The Emperor’s New Drugs
Thursday, June 4th, 2015 | Books
In The Emperor’s New Drugs Irving Kirsch argues that antidepressant do not work. Well, sort of. What he argues is that the majority of the benefits of antidepressants are also found in placebo and that the additional drug benefit of antidepressants is explained by the enhanced placebo effect generated by the significant side-effects of the drugs.
He begins by talking about the meta analysis he did showing that almost all of the benefits from antidepressants come from the placebo effect. When he published this, the response of the medical community was “yeah, well we know that already”.
The problem with working out if antidepressants work is because it is really difficult to control for. You can add a placebo group, but because placebos cause less side effects it is easy to break blind. Indeed, 80% of patients on the drugs in one study were correctable able to guess they were in the test group. Also, the increased side effects could produce an enhanced placebo effect rather than a drug benefit.
Comparator trials, trials in which you compare with other antidepressants, are also problematic because everyone knows they are on an active drug, and therefore everyone can benefit from a strong placebo effect.
If his claims were true, you could surely use any drug as an antidepressant as it just has to cause side effects to work though? As it turns out, that is basically what we have. SSRIs, NDRIs, beta-blockers, stimulants, depressants, they all produce the same success rate in treating depression.
But surely NICE, the National Institute of Clinical Excellence, are here to protect us? In theory yes, but Kirsch claims they have trouble getting all the data they need. Drug companies publish studies multiple times, publish summaries across different studies and combine them in different ways to make it incredibly difficult for NICE to work out whether they have duplicate data in their own meta analysis.
He also attacks the idea that depression is caused by a chemical imbalance. If it is, we should be able to cause healthy people to become depressed by lowering their levels of serotonin and other monoamines. But we can’t.
Also, we should be able to effectively treat the imbalance. But again, we can’t, people just respond consistently no matter what treatment we give them. For example 60% respond to SSRIs and 59% respond to NDRIs. If we assume that is is a chemical imbalance, there is no way those numbers should be able to add up to more than 100%.
In summary, Kirsch argues that antidepressants to not out-perform active placebo and therefore subjecting patients to the significant side effects is unjustifiable.
I would like to read some rebuttals before I come down on one side or the other, but it is certainly a thought-provoking book.
A note for antidepressant users
In this blog post I have tried to set forward the views expressed in The Emperor’s New Drugs. I honestly do not know what to believe. However in April, I wrote a blog post in favour of antidepressants. I have thought about whether I should reference the two, and I think it is probably best that I add this disclaimer.
This book does not claim that antidepressants do not work. They clearly do. The question is whether their entire effect is due to the placebo effect, or whether they have a drug benefit on top. This is not that relevant for a user of antidepressants though. If you are using them, and they help, keep using them, because the placebo effect is a real, clinically proven, measurable health benefit.
And of course, discuss any medication chances or concerns with your doctor.
In The Emperor’s New Drugs Irving Kirsch argues that antidepressant do not work. Well, sort of. What he argues is that the majority of the benefits of antidepressants are also found in placebo and that the additional drug benefit of antidepressants is explained by the enhanced placebo effect generated by the significant side-effects of the drugs.
He begins by talking about the meta analysis he did showing that almost all of the benefits from antidepressants come from the placebo effect. When he published this, the response of the medical community was “yeah, well we know that already”.
The problem with working out if antidepressants work is because it is really difficult to control for. You can add a placebo group, but because placebos cause less side effects it is easy to break blind. Indeed, 80% of patients on the drugs in one study were correctable able to guess they were in the test group. Also, the increased side effects could produce an enhanced placebo effect rather than a drug benefit.
Comparator trials, trials in which you compare with other antidepressants, are also problematic because everyone knows they are on an active drug, and therefore everyone can benefit from a strong placebo effect.
If his claims were true, you could surely use any drug as an antidepressant as it just has to cause side effects to work though? As it turns out, that is basically what we have. SSRIs, NDRIs, beta-blockers, stimulants, depressants, they all produce the same success rate in treating depression.
But surely NICE, the National Institute of Clinical Excellence, are here to protect us? In theory yes, but Kirsch claims they have trouble getting all the data they need. Drug companies publish studies multiple times, publish summaries across different studies and combine them in different ways to make it incredibly difficult for NICE to work out whether they have duplicate data in their own meta analysis.
He also attacks the idea that depression is caused by a chemical imbalance. If it is, we should be able to cause healthy people to become depressed by lowering their levels of serotonin and other monoamines. But we can’t.
Also, we should be able to effectively treat the imbalance. But again, we can’t, people just respond consistently no matter what treatment we give them. For example 60% respond to SSRIs and 59% respond to NDRIs. If we assume that is is a chemical imbalance, there is no way those numbers should be able to add up to more than 100%.
In summary, Kirsch argues that antidepressants to not out-perform active placebo and therefore subjecting patients to the significant side effects is unjustifiable.
I would like to read some rebuttals before I come down on one side or the other, but it is certainly a thought-provoking book.
A note for antidepressant users
In this blog post I have tried to set forward the views expressed in The Emperor’s New Drugs. I honestly do not know what to believe. However in April, I wrote a blog post in favour of antidepressants. I have thought about whether I should reference the two, and I think it is probably best that I add this disclaimer.
This book does not claim that antidepressants do not work. They clearly do. The question is whether their entire effect is due to the placebo effect, or whether they have a drug benefit on top. This is not that relevant for a user of antidepressants though. If you are using them, and they help, keep using them, because the placebo effect is a real, clinically proven, measurable health benefit.
And of course, discuss any medication chances or concerns with your doctor.