RESEARCH DOSSIER
Bipolar and Related Disorders
Abstract. This fundamental research monograph investigates bipolar I and II disorders, cyclothymia, related bipolar presentations, and the organization of manic, hypomanic, depressive, and mixed states over the life course. A structured critical narrative review integrates the historical development of nosology, ICD-11 and DSM-oriented assessment, longitudinal illness-course studies, international epidemiology, genomic evidence, circadian and cognitive theories, psychosocial mechanisms, randomized acute and maintenance trials, psychotherapy research, safety monitoring, and global recovery challenges. Particular attention is given to diagnostic validity, temporal measurement, phase-specific treatment evidence, reproductive and physical-health risks, suicidality, methodological bias, and clinically relevant uncertainty. The synthesis distinguishes observed associations, causal experimental findings, model-based explanations, and jurisdiction-specific guidance. It proposes a research agenda centered on prospective trajectories, meaningful functioning, equitable care, and validated human–artificial research tools. This is an evidence-oriented narrative monograph rather than an original clinical study, a registered systematic review, or a new meta-analysis.
Keywords: Bipolar; and; Related; Disorders
Introduction: The Scientific Object and Its Boundaries
Bipolar and related disorders comprise a group of recurrent or persistent mood conditions organized around clinically significant episodes of mania, hypomania, depression, and changes in activation, sleep, cognition, behavior, and functioning. Their defining scientific problem is temporal: an assessment of an individual at a single moment may describe depression, agitation, psychosis, anxiety, or apparent wellness, while the diagnosis rests upon the configuration and sequence of states across the life course. The research object of this dossier is therefore the longitudinal organization of bipolar illness, from episode phenomenology to recurrence, developmental context, biological susceptibility, differential diagnosis, treatment, and recovery. World Health Organization. (2026) and National Institute of Mental Health. (2024) provide accessible institutional descriptions, while World Health Organization. (2024) and National Institute for Health and Care Excellence. (2025) establish more precise diagnostic and management contexts.
The word bipolar denotes the capacity for clinically consequential changes in mood polarity, but polarity alone is insufficient to explain the disorder. A manic state entails changes in energy or activity, self-evaluation, sleep requirement, speech, thought organization, decision making, and interpersonal conduct. A depressive state may entail slowed action, anhedonia, hopelessness, fatigue, cognitive difficulty, altered appetite or sleep, and suicidal thinking. Subthreshold symptoms and periods of relative stability occupy substantial proportions of some longitudinal courses. No single biological assay currently replaces the clinical assessment of these trajectories. In examining mechanisms we therefore preserve the distinction between a validated clinical syndrome and hypotheses about its causes.
This monograph is a structured critical narrative synthesis of original longitudinal investigations, experimental and observational studies, randomized trials, epidemiology, genetic research, clinical diagnostic manuals, and evidence-based guidelines. It does not report original experiments, a prespecified systematic search, or a new meta-analysis. The primary question is how observable episodes, illness course, and functional outcomes can be explained and managed using evidence whose certainty varies by research design. The broader psychological setting includes sleep, relationships, developmental adversity, motivation, executive control, and social participation; clinical treatments are analyzed as empirical interventions under jurisdiction-specific requirements, without patient-specific prescribing.
The Formation of Bipolar Concepts in Clinical History
Descriptions of melancholy and excitement preceded contemporary nosology by many centuries, but historical resemblance cannot itself establish that an ancient description refers to the modern bipolar syndrome. The decisive modern transformation involved longitudinal clinical observation: clinicians began to classify recurrent mood disturbance by the pattern of episodes and recovery over time. Emil Kraepelin's manic-depressive conception assembled forms of periodic and recurrent affective illness under a broad clinical heading. Later distinctions between unipolar depressive illness and forms characterized by mania or hypomania progressively sharpened the bipolar category. The history matters because it shows that episode-based classification emerged from attempts to organize courses of illness, long before the current biomarker and genomic programs.
The treatment history also changed the scientific meaning of prognosis. In his original 1949 paper, Cade, J. F. J. (1949) described lithium salts in patients with psychotic excitement. His small, uncontrolled observations were historically seminal but cannot by contemporary standards establish generalizable effectiveness or long-term safety. Subsequent trials, monitoring procedures, and recurrence-prevention studies transformed lithium into a treatment supported by much stronger evidence. The contrast between an initial clinical observation and modern comparative trials is central to understanding evidence development, particularly when reconstructing treatment claims that appear convincing in historical narratives.
Modern international classification systems emphasize clinical descriptions, duration, impairment, exclusion of substances and medical causes, and episode patterns. The 2024 World Health Organization. (2024) provide internationally oriented requirements; the DSM-5-TR is a separate diagnostic framework with its own wording and clinical specifications. They overlap extensively but must be examined independently when making exact diagnostic or coding claims. Historical constructs such as a broad bipolar spectrum remain useful research proposals where operationalized; the diagnostic thresholds used in treatment studies are narrower and should never be silently replaced by a theoretical spectrum.
Episode Phenomenology: Mania, Hypomania, and Depression
A manic episode encompasses a persistent, marked change in mood and increased energy or activity, associated with a recognizable constellation of symptoms and substantial change in functioning. Elevated or expansive mood is one presentation; irritability can be prominent. Decreased need for sleep is distinct from insomnia because the person may sleep very little without the expected experience of fatigue. Accelerated speech, flight of ideas, distractibility, grandiosity, goal-directed overactivity, and unusually risky decisions can co-occur in varying configurations. Psychotic symptoms and hospitalization can occur in severe mania, and clinical urgency depends on overall danger, impairment, judgment, and capacity for safe care rather than mood valence alone.
Hypomania shares several symptom dimensions with mania while differing in severity and functional consequences. It involves an observable change from baseline and requires credible longitudinal evidence. Pleasant subjective energy, ordinary enthusiasm, irregular sleep, or creativity alone are weak substitutes for an episode assessment. The question is whether a distinct interval represents a clinically meaningful change in activity and mood accompanied by additional characteristic symptoms, and whether competing explanations account for it. Collateral accounts can be particularly valuable when a person remembers a high-energy interval positively or incompletely. National Institute for Health and Care Excellence. (2025) recommends a detailed history including episodes of disinhibition, changes in behavior, and information from family or carers where appropriate.
