CA NORML Health Information

The California NORML/MAPS Smoking Device Study

[Article by Dale Gieringer printed in MAPS Newsletter, 1996; revised Nov., 1999]

Contrary to popular impression, waterpipes don't necessarily protect smokers from harmful tars in marijuana smoke, according to a study sponsored by California NORML (National Organization for the Reform of Marijuana Laws) and MAPS (Multidisciplinary Association for Psychedelic Studies). The reason is that waterpipes filter out proportionately more psychoactive THC than they do other tars, thereby requiring users to smoke more to reach their desired high. The study does not rule out the possibility that waterpipes could have other benefits, such as filtering out noxious gases, but it suggests that other methods, such as the use of high potency pot, vaporizers, or oral ingestion, are needed to avoid harmful toxins in marijuana smoke.

The study, which was supported by the Drug Policy Foundation and private donors, was conducted at a research lab with expertise in smoking. Researchers tested the smoke from seven different sources: a regular rolled joint, a joint with a cigarette filter, three different waterpipes, and two vaporizers, designed to heat marijuana to a temperature where psychoactive vapors form without producing smoke. The waterpipes included a standard bong, a small portable device with a folding pipestem, and a battery-operated model with a motorized paddle to thoroughly mix the smoke in the water. The first vaporizer, commercially produced in Canada, consisted of a battery-powered metal hot plate inside a jar to trap the marijuana vapor. The second was a home-made, hybrid apparatus, in which vapors were produced by a hot air gun and then drawn through a beaker of water, thereby combining vaporization with water filtration. The smoke was produced from standard NIDA-supplied marijuana drawn through a smoking machine adjusted to mimic the puff length of marijuana smokers.

The study focused on two key components of the smoke: (1) total solid particulates, or tars, which are noxious waste byproducts of burning leaf like those from tobacco; and (2) cannabinoids, the chemicals distinctive to marijuana, including its major psychoactive ingredient, delta-9-tetrahydrocannabinol (THC), and its two commonest chemical relatives, cannabinol (CBN) and cannabidiol (CBD), which are only weakly psychoactive but may have medical benefits.

Like tobacco, marijuana tars are rich in carcinogenic compounds known as polycyclic aromatic hydrocarbons, which are a prime culprit in smoking-related cancers. However, cannabinoids themselves are not carcinogenic. An obvious way to protect smokers' health is therefore to minimize the content of smoke tars relative to cannabinoids.

One way to do this is to increase the THC potency of the marijuana. Assuming smokers adjust their smoke intake to the cannabinoid dosage, the higher the concentration of cannabinoids, the lower the amount of tars they are likely to consume.

Another strategy is to try to reduce the tars in the smoke with some kind of filtering device. Obviously, this is beneficial only to the extent that THC isn't also reduced, thereby inducing users to smoke more to compensate. A major aim of the study was to determine the efficacy of various smoking devices at reducing the concentration of tars relative to cannabinoids. The performance of each device was accordingly rated in terms of the cannabinoid-to-tar ratio in its smokestream.

Surprisingly, the unfiltered joint outperformed all devices except the vaporizers, with a ratio of about 1 part cannabinoids to 13 parts tar. This disturbingly poor ratio may be explained by the low potency of the NIDA-supplied marijuana used in the study, which was around 2.3%.

Disappointingly, waterpipes performed uniformly worse than the unfiltered joint. The least bad waterpipe, the bong, produced 30% more tar per cannabinoids than the unfiltered joint. Ironically, the pipe with the electric mixer scored by far the worst of any device. This suggests that water filtration is actually counterproductive, apparently because water tends to absorb THC more readily than other, noxious tars.

Like the waterpipes, the cigarette filter also performed worse than the unfiltered joint, by about 30%. Researchers speculate this is because cannabinoids are exceptionally sticky and adhere to other solids. Hence, any filtration system that picks up particulates is likely also to screen out cannabinoids.

The vaporizer results appeared more promising, but confusing. The two vaporizers were the only devices to outscore unfiltered joints in terms of raw cannabinoid/tar ratio. The electric hotplate vaporizer did best, with a performance ratio about 25% higher than the unfiltered joint. The hot air gun was just marginally superior, but might have done better had it not been for its water filtration component.

However, the situation was complicated by the fact that the cannabinoids produced by the electric hotplate vaporizer were unusually high in CBN, leaving 30% less THC. Since CBN is not psychoactive like THC, recreational users might be expected to consume more smoke to make up for the deficit. (The situation may be different for medical users, who could experience other, medicinal benefits from CBN). For this reason, it seemed advisable to recompute the performance efficiencies of the vaporizers in terms of THC, rather than all cannabinoids. When this was done, the electric hotplate vaporizer turned out to have a lower THC/tar ratio than the unfiltered joint, while the hot air gun was still marginally higher.

