1/7 I've commented on this in various places over the past few days, so I'll summarize here:

The VA data show that the suicide rate among Veterans with a mental health diagnosis who are in treatment is higher than those not in treatment.

Is that concerning? Absolutely. https://twitter.com/Deedoherty2/status/1200629823351279616

2/7 However, the actual *number* of suicides is higher for those Veterans not enrolled in VA than those that are (because most are not enrolled in VA).

3/7 The VA reports that Whitaker references do not include ANY data on antidepressant prescribing. Yes, that can be part of treatment, but so can psychotherapy, suicide screening & treatment planning, and encounters without formal treatment per se.

4/7 Whitaker makes the leap to implicate antidepressants which is pure speculation without any data (from the VA reports) to back it up that claim.

5/7 Note that the highest rates of suicide per MH diagnosis are w/ opiate use d/o's which are not treated w/ antidepressants per se. Another piece of evidence pointing away from ADs as culprit (though there could be a connection w/ bipolar d/o where rate is also quite high).

6/7 Whitaker has made a career out of conflating correlation with causation & not considering the likelihood that those with more severe symptoms would be expected to be more likely to seek out & be in treatment, which could account for higher rate in tx rather than the reverse.

7/7 Name any intervention (X).

If you now compare the rate of the disorder it is supposed to treat and its associated morbidity, it will almost always be higher among people receiving X in a naturalistic study.

Does that mean X worsens or causes the disorder?

No, it doesn’t.