Bipolar Overdiagnosis: Are You Swayed?

An excerpt from my book review on the new book by the Brafman brothers, Sway, in bookstores shortly:

The one place the authors don’t really sway me is their attempt to explain why bipolar disorder is diagnosed so much more often than it was a decade ago. Unmentioned by the authors is the fact that many other mental disorder diagnoses have also experienced a significant increase in their use from a decade ago.

They link the increase to two factors — the modern diagnostic system put into use in 1980 with the publication of the DSM-III, which “broadened” the bipolar diagnosis; and pharmaceutical advertising in the 1990s. Left out of this explanation are some of the reasons proffered by the actual researchers of the study (Moreno et. al, 2007).

So what did the researchers who actually penned the bipolar “forty fold increase” say? Well, they were far more cautious about suggesting possible causes for the increase in diagnoses. But they did note that many of the symptoms of bipolar disorder overlap with other mental diagnoses, which could also be, in part, reason for the increase. For example, in a study conducted in 2001, nearly one-half of bipolar diagnoses in adolescent inpatients made by community clinicians were later re-classified as other mental disorders. Here is what one of the researchers of the study actually said:

“It is likely that this impressive increase reflects a recent tendency to overdiagnose bipolar disorder in young people, a correction of historical under recognition, or a combination of these trends. Clearly, we need to learn more about what criteria physicians in the community are actually using to diagnose bipolar disorder in children and adolescents and how physicians are arriving at decisions concerning clinical management,” said Dr. Olfson.

Sway’s authors’ suggestion that the increase in bipolar diagnoses is related to the modern diagnostic system seems to be reaching. If the DSM-III was the cause of the forty fold increase from 1994 to 2003 in bipolar diagnoses, why did it take more than 14 years to even reach the lower 1994 levels, long before the increase occurred?

The authors also link the diagnostic system back to its founder, Emil Kraepelin, and imply that the DSM-III (and its current version, the DSM-IV) have no links to “hard science” (whatever that is). Of course that’s not true — the DSM-IV is nowadays very much based upon empirical data; Kraepelin’s original categories have largely been discarded in the modern version. Kraepelin’s concept of bipolar disorder in the early 20th century was that it included both the modern version of “major depression” and what we now call “bipolar disorder.” He did not, however, describe bipolar disorder as know it today and the authors’ implication that this diagnostic category remains largely unchanged for nearly a century is just ludicrous.

As for pharmaceutical advertising, that’s likely a stronger link to the increase in diagnoses. Advertising largely works, otherwise companies wouldn’t bother. This too wasn’t a hypothesis of the researchers.

But neither explanation really goes to any irrational behavior on anyone’s part. Yes, once a patient is diagnosed by a mental health professional, diagnosis bias kicks in — we tend to view the person only in the filter of their diagnosis (and most other professionals will adhere to the original diagnosis, perpetuating the bias).

What the Brafmans do show is that diagnosis bias can lead to the patient themselves changing their behaviors to also fit the diagnosis. Once people are labeled, they tend to live up (or down) to those labels, or take on the characteristics of the diagnosis. The authors call this the “chameleon effect,” which is a person’s taking on positive or negative traits assigned to them by someone else.

Except for this one section of one chapter, I otherwise found the book even more enjoyable than Ariely’s Predictably Irrational.

Read the full book review for Sway now.