Bipolar Disorder

Bipolar disorder is a chronic psychiatric condition defined by recurrent episodes of abnormally elevated, expansive, or irritable mood, called mania or hypomania, alternating with episodes of depression. Between episodes many people return to a relatively stable baseline, but for others mood symptoms persist or cycle rapidly. It is classified among the mood disorders and is distinguished from major depressive disorder mainly by the presence of at least one elevated-mood episode.

The condition matters because it typically begins in adolescence or early adulthood, tends to recur over a lifetime, and is associated with substantial functional impairment, elevated suicide risk, and reduced life expectancy. Its causes involve interacting genetic vulnerability, brain and biological rhythm factors, and environmental stressors, and it is treated mainly with mood-stabilizing medication combined with psychological and social interventions.

This entry first defines the disorder and its clinical features, then describes its subtypes, course, and epidemiology. It goes on to examine genetic, environmental, and neurobiological explanations, followed by diagnosis, pharmacological and non-pharmacological treatment, prognosis, the historical development of the concept, and the questions that remain open in research and clinical practice.

Definition and Core Clinical Features

The defining manic or hypomanic pole

Bipolar disorder is defined by the occurrence of elevated mood states, not by depression alone. Under DSM-5, a bipolar I diagnosis requires at least one lifetime manic episode, and the episode may come before or after hypomanic or depressive episodes. Hypomanic or depressive episodes are not required for a bipolar I diagnosis. Bipolar II, by contrast, requires a history of both a hypomanic episode and a major depressive episode. The absence of any elevated episode is what separates a presentation of unipolar depression from the bipolar spectrum, which is why the clinical history matters as much as the current mood.

Manic episodes

DSM-5 sets the threshold for mania in terms of both mood and energy. A manic episode requires a period of elevated, expansive or irritable mood together with persistently increased goal-directed activity or energy, lasting about a week or any length of time when hospitalization is needed. The mood disturbance must be severe enough to cause marked impairment in work or social functioning, to make hospitalization necessary, or to involve psychotic features. Three of seven symptoms are required when mood is elevated, and four when mood is irritable alone. These symptoms include inflated self-esteem or grandiosity, decreased need for sleep, pressured or more talkative speech, racing thoughts, distractibility, increased goal-directed activity or psychomotor agitation, and excessive involvement in activities with a high potential for painful consequences such as reckless spending or sexual indiscretion. A person may feel rested after only three hours of sleep and still not be sleeping abnormally by the criteria's standard, so the reduced need for sleep is judged against the person's own baseline. The symptoms must also represent a noticeable change from usual behavior. Exclusion criteria apply: an episode caused by a substance, a medication, or another medical condition does not count toward the diagnosis (the episode is not attributable to the direct physiological effects of a substance or another medical condition).

Hypomanic episodes

A hypomanic episode uses the same symptom list and the same mood-and-energy framework, but with a lower threshold. It lasts at least four consecutive days and does not produce marked impairment, hospitalization, or psychotic features. The four-day rule has been criticized as arbitrary and lacking empirical support, and some researchers argue that shorter elevated periods are common and clinically more similar to bipolar disorder than to unipolar depression (shorter-duration hypomanic episodes are highly prevalent and those experiencing them are clinically more similar to patients with bipolar disorder than to those with unipolar major depressive disorder). The DSM-5 change that made increased energy or activity a required element of the mood criterion also had measurable effects: in one analysis of patients from the STEP-BD study, only 52% of patients who met DSM-IV criteria for a manic or hypomanic episode also met the stricter DSM-5 criteria.

Depressive episodes and the risk of misdiagnosis

The depressive episodes of bipolar disorder are, by DSM criteria, the same as those of unipolar depression. Recognized symptoms include low mood most of the time, fatigue and low energy, loss of interest and pleasure, slow movements or speech, changes in weight and appetite, altered sleep, and excessive feelings of worthlessness, hopelessness, and guilt. Depressive episodes must last at least two weeks. Clinical comparisons suggest that bipolar depression shows more mood lability, more motor retardation, and greater time spent sleeping than unipolar depression. Nearly 50% of bipolar disorder cases first manifest with depressive episodes, which is a central reason for misdiagnosis. One large analysis reported that misdiagnosis of bipolar depression as major depressive disorder occurs in 60% of patients. Misdiagnosis carries a practical cost, because treatment with antidepressants can worsen the disorder, delay correct treatment, and lead to antidepressant-emergent mania and suicidality. Features that point toward bipolar disorder include early age of onset, a high frequency of depressive episodes, a greater percentage of time ill, and a relatively acute onset or offset of symptoms.

Mixed features and psychotic symptoms

Mixed features refer to episodes in which symptoms of mania or hypomania and of major depression occur together. DSM-IV treated this as a separate entity, whereas DSM-5 uses it as a specifier that can be applied to manic, hypomanic or depressive episodes. Psychotic symptoms, such as delusions or hallucinations, may accompany a manic episode, and their presence can raise its severity classification; they are also an important differential issue, because the criteria exclude psychotic disorders such as schizoaffective disorder and schizophreniform disorder as better explanations of the episodes. Psychotic features in bipolar depression are also recognized, although they are less frequent than in mania.

Subtypes and Patterns of Illness Course

Bipolar I, bipolar II and related forms

The subtypes are distinguished mainly by which mood poles a person has experienced. Bipolar I requires at least one manic episode, while depressive episodes are common but not required. Bipolar II requires at least one hypomanic episode and at least one major depressive episode, and no lifetime full mania. If a full manic episode later appears, the diagnosis changes to bipolar I. Hypomania is set apart from mania by its shorter minimum duration (at least four days, against roughly a week for mania) and by the absence of marked impairment or psychosis. Some researchers argue that the core symptoms other than psychosis are not unique to either subtype, so the boundary between them is contested (see diagnostic boundaries in the controversies section).

Cyclothymic disorder describes chronic mood instability: numerous hypomanic and depressive symptoms that never reach full episode thresholds. It requires two years in adults but only one year in children and adolescents, with symptom-free gaps no longer than two months. The ICD-11 framework differs in several respects. A single hypomanic or mixed episode no longer qualifies as a stand-alone diagnosis, and recurrent hypomanic symptoms can satisfy the cyclothymia criteria. Both systems allow episodes triggered by antidepressant treatment to count toward a bipolar diagnosis. The DSM-5 "with mixed features" specifier marks episodes in which depressive and manic or hypomanic symptoms coexist.

Rapid cycling

Rapid cycling is defined by frequency rather than speed: four or more affective episodes of depression, mania or hypomania within one year. The term is somewhat misleading, since it refers to the count of episodes across twelve months, not to mood shifts within days. Estimates of its prevalence vary with the sample; one review cites a point prevalence of 10% to 20% in clinical bipolar samples, and a multicenter Chinese cohort of 520 patients found about 9.4% currently met the criterion. Rapid cycling is associated with greater severity, longer illness duration, poorer global functioning and higher suicide risk, and no consensus treatment exists for it. Sources disagree on whether it is more common in bipolar II; some clinical summaries link it mainly to that subtype, while others report it in both.

Seasonal, peripartum and mixed patterns

Seasonality has been proposed mainly for manic episodes, possibly through changes in light exposure, temperature and circadian rhythms. A retrospective analysis of four years of admissions to a single Portuguese inpatient unit found that 58% of manic episodes occurred in spring and summer, with a secondary peak in autumn. Its single-site, retrospective design limits generalization. The peripartum period is the most clearly defined pattern. The DSM-5 onset specifier applies when a mood episode begins during pregnancy or within about four weeks after delivery, although advocacy groups argue for a longer window of up to a year. Recurrence rates in pregnancy are reported across a very wide range, and pregnancy has been suggested to be relatively protective, whereas the postpartum weeks carry high risk. One meta-analysis reported a postpartum recurrence rate of 35%. Perinatal episodes are most often depressive, and acute bipolar and unipolar depression are clinically hard to tell apart, so prior manic or hypomanic history is the key distinguishing feature.

Episode frequency and the course of illness

Course varies widely between individuals. Some people have only a few mood episodes in a lifetime, while others have them far more often. Patient-oriented sources commonly describe about two cycles a year as typical, but this is an approximate, not a research estimate. Across long-term follow-up, depressive episodes are generally reported to occupy a larger share of illness time and to recur more often than manic or hypomanic ones, so depression tends to predominate as the illness progresses. Whether episodes become more frequent or intervals shorten with each recurrence (the so-called progression or kindling idea) remains debated.

Presentation across the life span

Most people first show symptoms in young adulthood, but onset in adolescence is common and is often missed because mood changes are normal at that stage of development. In children and adolescents, irritability is prominent in the DSM mood criterion and is often more visible than elation, and mixed or rapidly shifting presentations are frequent. The shorter duration threshold for cyclothymia in young people reflects this. Clinicians therefore need to weigh developmental context, collateral information and the course of symptoms over time when assessing younger patients.

Epidemiology and Risk Groups

Estimates of how common bipolar disorder is depend heavily on the time window and the survey method. Moderate to high quality evidence suggests the lifetime worldwide prevalence of bipolar disorder is around 1%, and the one-year prevalence is around 0.5%, with bipolar I consistently more frequent than bipolar II. The World Health Organization's own count is lower in one sense: an estimated 36 million people (or 0.4% of the global population), including approximately 32 million adults, were living with bipolar disorder in 2023. Prevalence figures from the World Mental Health Surveys, which sampled high-, middle- and low-income countries, were lower again, with a lifetime rate of 0.6% for bipolar I and 0.4% for bipolar II.

