Diagnosis Concepts

Nassir Ghaemi MD

The manic-depressive spectrum

Until 1980, all recurrent severe mood episodes were called “manic-depression”.  It didn’t matter if the mood episodes were manic  OR depressive, the illness was called “manic-depressive illness.”  This mean the presence of manic OR depressive episodes, not manic AND depressive episodes. Thus, one could have 20 depressive episodes, and the diagnosis would be manic-depressive illness (MDI).

In 1980, this approach was officially changed in DSM-III and the very large and broad concept of MDI was divided into two: the much smaller concept of bipolar disorder and the broad concept of “major depressive disorder” (MDD).  Bipolar disorder meant manic AND depressive episodes, which is a much smaller group than MDI (which was manic OR depressive episodes).  Those who had only depressive episodes were called “unipolar” depression, which later got relabeled MDD. 

This distinction between bipolar and unipolar mood illness was based on research in the 1960s and 1970s which claimed that these conditions could be differentiated based not just on symptoms, but on other “diagnostic validators”, which means lines of evidence that are independent of symptoms but which indicate symptom differences represent different illnesses. For instance, pneumonia with cough is  different in symptoms than pneumonia without cough; but those symptom differences are not sufficient to say they are different illnesses.  Not all symptom differences represent different illnesses.  The idea was that other diagnostic validators could tell us if different symptoms represent different illnesses.

Those other diagnostic validators were genetics and course, and as a smaller validator, treatment response. 

Genetics:  Those studies found that bipolar illness was genetically specific: in families of bipolar patients, there was bipolar illness; in families without bipolar illness, there we no bipolar patients.  If patients had unipolar depression, they did not have bipolar illness in their families. 

Course:  Depression in persons with bipolar illness began around age 19, and episodes lasted months or less. Depression in persons with unipolar depression began around age 30, and episodes lasted 6-12 months or more. 

Treatment effects:  Antidepressants can cause mania in bipolar illness, but do not do so in unipolar depression  (25% occurrence in bipolar patients, <<<1% occurrence in unipolar depression).  Antidepressants are ineffective in depression in bipolar illness, and somewhat more effective in unipolar depression.  

SO, if you have depression, but no manic episodes, but you have a genetics of bipolar illness and/or a course of bipolar illness and/or mania with antidepressants or other medications, then you are a squared circle. You shouldn’t happen. The whole idea of having unipolar depression is that there would be no genetics of bipolar illness and the course of illness would be as described above. 

The problem we have now in psychiatric diagnosis according to DSM and the standard textbooks is that the claim is made that we should diagnose solely based on symptoms:  depression with or without mania.  If there is no mania, there is no bipolar.  But we ignore the genetics and course evidence which is the entire basis for making the claim that the symptom differences represent different illnesses.  In other words, the whole reason we think presence or absence of mania represents a different illness (bipolar versus unipolar) is because presence or absence of mania correlated with a certain genetics and course.  But if the genetics and course of a person is opposite the scientific evidence about the presence or absence of mania, then we have a paradox.

There are only two solutions to this paradox.

1.        It could be that the old MDI concept is correct, and we should not obsess about presence or absence of mania, and simply say that all recurrent mood episodes represent the same illness. It has been shown that lithium is effective for prevention of mood episodes of any kind, not just bipolar illness. It has been shown that dopamine blockers (antipsychotics) are effective for acute treatment of mood episodes of any kind (depression or mania).  Thus, the claim that depression versus mania makes a big difference in diagnosis or treatment (using “antidepressants” versus “antipsychotics” or “mood stabilizers” may be a false distinction.

2.       The other solution is to think of mood illness as a spectrum, with classic unipolar depression on one extreme, and classic bipolar illness on the other extreme, with a lot of people in the middle with mixtures of unipolar and bipolar features. We sometimes call the middle area the “bipolar spectrum” or we could just say the mood spectrum or the “manic-depressive spectrum.”  Many people are in the middle of this spectrum, but DSM and our textbooks try to force people into the bipolar or MDD categories, with corresponding simplistic decisions of giving antidepressants over and over again for the MDD category (which is the broadest and thus where most people get shoved by our textbooks and DSM).  If we accept a spectrum concept, we would be more flexible about using dopamine blockers or mood stabilizers for treating depressive conditions, and indeed those agents are effective for many depressive conditions that are not part of the narrow DSM defined bipolar disorder definition.

