Hacker News new | past | comments | ask | show | jobs | submit login

The book you should read on this is Robert Whitaker's Anatomy of an Epidemic. Whitaker's book goes the history of academic studies on antidepressants (as well as benzos and neuroleptics), and after extensive analysis of these studies concludes that over the long term antidepressants make depression worse. I know people are quick to talk about people who take antidepressants and feel so much better, but 1) placebo effect--note that many studies show antidepressants have very little benefit over placebo 2) this is in part explained by the analysis of when antidepressants start to worsen the course of the illness (which is ~2 years for neuroleptics...unsure about antidepressants) 3) people who do not take any drugs and don't have a "chronic" illness but rather a one-time episode (which Whitaker notes was the natural course of the illness before antidepressants) don't consider themselves "depressed" people and thus the sample size is skewed. edit: also, there's no control or comparison group. Because there's an assumption that the drugs just must do good worsening is assumed to just be the natural course of the illness.

The problem with this literature base is that its really difficult to measure outcomes in the first place, and especially when there are methodological flaws in studies (two examples of these skewing the literature base in favor of drug use are withdrawal trials and six week studies, an example skewing the literature away from drugs is that the longer term studies can be at best correlative with poorer outcomes because you can't keep people on placebo for 15 years). However I think Whitaker does a great job of analyzing the methodology and will help you evaluate the evidence.

There's alot of quacks in the anti-psychiatry literature base. Reading Thomas Szasz isn't going to help you feel any better. You can't trust CCHR and probably not Peter Breggin to accurately represent the evidence. But Whitaker is extremely qualified, has won alot of awards for journalism, and even the people who disagree with him aren't saying that he's a quack--Whitaker is straight up science, no ideological biases against 'drugs'--he just is looking at outcomes.

There's been several discussions of his work on HN, and several people here recommended it a 2010 book to read.

You can look at the annotated bibliography for his section on antidepressants here http://robertwhitaker.org/robertwhitaker.org/Depression.html




Thanks, it sounds good. I just requested it by ILL.

My last book recommendation from HN didn't go that well, though. I saw Seligman's "Learned Optimism" recommended, so read it soon after. While it had some great insights, I fear my take home message was: yes, I'm horribly depressed, and this is likely the result of a lifetime of training and the influence of my depressed mother, and thus there is little I can do about it. Rather the opposite of the intent of the book, I fear. We'll see how this one fares!


I'll second the Anatomy of an Epidemic recommendation, it's really quite an amazing book. The comment below by the "pharmacist" is odd in that he hasn't read the book and knows nothing about it even though it is a major work. In particular, it has around 1080 citations in the notes at the end, most of which are to medical journal articles.


> 2) this is in part explained by the analysis of when antidepressants start to worsen the course of the illness (which is ~2 years for neuroleptics...unsure about antidepressants)

Neuroleptics are an entirely different class of drugs - you absolutely cannot lump them together. They have varyingly opposing effects (e.g. blocking dopamine instead of increasing it). And just FYI, only old doctors call them neuroleptics anymore in practice - we just say antipsychotics. And there are two very distinct classes of antipsychotics that can't really be lumped together for most purposes - simply, first generation and second generation. The first generation (chlorpromazine, fluphenazine, thioridazine, haloperidol, prochlorperazine, etc) is much more likely to cause depression, whereas some agents in the second generation are actually indicated to treat depression (e.g. olanzapine, aripiprazole).

Looking at the page you linked, I have several comments: 1) As far as chronicity, depressed patients are likely to have multiple episodes throughout their lives. If they suffer an episode while on an antidepressant, it doesn't mean the antidepressant caused it. If they weren't on the antidepressant, they would likely be having even more episodes. Also, only one citation? From 1973? That doesn't stand in evidence-based medicine today.

2)"British researchers found that 50% of drug-withdrawn patients relapsed within six months." That would indicate that the drugs were working, don't you think? They can't relapse unless they first leave depression. Equally, this would imply that after stopping antidepressant therapy, 50% of patients /didn't/ relapse.

Replace depression with another disease state and think about it this way: 50% of patients with high blood pressure who stopped taking their medication (which lowered their blood pressure) had high blood pressure 18 months after stopping the drug. Is there anything surprising about this? No. The problem is the stigma associate with depression.

3) Patients with mental illnesses are more likely to suffer from chronic illnesses (if you wish, I will fetch citations for you). Therefore, quite logically, patients who take antidepressants suffer from more chronic illnesses becaues they have depression than patients who have never had a mental illness. Again, there is nothing surprising about this.

4) Cherry-picked journal articles. I know very few people read medical journals regularly, but you can find an article to prove anything. There are reputable journals, and disreputable journals. Standards are a big deal, and just because something got published doesn't mean it's accurate or true. For a nice ranking of journals, you can check out http://www.scimagojr.com/

Is this the entire bibliography for the book? Honestly, 8 page journal articles have more than this (I'm serious. Look up NEJM and pull any five articles over two pages. And then keep in mind that there is a citation /limit/ to get published in the journal).

If you would like to discuss any of the literature more thoroughly, I would be more than happy to.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: