Comments by A. Person

1. 'Growing Up in the Universe' now available free online

Comment #88074 by A. Person on November 14, 2007 at 1:57 pm

90s flashback.

Comment #86618 by A. Person on November 9, 2007 at 5:54 pm

Deep down, most people don't care about the "truth value" of claims.

3. Pat Robertson Says Giuliani Presidency Appears in Book of Revelation

Comment #86245 by A. Person on November 8, 2007 at 8:06 pm

Poe's Law, BAEOZ.

I'll be honest; I didn't know this was satire until the 9/11. Giuliani includes a noun and a verb along with 9/11 in his statements, which demonstrates the fakery.

4. Same Flea, Different Name?

Comment #86001 by A. Person on November 7, 2007 at 6:58 pm

Socratzsche, what was that quote from? It reads vaguely like Nietzsche, but it's not ringing any bells.

5. Judgement Day: Intelligent Design on Trial

Comment #85955 by A. Person on November 7, 2007 at 3:18 pm

My one regret is that actors playing Intelligent Design advocates are probably going to stutter and pause far less than they actually do, thus misrepresenting them as people who at least act like they know what they're talking about.

6. Jesus Rides the Number 7 Train

Comment #84773 by A. Person on November 3, 2007 at 2:02 pm

"A boss in Heaven is the best excuse for a boss on earth, therefore If God did exist, he would have to be abolished." - Mikhail Bakunin

In the long run, nothing changes about the church. Or the arguments against it.

7. Searching for God in the Brain

Comment #77015 by A. Person on October 8, 2007 at 7:26 am

To sum up that and what I've read of his book, Beauregard writes like a monk with severe ergotism whenever the chance to express personal opinion instead of experimental results presents itself. IIRC, a lot of it is very typically Buddhist/Hindu.

8. Searching for God in the Brain

Comment #76964 by A. Person on October 8, 2007 at 1:15 am

Am I the only one thinking "I wonder if you could put people on mescaline under an fMRI to compare?"

Because really, when they keep talking about artificially inducing religious experiences, which (aside from being their goal) would be experimentally useful here, I keep thinking of Huxley's essays on mescaline.

9. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74942 by A. Person on October 1, 2007 at 7:33 am

In the interest of continuing to be contentious, I'm going to re-re-re-iterate that I think clinical trials have many times greater explanatory power than all the evidence in favor of SSRIs combined.

10. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74845 by A. Person on September 30, 2007 at 9:56 pm

When I say "statistical" and "population" I'm thinking of any study that doesn't collect its own data but instead relies upon (for instance) the CDCs nationwide statistics. It's true that's not a very precise usage of the term.

It's not about placebos preventing suicide, it's about the course of treatment during which SSRIs are administered preventing suicide, which is different. The continued mantra is that correlation does not imply causation. The burden of proof ought to be heavily on those asserting it does, especially when suicidality has been shown to actually increase in experiments that are properly blinded.

I've got my own thoughts about experimental models regarding SSRIs and psychiatric meds in general, but I'm really not inclined to write them out on a weekend. It's depressingly negative to do nothing but point towards the inadequacies of existing statistics, but it also consumes far less energy. :P

11. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74829 by A. Person on September 30, 2007 at 6:12 pm

Sure. Statistical studies are worth a lot when more precise research can't be done, I agree, but especially with anything as intricate as suicide and depression it's very nearly impossible to pin down all the roving variables in population studies.

The problem is that, statistically, the apparent effect of SSRIs on suicide could likely be duplicated by sugar pills. If you gave out sugar pills to everyone seeking psychiatric help, 80% of suicides would not have sugar pills in their system because the people who kill themselves are not the ones acknowledging they have problems and seeking help. It's fairly basic statistics.

That's my objection to the population research across the board. It's hairy and difficult enough to control for even the basics like social class in population studies, when you get into such intangibles as tendency towards suicide it's very nearly completely outside of the ability for statistical models to explain or comment on meaningfully. Looking at trends is very useful to suggest information, as are anecdotes, but to really quantify what any actual program is doing or not doing meaningfully requires the mind-bogglingly expensive and time-consuming years spent in clinical trials.

