Skip to content

Diagnosis and the bipolar spectrum

TL;DR — Bipolar disorder is a longitudinal diagnosis anchored by mania or hypomania, not a blood test, scan, or depressive symptom profile. Bipolar I requires mania; bipolar II requires hypomania and major depression without mania, while cyclothymia and subthreshold presentations expose unsettled categorical boundaries (Nierenberg 2023, PMID 37815563; Vieta 2008, PMID 18199235; Van Meter 2012, PMID 22459786). Screening questionnaires can prompt a structured history but cannot establish diagnosis: a 21-study MDQ meta-analysis found sensitivity 0.62 and specificity 0.85 overall, with sensitivity only 0.37 when known bipolar cases were excluded (Wang 2015, PMID 26010478). Mixed-feature prevalence changes several-fold with the definition used, and DSM-5 and ICD-11 retain materially different approaches (Chakrabarti 2022, PMID 36579354; Grasso 2026, PMID 41448395). The central diagnostic task is to reconstruct episodic changes in energy, activity, sleep, cognition, behavior, impairment and context using the patient, collateral informants and records, while excluding substances, medicines and medical or neurological causes.

What establishes polarity

Construct Positive anchor Boundary that creates error Evidence implication
Mania Distinct elevated, expansive or irritable mood plus increased energy/activity, with marked impairment, hospitalization or psychosis Irritability, insomnia or agitation alone is nonspecific Establish an episode and change from baseline, not a symptom checklist in isolation (Nierenberg 2023, PMID 37815563)
Hypomania Same symptom family, observable change from baseline, but without marked impairment, hospitalization or psychosis Improved productivity may be ego-syntonic and poorly recalled Collateral history and behavioral probes can outperform asking only about feeling “high” (Angst 2003, PMID 12957719)
Bipolar I At least one manic episode Depression is common but not required for the categorical diagnosis Current depressive presentation cannot erase a documented manic episode (Nierenberg 2023, PMID 37815563)
Bipolar II Hypomania plus major depression, with no mania Four-day duration and impairment boundaries remain debated It is supported as a valid category, but is not simply a mild version of bipolar I (Vieta 2008, PMID 18199235)
Cyclothymic disorder Chronic fluctuating subthreshold hypomanic and depressive symptoms Overlap with temperament, personality constructs and developing bipolar disorder Heterogeneity and sparse prospective research limit prognostic precision (Van Meter 2012, PMID 22459786; Parker 2012, PMID 22553122)
Mixed presentation Clinically important opposite-polarity symptoms within an episode DSM-5 excludes some overlapping symptoms and ICD-11 retains mixed episodes Definition choice changes measured prevalence and eligibility (Verdolini 2015, PMID 26380368; Chakrabarti 2022, PMID 36579354)

Reconstructing an episode

A useful diagnostic interview separates enduring traits from an episodic departure from baseline. It dates onset and offset, establishes whether symptoms co-occurred, and asks what changed in observable functioning. Decreased need for sleep is more specific than insomnia: the relevant history is sleeping less without fatigue while activity increases. Spending, sexuality, driving, conflict, work output, social contact and goal-directed projects can provide behavioral anchors when mood adjectives fail (Angst 2003, PMID 12957719; Benazzi 2007, PMID 17696573).

Domain High-yield reconstruction Common false inference
Time course Distinct start/stop, duration, recurrence and recovery Treating chronic emotional lability as an episode
Energy/activity Observable acceleration, multiple projects, increased goal pursuit or agitation Equating subjective anxiety with hypomania
Sleep Reduced need without next-day fatigue Counting insomnia with exhaustion as decreased need
Speech/thought Pressured speech, increased output, racing or rapidly shifting thought Counting ordinary rumination alone
Judgment Risk-taking, disinhibition, grandiosity or consequential decisions Inferring mania from any impulsive act
Function Change noticed by others; impairment, hospitalization or psychosis separates mania from hypomania Calling a productive period pathological without syndromal context
Context Substance exposure, prescribed medicines, sleep loss, postpartum state, medical illness Assigning primary bipolar disorder before excluding an induced syndrome
Evidence source Patient account plus collateral history and contemporaneous records Treating retrospective self-report as complete

The depressive episode often brings a patient to care, whereas hypomania may not be experienced as illness. Bipolar II therefore remains vulnerable to under-recognition, yet broadening the spectrum too far risks relabeling recurrent unipolar depression or personality-related affective instability (Vieta 2008, PMID 18199235; Miller 2016, PMID 26830885).

