Click here to view next page of this article
Mood disorders are common, deadly and treatable. The prevalence for bipolar mood disorder is 1-3% of the population. Bipolar disorder accounts for very small subgroup of patients compared to those who have depression, who account for 8 to 15% of the population. Depression has a much worse outcome for untreated disorders than even some forms of cancer. 15-20% will die by suicide if untreated. The good news is that mood disorders are highly treatable depression and mania.
The frequency of suicide has a bimodal distribution. We have a low peak in the year over the spring months and a secondary peak in the fall. Hippocrates wrote that spring is the time the black pile runs, right? The black pile being melancholia and the reason Hippocrates said that is because his teachers taught him that it was autumn and you can see that both of them were right. We have two peaks during the year. This is very consistent and it’s not just the shape of the curve for suicide but also for suicide attempts. The new onset of affective episodes tends to be in the spring and the fall all over the world although the spring peak is obviously October/November.
People have looked very carefully at why this might be happening and you notice that the peaks are centered on May and October, months when you might say if ever psychiatrists should be there to return the phone calls. So mood disorders are common. Unipolar there’s a 2:1 ratio of female to male where bipolars, at least for most forms of bipolar illness and most studies find a 1:1 ratio. If you are ever asked about the gender ratio for rapid cycling, it’s 3:1, maybe even 4:1 in favor of females where bipolar IIs there may be a 2:1 females but overall most studies don’t find that. Most studies find that there’s a 1:1 ratio for most forms of bipolar illness.
Bipolar I you see it has got 1% of the population. When people ask questions about the prevalence of bipolar illness they often mean what is now termed bipolar I and we’re going to clarify that later. But for now I just want you to get the idea that there are subtypes within the mood disorders.
Because the importance of these numbers to me is not so much the numbers themselves. I don’t think we worry a lot about epidemiology but there are some important lessons to be learned from how these numbers are collected. We could do various kinds of psychiatric assessments. There’s great limits on the validity and reliability because there’s no gold standard. How long do episodes last? Another very important issue and this is data from my colleagues at Pittsburgh but it happens to be almost precisely the same as the data from our collection. Time to remission for pure manic episodes, half the episodes have remitted at 20 weeks. Now that may not sound so good to you and you should know their definition – that 20 weeks required.
So that’s kind of good news and then even better by 40 weeks, everybody who had an acute episode of mania is well. Maybe that doesn’t sound so good to you but look at the alternatives. What if you had an episode of depression? You are going to have a much more crowded course. For people with pure depressive episodes the median duration now is out at 40 weeks and we never get to here. At the end of a year, 42% of these patients are still ill. In our clinic at the end of two years 23% are still ill and you’re getting borderline.
If that isn’t bad enough, what if you have a mixed episode? If you have a mixed episode, now the median duration is out to 50 weeks and 45% are still ill at a year. Well, that’s pretty bad. So depression confers chronicity.
That’s a quick overview of development of the concept and the course of symptoms. Now let’s go and talk quickly about core concepts in the DSM. The single most important concept to make sense of the DSM-IV is that the DSM-IV distinguishes the episodes from the illnesses comprising those episodes and it also gives us so-called course specifiers. So these are three separate concepts to understand.
The course specifiers are rapid cycling, meaning the patients have had four episodes a year, post partum onset which is having an episode of mania depression within a month of delivery and then seasonal patterns simply means over several years the episodes have the same onset and offset in the time of the year predominantly. It doesn’t have to be that every episode occurred in the fall or winter but predominantly. You’re left to make a judgment about that. So those are the four specifiers.
These are specifiers because somebody with bipolar illness could be rapid cycling or a unipolar could have a seasonal pattern and they are not separate illnesses. Any unipolar or bipolar could have any of these.
The acute episodes, as you know well, are depression, hypermania, mania and mixed. These are acute episodes but you can’t give a diagnosis code for one of these four. They’re always coded relative to a disorder so the mood disorders are categorized as primary. We’re going to talk a lot about unipolar and bipolar in a minute but worth noting the other category secondary – induced by a substance or a general medical condition. This means you think that is etiologically important in the development of that episode. So if somebody were using alcohol or cocaine, antihypertensives, steroids, they would have a substance induced secondary episode but the episode would be diagnosed using the same criteria as it would be for a primary episode.
So depression is depression. Mania is mania. If you have a pancreatic cancer and get depressed we diagnose depression because it meets the criteria for major depression but we call it secondary because of the pancreatic cancer or this drug or the TLB.
"Residual mood disorder not otherwise specified" is seldom used but it simply means there are prominent mood symptoms and it doesn’t meet criteria for any of these.
On the unipolar side, looking at the primary mood disorders, we have major depression and its little brother dysthymia. Major depression and dysthymia are unipolar because notice there is never any period of abnormal mood elevation. For major depressive disorder we have to have at least one episode that meets criteria for major depression.
