About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Traffic Record | Main | A Nobel for Telomerase »

October 2, 2009

Placebos Can Work the Other Way, Too

Email This Entry

Posted by Derek

There's been a lot of valuable research into the placebo effect in recent years. That has interest in and of itself, and it also has a practical side. Understanding how people feel better on their own could tell us more about how to make our actual drugs work better, and it could also help us design clinical trials more efficiently. It would be a great help to know accurately how much of a positive effect is due to an investigational drug, without having to run thousands of people to separate that out statistically from a robust (but highly variable) placebo effect.

A new paper in the journal Pain (which has always gotten my vote for "Most To-the-Point Journal Title Possible") sheds some light on this issue, and on the mirror image "nocebo effect". The authors have looked over trials of several migraine drugs. In each case, there was a study arm and a placebo arm, and (since no one knew which group they were in), every patient got the lecture about possible side effects if you were in the treatment group.

The key point is that the migraine trials were investigating three different classes of drugs (anti-inflammatories, triptans, and anticonvulsants), and these three, not surprisingly, have different sets of possible side effects. The patients taking the drugs certainly manifested some of these, but what about the placebo groups?

Well, the placebo groups in the anti-inflammatory trials reported more dry mouth, nausea and vomiting than the placebo arms of the triptan studies. The placebo patients in the anticonvulsant trials, though, had a higher incidence of fatigue, sleepiness, and dizziness than the anti-inflammatory placebo groups reported. In short:

We found specific side effects in the placebo arms of anti-migraine trials when analyzing the three groups of drugs. We observed that the side effects that are expected for the active drug against which the placebo is compared, are also more frequent in the placebo group. In particular, anticonvulsant-placebos appear to have a higher rate of AEs (adverse events) than the other two classes of anti-migraine drugs. . .

. . .Moreover, it is also important to note that a larger number of patients in the anticonvulsant-placebo group discontinued the study (withdrawals due to AEs) than those in the triptan-placebo and NSAID-placebo groups. Both patients’ and experimenters’ expectations may have affected the AEs occurrence in the placebo groups. . .

This sort of thing has been observed before, but this is a particularly neat example. As a researcher (or a patient), it's important to remember that we tend to get what we think we're going to get. And we need to be aware of that, and be ready to correct for it if we have to.

Comments (28) + TrackBacks (0) | Category: Cardiovascular Disease | Clinical Trials | The Central Nervous System


1. Bill on October 2, 2009 9:04 AM writes...

So... Is it a logical extension to say that if a group hears that exposure to a certain chemical may give them some adverse effect (say hives) then some of them on exposure to what is really nothing (the placebo) will report hives?

Permalink to Comment

2. Vader on October 2, 2009 9:14 AM writes...

Yes, Bill, that's been observed.

I believe it underlies a lot of "allergies" and "multiple chemical sensitivities."

Clearly there are real allergies as well. I'm not as sure that there is a real physiological condition of "multiple chemical sensitivity" but am trying to keep an open mind.

Permalink to Comment

3. milkshake on October 2, 2009 9:27 AM writes...

and a placebo in a contraceptive pill study can even give you an unwanted pregnancy

Permalink to Comment

4. Mark on October 2, 2009 9:29 AM writes...

Where is this effect coming from? If the patients have no idea what class of drug they are taking (or not taking) and the people in contact with the patients have no idea, where does the effect arise?

Analysis of the data?


Permalink to Comment

5. Medicine General on October 2, 2009 10:37 AM writes...

It's good to be true, since many people need proper care and can not move especially if the case is of lower back pain or a terminal case, I think is the most advisable that this is so and benefit many people who happen thus, it must provide immediate solutions to the sick and targeted as does findrxonline with health issues.

Permalink to Comment

6. RB Woodweird on October 2, 2009 10:55 AM writes...

Clearly we must ban these dangerous placebos. I mean, has any clinical trial been done on just the placebo?

Permalink to Comment

7. less baffled on October 2, 2009 10:55 AM writes...

I wondered the same at first. However, the paper compared trials of three classes of anti-migraine drugs - there was no trial where a patient had the possibility of receiving drugs from more than one class. In other words, patients knew what side effects they would exhibit because they expected a particular drug (not one of several drugs)

Permalink to Comment

8. RB Woodweird on October 2, 2009 10:59 AM writes...

Clearly we must ban these dangerous placebos. I mean, has any clinical trial been done on just the placebo?