The reason for the excess CBN from the hotplate vaporizer remains unexplained. Because CBN is produced from THC by chemical oxidation, it has been suggested that the device somehow exposed the sample to too much oxygen. However, there is no evidence that this was the case. As for the second, hybrid vaporizer, it seems likely that its performance could have been improved by deleting its water component.

The results clearly indicate that more developmental work needs to be done on vaporizers. Theoretically, an ideal vaporizer could minimize production of tars by holding the temperature above the point at which THC vaporizes, but below that where carcinogenic hydrocarbons are produced by combustion[01] In practice, both vaporizers produced over ten times more tars than cannabinoids, indicating that there is plenty of room for improvement.

In the late 1970s, a vaporizer known as the Tilt appeared on the market. According to the manufacturer, laboratory tests showed that it released 80% more THC and 79% less tar than a regular pipe, a performance ratio almost ten times better than any observed in this study. It is to be hoped that these impressive results can be replicated in the future. Unfortunately, the Tilt was withdrawn from the market in the early 1980s due to the passage of anti-paraphernalia laws.

As for waterpipes, the prospects for improvement appear more dubious. It has been suggested that the performance of waterpipes could be improved by using liquids other than water or by changing the temperature of the liquid. However, it seems doubtful whether such tactics would circumvent the basic problem of separating the tars from the sticky cannabinoids.

The study results are obviously discomforting to waterpipe enthusiasts, many of whom prefer the cooler, milder smoke they produce, and have naturally assumed it is also more healthful. Unfortunately, however, the study indicates that waterpipes may actually be counterproductive in increasing consumption of carcinogenic tars.

Nonetheless, it is still premature to judge that waterpipes are actually unhealthful, since they may filter out other, non-solid smoke toxins occurring in the gas phase of the smoke, which was not analyzed in the study. Noxious gases known to occur in marijuana smoke include hydrogen cyanide, which incapacitates the lung's defensive cilia; volatile phenols, which contribute to the harshness of the taste; aldehydes, which promote cancer; and carbon monoxide, a known risk factor in heart disease. Previous studies indicate that water filtration may be quite effective in absorbing some of these [Nicholas Cozzi, "Effects of Water Filtration on Marijuana Smoke: A Literature Review," MAPS Newsletter, Vol. IV #2, 1993]. If so, waterpipes might still turn out to have net health benefits.

In the meantime, the easiest way for most smokers to avoid harmful smoke toxins may be simply to smoke stronger pot. This strategy is apt to be more effective than any smoke filtration device. By simply replacing the low, 2.3% potency NIDA marijuana used in this study with high-quality 12%-sinsemilla, smokers could presumably reduce their tar intake by a factor of five while still achieving the same high. Further improvements could be had by using pure THC or hash oil, which has been tested at potencies of 60%.

The notion that high-potency marijuana is less unhealthful directly contradicts official government propaganda, which maintains that marijuana has become more dangerous since the '60s due to increased potency. This claim appears to rest less on scientific evidence than on the desire to frighten the public. A careful analysis of government data by Dr. John Morgan has shown that the supposed increase in potency has been greatly exaggerated ["American Marijuana Potency: Data Versus Conventional Wisdom," NORML Reports (1994)]. In any case, however, there is no good reason to presume that higher potency marijuana is more harmful, given the potential respiratory benefits of reduced smoke consumption. The hazards of excessive potency are purported to be an increased risk of acute overdose and greater susceptibility to dependency. However, both problems can be avoided if users adjust their dosage to potency. For most users, such hazards may well be outweighed by the benefits of reduced smoke consumption.

The Australian government is currently conducting another study that may cast further light on the effects of potency variations. The study is designed to determine baseline THC, tar, and carbon monoxide levels from marijuana and marijuana-tobacco mixtures smoked through joints and waterpipes. The samples being tested come from police seizures in six different Australian states. Researchers say that they have observed "incredible" variations in tar and THC potency among different samples. Their report is expected shortly.

The MAPS-NORML study provides additional information on the efficiency of different devices in delivering THC from marijuana to the user. Previous studies have shown that 60% - 80% of the THC burned in joints or waterpipes is lost in slipstream smoke, adhesion to the pipestem and bowl, pyrolysis, etc. [Mario Perez-Reyes, "Marijuana Smoking: Factors that Influence the Bioavailability of Tetrahydrocannabinol," in C. Nora Chiang and Richard Hawks, ed., Research Findings on Smoking of Abused Substances, NIDA Research Monograph 99, 1990]. The percentage of total THC delivered to the user is called the THC transfer rate.