Dysthymia is a chronic condition where we have low level mood symptoms, disinterest, dysphoria and this goes on for a period of one year in children and adolescents, two years in adults where more than half their days they have either depression or disinterest and half their days two symptoms of major depression. Anytime that their mood is normal in this year or two year period lasts less than eight weeks and during this whole time they didn’t meet criteria for major depression. Now that could have happened to somebody when they were age 18 and now when they’re 24 they have major depression. They would have so-called double depression. They had the prior dysthymia and we don’t take dysthymia away because years later they got an episode of major depression. We would only take it away if during this period of one to two years they got major depression.
On the bipolar side, we have three defined conditions, bipolar I, bipolar II and their little brother psychothymia. Bipolar I requires at least one full manic episode. No depressions are required but mostly patients have had lots of depression. In fact, bipolar patients tend to have more depressions by far than they do manic episodes but it’s not required to diagnose bipolar I.
Bipolar II, on the other hand, patients suffer a hypomanic episode but never mania and must have at least one episode of major depression. Why do they require major depression for bipolar II and not bipolar I? Very simple. If you go to Yale at the end of four years there’s a 96% chance that you would endure symptoms sufficient to diagnose an episode of hypomania but very few of them actually get depressed. So what is that they were describing? They were describing falling in love, winning an award, getting good grades, the things that happen to them and there really isn’t much specificity to hypomania particularly during those years until you add that depression in. Then it’s a lot more meaningful particularly if it was recurrent and I know if Hugo Pakiska was here he would say you can always tell the difference between bipolar illness and falling in love because bipolar illness is always recurrent.
Cyclothymia, like dysthymia, the patient has had it a year for children or adolescents, two years for adult where they have frequent hypomanic symptoms. It doesn’t say frequent hypomania just frequent hypomanic symptoms and half the days are either too high or too low. That can mean you have 20 days of dysphoria and two days of mood elevation each month and you would be cyclothymic not dysthymic as long as any period of euthymia was less than eight weeks, you never meet criteria for mania or you would be bipolar I. During this period of time again you don’t meet criteria for a major depressive episode.
It sounds like it would cover the waterfront but it doesn’t. It’s worth knowing about bipolar NOS which is in the book. This is somebody’s got some period of clear-cut abnormal mood elevation. Maybe they’ve had one, two or three or more hypomanias but never depression, that’s a bipolar NOS not a bipolar II. Maybe they’ve had a period where they look cyclothymic but it only lasted 10 months. We’re not going to ignore that they’re bipolar NOS. Even if it’s for less than two months they’re bipolar NOS.
You often hear the term bipolar III bandied about. There are ten definitions in the literature for bipolar III. It is not in the DSM but when you hear it you can know that some people mean cyclothymia, some people mean antidepressant induced mood elevation but actually one that we find useful is this one. Bipolar III is somebody who themselves has never had any period of abnormal mood elevation but they have a first degree relative, a parent, child or sibling, who does have bipolar illness and they themselves have recurrent major depression. That is to say if my brother and I have both been depressed and you find out that my brother has been manic, when I come to you for treatment you might want to say to me, "Your chances of having the same illness your brother has are very high. We might characterize you as bipolar III and maybe we shouldn’t just give you an antidepressant alone for your treatment."
Now, I’ve been talking about 35 minutes and I’ve probably said the word "mood" over a hundred times giving you the idea that these are mood disorders. Why, if they’re mood disorders, do we see that 30-40% of these patients receive chronic treatments – at least bipolar patients – with antipsychotics. This is not in the days of atypicals. This is data from before that when every expert recommended not giving neuroleptics to affectively ill patients chronically.
Well, there are some reasons for that. One of them is that psychosis is not uncommon as a part of mood disorder. This is data for bipolars and you see in twenty studies since 1970 about half the patients in any episode had delusions, 15% hallucinations, 18% or just about 1 in 5 patients have the Schneiderian first rank symptom. Right? That’s what I was taught discriminates schizophrenia from affective psychosis which is not the case.
Over the course of a lifetime more than half the patients have some history of psychosis and you can see in some studies it’s three-quarters of the patients depending on how you define it when you break them up. So psychosis is common as part of affective illness and it tends to be more common the earlier the age the person is when they first get ill.
Diagnostic criteria for mania. Here we require a distinct period of abnormal mood that must be euphoric, expansive or irritable. If euphoria or expansive mood is part of it, you only need three of these associated symptoms but if it’s only irritable then you need four. Now remember when I told you the criteria for major depression. It says the symptoms need to be there most of the day nearly every day. That’s not what it says here. Associated symptoms must be present to a significant degree during a week. Now, that leaves you to figure out what "to a significant degree" means. That’s your judgment call. It doesn’t mean that they were constantly talking.