Mood Temperaments

Another aspect of diagnosis of depression and manic-depression is the concept of mood temperaments.  Temperament reflects the biological component of personality; it is about half genetic, and appears to be basically stable from childhood into later life.  Mood temperaments can be conceived as basically mild versions of depressive or bipolar illness, with three major varieties:

Hyperthymia reflects mild constant manic symptoms.  These people are energetic, active, workaholics, life-of-the-party, highly sociable, very funny, curious and even sometimes impulsive, sometimes risk-takers.  They can be very charismatic, creative, and productive. Often they are successful professionally, wealthy, and well-known.   They tend to need less sleep than the norm (ie, fine with 6 hours of sleep nightly or less), with a high level of energy.

Dysthymia reflects mild constant depressive symptoms.  These people are relatively low in energy, slow in acting and can be indecisive, shy, less sociable, and often thoughtful and introspective.

Cyclothymia reflects mild manic and depressive symptoms, alternating with each other on the order of days or less.

DSM definitions of the latter two definitions are not as I describe them here, and hyperthymia is not in DSM at all.  DSM uses these terms as “disorders”, on a par with bipolar illness or schizophrenia.  In fact, these conditions were always conceived as mild variations of the disease of manic depression, but not themselves as diseases.  They are formes frustes of manic-depression, i.e., very mild manifestations, like borderline elevated glucose which is related to but different than the diagnosis of diabetes.   Temperaments are extremes of personality traits that happen in families of persons with manic-depressive illness, or in those persons themselves in between their mood episodes.

Mood temperaments are genetically related to recurrent  depressive illness and bipolar illness, and thus can be thought of as reflecting what happens when someone gets some genes for those conditions, but not enough genes to cause the full-blown illness. 

It is relevant that people can have full-blown bipolar illness or severe recurrent depression, and they can also have mood temperaments, in between their mood episodes. Thus, someone can have recurrent severe unipolar depressive episodes, with hyperthymic temperament in between those episodes as their “normal” baseline.  Such persons are diagnosed with “MDD” by DSM methods, because DSM ignores the concept of mood temperament.  Yet, that type of presentation clearly includes manic symptoms (though not episodes) as part of the clinical picture, and thus is part of the manic-depressive spectrum.

We have found that about 50% of persons with bipolar illness have mood temperaments, most commonly cyclothymia, but 50% do not have any mood temperaments, i.e., they are in the normal range of personality traits. 

Mood temperaments can be important for diagnosis and treatment:

For diagnosis, having mood temperaments themselves can have benefits and harms.  On the harm side, the symptoms are mild but constant. Over time, they can cause problems in life, like divorce or losing jobs or friends.  These effects sometimes are not obviously related to the temperament, because they happen slowly over time. It is like Chinese water torture, with drips of water accumulating over time; in contrast, the mood episodes of full unipolar or bipolar depressive illness are like storms that come all of a sudden, inundate an area, and then go away.  Both cases lead to suffering, one mildly but consistently, the other severely but rapidly.

Also, if mood temperaments are ignored, people may be misdiagnosed as having “MDD” because the manic symptoms they experience constantly are ignored since they are not severe episodes that come and go, but rather mild chronic aspects of hyperthymia or cyclothymia.  By ignoring these constant manic symptoms, antidepressants are overused, with less efficacy and harmful consequences.

The last comment brings us to why temperament matters for treatment. Antidepressant use is the major issue here.  A number of studies show that if someone has hyperthymia or cyclothymia as a mood temperament, he/she will be more likely to experience a manic episode with antidepressant treatment.  Also, it is my view, though not yet well proven with studies, that antidepressants are less effective in such persons.  This is often the case by means of the presence of “mixed depression”, depression mixed with manic symptoms (but not a full-blown manic episode).  This kind of “mixed depression” is more common in persons who have hyperthymic or cyclothymic temperaments.  Antidepressants seem less effective in mixed than pure depression, and antidepressants seem to worsen mixed depression, causing more agitation and even sometimes suicide.  In contrast, dopamine blockers (neuroleptics) are effective in mixed depression, as we have shown in a randomized clinical trial with ziprasidone.

