Very interesting TVAA. I believe this may be what is called a strawman?
The abstracts you have quoted here do not support your argument that antidepressants are rarely better than placebos.
The Antonucci article is not the one I cited and the abstract outlines the issues discussed in the articles. Says nothing specific about the relative position of antidepressants vs placebo effect
The de Oliveira abstract doesn’t mention placebo effect at all.
You need to read the whole article, not google the reference and copy and paste the abstract.
auliya, I believe that the placebo effect is exceptionally strong, and ultimately may be found to explain the effects of antidepressant medications. However, that demonstration has not happened yet. I think that the Kirsch response to the Thase and other commentary articles regarding placebo effect previously cited sums the current state of our knowledge most appropriately.
As an aside, (and an indication of bias) I studied under Irving Kirsch at UConn. Of any single person, he is most responsible for the psychologist that I have become today. His work on the placebo response is consistently thoughtful, methodologically sound and insightful.
So, I think you are very much right to say that the placebo effect has not yet been demonstrated to be typically equivalent to the effects of antidepressants. Irving’s work suggests that the placebo effect itself is exceptionally strong, which I don’t think you have contested.
Not significantly better (although they are significantly better in a mathematical sense at times). As the article points out, antidepressant effects are relatively weak – and the comparison being made is to placebo. We don’t know why the placebo effect is so unusually potent in depression. I suspect it’s because it is generally NOT a problem with physiology.
I realize that’s not the article you mentioned. That’s precisely why I mentioned it – it came up when I performed a search on the paper’s main authors.
Look, in my little backwater subdivision of this field, we’re still bickering about whether or not “the structural model” is a useful clinical concept. No kidding. As a clinician working in the branch of psychoanalytically-oriented psychotherapy, I have my work cut out for me just keeping abreast of the latest developments in that field; I’m not involved in any sort of clinical research.
The general attitude of the psychotherapeutic community in Sweden regarding anti-depressants could be neatly summed up by the quote at the top of this post, but I feel (or have felt) a need to supplement this opinion with some up-to-date scientific studies. Many of the patients I meet imbibe one or another of the latest medications; they often ask me for my views on them. I would prefer those views to be anchored in the latest research rather than in the general view of a community that’s still debating the pros and cons of Freudian theory. Hence my curiosity, especially when you claimed access to 30 studies that run contrary to what I’ve been taught.
Having said that, I must tell you that in my judgement thus far TVAA has produced more compelling evidence in support of his views than you have regarding yours. Both of his assertions (1. As a general condition, depression is typically more responsive to the placebo affect than most other “disorders,” and 2. anti-depressive medications rarely produce results that are “significantly” better [in the non-statistical sense] than placebos) are borne out by the meta-studies he and Hentor have presented. And the statistical differences between drug and placebo are in fact, clinically insignificant. Although Hentor is right to point out that “the placebo effect has not yet been demonstrated to be typically equivalent to the effects of antidepressants,” this problem revolves upon the experimental design of most studies, which cannot be used to determine the extent to which improvements are a result of medication, and the extent to which they are a result of placebo effects. That is to say, the extent to which we can say that the affects are additive. However, if I’ve understood correctly, most studies assume that the affects are additive. Kirsch state quite categorically that, if such is the case, “the pharmacological effects of antidepressants are clinically negligible.”
I find some of the claims in Hentor’s linked article, “Antidepressants and Placebos: Secrets, Revelations, and Unanswered Questions” even more worrisome:[ul][li]”57% of the trials funded by the pharmaceutical industry failed to show a significant difference between drug and placebo;”[/li]
[li]”most of these negative data were not published and were accessible only by gaining access to U.S. Food and Drug Administration (FDA) documents;”[/li]
[li]many of the studies employ arbitrary “cut-off” points that are statistically misleading;[/li]
[li]en FDA spokesperson has confirmed that in some studies the data is “fudged,” leading to an overestimation of both drug and placebo response;[/li]
[li]in some studies, no significant difference was found between low drug dose, high drug dose, and placebo effects;[/li]
[li]the existence of negative trials was until recently kept from the public as a matter of policy by the FDA.[/ul][/li]
All of these factors, taken together, seem to support a skeptical attitude regarding anti-depressives. Kirsh concludes:
At the same time, the general public is mislead into believing quite opposite, that is to say, that we actually possess good knowledge of the beneficial effects of these medications. Once again, this strikes me as precisely the problem TVAA seeks to address.
auliya disputed both points 1) and 2) above; she claimed they were “rubbish” on the previous page.