The unfiltered joint scored surprisingly well in smoking efficiency, coming in second place with a transfer rate close to 20%. The portable waterpipe did slightly better, and the bong slightly worse. The other devices did notably worse. The vaporizers and electric waterpipe did especially poorly, with transfer rates less than one-third that of the top three devices. Thus heavy smokers could be literally blowing most of their money away with bad pipes!

Why Marijuana Smoke Harm Reduction?

The waterpipe study was undertaken as a first step toward marijuana harm reduction. It was motivated by concerns that, like tobacco, marijuana smoking poses hazards to respiratory health, such as increased risks of bronchitis, lung infection, and throat and neck cancers. These hazards are not caused by the psychoactive ingredients in marijuana, but by noxious vapors and solid particles, or tars, in the smoke produced by leaf combustion. In practice, these hazards can be eliminated by oral ingestion. However, smoking remains the most popular mode of consumption on account of its faster action, greater convenience, and easier adjustability of dosage. Surveys indicate that some two or three million Americans are daily marijuana smokers. Thousands of other Americans are currently smoking marijuana for medical purposes in the treatment of cancer, AIDS, glaucoma, and chronic pain and spasticity. To the extent their health is already compromised, such patients may be especially vulnerable to respiratory infections caused by marijuana smoking.

There accordingly appears to be a need for technology to reduce the respiratory hazards of marijuana. The most obvious fix is some form of smoke filtration device to reduce, suppress, or otherwise separate noxious byproducts from the cannabinoids in the smokestream. Among the vast variety of pipes, bongs, chillums, hookahs and other pot smoking devices available, three basic technologies are in use: water filtration, cigarette filters and vaporizers. The first two of these have been shown to be of some value for tobacco. However, there has been virtually no published research on any devices when used with marijuana. We accordingly undertook the present study to find out which if any currently available technologies were effective, and which offered most promise for further development.

Our research was guided by the philosophy of harm reduction, a concept popularized by advocates of needle exchange, methadone maintenance, and similar drug treatment measures. Harm reductionists hold that drug use is to some degree inevitable, so it is better to mitigate the harmfulness of drugs than to aggravate it through harsh and futile law enforcement efforts. So far, harm reductionists have focused on hard drugs, the major source of drug abuse problems. However, there is no reason harm reduction efforts should not also be applied to marijuana.

Unfortunately, research in marijuana harm reduction has been stifled by prohibitionist policymakers, who mistrust efforts to mitigate the adverse effects of drugs on the grounds they make illicit drug use acceptable. Not surprisingly, it proved impossible for us to interest the National Institute on Drug Abuse in supporting our project. They reminded us that the tobacco industry had spent billions developing a smokeless cigarette, only to withdraw it in the face of consumer distaste and active hostility from the anti-tobacco lobby. Sadly, the reduction of smoking-related harm is viewed as a threat by many anti-drug zealots, insofar as it undermines their rationale for prohibiting drugs in the first place. Thus the anti-drug lobby has actively impeded the development of marijuana harm reduction technology by lobbying for anti-paraphernalia laws, which outlaw the manufacture of devices for smoking controlled substances.

Harm reduction has equally little appeal to those marijuana enthusiasts who naively believe that marijuana, alone of all drugs, is a perfectly harmless herb. This delusion is quickly refuted by a review of the medical literature, which reveals extensive evidence of possible adverse effects of marijuana. From a physiological standpoint, these effects are mostly mild or of marginal significance, such as temporarily elevated heartbeat, slight and subtle impacts on immune cells, alleged changes in endocrine functioning, disputed and marginal influences on newborns, and so forth. Of considerably more consequence are the alleged psychological effects, including increased risk of accidents, impaired school and job performance, amotivation, heightened risk of drug abuse and sundry other social pathologies.

