TCPR: Dr. Baldessarini, can you describe some of the difficulties in distinguishing bipolar depression from unipolar depression?
Dr. Baldessarini: This is a great challenge. In the U.S., we tend to overvalue antidepressants, and are quick to diagnose major depression in persons we don’t know very well. We make the mistake of talking only to patients and not anyone close to them. Patients complain bitterly about depression and dysphoria, but they don’t tell you anything about the other half of their lives—the upswings of mood and energy—so you may not realize the need to suspect bipolar illness. Or you may find out the hard way when you give them antidepressants and they become hypomanic, manic, enter a mixed state, or become psychotic.
TCPR: So how do we evaluate someone who comes in with an episode of major depression to be sure we don’t miss bipolar disorder?
Dr. Baldessarini: The first, and perhaps most important, tip is to seek some corroborative history. It is hard to find the time to get a relative or close friend into the office, and patients are sometimes reluctant to share what they may consider secrets, but you should at least make an effort.
TCPR: Good advice. And what else?
Dr. Baldessarini: I learned from Dr. Joe Mendels back in the 1970s that there are certain characteristics in a depressed person that might signal bipolarity. [See Mendels J et al, Arch Gen Psychiatry 1979;36:845–846]. For example, if the patient started getting depressed as a child, adolescent, or young adult, as opposed to midlife, that’s a red flag. Also, if there is any family history of psychiatric illness—a “nervous breakdown,” alcoholism, depression, mania, or psychosis. Also, the pattern of the illness is important. If a person has relatively frequent recurrences of depression, or is dysphoric a lot of the time with major episodes of severe depression at least once a year, you should suspect bipolar disorder. Unipolar depression patients, on the other hand, typically have much slower cycling every couple of years, or even less.
TCPR: Is there a typical constellation of symptoms that you see in bipolar depression?
Dr. Baldessarini: The textbook story is that if you see someone with prominent psychomotor retardation or rapid fluctuations of mood, energy, speech, and activity—so-called “lability”—you should think about bipolarity. Unfortunately, it doesn’t always work out so neatly. But on average, an anergic/lethargic pattern and affective lability with prominent dysphoria are quite common in bipolar depression. Also, the so-called “atypical” depressive symptoms of increased sleep and appetite. People with unipolar depression tend to lean more toward anxiety, somatization, guilt, and rumination.
TCPR: And we also tend to think of lack of an obvious psychosocial trigger as being a clue to bipolarity.
Dr. Baldessarini: That’s the common teaching, but if you dig deep enough, you can almost always find some stressor that could have caused someone’s depression. But in general, bipolar disorder is at least theoretically more endogenous and autonomous than typical unipolar depression.
TCPR: What do you ask a depressed patient coming into your office for the first time in order to rule out bipolar disorder?
Dr. Baldessarini: I want to know about family history, when did this depression start, attack frequency, nature of the symptoms, what has helped or hurt, what other things are going on that might be contributing.
TCPR: How do you actually ask someone if they have had a manic episode?
Dr. Baldessarini: That can be difficult, particularly in people who have never been hospitalized. Nevertheless, you can ask directly if they have ever been impulsive, hypersexual, made bad decisions, or had really high mood or energy that lasted for at least several days. However, depressed patients often just want to talk about their depression, and you have to ask vigorously about even mild hypomania, for example, “Do you find that your work production or your relationships fluctuate month to month, or are you on a stable course?”
TCPR: Can you define what a “mixed state” is?
Dr. Baldessarini: It basically involves patients who are at least dysphoric, but also present anger, agitation, maybe insomnia—features that remind you of pre-mania or hypomania. Another clue is that antidepressants can push depressed patients into a mixed state, which you may confuse with worsening depression. Then, you may end up throwing more gasoline on the fire by giving them more antidepressant instead of relying on mood stabilizers. DSM IV requires simultaneous, full manifestations of mania and major depression to diagnose a mixed manic depressive state, but such criteria are narrow. Mixed states are very common in both types I and II bipolar disorders (Salvatore P et al, Harvard Rev Psychiatry 2002;10: 255–275).