Technically no, I’m trained in the “object-relations” school (Klien, Bion, Racker, Fairbairn and Guntrip) with heavy influence from the “independent” tradition (Winnicott, Meltzer, Bollas, Kohon and Casement, among others). As I understand it, traditional Freudians are a dying breed.
On the other hand the schools I’ve mentioned are outgrowths of classical psychoanalysis, so it depends a bit on what you mean by “Freudian.”
It doesn’t matter what it “sounds like”, it matters what it is.
If someone tells you that they’re the Second Coming of Christ, it’s reasonable to be skeptical, but much less so if the person is speaking from a fiery throne descending from the sky and borne aloft by seraphim who sing “Hosanna” in unspeakably harmonious tones.
I am not familiar with the functional conventions of therapy, either psychological or psychiatric. In other words, I don’t know much about how to be a clinician. I have intimate knowledge of how medicine is practiced, and I know a lot more about the theory and philosophy of logic and the scientific method than the vast majority of clinicians.
The responses I’ve gotten from the self-proclaimed medical authorities in these threads would seem to support those claims, eh?
No I didn’t **Mr. Svinlesha **. What I disputed was TVAAs claim that
Rereading my original post I see I quoted both sentences and not the one I was specifically refuting. My apologies, the references mentioned are not concerned with whether there is a greater placebo effect in depression that other disorders.
However my subsequent posts on the matter should have made it clear to you what I was talking about see this reply to yourself
(sorry I cannot seem to get the link to the actual post to work, you will have to scroll down -a)
Certainly Hentor and others had no problem understanding my point.
Delusions does seem a likely assessment. (Comparing onesself to the Second Coming of Christ is always a tip-off, BTW)
As to the question of placebo response and the efficacy of antidepressants:
the cited studies are really more opinion bits than original research. Maybe true or not, I am not intimately familiar with the research, but I can easily find data to contradict the same:
So, without personally performing a full review of te literature, I’ll believe that some of the newer meds have pretty good evidence of efficacy. But I am open to being convinced otherwise and the evidence may not be as strong as I’d like.
BUT BUT these articles point out the need and utility of the DSM approach. Without some agreement as to what depression is efficacy (or non-efficacy) of treatment, be it medication or cognitive therapy or neo-Freudian psychoanalysis, would be impossible.
** First, that’s “delusions do seem”. Subject-verb agreement, DSeid.
Secondly, if someone ever enters these threads comparing themselves to the Second Coming, I’ll keep your opinion in mind.
** It’s not an “opinion bit”. Haven’t you ever heard of meta-analyses?
When the available data is examined as a totality, we find that there’s really no reason to presume that pharmacological treatments actually treat any presumed underlying problem.
** The older medications were also claimed to be quite effective. Since it has now become clear that not only isn’t that claim true, but there was never sufficient evidence to support it in the first place, similar claims would at least seem to be somewhat suspicious, yes?
Oh, please. The DSM doesn’t merely attempt to define syndromes, it attempts to define conditions. If you can’t tell the difference, perhaps you should excuse yourself from this discussion.
Now TVAA I don’t have the actual articles, just what you posted. Here’s what I read there - The Antonuccio article doesn’t mention meta-analysis in the portion you posted. If it is in the body of the article then please post what the inclusionary and exclusionary critera were. And the de Oliveira IR presents a case report as a jumping off point for pontification. You do know what a meta-analysis is, don’t you? (hint: look at the Joffe article I posrted for an example of one)* :rolleyes:
Not that studies that show a strong placebo effect for mild to moderate depression do not exist, mind you. And that stronger effects of antidepressants have been seen in studies that have a weaker placebo effect for comparison. It may be true that the evidence for effectiveness in mild to moderate depression is over-rated. Or not.
