TCPR: And course of illness?
Dr. Phelps: This is a broad category and refers to multiple illness descriptors. The three strongest predictors are: post-partum onset of mood symptoms, the presence of psychotic features, and highly recurrent unipolar depression. Other descriptors that are less predictive are a rapid onset of depression, a relatively short duration of a depressive episode, and “atypical features” (hypersomnia, hyperphagia, leaden paralysis, rejection hypersensitivity).
TCPR: So, for example, if a patient were to present with highly recurrent depression with psychotic features, you would be quite suspicious of bipolar disorder even in the absence of a history of manic episodes.
Dr. Phelps: Right. And the final clue to bipolarity is response to treatment, which technically includes a positive response to lithium, but practically the most important is a negative response to antidepressants, particularly if that negative response includes hypomanic symptoms.
TCPR: Dr. Phelps, before we begin, I know that you have always been very up front about the fact that you speak for drug companies, and I wonder how you think this might influence what you have to say about the diagnosis of bipolar disorder?
Dr. Phelps: The bottom line is that it is very hard to know with certainty that one is not being influenced. When I give talks funded by Astra Zeneca, I wonder if I’m being perceived as a mouthpiece for the company. For example, today I’m sure we’ll talk a bit about my belief that antidepressants are not appropriate for bipolar depression. I might be perceived as indirectly promoting Seroquel by discouraging people to use antidepressants. I don’t see it this way — I’ve looked carefully at the evidence base and believe that Seroquel has some of the best evidence for the treatment of bipolar depression.
TCPR: I’m interested in your take on the idea that bipolarity lies on a spectrum. I read your recent article in the journal Bipolar Disorders, in which you and your colleagues at the International Society for Bipolar Disorders review the most recent evidence in support of this approach (Phelps, et al., Bipolar Disord. 2008 Feb;10:179-93). Elsewhere in this issue, there is an interview with Mark Zimmerman about his new paper presenting a different viewpoint, namely, that bipolar disorder is sometimes overdiagnosed. Do you think that psychiatrists are overdiagnosing bipolar disorder?
Dr. Phelps: To say there is such a thing as overdiagnosis means that there is a gold standard for bipolar disorder, but that doesn’t exist. The closest we have are the current DSM criteria for bipolar disorder, which are already 15 years old and there is contention about their validity.
TCPR: What is wrong with the DSM criteria?
Dr. Phelps: They are too narrow. I believe in a broader concept of bipolarity. The classic paper that I often refer to was written in 2002 by Ghaemi, Ko, and Goodwin and defined the notion of “bipolar spectrum disorder” (Ghaemi et al., Can J Psychiatry, 2002;47:125-134). The way they did this was to focus on two characteristics that very frequently accompany bipolar disorder — family history of bipolar disorder, and hypomania in response to antidepressants — and then to survey what other characteristics are also correlated with bipolar outcome. They defined 10 variables, but I find it useful to put these variables into 5 larger categories, which have been used by Gary Sachs as the basis for what the STEP-BD calls the “Bipolarity Index.”
TCPR: And what are they?
Dr. Phelps: The first is what most clinicians would consider to be the gold standard — the DSM-IV criteria for a manic or hypomanic episode. The other four are sometimes called “soft signs” of bipolar disorder. I often refer to them as “non-manic bipolar markers.” They are: 1) family history, 2) early age at onset of depression, 3) course of illness, and 4) response to treatment.
TCPR: Let’s start with family history of bipolar disorder. While this sounds like a pretty straightforward piece of information to attain, in practice it’s not so easy.
Dr. Phelps: This is true. When you ask about family history, you have to dig for information about that relative’s life and course of treatment, using the same earphones you use when listening to the patient’s personal history. For example, if you hear something vague like, “Uncle John had problems and was hospitalized,” you have to spend some time asking more detailed questions. What were Uncle John’s behaviors? How many times was he hospitalized? What medications did he take? You might find out that Uncle John was hospitalized 5 times for mania and was given lithium and Depakote. You’d be pretty certain of a family history of bipolar disorder.
TCPR: What should we be looking for in terms of age of onset?
Dr. Phelps: The studies suggest that if the first onset of depression is between 18 and 24, these patients are more likely to have a bipolar course over time.
TCPR: Some people have proposed that patients with hypomania in response to antidepressants are in a special category, “Bipolar Disorder, type 3.”
Dr. Phelps: Yes, and this reflects the results of one particularly intriguing study by Akiskal and colleagues in which they followed patients who had AD-induced hypomania over a period of time. They found that virtually 100% of these patients ended up meeting standard criteria for bipolar disorder (Akiskal et al., J Affective Disord 1983; 5:115-128).
TCPR: In my practice, however, it is not always easy to determine what constitutes a “hypomanic” response to an antidepressant. I often have patients reporting that they feel extremely good, even euphoric, for a brief period of time after being started on an antidepressant. That euphoria lessens over time, and then they settle into a typical antidepressant response. Should I be calling this antidepressant-induced hypomania?
Dr. Phelps: Probably not. But I have seen another variation in which mild euphoria does not diminish so quickly, lasts for two or three weeks, and when this goes away, the patient is often again depressed. I would classify that as a hypomanic response.
TCPR: The other difficulty is differentiating SSRI-induced agitation or akathisia from hypomania. When I go through a patient’s past psychiatric history, it is not uncommon to hear something like, “I once took Prozac and I felt like I was crawling out of my skin. I had to stop it right away.” Sometimes such patients will then report that they did fairly well on a different SSRI.
Dr. Phelps: My approach is to avoid jumping to conclusions in these cases. I’ll ask, “What happened when you took Prozac,” and then I’ll listen with unbiased ears, as best as I can manage. If a patient says, “Oh Prozac? That was awful — my mind was flying, and I couldn’t go to sleep.” To me, that is beyond akathisia, and it qualifies as a hypomanic response. My criterion is an “antidepressant misadventure,” as a primary care colleague put it: something that was really bad, not just jitteriness. But I won’t hang my eventual diagnosis on this alone. In my subsequent interview, if I find other soft signs of bipolar disorder, I’ll feel more confident that this was, indeed, an antidepressant response to be concerned about.
TCPR: To take a step back, there’s clearly a great deal of disagreement in the field about how broadly we should set the standards for what constitutes bipolarity. Many would ask, “What’s the point — isn’t this really hairsplitting?”
Dr. Phelps: In clinical terms, the meaning of the effort to assess bipolarity is, “How willing or unwilling should I be to give this patient an antidepressant?”
TCPR: If it turned out in five or ten years that the research clearly indicated that antidepressants are safe in bipolar disorder, would there still be any reason to worry about bipolar spectrum?
Dr. Phelps: To be frank, probably not. If there were no treatment implications, this would be an empty exercise in diagnostic semantics. In looking at the research over the years, I believe that giving patients with even soft signs of bipolar disorder an antidepressant is usually a mistake, at least in the long term. [Ed. note: For more information on the use of antidepressants in bipolar disorder, see the article “Antidepressants and Bipolar Disorder: An Update,” in this issue.]