A Stupid Study and an Even Stupider Headline

http://www.medwire-news.md/47/90919/Psychiatry/Hypomania_common_in_young_adults.html.

A news article posted at the above website, MedwireNews, had the headline, "Hypomania common in young adults."  It reported on a study about to be published in the Journal of Affective Disorders by Serge Brand and others (currently available on line) and states rather unequivocally, "Hypomania affects around one-fifth of young adults..." 

Hypomania is supposedly a mood episode characteristic of a disorder called bipolar II in which a person has an elevated mood, but does not have the severe, often psychotic form found in a full-blown manic epsiode.

Wait a minute. Twenty per cent of all young adults may have bipolar II?  I mean, unipolar hypomania (recurring hypomanic episodes without any history of episodes of clinical depression, otherwise known officially as Major Depressive Episodes), is nowhere to be found in any version of the diagnostic book in psychiatry, the DSM, past, present, or future. 

If these research subjects - and actually they were all college students in Switzerland - really were having hypomanic episodes, then they should all have "Bipolar II" disorder.  That would be a preposterously high proportion of this population.

Furthermore, the journal article described in the news story also distinguished two subtypes of hypomania, the "bright side" and the "dark side" types.  This is the way that the news article summarized this finding from the Brand article: "Participants with dark-side hypomania had significantly higher levels of depressive symptoms, sleep disturbances, stress, negative coping strategies, and lower self-efficacy."  Dark side hypomania?  Sorta an oxymoron.  Gee, it sounds more like symptoms of an agitated depression than hypomania to me.

Of course, in my opinion - after having been in practice in two states with a wide variety of clinical populations (private practice, private hospitals, medical school outpatient department including screening potential patients for studies, state hospitals, county hospitals, and a veterans' hospital) since finshing my training in 1977 - the whole diagnosis of bipolar II is a figment of Hagop Akiskal's imagination. 

Hagop Akiskal is an academic psychiatrist I used to know. He did some of his early work where I am now on the faculty, at the University of Tennessee (UT) Department of Psychiatry.  Not coincidentally, he is also one of two editors-in-chief of the Journal of Affective Disorders, along with a Dr. C. Katona - the journal that is publishing the study in question.

I had tentatively formed a rather skeptical opinion of Dr. Akiskal's diagnostic procedures and acumen in mood disorders from listening to a few stray remarks from a couple of people who worked with him when he worked at UT, but I never had a chance to directly observe him in action with patients or research subjects.  However, I did hear him speak at a "Grand Rounds" (academic speak for an invited lecture) not too many years ago.  He made a couple of what to me were amazing proclamations.

First, he said that if a depressed patient was given an anti-depressant like Prozac and had became more agitated as a result, he just knew that the patient was bipolar. So I guess benzodiazepines like Xanax or Klonopin must be a cure for bipolar disorder, because they make the side effect of agitation go away immediately.

Second, he said that if he was referred a depressed patient who immediately displayed an angry, nasty attitude when they first met (I can't recall his exact words, but that was clearly the gist of what he was saying), then he just knew the patient was bipolar. Of course, patients who have borderline personality traits, as well as other personality problems, act like that all the time, so it sounded to me like he might be either unable or unwilling to tell the difference.

He had also told me in private on an earlier occasion that he knew that a lot of his "bipolar" patients had been abused as children - another characteristic of patients who have personality problems but something not exceptionally common in patients with true bipolar disorder. I have never personally heard of him saying anything about this in public.

In any event, I became concerned that he might be making rather hasty diagnoses before he had even done a complete psychiatric evaluation, and was seeing bipolar disorder where there was no bipolar disorder at all.

Hagop Akiskal

But even if bipolar II exists, a fifth of all young adults?  Please.  Interesting, the MedwireNews article neglected to mention any of the limitations of the study that were in fact mentioned in the actual journal article.  One of these limitations: "The pattern of results [of this study] may be due to other characteristics such as psychiatric disorders (e.g., eating disorders, personality disorders, and addictions) or motivation, which were not assessed in the present study." 

The study was entirely based on a self report instrument, the HCL-32,  that clearly should be limited for use only in screening subjects for an actual diagnostic interview.   All such screening instruments, as I have mentioned in previous posts, are designed to have a lot of false positives, meaning that many of the people who seem to be positive for a disorder based on the test do not have it.  They are meant to cast a wide net so that the experimenters do not miss subjects who falsely appear to be negative for the disorder. 

In the article that "validated" this self-report instrument by Angst et. al. (Isn't that a great name for a psychiatrist?), also conveniently published in the Journal of Affective Disorders (88 [2005] 217-233),  it states:  "Despite the use of broader and slightly differing criteria for BP-II, the HCL-32 still showed good discrimination between the unipolar and bipolar samples. The cut-off of 14 offered the best trade off between sensitivity (true bipolars) and specificity (true non-bipolars) with the total scale showing a sensitivity of 80% and a specificity of 51% for both BP. 

In other words, the instrument totally mis-diagnosed at least half of the population it was validated against as bipolar when they did not have the disorder.  This is typical of a screening measure, as I mentioned above. 

Unlike some of the other phony instruments used for diagnosing mania, the HCL-32 at least tries to address the duration criteria for a hypomanic episode mentioned in the DSM.  In order to distinguish reactive mood changes due solely to environmental events from true bipolar disorder,  a manic or hypomanic episode has to go on non-stop for at least a few days, last all day every day, and be completely atypical of a patient's usual functioning. 

The HCL-32 asks the subjects about the "Length of your “highs” as a rule (on the average)."  However, it does not address any environmental events the subjects might have been experiencing at the time; nor does it even mention anything about how the "high" feeling they were asking about should essentially never abate throughout the entire episode.

Many of the symptoms listed on the HCL-32 may be characteristic of people who are highly engaged in what they are doing, such as going to college classes, when they are in fact actively engaged, but which may later go away when things slow down.  What follows are some of the symptoms as they are described in the test. Answers are either "yes" or "no" with no opportunity for subjects to qualify or describe any complexity in their answer:

Please try to remember a period when you were in a “high” state.  How did you feel then? Please answer all these statements independently of your present condition. In such a state:


2. I feel more energetic and more active
3. I am more self-confident
4. I enjoy my work more
8. I spend more money/too much money
10. I am physically more active (sport etc.)
20. I make more jokes or puns when I am talking
21. I am more easily distracted
22. I engage in lots of new things
29. I drink more coffee

Do you know any active people who do not have periods like this?  I don't think I do.

The instrument also asks subjects to answer only for those periods when they were not high on drugs.  I do not know about Swiss college students, but a lot of American students use stimulants like Ritalin and Adderall as "academic steroids."  I wonder how many of the Swiss students might have thought that using these meds, since they are prescription drugs, did not qualify as times during which they were high on drugs? 

We will never know, because they were never asked.  Sleep deprivation can also lead to some of these symptoms.  They were not asked about that either.  Know any students who party a lot and don't sleep as much as they should?

The MedwireNews article makes it sounds like the Brand study is conclusive.  I talk in more detail about the process involved in the journey of facts that have not been at all established as they become accepted by doctors and the public alike as if they were established facts in my book, How Dysfunctional Families Spur Mental Disorders. 

In my opinion, the bastardization of news may have become another one of big PhARMA's strategies to expand psychiatric diagnoses so they can sell more drugs.