These three diagnoses get confused constantly, including by clinicians, and the difference matters because the treatments diverge in important ways.
Sections
These three diagnoses get confused constantly, including by clinicians, and the difference matters because the treatments diverge in important ways. The short version is that bipolar I has full manic episodes, bipolar II has hypomania plus serious depression, and cyclothymia is a chronic lower-grade version of both. The diagnostic line between them lives in the highs, not the lows, which is counterintuitive because the lows are what bring most patients in. The cost of getting this wrong is that a meaningful fraction of patients labeled “treatment-resistant depression” are actually bipolar II that nobody caught, and the SSRI they’ve been on for two years is part of why they keep crashing.
What follows is how the distinction actually shakes out in real visits, because the textbook criteria are clear and the patient histories almost never are. The good news, since the rest of this post is mostly about how the field gets this wrong, is that once you’ve got the right diagnosis, the right meds work surprisingly well, and the patients who’d been written off as untreatable often respond to the second or third regimen they actually needed to be on.
Bipolar I
Bipolar I requires at least one manic episode in the patient’s history. Mania, the real thing, lasts at least a week, often involves loss of judgment in ways that have real consequences, and frequently lands the patient in the hospital. We’re talking about not sleeping for four nights in a row and feeling fine, racing thoughts that other people can hear come out as pressured speech, grandiose plans that look reasonable to the patient and crazy to everyone else, sometimes psychotic symptoms (delusions, hallucinations, the works). Spending sprees, sexual indiscretion, sudden cross-country moves, business ventures launched in a weekend with a credit card and a vision. Mania is usually obvious to the people around the patient even when the patient is convinced they’re having the best month of their life.
The depressive episodes in bipolar I are clinically similar to major depression but they tend to be more severe and more recurrent. Treatment for bipolar I is centered on a mood stabilizer, usually lithium (an old, cheap, well-studied mineral salt that’s been the gold-standard mood stabilizer for fifty years, requires regular blood monitoring because the therapeutic window is narrow), sometimes Depakote (valproate, an anticonvulsant that also stabilizes mood), often with an atypical antipsychotic (the newer class of antipsychotic, like Abilify or Latuda or Vraylar) for acute mania and sometimes long-term. Antidepressants alone are generally a bad idea here because they can trigger mania, which is why the diagnosis matters before anyone starts an SSRI (selective serotonin reuptake inhibitor, the standard antidepressant class).
Bipolar II
Bipolar II is the diagnosis that gets missed the most because the highs are subtle. Hypomania (literally “below mania,” the milder version that doesn’t tip into psychosis or hospitalization) lasts at least four days, doesn’t usually wreck the patient’s life, often feels great, and in middle-aged guys frequently looks like a productive stretch. Sleep drops to four or five hours and the patient feels fine. Energy is up, libido is up, mood is good, they’re getting things done at work, knocking out projects they’d been putting off for months. No psychosis, social consequences are usually minor, and the patient thinks of it as a good week or a productive phase, not a symptom. Most guys would describe it as the best they’ve felt all year and then wonder why two months later they’re flat on their back.
Then the depression hits, often months later, lands like a brick, and lasts three to six months. That’s what brings them in. Most of these patients have been treated for unipolar depression for years before anyone catches the bipolar piece, because nobody asked the right questions about the highs. The honest reason this gets missed is that fifteen-minute medication-management visits don’t have time for the full mood history, the patient isn’t going to volunteer “I felt great in February,” and the prescriber is incentivized to start an SSRI and move on. The system is designed to miss this.
Treatment for bipolar II is usually Lamictal (lamotrigine, an anticonvulsant that works well for the depressive side of bipolar II without much risk of triggering hypomania, but requires a slow titration over six weeks because of a rare but serious rash), sometimes lithium, occasionally lurasidone (Latuda, an atypical antipsychotic with a decent track record for bipolar depression) for breakthrough depressive episodes. Antidepressants alone aren’t usually the answer here either, though the conversation is more nuanced than for bipolar I, and some patients do okay on an SSRI plus a mood stabilizer.
Cyclothymia
Cyclothymia is the lower-grade chronic version. The patient has had at least two years of fluctuating moods with hypomanic-ish and depressive-ish symptoms that don’t fully meet criteria for either pole, but they’re real and they’re disruptive. These patients often describe their whole adult life as moody, with weeks or months of being up followed by weeks or months of being down, without a clear baseline. Their relationships often suffer because their partner can’t predict which version of them is going to be at dinner, which is the kind of thing that erodes a marriage slowly over years rather than blowing it up in one event.
Treatment is usually a mood stabilizer at lower doses, often Lamictal, combined with therapy focused on stabilizing routines like sleep, meals, exercise, and the small daily rhythms that make the up-and-down swings less extreme. The pharmacology is less dramatic than for full bipolar disorder, but the structural intervention on routines often does the heavy lifting. Guys with cyclothymia who actually get sleep on a consistent schedule, eat at consistent times, and exercise regularly often find the mood swings dampening down to something they can live with, which is the part of the treatment nobody is selling because there’s no pharma rep walking into prescribers’ offices to push consistent sleep schedules.
Naming it instead of slapping the wrong label on
You can’t change shit you won’t name. A patient who’s been told for five years that he has treatment-resistant depression doesn’t have treatment-resistant depression, he has bipolar II that nobody asked about. The label that’s been put on him for years is the wrong one, and as long as it’s the label in his chart, the treatment is going to keep being aimed at the wrong target. The diagnostic step is most of the work here, and most of the diagnostic step is asking better questions and bringing in the spouse for the history.