Mixed depression

Just as there is a spectrum of mood diagnosis, i.e., from bipolar to unipolar, there is a spectrum of mood episodes, from manic to mixed to depressive.  The DSM approach to is define mania very narrowly, and depression very broadly. So you can have 1, 2, or 3 manic symptoms for weeks or months on end, along with depression, and yet still be diagnosed with a “major depressive episode”.  Or you could have 4, 5, or 6 manic symptoms for 1, 2, or 3 days (but not 4!) and still be diagnosed with a “major depressive episode”.  These manic symptoms don’t exist, as far as DSM is concerned, and so everyone is diagnosed with “major depression” even though they have racing thoughts, and high libido, and talk very fast, and are very angry and labile in mood (angry one second, anxious the next sad, crying the  next) and agitated.  But in reality this kind of high “psychomotor excitation” is a different kind of depression than someone who is very slowed down, talking and thinking very slowly, not angry at all, not anxious, cannot get out of bed, and has zero interests in anything. The latter presentation used to be called “melancholia”, and it contrasts with “mixed depression”, which reflects the agitated/labile description above. 

A great thinker in this field was Athanasios Koukopoulos, and his articles explain mixed depression well.  His research and that of Jules Angst, another great researcher in depressive and bipolar illness, is the basis for my description. Angst has found that about one-half of all patients with “major depressive episodes”, mostly with “MDD”, have mixed depression. This mixed depression is associated with a much higher rate of bipolar family genetics, and a very high rate of antidepressant-induced mania. Koukopoulos always emphasized that these patients get more mixed and more depressed with antidepressants, and in many cases, more suicidal.  The risk of suicide is high in these patients, and it is probably increased with antidepressants even further, in his view, with which I agree.

Dopamine blockers are effective, as are lithium and some (not all) anticonvulsants, given without antidepressants. 


Cognitive impairment (You don't have ADHD)

There is a common misconception that bipolar illness is a progressive illness that gets worse over time (“kindling”). This is a false belief.  I’ve reviewed this entire literature and summarized it in the Course chapter of the main textbook in this field, Goodwin and Jamison’s Manic-Depressive Illness.  We concluded that about one-quarter of patients may have worsening of mood episodes over time, but for the majority, the illness is stable, with the same basic frequency of episodes and severity of symptoms.  This is the natural course without treatment.  With treatment, one-third of patients completely recover without any further symptoms in response to a single mood stabilizer, and most of the remaining patients improved with multiple mood stabilizers. 

 There is a separate problem though from kindling, and that is that there appears to be cognitive impairment over time in bipolar illness. This means problems with attention and organization primarily (“executive function”), as well as short-term memory problems.  This cognitive impairment is frequently misdiagnosed as “adult ADHD”.  It is clearly incorrectly diagnosed as “ADHD” when a person has active depressive or manic symptoms, since both depression and mania have, as core DSM criteria, impaired concentration as part of their symptom profiles.  But this cognitive impairment happens even when patients are well, in between manic or depressive episodes.

The theory about why this cognitive impairment happens is that each manic or depressive episode is associated with overactivation of the adrenal glands, which leads to excessive production of steroids, which are harmful to the brain (they kill neurons: “excitotoxicity”).  Each mood episode harms some neurons, so that, over time, multiple mood episodes will lead to neuronal death, and the most sensitive part of the brain to steroid-induced excitoxicity is thought to be the hippocampus, which is the main region involved with short-term memory.  Hence hippocampal atrophy occurs over time in mood illnesses.  Thus, even when mood episodes eventually are controlled with adequate treatment, this long-term cognitive impairment, the result of past mood episodes, can persist.  It may be an important factor behind the fact that depression and bipolar illness increases the risk of eventually developing Alzheimer’s dementia 2-4 fold. 

