I.

The New York Times reports the story, which of course is only half the story.

According to the FDA website, an analysis of almost 200 clinical trials of 11 antiepileptic drugs finds that the suicide rate is almost double that of placebo. 

On first glance, this seems like the FDA went to the semiotic school of medicine:  "if SSRIs double the rate of suicide, and SSRIs are antidepressants, then antiepileptics, which are also antidepressants, should also double the rate... holy crap, we'd better check."

Popper would be appalled, but psychiatry lives in this world, and so does the FDA.  It's insane, but if that's their  worldview at least they should be true to it and look for suicide in "logical" places.

Nope.  Not even honest in their wrongness.  Semiotics isn't what drove the FDA to these conclusions-- it was even more basic than that.

Lawyers.

One lawyer, anyway.  Andrew Finkelstein, personal injury lawyer from NY, was handling some Neurontin suicides.  How do you make big money?  First, you send the FDA all the adverse event reports you can get; then ask them to put a black box warning on it.  Black box warning is gold: it means the suicide link exists, it's true-- indisputably so (preposterousness of it aside, of course.)  Sprinkle in a dash of off label promotion by Big Pharma and you have yourself a lawsuit masquerading as a self-righteous public service announcement...

It helps to pretend to be for the little guy: "The FDA's complete inaction in protecting the health and safety of United States' citizens from a known serious risk..."

And so, the FDA got cracking...

II.

Let's take a look.  The FDA report finds the risk of suicidal thoughts and behaviors is double that of placebo-- just like the doubling of suicide rates for antidepressants.

But not really.  The rate increased from 0.22% placebo to 0.43% for anticonvulsants.  That's tiny.  But it's also not comparable to antidepressants, because the number needed to harm for antidepressants is 50, for anticonvulsants it's 500.  In other words, antidepressants are ten times more likely to cause suicide than anticonvulsants.

In the 28,000 people in the seizure drug arms, only 4 committed suicide (none on placebo did.) 

I'm not buying the increased suicide risk with SSRIs, but I am sure as hell not buying this nonsense for anticonvulsants.

III. 

So, why did it take 4 years to release the data?  (Finkelstein has been at them since at least 2004.)

I'm not sure I should even mention this because I do not fully understand it, but I will, if only to give a starting point to others with more time.

Russell Katz, MD is the FDA's chief of neurologic drugs; it is his report on anticonvulsants and suicide that is on the FDA website.  He also handled the antidepressant-suicide investigation.

According to an old story (2004) from my San Francisco Chronicle, Dr. Katz may have been involved in suppressing/delaying the release of the antidepressant-suicide data.

Make of that what you will.  I'm going to bed.