Worst Pills, Best Pills

An expert, independent second opinion on more than 1,800 prescription drugs, over-the-counter medications, and supplements

Codeine: The Drug With Multiple Personalities

Worst Pills, Best Pills Newsletter article June, 2008

Whether a patient gets the right amount of pain relief from codeine, inadequate pain relief or a dangerously increased effect depends on the person’s genes and whether they are taking other drugs that inhibit codeine’s effectiveness.

Below, Public Citizen explains the way codeine works and how the same dose can affect people in very different ways.

How does codeine work?

Codeine is routinely converted to morphine in the body in order for it to be an effective painkiller. The metabolism of codeine to morphine takes place in the liver through the actions of an enzyme called CYP2D6. Most people have normal CYP2D6 activity and their response to codeine is as expected. However, a substantial minority of people— differing by national origin and race (see below)—have CYP2D6 activity that is higher or lower than normal, potentially resulting in excessive (higher activity) or inadequate (lower activity) response to codeine.

What genetic factors are at work?

The effect of genetics on the CYP2D6 enzyme has been extensively studied, and we now know that total CYP2D6 deficiency occurs in about 6 to 10 percent of Caucasians, 3 to 6 percent of Mexican Americans, 2 to 5 percent of African Americans and about 1 percent of Asians Americans.

People deficient in CYP2D6, called “poor metabolizers” (PMs), have inherited a nonfunctional gene from each parent, and have no CYP2D6 activity.

The next group is called “intermediate metabolizers” (IMs) and they have inherited one nonfunctional gene; they have some CYP2D6 activity, but it is lower than normal.

Most people have at least one functional gene, and they have normal CYP2D6 activity. They are called “extensive metabolizers” (EMs).

Finally, there are some people who have higher than normal CYP2D6 activity, and they are called ultrarapid metabolizers (UMs). As with poor metabolizers, the likelihood of being an ultrarapid metabolizer depends on one’s ethnic background. Ultrarapid metabolizers are found in about 1 percent of people from Finland and Denmark, about 4 percent in Caucasian North Americans, about 10 percent of people from Greece and Portugal, about 20 percent in Saudi Arabia and almost 30 percent of people from Ethiopia.

How do these genetic differences affect codeine?

People who do not have any CYP2D6 activity (PMs) are not able to convert codeine to morphine; thus, codeine is unlikely to be an effective painkiller for these people. So if you have taken codeine for pain and it didn’t help, you may be in the PM category.

Intermediate metabolizers (IMs) who have some CYP2D6 activity may have some analgesic response to codeine, but it would probably be inadequate in most cases.

People with normal amounts of CYP2D6 (EMs) will generally have a normal response to codeine. Keep in mind that there is a substantial placebo effect when drugs are taken for mild to moderate pain, so it would be difficult to assess one’s CYP2D6 status based only on their response to codeine.

The potentially most dangerous situation is when codeine is used in ultrarapid metabolizers (UMs). Since UMs have increased amounts of CYP2D6—sometimes markedly increased amounts—they produce greater amounts of morphine. So UMs can have severe reactions to codeine with excessive sedation, severe constipation, and other side effects of excessive morphine levels in the blood.

What happens when codeine is used with drugs that inhibit CYP2D6?

Even if a patient has a normal amount of this codeine-tomorphine metabolizing enzyme, a number of commonly used drugs are capable of substantially inhibiting CYP2D6, and codeine is not likely to be effective in people taking such drugs. Indeed, studies have shown that codeine is not very effective as an analgesic in people receiving potent CYP2D6 inhibitors such as quinidine (DURAQUIN, QUINAGLUTE DURA-TABS, QUINIDEX) and other drugs. See the accompanying table for a list of CYP2D6 inhibitors; codeine is especially not likely to work well in people taking these drugs regularly. Note that one of these drugs (diphenhydramine, brand name BENADRYL) is a nonprescription medication. When one adds the number of people genetically deficient in CYP2D6 to the number taking medications that inhibit CYP2D6, a substantial percentage of people are at risk of a suboptimal response to codeine.


 Table 1. Drugs That Inhibit CYP2D6.

 Generic Name

 Brand Name

 Amiodarone  CORDARONE
 Bupropion  WELLBUTRIN
 Chloroquine  ARALEN
 Cinacalcet  SENSIPAR
 Diphenhydramine  BENADRYL
 Fluoxetine  PROZAC
 Haloperidol  HALDOL
 Imatinib  GLEEVEC
 Paroxetine  PAXIL
 Propafenone  RYTHMOL
 Propoxyphene*  DARVON
 Quinidine  QUINIDEX
 Terbinafine  LAMISIL
 Thioridazine*  MELLARIL

*Do Not Use in Worst Pills, Best Pills

What about other codeinelike painkillers?

While they are not as well-studied as codeine, other, codeine-related painkillers may not work as well if the patient is genetically deficient in CYP2D6 or if they are using the drugs listed to the right that can inhibit the enzyme. Such other codeine-related drugs include oxycodone (OXYCONTIN), hydrocodone (VICODIN), tramadol (ULTRAM) or combination drugs including any of these.

What You Can Do

For patients taking codeine or one of these related drugs who do not not have the expected painkilling effect, one possible explanation is reduced CYP2D6 activity. This decreased effectiveness should be reported to the prescribing physician so another pain-killer can be recommended. There is also some evidence of excessive toxicity from oxycodone or hydrocodone in ultrarapid metabolizers of CYP2D6, but more data are needed to establish whether this is a problem.

People taking codeine or related drugs for the first time should be alert for an excessive response (marked sedation, respiratory depression, gastrointestinal effects such as severe constipation) which might indicate that they are ultrarapid metabolizers of these drugs. Converserly, if these drugs are ineffective, the lack of the metabolizing enzyme or the concurrent use of inhibitors of the enzyme may be the cause. Until routinely available and relatively inexpensive CYP2D6 enzyme testing is widely available, excessive or inadequate responses to these drugs need to be reported to a physician.   


 Breastfeeding while taking codeine?

A tragic case was reported from Canada last year in which a healthy newborn infant began having lethargy and difficulty feeding when he was about 7 days old. On day 12, he had gray skin, and he died on day 13. When tests were done to measure morphine in the infant’s blood after his death, the amount was around 50 times higher than normal for a infant whose mother is taking codeine. Doctors also measured for morphine in the mother’s milk, which she had stored on day 10; the amount was much higher than expected.

Codeine has been considered safe for nursing mothers, and it probably is safe in most cases. But unless the mother has had genetic testing for CYP2D6 activity—and very few people have—she may be putting her infant at risk by taking codeine.

It would be prudent to simply avoid codeine in nursing mothers, or at least limit codeine use to no more than two or three days to avoid a build-up of morphine in the infant. If the mother does use codeine, the ideal approach would be to do genetic testing for CYP2D6 so that ultrarapid metabolizers can be identified and codeine avoided. Unfortunately, such testing is not readily available in most clinics.

In any case, nursing mothers and their infant must be very carefully monitored for toxicity if the mother is taking prescription codeine.