The spread between studies is wide. A pooled analysis of community surveys put lifetime prevalence at 1.06% for BP type 1 and 1.57% for BP type 2, while one cross-study review found country-level estimates ranging from 0.1% in Nigeria to 3.3% in the U.S. Differences in instruments, diagnostic criteria and sampling explain part of this range. The Global Burden of Disease team warns that unusually low national figures may reflect underdiagnosis, sparse input data, and modeling assumptions rather than a true absence of illness, which means low numbers from some regions should be read as uncertain rather than reassuring.

Age of onset is also reported inconsistently. One review gives a median near 33 years, and describes a trimodal distribution, with 45% of people with bipolar disorder showing an early-onset age (~17 years), 35% showing a mid-onset age (~26 years), and 20% showing a late-onset age (~42 years). Other large studies place the average nearer 25, and a meta-analysis found an average around 22.8 years. Proportions beginning in youth are substantial. In a representative US sample of 1,044 people with a history of the disorder, childhood-onset BD was reported by 19.83% of participants, adolescence-onset by 21.65%, and adult-onset by 58.52%, meaning roughly four in ten began before adulthood. Early onset matters clinically because early-onset BD has been associated with a worse prognosis than late-onset BD, including more psychotic features, drug and alcohol abuse disorders, comorbidity with panic and obsessive-compulsive disorders, rapid cycling, lower lithium-response, and more suicide attempts. Since the illness is frequently missed for years, the gap between onset and recognition is itself a major epidemiological issue, taken up in the section on diagnosis.

Sex patterns are among the most debated findings. The WHO notes that the prevalence of bipolar disorder among men and women is approximately equal, while available data indicate that women are more often diagnosed. A large American claims-based study reported that women are more frequently diagnosed with BD, although other studies found similar rates across sexes. Subtype seems to matter: Bipolar II disorder, characterised by hypomanic episodes, appears to be more common in women, while bipolar I disorder shows no significant sex disparity. Researchers disagree on why. One possibility is that men are underdiagnosed, since men are less likely to seek help for issues related to mental health, and bipolar II's milder hypomanic episodes may go unrecognised. Clinical profiles also differ. In a sample of bipolar I inpatients, bipolar illness in women was characterised by the predominance of depression, as indicated by a depressive polarity at onset, higher rates of mixed mania, more suicidal behaviour, while the manic component was found to predominate in men. Whether men have earlier onset remains contested, since some studies report no significant sex differences, while others suggest that men experience an earlier onset of symptoms.

Comorbidity follows its own sex pattern. Comorbidity, particularly thyroid disease, migraine, obesity, and anxiety disorders occur more frequently in women than men, whereas substance use disorders are more common in men. Bipolar disorder is rarely isolated; patients frequently present with multiple psychiatric and physical comorbidities, complicating clinical management. Early-onset cases carry a heavier load, with higher rates of comorbidity (including childhood psychopathology). Attention deficit hyperactivity disorder is also commonly recorded alongside the illness, particularly in young people, and overlapping features such as restlessness and distractibility can lead to a child being misdiagnosed with ADHD, or with depression, before the bipolar pattern emerges.

Access to care differs as well. Patterns of treatment vary by sex: in one national cohort, women were more likely to receive only out-patient treatment and have comorbid anxiety whereas more men had substance use disorder, were convicted for crimes when unwell, received compulsory treatment orders and received in-patient treatment. Service contact rates were also higher in women in that study. Across health systems, the same factors that make prevalence hard to estimate, including screening practices, clinician familiarity and the availability of specialist services, shape who is identified and treated, so epidemiological figures partly reflect the systems that generate them.

Genetic and Environmental Causes

Bipolar disorder has no single cause. The best-supported picture is a substantial inherited vulnerability that is shaped, and often triggered, by environmental experiences. Much of the evidence comes from genetic and epidemiological designs, which show how strongly risk runs in families but cannot by themselves identify which exposures cause episodes.

Family, twin and adoption evidence

Family studies have found that first-degree relatives of affected people carry an excess risk of the disorder, and twin studies suggest that genes account for much of this familial clustering. Twin studies (and to a lesser extent, adoption studies) suggest that genes are largely responsible for this familial aggregation. Monozygotic twins are more concordant than dizygotic twins, although the figures vary between studies. One review gives monozygotic concordance of 45% and 75% and a heritability of between 60% and 80%, while other reviews report higher figures of 80–90%. Adoption studies point the same way: biological relatives of bipolar patients are substantially more likely to have the disorder than are adoptive relatives.

These estimates need careful reading. Heritability is a population-level statistic describing how much variation in liability tracks genetic differences within a particular population and environment; it says nothing about how much a given individual's risk is genetic. The concordance of monozygotic twins is well below 100 percent, which is itself evidence that environment and chance matter. Based on this imperfect concordance rate, it is clear that gene x environment interactions play a key role in the risk and onset of bipolar disorder.

Polygenic architecture and shared risk

Genome-wide association studies (GWAS) have shifted the focus from families to the DNA itself, but they have so far explained only part of the inherited risk. Only a modest portion of the genetic variance of either disorder has been explained at the molecular level. The genetic signal is spread across very many common variants, each contributing a small effect. Early large studies identified a few loci that reached genome-wide significance, including CACNA1C, and later work has built polygenic risk score (PRS) analysis using genome-wide association data from large case–control samples. Polygenic risk scores have shown that schizophrenia-associated variants are over-represented in bipolar disorder: after re-imputation and correction for sample overlap, 22 of 107 investigated SCZ SNPs showed nominal association with BD. The number of shared SCZ-BD SNPs was significantly higher than expected. Shared risk also extends to major depression, with 236 distinct genetic loci shared between BD subtypes and both SCZ and MDD identified in one analysis of bipolar subtypes.

Overlap in risk does not mean the disorders are the same illness. Diagnostic and molecular distinctions still remain despite the shared genetic architecture. Polygenic scores have also been used to ask clinical questions: in one study, the SZ-PRS was significantly higher for BD-I cases with manic psychosis than BD-I cases with depressive psychosis.

Childhood adversity and traumatic experience

Early adversity is one of the most consistently reported environmental risk factors. In a large US survey, childhood physical abuse and sexual maltreatment were associated with significantly higher risks of both first onset mania, with an odds ratio of 2.23 for abuse. A review of 19 studies found that childhood abuse (particularly physical abuse) was related to earlier onset, rapid cycling, psychosis, suicidality, impulsivity, aggression, symptom severity, and more mood episodes. Other work links at least one form of CT (emotional abuse, sexual abuse, and emotional neglect) was strongly linked to a younger age at which bipolar disease first manifested. Findings are not uniform: in one study, childhood abuse was correlated with the severity of depressive, but not manic, symptoms within bipolar disorder, and the specific patterns across symptom domains remain incompletely mapped.

Stress, life events and triggers

Adult stressors have also been linked to episode onset. In a five-year follow-up of offspring of parents with bipolar disorder, each negative life event increased risk of onset by approximately 10%. In a general adult sample, past-year stressors in the domains of interpersonal instability and financial hardship were associated with a significantly higher risk of incident and recurrent mania. Stress also appears to operate through a chain: in a prospective bipolar I sample, trauma history predicted greater chronic stressors across time, and those chronic stressors then predicted greater symptoms of depression but not mania.

Sleep disruption and substance use are widely reported as precipitants of episodes, and clinicians routinely ask about them. Establishing their causal role is difficult, because insomnia and substance use are also early features of an emerging episode and are common in people with the illness, so the direction of effect is often unclear.

Gene–environment interaction and stress sensitization

The stress-sensitization model proposes that vulnerability is expressed through the interaction of an inherited predisposition with life experience. According to this model, the influences of adult stressors on the course of bipolar illness are magnified among individuals with a history of childhood adversities. Among the evidence, exposure to childhood adversity potentiated the effects of recent stressors on adult mania. A one-year study reported that the interaction of early adversity severity and stressful life events significantly predicted recurrence in a manner consistent with the sensitization hypothesis, and that individuals with early adversity also had a significantly younger age of bipolar onset. However, the evidence is limited and mixed. The same study found no significant interaction between stress and episode number in the prediction of bipolar recurrence, so the idea that each prior episode makes later episodes more likely to be triggered by smaller stressors has weaker support. Multiple-hit models similarly propose that different neurodevelopmental trajectories are determined by the interplay between all these variables ultimately giving rise to BD only in a relatively small portion of individuals.

Neurobiological and Biological Mechanisms

No single pathophysiological model of bipolar disorder is widely accepted. Most candidate mechanisms rest on pharmacological, imaging and post-mortem data that do not map cleanly onto the illness, and many findings are small or inconsistent across studies.