TCPR: Whether antidepressants cause patients to switch into mania is a perennially controversial topic. Can you share some of your thoughts?
Dr. Baldessarini: Some psychiatrists argue that when a depressed patient’s mood switches to mania or hypomania, the original diagnosis of major depression was wrong, and that you have uncovered “latent bipolar disorder” by giving an antidepressant. I personally lean in that direction. However, other experts believe that such mood shifts may be the toxic effect of an antidepressant, and may not mean that a patient has a bipolar disorder. They say: Back off on the dose or stop the antidepressant, and the patient will be fine.
TCPR: You and some colleagues recently published a meta-analysis of research done on mood switching (Tondo L et al, Acta Psychiatr Scand 2010;121:404–414). What did you discover?
Dr. Baldessarini: We found that, in patients with bipolar depression, antidepressant [AD] use was associated with a slightly higher risk of mania or hypomania—15.3% with an AD versus 13.8% without an AD. Interestingly, ADs had relatively much more of an effect on switching in patients diagnosed with apparent unipolar depression—with switch rates increasing by five times, to 6% from a baseline of only 1.2%.
TCPR: This finding seems counterintuitive, in the sense that one might assume ADs would be more likely to cause switching in bipolar depression than in unipolar depression. What do you make of these numbers?
Dr. Baldessarini: We suspect that the paradox reflects a ceiling effect in bipolar disorder. By this I mean that the rate of spontaneous switching in bipolar depression is already quite high, so adding an antidepressant is unlikely to increase the switch rate greatly. However, in apparent unipolar depression, the baseline switch rate is very low, so adding an antidepressant appears to increase the rate more dramatically.
TCPR: But doesn’t this finding challenge the common opinion that antidepressants are likely to cause manic switching in patients with bipolar depression? After all, antidepressants did not significantly increase the switch rate in these patients.
Dr. Baldessarini: Yes it does, but that is nonetheless what we found in our review. In some of these studies patients were also on mood stabilizers, which might have lowered the average switch rates. We also found evidence that tricyclic ADs (TCAs) have a particularly high switch risk, and my clinical impression is that switching is a dose dependent phenomenon, meaning that you should use low AD doses whenever possible in a depressed bipolar disorder patient. Overall, despite our findings, I still believe that we should be very cautious about using an antidepressant in a depressed patient with known or suspected bipolar disorder. Even if the relative increase in risk with the AD may not be as great as many colleagues expect, it is high, and the situation can be clinically dangerous.
TCPR: Is there much information available on the long term effects of antidepressants in bipolar disorder?
Dr. Baldessarini: We did a meta-analysis in 2008 looking specifically at the effects of long term AD treatment in bipolar depression. Although the available studies on this topic are very limited, it appeared that long term AD use after recovering from an episode of bipolar depression yielded a small prophylactic benefit, perhaps reducing depression recurrence or chronicity by 20% to 30%. On the other hand, such treatment was associated with increased mood switching and emotional destabilization of 70% or more (Ghaemi SN et al, Acta Psychiatr Scand 2008;118:347–356). These findings do not indicate a favorable benefit/risk ratio long term, but given the dearth of data, the jury is still out.
TCPR: Another clinical pearl is the idea that we should never prescribe an antidepressant to a patient with bipolar disorder without a mood stabilizer on board. Has this stood the test of time?
Dr. Baldessarini: In this meta-analysis, we also looked specifically at whether an anticonvulsant or lithium given with an antidepressant reduces risk of mood switching with antidepressants in bipolar disorder patients. Again, there was a paradoxical finding, in that the risk of switching was greater with a mood stabilizer than without, and this outcome was totally unexpected! The basis of this paradox almost certainly is methodological, since most of the data we reviewed were clinical and not from randomized trials. It is likely that the clinicians in these studies prescribed mood stabilizers to patients who had been, and seemed most likely to become, manic. So the apparent association of mood stabilizers with more switching may be an artifact of this clinical practice. To fairly test the widely assumed value of mood stabilizers, we need prospective studies that randomize bipolar depression patients to AD treatment with and without mood stabilizers.
TCPR: Thank you, Dr. Baldessarini.