U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Mack A, Joy J. Marijuana as Medicine? The Science Beyond the Controversy. Washington (DC): National Academies Press (US); 2000.

Cover of Marijuana as Medicine?

Marijuana as Medicine? The Science Beyond the Controversy.

Show details


Pain is the alarm of disease, the symptom that announces that all is not right with our bodies. Whether due to accident or illness, it is the most common reason that people seek medical assistance. But because pain has many causes, some of which are poorly understood, it is often a vexing problem to treat. There are no truly effective medicines for certain types of pain, and sometimes relief comes only at the expense of debilitating side effects. Thus, the search for new and better pain relievers, perhaps the oldest form of medicine, continues unabated.

Early in that pursuit, people discovered the pain-relieving properties of marijuana. It has since been used to treat a wide variety of painful conditions, from headache to the pain of childbirth. Many of the medical marijuana advocates who spoke at the public sessions held by the IOM—among them cancer and AIDS patients, migraine sufferers, and people with spastic and movement disorders—described how marijuana helped relieve their painful symptoms (see Chapter 2). Because marijuana is used to treat pain under such diverse circumstances and because the IOM team determined that marijuana appears to be a promising source of analgesic medications, the next chapter is devoted to discussing the performance of marijuana and cannabinoids in clinical studies of pain relief.

The nerve signals that our brains interpret as pain originate in receptor-bearing cells that become activated by temperature, touch, movement, or chemical changes in their environment. Pain signals travel to the brain by one of three main pathways, described in Box 4.1. Pain may be acute—short lived and intense—or chronic, persisting for days to years. For acute pain, such as the discomfort that follows surgery, doctors typically prescribe opiates: narcotic drugs derived from, or chemically similar to, opium. For chronic pain, however, opiates rarely bring relief. Even when they are effective, opiates often cause nausea and sedation that become a burden to the long-term user. At the very least, people with chronic pain develop tolerance to opiates over months or years and so must continually increase their dosage. Clearly, better pain medications would be welcome. Might marijuana be a source of these sought-after drugs?

Box Icon

Box 4.1

Types of Pain. Pain signals arise and travel to the brain by one of three main pathways, each of which produces different pain sensations: Somatic pain is the feeling most people imagine when they think about pain: a message sent by receptors located (more...)

Cannabinoids have shown significant promise in basic experiments on pain. Peripheral nerves that detect pain sensations contain abundant receptors for cannabinoids, and cannabinoids appear to block peripheral nerve pain in experimental animals. Even more encouraging, basic studies suggest that opiates and cannabinoids suppress pain through different mechanisms. If that is the case, marijuana-based medicines could perhaps be combined with opiates to boost their pain-relieving power while limiting their side effects.

But because of the ethical and logistical difficulties of conducting pain experiments on human volunteers, marijuana's potential to relieve pain has yet to be conclusively confirmed in the clinic. Only a few such studies have been conducted and only one since 1981. Most tested the ability of cannabinoids to relieve chronic pain in people with cancer or acute pain following surgery or injury. Unfortunately, few of these studies are directly comparable because the methods used to conduct them varied greatly and in some cases appear to have been less than scientifically sound. However, after critically reviewing existing research on THC and pain relief, the IOM team concluded that cannabinoids can provide mild to moderate relief from pain, on a par with codeine. The IOM team also determined that the body's own cannabinoid system likely plays a natural role in pain control.

By contrast, some clinical studies not only have failed to demonstrate that THC relieves pain but have also found that the drug has the opposite effect. In these experiments, volunteers who experienced painful shocks, heat, or pressure from a tourniquet reported that THC actually increased their sensitivity to pain.1 Another clinical study found that THC merely failed to relieve pain induced by either electrical shock or pressure, but the experiment was flawed in two respects.2 First, the researchers measured responses to extremes of pain, rather than to more typically painful sensations. Participants were exposed to shocks or pressure over a range of intensities but were only asked to note when they first felt pain and the maximum intensity of pain they could withstand. Since most people take medication for moderate pain, it would have been more useful to evaluate the ability of THC to relieve pain between the extremes that were actually measured (researchers commonly do this by asking participants to use a numerical scale to rate the pain they feel under various conditions). The second problem with this study is that the researchers failed to demonstrate that other painkillers could work under their experimental conditions. Without this standard of comparison, the results on THC have little meaning. They may conflict with those of other studies simply because of the methods the researchers used.