The thing is, those trials have already been done, and they're the reason for the initial concern. If anything, the genetics research the initial article here is on seems to support the thesis that there's a component outside of simple "extra energy" increasing suicidality in reaction to SSRIs, which is very troublesome. Supposing the people with 15x the likelihood of a violently bad reaction to the medication are given it and then cut loose for months, doesn't this seem like a serious and severe problem? It's an extremely common practice, and it's, well, bad shit. Genetics research to figure out who is at risk is a good stopgap measure, but the perpetual optimist in me sees this as an avoidable problem brought on by a lack of realistic evaluation for how the medication is given out and affects people, in line with many past blunders.

12. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74676 by A. Person on September 30, 2007 at 12:19 am

Oh, sorry. The comment on clinical trials against placebo was in response to your third paragraph (or thereabouts) where you were speculating (or it seemed) about whether the suicidality link could be correlation rather than causation. Should've been clearer, it wasn't about your story, which was properly noted as such and informative. Anecdotes are good. If you've got enough good anecdotes and thinking you're well on your way to understanding things.

I'm still hyping up clinical trials, it's population studies (without a placebo group) that I hate. Clinical trials are considered the gold standard of testing in medication, the problem is that they don't show antidepressants to be safe. They are, in fact, what convinced the FDA they were dangerous.

I think I slipped much closer (in retrospect) to ad hominem when bringing up ice pick lobotomies and Freud's love of cocaine. They're a fairly, ah, dramatic selection for examples of unblinded observations gone haywire, to say the least. I should have picked something more neutral, probably, but I'm not sure anything else in the same vein would be quite as fitting and I was already thinking of psychiatrists.

13. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74639 by A. Person on September 29, 2007 at 7:06 pm

I'd like to note that saying that "scientologists have succeeded in their plot" in response to my issues, which the FDA backs up, is in fact tarring me (and the FDA) unreasonably with the same brush as scientology. Arguing from authority because I'm a student who you assume is "unaware of the intricacies of this type of research", along with the sarcastic "professor", is simply stupid and an ad hominem attack. I'm certainly not going to bother with undoing my anonymity any further than that if you're immediately eager to argue ad hominem at the tiniest hint that I don't have your high authority when it comes to medications.

Seeing as you're willfully applying ad hominem attacks, though, I can happily list off mindless counters to your bullet points, which I nevertheless already addressed. It's going to get repetitious here, but I don't feel like being accused of not arguing against your points again.

1-6 Correlation does not imply causation. Since we're being mean, anyone who took a week of statistics 101 should know these studies prove nothing. Those doing the studies mostly know it, too. Studies with inconclusive results are done to suggest further, more conclusive research before anything is done. Suicide rates have trends up and down regardless of drugs. Clinical trials demonstrate 35% efficacy for depression in placebo, which is extremely high. These results do not demonstrate that suicide rate drops because of antidepressants rather than during the course of the illness during which people can and do seek help, and the more scientifically sound double-blind research that prompted the black box warnings seems to disagree with conclusions of safety. Additionally, the segments applying to adults are not as relevent, as the black box warning was initially and primarily for children and adolescents.

7- One year does not establish a trend. Suicide rates are measured in a few per hundred thousand, it's not necessarily the warnings doing it. More importantly, the "black box warning" could be prompting people to stop taking the medication abruptly, leading to severe withdrawal- which is a problem with the patients and the media and the warnings, not a lack of problem with the drug. Even if the black box warnings are doing damage, it isn't necessarily because SSRIs are good so much as that abrupt, unsupervised termination is bad. And, lest we forget: Correlation does not imply causation.

8- Correlation does not imply causation, again. Those who killed themselves never sought SSRIs, but also probably never sought ANY kind of help. This doesn't necessarily mean SSRIs would have saved them, since if they didn't even have SSRIs they probably weren't doing anything at all. Remember there's a massive placebo effect, and it's not like they were chosen in a trial not to get SSRIs- they had to be people who chose not to get help, which is very possibly - even probably - a bigger indicator of being a suicide risk than SSRIs could be. It's known as a self-selecting sample, it's completely unscientific to rely on it, and it's why we have double blinds.