Depressive episodes in bipolar illness share many features with episodes diagnosed in major depressive disorder. The defining differentiator is the broader longitudinal pattern, rather than a unique symptom that establishes bipolarity during depression. A history of mania or hypomania, family history, sleep–wake changes, prior treatment-emergent activation, and age and pattern of previous episodes can inform the assessment, while none is individually diagnostic. The distinction matters because the evidence supporting treatment of unipolar depression cannot be transferred wholesale to bipolar depression. Sachs, G. S., Nierenberg, A. A., Calabrese, J. R., et al. (2007) addressed this difference experimentally in the STEP-BD research program.
Diagnostic Architecture: Bipolar I, Bipolar II, Cyclothymia, and Other Conditions
Bipolar I disorder is defined by a history of at least one manic episode under the applicable diagnostic system. Major depressive episodes are common but are not the defining requirement for every bipolar I diagnosis. Bipolar II disorder entails hypomanic and major depressive episodes without a history of a full manic episode; an eventual manic episode changes the diagnostic formulation. Bipolar II is not properly understood as a uniformly minor illness: the associated depressive burden and long-term impairment can be substantial. Judd, L. L., Akiskal, H. S., Schettler, P. J., et al. (2003) documented the prolonged symptomatic course of a bipolar II cohort, providing a longitudinal counterweight to classifications based only on the maximum intensity of activation.
Cyclothymia refers to persistent fluctuations involving numerous periods of elevated and depressed symptoms that do not necessarily meet full episode thresholds, within the duration and exclusion rules of the applicable classification. Other specified and unspecified bipolar-related categories accommodate clinically consequential presentations when full criteria are unmet or available information is insufficient. The distinctions among these categories depend on episode duration, symptom count, impairment, developmental timing, observed change, and other diagnostic requirements. In research, these are measurement decisions with consequences for sample composition and treatment generalization, not merely administrative labels.
The scientific problem of a bipolar spectrum remains open to construct analysis. Broad-spectrum definitions capture forms of activation and instability excluded from strict categories, potentially improving sensitivity at the cost of specificity and increasing overlap with personality, anxiety, attention, sleep, and substance-related phenomena. Reliable clinical research therefore declares whether it studies DSM or ICD diagnoses, subthreshold syndromes, screening-positive groups, or transdiagnostic symptom dimensions. Claims made for one population should retain that population in the statement of evidence.
Mixed Features, Psychosis, Rapid Cycling, and Diagnostic Specifiers
Mixed presentations combine manic or hypomanic features with depressive symptoms, or depressive episodes with features of activation. They are clinically significant because high energy, agitation, impulsivity, hopelessness, and suicidal risk may coexist. A mixed-features specification is not identical to every historical account of a mixed state; precise application requires the relevant classification. A clinician's assessment examines the temporal overlap of specific signs, course, treatment history, and potential medication or substance effects. National Institute for Health and Care Excellence. (2025) treats mixed affective states as requiring special vigilance, including monitoring for emergent depression.
Psychosis can accompany severe mood episodes. Delusions or hallucinations do not automatically identify schizophrenia, and conversely the presence of mood symptoms does not settle a psychotic-disorder differential diagnosis. The timing of psychosis relative to mood episodes, the proportion of longitudinal illness occupied by psychosis, functional change, substance exposure, and evidence across settings all matter. Early diagnostic assignments may be provisional; the relevant scientific task is longitudinal diagnostic adjudication. This is one reason to distinguish episode assessment from a lifetime clinical formulation.
Rapid cycling is a course descriptor under specified diagnostic rules and is commonly operationalized through the frequency of qualifying mood episodes within an interval. It should not be equated with ordinary hour-to-hour emotional changes. An account of apparently rapid mood variability requires clarification of sustained episode boundaries, intervening remission, mixed presentations, sleep disruption, medications, and possible comorbid processes. Such precision is critical to comparing research samples, as treatment-response estimates may change when operational definitions are inconsistent.
Longitudinal Course and Subsyndromal Burden
Episode counts alone substantially underrepresent the duration and texture of bipolar illness. The prospective program by Judd, L. L., Akiskal, H. S., Schettler, P. J., et al. (2002) used weekly symptom status to demonstrate extensive depressive and subsyndromal periods in bipolar I disorder; its counterpart for Judd, L. L., Akiskal, H. S., Schettler, P. J., et al. (2003) followed a smaller cohort over many years. These studies are important because they separate time spent with symptoms from the number of dramatic episodes that bring a person into acute care. Their samples, assessment schedules, clinical era, and attrition require consideration before extrapolating exact proportions to contemporary populations.
Illness trajectory is heterogeneous. Some people experience infrequent episodes and substantial periods of recovered function. Others encounter recurring depression, persistent residual symptoms, frequent polarity changes, comorbidity, or treatment complications. Age at onset, early episode severity, access to appropriate care, substance use, sleep disturbance, social conditions, and previous response may be informative, but prognostic correlations can be distorted by clinical ascertainment, follow-up losses, and confounding. The task of a longitudinal model is to predict individual outcomes with calibration, not to turn a population average into a personal forecast.
Functional recovery and symptomatic recovery also diverge. Reduction in syndromal symptoms may coexist with debt, disrupted education, occupational loss, damaged relationships, executive difficulty, or the accumulated effects of stigma. Longitudinal assessment should therefore follow participation, cognition, quality of life, physical health, and patient-defined goals alongside mood symptoms. This broader perspective shifts treatment evaluation from an exclusive focus on episode suppression to sustained opportunities for meaningful participation.