That’s the bad news.  The good news is that lithium prevents all the above harm.  And getting mood episodes under control early in life can prevent all the above harm by preventing the occurrence of too many mood episodes.  In my experience, this kind of cognitive impairment is uncommon in people with mood illnesses in their twenties or thirties or younger, but it is common in the fourth decade and later.  If mood illness is treated before a person’s thirties, this kind of cognitive impairment can be prevented.  If not, then lithium is an important part of treatment to prevent it and to retard its progression in the fifth decade of life and later.  It is well proven in many animal and human studies that lithium is neuroprotective: it keeps neurons alive longer and even helps some neurons regrow that have previously shrunk.  Other medications, like antidepressants and antipsychotics, do not consistently show such neuroprotective effects.  Some animal and human data exist showing such benefits with those other drug classes, but other human studies do not find such benefits.  Some antipsychotics appear to have at least some harmful effects in humans leading to neuronal loss.  Amphetamines have clearly been shown in many replicated animal studies to be neurotoxic, to lead to neuronal death.  Human data are limited.  In sum, only lithium has been shown repeatedly and consistently in humans to have benefits in keeping neurons alive,  and thus could be a long-term aid for cognitive impairment.   There is even evidence that lithium prevents Alzheimer’s dementia in persons with mood illnesses, reducing the incidence of depression to equal to or less than the normal population. Even low doses of lithium may have this dementia prevention effect, thus low dose lithium may be useful in most persons with mood illnesses in their fifties or later primarily for dementia prevention purposes, even if it is not used for mood prevention benefits.

A common mistake is to take amphetamines to improve the attentional impairment and executive dysfunction that occurs as part of the cognitive impairment of depressive and bipolar illness. Although amphetamines symptomatically improve attention, they are neurotoxic themselves in animal studies, that is, they kill neurons directly and thus can worsen the hippocampal atrophy and other neuronal degeneration that is the basis of the cognitive impairment of mood illnesses. Short-term symptom benefit at the cost of probable long-term neurobiological harm is not a good trade off.

In sum, the best approach to preventing cognitive impairment in mood illness is to get appropriate treatments in the first decades of life and to prevent the occurrence of numerous mood episodes.  In middle age and later, the best approach is to use lithium, in standard or low doses, to help neuronal regrowth and prevent further neuronal loss and to prevent Alzheimer’s dementia.  Amphetamines at all stages should be avoided.    Further details are provided in the Treatment section on the sidebar. 

For children and adolescents: You still don’t have ADHD if you have depression or bipolar illness

Often a person has attentional impairment, but this is is the case in anyone who has mood symptoms of mania or depression or even anxiety.  To diagnose “ADHD” for attentional impairment while ignoring these other causes is like diagnosing “fever disorder” for a fever.  ADHD does NOT cause increased energy.  “Hyperactivity” is a misnomer; it is not due to increased energy but due to decreased attentional capacity.  One does different things because one cannot focus, not because one has a very high energy level that allows one to focus on many different things.   Especially in the setting of  bipolar genetics, such mood and impulsivity symptoms are most scientifically validly considered to be mood symptoms, not ADHD.

Most child psychiatrists and mental health clinicians will disagree with this kind of evaluation, because they diagnose symptoms, and their “disorders” are merely names for symptoms, hence ADHD, or ODD, as described below.

The scientifically legitimate medical approach to illness is to see if any diseases can cause symptoms; for instance, if there is a genetic family history of bipolar illness, then the disease of bipolar illness is highly likely as a cause of attentional symptoms.