Neurotransmitter hypotheses

The dopamine hypothesis is the most extensively studied. Pharmacological and imaging work points toward excess dopaminergic signalling in mania, with raised D2/3 receptor availability and an overactive reward network Converging findings from pharmacological and imaging studies support the hypothesis that a state of hyperdopaminergia, specifically elevations in D2/3 receptor availability and a hyperactive reward processing network, underlies mania. In bipolar depression, the evidence is less coherent. Imaging has reported higher dopamine transporter levels, while other dopaminergic markers vary between studies In bipolar depression imaging studies show increased dopamine transporter levels, but changes in other aspects of dopaminergic function are inconsistent. Drug responses also complicate the model, since both dopamine agonists and antidopaminergics can improve bipolar depressive symptoms. Antimanic antipsychotics are mainly D2 receptor blockers, which is a key reason the hypothesis persists As current antimanic antipsychotics are all dopamine D2 receptor blockers, their use for mania has long supported the involvement of dopamine in mania. Studies in non-psychotic mania have found that dopamine synthesis and receptor density do not clearly differ from healthy controls Dopamine synthesis and receptor density appear to remain unchanged, at least in non-psychotic mania patients compared with HC.

Evidence for noradrenaline is more limited. In a small study of 24 patients with bipolar I disorder, plasma MHPG, a metabolite linked to brain noradrenaline, was lower across the manic, response and remission stages BDI patients had significantly reduced plasma levels of MHPG throughout manic syndrome, response and remission stages. Because antipsychotics also lower MHPG, whether these phenomena are caused by antipsychotics or by the pathophysiology of BDI is not known.

Glutamatergic findings are heterogeneous. A meta-analysis of proton magnetic resonance spectroscopy studies, which included 33 studies, found that NAA levels were significantly lower in the left white matter PFC (wmPFC) of depressive and remitted BD patients compared to controls, while Gln levels were significantly higher in the ACC of remitted BD patients compared to in HC. The authors suggest that the NAA changes may reflect altered neuroplasticity, but this remains a hypothesis. Serotonin and GABA have also been implicated, but with weaker and less replicated evidence. Monoamine-based models in general have faced criticism; in depression research, serotonin and catecholamine depletion did not consistently include depressive symptoms.

Fronto-limbic circuitry and brain imaging

Neuroimaging has most consistently implicated the fronto-limbic network. A meta-analysis of 49 fMRI studies, comprising 999 people with bipolar disorder and 1027 healthy controls, found robust activity disturbances in brain regions, mostly within the fronto-limbic network. These abnormalities are present in both euthymic and symptomatic individuals. The affected regions shift with mood state: altered activation patterns in limbic structures in the euthymic state, whereas more frontal alterations were found during manic and depressive mood states. The circuit typically includes the hippocampus, amygdala, and thalamus, along with anterior cingulate cortex, ventrolateral and ventromedial prefrontal cortices. Structural data from the ENIGMA-BD group found that, compared with healthy controls, untreated patients had smaller hippocampal and thalamic volumes, along with larger lateral ventricles. Lithium-treated patients showed a different pattern in a diffusion imaging analysis of 3033 subjects, with higher widespread frontolimbic connectivity in lithium-treated BD subjects compared to those not treated with lithium. The authors of a 2026 systematic review caution that correlations between neuroimaging findings and clinical response remain limited. Because treatment, illness duration and medication status are confounded in most cohorts, these imaging differences cannot yet be read as causes of the illness.

Intracellular signalling and mood stabilizers

The inositol depletion hypothesis was proposed to explain how lithium works. Lithium directly inhibits IMPase activity by displacing the Mg2+ cofactor(s), which consequently decreases inositol. Valproate also depletes inositol, but through a different route: VPA does not affect IMPase but indirectly inhibits MIPS. A 2023 study in yeast suggested that the mechanism whereby VPA causes inositol depletion is by reducing expression of the rate-limiting enzyme MIPS. The hypothesis links these effects to a proposed overactivity of inositol-mediated calcium signalling in neurons. Mood stabilizers therapeutically reduce or buffer abnormally high IP3-induced Ca2+ release from the mitochondria of hyperexcitatory neurons, as proposed by Berridge and colleagues.

Another proposed target is glycogen synthase kinase-3 (GSK-3). Lithium and VPA both inhibit GSK3β, which is required for optimal MIPS activity in yeast. However, experiments on sensory neuron growth cones have challenged the single-target view: These effects do not depend on GSK3 or HDAC inhibition. The same study found that inositol reverses these drug effects, which supports inositol depletion as one contributing mechanism, while the overall picture remains unresolved.

Circadian rhythms

Disturbed circadian regulation is often linked to episode onset. Sleep loss and abrupt changes in sleep timing are commonly reported triggers of mania, and light exposure can shift the body clock. Lithium has been reported to lengthen the period of circadian rhythms in experimental systems, which has prompted interest in clock genes as a shared pathway between drug action and mood regulation. The direction of causality, and whether clock abnormalities are a trait marker or a consequence of episodes, are still under investigation.

Immune and inflammatory processes

Some studies have reported raised inflammatory markers, such as certain cytokines and C-reactive protein, in people with bipolar disorder, particularly during episodes. These findings vary considerably between studies and are confounded by smoking, obesity, medication and substance use. Trials of anti-inflammatory agents as adjunctive treatments have been small, so whether inflammation contributes to the disease or simply accompanies it is not yet established.

Diagnosis and Differential Assessment

Bipolar disorder is diagnosed clinically, not by a laboratory test. The diagnosis rests on the patient's history of manic, hypomanic, and depressive episodes and on the longitudinal course of illness. The central task is therefore to establish whether a manic or hypomanic episode has ever occurred, which often means reconstructing past mood states over years.

Clinical interview and collateral history

A full assessment covers current symptoms, the pattern of mood across time, functional impairment, and psychosocial circumstances. Clinicians should also obtain information on prior mood episodes, psychiatric and medical comorbidities, medication history and adherence, and substance use. Features that point toward the diagnosis include episodic mood change, reduced need for sleep, increased goal-directed activity, and impulsive or risky conduct. A history of psychotic features, rapid mood shifts, or elevation of mood after antidepressant use may also raise suspicion.

Because people in a manic or hypomanic phase often lack insight, relatives and caregivers are frequently the best source of information about those episodes. Collateral information from family members or other informants is often essential. Such input is usually gathered with the patient's consent. Physical examination is typically unremarkable, but its purpose is to detect neurological or medical illness that could either cause mood symptoms or imitate bipolar disorder. Standard initial tests include thyroid function, a complete blood count, chemistry panels, and urine toxicology for illicit drugs.

Structured instruments and mood charts

Screening tools help clinicians flag possible cases and track change. The Rapid Mood Screener is a six-item, self-report instrument that screens for mania and depression (but not hypomania). The Mood Disorder Questionnaire is a self-report alternative. In one small study comparing patients with confirmed bipolar disorder against those with unipolar major depression, when four or more questions were answered yes, the sensitivity and specificity were 88 and 80 percent, although such figures depend on the population studied. Structured interviews such as the CIDI are used to cover DSM criteria systematically. Life charts, a visual timeline of mood episodes, hospitalizations, and treatment responses across the lifespan, help reconstruct the course of illness. Patients are often asked to keep daily records of mood and sleep, which can reveal patterns that memory misses.

Differential diagnosis

The most consequential confusion is with major depressive disorder. DSM-5 criteria for the depressive phase of bipolar II disorder are identical to those required for a major depressive episode, which makes the two cross-sectionally indistinguishable. Certain features are more suggestive of one or the other. Hypersomnia, increased napping, hyperphagia, and weight gain point toward bipolarity, whereas initial insomnia and appetite or weight loss point toward unipolar depression. Lability of mood, irritability, psychomotor agitation, and racing thoughts also favour bipolarity.

Borderline personality disorder shares mood instability, but the two differ in structure. Bipolar disorder involves distinct periods of depression, mania, or hypomania, which are not part of BPD. In bipolar disorder these episodes can last for days, weeks, or even months. In ADHD, the underlying mood is typically stable, whereas in bipolar disorder it fluctuates through distinct cycles. Among adolescents, this can be mistaken as the impulsiveness or hyperactivity of ADHD. Episodes with hallucinations or delusions can be misread as schizophrenia, and when this happens, it can be misdiagnosed as schizophrenia.

Substance- and medication-induced mood states must be excluded. Alcohol and stimulants can produce symptoms that mimic mood episodes. DSM-5 criteria require clinicians to look for periods of mood and psychotic symptoms during times without substance use, and in some cases no periods without substance use can be identified beyond childhood. Medical causes, including traumatic brain injury, multiple sclerosis, stroke, and Cushing's syndrome, are considered in the same way.

Reasons for diagnostic delay

Delay is common and long. One primary-care-focused study reports that the mean delay for bipolar disorder diagnosis is 10 years. A large population-based cohort in the Netherlands found a mean of 6.46 years from first specialist contact to diagnosis, with a median of 6.85 years (the cohort was not nationally representative). In a patient survey, 69% of patients with bipolar disorder reported an initial misdiagnosis.

Several factors contribute. Many people first seek care during a depressive episode, and people who present with depression may not spontaneously mention past periods of elevated mood, especially if those periods didn't feel problematic at the time. After unrelenting episodes of depression, it is understandable that many would perceive hypomania as a return to baseline. Patients may also lack basic knowledge of what hypomania/mania is, and/or have limited insight into these symptoms.