Design flaws also compromise a study that tested smoked marijuana's ability to relieve heat-induced pain in human volunteers. In this experiment, habitual marijuana users were hospitalized and allowed free access to marijuana cigarettes for a period of four weeks. During this time, volunteers consumed an average of four to 17 marijuana cigarettes per day and were tested periodically to gauge their response to painful heat applied to the skin. But since these tests were only performed “approximately every one to two weeks,” it is quite likely that the participants had already developed tolerance to the pain-relieving effects of THC by the time the tests were performed. It is therefore not surprising that THC failed to relieve pain under these conditions.

Two studies have examined the effectiveness of THC and levonantradol, a synthetic compound similar to THC, in relieving acute postoperative pain. In the first, volunteers who each had four molars extracted on separate occasions received the local anesthetic lidocaine plus one of the following treatments, given intravenously, with each successive tooth extraction: two different concentrations of THC, the sedative tranquilizer diazepam (Valium), and a placebo. Twenty-four hours after surgery the patients were asked to rate how much pain they felt during the procedure. Based on these ratings the researchers concluded that THC had no effect on surgical pain. There are several reasons to question this conclusion, however. Most importantly, the scientists once again failed to check whether another pain reliever, rather than a sedative, would have fared better than THC in the test. Lidocaine almost certainly diminished the patients' perceptions of pain, which were further compromised because they were not reported until 24 hours after surgery.3

The study on levonantradol is less problematic. Researchers gave the drug by intramuscular injection to 56 volunteers 24 to 36 hours after they were treated for injuries or underwent surgery. To eliminate the possibility that prior drug exposure would influence the patients' experience, the researchers did not test people who had a history of drug abuse or addiction or those who were taking prescription drugs that might interfere with their ability to perceive pain. On average, the researchers reported, patients who received levonantradol after surgery experienced significantly greater pain relief than those who got the placebo. The extent to which patients varied in their response to the drug is not clear, however. The authors do not reveal whether all patients who took the THC analog felt its effects to some extent or whether some people obtained great relief while others found it had little or no effect on their postoperative pain.4

The most encouraging—and believable—clinical studies of cannabinoids focus on chronic pain in cancer patients. Cancer causes pain in a variety of ways, including inflammation, nerve injury, and the invasion of bone and other sensitive tissue by growing tumors. Cancer pain tends to be severe, persistent, and resistant to treatment with opiate painkillers. For this reason, researchers hope to discover pain relievers that act on the body in a different way than opiates do.

In one such study, 10 patients with advanced cancer received THC pills in four different doses as well as a placebo. Each patient received the entire range of pills, which were identical in appearance, over successive days. On days when patients received the two highest doses—15 and 20 milligrams of the drug, as compared with 0, 5, or 10 milligrams—they reported significant pain relief. (By comparison, when patients take Marinol for AIDS wasting, an approved indication, they commonly take it in 5-milligram doses, with a maximum dosage of 20 milligrams per day. Marijuana cigarettes contain highly variable amounts of THC, typically between 30 and 150 milligrams, but much of that THC is lost in uninhaled smoke.) The study did not, unfortunately, compare THC with any other painkiller.5

Although they reported feeling less pain, patients who received the highest dose of THC in this study were also heavily sedated. They appeared dreamy and immobile; their thoughts were disorganized and they described feelings of unreality. Moreover, during the process of selecting patients to participate in the study, five of 36 volunteers became intensely anxious after receiv ing 10 to 20 milligrams of THC and as a result were excluded from the experiment. If this experiment is any indication, THC's side effects—though somewhat different—are as problematic as those of opiates.