The hypomania-duration controversy

DSM editions have used a four-consecutive-day threshold. Reviews find that shorter hypomanic episodes are common and may identify a phenotype intermediate between unipolar depression and DSM-defined bipolar II, but they do not establish a universally valid replacement threshold (Miller 2016, PMID 26830885). The dispute is not semantic: reducing duration increases sensitivity, changes prevalence, and may alter treatment exposure; retaining four days may sacrifice clinically important cases. Prospective studies tied to course and treatment response are more informative than cross-sectional symptom similarity.

Mixed features: definition drives the number

The DSM-5 mixed-features specifier sought to recognize opposite-polarity symptoms without retaining the older mixed-episode construct, but critics argue that excluding irritability, distractibility and psychomotor agitation reduces sensitivity (Koukopoulos 2014, PMID 23600771; Verdolini 2015, PMID 26380368). ICD-11 retained a mixed-episode category and differs in its manic, hypomanic and depressive episode requirements; whether this produces the optimal sensitivity–specificity balance is unresolved (Chakrabarti 2022, PMID 36579354).

Synthesis MDD episode Bipolar depression Bipolar mania/hypomania Interpretation
DSM-5-era meta-analysis, 17 studies 11.6% (95% CI 7.9–16.7) Included within MDE estimates 26.8% (17.0–39.5) Moderate-to-high risk of bias and substantial heterogeneity (Na 2021, PMID 33418368)
Updated review, 26 studies 18.1% (17.5–18.7) 32.8% (31.6–34.0) 34.8% (32.8–36.9) Opposite-polarity features were about twice as frequent in bipolar as MDD episodes (Grasso 2026, PMID 41448395)
Updated review, DSM-5 versus ≥3 features 3.6% vs 23.8% 29.3% vs 35.0% 31.2% vs 36.9% Operational criteria, not only biology, produce the prevalence spread (Grasso 2026, PMID 41448395)

Screening tools: case finding, not diagnosis

The MDQ was validated in 198 mood-clinic patients; a score of at least seven yielded sensitivity 0.73 and specificity 0.90 against a blinded structured interview (Hirschfeld 2000, PMID 11058490). Performance changed across settings and bipolar subtypes.

Study / setting Sensitivity Specificity Diagnostic lesson
Initial psychiatric outpatient validation, n=198 0.73 0.90 Spectrum was already enriched; predictive values do not transfer unchanged to primary care (Hirschfeld 2000, PMID 11058490)
Clinical replication, 37 bipolar + 36 unipolar 0.58 overall; 0.69 bipolar I; 0.30 bipolar II/NOS 0.67 Lower insight and rating mania as non-severe contributed to false negatives (Miller 2004, PMID 15306144)
MDQ meta-analysis, 21 studies 0.62 0.85 When studies excluded known bipolar disorder, sensitivity fell to 0.37 while specificity was 0.88 (Wang 2015, PMID 26010478)
Argentina, mood clinics 0.70 bipolar I; 0.52 bipolar II; 0.31 NOS 0.97 BSDS was more sensitive for bipolar II/NOS but less specific at 0.81 (Zaratiegui 2011, PMID 21440943)
Postpartum specialty sample, n=125 75.4% traditional; 87.7% modified 86.8% traditional; 85.3% modified Removing supplementary impairment/clustering questions improved sensitivity (Sharma 2011, PMID 21185082)
Pregnancy/postpartum referral sample, n=150 39% traditional; 89% modified 91% traditional; 84% modified Eighteen participants (12%) had bipolar disorder; algorithm choice changed detection sharply (Frey 2012, PMID 23146292)
Korean mood-disorder sample, n=452 Optimal cutoff 5 — Current depression, euthymia or mania/hypomania did not significantly change ROC area (Wang 2020, PMID 31958902)

A meta-analysis of 103 studies and 127 effect sizes across 14 scales found major heterogeneity within and between samples, although four instruments performed similarly well after design adjustments (Youngstrom 2018, PMID 29389179). A positive screen should therefore trigger a structured diagnostic assessment; a negative result cannot confidently exclude bipolar II or subthreshold illness.