The renaming is uncomfortable for patients because “bipolar” still carries cultural weight that “depression” doesn’t. Patients sometimes push back on the diagnosis because it feels heavier than what they came in expecting. The honest version of that conversation is that the diagnosis isn’t a worse problem than what they came in with, it’s just the actual problem, and naming the actual problem is what makes the actual treatment possible. The four years of failed SSRI trials wasn’t a worse outcome than getting the right diagnosis on day one, it was the same disease running unchecked for longer.
The diagnostic line between them lives in the highs, not the lows, which is counterintuitive because the lows are what brings most patients in.
What’s nice to hear about this one
The reason it’s worth chasing the right diagnosis is that the right meds for bipolar II actually work, often dramatically. Lamictal for bipolar II is one of the genuinely good drugs in psychiatry. The titration takes six weeks because of the rash risk, but once the patient is at a real dose, the depressive episodes thin out, the hypomanic swings flatten, and most patients report feeling like themselves for the first time in years. Lithium for bipolar I is even older and even more effective for the patients who tolerate it. These aren’t experimental treatments. They’re decades-old drugs with strong evidence behind them, and the patients who get the right diagnosis usually do well. The horror story of years of failed treatment usually ends, in the cases that get caught, with a clean response to the right medication and a patient who’s surprised by how stable life can actually be.
The pattern that comes up most
Say you’ve got a guy in his mid-thirties, sales or finance or some other role with measurable ups and downs, married, kids or not yet. Comes in for his fourth depressive episode in five years. Has been on three SSRIs across that span, none of which worked particularly well, and the prior prescriber labeled him treatment-resistant unipolar depression. He’d been told to try ketamine (a newer treatment-resistant depression option, an old anesthetic repurposed for severe depression at sub-anesthetic doses).
Most of the first appointment goes to finding it. Asking specifically about stretches of feeling great, sleeping less, being on a roll. Yeah, he says, he has those. Two or three a year, usually three or four days at a time. He’d thought of them as the weeks he made his numbers. The wife, when she comes in separately, says those stretches scare her a little because he gets intense and doesn’t sleep and makes decisions she doesn’t always agree with. That’s bipolar II. Stop the SSRI, start Lamictal, titrate up to 200mg over six weeks. Three years later he hasn’t had a depressive episode and the hypomanic stretches are gone too. He isn’t treatment-resistant. He had the wrong diagnosis.
Why this gets missed so often
Two reasons. First, patients don’t volunteer the hypomanic stuff because it doesn’t feel like a problem, so unless the clinician asks specifically about it, it doesn’t come up. The patient thinks of those weeks as “good weeks” and reports them under “no” when asked about mood symptoms. Second, the spouse is usually the better historian for the highs, and most psych appointments don’t include the spouse, so the data is missing. Bringing the partner in for the first appointment whenever possible is the single change that catches a lot of bipolar II that would otherwise get labeled depression. The wife knows. She’s been watching this run for years.
The other reason worth naming is that fifteen-minute medication-management visits aren’t set up to catch this. The system rewards starting a med and moving on, not slowing down for the longer mood history. A guy who’s been getting his SSRIs refilled at fifteen-minute intervals for four years has never had the appointment where somebody actually sat with him for an hour and walked through the highs and the lows year by year. That appointment is where bipolar II gets caught, and the system doesn’t reimburse for it well, which is most of why it doesn’t happen.
Full mania, treatment is stabilizer + antipsychotic
At least one true manic episode, lasting a week or more, often with psychosis and hospitalization. Lithium or valproate, often plus an atypical antipsychotic. Antidepressants alone are risky because they can trigger mania.
Hypomania + heavy depression, mostly missed
Four-day or longer hypomanic stretches, no psychosis, often look like “good productive weeks.” Big depressive episodes follow. Lamictal is the workhorse, slow titration over six weeks. Frequently labeled treatment-resistant depression until the highs get asked about.
Chronic ups and downs that don’t hit threshold
At least two years of mood swings that don’t fully meet bipolar I or II criteria but are real. Treatment is low-dose mood stabilizer plus structural work on sleep, meals, exercise, and routine. Less dramatic pharmacology, more routine work.
Bottom line
If you’ve been treated for depression for years and the SSRIs aren’t working, or if you have a pattern of crashes that come after weeks of feeling unusually great, the bipolar workup is worth doing. The treatment is different, the right meds work, and the wrong meds can make things worse. The diagnostic step is the whole game here, and most of the time it just takes asking better questions about the good weeks and bringing in the person who watched the good weeks happen.
Sources
- Yatham LN, Kennedy SH, Parikh SV, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) 2018 guidelines for the management of patients with bipolar disorder. Bipolar Disord. 2018;20(2):97-170. PMID 29536616.
- Hirschfeld RM. Differential diagnosis of bipolar disorder and major depressive disorder. J Affect Disord. 2014;169 Suppl 1:S12-6. PMID 25533909.
- Hirschfeld RMA, Vornik LA. Recognition and diagnosis of bipolar disorder. J Clin Psychiatry. 2004;65 Suppl 15:5-9. PMID 15554789.
- Keramatian K, Chithra NK, Yatham LN. The CANMAT and ISBD guidelines for the treatment of bipolar disorder: summary and a 2023 update of evidence. Focus (Am Psychiatr Publ). 2023;21(4):344-53. PMID 38695002.