Suicide risk, psychosis, and urgency

At diagnosis, clinicians assess suicide risk, including prior attempts and family history. Family history of suicide among first-degree relatives has been significantly linked to suicide deaths in this population. Psychotic features and mixed states, meaning symptoms of mood episodes of opposite polarity, both affect the level of urgency. Where risk is acute, the need for urgent care, hospitalization, or intensive monitoring takes priority over completing a full diagnostic history.

Pharmacological Treatment

Acute mania

Medication is the main treatment for acute mania. Recent guidelines recommend various second generation antipsychotics (SGAs), lithium, and valproate as first-line monotherapy for adults with acute mania. A network meta-analysis of double-blind trials found that the antipsychotics, carbamazepine, lithium, tamoxifen and valproate all showed efficacy in acute mania, but only aripiprazole, olanzapine, quetiapine and risperidone were better tolerated than placebo. Lithium and valproate are often compared directly. A 12-week open-label trial randomised 268 adults with manic or mixed episodes to valproate or lithium, with lithium dosed at 600 to 900 mg/day and adjusted to a plasma target of 0.8 to 1.2 mmol/L. Systematic reviews of such comparisons generally find similar efficacy.

Valproate's antimanic evidence is strong against placebo but weaker against newer drugs. A Cochrane review found high-quality evidence that valproate outperforms placebo in adults. Its comparison with olanzapine is less clear: olanzapine was better at reducing manic symptoms in high-quality data. In young people, valproate may be less effective than risperidone. Valproate may also be preferred for acute mixed mania, a small body of early evidence suggesting these patients respond better to it than to lithium.

Choosing between lithium and valproate is partly guided by clinical profile. Lithium may suit patients with a family history of bipolar disorder, few previous episodes, no comorbidities, or elevated suicide risk. Valproate may suit patients with many earlier episodes or psychiatric comorbidity.

Maintenance and relapse prevention

Lithium remains the reference maintenance drug. It has been used for more than fifty years and reduces the severity and frequency of manic states. Many patients take it long term, and stopping it sharply raises the chance of recurrence. In a placebo-controlled trial of patients with bipolar I disorder who had recently had a manic or hypomanic episode, lithium was superior to placebo in preventing mood relapse. Lamotrigine was also superior to placebo in that trial, but only lithium significantly delayed the next manic, hypomanic or mixed episode.

Lamotrigine is licensed for maintenance treatment. Its main strength is preventing depressive relapse rather than manic relapse. A Cochrane review found that, compared with lithium, lamotrigine produced more manic recurrences at one year, while people tolerated it better over the long run. A network meta-analysis of maintenance trials reported that lithium and quetiapine, among other agents, outperformed placebo for depressive recurrence. Current CANMAT and ISBD guidance lists lithium, quetiapine, divalproex, lamotrigine and several other drugs as first-line options for bipolar I maintenance, while quetiapine, lithium and lamotrigine are first-line for bipolar II.

Bipolar depression and the risk of switching

Bipolar depression is harder to treat than mania. CANMAT and ISBD guidance lists quetiapine, lurasidone, lithium and lamotrigine as first-line monotherapies. Quetiapine's benefit comes with sedation and weight gain. Lurasidone is FDA-approved for depression in bipolar I, but not for mania or for preventing recurrence. Lurasidone is generally used together with lithium or divalproex rather than alone.

Antidepressants are cautiously used. The guidelines advise against antidepressants in bipolar I disorder. In bipolar II, they are second-line and reserved for pure depression without mixed features. The main concern is treatment-emergent switching into mania or hypomania. Lamotrigine has been described as stabilising mood without pushing patients toward mania, and analyses have not found it to raise switch risk.

Monitoring, side effects and adherence

Each drug carries distinct monitoring needs. Lithium can affect the kidneys and thyroid, and its blood level must be checked to prevent toxicity. Lamotrigine can cause serious skin reactions such as Stevens-Johnson syndrome, which is why its dose is increased gradually. Olanzapine and quetiapine are associated with weight gain and abnormal glucose metabolism. Because most drugs are used for years, adherence is central, and interruption is one of the most common triggers of relapse.

Special populations

In children and adolescents, lurasidone is the only first-line option for acute depression, while lithium and lamotrigine are second-line. Evidence for paediatric mania is thinner than for adults, as noted above.

Several general principles apply to other groups, although the sources consulted for this section did not address them in detail. In people who are pregnant or who may become pregnant, lithium and valproate carry known fetal risks, so drug choice weighs teratogenic risk against the danger of relapse, and the decision is individualised. In older adults, reduced kidney function can raise lithium levels, so lower doses and closer monitoring are typical.

Psychological, Somatic, and Supportive Treatments

Psychosocial interventions

Psychological treatments are generally used alongside mood-stabilizing medication rather than in place of it. A review of randomized trials reported a strong evidence base for psychoeducation, cognitive-behavioral therapy (CBT), family-focused therapy (FFT), interpersonal and social rhythm therapy (IPSRT), and peer-support programs. Trials consistently found that adding these therapies to medication, compared with medication alone, shortened time to remission, delayed recurrence, and improved functioning. Evidence is uneven, however. Some modalities, such as neurocognitive training, have shown little benefit, and the effects of family interventions on patients themselves remain uncertain.

Psychoeducation teaches patients and families about the course of the illness, mood monitoring, early warning signs, and medication. Trials that combined it with active practice of skills such as routine regulation and family communication reduced one-year recurrence compared with usual care. CBT focuses on identifying and restructuring distorted thinking and on coping with stress, and it has been linked to longer time to relapse. FFT is a time-limited, modular program that targets high levels of family criticism and conflict, which are associated with relapse. In a two-year comparison, FFT was associated with fewer hospitalizations and recurrences than individual treatment. IPSRT is built on the idea that disruption of daily and sleep routines can trigger episodes. It stabilizes these routines and addresses key interpersonal problems. In one randomized trial of 175 people, those who received IPSRT during the acute phase had longer periods without new episodes, and greater regularity of social rhythms was linked to lower recurrence.

Electroconvulsive therapy

Electroconvulsive therapy (ECT) involves brief general anesthesia and muscle relaxation, followed by a small electrical current applied to the scalp to induce a generalized seizure. It is a standard treatment for unipolar major depression and is also used for bipolar depression. It is less commonly used for mania and mixed states. Typical indications include failure to respond to several antidepressant trials and acute, severe illness, such as serious suicidality, that requires a rapid response. Delirious mania, marked by sudden delirium, psychosis and autonomic instability, is widely regarded as a medical emergency in which ECT should be considered. An older commentary cited a review of about 50 years of use in which roughly 80% of 589 manic patients remitted, though such figures come from heterogeneous, mostly observational studies.

ECT's evidence base for mania is thinner than for depression. Randomized trials are few, and manic episodes make up only a small share of ECT use in many settings. In the United States, the FDA's 2018 reclassification of ECT devices covered major depressive episodes and catatonia, and mania is not in the labeled indications, so its use there is off-label. ECT can, rarely, precipitate a manic or hypomanic switch, and case reports describe mania emerging after ECT was started, particularly when a mood stabilizer had been withdrawn. Clinicians therefore usually maintain or add mood-stabilizing medication during an ECT course.

Hospitalization and crisis care

Inpatient admission is used when acute mania, severe depression, psychosis, or suicidal risk cannot be safely managed in the community. Partial hospital programs offer structured daytime treatment, which can bridge the gap between discharge and full return to community life. Crisis services, including emergency psychiatric assessment and crisis helplines, are the entry point for people in acute distress. Their goal is to ensure safety and to connect patients to ongoing care, so that short-term stabilization is followed by continuity of medication and psychosocial treatment.

Co-occurring substance misuse and comorbidities

Substance misuse is common in bipolar disorder and tends to worsen the course of illness, including relapse and hospital use. Clinical practice generally favors integrated treatment, in which mood disorder and substance use are addressed together rather than one after the other. Other comorbid conditions, such as anxiety disorders and ADHD, are usually treated within the same plan, with attention to how medications for one condition may affect the others. Medication adherence is a central concern throughout, since missed doses are a common route to relapse.

Social support, employment, and peer services

The World Health Organization describes recovery-oriented psychosocial interventions as including supported employment, supported housing, peer support, and social and life-skills training, with the aim of restoring hope, autonomy, and social inclusion. Work can be protective when workplaces provide reasonable adjustments, and it has been associated with better functioning, fewer symptoms, and higher self-esteem. Assertive community treatment, which includes an employment specialist on each team, is one model that integrates vocational support into intensive community care.

Peer support, in which people with lived experience offer one another informational and emotional support, is a growing component of care. A pilot trial based on a certified peer specialist model aims to foster an expectation of recovery, self-monitoring, self-management, and a return to work and other social roles. A qualitative study of people with bipolar disorder in Hong Kong found that empathic understanding from peer workers mattered more to participants than the transfer of knowledge. Some participants, however, regarded medication as the main route to recovery and found peer support less useful, which suggests that peer services complement rather than replace clinical care.