Differential diagnosis and comorbidity

Overlap is not exclusion. ADHD, substance-use disorders, anxiety, obsessive-compulsive symptoms, autism, personality disorders and bipolar disorder can coexist, but an episodic change from baseline remains the discriminator for a bipolar mood episode.

Alternative or co-occurring condition Overlap Discriminating emphasis Evidence boundary
Major depressive disorder Identical major-depression syndrome may occur Lifetime mania/hypomania, family history, longitudinal activation No depressive symptom pattern is a standalone polarity test (Nierenberg 2023, PMID 37815563)
ADHD Distractibility, talkativeness, activity, impulsivity ADHD is trait-like and developmentally persistent; bipolar activation is episodic Pediatric presentations require parent/child interviews and systematic exclusion (Lazaratou 2012, PMID 23399752; Kowatch 2016, PMID 27570927)
Borderline personality disorder Affective instability, impulsivity, self-harm, anger Relationship-triggered rapid shifts and pervasive interpersonal pattern versus sustained mood episodes Longitudinal records and collateral history matter more than one crisis assessment
Substance/medication-induced disorder Activation, insomnia, psychosis or depression Temporal relation to intoxication, withdrawal or treatment and persistence after exposure Exposure does not automatically exclude an underlying bipolar diathesis
OCD Intrusive thoughts and repetitive behavior may fluctuate with mood Determine whether obsessive-compulsive symptoms are mood-state dependent A systematic review found OCD symptoms often appeared secondary to bipolar course (Amerio 2014, PMID 24506190)
Autism spectrum disorder Sleep, activity, irritability and communication changes Establish change from the person’s neurodevelopmental baseline Review estimates place bipolar comorbidity at 5–8%, but standard tools are poorly validated in ASD (Dunalska 2021, PMID 35472236)
Behavioral-variant frontotemporal dementia Disinhibition, apathy, altered judgment, compulsivity Progressive cognitive/behavioral change, neurological assessment and imaging About 50% of bvFTD cases had a prior psychiatric diagnosis and average delay reached 5–6 years (Ducharme 2020, PMID 32129844)
Inborn errors of metabolism Depression, psychosis, mania-like and cognitive syndromes Atypical age/course, neurological or systemic signs, family history A systematic review included 88 studies and proposed multidisciplinary screening pathways (van de Burgt 2023, PMID 36436739)

Age and reproductive context

In children and adolescents, chronic irritability and overlapping ADHD symptoms should not be converted into mania without a distinct episode; systematic family history, developmentally informed interviews and exclusion of other disorders are central (Kowatch 2016, PMID 27570927). In the perinatal period, depression may be the presenting state and questionnaire scoring behaves differently, so screening must be followed by diagnostic interview rather than automatic treatment assignment (Sharma 2011, PMID 21185082; Frey 2012, PMID 23146292).

What diagnosis does not currently provide

  • It does not identify a specific molecular mechanism in an individual.
  • It does not predict which mood pole will dominate future illness.
  • It does not select lithium, an anticonvulsant, an antipsychotic or psychotherapy.
  • It does not make a positive screen equivalent to a disorder.
  • It does not eliminate diagnostic revision as longitudinal evidence accumulates.

Minimum research-grade documentation

Element Why record it
Diagnostic system and version DSM and ICD mixed-state and episode definitions are not interchangeable (Chakrabarti 2022, PMID 36579354)
Interview type and informants Structured interview, routine clinician diagnosis and self-report screen have different error structures (Youngstrom 2018, PMID 29389179)
Episode dates and duration Duration determines whether activation crosses the hypomania threshold (Miller 2016, PMID 26830885)
Impairment, hospitalization and psychosis These observations separate mania from hypomania more reliably than symptom names alone (Benazzi 2007, PMID 17696573)
Current mood state and treatment exposure State and medication can alter recall and observed behavior even when MDQ ROC performance is similar (Wang 2020, PMID 31958902)
Exclusion assessment Substance, medication, medical and neurological causes require explicit evaluation rather than silent exclusion (van de Burgt 2023, PMID 36436739)

Age changes the diagnostic problem

The pediatric literature contains a real boundary dispute. A meta-analysis of 28 studies and 115 effect sizes found strong aggregate separation by symptom-index tests (g=1.30, 95% CI 0.98–1.62), but accuracy depended on comparator, informant and scale content; caregiver-reported mania content performed better, and clinically ill comparison groups were more informative than healthy controls (Alcaíno 2024, PMID 39103165). This is evidence for structured assessment support, not a stand-alone screening diagnosis.