You have yet to 'prove' in any meaningful way that the benefits of antidepressants outweigh the costs, especially given the complicated nature of antidepressants and suicide. The double blind reseach is not there, I know it, the FDA knows it, and a majority of your sources know it. Non-double blind research has limits, and those limits are, specifically, that it can't actually 'prove' anything, only suggest it, because self-selecting samples are extremely unreliable and lurking variables- causes you aren't looking and adjusting for- are abundant. Ignoring these facts results in incorrect conclusions about cause and effect, and is the result of wishful thinking and confirmation bias.

Back outside of the very basics and repeating myself for your bullet-points, inpatient facilities are much, much safer and closer-monitored than elsewhere. I shouldn't even need to source that many millions of those receiving treatment aren't in inpatient, and I'm going to add that they aren't monitored very well. I'm not even going to cite any research besides the study you linked that gives two months as a standard period of neglect. Insisting that antidepressants are safe because they're safe in inpatient or in clinical trials is poor logic. That guns are safe in gun ranges does not make guns safe in reality. Proving they're safe enough to be used in the setting they're actually used in should be placed at the feet of those who say they are, especially when a lot of data suggests that they aren't.

14. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74587 by A. Person on September 29, 2007 at 4:03 pm

Alright.

From the start, I'm going to have to boggle at the Scientology reference. The FDA aren't scientologists, I don't think, and it's their double-blinded studies and warnings that largely cause the concern over SSRIs. Wariness regarding SSRIs does not require or even correlate with a conspiracy theory. The worst I could say for those reliant on medications is that they overuse them because when all you have is a hammer, everything looks like a nail. Drug companies I'm less kind to, out of the belief that they suppress unfavorable information in the name of the bottom line- which is a well-documented phenomenon. But that's all. Beyond that lies madness and Scientology.

But even the father of the ice-pick lobotomy was deathly convinced to his grave he'd done wonderful work for mankind. Intent doesn't prove that you're correct, nor does being hated by Scientologists.

Regarding the studies, you'll note placebo works 35% of the time in double-blind studies compared with 60-70% with prozac, resulting in a 25-35% increase in improvement on the drug, compared with a 2% increase in suicidality, which is about what I remembered- a fraction of the number. For the simple reason of the huge placebo effect, however, all the population studies- those without double blinds- are more or less useless. They don't do much but look good on paper and make for a nice sound byte. Several of them even have disclaimers to this effect towards their end- "Gibbons and colleagues also point out that their aggregate data do not prove a causal relationship.", and similar. In the social sciences, those kinds of studies are known to use the exact same sets of information to argue exactly opposite conclusions, usually (in my experience, anyway) in keeping with the political leanings of those doing the analysis.

... before you leap on this, I still don't believe it's a conspiracy. I just think people expect their point of view to be correct, which is understandable, but results in confirmation bias. The studies are, regardless, still not double-blinded, still horrifically unscientific, and still useless.

One of your articles also highlights one of the problems; "In contrast, toxic side effects are rare in SSRIs. Physicians often prescribe the drugs in larger doses and may not see the patient again for up to two months. This scenario, Licinio warns, can set the stage for suicide risk."

It's not uncommon, it's done, and it means that the 2% of patients that get much, much worse on SSRIs within the first few weeks are much more likely to kill themselves than a closely-monitored clinical trial would indicate. If you're a psychotherapist (I took this from context and may be wrong), you see your patients often enough that they're at much lower risk, which is great but says little about common practice. Suicide is measured as a few per hundred thousand, it only takes a small fraction of patients to spike the rate considerably.

If you want a textbook example, one of the Columbine shooters was on Luvox for a year. It was entirely insane and stupid to pull the drug simply because he was on it, because Columbine was a one in a million occurrence. This is the same public outcry that targeted all kinds of media at half-random, after all. That said, it illustrates nicely that when teenagers are going off the deep end, they will not necessarily tell people, even though it's generally 'just' a suicide at the end instead of a shooting spree, and it's completely impossible to directly point to antidepressants as the cause of anything.

