Nassir Ghaemi M.D., M.P.H.

Mood Swings

Misdiagnosed? Bipolar Disorder Remains Ignored

A 40-fold increased diagnosis still means a fraction of 1 percent diagnosis

Posted Dec 10, 2012

"From 1993-2004, the diagnosis of bipolar disorder in children increased by 40 fold (Moreno, et al 2007)".

This phrase must have been repeated by now thousands of times, and the same tired conclusion is drawn:  Childhood bipolar disorder is so overdiagnosed!  See?

But we never hear the full story, the scientific truth. This is only half the story, like the advertising for laundry detergent that doesn't tell you what it doesn't clean.

What was the baseline rate of diagnosis of bipolar disorder in 1993?   No one tells this part of the story.  It is in the same paper, however, in a nice graph:  It was basically nonexistent, as close to ZERO as you can get in this kind of research:  0.01%.

What was it after the amazing 40-fold increase, which so clearly proves overdiagnosis?  Again, it is in the paper in that nice graph: 0.4%.

It is arithmetic, as Jon Stewart would say:  40 times practically nothing, is still practically nothing.  Or 0.4% to be exact.  Less than half of one percent.

That's an amazing overdiagnosis rate.

Ask another question:  What is the "correct" diagnosis rate of bipolar disorder in childhood (assuming one is not a fanatic who refuses to ever believe that Nature would create this illness until you are old enough to vote).  According to the National Comorbidity Survey  (NCS) epidemiologic study of adolescents in the general population in the US, the observed rate of bipolar disorder, using standard DSM-IV criteria applied by epidemiological researchers (not based on clinician practice patterns, which is the issue at question) is 2.6%.  That's alot more than 0.4%, and the Moreno study included adolescents up to age 19. 

In another NCS study assessing age of onset of psychiatric disorders in adults who were diagnosed with them in the general population using standard accepted epidemiological methods, it was found that there was a lifetime risk of bipolar disorder of 5.1%.  This was associated with 13% onset before age 10. What's 13% of 5.1%?  




Seems rather close to me, even a bit low still.

How does this prove overdiagnosis?

It seems to me that we are setting right a long-standing underdiagnosis, as stated earlier in the post to which I am responding. 

By the way, it's simply false to say that bipolar disorder is a "progressive" disease. I co-wrote the chapter on natural history and course in the main text in the field, Manic-Depressive Illness, and we reviewed the literature concluding that such generic claims are not the case.  Also, the fact that someone improves who has bipolar disorder, without specific treatment for it, is far from showing that person does not have bipolar disorder. In fact, this is what one would expect: the episodes are episodes, which means they always end.  The average episode lasts weeks to months, even without treatment. Again, we have documented all this natural history evidence above.  

Science is about looking at the whole truth, not the half which agrees with our beliefs. 

More Posts