Permalink to Comment

9. DC on October 2, 2009 11:26 AM writes...

lol@ Woodweird

Permalink to Comment

10. DC on October 2, 2009 11:28 AM writes...

and wouldn't you then need a placebo, for the placebo?

Permalink to Comment

11. website on October 2, 2009 1:07 PM writes...

I would think one of the most significant factors would be whether the patients have contact with one another. Its well known that people are extremely suggestible and most will take on any malady someone near them claims to have. One person in an office building claiming to smell musty air and having trouble breathing can lead to an epidemic that evacuates the building for absolutely no reason apart from widespread idiocy and the extreme reliance on emotion to judge everything. Go around and tell 5 people at your place of work that youre feeling nauseous and I guarantee youll get some people feeling nauseous.

Permalink to Comment

12. retread on October 2, 2009 3:34 PM writes...

#12 -- quite true. However as an experiment recall that reading websites all day long causes hemorrhoids because of prolonged sitting. Think about it. Get up and stretch and move about. I'm sure you'll feel better. Write in and let us know what happened.

Permalink to Comment

13. drcharles on October 2, 2009 6:12 PM writes...

It brings to mind the maxim that the very process of observing something changes its outcome.

Permalink to Comment

14. Dewey, Cheatem & Howe on October 2, 2009 8:10 PM writes...

We at the law firm of Dewey, Cheatem & Howe are announcing the filing of a class action lawsuit against the makers of Placebos. We claim that Placebos have caused pain and suffering in people who relied on these medicines for relief and who believed they would be safe and effective. These patients must be compensated for their pain and suffering.

Permalink to Comment

15. ar-lock on October 2, 2009 8:20 PM writes...

Placebo effect is ubiquitous, it effects every action every day of your life, its your mind. Just try to wipe the slate free of preconceptions, its impossible

Have a nice day.

Permalink to Comment

16. Anonymous on October 2, 2009 10:35 PM writes...

Are all the possible side effects of all drugs listed when beginning the study?

In other words, if I were in a study that said that dry mouth was a possible side effect, but didn't mention sleepiness as a possible side effect, I would be more likely to notice and report the former than the latter, even if both occurred.

Permalink to Comment

17. REtread on October 3, 2009 8:14 AM writes...

A lot of the commenters appear to feel that they are somehow immune to the placebo effect (possibly from a sense of superiority to the mass of humanity -- see #11 "widespread idiocy" etc. etc. ). Those who got up thanks to comment #12 and felt better, should know that they just experienced the placebo effect, as there is no evidence (to my knowledge) that prolonged sitting produces hemorrhoids.

Permalink to Comment

18. DrSnowboard on October 4, 2009 9:23 AM writes...

By coincidence, I read the placebo chapter in Ben Goldacre's Bad Science yesterday (not spam, I really did). A very good summary of human susceptibility to influence by number of tablets, colour of tablets etc and all referenced to the lit.

Permalink to Comment

19. CMCguy on October 4, 2009 9:47 AM writes...

DrSnowboard your comment brings to mind many complaints I have heard from Clinical types as to the "noise" in human trials and all the subjective factors that come into play. There seems to be arguments back and forth on the reliability and objectivity of animal studies vs humans since the latter can report many things that hard to know about in non-verbal species. Certainty comes into play when creating drug labels which is challenging enough without placebo effect observations.

Permalink to Comment

20. Still Scared of Dinosaurs on October 4, 2009 9:54 AM writes...

There is more than one type of "placebo effect" relevant to the analysis of clinical trial data, and the effects of patient expectations is rarely the most important (though it must be noted that this effect is strongest in pain trials).

The most important is so simple most people ignore it - the fact that most conditions resolve on their own. Think of FluMist - without it flu resolves in 7 days, with it, in 6 days (or something like that). You need a placebo group to detect the difference and probably a lot of patients to reach significance, but no expectations need apply.

In addition, the effects of differing enrolment criteria has an effect on the AEs you'll see in a placebo group. The worse off patients are the more AEs of all types you'll see. In this context the comparison I'm interested in is not between studies where patients were warned of a condition in one trial but not the other but in the rates of AEs between placebo groups for AEs that were not mentioned in any of the trials. This can also be confounded by differing length of follow up - follow people long enough and you'll approach 100% for any AE, even death.

