You cannot reliably tell depression from bipolar disorder by comparing symptom lists, because both share the same low mood, sleep loss, and anhedonia. The deciding signal is whether you have ever had a distinct episode of elevated, expansive, or irritable mood with increased energy—mania or hypomania—lasting days and impairing function. If no such episode has ever occurred, a depression label is provisional; doctors call this "unipolar" only over time, not in one visit.
Depression and bipolar disorder overlap so heavily that clinicians treat the overlap as the diagnostic problem rather than a footnote. Major depressive disorder affects an estimated 5 percent of adults worldwide in any given year, per WHO figures, and bipolar spectrum disorder around 1 to 2 percent, though screening suggests higher lifetime rates. That unequal base rate matters: a generic "depression" presentation is far more likely unipolar than bipolar before any episode of elevated mood is found.
The clinical cost of getting it wrong is not cosmetic. Treating bipolar depression with a standard antidepressant alone can trigger hypomania, accelerate cycling, or produce mixed states—one reason current guidance discourages antidepressant monotherapy in bipolar disorder. The diagnosis therefore turns on history-taking quality, not on a test result.
Shared neurobiology, divergent switches
Both conditions involve disrupted mood-regulation circuits linking the prefrontal cortex, amygdala, and ventral striatum, along with dysregulated stress-hormone signaling and altered circadian timing. The difference is not a single broken molecule. It is how the system responds to perturbation.
In unipolar depression, the dominant pattern is a sustained lowering of positive-affect and reward circuitry—reduced ventral striatal response to reward, flattened dopamine signaling, and hyperactivity of the hypothalamic-pituitary-adrenal axis. The system settles into a low-energy state and tends to stay there.
In bipolar disorder, mood regulation behaves more like a bistable switch: the same circuitry flips toward elevated energy, reduced need for sleep, and reward-seeking during mania. Sleep loss is especially revealing because it is both a trigger and a symptom. Circadian disruption can precede episodes, and mood stabilizers such as lithium act partly by stabilizing circadian and reward signaling.
One level deeper: the two conditions likely diverge in how the brain's predictive reward model updates. Depression looks like persistently low reward prediction error—nothing feels better than expected. Bipolar mania looks like runaway positive prediction. That framing explains why "just depression" and "depression that is actually bipolar" can look identical from the low side.
Plain-language version: the brain's mood thermostat is broken in both, but only in bipolar disorder does it also swing violently to the high end.
How often does depression turn out to be bipolar?
Base rates matter because diagnosis is a probability problem. Community screening studies suggest that among people treated for a major depressive episode, a meaningful minority—commonly described as roughly 1 in 10 to 2 in 10—later meet criteria for a bipolar spectrum diagnosis. Figures vary widely by setting and screening instrument, so treat them as a signal, not a precise conversion.
A commonly cited lag is striking: people with bipolar disorder often wait years between first mood episode and correct diagnosis, with estimates in the range of 6 to 10 years in many health systems. That delay is not random. First episodes are usually depressive, and the hypomanic episodes that define bipolar II are easy to miss because they can feel productive rather than pathological.
Age of onset nudges the odds. Depression beginning in the teens or early twenties, especially with a strong family history, carries more bipolar risk than first onset after 40. A family history of bipolar disorder or of lithium response raises the prior further.
Quantified context also includes treatment response. A depression that does not respond to two adequate antidepressant trials—the definition of treatment-resistant depression—warrants re-examination for bipolar spectrum features. Note the limit: nonresponse is not proof of bipolar disorder; it only raises the index of suspicion.
Edge cases that blur the line
Mixed features are the most consequential edge case. A person can have depressed mood plus racing thoughts, agitation, and insomnia—features that belong to both poles. Modern criteria allow "mixed features" specifiers on both depressive and bipolar diagnoses, which means a mixed episode can be coded either way and still be wrong in spirit.
Hypomania is easy to miss because it often looks like a good week. Increased productivity, less sleep without fatigue, and elevated confidence can be adaptive. The clinical threshold is a distinct change lasting at least four consecutive days for hypomania, versus at least one week for mania, with observable change and impairment or hospitalization for mania. Four days is a short window to recall accurately months later.
Subthreshold bipolarity is another trap. Some people never reach full hypomania criteria but show recurrent depressive episodes with brief, recurring energetic periods and a family history. This "bipolar spectrum" concept is clinically useful but not a discrete diagnosis in the main manuals, so labels vary by clinician.
Medical mimics deserve explicit mention. Thyroid disease, substance use, some medications, and neurological conditions can produce mood swings that imitate bipolar disorder. A bipolar diagnosis made without ruling out these contributors is premature.
Counterintuitive detail: antidepressant-induced hypomania can itself establish bipolarity. A switch into elevated mood during antidepressant treatment is a recognized indicator of underlying bipolar diathesis, not just a side effect to manage and forget.