The broad pre-pubertal phenotype—chronic irritability and explosive temper, often alongside ADHD—has not been supported by prospective high-familial-risk studies as equivalent to classical episodic bipolar disorder. Those cohorts instead locate the peak onset in adolescence/early adulthood, often after sleep and internalizing symptoms and adolescent depression (Duffy 2020, PMID 32307651). Against an overly restrictive reading, a separate validity review concluded that considerable evidence supports bipolar disorder in children and adolescents, while locating the remaining disputes in elated versus irritable mood and where to set spectrum boundaries (Youngstrom 2008, PMID 18199237). The controversy is thus about phenotype and threshold, not whether youth can ever have bipolar disorder.

Psychosis does not by itself settle the differential. Across pediatric studies, reported psychotic-feature prevalence ranged from 16% to 87.5%, largely reflecting age, sampling and interview differences; mood-congruent grandiose delusions were common, and temporal confinement of psychosis to affective episodes favored bipolar disorder over schizophrenia (Pavuluri 2004, PMID 15094254). At the other end of life, first mania in older age warrants psychiatric and somatic investigation, including brain imaging, because late-onset mania has a broader secondary differential and adult treatment evidence cannot simply be extrapolated (Dols 2020, PMID 32222302).

Open questions

  • What duration and impairment threshold for hypomania best predicts course and treatment response without materially increasing false-positive diagnosis? (Miller 2016, PMID 26830885; Vieta 2008, PMID 18199235)
  • Can a prospective diagnostic model outperform structured longitudinal assessment in people presenting with a first depressive episode? (Nierenberg 2023, PMID 37815563)
  • Which mixed-state definition has the best inter-rater reliability and predictive validity across cultures? (Chakrabarti 2022, PMID 36579354; Grasso 2026, PMID 41448395)
  • How should screening thresholds change with prevalence, language, setting and the cost of false positives versus false negatives? (Wang 2015, PMID 26010478; Youngstrom 2018, PMID 29389179)
  • Which features distinguish episodic bipolar activation from neurodevelopmental or personality-related affective instability in prospective cohorts? (Kowatch 2016, PMID 27570927; Dunalska 2021, PMID 35472236)