Epidemiology, Global Burden, and Measurement Differences
The magnitude of bipolar illness depends on the case definition, time frame, assessment instrument, population, and service context. In the cross-national World Mental Health initiative, Merikangas, K. R., Jin, R., He, J.-P., et al. (2011) analyzed household interviews with 61,392 adults in eleven countries and reported aggregate lifetime prevalence estimates of 0.6% for bipolar I, 0.4% for bipolar II, and 2.4% for a broader bipolar spectrum. These results describe that survey's definitions and sampled populations. They should not be combined uncritically with health-system diagnoses or contemporary global model estimates.
The World Health Organization. (2026), reports an estimated 36 million people affected worldwide in 2023, corresponding to roughly 0.4% of the global population under its estimation method. The same report emphasizes disability, the treatment gap, stigma, comorbidity, and premature mortality. A model-based point estimate of people currently living with the condition and a household-survey lifetime prevalence are different quantities, so their numerical contrast is not an epidemiologic contradiction. In the United States, National Institute of Mental Health. (n.d.) provides older survey-based estimates that must be explicitly dated rather than presented as contemporaneous incidence.
An epidemiological model has to consider help-seeking, access to diagnosis, cultural interpretation of activation, diagnostic recognition of hypomania, underrepresentation of disadvantaged groups, and changes in reporting standards. Stigma can both limit service contact and shape self-report. Hospitalized samples oversample severe illness; online community samples can oversample self-identified presentations; routine-care databases reflect coding practices and insurance access. International comparisons become credible when the denominator, age range, observation period, measurement method, and uncertainty are specified. Research should also differentiate diagnostic prevalence from the social burden caused by untreated symptoms.
Genetic Architecture and Familial Susceptibility
Family aggregation and twin studies support a substantial heritable contribution to bipolar disorder, while inheritance is polygenic and environment-dependent. The most consequential recent discovery is the multi-ancestry genomic analysis by O'Connell, K. S., Koromina, M., van der Veen, T., et al. (2025), which combined 158,036 cases and approximately 2.8 million controls, identifying 298 genome-wide significant loci and prioritizing 36 credible genes using several mapping approaches. The investigators also reported differences by ancestry, ascertainment, and bipolar subtype. These findings strongly support distributed genetic susceptibility rather than a single diagnostic gene.
Genome-wide association does not identify a universal molecular cause. Association loci may map imperfectly to genes; effects are generally small, linkage patterns and allele frequencies differ across ancestry groups, and genetic correlations between psychiatric categories do not erase diagnostic differences. Polygenic scores are sample-dependent statistical predictors, with performance affected by training ancestry, prevalence assumptions, phenotype accuracy, and the target population. Current research does not justify diagnosing an individual with bipolar disorder from a polygenic score alone.
The relationship between inherited vulnerability and course is also conceptually difficult. Genetic susceptibility may influence initial episode liability, circadian characteristics, other psychiatric dimensions, medication response, or intermediate phenotypes. Environmental exposures and social support may shape whether, when, and how these liabilities become clinically expressed. A comprehensive causal account requires longitudinal and mechanistic work that distinguishes correlation from gene–environment interaction, and that avoids treating genetic association as inevitability.
Neural Systems, Reward, and Affective Regulation
Neurobiological accounts examine neural networks supporting reward valuation, salience, affective reactivity, executive control, motivational pursuit, and self-regulation. Manic activation has been investigated through altered reward sensitivity and approach behavior, while depressive states are associated with diminished reward engagement and negative cognitive-affective processing. Such explanatory families are research models that organize converging observations; they are not person-level biomarkers with sufficient specificity to establish a clinical diagnosis. Similar task and imaging findings can appear in depression, psychosis, substance-use disorders, attention conditions, and healthy volunteers under particular circumstances.
Functional neuroimaging provides valuable measurements of blood-oxygen-level-dependent responses and network organization, yet these signals are indirect indicators of neural activity. Cross-sectional studies cannot determine whether an observed difference preceded illness, emerged during episodes, resulted from medication, or reflected sleep deprivation or comorbidity. Small samples, site differences, motion, analytic flexibility, and partial independence of participant datasets have historically limited many psychiatric imaging literatures. Experimental task manipulations, within-person repeated imaging, harmonized multisite design, and preregistered analysis can improve mechanistic discrimination.
Bipolar disorder also exposes the limits of a single-level explanation. Neural activity occurs within an organism whose sleep timing, interpersonal environment, reward contingencies, financial risks, and medication exposure change over time. Mechanistic progress requires demonstrating how candidate processes predict episode transitions, and whether manipulating a process reliably changes clinically meaningful outcomes. A molecular finding, a cognitive mechanism, and a causal clinical intervention must remain evidentially distinct.
Circadian Timing, Sleep, and Social Rhythms
Sleep and circadian organization occupy a central position in bipolar research. Decreased need for sleep can be an episode symptom, while prolonged sleep loss may also precede or amplify activation in vulnerable people. Circadian phase, regularity of light exposure, shift work, travel, social activity timing, and medication-related sedation may influence observable rhythms. These pathways are intertwined, complicating simple causal claims: sleep disturbance can be a prodrome, a maintaining process, a consequence of an episode, or a signal of a separate medical or sleep disorder.
The social-rhythm hypothesis contributed to the development of interpersonal and social rhythm therapy, which seeks to stabilize daily routines while addressing interpersonal changes and stress. Frank, E., Swartz, H. A., & Kupfer, D. J. (2000)00969-0) described its theoretical framework; Frank, E., Kupfer, D. J., Thase, M. E., et al. (2005) investigated longitudinal outcomes in a randomized program. The clinical value of these studies is tied to their structured intervention and outcome assessments. Interpretation is constrained by participant selection, treatment context, and the difficulty of isolating rhythm regularity from the other therapeutic components.
Rhythm measurement now includes diaries, ecological momentary assessment, actigraphy, and wearable sensors. These provide complementary representations: self-reported need for sleep, observed rest–activity patterns, and circadian physiological phase are not interchangeable. Digital sensors can identify within-person deviation but also generate false alarms from work schedules, exercise, battery failures, or device removal. The most useful research designs establish stable individual baselines, characterize context and medication effects, and test whether rhythm signals add predictive value beyond existing clinical information.
