-exposed babies is genitourinary malformation (Chasnoff et al., 1988; Chavez et al., 1989; Rosenstein et al., 1990)and even this relationship may be spurious because these studies do not adequately control for the use of other drugs. [back]
44. A similar conclusion has been reached regarding fetal abnormalities found in offspring of heroin users: the cumulative effects of the addict's lifestyle are more detrimental than heroin itself (Alexander et al., 1985; Forfar and Nelson, 1973; Neumann, 1973). [back]
45. Most of the research does not distinguish between crack and powder users. Wang et al. (1985) suggest that absorption of drugs through inhalation may be enhanced during pregnancy, but there is as yet no evidence that this alters a drug's impact on the fetus. [back]