Skip to main content

Verified by Psychology Today


Determining the Real Side-Effects of Drugs

How to make sense of the lists of drug side-effects.

There are many people who worry excessively about their health. They carry various diagnoses: hypochondriasis, somatization disorder, obsessive-compulsive disorder, and, often, depression. It turns out that these same people worry considerably about drugs and their side-effects, often to the point where they take untested, but “natural” substances, rather than drugs that the manufactures have spent hundreds of millions of dollars testing. I run groups treating these cases of “health anxiety.” There is an exercise I like to conduct in these classes. I take down a glossy pamphlet which is marked “Prozac” on the outside. I ask the group, “would you be willing to take a drug that has these side effects?” and I read to them out loud from the pamphlet:

Headache—15.5% of all the patients taking this drug.




Anxiety –5.5%



Dry Mouth—6.0%


Usually, by the time I get half-way through this list, they are shaking their heads, “no.”

Then I tell them that I have been reading from the list of side-effects in the placebo arm of the study! These are the side-effects patients get when they take a sugar pill that they think might be the real pill. Of course, my patients, who are especially suggestible, arestill more likely to develop these placebo responses.

So, that leaves the question, what are the real side-effects of Prozac? I read from a second list, the patients who were taking Prozac:

Headache—20.3% about one third higher than the placebo-responders. Therefore, there is a tendency (slight) for Prozac to cause headaches.

Nervousness—14.9% again, about one third higher than the placebo-responders. Therefore, there is a real tendency for the drug to cause nervousness. (Usually, this effect disappears in about 3 weeks.)

Nausea—21.1% vs. 10.1 % for placebo responders. There is, therefore, a distinct tendency for nausea to come from taking Prozac. (This symptom, too. is likely to go away after three weeks.)

Diarrhea—12.5% vs.7.0% for placebo responders, indicating a mild tendency for Prozac to cause diarrhea.

Sweating-- 8.4% vs. 3.8% for placebo responders. A very mild tendency to cause sweating.

All the other side effects for Prozac are close in range to the side-effects in the placebo groups, indicating that the drug does not cause these side-effects. (Any more than a sugar pill would.)

What does all this mean? It means, first of all, that someone who develops these side-effects upon taking a new drug may not be reacting to the drug itself! They may be reacting to the idea of the drug. Yet, if a patient is told that the particular physical reaction they are having after taking the drug cannot be reasonably attributed to the effect of the drug itself, they often become offended.

“This is real. This is not in my imagination.”

These symptoms feel overwhelmingly real, and they are real. All the symptoms described above in the placebo arm of the study are real; they are just not due to the pharmacology of the drug. They are real, and they must be managed, but they should not be managed by stopping the drug. There is nothing ridiculous about developing a placebo response. Everyone is subject to having these reactions under the right circumstances. If the patient is facing a serious illness, or if the drug is given by a charismatic doctor, or, if medical circumstances are dire in some other way, everyone will develop a placebo response to some extent. Sometimes these responses are positive and sometimes they are negative, causing side-effects. Studies have suggested that the areas in the brain ordinarily activated by a narcotic, for instance, are similarly activated by a drug the patient thinks is a narcotic. Similarly, if the patient starts off afraid of a bad effect from taking a drug, that drug is more or less likely to produce that result.

Since placebo effects often disappear after the first few days, it is not prudent to stop taking a drug right away because of these side-effects. There is plenty of time a few days later to stop the drug if side-effects persist. In my practice, I have often seen patients who have tried and stopped a number of drugs before giving any of them a chance to work. They end up thinking that they are especially sensitive to medications.

Often, I can get the patient to try once again to take a medicine related to those he/she has abandoned. If the patient stays on the drug long enough for it to work, I am likely later on to have similar problems taking the patient off the drug. Patients are then afraid to be off it as they were once afraid to go on it. In a similar sort of placebo response, the patient is likely to develop withdrawal symptoms unless there is a very slow and measured withdrawal. I have seen withdrawal symptoms when the half-life of the drug is so long, the level of the drug has not yet dropped at all. © Fredric Neuman 2012 Follow Dr. Neuman’s blog at

More from Fredric Neuman M.D.
More from Psychology Today