Cognition, Attention, and Functional Performance
Cognitive difficulties in bipolar disorder may include attention instability, changes in processing speed, verbal learning, working memory, and executive function. Their expression varies across manic, depressed, and relatively euthymic states. Research must separate the transient consequences of an episode from enduring differences, and both from medication effects, disrupted sleep, education, and co-occurring neurodevelopmental or substance-related problems. A deficit measured during an acute admission is not a direct measure of long-term cognitive capacity.
Standard neuropsychological instruments offer reliable procedures, but score interpretation depends on normative representation, repeated testing effects, motivation, symptom severity, and clinical context. A longitudinal cognitive study should distinguish practice-related improvement from functional recovery, and measured laboratory performance from effectiveness in everyday life. Individual profiles may be uneven, so group means should not be translated into generalized incapacity or an assumption about work, creativity, or responsibility.
Cognition also shapes illness recognition and treatment participation. Executive difficulty may interfere with medication management, sleep planning, and maintaining follow-up appointments. Conversely, environmental supports, simplified routines, shared planning, accessible communication, and workplace adjustments can improve participation without requiring every cognitive measure to normalize. The central outcome is often whether a person can sustain self-directed activity within a supportive environment.
Psychosocial Stress, Interpersonal Context, and Social Determinants
Interpersonal stress can coincide with episode onset, recovery, or recurrence. Family environments, bereavement, conflict, discrimination, insecure housing, and financial strain can affect sleep, access to treatment, and coping resources. Because episodes themselves can disrupt relationships and employment, observational associations face substantial bidirectionality. A study linking stressful life events to later symptoms must establish temporal order, assess preceding prodromal signs, and account for the possibility that early changes in behavior contribute to the apparent stress exposure.
The empirical literature on family-focused psychotherapy demonstrates that social context can be an intervention target, even when it is not assigned sole causal responsibility for the disorder. Rea, M. M., Tompson, M. C., Miklowitz, D. J., Goldstein, M. J., Hwang, S (2003) compared a manualized family intervention with individually focused care among recently hospitalized patients, while Miklowitz, D. J., George, E. L., Richards, J. A., Simoneau, T. L., & Sud (2003) studied family psychoeducation in outpatient management. Such trials support the value of structured family communication and relapse planning in relevant contexts; they do not imply that families create bipolar illness.
The structural dimension deserves its own analytic treatment. Health-system access, paid leave, medication costs, continuity of care, transport, culturally appropriate assessment, and safe housing affect the feasibility of recommended treatments. Observational evidence about social gradients should be interpreted alongside policy and service design, not compressed into individual behavior explanations. A recovery-oriented research program measures opportunity, autonomy, social connection, and treatment access as outcomes that matter independently of symptom scales.
Development, Childhood, Adolescence, and Diagnostic Caution
Bipolar illness can begin in youth, but assessing elevated mood, irritability, sleep disturbance, activity, and impulsivity in children requires developmental specificity. Ordinary developmental variation, attention-deficit/hyperactivity disorder, trauma-related activation, substance exposure, depression, and disruptive behavior conditions may overlap with components of bipolar presentations. A sustained, episodic departure from baseline is crucial to the clinical reasoning. A single brief encounter, questionnaire, or parent description cannot replace longitudinal, multi-informant assessment.
Research in adolescents is further complicated by developmental changes in sleep timing, school demands, social relationships, and treatment tolerability. Trial populations may differ in episode history, age, psychosocial support, and comorbid conditions. Adult pharmacological findings cannot be presumed to generalize across all pediatric ages. Miklowitz, D. J., Axelson, D. A., Birmaher, B., et al. (2008) examined family-focused treatment for adolescents using a randomized design, providing relevant evidence for adjunctive care while leaving questions of access, developmental tailoring, and longer follow-up.
Early identification is most helpful when it improves accurate assessment and timely support without prematurely fixing a lifelong identity on uncertain symptoms. Prospective work is needed to understand which early patterns predict later confirmed bipolar episodes, how often initial formulations change, and what interventions improve outcomes among different risk groups. Prediction metrics must be judged by calibration, false positives, burden, and actionability, not merely group separation.
Pregnancy, Postpartum Periods, and Reproductive Safety
Pregnancy planning and the postpartum period raise clinically important questions of episode vulnerability, sleep disruption, medication exposure, and the consequences of discontinuation. Individual risk depends on previous episode polarity and severity, prior postpartum events, treatment response, current stability, available care, and personal reproductive goals. Evidence-based decision making requires coordinated psychiatric and reproductive-health expertise. Abrupt changes in maintenance treatment can carry relapse risk, while some medications have substantial fetal or reproductive hazards; these considerations should be discussed prospectively wherever feasible.
Valproate illustrates the importance of distinguishing therapeutic evidence from regulatory permissions. The Medicines and Healthcare products Regulatory Agency. (2025) has introduced restrictions and safety measures concerning reproductive risks in female and male patients, with updates in 2025. These rules are jurisdiction-specific and should not be generalized as identical legal requirements everywhere. National Institute for Health and Care Excellence. (2025) links to current safety requirements. Published information on embryo-fetal risk, interactions, and alternative options must be kept current in any clinical review.
This field also illustrates why observational comparisons of exposed and unexposed pregnancies are difficult. Illness severity, concurrent medications, treatment changes, indication, smoking, substance exposure, and the availability of specialist follow-up can confound outcomes. Randomization is often unethical or infeasible. Registries, active comparators, carefully documented timing of exposure, and transparent uncertainty are therefore essential. The objective of research synthesis is to inform informed, collaborative clinical decisions rather than issue universal individual medication instructions.
Comorbidity and Differential Diagnosis
Anxiety disorders, substance-use disorders, attention problems, trauma-related conditions, sleep disorders, and physical illnesses frequently intersect with bipolar presentations. Comorbidity is clinically important because it can modify symptom severity, adherence, suicidality, functional outcomes, and the interpretation of treatment response. It also presents a measurement challenge: diagnostic interviews must determine whether a symptom occurs only during a mood episode, persists outside episodes, or belongs to another independent syndrome.