Nevertheless, from the standpoint of physical health, the single best established hazard of marijuana use appears to be an increased risk of lung disease from smoking. According to Dr. Lester Grinspoon, "After carefully monitoring the literature for more than two decades, we have concluded that the only well-confirmed deleterious physical effect of marihuana is harm to the pulmonary system.[02]

This should come as no surprise to any naive non-smoker who has exploded in a paroxysm of coughing after inhaling his first toke of pot. Chemically, marijuana and tobacco smoke are quite similar, aside from their psychoactive ingredients: both arise from the combustion of leafy material, which produces a host of noxious gases and solid particulates, or tars, that are known to be hazardous to respiratory healt[03] .Dating back to the British Indian Hemp Drugs Commission a century ago, observers have noted a high rate of bronchitis and other respiratory diseases among chronic ganja smokers in India, Jamaica and elsewhere[04] however, interpretation of the data has been clouded by the subjects' high rate of tobacco use, making it impossible to determine whether cannabis itself was responsible. This issue has been resolved thanks to modern clinical research by Dr. Donald Tashkin at UCLA, who has followed separate cohorts of marijuana-only, tobacco-only, marijuana-and-tobacco, and non-smoking subjects. Dr. Tashkin's work indicates that heavy daily pot smokers are more susceptible to respiratory disorders such as coughing, bronchitis, impaired lung immune function, and potentially precancerous cell changes[05]

In the last couple of years, there has also emerged epidemiological evidence of marijuana's respiratory hazards. A prospective study of 902 subjects by the Kaiser Permanente Center found that daily marijuana-only smokers had a 19% higher rate of respiratory complaints than non-smokers[06] They also found a 30% higher rate of injuries, perhaps reflecting an increased risk of accidents. Surprisingly, those subjects who had used marijuana for the longest time (>15 years) showed no increase in respiratory illness but a higher risk of injuries, while those who had used marijuana for less than 15 years suffered more respiratory complaints, but not injuries! The Kaiser study was not large enough to detect changes in mortality, but a larger study is in progress.

In the meantime, an important, unsettled concern is that of lung cancer. Despite the fact that epidemiological data are still insufficient, there is strong circumstantial evidence to suppose that marijuana smoking increases the risk of cancer, especially in the throat and upper respiratory tract[07] To begin with, the tars from marijuana contain most of the same carcinogens as tobacco, to a greater or lesser extent[08] It has been argued that marijuana is even more carcinogenic than tobacco because it contains some 50% more of the highly potent carcinogens known as polycyclic aromatic hydrocarbons, byproducts of incomplete combustion which are thought to be a prime culprit in lung cancer. In reply, hempsters contend that tobacco is more dangerous because it contains far more radioactive carcinogens, particularly polonium-210[09] However, this point seems moot in the light of experiments by the Leuchtenbergers and others, showing that marijuana tars, like those of tobacco, produce carcinogenic changes when applied to both animal and human lung tissue cultures[10]

The most compelling evidence of marijuana's potential carcinogencity come from recent clinical reports of throat and neck cancer in young marijuana-using males. This was first discovered by oncologist Dr. Paul Donald at the University of California at Davis, who in examining six patients who had contracted throat and neck cancer at the unusually early age of under 40, found that every one had a history of marijuana use[11] Although most of the patients also had other risk factors such as tobacco smoking or heavy drinking, marijuana use was the only one common to them all. Subsequent investigations by Dr. Donald and other oncologists have continued to find suspiciously high rates of marijuana use among younger throat, neck and tongue cancer patients, suggesting the possibility of a significant upsurge in upper respiratory tract cancers in coming years as the sixties generation ages[12]

The link between marijuana and throat cancer seems especially compelling in light of Dr. Tashkin's work, which indicates that cannabis smoke tends to concentrate in the larger, upper passages of the respiratory tract[13] In contrast, cigarette smoke is more likely to penetrate to the smaller, lower air passageways, where most tobacco-related lung cancers originate. It is still unclear whether pot plays a significant role in cancer of the lower lungs. However, Dr. Tashkin warns that the total tissue area in the upper respiratory passages is much smaller than that in the lower passages, so that pot smokers may well be exposing their throats to a proportionately much greater concentration of carcinogens. It is therefore possible that pot is a greater risk to the throat than cigarettes to the lungs. On the other hand, pot appears to be a much lesser factor in emphysema, which originates in the lower lungs.

It is tempting to try to compare marijuana and cigarette smoking. An exact comparison is hard to make, given that pot and tobacco affect different parts of the respiratory system differently. Anti-pot propagandists like to say that one joint per day is equivalent to one pack a day of cigarettes. This myth misrepresents a study by Dr. Tashkin, which found that one-joint-per-day pot smokers experienced a "mild but significant" increase in airflow resistance in the large airways greater than that seen in persons smoking 16 cigarettes today. However, the same study found that marijuana smokers did much better in other measures of respiratory health. A more accurate comparison based on studies by Dr. Tashkin's group is that pot smokers absorb four times as much tar in their lungs than cigarette smokers per weight smoked.[14] Given that a typical joint weighs about .4 - .5 grams, one-half as much as a tobacco cigarette, a rough equivalence is 2 cigarettes = 1 joint.