Course, Prognosis, and Mortality

Recurrence and cumulative burden

Bipolar disorder usually runs as a recurrent illness, and whether each successive episode leaves a lasting mark on the brain and on day-to-day functioning remains an open question. Some researchers have proposed a progressive, "kindling-like" model in which untreated episodes shorten the intervals between relapses and erode outcomes. Others argue that much of the apparent decline reflects residual symptoms, medication effects, substance use, or the confounding of illness severity with treatment access, and that the evidence does not yet separate these influences cleanly. Episodes with mixed features are tied to a longer overall illness, a higher likelihood of relapse, and a heavier load of co-occurring conditions.

Functioning between episodes

Many people with the disorder work, study, and maintain relationships in the intervals between episodes, but residual impairment is common. Difficulties with sustained employment, interpersonal functioning, and attention or memory often persist, and they tend to be most pronounced in people with frequent recurrences or prolonged subsyndromal symptoms. How much of this impairment is caused by the illness itself, and how much by the side effects of treatment, is still debated.

Suicide risk

Suicide is the most consequential cause of premature death in the condition. A systematic review of 31 studies estimated the standardized mortality ratio for suicide at about 14, meaning that people with bipolar disorder die by suicide at roughly fourteen times the rate of the general population, and summary estimates for all causes of mortality showed elevated mortality compared with the general population. A later meta-analysis of 57 studies, covering more than 678,000 people, produced a similar relative risk for suicide of about twelve. Adults with the disorder show suicide rates above 200 per 100,000 per year, against roughly 11 in the general adult population.

The timing of risk is more nuanced than often assumed. Although mixed states carry a higher relative risk of suicide attempts than depressive episodes, the absolute risk is greater during depressive phases simply because patients spend far more time in them. Some clinicians regard a depressive episode with any degree of contamination by hypomanic or manic features as among the most dangerous clinical states, and the mixed states literature continues to debate whether mixed features add risk beyond the severity of depression. A recent analysis suggested that the association between mixed symptoms and suicidal ideation may primarily reflect the severity of underlying depressive symptoms rather than an independent effect. Alcohol is the substance most strongly linked with suicide attempts, with impulsivity possibly mediating the relationship between alcohol use and attempts. Seasonal patterns have also been reported, with more suicide deaths occurring in spring and summer than in fall and winter.

Physical morbidity and mortality

Premature death is not confined to suicide. Pooled life expectancy estimates place the average person with bipolar disorder at about 67 years, with a weighted average of about 12.9 years of potential life lost, a measure of the years lost before expected death, as summarized in years of potential life lost analyses. Men appear to be more affected: life expectancy was significantly lower in men than in women with the disorder in subgroup analyses. Among natural causes, the largest excess is in circulatory and respiratory disease, followed by infection and neoplasm, with circulatory deaths showing a summary ratio of about 1.7 and respiratory deaths about 2.9.

Cardiovascular disease is central to this excess. Bipolar disorder itself appears to confer risk for cardiovascular disease independent of the treatments used to manage the disorder, and the condition is linked to increased risk of medical comorbidities such as autoimmune disorders, obesity, diabetes and cardiovascular disease. The relationship appears to run in both directions and may be partly mediated by immune dysfunction. Medication contributes as well. Second-generation antipsychotics are associated with significant cardiovascular risk factors, including weight gain, dyslipidemia, and insulin resistance, and lithium can also have direct cardiovascular effects, mainly dysrhythmias. Because treatment-related metabolic harm is real but not the sole cause, researchers emphasize that systematic metabolic monitoring is crucial to mitigate treatment-related risks, alongside lifestyle measures such as diet and physical activity. Barriers to physical care, including limited health knowledge, difficulty navigating services, and poor adherence, also contribute to the gap.

Factors associated with better long-term outcomes

Outcomes are generally better when treatment begins early, when maintenance medication is continued consistently, and when care is sustained rather than episodic. Avoiding or treating co-occurring substance use, reducing exposure to sleep disruption, and building stable social and occupational routines are also associated with fewer relapses and better functioning, although many of these associations come from observational data and cannot establish cause and effect.

Historical Development of the Concept

Early descriptions

Accounts of alternating excitement and gloom are older than any modern diagnosis. Hippocratic writings already used the terms melancholia and mania, but the idea that the two were phases of one illness came later. Aretaeus of Cappadocia, writing in the first or second century AD, observed people moving between excited and depressed states and suggested that the two could represent different stages of the same disease. Though there is mention of 'melancholia' and 'mania' in the texts of Hippocrates, it was later in the works of Falret and Baillarger that BPAD was identified as a cyclical disease. Between antiquity and the nineteenth century, these descriptions remained scattered and did not form a coherent category.

Falret, Baillarger, and "circular insanity"

The modern concept emerged in Paris in 1854. Jules Baillarger and Jean-Pierre Falret presented closely similar descriptions to the Académie de Médecine within weeks of each other. Baillarger named the condition folie à double forme (dual-form insanity), while Falret called it folie circulaire (circular insanity). The two men disputed priority. Falret also noticed that the disorder ran in families, which is an early recognition of its hereditary basis. Psychiatrist Karl Kahlbaum was among the first to link symptoms and outcomes of people who had mood disorders.

Kraepelin's unification

In the 1890s, the German psychiatrist Emil Kraepelin brought recurrent mood disorders together under the heading of manic-depressive insanity. He set this group against dementia praecox, the illness now called schizophrenia, which he distinguished by its periodic course and a more favourable prognosis. Kraepelin's category was broad. It covered what is now bipolar disorder and also recurrent unipolar depression, and he documented the symptom-free intervals between episodes that remain central to the modern definition. Kraepelin's classification was later refined, but its basic split between mood and chronic psychotic illness shaped twentieth-century psychiatry.

Twentieth-century classification

The second edition of the American Diagnostic and Statistical Manual (DSM-II, 1968) placed the illness among the affective disorders and renamed it manic-depressive illness, recognising manic, depressed, and circular forms. In the mid-20th century, recurrent (unipolar) major depressions were differentiated from those with manic episodes (bipolar I disorder), encouraged in part by new treatments. A bipolar II syndrome, defined by recurrent depression with hypomania, followed in the 1970s. The decisive change came with DSM-III in 1980, which introduced the term "bipolar disorder" and assigned separate criteria to manic and depressive episodes. That split the mood disorder into two categorical diagnoses: major depressive disorder, on the one hand, and bipolar disorder, on the other. The dichotomy persisted through DSM-IV, DSM-IV-TR, DSM-5 and DSM-5-TR. The International Classification of Diseases (ICD) followed a similar route, but with differences. ICD-11 introduced an independent category of bipolar II disorder, while DSM-5 dropped mixed episodes as a separate diagnosis, a choice ICD-11 did not follow.

Lithium and the treatment turn

The introduction of lithium changed how the illness was understood and managed. In 1949, the Australian psychiatrist John Cade gave lithium carbonate to patients with mania. He had hypothesised that a toxic substance related to uric acid might produce manic excitement, and he reasoned from the use of lithium in gout. He followed this accidental discovery with his first clinical report in 1949, where he stated that lithium improved mania in mood disorders and also reduced restlessness in psychotic patients. His paper initially attracted little notice. Its impact grew after Erik Strömgren urged Mogens Schou to run a randomised trial, and Schou's 1954 results gave lithium a firm evidence base. US FDA approval for treatment of mania came in 1970, followed by approval for maintenance therapy in 1974. For several decades lithium was the leading treatment, and its use in preventing relapse made the disorder look less like a series of isolated crises and more like a chronic, recurrent illness requiring long-term management.

Terminology and public understanding

The names used for the condition have shifted with each change in classification. "Manic-depressive insanity" carried connotations of madness, and the move to "manic-depressive illness" in 1968 was seen as more favourable. The 1980 adoption of "bipolar disorder" placed the emphasis on the two poles of mood rather than on insanity, though the older term remained in use among clinicians and in popular writing for years. These changes reflect a broader shift from describing the illness as a form of madness toward framing it as a medical condition with defined criteria and treatments. Public understanding has followed these shifts unevenly, and the older labels still appear in historical texts and in everyday speech.

Diagnostic Boundaries and Controversies

The boundary of bipolar II

The edges of bipolar II disorder are the most contested part of the diagnosis. Disputes over how hypomania should be defined go back at least to the 1970s, and they center on how long an episode must last, which symptoms form its core, and how many symptoms are needed. DSM-5 requires hypomanic symptoms to persist for at least four consecutive days, whereas ICD-11 uses a looser phrase, "at least several days." A two-day threshold has been proposed as an alternative.

Jules Angst and colleagues, working with an epidemiological cohort in Zurich, argued for a broader concept. They proposed that overactivity should sit alongside euphoria and irritability as a core feature, that episode length should no longer be required if three of seven symptoms are present, and that a change in functioning should remain mandatory. Their broad definition of bipolar II gave a cumulative prevalence of 10.9%, compared with 11.4% for broadly defined major depression.

Critics point in the opposite direction. Prospective data suggest that subthreshold bipolarity is a weak predictor of later threshold illness, with a sensitivity of 28.6% and a specificity of 98.9% reported in one analysis. Commentators also note that no studies have examined the potential impact this change would have on diagnosis and outcome in real-world clinical practice. Some researchers regard bipolar II as the most common bipolar phenotype, yet its treatment remains controversial and poorly evidence-based.