The general issues  around childhood bipolar disorder are extremely complicated, and my comments here are not meant to presume assent of other clinicians.  However these are my views:  I think it is a scientific fact that bipolar disorder happens in childhood, at least in adolescence down to age 12.  The only question is how it presents before age 12.  The viewpoint that it NEVER happens is a belief-system, not a scientific hypothesis.  If it is the latter, it is disproven, because adult criteria for mania can be shown in children below age 12.  Some will claim that those criteria, even with classic symptoms like grandiosity or euphoria, cannot be known to happen in children, since children can tend to be “silly” or grandiose “normally”.   It is hard to prove something when all evidence is rejected as proof.   But at least one can ask another claim:  If symptoms that are “manic-like” happen in a child who has a parent or other family members with adult bipolar disorder, it is scientifically highly probable that the child also has bipolar disorder. This is because bipolar disorder has about 80% heritability, which is quite high, and similar to schizophrenia and height.  It is completely illogical, for instance, to claim that a child had “major depressive disorder” (MDD) or ADHD when there are immediate family members with bipolar disorder.  In the 1970s, the whole distinction between bipolar and unipolar depression had to do with genetics and course:  MDD was defined as depression that happens in people without family histories of bipolar disorder.  If depression happened in persons with family histories of bipolar disorder, then it was bipolar disorder.  In fact, the old manic-depressive illness concept ignored all this: recurrent depression was manic-depressive, even without any manic episodes.  If there is a family history of manic episodes, then this strengthens even further the view that a patient’s depression is biologically similar to bipolar disorder.  Similarly, the genetic research indicates that most people with bipolar disorder do not have ADHD in their families, and vice versa. Thus, it is scientifically highly uncommon to have ADHD with a family history of bipolar disorder.

In short, in children, family history is highly diagnostic: It tells you what the symptoms really are - underneath their complex childhood manifestations; it tells you, ahead of time, what will become clear in adolescence and young adulthood.  A family history of bipolar disorder means psychiatric manifestations of depression and anxiety and impaired attention are highly likely to be biologically related to bipolar disorder. If we do not want to give the diagnostic label of bipolar disorder, that’s fine, as long as we draw the practical conclusions that those symptoms will respond as they do in people with bipolar disorder, meaning, as described further below, that antidepressants and amphetamines are likely to be either ineffective or harmful. 

“Depression” is not a diagnosis because it is like fever, a set of symptoms. It is caused by diagnoses, just like fever; it is not a diagnosis itself, just like “fever disorder” is not a meaningful medical diagnosis. The current mainstream view is that there are two kinds of depression, unipolar and bipolar.  Unipolar depression (MDD) has a mean age of onset of 30 years; thus it is very unlikely to begin in a child or adolescent as in this case.  Bipolar  illness has a mean age of onset of 19 years, defined as the first onset of a manic episode, with about one-third of first manic episodes happening in adolescence, and about 10%, based on the National Comorbidity Survey, happening before age 12.  About one-third to one-half of persons with bipolar illness have depressive episodes before the first manic episode, and in some cases, 3 or more depressive episodes can occur before the first manic episode.  Thus, it is not at all uncommon for a child or adolescent to have multiple depressive episodes for 3-5 years, and then finally a first clear manic episode in late adolescence or around age 19.

There is a cultural zeitgeist among child psychiatrists against the diagnosis of bipolar disorder, and instead symptoms are given diagnostic labels, which is scientifically meaningless.  Oppositional defiant disorder (ODD)-in and intermittent explosive disorder only mean that a child has irritability and aggressive behavior; ADHD only means that a child can’t concentrate; MDD means the child has depressive episodes. None of these definitions are diagnostically meaningful.  Let me explain one by one:

  1. MDD – Depression in childhood, in the absence of manic episodes, is typically diagnosed as MDD.  But the average onset of mania is age 19.  It is common for individuals with bipolar disorder to have depression as their first mood episode, followed by mania later.  Thus, in prospective studies of children aged around 10, with about 10 year follow up into young adulthood by around age 20, researchers find that 25-50% of those children develop manic or hypomanic episodes. What this means is that about 25-50% of all children with “MDD” actually have bipolar disorder.  A good hint about who has bipolar disorder, and who has unipolar depression, is, again, family history of bipolar disorder.
  2. ADHD – About 90% of all children who meet mania criteria also meet ADHD criteria, because distractibility is one of the core criteria for mania.  To diagnose ADHD whenever there is attention impairment is like diagnosing “fever syndrome” whenever there is fever during pneumonia. It is consequence, not cause, in those cases.  Rather, to diagnose ADHD, one should rule out bipolar disorder first.
  3. ODD and intermittent explosive disorder – these aggressive and irritable symptoms can happen for a host of reasons, with mania being one cause.  By themselves, these diagnoses are merely restatements of the symptoms, like fever syndrome, rather than scientifically and biologically valid disease entities.
  4. “Manic-like” symptoms –  A recent prospective 4 year outcome study followed children who are about age 10 have manic-like symptoms  (meaning some manic symptoms but not enough in duration to meet full adult criteria of 4 days to one week of mania or hypomania or longer) (1).  When followed for about 5 years to mid-adolescence, about 1/3 of these children develop full hypomanic or manic episodes meeting adult criteria.  This progression to bipolar disorder is more (about 1/2 ) in the presence of a family history of bipolar disorder, and less (about ¼) in the absence of a family history of bipolar disorder.  Thus, with all the conflict and passion about whether such brief manic symptoms represent bipolar disorder, we can now say with solid data that the answer is: Sometimes. Again, the best predictor is: family history of bipolar disorder.

Time course of ADHD:  It is also relevant to point out that many clinicians treat ADHD with amphetamines, and then never stop the medications. They act as if the ADHD will never go away and will persist into adulthood in everyone. In fact, the literature from the 1980s and 1990s showed that ADHD in children was not diagnosable in 90% by age 20.  Even more recent studies by persons who are highly supportive of the diagnosis of adult ADHD show that 2/3 of children no longer meet criteria for ADHD by age 18.  Either way, in the majority of persons, ADHD goes away during childhood.  Since patients often are diagnosed by age 8-9, it makes sense to revisit the use of these medications at the very least by early adolescence. Given the concerns about neurobiological harm from amphetamines, especially in the developing brain (see below), it is even more important to avoid an attitude of denial about long-term risks with amphetamines. 

The standard of practice in child psychiatry today is to diagnose one of the above conditions, which are merely restatement of symptoms, and then to treat those symptoms with drugs for those symptoms:  antidepressants for depression, amphetamines for concentration, and antipsychotics for aggression.  None of this is scientifically valid. Scientific medicine consists of the process of finding the biological diseases that cause symptoms.  For instance, bipolar disorder could cause all the above symptoms.   I am not saying it always does, I am just saying it could.  Again, how do we know when it is more likely that bipolar disorder might be the underlying disease in children, where symptoms are so complex?  Family history is the best resource. 

Besides the references at the end of this section for treatment studies with amphetamines (see below), general references published by my group that provide the evidence and citations for many of my diagnostic comments above are also provided at the end of this consultation (2, 3).

“Dysphoric mood dysregulation disorder” is a diagnosis that has been added to DSM-5 so as to avoid diagnosing bipolar illness in children, and thus to avoiding treating bipolar illness. This new label is itself based on limited research evidence from one research group, without any follow-up course data into adulthood.  Thus, it is not well-based scientifically any more than studies on childhood bipolar disorder; in fact, in my judgment, data on bipolar illness being diagnosable, based on adult criteria for manic episodes, are more validly demonstrated in the scientific literature than the this new label. 

In the section on treatment, you can read about the harms and risks of amphetamines, particularly their neurotoxic effects (harm to the brain) and their worsening of bipolar illness. 




1.            Birmaher B, Axelson D, Goldstein B, Strober M, Gill MK, Hunt J, Houck P, Ha W, Iyengar S, Kim E, Yen S, Hower H, Esposito-Smythers C, Goldstein T, Ryan N, Keller M: Four-year longitudinal course of children and adolescents with bipolar spectrum disorders: the Course and Outcome of Bipolar Youth (COBY) study. Am J Psychiatry 2009; 166(7):795-804

2.            Vergne DE WE, Barroilhet S, Fradkin Y, Ghaemi SN: Adult ADHD and amphetamines: a new paradigm. Neuropsychiatry 2011; 1(6):591-598

3.            Wingo AP, Ghaemi SN: A systematic review of rates and diagnostic validity of comorbid adult attention-deficit/hyperactivity disorder and bipolar disorder. J Clin Psychiatry 2007; 68(11):1776-84