References

  1. Nierenberg AA, et al. Diagnosis and Treatment of Bipolar Disorder: A Review. JAMA. 2023;330:1370–1380. PMID 37815563
  2. Hirschfeld RM, et al. Development and validation of a screening instrument for bipolar spectrum disorder: the Mood Disorder Questionnaire. Am J Psychiatry. 2000;157:1873–1875. PMID 11058490
  3. Angst J, et al. Diagnostic issues in bipolar disorder. Eur Neuropsychopharmacol. 2003;13 Suppl 2:S43–S50. PMID 12957719
  4. Miller CJ, et al. Sensitivity and specificity of the Mood Disorder Questionnaire for detecting bipolar disorder. J Affect Disord. 2004;81:167–171. PMID 15306144
  5. Benazzi F, et al. Bipolar II disorder: epidemiology, diagnosis and management. CNS Drugs. 2007;21:727–740. PMID 17696573
  6. Vieta E, et al. Bipolar II disorder: arguments for and against a distinct diagnostic entity. Bipolar Disord. 2008;10:163–178. PMID 18199235
  7. Sharma V, et al. Screening for postpartum bipolar disorder: validation of the Mood Disorder Questionnaire. J Affect Disord. 2011;131:408–411. PMID 21185082
  8. Zaratiegui RM, et al. Sensitivity and specificity of the Mood Disorder Questionnaire and the Bipolar Spectrum Diagnostic Scale in Argentinean patients with mood disorders. J Affect Disord. 2011;132:445–449. PMID 21440943
  9. Van Meter AR, et al. Cyclothymic disorder: a critical review. Clin Psychol Rev. 2012;32:229–243. PMID 22459786
  10. Parker G, et al. Cyclothymia. Depress Anxiety. 2012;29:487–494. PMID 22553122
  11. Frey BN, et al. Sensitivity and specificity of the Mood Disorder Questionnaire during pregnancy and the postpartum period. J Clin Psychiatry. 2012;73:1456–1461. PMID 23146292
  12. Lazaratou H, et al. Attention-deficit hyperactivity disorder or bipolar disorder in childhood? Psychiatriki. 2012;23:304–313. PMID 23399752
  13. Koukopoulos A, et al. DSM-5 criteria for depression with mixed features: a farewell to mixed depression. Acta Psychiatr Scand. 2014;129:4–16. PMID 23600771
  14. Amerio A, et al. Diagnostic validity of comorbid bipolar disorder and obsessive-compulsive disorder: a systematic review. Acta Psychiatr Scand. 2014;129:343–358. PMID 24506190
  15. Wang HR, et al. The validity of the Mood Disorder Questionnaire for screening bipolar disorder: a meta-analysis. Depress Anxiety. 2015;32:527–538. PMID 26010478
  16. Verdolini N, et al. The state of the art of the DSM-5 “with mixed features” specifier. ScientificWorldJournal. 2015;2015:757258. PMID 26380368
  17. Miller S, et al. The prevalence and diagnostic validity of short-duration hypomanic episodes and major depressive episodes. Curr Psychiatry Rep. 2016;18:27. PMID 26830885
  18. Kowatch RA, et al. Diagnosis, phenomenology, differential diagnosis, and comorbidity of pediatric bipolar disorder. J Clin Psychiatry. 2016;77 Suppl E1:e1. PMID 27570927
  19. Youngstrom EA, et al. Improving the global identification of bipolar spectrum disorders: meta-analysis of diagnostic accuracy of checklists. Psychol Bull. 2018;144:315–342. PMID 29389179
  20. Wang HR, et al. The influence of current mood states on screening accuracy of the Mood Disorder Questionnaire. Clin Psychopharmacol Neurosci. 2020;18:25–31. PMID 31958902
  21. Ducharme S, et al. Recommendations to distinguish behavioural variant frontotemporal dementia from psychiatric disorders. Brain. 2020;143:1632–1650. PMID 32129844
  22. Na KS, et al. Prevalence of DSM-5 mixed features: a meta-analysis and systematic review. J Affect Disord. 2021;282:203–210. PMID 33418368
  23. Dunalska A, et al. Comorbidity of bipolar disorder and autism spectrum disorder. Psychiatr Pol. 2021;55:1421–1431. PMID 35472236
  24. Chakrabarti S, et al. Bipolar disorder in ICD-11: a review of the changes, their basis, and usefulness. World J Psychiatry. 2022;12:1335–1355. PMID 36579354
  25. van de Burgt N, et al. Psychiatric manifestations of inborn errors of metabolism: a systematic review. Neurosci Biobehav Rev. 2023;144:104970. PMID 36436739
  26. Grasso V, et al. Mixed features in depressive and bipolar disorders: an updated systematic review. J Affect Disord. 2026;398:120929. PMID 41448395
  27. Alcaíno C, et al. Discriminant diagnostic validity of paediatric bipolar disorder screening tests: a systematic review and meta-analysis. Early Interv Psychiatry. 2024;18:669–697. PMID 39103165
  28. Duffy A, et al. Pre-pubertal bipolar disorder: origins and current status of the controversy. Int J Bipolar Disord. 2020;8:18. PMID 32307651
  29. Dols A, Beekman A. Older age bipolar disorder. Clin Geriatr Med. 2020;36:281–296. PMID 32222302
  30. Pavuluri MN, et al. Psychotic symptoms in pediatric bipolar disorder. J Affect Disord. 2004;80:19–28. PMID 15094254
  31. Youngstrom EA, et al. Pediatric bipolar disorder: validity, phenomenology, and recommendations for diagnosis. Bipolar Disord. 2008;10:194–214. PMID 18199237