Regarding drug trials, it's my understanding that they screen out either suicidal or previously suicidal patients fairly frequently as a matter of course. It's not going to skew the results terribly, it's true, but completed suicide almost never happens. I'm sorry if "miracle stories" seems like hyperbole, but personally I think being twice as effective as placebo is miracle drug territory. I know it was when Prozac came out, has this changed? Noone's ever doubted that doubling the number of people who improve is a wonderful result- it's just that it also doubles suicidal ideation, which is a much smaller group and smaller effect but very worriesome. Suppose (completely hypothetically) that the drug saves or even just cheers up two or ten teenagers, but kills one. Isn't that a serious problem that needs to be addressed? The ethical dilemma gets harder the worse the ratio between people saved and harmed gets and the milder the help is compared with the harm- suicide compared against wellbeing. The dilemma becomes nearly impossible (back in reality) when it becomes difficult to actually quantify help versus harm and total effects meaningfully. The default scientific position is ignorance, and if you have evidence for both help and harm clinically, with largely unmeasurable practical effects, it's far saner to back off and be cautious.

The only set of statistics there that's actually worrying about the black box warning is the well-documented spike in 2004 suicides, especially among young girls and by hanging. If I remember (and I'm sorry, but I'm not going to source this) prescription numbers didn't drop until a year later; if the 2004 suicide hop isn't a fluke and really is related to the black box warnings, it's probably (IMO) because patients got spooked at the black box warning and stopped taking their medication against their physicians' advice, which results in withdrawal and is a rather different problem than fewer prescriptions. This is a problem with the patients and the media, though, not the scientific studies and concerns about SSRIs.

15. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74456 by A. Person on September 29, 2007 at 12:12 am

irate_atheist: The increase in suicidal thoughts is from double-blind clinical trials compared with placebo, not normal case histories. It's only observed in a fraction compared with the number that improve (IIRC), and all antidepressants have a number of very real miracle stories associated with them, which is why they're in such wide use.

Oxytocin: I look forward to it. I'm going to hold off further commentary of my own until I've heard more of your side. It's a very interesting issue, and certainly deserves (even requires) the extra depth debate can add to it. I'm a student.

16. Genes Tied to Bad Reactions to Antidepressant Drug

Comment #74395 by A. Person on September 28, 2007 at 2:17 pm

Antidepressants aren't the subject of "hysteria", it's a serious and well-established concern. The FDA isn't subject to hysteria, they're a serious organization. In clinical trials with most SSRIs, those getting the drug are roughly twice as likely to think of killing themselves or develop psychotic symptons (hallucinations, delusions, etc) as those getting placebo. The simple explanation is that the drugs make people want to kill themselves. Anywhere outside of a drug trial, if someone thinks of killing themselves, it is considered more likely that they will then kill themselves.

However, in the interest of making the drug sound safe, they insist that suicidal ideas from a drug in no way contribute to suicidal actions. This is basically the same as asserting that, if I have two friends, and one says he's thinking of blowing his brains out, and the other one says he's feeling depressed, I should worry equally about the likelihood that they will kill themselves. It is, in other words, stupid.

The reason suicides almost never occur during drug trials are that they screen patients looking for relatively mild cases, and that the patients are monitored very closely. Most drug trials have checkins weekly, if that infrequently, whereas a real psychiatrist who has given a suicidal-ideation-inducing dose of medication will likely not see that patient again for months to years unless the patient wants to come in. When you consider that the patients are already ill, medicated, and often children, it's not surprising they often fail to take this initiative.

If those doing research want to assert that it's simply "extra energy" that causes extra suicidal ideation, and the drugs are really safe, they can try to produce evidence to support that theory. Given that it's a fairly unsupported hypothesis in antidepressants and that it is far more likely that suicidal thought leads to suicidal actions, it ought to require special effort on the part of those making the assertion. This is something that has not been done.

It's worth remembering that psychiatry proper is an extremely young science, fifty years removed from the era of lobotomies, which oodles of anecdotal reports insisted worked great, thanks to how nonscientific observations can and do work. Psychiatry is also a little over a hundred years from Freud promoting cocaine and radical nose surgery as cure-alls, though he was part of the cure for that problem as well. It's also worth remembering that drug companies, apart from much idiot vaccine hysteria, really have quashed drug trials in many other classifications regarding the potential dangers of their products.

... also, hi. I made an account to say this.