And Mr. S., I’ve NEVER heard anyone claim that antidepressants *cure/i] depression, have you? I’ve heard that they treat depression, that they control depression. Of course neither does insulin cure Type 1 diabetes to use the medical analogy.
I will excuse myself, I believe, only because the points have been said and don’t bear to be repeated ad infinitum. I am happy to play the critic of the DSM model in many a forum, but your failure to have the faintest understanding of its intent and utility is as mind-boggling as your delusions of intellectual greatness.
TVAA, you have to be the worst poster I have come across on the straighdope boards. You complain about a subjuct’s intrinsic characteristics, and you have no solutions to any real or percieved problems.
for example:
If I give several people the instructions to “go north for a
while, then head east, then walk south-southeast until you
reach the destinationi”, and they all end up in the same place
even though they travelled independently, it’s clear they’re
not really following the instructions – they’re following a
more specific set of instructions that isn’t obviously
incompatible with my much looser set.
That is not a valid conclusion. If the destination was a sufficiently wide area, every person possibly could end up at the same destination, even if they were following exact compass directions. You need to think before you type TVAA.
You still haven’t answered my question from the earlier thread, what is a “known disease process?”
You relentlessly slam the DSM, yet you give no worthwile suggestions as to a replacement. The funniest, and most aggrivating, part of the original thread was were you suggested “we need an objective test for autism.” Idiot! What do you think the DSM is trying to do? I’m not going to go over again how complaining that something is vague when there is no possible way of measuring it anyway is pointless at best. You are obviously not smart enough to grasp the concept.
Valid: (from the online webster dictionary)
having legal efficacy or force; especially : executed with the proper legal authority and formalities <a valid contract>
well-grounded or justifiable : being at once relevant and meaningful <a valid theory> b : logically correct <a valid argument> <valid inference>
appropriate to the end in view : EFFECTIVE <every craft has its own valid methods>
of a taxon : conforming to accepted principles of sound biological classification
Posted by TVAA:
I’m saying that the claims that mental illnesses are
physiological diseases with known etiologies aren’t valid.
Lets walk through this shall we? Is the claim that mental illnesses are phsyiological dieases with known causes well-grounded and justifiable? From the previous thread there was an impressive amount of evidence that many mental disorders could be caused physically. Such as ADHD. In addition, there is the blatently obvious effects of “uppers and downers” on a patients mental state. If a drug can cause a patients brain to function incorretly, such decreasing the rate a certain brain-chemical or hormone is processed or recieved. Then it is also possible that a defect could cause this to happen in another patients brain, much like a mental diabeties.
It can also be a logical conclusion, since a mental state can cause a physical state, therefore it is possible that by reproducing the physical state(or at least the correct hormones and neurochemicals) you can also cause the mental state.
Is it appropriate to the end in view? I assume it aims at curing or at least easing the discomfort of those affected. If studying the physiological aspects brings us a better understanding of mental disorders, why not. If a medication can release a patient from suffering, or at least aid in its therapy, why deny it to them?
Is this approach effective? Considering the subject matter, it the most effective thing we’ve got. A better concept may be discovered in the future, but until then we must work with what we have.
I believe that the claim that mental disorders can have physical causes can be considered quite valid. Its justifiable in respect to current medical knowledge. Its logical in the respect that cause and effect may be interchangable. It is appropriate because it can bring insight and solutions to a vague subject matter. It is effective since it is a measureable approach to a difficult problem.
A meta-analysis, in the simplest terms, is a study of studies: a researcher examines the findings of many different studies, attempting to compensate for possible differences in methodology, in an attempt to find what the current available evidence across a field suggests.
You’ve never heard anyone claim that antidepressants cure depression? What about the claim that they compensate for the underlying biochemical imbalances responsible for it?
Way to ignore the very real questions brought up by those articles. (By the way, if you can’t access them, why are you discussing their contents?)
You keep talking about what the DSM should be used as instead of how it actually is used, not only in research but within clinical practice and the general culture. It’s not merely a means for clinicians to describe the problems they deal with. It’s based in theory, set by APA consensus, and used to suggest a degree of objectivity not genuinely found in psychiatric diagnosis.