Children and adolescents

Whether bipolar disorder affects prepubertal children in substantial numbers has been debated since the 1990s. The central question is whether the first hypomanic or manic episode almost always occurs after puberty. The answer matters practically: if young children need medication, withholding it carries risk, but if they do not, drug treatment exposes them to avoidable iatrogenic harm. Proponents of a childhood form have suggested that mania may look different in young people, for example as ultradian mood cycles lasting less than a day, or as chronic irritability without clear elevation of mood.

Epidemiological estimates have moved considerably. A widely cited meta-analysis of a dozen surveys suggested a global community prevalence of 1.8%, and an update that added eight further surveys raised the figure to 3.9%. Critics have argued that these heterogeneous surveys do not support a pooled estimate at all. The worry about overdiagnosis led the authors of DSM-5 to create disruptive mood dysregulation disorder, intended to separate persistently irritable children who had often been labelled bipolar. A commentary from the Hastings Center also argued that many children receiving the diagnosis do not closely match DSM criteria, and that their medication risk-benefit profiles are not well established.

Not all evidence points to overdiagnosis. One controlled study of children and adolescents found that bipolar disorder was underdiagnosed rather than overdiagnosed, while ADHD was overdiagnosed, but its authors called for larger samples to confirm the result.

Continuum versus discrete categories

The Kraepelinian split between manic-depressive illness and dementia praecox has been challenged for decades. Some proponents describe a continuum running from unipolar depression through bipolar disorder to schizophrenia. Supporting arguments include a high degree of genetic transmissibility in both disorders, and some family and twin data suggesting hereditary overlap between the two. Shared pharmacology is also cited, since many atypical antipsychotics developed for schizophrenia are also used in bipolar disorder.

Clinical overlap is a practical concern. Psychotic mood episodes can be phenotypically indistinguishable from schizophrenia, which means they may be misdiagnosed as schizophrenia. Defenders of a dichotomy point to brain MRI abnormalities, chronicity of course, mood-incongruent delusions, neurocognitive deficits, and the deficit syndrome as features that separate the two. One proposed compromise is a classification that accommodates both a dichotomy and a continuum. Imaging studies are also contested: a 2023 MRI study of 412 participants reported a gradual increase in hippocampal subfield volumes from schizophrenia to depression to bipolar disorder, a result that fits a dimensional view but is limited by its size.

Reliability across clinicians and settings

Diagnostic consistency is another open problem. The Structured Clinical Interview for DSM-IV asks about hypomania through a mood-based stem question; a negative answer moves the assessment away from bipolar disorder, and the interview does not systematically probe hypomanic symptoms during depressive episodes, which can cause mixed depressions to be missed. Clinicians have also been found to apply DSM duration thresholds flexibly, so recurrent short hypomanic episodes may still lead to a bipolar diagnosis in practice. Inconsistent application of criteria, in turn, undermines research that relies on diagnostic labels. In children, reliable diagnosis appears to require longitudinal assessment and reports from informants, and in adults, borderline personality disorder, which is linked to childhood trauma, is frequently mistaken for bipolar disorder.

The stakes of these disputes are clinical as well as conceptual. Over-diagnosis can expose people to mood stabilizers and stigma, but under-recognition also carries costs: early onset and delayed first treatment are both independent predictors of poor adult outcome. Debates about whether medication or psychological treatment should come first, particularly for young people and those with milder forms of the illness, remain unresolved.

Research Frontiers and Unresolved Questions

Bipolar disorder still has no laboratory test that clinicians can use to confirm a diagnosis, estimate an individual's course, or choose a drug. Current research is working on each of these gaps, but most results are preliminary.

Biomarker discovery

Work on biomarkers spans metabolomics, proteomics, neuroimaging and peripheral blood tests. One systematic review of 27 metabolomics studies pointed to disturbances in glucose, amino acid and fatty acid metabolism, and its authors suggested that metabolite panels could help with diagnosis. A proteomic meta-analysis of blood and other peripheral fluids identified hundreds of proteins that differed between patients and controls, mainly in complement, coagulation and lipid pathways, but the authors stressed heterogeneity between studies and the need for larger validation cohorts. Research on neuroprogression asks whether ventricular enlargement, grey-matter loss, and peripheral inflammatory or neurotrophic markers such as BDNF track illness stage over time. These associations appear in several reviews, yet small and inconsistent samples prevent firm conclusions. Magnetic resonance spectroscopy findings on glutamate-related signals have also been proposed as a diagnostic aid, but they have not entered routine practice.

Pharmacogenomics and precision psychiatry

Lithium is the clearest case of both promise and limitation. About one-third of patients show an optimal response, and no clinical feature has been found that separates likely responders from non-responders before treatment starts. Pharmacogenomic work has combined genome-wide association studies, gene-expression analyses and polygenic scores, and several loci and pathways have been reported, but no candidate gene has been validated as a predictor. In a cohort of 2,367 patients from the International Consortium on Lithium Genetics, a multi-trait polygenic score explained about 5% to 9% of variation in lithium response, a clear improvement over single-trait scores but still a modest share of the total. Models that add clinical variables to genetic scores perform better in some analyses, and reviewers conclude that clinical use remains a long way off. The authors of these studies call for larger and more diverse samples, because scores derived mainly from one population often transfer poorly to others.

Digital monitoring and wearable devices

Smartphone-based ecological momentary assessment has proven feasible in bipolar disorder. In a small pilot, participants answered prompts twice a day for two weeks, and their ratings of mood, thought speed and impulsivity varied more than those of healthy controls. Wearable devices can record sleep duration and movement continuously, and systems such as MONARCA combine phone sensors, self-report and clinician web portals to detect early warning signs. The central problem is interpretation. Decreased need for sleep is a hallmark of mania and can be measured by a wrist device, but evidence that these signals reliably indicate a particular mood state is still inconclusive. Self-monitoring also depends on insight and sustained engagement, and adherence often lapses. A 2025 systematic review found that passive digital tools had not yet produced clear improvements in patient outcomes.

Unresolved mechanistic questions

Several basic questions remain open. Researchers do not fully understand why episodes cycle, why some people cycle rapidly while others rarely switch, or what determines whether a person who is stable relapses. It is unclear whether triggers such as sleep loss act through the same pathways as spontaneous episodes. The neurotransmitter, circadian and immune hypotheses each explain part of the picture, but studies have not clearly separated the causes of episodes from their consequences. Likewise, the basis of treatment response remains unclear: the same drug can work well for one person and fail for another, and the biological reasons are not established.

Underrepresented populations and access to care

Much of the evidence on biomarkers, genetics and digital monitoring comes from relatively small clinical samples, often in high-income settings. Biomarker work in late life, for example, has identified candidate signals in genomic, blood-based and neuroanatomical studies, but these findings still require replication in broader groups. Digital tools also assume access to smartphones and stable data plans, which may limit their reach in lower-resource settings. Closing these gaps will require research in diverse populations and health systems, as well as evaluation of whether new tools actually improve access to evidence-based care.

