One of the main arguments for continuing drug treatment for depression, psychosis and bipolar disorder is that you will get worse from stopping the drugs, especially if they are stopped abruptly. These are findings from mainstream psychiatry. However, if we combine this information with the methodology of the randomized controlled trial, we may see that these drug trials do not show efficacy of drugs, and may not be usable to show safety. The positive side to this is that the trials may actually demonstrate the healing power of our own minds.
When drugs are tested, patients are given a placebo for 4 to 14 days in order to wash out the drug that they have been taking before the trial. Patients do not know that they are getting a placebo, and many of them may think that they are lucky and have been given the new wonder drug.
During this washout period, many individuals react very well, so well that they do not have enough psychological problems to enter the real drug trial. These are so-called placebo responders. They are removed from the trial, since they are not depressed or psychotic enough to take part in the research project.
This seems like it is going to give the drug and unfair advantage. However, if the drug group and the placebo group are of the same size, this does not actually provide an advantage for the drug. But something much more significant is happening.
The placebo washout period is used to ensure that all participants have brains free from the previous drug they were taking before they start treatment with the new drug.
That means that at the end of the placebo washout period, all the patients are in acute withdrawal. One would reasonably expect these patients to do quite poorly, to be quite disturbed. Some of them are , but many who had a problem when they were on the drug, actually get well from stopping the drug cold turkey.
However, those who react badly to the cold turkey withdrawal get to continue. Half of them then get a drug very similar to the one they had abruptly stopped, and the rest, the placebo group, get to continue on cold turkey withdrawal.
So this is what we are testing: Difference between continuing cold turkey withdrawal and getting back a drug very similar to the one you were dependent on.
So the classic RCT is not at all testing the difference between drug and no-drug.
This may actually explain some strange effects seen in drug trials:
- The positive effects of the SSRIs start immediately, even if doctors tell their patients that the drugs have to be taken from 4 to 8 weeks to feel better
- Patients in the placebo group get a lot of negative effects such as sleeplessness, and inner turmoil (akathisia). One would not expect such negative placebo effects (nocebo) from a pill that patients think will actually relieve their problems.
- Effect only in the severely depressed. Many doctors seem to accept that SSRIs don’t have an impressive effect on the moderately depressed, but justify their use on the severely depressed because of a larger difference between placebo and drug for this group. We know that patients in this severely depressed group must have taken larger doses of their previous drug for longer times. It is then quite obvious there will be more severe withdrawal reactions in the placebo group. The group getting back a similar drug will be even more relieved, since they were very addicted to the previous drug.
- Drugs not working in children and adolescents. This may be the only case where participants have not been on a similar drug before they start the trial, and here we don’t see a difference between drug and placebo.
So what is being tested in an RCT is not at all the difference between a drug and an inactive pill, we are testing how good it feels to get back something very similar to the drug you were dependent on, compared to going cold turkey! Anybody who has had a serious hangover and then “repaired” by taking a few drinks knows this difference. The relieving drug doesn’t have to be the same: Acute alcohol delirium can be cured with benzodiazepines since they are similar to alcohol in effect on the nervous system.
So the first conclusion may then be that none of these RCTs prove that the drugs work.
The second conclusion may even be more startling: We cannot say anything about the safety when we compare a drug to abrupt withdrawal.
One of the reasons drug companies give for not leaving depression untreated, is the the untreated patient may die from suicide. One could then think that taking the drug away from a patient would increase the risk for suicide. So according to drug company logic, a patient who has abruptly stopped the life-saving drug, would have a higher risk for suicide than the medicated patient.
This is confirmed in a very large study reported in Journal of Clinical Psychiatry in 2009. The abstract states:
“Antidepressant discontinuation showed a significant risk for suicide attempt as did the period of an abbreviated trial, that is, stopping before a therapeutic regimen of 56 days had been reached. The highest risk was associated with initiation, a finding consistent with other studies, closely followed by periods of dosing changes and discontinuation.”
Slowly increasing the dose (titration up) increased the risk to 262% of normal risk. Decreasing dose gradually lead to a 219% increased risk according to the article.1
This means that patients on placebo have at least a doubled suicide risk after 2 weeks in withdrawal. The 219% risk was on gradual withdrawal, not abrupt, which would possibly be much worse.
RCTs compare suicidal ideation on the drug to placebo. Many studies find a doubled risk. One study found a 6-fold risk for Paxil2. Since the risk in the placebo group is already doubled, the increased risk from the drug may be 4 to 12 times the risk of unmedicated depression. The conclusion is difficult to avoid: antidepressants are extremely hazardous to our health.
However, the flipside of this problem is quite positive: avoiding medication for depression is associated with much less suicide!
And now we come to the truly amazing power of hope in the form of a placebo. Let us take a typical antidepressant trial with some simple numbers: 100 participants in the placebo group and 100 in the drug group. Let us be optimistic for the drug and assume that 60% get well. 50% typically get well on the placebo. This means that 100 out of 200 patients actually would get well from the placebo, since 50 patients in the drug group also are expected to get well from just popping a pill whether it is sugar or drug. This means that only 10% got anything out of the drug. The rest get well on a placebo. The results are thus that placebo is 5 times more effective than the drug.
If we combine this with our first observation that placebo treatment is actually abrupt withdrawal, it becomes really amazing. Going cold turkey off the previous medication makes 50% of patients well, often instantly in the case of those who respond in the washout phase. This is very bad advertising for the makers of the previous drugs. And the new drug is only one fifth as effective as cold turkey.
There are so many other things wrong with RCTs, such as publishing only the positive trials, that we cannot even be sure of this 10% drug effect.
In addition to this, even the very pro-drug Dr. Gibbons’ meta-analysis of industry-sponsored trials showed that the placebo reached the exact same level as the drug just four days later. All this shows that we can trust our own minds to heal themselves as long as we don’t add chemicals.
To conclude: From this analysis it becomes clear that we don’t have ANY research showing the effects or safety of SSRIs since we are testing cold turkey withdrawal against getting back a me-too drug. The research shows the opposite of what the drug makers hoped to find: placebo is extremely effective and has no side effects. Hope is curing us.
* * * * * *
- Valuck RJ, Orton HD, Libby AM. Antidepressant discontinuation and risk of suicide attempt: a retrospective, nested case-control study. J Clin Psychiatry. 2009 Aug;70(8): 1069-77. doi: 10.4088/JCP.08m04943.