So how exactly are those cited studies meta-analyses oh wise one? You do know that a meta-analysis pools data from a variety of studies that meet inclusionary critera according to strict rules and then statistically analyzes the pooled data? Not just a literature review. Or a case report with the author extrapolating to the whole of depression treatment from the n of 1. Do you actually have any knowledge of how science and research works?
No, compensating for underlying biochemical imbalances treats the imbalance, but it doesn’t make the imbalance go away, it doesn’t cure. Again, insulin does not cure Type 1 diabetes, it treats it.
Your reading comprehension is lacking. I didn’t claim those studies were meta-analyses. I asked whether you’d heard of them.
Yes, you can find a study that shows that some antidepressant has an effect significantly (both mathematically and clinically) greater than placebo. You can also find a person who, after tossing a coin in the air a hundred times, had a hundred heads. But that’s not important – it’s the results of all the tossings that matter.
Didn’t you pay any attention to the complaint that pharmaceutical companies only reported the studies that suggested their products were helpful?
And you are completely missing the point. We know what the biochemical problem is with Type I Diabetes, and supplying insulin corrects it. We have no idea what any biological problems exist with the mental disorders, if any, and we have no knowledge that any treatments work by affecting them (if they exist). The only known biochemical imbalances in such individuals are those caused by medication.
Oooookay. So when I challanged your cites as “opinion bits” and you responded
you were just curious if I knew what a meta-analysis was at that moment because it happened to hop into your expansive mind to be curious about the limited extent of my feeble knowledge base? You were not claiming that those studies were meta-analyses. Is there a way to do rolleyes to some exponential power?
Let me write this in simple words that you might comprehend. As you said
And that is what meta-analysis does. The meta-analysis by Joffe et al is the results of all the tossings and shows a real effect. An opinion bit, even one that reviews a variety of studies, does not count all the tossings. It selects studies to support a POV. And a case report(!!!) is hardly something to extrapolate off of.
Do you even comprehend your own posts? The pharmas reported the studies to the FDA. Negative studies are alleged to not have been made public. Now are these studies of meds that were approved for indications that they had negative results in? Or are these studies of products that failed to meet FDA muster for those particular indication? I don’t know, but I really doubt that the FDA is in on a big conspiricy to get ineffective medications to market. But maybe you have more than just delusions of granduer.
We have lots of ideas of what biological problems exist in many mental disorders. We do not have a complete understanding.
You fail to understand the most basic points. Before insulin deficiency was determined to be the proximate cause of Type 1 DM the condition was defined by clinical signs. No one had any idea what the cause was. (Truth be told, we still don’t understand what causes the insulin deficiency …) The first step was to define the condition … frequent urination, insatiable thirst, sweet tasting urine (yes they tasted it) and the usual clinical course. Noting that these things seemed to travel together. Then it could be studied and better understood. Different types of DM could be delineated. You don’t wait to study or to treat the best you can until you have all the answers.
You have made many statements that are just patently untrue. And the studies that document the untruth have been cited in these threads many times. It is okay to be as ignorant as you are. But to be so obnoxiously arrogant while being so ill-informed and uneducable is another matter altogether.
(I dislike arrogance in all cases, but worst of all is an arrogant idiot. But then again almost every arrogant posturer turns out to be an idiot once you talk to them a little while …)
** No, or I would have used it to respond to you previously.
As you said, you can find a study that suggests antidepressants are greatly superior to placebos. The point is that you can find the reverse equally easily. Meta-analyses are necessary to examine the issue when individual studies reach conflicting conclusions.
That case study brought up a dozen issues that are frequently ignored. Did you actually think about them at all, or did you dismiss them because the author didn’t provide cites? There were cites for those articles, by the way – I don’t suppose you checked any of them out?
** It’s all in the public information given out…
And you do realize that there have been several relatively high-profile cases where it was found that heavily-promoted and expensive drugs weren’t any better than the older and cheaper drugs that were available (and in some cases, they were actually considerably worse)?
The drug companies pay for the studies, they perform the studies, and they decide what studies are sent to the FDA. They don’t give out all those pens, tablets, and vacations for nothing.
** We have countless hypotheses. We have no understanding. More to the point, we (and I am expressly including you in that ‘we’) accept explanations that have been scientifically disproven.