References

1. DSM-5 Criteria and How It Differs from Bipolar IIneurodivergentinsights.com
2. DSM 5 Criteria for Mania or Manic Episodepsychologynepal.com
3. DSM-5 Criteria for Maniacarepatron.com
4. Manic Episode Criteria - Dsm-5 Criteriapediatriconcall.com
5. The Prevalence and Diagnostic Validity of Short-Duration Hypomanic Episodes and Major Depressive Episodes - PubMedpubmed.ncbi.nlm.nih.gov
6. Frontiersfrontiersin.org
7. Depression vs. bipolar: Symptoms, diagnosis, & treatmentrula.com
8. Strategies to Reduce Misdiagnosis of Bipolar Depressionpsychiatryonline.org
9. Differentiating Bipolar Depression from Unipolar Depression in the First Presentation of a Depressive Episodelink.springer.com
10. Is it depression or is it bipolar depression? : Journal of the American Association of Nurse Practitionersjournals.lww.com
11. DSM-5 criteria for a manic episode (Bipolar I disorder) — The Cardiovascularcardvasc.org
12. Bipolar DSM-5 Diagnostic Criteria for Healthcare Professionalsmedcentral.com
13. Bipolar II Disorder: DSM-5 Criteria and Key Symptomsneurodivergentinsights.com
14. Bipolar Definition and DSM-5 Diagnostic Criteriahealthcentral.com
15. Diagnosing bipolar disorders: ICD-11 and beyondncbi.nlm.nih.gov
16. Bipolar Disorder DSM-5 Criteria: Diagnosis Guidereachlink.com
17. Bordering on the bipolar: a review of criteria for ICD-11 and DSM-5 persistent mood disorderscambridge.org
18. The bipolar two-syndrome concept: questioning the shaping of a circular argument for subtyping a dimensional disorderncbi.nlm.nih.gov
19. Rapid cycling bipolar disorder: Literature review on pharmacological treatment illustrated by a case report on ketamine - PMCpmc.ncbi.nlm.nih.gov
20. Sociodemographic, clinical and treatment characteristics of current rapid-cycling bipolar disorder: a multicenter Chinese studyncbi.nlm.nih.gov
21. Rapid Cycling Bipolar Disorder: An Overview - Lincoln Prairie Behavioral Healthlincolnprairiebhc.com
22. Seasonality in bipolar disorder, analysis of an inpatient unitncbi.nlm.nih.gov
23. Striking the Balance: Bipolar Disorder in the Perinatal Periodpsychiatryonline.org
24. Treatment of Peripartum Bipolar Disorder - ScienceDirectsciencedirect.com
25. Bipolar Disorder in the Postpartum Period: Management Strategies and Future Directions - Carley J Pope, Verinder Sharma, Dwight Mazmanian, 2014journals.sagepub.com
26. Postpartum mood episodes in bipolar I vs. II disorder: a retrospective observational analysis of clinical correlateslink.springer.com
27. Bipolar disorder rapid cycling: Causes, who has it, and mood changesmedicalnewstoday.com
28. What Is Rapid Cycling Bipolar Disorder? - GoodRxgoodrx.com
29. Worldwide prevalence of bipolar disorder - NeuRA Librarylibrary.neura.edu.au
30. Bipolar disorderwho.int
31. Prevalence and Correlates of Bipolar Spectrum Disorder in the World Mental Health Survey Initiative - PMCpmc.ncbi.nlm.nih.gov
32. Brazil - Bipolar disorder prevalence: a systematic review and meta-analysis of the literature Bipolar disorder prevalence: a systematic review and meta-analysis of the literaturescielo.br
33. Comparative Analysis of the Burden of Bipolar Disorder Among Individuals Aged 15–49 Years in China and Globally, 1990–2021: An Ecological Trend Analysis Based on GBD 2021 - PMCpmc.ncbi.nlm.nih.gov
34. Age at onsetlibrary.neura.edu.au
35. Age of onset of bipolar disorder: Association with lifetime psychiatric disorders and health-related quality of life - PubMedpubmed.ncbi.nlm.nih.gov
36. Admixture Analysis of Age of Onset in Bipolar Disorder and Impact of Anxiety Comorbidityncbi.nlm.nih.gov
37. When Does Bipolar Start? Onset Age and Early Warning Signs - ScienceInsightsscienceinsights.org
38. The Role of Sex in Clinical Characteristics and Pharmacological Treatment of Bipolar Disorder - PMCpmc.ncbi.nlm.nih.gov
39. Risks associated with gender differences in bipolar I disorder - ScienceDirectsciencedirect.com
40. Gender differences in bipolar disorderpubmed.ncbi.nlm.nih.gov
41. Gender and mental health service use in bipolar disorder: national cohort studycambridge.org
42. Average Age Diagnosis Bipolar Disorder - Consensus Academic Search Engineconsensus.app
43. Sex Differences in Brain Structures and Functional Connectivity among Patients with Bipolar I Disordermedrxiv.org
44. Gender and Sex Issues in Bipolar Disorderpsychiatrictimes.com
45. Family, twin, and adoption studies of bipolar disorder - Smoller - 2003 - American Journal of Medical Genetics Part C: Seminars in Medical Genetics - Wiley Online Libraryonlinelibrary.wiley.com
46. Genetic factors in the etiology of bipolar disorder - ScienceDirectsciencedirect.com
47. Family, twin, and adoption studies of bipolar disease - PubMedpubmed.ncbi.nlm.nih.gov
48. Genome-wide association study implicates NDST3 in schizophrenia and bipolar disordernature.com
49. Association of schizophrenia polygenic risk score with manic and depressive psychosis in bipolar disordernature.com
50. Identification of shared risk loci and pathways for bipolar disorder and schizophreniancbi.nlm.nih.gov
51. Characterizing the polygenic overlaps of bipolar disorder subtypes with schizophrenia and major depressive disorder - ScienceDirectsciencedirect.com
52. Polygenic dissection of diagnosis and clinical dimensions of bipolar disorder and schizophrenianature.com
53. Contributions of the social environment to first-onset and recurrent mania - PMCpmc.ncbi.nlm.nih.gov
54. The Influence of Trauma, Life Events, and Social Relationships on Bipolar Depression - PMCpmc.ncbi.nlm.nih.gov
55. Childhood trauma and bipolar affective disorder: Is there a linkage? - PMCpmc.ncbi.nlm.nih.gov
56. The stress sensitization hypothesis: understanding the course of bipolar disorder - PubMedpubmed.ncbi.nlm.nih.gov
57. Childhood adversities and bipolar disorder: a neuroimaging focusncbi.nlm.nih.gov
58. Genetic research into bipolar disorder: the need for a research framework that integrates sophisticated molecular biology and clinically-informed phenotype characterization - PMCpmc.ncbi.nlm.nih.gov
59. Family, twin, and adoption studies of bipolar diseaseresearchgate.net
60. Patterns of childhood adverse events are associated with clinical characteristics of bipolar disorder - PMCpmc.ncbi.nlm.nih.gov
61. The dopamine hypothesis of bipolar affective disorder: the state of the art and implications for treatmentnature.com
62. The dopamine hypothesis of bipolar affective disorder: the state of the art and implications for treatmentncbi.nlm.nih.gov
63. (PDF) The dopamine hypothesis of bipolar affective disorder: the state of the art and implications for treatmentresearchgate.net
64. The Noradrenaline Metabolite MHPG Is a Candidate Biomarker from the Manic to the Remission State in Bipolar Disorder I: A Clinical Naturalistic Studyncbi.nlm.nih.gov
65. Glutamatergic and N-Acetylaspartate Metabolites in Bipolar Disorder: A Systematic Review and Meta-Analysis of Proton Magnetic Resonance Spectroscopy Studiesncbi.nlm.nih.gov
66. Editorial: The glutamate hypothesis of mood disorders: Neuroplasticity processes, clinical features, treatment perspectivesncbi.nlm.nih.gov
67. Association Between the Fronto-Limbic Network and Cognitive and Emotional Functioning in Individuals With Bipolar Disorder: A Systematic Review and Meta-analysisjamanetwork.com
68. Association Between the Fronto-Limbic Network and Cognitive and Emotional Functioning in Individuals With Bipolar Disorder: A Systematic Review and Meta-analysis - PubMedpubmed.ncbi.nlm.nih.gov
69. ARTICLE IN PRESS ARTICLE IN PRESS https://doi.org/10.1038/s41398-026-03868-znature.com
70. Lithium effects in the frontolimbic circuitry: a systematic review of neuroimaging findings in bipolar disordernature.com
71. Valproate regulates inositol synthesis by reducing expression of myo-inositol-3-phosphate synthasencbi.nlm.nih.gov
72. Valproate regulates inositol synthesis by reducing expression of myo-inositol-3-phosphate synthasenature.com
73. Valproate regulates inositol synthesis by reducing expression of myo-inositol-3-phosphate synthase - PMCpmc.ncbi.nlm.nih.gov
74. A common mechanism of action for three mood-stabilizing drugsnature.com
75. Bipolar Disorder Differential Diagnosesemedicine.medscape.com
76. Bipolar Disorder Clinical Presentation: History, Physical Examinationemedicine.medscape.com
77. Bipolar disorder in adults: Assessment and diagnosis - UpToDateuptodate.com
78. Bipolar ICD-10 Coding & Clinical Essentials - Blueprintblueprint.ai
79. Special Report: Bipolar Disorder II—Frequently Neglected, Misdiagnosedpsychiatryonline.org
80. Bipolar Disorders. 2018;20:97–170. wileyonlinelibrary.com/journal/bdicanmat.org
81. Why Bipolar Disorder Can Be Misdiagnosedcerebral.com
82. Why Bipolar Disorder is Often Misdiagnosed in Teens - Child Mind Institutechildmind.org
83. Misdiagnosis of Bipolar Disorderajmc.com
84. Clinical Medication Development for Bipolar Disorder and Alcohol Use Disorderscdn.clinicaltrials.gov
85. Evaluation of the 13-item Hypomania Checklist and a brief 3-item manic features questionnaire in primary carencbi.nlm.nih.gov
86. Bipolar Disorder Diagnosis: What to Expect and Why It Takes So Longswayhealth.co
87. Bipolar disorder - Diagnosis and treatment - Mayo Clinicmayoclinic.org
88. Exploring factors of diagnostic delay for patients with bipolar disorder: a population-based cohort study - PMCpmc.ncbi.nlm.nih.gov