The differential diagnosis of unipolar depression is particularly consequential because many people first enter care during depression. A careful interview explores earlier periods of activation, duration, changes in sleep need, family observations, and treatment history. Mood Disorder Questionnaire screening can direct attention to relevant lifetime symptoms, but the Hirschfeld, R. M. A., Williams, J. B. W., Spitzer, R. L., et al. (2000) does not authorize using a screening threshold as a stand-alone diagnostic decision. Predictive values vary with setting and prevalence, and sensitivity may be different across bipolar subtypes.
Medical and medication-related mimics include thyroid conditions, neurologic problems, intoxication or withdrawal, stimulant effects, and other pharmacological exposures. Severe agitation or insomnia can occur in several disorders and in response to environmental stress. National Institute for Health and Care Excellence. (2025) explicitly calls for evaluation of physical health, medication history, substance use, and competing psychiatric diagnoses. The sound diagnostic question concerns the complete temporal and causal formulation, including what evidence would alter the working hypothesis.
Assessment Methods, Screening, and Measurement Validity
A comprehensive assessment begins with a chronological interview covering depressive and elevated episodes, their frequency and duration, observable functional change, impairment, psychosis, suicidality, treatment exposure, and the consequences of symptoms. Records, previous clinical notes, and collateral accounts can improve reconstruction of past states when accessible and appropriate. The interviewer needs to distinguish self-described energy from objectively changed activity and to examine context without presupposing a diagnosis. Diagnostic certainty should be revised when new longitudinal information becomes available.
The Young, R. C., Biggs, J. T., Ziegler, V. E., & Meyer, D. A. (1978), provides clinician-rated assessment of manic symptom severity. Depression scales, clinician global ratings, patient-reported outcome tools, structured diagnostic interviews, and daily mood charts each address different targets. The Mood Disorder Questionnaire is a screen; symptom scales quantify current burden; standardized interviews establish operational diagnostic categories. An instrument can demonstrate high inter-rater reliability yet lack specificity for a particular clinical population or outcome. Construct validity, reference standards, responsiveness, and measurement invariance should be assessed separately.
Repeated ecological assessments reduce dependence on long retrospective recall, but introduce compliance and reactivity challenges. Smartphone reports can oversample responses from those who remain engaged, and wearables cannot infer subjective mood directly from activity or sleep. Studies should document missingness, selection, device error, medication changes, prespecified outcome windows, and how momentary observations are transformed into episode-level judgments. In an illness marked by time-varying states, the measurement schedule is part of the scientific theory being tested.
Risk Assessment, Suicide, Self-Harm, and Acute Safety
Bipolar disorder is associated with elevated risk of self-harm and suicide, although risk varies markedly across individuals, episodes, comorbidity, and care contexts. Acute depression, mixed features, agitation, psychosis, substance use, previous self-harm, access to lethal means, and loss of supports may require urgent assessment. Risk evaluations should include current intent, planning, immediate safety, protective factors, and capacity to participate in care, while recognizing that actuarial prediction of rare events has major limitations. National Institute for Health and Care Excellence. (2025) includes comprehensive risk assessment as part of psychiatric care; World Health Organization. (2026) emphasizes the importance of accessible services.
Lithium's possible effects on suicidal behavior illustrate evidence triangulation. A Swedish within-person registry study by Song, J., Sjölander, A., Joas, E., Bergen, S. E., Runeson, B., Larsson, (2017) found lower rates of suicide-related events during lithium treatment intervals than untreated intervals, but such a design cannot completely eliminate time-varying confounding. A randomized-trial synthesis by Nabi, Z., Stansfeld, J., Plöderl, M., Wood, L., & Moncrieff, J. (2022) assessed relatively uncommon suicide outcomes, demonstrating the challenge of low event counts. A later synthesis of randomized trials by Wang, J. X., Le, G. H., Wong, S., Teopiz, K. M., Kwan, A. T. H., Rosenbl (2025) reported point estimates favoring lithium without statistically definitive reductions in several suicide outcomes. These bodies of evidence can be consistent with potential benefit while differing in their certainty about causal protection.
Emergency psychiatric care concerns a person's immediate condition, not a general internet screening score. Severe mania with impaired judgment, imminent self-harm risk, inability to meet essential needs, and major psychotic disturbance require timely professional evaluation. This research dossier analyzes these issues at the population and service level. It supplies neither individualized risk certification nor a substitute for emergency assessment.
Acute Mania: Treatment Evidence and Decision Structure
Acute mania management addresses symptom reduction, protection from immediate harm, restoration of sleep and functioning, and the causes or complications of the episode. Clinical guidelines distinguish medication already being taken, illness severity, psychosis, agitation, previous response, medical comorbidity, and patient preferences. According to the Keramatian, K., Chithra, N. K., & Yatham, L. N. (2023), several antipsychotics and mood-stabilizing treatments have evidence for acute mania, with some combinations appropriate in specified settings. U.S. Department of Veterans Affairs & U.S. Department of Defense. (2023) organizes its guidance through diagnosis, specialty care, mania/hypomania, and bipolar depression.
Evidence of acute antimanic effect is not equivalent to long-term recurrence prevention or tolerability. A short randomized trial may establish improvement in a mania scale while providing insufficient evidence about diabetes risk, renal outcomes, sedation, hospitalization, or sustained occupational recovery. Head-to-head comparisons, long-term registries, and direct patient-reported measures are needed alongside placebo-controlled efficacy trials. Antipsychotic side effects may include extrapyramidal symptoms, akathisia, metabolic change, weight gain, and sedation, with substantial differences across molecules and individuals.
Treatment interpretation also depends on context. Emergency admission, inpatient monitoring, involuntary interventions under local law, and adherence challenges can change the populations represented in trials. Clinical guideline recommendations are conditional on available licensing, monitoring, reproductive safeguards, and clinical judgment. National Institute for Health and Care Excellence. (2025) is explicit that lamotrigine is not an acute antimanic treatment. A well-structured evidence review preserves this phase specificity rather than treating all substances described as mood stabilizers as interchangeable.
