Key moments, in order
The line with mania: was there anything destructive, damaging or dangerous, something the person would really want to avoid happening again? If so, that points to a manic episode and bipolar I.
Admit up front that ruling out hypomania is a fool's errand. At the low end it is nearly indistinguishable from a really good day, and that is the fundamental problem of the diagnosis.
Hypomania is not always euphoric, even at a very subtle extent.
In a year-long study using weekly retrospective ratings in bipolar II, about half the time was euthymic, the rest depression, and only about 1 percent of the time was recognized as hypomania. So hypomania is subtle and rare in lived experience.
There is no magic screening question, and no gold question tells you the answer. Gather all the other data to adjust your probabilities, aiming for the greatest confidence you can get for the treatment decision.
Do not feel pressure to make a diagnosis at the end of a first appointment. Tell the patient what is on your differential; with experience, uncertainty goes up, and one 45-minute conversation cannot name all of someone's problems.
Beyond symptoms, weigh four non-manic markers: age of onset, family history, course of illness and response to treatment. The Bipolarity Index gives each of the five dimensions 20 points on a 100-point scale.
Age of onset means the first episode of depression, and earlier onset raises the probability.
Family history is the biggest pointer: a sprawling family history of bipolar disorder really raises the chances. A first-degree relative with a clear diagnosis counts most; vaguer stories are the realm of fuzziness.
Cyclicity is more important than polarity. Defining bipolar disorder by whether you can find a manic episode is the wrong approach to the mood spectrum: look at how bumpy the course is.
Make cycling the treatment goal: "We're focusing on fewer ups and downs, not on how you're feeling on a particular day."
Do not treat the mood du jour. Stop the cycling, then reassess: there may still be a depressive component, such as a dysthymia tied to childhood trauma, to address without restarting the cycling.
Response to treatment, as a bipolar clue, really means an adverse reaction to an antidepressant: something weird happened, an antidepressant misadventure.
How far to carry on through early activation depends on your suspicion of bipolarity before you started. If it was very low, you and the patient can tolerate some of it and see where it goes.
Show the patient your thinking and ask what they want to do, within certain parameters.
Think of a mood spectrum rather than a bipolar-or-not split. There is no dividing line between the bipolar and unipolar parts: bipolarity is like pigment in white paint shifting it toward red, so the question is not whether the patient is bipolar but how bipolar they are.
Bipolarity reflects more of five things: genes, inflammation, metabolism, reward hypersensitivity (latching onto potential rewards and overestimating good outcomes) and circadian differences.
The biggest treatment question bipolarity raises is how likely an antidepressant is to make the patient worse. The more you worry about that, the more a mood-stabilizer approach makes sense; simplified social rhythm therapy is the number one mood stabilizer, then lamotrigine and low-dose lithium.
In bipolar disorder, cognitive processing therapy, written exposure therapy or even EMDR are fine for trauma: none is likely to make bipolar worse, and they may help.
If there is still doubt about bipolar disorder, try hard to wait before committing to a medication approach, and go get collateral from some source.
Listen rather than run a checklist. Ask about phases of higher energy (anxious, productive, racing thoughts, energy under the skin), then follow whatever you hear: tell me more about that.
Collect a full range of evidence rather than relying on one story or event.
If you have a bipolar lens, it is on you to try to disprove your leap to the diagnosis. An expert is someone who can find what they are an expert in, in anyone.
Your setting shapes your hierarchy. Child psychiatrists rightly rank ADHD higher because the other conditions are much rarer before 17, and an expert can find their expertise in anyone.
Mixed depressions are the big problem because they sit right next to unipolar depression. With "major depression with mixed features," DSM-5 got as close as it could to saying that mood disorders are a spectrum.
Look for the four A's: anxiety, anger, physical agitation and attention problems. Once you accept that these can be manic symptoms inside a unipolar picture, they are part of a bipolar mixed state. The four A's work well when talking with patients too.
A mixed state often gets read as depression plus something else: plus PTSD with an agitated state, plus generalized anxiety disorder with an anxious state, plus ADHD with irritable phases, or plus borderline personality disorder with phases of varied symptoms.
Recognize that the symptom picture is a cloud, and on that basis accept that you will not be certain. Uncertainty is wisdom.
Share your uncertainty with the patient and recruit them, and their family, onto the "bloodhound team" to look for mixedness such as the four A's.
The higher the stakes and the more severe the symptoms, the more information you need to gather and the more certainty you need.
A busy clinician can send the patient home with an HCL-32 PDF to talk over with their people and pick it up at the next visit. As long as you pursue the information before you set out on a treatment course, you are doing your due diligence.
A lot of substance use disorder in the family points a little more toward bipolarity.
Ask about the offset of depressions: "Have you ever gone, boom, out of it?"
Listen for three or more depressive episodes as the frequency part of cyclicity. Postpartum onset of depression also raises the probability of bipolarity.
Our minds are meaning-making machines: they automatically find things to cling to to explain each episode.
Of the atypical features, leaden paralysis (a profound lack of energy) raises the probability that a depression is bipolar rather than unipolar a little. The overlap is too big for it to be very precise. Rejection hypersensitivity overlaps too much with borderline personality to be worth sorting out.
Think of it as a trait on a spectrum, "borderlineity," with no real cutoff. Toward the milder end it becomes a temperament, much like the temperament studies in bipolarity, and people can learn to combat rejection sensitivity, overvaluation and devaluation, and crisis-oriented thinking.
BPD is not binary, which gives you a way to talk about traits when full criteria are not met. A spectrum view lets you say this is not a mood disorder without saying their personality is broken.
Typical episode length (about six months for major depression versus two or three months for bipolar) has not been particularly helpful.
Akiskal's rule of threes (three marriages, three languages, three credit card defaults, three degrees) captures the flavor of hypomanic history.
Learn how to do the simplified version. You are already encouraging a regular bedtime and a regular rise time; make it part of your standard approach to everyone with a mood disorder.
It gets you out of choosing between antidepressants and antipsychotics. Weighed against an antipsychotic, social rhythm therapy is a no-brainer on risk and benefit, and lamotrigine can be seen in a similar light.
We are big fans of lamotrigine and low-dose lithium. In bipolar I, use lamotrigine as an adjunct to other tools: it does not have enough anti-manic prophylactic value to be monotherapy.
Build in systematic information before you sit down with anyone. PHQ-9s are common, but you also need a rapid mood screener, and you want the modified version.
Instruments such as the Rapid Mood Screener, Mood Disorder Questionnaire, Bipolar Spectrum Diagnostic Scale and Hypomania Checklist help because they pre-write the questions better. Use the modified Rapid Mood Screener; the original's sensitivity is too low.
Each moment is a point made in the episode, linked to the second it was said. The topic links open everything the podcast has said on that topic.