With this information in mind, we undertook to explore various ways of filtering marijuana smoke. Waterpipes were the most obvious candidate, being widely available in head shops and popular with many users on account of the apparent mildness of their smoke. We were especially encouraged by research showing that waterpipes could be highly effective in filtering tobacco[15]. Unfortunately, we were to discover that these results did not hold up for marijuana. A second candidate technology that would likewise prove disappointing was cigarette filters, which are widely available and can be easily adapted to marijuana by means of a simple homemade filter holder. We did not consider the more advanced "smokeless cigarette" developed by RJ Reynolds, due to the fact that it is not actually a smoke filtration device, but rather an inhaler for artificially flavored nicotine, which is of no use for marijuana. Instead, we turned our attention to vaporizers, which have been touted as a possible ideal solution to the cannabis smoking problem. Unfortunately, because vaporizers can't be used with tobacco, they are prohibited under US paraphernalia laws, and users must accordingly resort to homemade designs. We obtained one such device from the San Francisco Cannabis Buyers' Club. Another we managed to obtain from a Canadian supplier, who is selling them on that country's newly emerged, illegal but tolerated "gray market" in Vancouver. Although neither device performed close to the smokeless ideal, our study left reasonable hope that substantial improvement is possible. Given the evident need, vaporization merits further research and development.

[01] Contrary to the initial version of this article, which erroneously stated that THC vaporizes at 155º C, the Merck Manual lists the vaporization point of THC as 200º in vacuum. The vaporization point at normal atmospheric pressure appears to be unknown, but is thought to be in the range 250-400º.

[02] Lester Grinspoon and James Bakalar,Marihuana: The Forbidden Medicine (Yale U., 1993), pp.151-2.

[03] One important issue that has not been settled is whether there are significant differences in chemistry between the government-supplied marijuana leaf used in laboratory studies and the sinsemilla bud that has become popular in recent decades.

[04] Helen C. Jones & Paul Lovinger, The Marijuana Question (Dodd Mead & Co., NY 1985), pp. 16-21 .

[05] Donald Tashkin et al, "Effects of Habitual Use of Marijuana and/or Cocaine on the Lung," in Research Findings on Smoking of Abused Substances, NIDA Research Monograph 99 (1990).

[06] Michael R. Polen et al. "Health Care Use by Frequent Marijuana Smokers Who Do Not Smoke Tobacco," Western Journal of Medicine 158 #6: 596-601 (June 1993).

[07] Donald Tashkin, "Is Frequent Marijuana Smoking Hazardous To Health?" Western Journal of Medicine 158 #6: 635-7 (June 1993).

[08] National Academy of Sciences , Marijuana and Health (report by a Committee of the Institute of Medicine, National Academy Press 1982), pp. 14-7.

[09] Jack Herer, The Emperor Wears No Clothes, 1991 Edition, pp. 81, 114-5. The difference in radioactivity between cannabis and tobacco may have less to do with plant genetics than with differences in the soil and fertilizer used for the two crops.

[10] Cecile and Rudolf Leuchtenberger, "Cytological and cytochemical studies of the effects of fresh marihuana cigarette smoke on growth and DNA metabolism of animal and human lung cultures", in MC Braude and S Szara, ed: The Pharmacology of Marihuana (Raven Press, NY 1976).

[11] Paul Donald, "Marijuana smoking- Possible cause of head and neck carcinoma in young patients," Otolaryngology - Head and Neck Surgery; v.94: 517-21 (April 1986).

[12] Frank M Taylor, "Marijuana as a potential respiratory tract carcinogen: A retrospective analysis of a community hospital population," Southern Medical Journal 81:1231-6 (1988); Gideon Caplan and Brian Brigham, "Marijuana smoking and carcinoma of the tongue: Is there an association?" Cancer 66:1005-6 (1990); PJ Donald, "Advanced malignancy in the young marijuana smoker," Advances in Experimental Medicine and Biology, 288: 33-56 (1991).

[13] Donald Tashkin et al, "Effects of Habitual Use of Marijuana and/or Cocaine on the Lung," loc. cit.

[14] T-C Wu, D. P. Tashkin, B. Djahed and J.E. Rose, "Pulmonary hazards of smoking marijuana as compared with tobacco," New England Journal of Medicine 318: 347-51 (1988).

[15] D. Hoffmann, G. Rathkamp and E. Wynder, "Comparison of the yields of several selected components in the smoke from different tobacco products," Journal of the National Cancer Institute 31#1-6: 627-35 (Jul-Dec 1963).

CA NORML 2215-R Market St. #278 San Francisco CA 94114 (415) 563-5858