Bipolar Depression and the Problem of Phase-Specific Efficacy
Bipolar depression contributes greatly to overall morbidity and is often the phase in which patients seek help. Its treatment presents a different evidence structure from acute mania, reflecting distinct symptom domains and different trial end points. Keramatian, K., Chithra, N. K., & Yatham, L. N. (2023) and National Institute for Health and Care Excellence. (2025) recommend selected pharmacological and psychological approaches but differ in hierarchy, licensing context, and the way evidence is operationalized. A scientific synthesis should state which phase and diagnostic subtype each trial studied before comparing recommendations.
In an updated network meta-analysis and systematic review, Yalin, N., Yildiz, A., Siafis, S., Vieta, E., & Leucht, S. (2026) described evidence supporting several second-generation antipsychotic regimens and lamotrigine among studied interventions, while noting lower confidence for other findings. A subsequent dose-related network analysis by Fornaro, M., Di Lorenzo, C., Oliva, V., De Prisco, M., & Vieta, E. (2026) included 103 randomized trials and 20,941 participants, but the conclusions depended on sensitivity analyses and important clinical effect modifiers. Network rankings should be interpreted through certainty, transitivity assumptions, dropout patterns, outcome definitions, and dose-specific design. Apparent differences between publications may reflect population selection and methodology rather than direct contradictory evidence.
Antidepressant trials require particular care. The Sachs, G. S., Nierenberg, A. A., Calabrese, J. R., et al. (2007) randomized participants receiving a mood stabilizer to an additional antidepressant or matching placebo and did not find the commonly presumed advantage in durable recovery from depression. That result does not show that every antidepressant, combination, or patient subgroup has identical effects; it highlights why unipolar depression practice cannot substitute for bipolar-specific trials. Concerns about mood switching, mixed activation, and cycle destabilization remain relevant in shared decision making and in phase-specific guideline interpretation.
Maintenance Treatment: Recurrence Prevention and Sustained Recovery
Maintenance research investigates prevention of new episodes, reduction of symptom burden, tolerability, safety, functioning, and persistence with care. Trials differ in whether patients were stabilized with the same drug before randomization, whether discontinuation effects are controlled, and which polarity of recurrence is counted. These distinctions matter because a study enriched with acute responders may overestimate effectiveness in a general population, while abrupt withdrawal in a comparator group may inflate apparent relapse risk. A credible maintenance synthesis explicitly identifies such designs rather than comparing all hazard ratios as though produced by equivalent experiments.
In the The BALANCE investigators and collaborators. (2010)61828-6), lithium monotherapy and lithium-plus-valproate combination therapy were more effective than valproate alone for the prespecified need for new intervention over follow-up, while the trial could not establish a reliable advantage of combination treatment over lithium monotherapy. The open-label design and outcome definition are essential to interpreting this result. Earlier Geddes, J. R., Burgess, S., Hawton, K., Jamison, K., & Goodwin, G. M. (2004) and network evidence summarized by Miura, T., Noma, H., Furukawa, T. A., Mitsuyasu, H., Tanaka, S., Stockto (2014)70314-1) support lithium's role in prevention across mood polarities, subject to safety and treatment suitability.
The contemporary evidence base has expanded substantially. The living umbrella review of De Prisco, M., Oliva, V., Miola, A., et al. (2026) synthesized 77 reviews across multiple treatment modalities, explicitly grading the certainty of evidence for phase-specific outcomes. It found that different interventions have different outcome profiles across acute depression, mania, and maintenance. The appropriate scientific conclusion is to compare absolute recurrence, harms, treatment acceptability, and preferences within the same clinical phase and population, rather than search for a universal ranking of all available treatments.
Lithium: Efficacy, Safety, and Monitoring
Lithium occupies an unusual position because its clinical use dates back decades and its maintenance evidence spans randomized trials, observational cohorts, and international guidelines. Its properties cannot be reduced to a single hypothesized molecular mechanism. The clinical evidence concerns prevention of mood recurrence, antimanic effects, and possible suicide-risk reduction; the mechanistic literature evaluates multiple intracellular and neurobiological pathways with incomplete causal resolution. National Institute for Health and Care Excellence. (2025) recommends lithium as a first-line long-term pharmacological option in the clinical contexts it covers.
Lithium also requires structured safety assessment. Renal and thyroid function, calcium, drug interactions, hydration, concurrent illness, toxicity symptoms, and blood concentrations are essential to medical management. Adverse effects can include tremor, gastrointestinal symptoms, thyroid dysfunction, renal complications, and neurological toxicity, with variation in reversibility and individual risk. The existence of a therapeutic laboratory range does not remove the need for individualized monitoring and interpretation. This research monograph does not reproduce prescribing doses or monitoring schedules as universal instructions because authoritative requirements depend on patient factors, the actual preparation, and jurisdiction-specific clinical protocols.
A persistent research issue is separating medication effectiveness from the experience of sustained care. People taking lithium successfully may have different adherence, service contact, illness history, and preferences from nonusers. Within-person registry methods partially reduce stable confounding, whereas randomized designs address allocation more directly but can be limited by tolerability, attrition, and event rarity. Converging evidence warrants attention without dismissing uncertainty in unusual outcomes or specific populations.
Anticonvulsants, Antipsychotics, and Comparative Harms
Valproate, lamotrigine, carbamazepine, and antipsychotic medications differ in their biological actions, supported treatment phases, regulatory permissions, and harm profiles. The term mood stabilizer is a clinical grouping that can conceal these distinctions. Lamotrigine's role in some maintenance and depression-related contexts does not imply evidence for treating acute mania. Valproate has evidence for certain manic and maintenance outcomes but carries major reproductive and other safety concerns. Antipsychotic drugs vary in efficacy, sedation, movement effects, metabolic consequences, and availability of long-acting formulations. The Yatham, L. N., Kennedy, S. H., Parikh, S. V., et al. (2018) and the Keramatian, K., Chithra, N. K., & Yatham, L. N. (2023) provide phase-oriented comparative frameworks.