What thin explainers and reviews get wrong
Thin articles imply that bipolar disorder means "mood swings" and depression means "sadness." Real bipolar disorder requires episodes, not rapid hourly swings; ultrarapid cycling is a recognized but narrower phenomenon. Emotional volatility alone is far more common than bipolar disorder.
Reviews often present screening tools as diagnostic. The Mood Disorder Questionnaire is a screening instrument that flags possible bipolarity; it does not confirm it. High scores generate a conversation, not a diagnosis.
Another common error: treating bipolar II as "mild." Bipolar II carries substantial suicide risk and frequent depressive burden, even though its highs are less extreme than bipolar I. Severity is not determined solely by the height of the mania.
A professional caveat: retrospective self-report is unreliable in both directions. People underreport hypomania because it felt good, and overreport it because a few energetic days feel abnormal when framed by a questionnaire. Collateral history from a partner or family member often changes the picture more than any self-test.
Finally, beware certainty sold online. No symptom checklist, including this one, can exclude bipolar disorder. Diagnosis is longitudinal and clinical.
When to seek urgent evaluation
Thoughts of suicide, self-harm, or harming others at any point. Little need for sleep for several nights with escalating energy or agitation. New or worsening psychosis: hearing or seeing things others do not, or fixed false beliefs. Rapid mood swings within days combined with agitation and impulsivity. Inability to function at work, school, or home, or dangerous spending or driving. A depressive episode that follows starting or increasing an antidepressant.
Red flags for bipolar spectrum, and what actually distinguishes the two
The single most useful screen is a careful lifetime timeline of elevated-mood episodes. Ask specifically about periods of several days with less sleep but high energy, unusual talkativeness, racing thoughts, increased goal-directed activity, and risky decisions. Then check whether those periods were noticed by others.
| Feature | Major depressive disorder | Bipolar disorder (depressive episode) |
|---|---|---|
| Lifetime elevated mood episode | Absent | Required (mania or hypomania) |
| Minimum duration of elevated episode | Not applicable | 4 days (hypomania) or 1 week (mania) |
| Typical onset age | Any, often 20s–30s | Often teens to early 20s |
| Family history of bipolar or lithium response | Less specific | Stronger signal |
| Antidepressant monotherapy | Standard option | Generally avoided; risk of switch or cycling |
| Sleep during low phase | Insomnia or hypersomnia | Same, plus sleep loss often triggers highs |
| Response to mood stabilizer | Not the primary lever | Core treatment target |
Practical actions that sharpen the answer
Start a dated mood log for at least four to eight weeks. Record sleep hours, energy, irritability, goal-directed activity, and any unusually confident or risky decisions. The sleep column is the highest-yield data.
Recruit one informant. A partner, parent, or close friend can answer whether they ever observed a distinct multi-day period of elevated energy or irritability. Their recollection often beats yours.
Bring a medication timeline to your appointment. Note every antidepressant, the dose, and whether mood worsened, switched, or cycled after starting it. That history changes diagnostic reasoning.
Get baseline medical testing before accepting a mood diagnosis. Thyroid function, basic metabolic panel, and a substance-use review are standard parts of a workup, not optional extras. Consider referral to a psychiatrist when bipolar features are plausible.
Prepare for a longitudinal diagnosis. Expect the label to be refined across visits rather than fixed at one. Ask your clinician explicitly what would change their mind—that question tends to surface the evidence they are weighing.
Timeline table for a typical diagnostic path:
| Stage | What often happens | Practical move |
|---|---|---|
| First depressive episode | Depression diagnosed; bipolar not yet visible | Start mood and sleep log |
| Months to years of treatment | Antidepressant trials, partial response | Track response and any switch |
| First suspected hypomania | Energy, sleep change, impulsivity noted | Collect collateral history |
| Screening | Mood Disorder Questionnaire or similar | Treat score as a prompt, not proof |
| Psychiatric evaluation | Lifetime episode timeline built | Bring logs and medication history |
| Medical rule-out | Thyroid, substances, other causes checked | Complete baseline labs |
| Reassessment | Diagnosis refined over time | Ask what evidence would revise it |
Limits of what any self-assessment can tell you
No article, checklist, or app can diagnose you. The difference between depression and bipolar disorder rests on a lifetime history that requires clinical interviewing and often collateral information. Self-screening tools raise the right questions but have meaningful false-positive and false-negative rates.
Base rates also limit certainty. Because unipolar depression is several times more common than bipolar disorder, most people with depressive symptoms do not have bipolar disorder—even though the cost of missing it is high. That asymmetry justifies screening, not self-diagnosis.
Changing clinical guidance is another moving target. Treatment recommendations evolve, so any specific drug strategy belongs in a conversation with a prescriber, not in a general explainer. What stays stable is the diagnostic logic: find the lifetime elevated episode, rule out mimics, and reassess over time.
Seek professional evaluation now if depression has not responded to two adequate treatment trials, if anyone has observed a distinct multi-day high-energy period, or if any red-flag bullet above applies. Those thresholds are concrete triggers for a specialist conversation, not a verdict.