89. Pharmacological treatment for bipolar mania: a systematic review and network meta-analysis of double-blind randomized controlled trialsnature.com
90. Treatment of bipolar disorder: A focus on medication therapy for maniamanagedhealthcareexecutive.com
91. Lithium and Valproate in Bipolar Disorder: From International Evidence-based Guidelines to Clinical Predictorscpn.or.kr
92. CD004052 valproate acute maniacochrane.org
93. An Update on the Role of Valproate in Bipolar Disorderpsychiatrictimes.com
94. Maintenance Treatment for Bipolar: Lamictal & Lithium,webmd.com
95. A Placebo-Controlled 18-Month Trial of Lamotrigine and Lithium Maintenance Treatment in Recently Manic or Hypomanic Patients With Bipolar I Disorderjamanetwork.com
96. Lamotrigine as a treatment for prevention of recurrence of bipolar disordercochrane.org
97. Lamotrigine in the maintenance treatment of bipolar disorder - Hashimoto, Y - 2021cochranelibrary.com
98. Mood stabilizers and/or antipsychotics for bipolar disorder in the maintenance phase: a systematic review and network meta-analysis of randomized controlled trialsncbi.nlm.nih.gov
99. Quetiapine for Bipolar Disorder - NCBI Bookshelf - NIHncbi.nlm.nih.gov
100. The CANMAT and ISBD Guidelines for the Treatment of Bipolar Disorder: Summary and a 2023 Update of Evidence - PMCpmc.ncbi.nlm.nih.gov
101. Quetiapine for Bipolar Disordercanjhealthtechnol.ca
102. Bipolar Depression Medication and Treatment Optionsmedcentral.com
103. CANMAT-ISBD Guidelines for Bipolar Depressionpsychopharmacologyinstitute.com
104. Add-on Lamotrigine Treatment for Subsyndromal Depression after Manic or Mixed States in Bipolar Disorder Improved the Quality of Life - PMCpmc.ncbi.nlm.nih.gov
105. Frontiersfrontiersin.org
106. A Phase III Study of SM-13496 in Patients With Bipolar I Depression.cdn.clinicaltrials.gov
107. www.cochrane.orgcochrane.org
108. Lamotrigine Therapy for Bipolar Depression: Analysis of Self-Reported Patient Datancbi.nlm.nih.gov
109. Evidence-Based Psychotherapies for Bipolar Disorder - PMCpmc.ncbi.nlm.nih.gov
110. Practical Psychosocial Management for Patients With Bipolar Disorderpsychiatryonline.org
111. A Review of Evidence-Based Psychosocial Interventions ...psychiatrist.com
112. Family-Focused Therapy: An Emerging Approach on the Treatment of Bipolar Disorderbiotech-asia.org
113. Psychosocial treatment and interventions for bipolar disorder: a systematic reviewlink.springer.com
114. Bipolar disorder in adults: Indications for and efficacy of electroconvulsive therapy - UpToDateuptodate.com
115. Electroconvulsive Therapy in Mania: A Review of 80 Years of Clinical Experiencepsychiatryonline.org
116. Electroconvulsive Therapy-Induced Mania in Bipolar Disorder: A Case Report - PMCpmc.ncbi.nlm.nih.gov
117. Electroconvulsive Therapy for Bipolar Disorder: Evidence Supporting What Clinicians Have Long Knownpsychiatrist.com
118. Electroconvulsive therapy for acute affective episodes in people with bipolar disorder - PMCpmc.ncbi.nlm.nih.gov
119. Pilot Trial of Peer Support for Bipolar Disorderclinicaltrials.gov
120. Combining technical and expert-by-experience knowledge in the quest for personal recovery from bipolar disorder: a qualitative studyncbi.nlm.nih.gov
121. A systematic review and meta-analysis of premature mortality in bipolar affective disorder - PubMedpubmed.ncbi.nlm.nih.gov
122. Prevention of Suicidal Behavior in Bipolar Disorderpsychiatryonline.org
123. Suicide Assessment and Prevention in Bipolar Disorder: How Current Evidence Can Inform Clinical Practice - PMCpmc.ncbi.nlm.nih.gov
124. Suicide Attempts and Completions in Patients With Bipolar Disorderpsychiatrictimes.com
125. The influence of depressive and manic symptoms on suicidal ideation in mixed mood statesncbi.nlm.nih.gov
126. Life expectancy and years of potential life lost in bipolar disorderovid.com
127. Life expectancy and years of potential life lost in bipolar disorder: systematic review and meta-analysiscambridge.org
128. Cardiovascular Disease and Bipolar Disorder: Risk and Clinical Implicationspsychiatrist.com
129. Medical comorbidity in bipolar disorder: The link with metabolic-inflammatory systems - ScienceDirectsciencedirect.com
130. Cardiovascular disease and bipolar disorder: A review of pathophysiology and treatment implications - Mayuresh Chaudhari, Luis Mendez, Rene L. Olvera, Sudha Seshadri, Antonio L. Teixeira, 2025journals.sagepub.com
131. Managing metabolic syndrome in bipolar disorder: multidisciplinary monitoring and treatment strategies - PubMedpubmed.ncbi.nlm.nih.gov
132. All-cause and cause-specific mortality among people with bipolar disorder: a large-scale systematic review and meta-analysisncbi.nlm.nih.gov
133. Bipolar disorder and the risk of cardiometabolic diseases, heart failure, and all-cause mortality: a population-based matched cohort study in South Koreanature.com
134. History of Bipolar Manic-Depressive Disorderresearchgate.net
135. History of Bipolar Disorderpsychcentral.com
136. CONTENTS Contributors ix Introduction to Psychological Assessment of Bipolarapa.org
137. Letter to the Editor: CONVERGENCES AND DIVERGENCES IN THE ICD-11 VS. DSM-5 CLASSIFICATION OF MOOD DISORDERS - PubMedpubmed.ncbi.nlm.nih.gov
138. Lithium: Past, Present, and Futurepsychiatrictimes.com
139. Lithium use in special populationsncbi.nlm.nih.gov
140. History of bipolar disorderen.wikipedia.org
141. A Short History of Bipolar Disorderpsychologytoday.com
142. A Brief History of Bipolar Disorderrestoringwellnesssolutions.com
143. History of Bipolar Disorderbipolarlife.org.au
144. History of bipolar disorder — Grokipediagrokipedia.com
145. The history of lithium therapyresearchgate.net
146. A Brief History of Lithium Treatment in Psychiatrypsychiatrist.com
147. The history of lithium therapy - Shorter - 2009 - Bipolar Disorders - Wiley Online Libraryonlinelibrary.wiley.com
148. Bipolar disorders in the new DSM-5 and ICD-11 classificationselsevier.es
149. Affective disorders-developments of ICD-11 in comparison with ICD-10 - PubMedpubmed.ncbi.nlm.nih.gov
150. Bipolar II disorder: a state‐of‐the‐art review - PMCpmc.ncbi.nlm.nih.gov
151. Toward a re-definition of subthreshold bipolarity: epidemiology and proposed criteria for bipolar-II, minor bipolar disorders and hypomania - PubMedpubmed.ncbi.nlm.nih.gov
152. Would Broadening the Diagnostic Criteria for Bipolar Disorder Do More Harm Than Good? Implications From Longitudinal Studies of Subthreshold Conditionspsychiatrist.com
153. Bipolar II disorder: arguments for and against a distinct diagnostic entity - PubMedpubmed.ncbi.nlm.nih.gov
154. Bipolar disorder and schizophrenia: distinct illnesses or a continuum? - PubMedpubmed.ncbi.nlm.nih.gov
155. Diagnostic issues in bipolar disorder - ScienceDirectsciencedirect.com
156. Bipolar II Disorder: Current Issues in Diagnosis and Managementpsychiatrictimes.com
157. Pediatric Bipolar Disorder: The Controversy and the Reality - International Bipolar Foundationibpf.org
158. ‘Pediatric Bipolar Disorder’ rates are still lower than claimed: a re-examination of eight epidemiological surveys used by an updated meta-analysisjournalbipolardisorders.springeropen.com
159. ‘Pediatric Bipolar Disorder’ rates are still lower than claimed: a re-examination of eight epidemiological surveys used by an updated meta-analysisncbi.nlm.nih.gov
160. Misdiagnosis of bipolar disorder in children and adolescents: a comparison with ADHD and major depressive disorder - PubMedpubmed.ncbi.nlm.nih.gov
161. Paediatric bipolar disorder and its controversycambridge.org
162. Controversies concerning the diagnosis and treatment of bipolar disorder in childrenncbi.nlm.nih.gov
163. Pediatric Bipolar Disorder: Challenges in Diagnosis and Treatmentlink.springer.com
164. This Issue: Schizophrenia and Bipolar Disorder: No Dichotomy, A Continuum, or One Disease?healio.com
165. Volumes of hippocampal subfields suggest a continuum between schizophrenia, major depressive disorder and bipolar disorderncbi.nlm.nih.gov
166. Biomarkers of bipolar disorder based on metabolomics: A systematic review - ScienceDirectsciencedirect.com
167. Systematic Review and Meta-Analysis on MS-Based Proteomics Applied to Human Peripheral Fluids to Assess Potential Biomarkers of Bipolar Disorderncbi.nlm.nih.gov
168. Biomarkers of neuroprogression and late staging in bipolar disorder: A systematic review - PMCpmc.ncbi.nlm.nih.gov
169. Full article: Biomarkers for bipolar disorder: current insightstandfonline.com
170. Biomarkers of Bipolar Disorder in Late Life: An Evidence-Based Systematic Reviewlink.springer.com
171. Pharmacogenomics and response to lithium in bipolar disorder: Pharmacogenomics: Vol 25 , No 16-18 - Get Accesstandfonline.com
172. Using multi-trait polygenic scores to predict lithium responsiveness in patients with bipolar disorder - PMCpmc.ncbi.nlm.nih.gov
173. Using multi-trait polygenic scores to predict lithium responsiveness in patients with bipolar disordermedrxiv.org
174. Daily mood monitoring of symptoms using smartphones in bipolar disorder: A pilot study assessing the feasibility of ecological momentary assessment - PubMedpubmed.ncbi.nlm.nih.gov
175. Digital Biomarkers for Passive Remote Monitoring of Bipolar Disorder: Systematic Reviewmedrxiv.org
176. Frontiersfrontiersin.org