Medical risks are not merely an appendix to efficacy. A drug producing strong mean improvement may be less acceptable where akathisia, weight gain, endocrine effects, reproductive concerns, or cognitive sedation are substantial. Trial discontinuation because of adverse events is a different outcome from discontinuation for any reason, and metabolic or renal harms can emerge outside the time window of acute efficacy trials. The clinical task therefore concerns benefit–harm trade-offs under appropriate monitoring, supported by a realistic understanding of both short- and long-term evidence.
The valproate case again shows why scientific and regulatory layers should be documented independently. Official Medicines and Healthcare products Regulatory Agency. (2025) address reproductive risk and restrictions applicable in the United Kingdom. A published trial demonstrating symptom efficacy cannot override such constraints, and guidance issued in another country may apply different license conditions. In scientific writing, claims about a treatment's efficacy, guideline position, and legal authorization must be sourced separately.
Psychological Treatments and Family-Focused Care
Psychological treatment in bipolar disorder addresses psychoeducation, relapse awareness, sleep and social routine regularity, medication-related decision support, symptom coping, family communication, and recovery goals. A psychosocial intervention can be valuable as an adjunct to medication, and its benefits may be expressed in time to recurrence, functional quality, treatment engagement, or reduced depressive symptoms. The therapeutic mechanisms are often multicomponent: education can support recognition of warning signs, while family work can reduce communication stress and increase coordinated help-seeking.
Randomized evidence includes the early Miklowitz, D. J., George, E. L., Richards, J. A., Simoneau, T. L., & Sud (2003) family psychoeducation program, the Rea, M. M., Tompson, M. C., Miklowitz, D. J., Goldstein, M. J., Hwang, S (2003) family-focused comparison, Frank, E., Kupfer, D. J., Thase, M. E., et al. (2005) on interpersonal and social rhythm therapy, and Miklowitz, D. J., Otto, M. W., Frank, E., et al. (2007) on psychosocial treatment of bipolar depression in STEP-BD. These studies have different comparison conditions and target populations. One trial's benefit on relapse cannot automatically be converted into another study's claim about acute depression response.
In their component network meta-analysis, Miklowitz, D. J., Efthimiou, O., Furukawa, T. A., et al. (2021) integrated randomized psychotherapy evidence and examined how intervention components relate to outcomes. Network designs allow wider comparison but depend on adequate similarity among trials. Individual and family preferences, intervention burden, service availability, therapist training, and cultural suitability affect real-world application. Psychological treatment should be evaluated by outcomes important to people living with the disorder, alongside standardized symptom scales.
Recovery, Physical Health, and Quality of Life
Recovery is a multidimensional process involving symptom stability, functioning, agency, housing, relationships, employment, education, and subjective well-being. The severity of an episode may be high while the person retains valued competencies and opportunities; conversely, a relatively stable mood score can coexist with major social loss. Rehabilitation and supported participation are therefore substantive elements of care research. World Health Organization. (2026) identifies structural barriers and stigma as contributors to unmet need and explicitly emphasizes recovery-oriented support.
Physical health is a major research and clinical priority. Cardiovascular and metabolic conditions, smoking, sleep disturbance, and inequities in preventive care intersect with both bipolar illness and medication exposure. Observed excess mortality cannot be assigned entirely to medication, illness biology, individual behavior, or health-system failures without careful causal analysis. Prospective research should model these processes together and assess whether integrated physical-health services improve measurable outcomes.
Quality-of-life measurement requires attention to who selects the outcome, the time horizon, and the construct being represented. Common symptom scales may underdetect improvements in self-determination, family stability, economic security, or participation. Patient-reported outcomes, objective role functioning, and qualitative studies can complement randomized clinical end points. A scientifically complete account of recovery asks whether interventions permit sustained lives the participants themselves consider meaningful.
Research Designs, Bias, and Evidence Hierarchies
Bipolar research relies on several complementary designs. Diagnostic field studies assess agreement and clinical utility; household surveys estimate prevalence under operational definitions; longitudinal cohorts describe trajectories and predictors; family and genomic studies examine susceptibility; laboratory studies interrogate intermediate processes; randomized trials test interventions; and systematic reviews compare accumulated outcomes. No design answers every question. Causal inference is strongest where exposure allocation and important biases are controlled, while representative epidemiological studies may be better suited to population frequency and service disparities.
Important biases include confounding by indication, immortal-time bias, nonrandom attrition, repeated use of the same cohorts in meta-analysis, outcome switching, selective reporting, enrichment of maintenance trials, and measurement of symptoms without functional outcomes. Studies with acute mania may exclude severe suicidality, pregnancy, serious comorbidity, or active substance use, even though these are clinically relevant populations. A claim of evidence-based treatment should therefore state which people and outcomes the evidence actually covers.
Comparative pharmacological syntheses face additional assumptions. Network meta-analysis combines direct and indirect evidence and depends on the plausibility of transitivity and consistency across trials. A network with many drugs, doses, clinical phases, and heterogeneous outcomes may produce deceptively precise rankings when certainty is low. De Prisco, M., Oliva, V., Miola, A., et al. (2026) explicitly graded a wide distribution of evidence certainty; this is a methodological strength because it prevents statistical significance from being mistaken for uniform reliability.
This monograph follows a transparent narrative method: evidence is grouped by research function, causal claims are contrasted with association, the clinical phase is retained when reporting treatments, and disputed findings are discussed with their methodological limitations. A prespecified systematic search across all databases, independent double screening, extraction tables, risk-of-bias adjudication for each included trial, and new quantitative pooling were not performed. The work is therefore a critical research monograph, with future systematic updates able to test and refine the synthesis.
Controversies: Spectrum Boundaries, Switching, and Treatment Heterogeneity
The boundaries of bipolarity are a persistent scientific controversy. Narrow operational diagnosis supports reproducibility but may miss clinically meaningful subthreshold activation. Broad-spectrum classifications may capture vulnerable populations while reducing specificity and increasing risk of overdiagnosis. The appropriate research response is to compare the predictive validity, reliability, treatment response, and patient consequences of alternative definitions rather than assume that the largest category must be the most explanatory. Longitudinal data are particularly important because early presentations often contain incomplete information.