** Nonsense. It’s caused by damage to the pancreas. We don’t know what causes that damage, though…
Still, diabetes had plenty of objective symptoms, and it was clearly a distinct condition. Mental disorders aren’t so objective, nor are they as distinct.
I have NOT made untrue claims (except that the earliest practitioners of psychiatry didn’t have professional degrees, and you collectively couldn’t even correct me properly). You’ve failed to produce effective counter-arguments and treated unsupported statements as if they were fact.
I keep trying to educate you, but I suppose you’re just not willing to think critically about this topic. It’s a shame, really – let us know when you change your mind.
I hope you don’t think I’m picking on you by continually returning to this point, but a lot of this discussion seems to revolve around it. I’ll just write out a quick summary of what I’ve understood regarding the question thus far, and then you, H. the B., and/or Dseid can correct me if I’m wrong. The question is, “Do anti-depressives perform significantly better than placebos in the majority of clinical tests that have been conducted?”
Clearly that question cannot be answered by looking at one or two clinical studies taken in isolation. It can only be resolved by comparing many studies taken together. And it is undisputed among experts that, if one reviews the entire body of tests thus far conducted, one discovers a wide variety of outcomes: some reveal a large difference between placebo and drug effect, some reveal a small difference, and some find no difference at all. Overall, however, the difference between drug and placebo effect is believed to be “small.” For example, Michael Thase, commenting on Kirsch et. al.’s meta-study “The Emperor’s New Clothes,” notes:
Note the very careful manner Thase phrases this observation, by the way. He is aware that in conducting these tests, the pharmaceutical industry screens out large numbers of subjects who would otherwise be included in the total population of depressed persons, whom the psychologist/physician would normally meet in routine clinical practice, and for whom these medications are often prescribed (see Zimmerman, et al., Are Subjects in Pharmacological Treatment Trials of Depression Representative of Patients in Routine Clinical Practice?, Am. J. Psychiatry 2002 (159), 469-473. In a later interview, Zimmerman flatly states that the industry excludes “mildly depressed” patients from its studies because it fears these patients will fail to exhibit any measurably different response to the drug when compared to the placebo).
Kirsh responds to the comments on his study with the observation that:
Part of the problem you (or any researcher) faces regarding this point is that most of the studies that produce negative results never get published. Negative tests are registered with the FDA and then simply not reported in the literature. That means that, up until recently, surveys or meta-analyses of studies have only had access to those trials that produce positive, measurable, statistically significant results! No wonder such surveys (like the two you produce, or those produced by DSeid) conclude that there is strong evidence for the efficacy of anti-depressives above and beyond the placebo effect.
Regardless, if Kirsch is to be believed, more than half of the studies conducted on the effects of anti-depressive medications fail to demonstrate a statistically significant difference between drug and placebo. If we look at the raw data of his particular study, we find that of 47 clinical trials, 9 reported no results, ostensibly because the studies failed to produce statistically significant drug effects. Regarding those 9, “the statistical or medical reviewers stated that no drug effect was found.” Kirsch therefore excluded the drugs that had been tested in these trials from his calculations of drug effect. Both the placebo and drug effects were measured as differences on a HAM-D scale administered first before, and then after, the treatment; Kirsch estimates a “weighted mean difference” between placebo and drug treatments of 1.8 points on the HAM.
82% of the drug effect was matched by placebo treatments. That means that if we were to convert the results to a 100-point scale, where everyone in the test started at zero, then we would find (on average) that at the end of the test, those taking the drug would score around 100, while those taking a placebo would score around 82. To express it differently: of 100 people being tested, 18 would score 1.8 points higher on the HAM scale at the end of the test. Kirsch judges this effect to be statistically significant. However, the difference in HAM score is of questionable clinical significance; in other words, if I score 1.8 points lower (or higher) on the HAM scale than you do, we would nevertheless still be pretty equally depressed. Kirsch summarizes these results as follows:
In other words, the remaining 18% difference between drug and placebo could be the result of the exclusion of 9 studies that showed no significant difference between placebo and drug treatment, combined with the bias inherent in the “last observation carried forward” method, which Kirsch found tended to exaggerate differences between drug and placebo effect (when compared to the more conservative “observed cases” method). In addition there is the “broken blind” problem; studies have shown that test subjects are sometimes able to ascertain whether or not they are receiving a placebo, as opposed to the real thing, based upon whether or not they spontaneously begin to suffer from known side-effects (Rabkin, et. al.).