Interestingly, during this study none of the patients experienced nausea or vomiting and more than half reported that their appetite increased, which suggests that oral THC acted as an antiemetic and an appetite stimulant, as well as a pain reliever. The authors also noted that some patients who appeared calmer after taking THC reported that it had not relieved their pain; other patients said that while their pain remained the same it bothered them less. These impressions resemble several anecdotal reports from marijuana users, who told the IOM team that marijuana did not take away their pain but helped them cope with their discomfort.

In a subsequent study the same researchers compared the effects of a single potent dose of THC with that of a relatively weak narcotic pain reliever, codeine. They found that 10 milligrams of THC gave the same pain relief as a 60-milligram (moderately strong) dose of codeine and that 20 milligrams of THC worked as well as 120 milligrams of codeine. The two drugs produced similar side effects, but THC appeared to be more sedating than codeine. On the other hand, patients tended to have a greater sense of well-being and less anxiety after taking THC than they did under the influence of codeine.6

Another group of researchers compared two conventional painkillers, codeine and secobarbital (a short-acting barbiturate), with a synthetic compound similar to THC. This THC analog had previously been shown to block pain in animals, so it was being tested for its ability to relieve moderate to severe pain in cancer patients. Both comparisons were conducted in cancer patients who suffered moderate to severe pain. In one trial 30 such patients were given three different treatments, in random order, on consecutive days: a moderately strong dose of codeine, a standard dose of the experimental cannabinoid, and a placebo. Patients then rated the intensity of their pain on a three-point scale (none, slight, moderate) every hour for six hours. The second trial, which compared the cannabinoid with secobarbital in 15 patients, followed the same procedure. On average, participants found that the THC analog relieved mild, moderate, and severe pain as well as the codeine and better than the secobarbital.7

In addition to the clinical trials already discussed, a handful of case studies and surveys have addressed the ability of marijuana or cannabinoids to relieve pain. The case studies are generally unconvincing, but survey responses suggest that marijuana—and by extension cannabinoids—can ease certain chronic pain syndromes. For example, in a recent survey of more than 100 regular marijuana users with multiple sclerosis, nearly every participant reported that marijuana helped relieve spasticity and limb pain (see Chapter 7).8 Likewise, many paraplegic patients interviewed in an earlier survey stated that smoking marijuana relieved phantom limb pain and headache.9

Similar anecdotal evidence has accumulated for the treatment of migraine headaches with marijuana, and marijuana is often mentioned as a “cure” for migraines. Yet the IOM team located only one scientific report on that subject published since 1975. It consists of a description of three cases in which people suffered migraines after quitting their daily marijuana habits.10 This is hardly convincing evidence that marijuana relieves migraine pain, since it is equally likely that the headaches were caused by withdrawal from the drug. Exploring the possibility of using marijuana-based medicines to relieve migraine pain will require rigorous clinical experiments designed to control for factors that can bias the results.

A possible link between cannabinoids and migraine has been revealed, however, in studies of cannabinoid receptors in the brain. These receptors occur in abundance in the periaqueductal gray (PAG) region, an area where migraines are suspected to arise. But it remains to be determined what effect cannabinoids exert on the PAG and whether they might prevent migraines from occurring. Such research would be worth doing since the best medicine currently available for migraines, sumatriptan (Imitrex), fails to provide complete relief for more than one in four of the patients who use it. An estimated 11 million people in the United States suffer from moderate to severe migraines.

