Furious Seasons

November 03, 2008

Major Study Of Zoloft And CBT For Kids' Anxiety Has Its Own Issues

Last week an interesting study of anxiety treatment in kids and teens came out in the New England Journal of Medicine and it's gotten quite a bit of press attention. The study involved using Zoloft, an SSRI, and cognitive behavioral therapy to treat anxiety disorders in teens and kids as young as seven and as old as 17. Look, I know how much some readers hate SSRIs and how strongly they react to any news about the drugs, pro or con, but for the moment put those kinds of concerns aside. My point in picking apart this study isn't to argue that no kid should ever be diagnosed with anxiety or that they shouldn't take meds of that therapy is useless or anything of the sort. My point is that this study isn't quite all that it's cracked up to be, although it does have its moments.

In the study the authors assert that the findings have major public health implications, meaning the findings can be extrapolated to the entire population of kids and teens diagnosed with a range of anxiety disorders. If true, it would affect approximately 7.5 million to 15 million kids and teens in America (the authors claim that anxiety disorders affect 10 percent to 20 percent of Americans under 18). That's a whole lot of kids. The authors also claim that anxiety disorders in youth lead to anxiety and depression as the kids turn into adults--one of the usual arguments for giving kids psychiatric treatment.

Certainly, the degree to which CBT was shown to be useful is valuable information as was the fact that CBT-only treatment outperformed Zoloft-only therapy by about 10 percent. But the fact that Zoloft had a fairly typical effect size of about 30 percent over placebo was hardly a ringing endorsement of meds-alone treatment, which is much the standard in America these days for kids (and adults).

The study involved 488 kids and teens diagnosed with social phobia, generalized anxiety disorder and separation anxiety and lasted for three months (there will be a longer term version of this study published later). Patients given placebo improved--the authors' terminology--23.7 percent of the time; patients given Zoloft improved 54.9 percent of the time; patients given CBT improved 59.7 percent of the time (so psychotherapy beat an SSRI); and, patients given a combination of Zoloft and CBT improved 80.7 percent of the time. In other words, the SSRI plus psychotherapy approach had a 57 effect size over placebo and a 20 percent to 25 percent effect size over either meds or therapy alone.

The result for Zoloft plus CBT is one of the most robust I have ever seen in any psychiatric study, so robust that it gives me doubts.

My concern about this study and the way it's being portrayed in the media, however, involves a problem in the study design, which overall is actually quite good. But here's the deal: if you are going to go around claiming that Zoloft plus CBT is now indicated for anxiety disorder in kids and teens--and that's what the study authors claim--then you need a way to compare the Zoloft plus CBT arm of the study to a placebo plus CBT group of patients in order to allow for whatever placebo effect--likely 20 percent or more--was present in the Zoloft plus CBT group. There was no such group in the study. Instead, we got a placebo-only group, a Zoloft-only group, a CBT-only group and a CBT plus Zoloft group.

It's a bit remarkable to me that if, as the authors note, they were trying to tease out of this study how powerful a therapy plus meds approach was that they didn't include a placebo plus therapy group. It would make comparisons and conclusions much easier and more powerful. What's more, my guesstimate is that such a group would've ended up with about a 65 percent to 70 percent "improvement" rate, leaving the Zoloft plus CBT group (at about an 80 percent improvement rate) with a 10 percent to 15 percent effect size. In the real world, that isn't much.

As a result of this, I don't think the study's conclusions on meds plus therapy are nearly as conclusive as some commentators do.

"The new study, paid for by the National Institute of Mental Health, is the largest examining treatment of childhood anxiety disorders, said co-author Dr. John March of Duke University. Dr. Thomas Insel, the institute's director, said the study provides strong evidence that combined treatment is 'the gold standard,' but that sertraline or therapy alone can be effective."

Yes, but the question is how effective.

Doug Bremner, a psychiatrist at Emory University, argues that the results are not remarkable at all:

"So now let's turn to the "incredible" results of this week's study of sertraline (Zoloft) in kids. Although there was a difference in "responders" based on much improved on the CGI of 60% versus 24% for placebo, when you look at the actual data, the Pediatric Anxiety Scale, a 30 point scale, went from 18.8 at baseline to 9.8 in the zoloft group, and from 19.6 to 12.6 in the placebo group, a difference of 9%."

While I'm no expert on this particular scale, it does strike me that the results of this study aren't nearly as far-reaching as some would like them to be, or as some fear they might be.

Then, there's the whole tricky issue of giving SSRIs to kids as young as eight--fully 75 percent of the study participants are 13 or younger. While I'm not 100 percent opposed to giving SSRIs to kids, I'm like 90 percent against it, especially for minor cases of anxiety such as some kids in this study had (you're going to, as a parent, willingly give your kid Zoloft for social phobia or separation anxiety? Whatever). I don't think the fact that there were few side effects in any of the study's arms should give anyone too much peace--there were perhaps 200 kids on meds in the study (so it's hardly population based) and it's clear the docs did a commendable job of titrating the kids onto Zoloft (real world docs tend to be a lot more sloppy and I think that leads to all sorts of problems). I certainly wouldn't use the low side effects outcome of this study to project onto, say, 50,000 kids with anxiety disorders. And it's when you get into populations of that size that you begin to see the real problems with SSRIs, even if they are relatively few on a percentage basis.

Nonetheless, I think the study authors did a commendable job of monitoring and reporting side effects--the paper is certainly the most aggressive on this point that I've seen in recent years in the use of SSRIs. While I know some of you will want to nitpick me over that assessment, please remember how studies like this used to address side effects issues in the past--hardly at all. So this new study represents an improvement on that front. For what it's worth to you all, CL Psych agrees with me on this point and there is no sterner judge of research methodology than that anonymous blogger.

All the same, my jaw is simply on the floor over something else that underlies the study's social context. Twenty years ago, children were not medicated for anxiety disorders (can you imagine giving an MAOI or tricyclic to a child for any disorder?), except perhaps in rare instances, nor were they packed off to a psychiatrist if they exhibited signs of anxiety. And yet here we are today with psych researchers talking about implications for American kids as a whole and while I am pleased to see a solid emphasis on psychotherapy, I am very leery of the broader message getting out there that kids with anxiety should be on an SSRI and CBT. Because in the real world of our wonderful economy and broken health care system, the kids aren't going to end up getting psychotherapy unless their parents are fairly well off. Zoloft in generic form is quite cheap. CBT can easily run $120 or more a session. Guess which option insurance companies will go for?

And after 20 years of our culture pushing SSRIs and other psych meds down the throats of adults with depression, anxiety and other disorders, can anyone point me to where this has improved things and what the argument would be for thinking it'll be any better for kids?

As usual, perhaps I am pushing the envelope a bit hard by asking these questions and raising these issues, but someone has to do it.

Posted by Philip Dawdy at November 3, 2008 12:05 AM

del.icio.us Digg it reddit

Comments

"then you need a way to compare the Zoloft plus CBT arm of the study to a placebo plus CBT group of patients in order to allow for whatever placebo effect--likely 20 percent or more--was present in the Zoloft plus CBT group. There was no such group in the study. Instead, we got a placebo-only group, a Zoloft-only group, a CBT-only group and a CBT plus Zoloft group."
Interesting. While I do agree that a placebo/cbt group would be great, I don't think it would have such a marked effect to bring it up to 70 percent. What if CBT acted like a placebo as well? That would mean that the result would stay the same. Ok, there's something to be said about the belief in popping a pill, but it wouldn't stack on to of the CBT IMHO.

Before I'd buy Doug Bremner's argument, I want to know several things
First was this the 24% for placebo that went down to 12%? If so, who cares?
And what was the standard deviation and range for both those who improved and those who didn't?
The claim sounds grand, but only because it is ambigous.

Posted by: Felix at November 2, 2008 09:34 PM

One other thing to consider is that here's no way to tell which kids will be "Zoloft responders" and which ones will not, or may even worsen.

Other types of drugs, antihypertensives to anti-cholesterol meds, are subject to the same peril to some extent, but psych meds are particularly troublesome in this way (and with, say, Zocor, at least you can count on someone's cholesterol not actually getting higher as a result of the treatment). And this is independent of the fact that it's difficult to objectively assess whether someone's depression has genuinely abated using available metrics.

What I take from this study is that it's a good idea to get kids who need help into therapy whenever possible. There's your starting point; whether or not to mix in an SSRI should be tabled for later.

Nice assessment, in any event. The lack of a CBT plus placebo group is not merely a mild oversight, it's a crippling design flaw IMO.

Posted by: bexter at November 2, 2008 10:26 PM

I interviewed the primary researcher in this study about 2 years ago, when I saw it up up on the FDA site.

In a 90 minute telephone conversation, he assured me the study was to "not have what happened to your daughter, happen to another child". When I called him on using Zoloft as a treatment in the study.

I'll find the notes, back then I threatened to shut this study down due to the Zoloft use in children, CBT or not those parents were paid to enroll their kids to use a drug that sent mine through the roof, and I was not happy with it.

This is why they contacted me to talk. They even wanted to talk to my daughter.

Oh well, what do I know.

Posted by: Stephany at November 2, 2008 10:32 PM

Bremner might want to check out Cl Pysch's blog re: Nemeroff, due to co -authouring 200 or so papers with him. Readers might want to check out Bremner's site, he specializes in PTSD.

In the meantime, I will be contacting my source re: this study and post the interview on my blog.

Posted by: Stephany at November 2, 2008 10:35 PM

PS-Philip do you remember me telling you about this study?!!!!!!!

Posted by: Stephany at November 2, 2008 10:39 PM

Mr. Bremner has some interesting findings such as "subjects with PTSD had a 5% increase in hippocampal volume and a 35% increase in memory function (Bremner, 2006)*." This hippocampal volume is also increased on London cab-drivers when they prepare themselves to the "The Knowledge" test.
I had a discussion on Fiddaman's Seroxat Sufferers (I'm having trouble to download this page) and was appalled by some of his findings. He claims that withdrawing Paxil takes a couple of months and is not aware of some already known SSRIs side effects such as PSSD.
I believe that the "Center for evaluation of risks to human reproduction" that is concerned with SSRIs being one cause of infertility should take some measures regarding children and teenagers being put on SSRIs.
This is beyond comprehension this use of Zoloft for anxiety specially in children.
A pill for separation anxiety?
This is absurd. These kids will have depression when they get older? Profilatic use of SSRI?
How can you predict it?
Zoloft was the drug Eric Harris, Columbine, was taking and was changed to Luvox by the time of the shootings.
I'm totally exhausted seeing all these absurds.

Posted by: Ana at November 3, 2008 06:20 AM

What really concerns me about this trial, to be honest, is the failure to provide more detail about concomitant medication and/or prior treatment. The study did not rule out children diagnosed with ADHD on "stable" doses of stimulants and, according to a table in the study, nearly 12% of the kids had ADHD. Over 2% of the kids had "tic disorder," according to that table, which can be a red flag for treatment with stimulants and/or neuroleptics and even antidepressants. They did exclude children who had not had "adequate responses" to two prior treatments of SSRIs, they claim in the text of the article, which means they could have been including children who did have "adequate responses" or else had just had one prior course of SSRIs. This throws in the distinct possibility that at least some of these events occurring in the trial might have been related to withdrawal rather than anxiety. Looking at the detailed adverse event chart available through a link on Clinical Psych's blog might make one think that. Whether or not that is the case the adverse events of kids on active treatment are qualitatively more serious than either of the "inactive" arms, including two incidents of homicidal ideation and a total of 21 serious adverse events out of 133 kids in the combo arm which gives me pause. As far as I'm concerned I'm afraid the failure to provide adequate detail on concomitant or prior treatment factors and how they related to adverse events nearly voids the whole study. I realize that this practice of ignoring concomitant (to say nothing of prior) treatment is very standard in academic research but that doesn't make it right.

75% of the kids in the trial were under 13 and the mean age was 10. Personally I think it's criminal to give children this age Zoloft. The 12 week window is of course the"honeymoon" period on antidepressants. This study is going to go out to 6 months and I suspect the results will be more favorable to therapy alone at that time. But after 6 months the kids on Zoloft will be well on their way to a diagnosis of "bipolar" and additional cocktails of medication. So yes, I am one of those who jump on studies like this and see red when I'm reading them, but I do think the science here, however "well reported" it appears to be is not the whole story.

Posted by: Sara at November 3, 2008 06:23 AM

Wow!! The homicidal ideation had two occurrences among 133 subjects in the Zoloft alone category. So this is Frequent, according to the criteria for adverse reactions - up to one in one hundred is Frequent and this is one in 66. No wonder www.SSRIstories.com has hundreds of cases of children in prison.

Also, Suicidal ideation on Zoloft combined with CBT had 12 compared to 2 on placebo. This is six times as many! I imagine the strain of undergoing therapy while on Zoloft is pretty darn high.

On Zoloft alone there were 5 who had suicidal ideation and only 2 who had suicidal ideation on placebo. This is about the same ratio that pushed the FDA to place the Black Box warning on antidepressants & suicidality for those under 25.

Also, Aggression was 2 for Zoloft alone and 9 for Combo but 0 for placebo. And "disinhibition" was 19 for combo & 3 for placebo and "defiant behavior" was 12 for Combo and 2 for Placebo - plus hostility was 7 for Combo and 2 for placebo, plus "Habits/Mannerisms/Rituals was 12 for Combo and 0 for Placebo. Also "Increased Motor Activity" was 21 for Combo and 5 for Placebo.

The only message I get from this chart is to be really, really careful with Combo!!

Posted by: Rosie at November 3, 2008 06:41 AM

This sort of study reminds me of the problem with studies years ago that "found" that post menopausal women who took estrogen supplements were healthier than pm women who didn't take it. It turned out that the researchers had the causation wrong. Years later it turned out that women who took estrogen were generally more affluent and thus able to afford medical care, to have healthier diets, and to exercise and just in general to afford the kind of recreation and mental stimulation that keeps a person young. Women who just took estrogen and were poor generally had a higher incidence of stroke, cancer and heart attacks than poor women who didn't take it.

What if these studies indicate that the experience of being in a study and being told that you are not bad for being anxious but "sick" is at least temporarily ameliorating the kids anxieties or if the kids, in unhappy families, simply are happy and relieved to be at the pshrinks getting relatively positive attention, or to get the attention they get from their parents or someone as they are observed in the study, or of course coming from poor families are happier and treated better while in the study because of the money the family is receiving? I wonder what results a control group of people whose only participation in the study was being paid would be.

Of course I agree with your points too and wonder how drugging a kid will help him if he's going to go on to have anxiety disorder as an adult, and of course how the kid is going to be harmed by being told he has an anxiety disorder in the long run. As many of your readers know, at first when you are in emotional distress it's comforting to be told you have a medical disease but in the end it's dehumanizing, invalidating and harmful to many of us, though I know others seem to find such a world view life saving.

And of course there's the problem of violence. Expect to see more studies showing that adolescents with anxiety disorders are more violent used as an argument for pushing these drugs when research really indicates that kids who take psych drugs are more violent and there's no real way to diagnose "anxiety disorders."

Can these drugs be helpful in limiting situations in children, probably. Perhaps with real trauma surviving ptsd kids in the short run even combined with therapy to work through the trauma and teach coping skills, but indefinitely, I'm not sure any study indicates such a thing.

And what about the sexual side effects? What are the emotional and psychosocial implications of chemically castrating teenagers?

Posted by: Sally at November 3, 2008 07:11 AM

Re-reading the article there was actually no change on the anxiety scale in any group except the CBT only group. I revised my own post on this.

http://www.beforeyoutakethatpill.com/2008/10/sertraline-kids.html

As for the other comments, those were previously addressed on my website under the post 'disclosure, seroxat, etc'. And the number of co authored articles is 8 (not 200).

Posted by: Doug Bremner at November 3, 2008 09:23 AM

I am curious as to the criteria for improvement from anxiety. "Positive outcomes..." Um, specific? I am curious also if narrowing the scope of a study like this to anxiety works the system to a more favorable outcome, closer to hypothesis.

Posted by: Sophia at November 3, 2008 05:40 PM

Doug,

I thought I found the 200 co-authored articles stat on your site; sorry for the mis-information of facts on my part here.

Though Google scholar is and easy reference for those interested in 'Nemeroff Bremner'.

Zoloft, of course we know was the drug Chris Pittman was taking in 2001 when he killed his Grandparents, and it's the one drug that sent my then 13 year old into suicidal/homicidal rage/attempt, etc.

My concern with the study when I spoke to one of the researchers in 2006 was that withdrawals, dosages etc were variables most parents had no idea of how to chart, or handle, or expect.

They also changed the title of that study several times, as the years have passed. At any rate, I speak from a personal perspective, we all have different life events that influence research or passionate discussion in this forum, and for me, this is one of them.

The researcher I spoke with in length also wanted to speak to my daughter about her Zoloft experience, yet unfortunately she was unable to due to becoming non-verbal. A decade later, I reside in a complex world with my daughter's care.

Respectfully

Stephany

Posted by: Stephany at November 3, 2008 06:40 PM

Bremner Nemeroff google search for starters.

Charles Nemeroff, of course was in recent news re: Emory, pharma money non-disclosure.

Bremner is a collegue at Emory.

Posted by: Stephany at November 3, 2008 06:55 PM

I've visited Mr. Bremner's blog and found nothing.
No substantial data on the most important issues.
It's really funny the new way of promoting the use of SSRIs and other psych-drugs: "yes! I know about ghostwriting, kols, conflicts of interest and the influence of pharmaceutical industry... blah blah .. .
but they work!
has little side effects... never heard or had a patient who had experienced withdrawal symptoms and I don't have a clue about PSSD.... go to my blog and you'll find the answers..."

"However, to my knowledge it isn't known if this effect persists after discontinuation. What else? That it can cause prolonged sexual dysfunction after discontinuation? In my experience sexual function improves after discontinuation, although as I have said before I am always interested in hearing about new problems that people have with medications. Do you have that problem?"
Dr. Doug Bremner on Fiddaman's
http://fiddaman.blogspor.com/2008/10/more-correspondence-with-doug-bremner.html

I've wrote you a post Mr. Bremner explaining PSSD, a condition I have already told you that is even on Wikipedia:
http://justana-justana.blogspot.com/2008/10/pssd-condition-still-not-known-by.html

I believe that suicide and violent behavior is already known.
Withdrawal symptoms are highly reported and you have plenty testimonies on the WWW kingdom and it's appalling that still, you have not a clue about it and can withdraw Paxil in a couple of weeks.
I hope you share your expertise on withdrawing with others.
I would be glad if you could answer about these serious problems.
According to David Healy it's impossible for some people to withdraw SSRIs

"Finally however there will be a group of people who are simply unable to stop whatever approach they take. Some others will be able to stop but will find problems persisting for months or years afterwards. It is important to recognize this latter possibility in order to avoid punishing yourself. Specialist help may make a difference for some people in these two groups, if only to provide possible antidotes to attenuate the problems of ongoing SSRIs such as loss of libido."
Thank you.

Posted by: Ana at November 4, 2008 12:15 AM

Stephany,

Using google scholar is not a good way to count publications because it magnifies whatever has been posted on the internet. Pubmed is also on the internet, free, and gives an exact count.

Ana,

I have responded to you repeatedly before on other web sites. As I have said before, I am not a spokesperson for paroxetine or SSRIs, but I am willing to answer your questions. First of all, I never said that SSRIs should be withdrawn in a matter of weeks. I taper over the course of two months, and if there are any issues I work with the patient to modify the withdrawal schedule. As to your being "appalled" that I was not aware of the syndrome of sexual dysfunction that persists permanently after discontinuation of SSRIs, I looked at the PSSD wikipedia entry you mentioned and the Bahrick et al 2008 article states
"The literature poorly captures the full spectrum of SSRI/SNRI sexual side effects, and a lack of systematic follow-up in the sexual side effects research precludes detection of post SSRI/SNRI sexual dysfunction, leaving the formal knowledge base inadequate..."

Since this article only recently came out, and since the literature on the topic is not well developed, although I am not precluding that you, or others, may have had this problem, I think that it hardly warrants such a reaction that I had not heard of such cases, although as I said before I am always interested to hear about new side effects of drugs that were not previously acknowledged.

Anything else?

Posted by: Doug Bremner at November 4, 2008 10:02 AM

Mr. Bremner,
I'm sorry I said "a couple of weeks" instead a couple of months. Since Fiddaman has said to you that is took 18 months to get rid off Paxil and it took me 2 years to get off of Effexor and there are many testimonies on the Web - I have some on my blog - and videos on YouTube and many evidences 8 months is nothing.
As I mentioned before on Fiddaman's blog I've only put the Wikipedia article because it's very pro-labs and the fact that PSSD is left there - PANES has been taken away a long time ago - is because there are serious concerns about this condition.
I don't know if you've noticed that I've put a link of a good article I've put for you on my blog written by Audrey S. Bahrick, Ph.D. on PSSD but you've only read the Wikipedia article.
"The literature poorly captures the full spectrum of SSRI/SNRI sexual side effects, and a lack of systematic follow-up in the sexual side effects research precludes detection of post SSRI/SNRI sexual dysfunction, leaving the formal knowledge base inadequate.

"Consumer reported information about persistent sexual side effects comes from the SSRIsex internet community. Founded in January of 2005, SSRIsex now includes a diverse membership of over eight hundred men and women who are struggling with sexual side effects that reportedly began on an SRI/SNRI, but that have persisted months and years after stopping the medications.
I believe you should take a look on this group to be aware of this little problem.
I'm sorry but a psychiatrists who is not aware of sexual side effects caused by SSRIs seems to be out of this planet.
Anything else?
Plenty!
I have raised many questions to you but as usual... you are not aware of many of them, never heard of a patient complaining of this side effect... - Wow! Haven't some of us been there with our psychiatrists two or three years ago? -
You're not concerned about suicidal or violent behaviour and plenty problems cause not only SSRIs but by other psych-drugs.
I'll not have the same kind of discussion with you as it was on Fiddaman when you was evasive and treated me with a little bit of sarcasm like the "Anything else?" you've just asked.

Posted by: Ana at November 4, 2008 12:44 PM

Errata: "evidences 8 months is nothing"
evidences 8 weeks is nothing

Posted by: Ana at November 4, 2008 02:16 PM

Yes, thanks for the research info site, I read a lot, (more than google searches).

Posted by: Stephany at November 4, 2008 09:50 PM

"My problem is this: It is now nearly three months since I stopped taking the Zoloft. Once the "physical" signs of withdrawal began to diminish (e.g., I haven't needed to take anything to sleep for three weeks now; I no longer experience the "itching" or the "passing-out sensation followed by muscle spasms") -- now, however, I seem to be mired in one of the worst depressions of my life. I've put on somewhere between 10-15 pounds over the past two months. My energy is low; I feel like I'm slogging through water much of the time. Basically, I'm sure I have all the classic symptoms of major depression, including powerful feelings of anguish, rage, and hopelessness -- except, thanks to many years working with a wonderful therapist and my own insight into antidepressants, I also recognize the perhaps iatrogenic nature of all this and am trying to ride it out, not to take it too personally."
" During the worst of it, my own doctor threw up his hands. He's a good guy, but was totally stymied. He didn't know, and the Neuropsychiatric Department at UCLA said no research into this subject is even being done (but I could come down, get assessed for meds!).
I believe it's an issue of tremendous importance -- to people like me, certainly, but also to the thousands if not millions of people who thoughtlessly and trustingly pop their antidepressant pills just to make it to work every day. Not to mention the future generations of children who, if they have their way and start using these drugs for so-called "early childhood" psychiatric illness, may end up having their brains scrambled forever. The spectre of permanent, unalterable damage is truly frightening. But I'll get off that pulpit and stick to my questions. I hope you'll find the time to reply. I've been perusing your site for some weeks before choosing to write you. Quite frankly, I was getting to a point where I wanted to call 60 Minutes in, immediately, to do a "whistle-blowing" story on what may be one of the most widespread medical horror stories of our time. Think, with millions of people popping these pills, they'd have an interested viewing audience? ... I'm selfishly concerned about me, also about the prevalence and damage from SSRIs to the lives of a much wider group of people."

This is part of a testimony I've saved on 07/29, 2005 but unfortunately I didn't put the link amd could not find it searching. It seems it's not online or it's from Socialaudit discussion board that is closed for the moment.

I'm just putting it here because I cannot understand how can it be that these problems are not known by researchers and some psychiatrists.
Keep on denying that these problems don't exist disappoint me on an unbearable degree.
The word I have to classify this behavior hurts. So I'll hold this terrible feeling I'm having because of these denials, that hurts me and I'm sure many others, to myself.
I'll try hard not to think about these issues for some days because it's overwhelmingly hard to bare.
And, Jesus, these people who are not aware of it all are those who will keep on making these drugs being put on the market for no matter what problem.

Posted by: Ana at November 6, 2008 04:06 AM

depression is bad. like, it's really REALLY not good

Posted by: corbin at January 1, 2009 06:39 PM