Bipolar disorder is both overdiagnosed and underdiagnosed, depending on the clinical setting and patient population. Studies estimate that overdiagnosis rates range from about 5% to as high as 67% in certain outpatient psychiatric clinics, while roughly 40% of people who actually have bipolar disorder are initially misdiagnosed with plain depression. The real problem isn’t simply “too many diagnoses” or “too few.” It’s that the wrong people are getting the label while others wait an average of nine years for a correct one.
Where Overdiagnosis Happens
Most evidence of overdiagnosis comes from outpatient psychiatric settings, where patients arrive with mood complaints and leave with a bipolar label that doesn’t hold up under closer evaluation. A critical review of the literature found that when researchers re-evaluated patients carrying a bipolar diagnosis using structured clinical interviews (the gold standard), a significant portion didn’t actually meet the criteria. The variability is wide, but the pattern is consistent: clinicians in busy practices sometimes assign the diagnosis too liberally.
One population particularly affected is people with borderline personality disorder (BPD). Nearly 40% of patients with BPD report having previously been misdiagnosed with bipolar disorder, compared to about 10% of patients without BPD. The two conditions share surface-level features like mood instability, impulsivity, and relationship difficulties. But the underlying patterns differ. Mood shifts in BPD are typically triggered by interpersonal events and last hours to days, while bipolar mood episodes follow a more sustained, cyclical pattern lasting weeks or months. When clinicians don’t probe deeply enough, the quicker label wins.
The Pediatric Surge
The most dramatic shift in bipolar diagnosis has been among children and adolescents. Between the mid-1990s and the early 2000s, office visits with a bipolar diagnosis in youth increased roughly 40-fold, jumping from 25 to over 1,000 visits per 100,000 population annually. During the same period, adult diagnoses only doubled. This spike raised serious questions about whether clinicians were applying adult diagnostic frameworks to normal childhood mood variability or to conditions like ADHD and oppositional behavior that can look similar on the surface. The trend has prompted ongoing debate about where the boundaries of bipolar disorder should sit in younger populations.
Where Underdiagnosis Is the Bigger Problem
For every patient incorrectly labeled bipolar, there are others living with the condition who don’t know it. The core issue is that bipolar disorder usually announces itself with depression, not mania. People seek help when they feel terrible, not when they feel unusually energetic or productive. So clinicians see the depressive episode, diagnose major depression, and prescribe accordingly. Some antidepressants can actually destabilize mood in bipolar patients, making the misdiagnosis actively harmful.
About 40% of people with bipolar disorder are initially misdiagnosed with unipolar depression. One multicenter study found that among 313 patients being treated for major depressive episodes, 18% actually met full diagnostic criteria for bipolar I or bipolar II disorder. When researchers applied a broader screening tool that captured subtler bipolar features, that number jumped to nearly 54%. The gap between those two figures illustrates how much depends on which questions get asked and how strictly the diagnostic lines are drawn.
The average person with bipolar disorder waits about nine years from symptom onset to correct diagnosis. That delay is longer for people with bipolar II (the form characterized by less intense “highs”), younger people, and those whose illness started with depression rather than mania. Nine years is a long time to receive the wrong treatment.
Why Screening Tools Fall Short
One factor driving both over- and underdiagnosis is the limited accuracy of common screening instruments. The Mood Disorder Questionnaire (MDQ), the most widely used bipolar screening tool, illustrates the problem clearly. In a large community sample, the MDQ caught only about 43% of people who actually had bipolar disorder (its sensitivity). It was good at ruling out people who didn’t have it, with a specificity of 96%. But here’s the striking number: its positive predictive value was just 3.3%. That means for every 100 people who screened positive, only about three actually had bipolar disorder.
This doesn’t mean the MDQ is useless in all settings. In populations where bipolar disorder is more common, like psychiatric clinics, the math shifts and positive results become more meaningful. But in primary care or community settings, a positive screen is far more likely to be a false alarm than a true case. If clinicians treat the screen as a diagnosis rather than a starting point for deeper evaluation, overdiagnosis follows naturally.
The Role of Drug Marketing
The expansion of bipolar diagnoses has not occurred in a vacuum. Researchers have pointed to pharmaceutical marketing as a contributing factor, particularly in the United States, where direct-to-consumer advertising is legal. As newer medications gained approval for bipolar disorder (especially drugs originally developed as antipsychotics), marketing campaigns broadened public and clinical awareness of the condition. Patient-directed advertising encouraged people to ask their doctors about bipolar disorder, sometimes prompting diagnoses that may not have been appropriate. This doesn’t mean the condition is fictional or that awareness is bad. It means commercial incentives can subtly shift diagnostic thresholds.
How Diagnostic Criteria Have Shifted
The transition from DSM-IV to DSM-5 made some structural changes to how bipolar disorder is classified, though the core requirements remained similar. The older system listed six subtypes of bipolar I based on the most recent episode type. The current system simplified this, requiring only that a person has met criteria for at least one manic episode. Depression and hypomania are noted as common but are no longer required for the bipolar I diagnosis. The mixed-episode category was replaced with a “mixed features” specifier that can be applied to either manic or depressive episodes.
These changes didn’t dramatically widen the diagnostic net for bipolar I. But the broader conversation about a “bipolar spectrum,” which includes conditions that fall short of full diagnostic criteria but share some features, has blurred the edges. When epidemiological studies use wider criteria, estimated prevalence climbs from the traditionally accepted 0.5% to 1.5% range to figures as high as 10%. By contrast, the World Health Organization estimates that about 0.5% of the global population, roughly 37 million people, lives with bipolar disorder as formally defined.
What This Means in Practice
The overdiagnosis debate isn’t purely academic. A false bipolar diagnosis carries real consequences. Mood stabilizers and antipsychotic medications have significant side effects, including weight gain, metabolic changes, and sedation. Being told you have a lifelong psychiatric condition shapes how you see yourself and how others see you. And if the real problem is borderline personality disorder, ADHD, or recurring depression, those conditions have their own effective treatments that get delayed or never tried.
At the same time, a missed bipolar diagnosis means years of ineffective treatment, preventable mood episodes, and the cumulative damage that untreated illness does to relationships, careers, and physical health. The answer to “is bipolar disorder overdiagnosed?” is that it’s frequently misdiagnosed in both directions, and the consequences are serious either way. The quality of the diagnostic process, not the label itself, is where things go right or wrong. Structured clinical interviews, careful history-taking that specifically asks about past manic or hypomanic symptoms, and collateral information from family members all improve accuracy. The problem is that these steps take time most clinical settings don’t easily accommodate.