These associated features include increased self esteem or grandiosity, decreased need for sleep as opposed to simple insomnia, pressured speech or just going on and on and saying more than anybody else in the room. Flight of ideas or racing thoughts, distractibility, increased goal directed activity and then there is my favorite criteria, participation in pleasurable activities with potentially painful consequences. This is risk taking and this is here because of the word "potentially".
Why? Because if you are participating in activities that are potentially dangerous, whether you get caught or not is immaterial to the diagnosis. So if you are embezzling money from your hospital and you don’t get caught, you still meet this criteria. If you were driving down the street here in the city at 90 mph and you don’t get a ticket, you meet this criteria. Remember the word is "potentially". That means we have an opportunity to intervene before there’s a disaster if we make the diagnosis. This criteria is often what helps you discriminate mania from hypomania as you’ll see in a bit.
Lots of times patients have no idea how to evaluate the risks that they’re taking. This is a guy that I always wondered did he ever have a psychiatric exam. Remember David Wolf the astronaut physician who went up and just before he went up on the shuttle to dock with MIR he said, "Flying MIR seems perfectly safe to him."
Sometimes patients present with excessive talking and when you ask them about pressured speech they say no. This gives you some idea about insight but more often they will endorse pressured speech outside of the interview if you simply ask about it. They’ll tell you that people cross the street to avoid having conversations with them. They’re often very aware of it. Don’t rate that as absent because you don’t see it in the office. Take their word for it. They’re telling you the truth.
Sometimes you will be the target of the patient’s being easily annoyed by everything you say but more often it’s somebody else so ask about that even if they seem perfectly fine to you if any little thing gets them going. Sometimes it’s a cop and often the art of interacting with these patients is when you see that you’ve just waved that red flag. A couple of words here that fit over all the patients but most particularly for adolescents where we get into this question of differential diagnosis versus comorbidity. Once the patient has met criteria for mood disorder, they’re very likely to have a comorbid condition. The only one on this list that is inconsistent with the mood disorder diagnosis is schizophrenia. Patients often have disruptive behavior disorders, anxiety disorders, substance abuse.
What is the significance of having that comorbidity? Again, this is one of the key points to remember. Remember those epidemiological studies? Remember all the people that got diagnosed? Well, 86% of people who got any diagnosis only met it for a period for one or two. The other 14% who had the diagnosis actually had three or more diagnoses. What does that mean? Well, if you focus on this slice of the pie, what you see is if you follow them up for a year this group accounts for nearly 60% of those who have a new episode in any year and if we look just at severe episodes, 90% of severe episodes come out in this group.
So when you have a patient, when you do your evaluation and you say, "Oh, my goodness. They have an anxiety disorder. They also seem to have a comorbid ADHD and they clearly meet criteria for mood disorder, that’s not good." They’ll need to be watched closely. Maybe that shows us our cookie cutter diagnostic system is inadequate because if you look carefully at the symptom overlap for mania, ADHD and let’s say conduct disorder, what might impress you is for instance if you have increased goal directed activity and psychomotor agitation that overlaps with six symptoms of ADHD, three of conduct disorder and this pleasurable activities overlaps with ten symptoms in conduct disorder.
So if you were to have five, six and seven, you would definitely meet criteria for all three of these. Right? That would be kind of worrisome. Can we disentangle them? Well, data from Janet Wolzniak, I think, showed you that in the case of ADHD you really can. She looked at consecutive referrals to the child psych unit with mood problems and when they had bipolar illness 96% of those who had hypomania or mania also met criteria for ADHD. But when she went down to those referred in for ADHD, only 19% met criteria for bipolar illness.
What does that mean? Well, if you think about this for a minute, I think we might agree, what we can say here is that if people really seem to meet criteria for both, the least likely thing is ADHD alone. They may just have bipolar illness, they may have both but the least likely thing is this is just an ADHD kid. So when you’re confused the first thing to do is not reach for the Ritalin. It’s better to reach for the Lithium or Depakote to treat that and see how much improvement that brings and then add stimulant if you still need to.
Also from Jo Mirian’s group we see the risk of new onset of bipolar illness over a four year followup, presentation has all three of these disruptive behavior disorders – ADHD, oppositional defiant and conduct disorder. If you diagnose that, you may not want to be around for the next four years treating this patient because within four years more than 40% of these patients with have a hypomanic or manic episode. But if we just take away the conduct disorder you see it’s less than 10% if they just have oppositional defiant ADHD. ADHD alone is about 6%. So having these are bad but really the worst thing is to have a comorbid conduct disorder. That foreshadows for you where the difficulties are.