Granting that it is still difficult to correctly interpret all of this statistical data, it nevertheless seems clear that the question of the efficacy of anti-depressives is unresolved, despite all the published evidence to the contrary. In particular, studies prior to the Freedom of Information Act didn’t have access to the negative and inconclusive trials registered with the FDA.
At any rate, it certainly seems to be case that in the majority of trials, anti-depressives failed to achieve clinically significant results beyond that of the placebo effect; and there were even trials in which the placebo out-performed the anti-depressive (although not in a statistically significant sense). The answer to the question I posed above, then, appears to be “No.” DSeid:
First off, thanks for asking! He was a bit spotted last week, but that went over, thank God. Otherwise, he’s healthy and happy. He’s managed to learn one English word: “cookie!”
Anyway, then, to the topic at hand:
The 1996 meta-analysis you’ve cited probably suffers from the biases mentioned by Kirsch, namely, the unavailability of unpublished studies that showed little or no difference between drug and placebo effect.
Your second study (Casacalenda, et. al.), while interesting, doesn’t address the placebo effect at all; we cannot tell what percentage of those who responded positively to anti-depressive medication did so due to that effect.
Well, what I hear is some lack of clarity, most of the time. Some doctors/psychiatrists would maintain that depression can’t be “cured” at all; it can only be “treated,” but because they basically believe the disorder to be biological in nature, “treatment” = “cure.” At any rate, there is no point trying to solve this biological illness with therapy.
Although I must admit I’ve noticed something of a “sea-change” over the last year or so, in which the original enthusiasm for these medications among doctors is shifting into something more ambivalent. I think many doctors are discovering that the medicines aren’t really performing as promised, and that their patients become dependent on them without ever really getting better. That’s just my anecdotal experience, however.
Sidestepping your fisticuffs with TVAA, I would just like to point out that the FDA meta-study I’ve cited is a meta-study, not opinion pieces. Regarding this:
Kirsch explains:
These requirements mean that a large number of trials which show no significant difference between placebo and drug effect can be conducted and still have no influence on the FDA’s decision to approve the drug. In addition, the difference between placebo and drug might be significant in a technical sense, i.e. statistically, but not significant in a clinical sense. On top of that, until recently, it was the policy of the FDA to not report the results of negative studies. Vorlan:
DSeid explained that he believed anti-depressives to be effective, he provided studies to explain why he believed that to be the case, and he declared himself open for the possibility that he might be wrong. I cannot see what more one could possibly expect of a debating opponent.
In this discussion both here and in the Pit you’ve mischaracterized at least three clinical psychologists and one MD as “idiots.” Clearly they aren’t, even if they don’t agree with your point of view. Your haughty attitude does disservice to your argument, and creates unnecessary conflicts.
Mr. S, I do not particularly care what someone “believes”. DSeid and several others have repeatedly suggested that not only are they personally familiar with the topic at hand, but that their knowledge allows them to judge that I don’t know what I’m talking about.
They don’t actually understand how research in clinical psychology occurs. They’re not even rudimentarily familiar with some of the most significant debates within the profession, or the discontinuity between what is known and the impression most people have. They even continue to insist that various psychological disorders are known to be caused by altered levels of certain neurotransmitters, although the experiments showing that simply is not the case are decades old.
They’ve also suggested that the few people who support, even incidentally, statements I’ve made are brainwashed puppets. That includes you.
I am more than willing to respect honest differences of opinion. But when people who both lack the relevant knowledge and who refuse to seek it out suggest that I am a liar, a fool, and a charlatan for stating known facts, my patience ends.
Even now, Zoe is lying about what I’ve said and what I’ve claimed to be. Perhaps she’s merely mistaken, and too angry about the things I’ve supposedly said (that she’s misunderstood) to recognize this fact.
That is not intelligent behavior. At best, it’s stupidity; at worst, she knows what she’s doing and is simply trying to discredit ideas she can’t tolerate. What should my response be to this?