Much of what medical scientists have learned about marijuana's pain-relieving potential warrants further study, according to the IOM team. A logical next step in basic research would be to determine whether existing cannabinoids could be modified to retain their analgesic properties while reducing or removing unwanted side effects such as amnesia and sedation. But some of those side effects may make marijuana an especially useful pain reliever. Cannabinoids appear to reduce nausea, vomiting, and appetite loss as well as pain. And the euphoric lift that attracts recreational users to marijuana could benefit people with anxiety-producing disorders such as AIDS or cancer. In fact, for that reason the IOM team recommended that researchers undertake clinical studies of cannabinoid medications among cancer patients on chemotherapy and AIDS patients suffering from wasting or significant pain. The IOM also recommended that the following groups of patients be included in such studies:

  • Surgical patients. In this case, cannabinoids should be administered along with opiates to determine whether cannabinoids reduce the nausea and vomiting associated with opiate painkillers.
  • Patients with spinal cord injury or other pain caused by nerve damage.
  • Patients with chronic pain who also suffer from insomnia.

All of the above patients are currently treated with opiate drugs, which produce tolerance and dependence as well as undesirable side effects. Could lower doses of opiates give these patients the same degree of relief when supplemented with cannabinoids? The answer lies in carefully conducted clinical experiments. Clinical trials could also determine whether THC is the sole—and, if not, the best—pain-relieving compound in marijuana. If additional cannabinoids relieve pain, researchers must then consider which cannabinoids or combinations thereof work best.

Although there has been very little clinical pain research on marijuana, the findings support positive results from animal and other basic experiments. Further clinical research appears to be well worth pursuing if it leads to a new class of drugs to complement existing painkillers or medications that could simultaneously relieve pain and nausea or appetite loss. The latter would be especially useful to people with AIDS and cancer, as described in the next two chapters.

But these future prospects offer little comfort to people with chronic pain that defies conventional treatments. Accordingly, the IOM researchers recommended the creation of an individual clinical trial program that would allow such patients to smoke marijuana under carefully controlled conditions for a limited period of time. Note that this is not the same as reopening the marijuana Compassionate Use Program that was closed in 1991 (see Chapter 11). As described in the IOM report, individual trials would be used to gather information to help develop alternative delivery methods for cannabinoid medications. Participants, who would be fully informed of their status as experimental subjects and the harms inherent in using smoking as a delivery system, would have their condition documented while they continued using marijuana. By analyzing the results of such trials, medical scientists could significantly increase their knowledge of both the positive and the negative effects of medical marijuana use.


1. Institute of Medicine. 1999. Marijuana and Medicine: Assessing the Science Base. Washington, DC: National Academy Press, p. 141.

2. Raft D, Gregg J, Ghia J, Harris L. 1977. “Effects of intravenous tetrahydrocannabinol on experimental and surgical pain: Psychological correlates of the analgesic response.” Clinical Pharmacology and Therapeutics 21:26-33.

3. Ibid.

4. Jain AK, Ryan JR, McMahon FG, Smith G. 1981. “Evaluation of intramuscular levonantradol and placebo in acute postoperative pain.” Journal of Clinical Pharmacology 21:320S-326S.

5. Noyes R Jr, Brunk SF, Baram DA, Canter A. 1975a. “Analgesic effect of delta-9-tetracannabinol.” Journal of Clinical Pharmacology 15:139-143.

6. Noyes R Jr, Brunk SF, Baram DA, Canter A. 1975b. “Analgesic effect of delta-9-tetracannabinol and codeine.” Clinical Pharmacology and Therapeutics 18:84-89.

7. Staquet M, Gantt C, Machin D. 1978. “Effect of a nitrogen analog of tetrahydrocannabinol on cancer pain.” Clinical Pharmacology and Therapeutics 23:397-401.

8. Consroe P, Musty R, Rein J, Tillery W, Pertwee RG. 1997. “The perceived effects of smoked cannabis on patients with multiple sclerosis.” European Neurology 38:44-48.

9. Dunn M and Davis R. 1974. “The perceived effects of marijuana on spinal cord injured males.” Paraplegia 12:175.

10. El-Mallakh RS. 1987. “Marijuana and migraine.” Headache 27:442-443.

Copyright 2001 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK224384


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...