How can you tell ADHD from bipolar disorder? It’s pretty difficult. People will look at the hyperactivity and say that’s more goal directed for bipolar and more scattered with ADHD. The sleep disturbance is episodic with bipolar, consistent with ADHD. Impulsivity here is reactive versus it seems consistent and self generated.
Those three things, to be completely honest with you, I think are hit and miss. I can’t really feel good relying on them. What I rely on if I have to make this discrimination is the last one – social or academic function – and I don’t ask about the year that they failed out. I want to know about the year that they did best, socially and academically because an ADHD kid, the bipolar kid will definitely have bad times that look indistinguishable but the bipolar kid often has periods of above average function even without treatment. So somebody who is on the outs with all their friends in fifth grade but in eighth grade was voted most popular, that’s rare with an ADHD kid. If they were spelling bee champion in the fifth grade and failing in the sixth grade, that’s rare for an ADHD kid. That ADHD kid they failed both grades without treatment. So we’re looking for the period of best function and I do find that to be helpful.
I can’t really find anything helpful to discriminate conduct disorder from bipolar disorder and this may be telling us something but lifetime comorbidity is about 70% in the studies from Kovax and Pollock. Episode comorbidity is 54% with any one episode. Conduct disorder gets diagnosed first in about 40% of kids and about a quarter of them bipolar first and maybe that’s a little bit helpful.
How to do the assessment. You want to have multiple sources of history if at all possible. We do not rely on a patient’s self reporting for the diagnosis and we will not treat people although we will do a consult and that will be it if we can’t have another source of information. You want to get a good family history, a good medication and substance abuse inventory.
Physical exam, you don’t just go and do a head to foot evaluation of everybody but your physical is directed by the findings of your history. You want to do a good cognitive status exam. At least some brief cognitive test and evaluate their mood state.
Then what about labs? Well, labs are going to help you sort out the secondary issue very often. Here I think as a minimum you’re going to get serochemistries, a TSH and a CBC and then if necessary you can get a sed rate, BDR and all, HIV, toxicology screen and then some people would say imaging. Some people would also say you should get an EEG. I really believe that all of these should be done at least once per patient. You don’t have to do them every episode but if somebody’s presented with mania, you might want to do this and see what’s going on there.
Differential diagnosis. In most areas we concentrate on the psychiatric illness first and you always start by ruling out the secondary causes so we are looking to our physical exam and labs to help us figure out what of these things might be going on.
We can also look at the difference between drug and placebo of these trials. 15, 23 and 18%. Not significantly different and then we can look between any two drugs of the same class, less than 2%, about 2%. Here this is the difference between bupropion and a tricyclic. Actually favored bupropion being 7% more effective. So we can see whichever of these criteria we set there’s not much difference between these drugs. That is not to say the individual wouldn’t respond preferentially. It is to say that if you took a thousand patients and you gave them all any of these treatments you would get a very similar percentage getting well. But any one patient that’s not necessarily the case.
We can use our definition to generate this "menu of reasonable choices" and so you can see here there are ten or so first line choices based on our low risk of serious morbidity, tolerable side effects and pretty good efficacy. If you look at this list, I don’t know if it jumps out at you, but when I show this around the country people say, "Look, you guys have missed the boat on something very important here." Drug safety in overdose.
If you use that criteria you must eliminate the two tricyclics because these drugs are in fact deadly in overdose. We felt we should leave them on because of lot of physicians feel like they can manage this by assuring they give limited supply of medication. So if you’re going to use that medication you need to know that these are deadly in overdose.
So in each case here, with the exception of these two, we have medications where there a drug with similar profiles that’s better tolerated over here. The reason ECT is on this list of available and preferred is because some people think that it is an ethical imperative to offer the safest, fastest most effective treatment we have to any patient who’s not doing well. So you may start with one of these but if this isn’t working you should make patients aware that ECT is available.
And there are also other reasons why you might prefer one of these. You might prefer it because the patient’s sister responded to lithium so you’re going to use it. The patient might have ulcers and allergies so you might decide to use doxepin. They might have sleep apnea so you might decide you’re going to use protriptyline. There could be lots of reasons to choose.
In this case, second line treatments there is an issue. These drugs all have some significant risk that’s completely avoidable so why would you start with amoxapine when it has a risk of tardive dyskinesia even if it’s small when you have two dozen other drugs to choose from? Clomipramine I’m not so worried that it’s not approved for depression but it has the highest rate of seizures of any drug up here. The MAOIs, we have a risk of hypertensive crisis.
How do you decide what to do? We can look at these various guidelines and we come up with our own menu of reasonable choices. I’m not trying to convince you to use the one I’ve just showed you but just to come up with one of your own. If you did, you might have a list like this that you show patients and this is what we actually do at our clinic. I will show a little profile of these drugs that I regard as first line treatments.