Permalink to Comment

21. Retread on October 4, 2009 10:43 AM writes...

It's worth remembering that, in the case of pain, naloxone (a narcotic == opiate antagonist) blocks the placebo effect. This implies that the placebo effect for pain is using endogenous mechanisms. Presumably a well trained mind could learn to do this without exogenous opiates. Has anyone tried biofeedback to produce this effect? If so, I'm unaware of it. It seems like a worthwhile thing to study.

Naloxone blockade of the placebo effect on pain is pretty well established both clinically and with fancy tests such as functional magnetic resonance imaging (fMRI), positron emission tomography (PET) using 11-C labelled opiate receptor ligands which emit positrons [ Neuron vol. 63 pp. 533 - 543 '09 ] will get you started on this literature.

Permalink to Comment

22. Kieth on October 4, 2009 5:30 PM writes...

Some years back Holman Jenkins (columnist for the WSJ) wrote a short piece commending the virtues of placebos and making the point that for all the good they did, the monetary investment was zero for research and negligible for manufacturing. He did not foresee the perils of AE.

Permalink to Comment

23. Jose on October 5, 2009 9:07 AM writes...

[Rant} At the risk of the becoming the "grouchy old man yelling at the kids from the front porch"-guy, I am again shocked at the current C&EN. It is a disgrace that the pharma industry is ramping up for a dual wave of massive, maybe unparalleled, layoffs, via two mega-mergers, and yet we get articles on boutique outdoor gear! and and phunny philosophy! and and witty chemical sculptures! wheeeeeeee! Thanks for nothing, clowns; thanks for fiddling while Rome burns. [Rant off]

Permalink to Comment

24. dave on October 5, 2009 4:16 PM writes...

what a poor choice to use FluMist as an example here. FluMist is not a drug that you take once you're sick, but it's a vaccine that you take well in advance to prevent getting sick!
Not sure what placebo effects look like in vaccine trials...would be kinda interesting...
It's hard to see how someone getting the placebo vaccine would still feel like she is protected from, say, chickenpox even though she fully contracted the virus and has red dots all over her body... :-))

Permalink to Comment

25. Dave L on October 7, 2009 9:06 AM writes...

Placebo = the power of suggestion

By the way, what's the true baseline in these studies (without suggestion). For example, I felt pretty nauseated this morning for no apparent reason. If I happend to coincidentally be in a study, I'd mark down that I was feeling nauseated. Take 1000 people...put them in a room...and have them mark down how they feel. How many are nauseated, have sore joints, have dry mouth, have a headache?

Permalink to Comment

26. Rich on October 8, 2009 9:39 AM writes...

Dave L - I was feeling fine until you asked - now you've made me aware of myself, I am feeling more tired today than usual and I think a headache might be building.

Permalink to Comment

27. Glenn on October 15, 2009 3:53 PM writes...

All I can say, is that when my job review comes up .. if my boss starts of with endless praise .. it guarantees there is no raise possible. If it starts out with a lot of criticism .. I know that I'm up for a raise and it's her job to make sure I do not feel I have access to it.

Placebo is the same thing -- it is all about conditioning your expectations and perceptions of the "reality" -- and once that situation has been made real (to you) the outcome will follow that path.

A hypnotist I bet would say it is a simple form of mind control.

Permalink to Comment

28. Dr. Z on March 18, 2010 2:45 PM writes...

It seems to me this effect comes from poor experimental design. Obviously the investigators explained only these side effects to each group that were associated with the particular class of drugs. What they should have done is make a union set of all side effects of all drug classes and then explain all of them to all participants. In this way the placebo effect bias will be at least systematic and randomly and uniformly distributed. To subtract the placebo effect after the experiment will be easy, just do statistical analysis comparing the expected density distribution of the placebo effects (uniform) to the real distribution. If the real distribution is uniform too, there are no side effects. If it is not, the discrepancy will show where the side effects are. Of course, this all holds for large enough group of participants, randomly picked up from different places etc. One have to make sure that the "freak out about imaginary side effects" probability of each member of the test population is about the same :).

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry