Tuesday, June 18, 2013

The Treatment of Early Age Mania Study Revisited

The Treatment of Early Age Mania (TEAM) study is not news, and more diligent and timely bloggers have already written about it. However, it is one of the more infuriating outcomes of the whole pediatric bipolar disorder phenomenon; in the last few years, few publications have irked me as much as the ones from this study. Thus, I'd like to chime in as well. 1 Boring Old Man had a lengthy blog post (the sound and the fury...) summarizing the main findings [I added the links]:
The first report shows that these children respond better to Risperdal® than either Lithium or Depakote®. That’s no surprise. They don’t respond to the traditional anti-manic treatments [that suggests to me that they don't have mania]. They do respond to Risperdal®. That’s something we knew before we ever heard of Risperdal® – you can control disruptive behavior with antipsychotic medications. It also showed that the metabolic side effects of Risperdal® were already apparent at only 8 weeks. Notice that there’s no placebo group in this study so we can’t really say that the Lithium or Depakote® responses were clinically significant.

The second report set out to define moderators of response. What it ended up showing was the extremely high overlap between ADHD and the presumed Bipolar Mania and the more ADHD, the greater the likelihood of a response. But there was another moderator of response – site. What in the hell does that mean? To me it suggests that there is bias in making this diagnosis or in measuring the response. I think that says something about the study and the diagnosis, not the afflicted. So in my reading, one thing it doesn’t mean is that Bipolar Disorder has some intrinsic regional difference.
1BOM used the study to illustrate how a fad diagnosis could become mainstream, to the detriment of children and their families, while benefitting pharmaceutical companies and the researchers (such as MGH's Joseph Biederman) they support. It's worth noting, though, that the TEAM study did not include the MGH group and was funded by the NIMH, not by pharma. The first author of the initial publication is Dr. Barbara Geller, and I've written about how her conception of childhood bipolar is different from Wozniak/Biederman's chronically irritable patients, with Dr. Geller trying taking into account more classical manic symptoms such as grandiosity. Thus, I've often wondered "what the hell happened?" when pondering this study.

Fortunately, Dr. Stuart Kaplan, child psychiatrist and author of Your Child Does Not Have Bipolar Disorder, has an excellent series of posts on his Psychology Today blog with additional insights into the TEAM study. In Dr. Kaplan's first post (The World Series of Child Bipolar Disorder), he describes a session from the 2011 AACAP meeting in which TEAM researchers talked about their study:
During the discussion, another nationally known presenter gave a wildly incorrect interpretation of defiance. The presenter claimed that defiant children are psychotic because they have a delusional belief that they can take on the far stronger adult world. Defiant children are not psychotic based on their defiance alone. They are mistaken in their belief that they can overpower the adult world, but this is a mistaken belief not a delusion. If the investigators believe that defiant children are delusional, this may explain how they found the high rates of psychosis in the children they studied (77%).
If defiance in children counts as psychosis, then my partner and I are both psychotic every time we argue, because we each have a false belief that we can convince the other with our arguments. Maybe some of the TEAM investigators would consider this folie à deux?

Dr. Kaplan's second post (Credulity Stretched) highlights the reasons why the children included in the study probably did not have bipolar 1 disorder, given the >90% comorbidity with ADHD, the 99.3% of patients with "daily rapid cycling" moods, and the fact that the average "manic" episode in the study lasted 4.9 years, which is about half the life of the average study participant (mean age 10.1 years). His third post in the series (Location, Location, Location) replicates the table from the second publication showing just how wildly variable the treatment response was at the various sites: "This was not a minor statistical artifact, but was the central finding of the study."

Besides highlighting the incredible (as in, not credible) aspects of the study that Dr. Kaplan already wrote about, I wanted to provide one additional anecdote: Several years ago, one of the renowned lead investigators of the TEAM study gave a talk at another institution. This mood disorder expert claimed that a 3-year-old who masturbates may be exhibiting the hyper-sexuality seen in mania. When audience members pointed out that a 3-year-old masturbating is actually normal behavior, the investigator appeared flabbergasted. Which makes me wonder if they thought 6-year-olds who were repeatedly touching themselves in defiance of parents telling them to stop were having manic episodes with psychosis.

One of the biggest problems in the field today is how biologically-oriented psychiatrists look at behavior in a vacuum without considering developmental, social, or familial factors. This study is one of the most egregious examples, not just of that problem, but also of how researchers at prestigious institutions, backed by NIMH funding, can get even the most ridiculous studies published. If I see a 16-year-old who is truly manic, I'm still going to seriously consider lithium over risperidone, "evidence-base" be damned.