Treatment-emergent switching is another disputed area. Reports of mania or hypomania following antidepressant exposure may reflect drug effects, spontaneous episode progression, confounding by clinical severity, or differences between bipolar subtypes and medications. Randomized and observational findings must be compared carefully, with attention to exposure timing, standardized switching definitions, concomitant mood-stabilizing treatment, and the duration of observation. The Sachs, G. S., Nierenberg, A. A., Calabrese, J. R., et al. (2007) provides one important controlled comparison but cannot establish a universal causal claim for every drug and clinical subgroup.
The idea of a single optimal treatment conflicts with the evidence that acute mania, bipolar depression, maintenance prevention, and recovery use distinct outcomes and mechanisms. Evidence hierarchies also evolve as new trials emerge. The De Prisco, M., Oliva, V., Miola, A., et al. (2026) illustrates the value of regularly updating comparative conclusions and making confidence visible. Future models must integrate heterogeneity without substituting individualized algorithmic scores for independently validated clinical outcomes.
Artificial Era: Longitudinal Data, Artificial Systems, and Care Research
In the Artificial Era, computational systems can participate in the collection, organization, and interpretation of longitudinal psychological information. Bipolar research is an especially demanding test because clinically meaningful evidence is distributed through time: episodes, prodromes, sleep patterns, medication changes, family observations, and functional consequences may be separated by months or years. Artificial systems can assist research teams in data harmonization, literature surveillance, temporal visualization, and transparent linkage between claims and studies. These capacities concern documented system performance and require validation against appropriate clinical standards.
The main methodological opportunity is person-specific time-series analysis combined with externally validated clinical assessment. A model might test whether a deviation from an individual's usual sleep or activity pattern predicts clinician-confirmed episode change beyond a simpler baseline. To establish usefulness, researchers need prospective cohorts, preregistered prediction windows, meaningful comparison models, uncertainty estimates, privacy-preserving data handling, calibration across populations, and direct evaluation of false alarms and missed episodes. More elaborate computation without these controls may create persuasive but clinically misleading outputs.
These questions connect bipolar psychology to research on human–artificial relations, cognitive offloading, trust calibration, and Artificial Agency. In this monograph, such terms identify interdisciplinary research directions rather than established mechanisms of bipolar disorder or proofs about the psychological status of artificial systems. A research participant's relationship with an application can influence engagement and reporting; the algorithm's diagnostic or therapeutic effectiveness must nevertheless be established empirically. Human care responsibility, informed consent, and transparent limits remain indispensable in the study of clinical decision support.
Unresolved Questions and a Discriminating Research Agenda
Several major questions remain unresolved. First, can integrated longitudinal measures distinguish vulnerability, prodrome, and active episode more accurately than current clinical information alone? The relevant studies should measure incremental predictive value, external validation, false-positive burden, and whether acting on alerts improves outcomes. Second, how do genetic susceptibility and circadian, social, and neurocognitive processes jointly contribute to episode transitions? This requires repeated within-person measurement linked to biological and contextual variables, rather than a static list of correlated risk factors.
Third, which therapies improve functional recovery and mortality outcomes beyond reducing short-term symptom scores? Trials should include psychosocial participation, adverse effects, physical health, and patient-valued recovery measures with appropriately long follow-up. Fourth, how can treatment evidence be generalized to populations commonly excluded from efficacy trials, including people with complex comorbidity, reproductive risks, and limited access to coordinated care? Pragmatic trials, registries, and transparent causal models can complement conventional controlled experiments.
Fifth, what governs apparent discrepancies between observational medication studies and randomized findings for rare outcomes such as suicide? Prospective designs should explicitly account for treatment discontinuation, indication, exposure adherence, outcome rarity, and time-varying confounding. Sixth, how should dimensional bipolar-spectrum models be tested against existing categories? Comparisons should emphasize prognosis, treatment utility, and reliability rather than conceptual appeal alone. The research agenda is unified by one requirement: clinical claims must survive testing against meaningful longitudinal outcomes.
Conclusion
Bipolar and related disorders are best understood scientifically through the organization of mood, activation, cognition, sleep, behavior, and functioning across time. Classification establishes recognizable episode patterns; longitudinal studies show the importance of depressive and subthreshold burden; genetic findings reveal broadly distributed susceptibility; psychosocial studies demonstrate the relevance of routine, family interaction, and access to support; and randomized treatment evidence specifies phase-dependent benefits and risks. The most informative account connects these domains while preserving the distinction among association, plausible mechanism, causal evidence, and clinical recommendation.
The present synthesis supports several durable conclusions. A lifetime diagnosis requires longitudinal evaluation and differential reasoning. Bipolar I and bipolar II have different defining episode patterns, while both may carry profound disability. Subthreshold symptoms, physical health, cognitive participation, and social circumstances are central research outcomes. Maintenance treatment, including lithium where appropriate, rests on meaningful evidence that must be balanced against monitoring and safety obligations. Psychological interventions contribute to relapse management and recovery, while treatment decisions remain specific to episode phase, patient circumstances, and relevant regulatory context.
Future progress depends on studies capable of tracking transitions, testing mechanisms prospectively, validating treatment effects across diverse populations, and measuring outcomes that matter beyond symptom remission. Artificial systems may improve research infrastructure and monitoring, provided their clinical utility and safety are demonstrated rather than presumed. The scientific ambition is a more precise, globally applicable, longitudinal psychology of bipolar illness that remains accountable to original evidence and to the lives it seeks to understand.
References
- Cade, J. F. J. (1949). Lithium salts in the treatment of psychotic excitement. Medical Journal of Australia, 2(10), 349–352. https://doi.org/10.5694/j.1326-5377.1949.tb36912.x
- De Prisco, M., Oliva, V., Miola, A., et al. (2026). Evidence based interventions for bipolar disorder across phases and age groups: Living umbrella review, evaluation, analysis, and communication hub (U-REACH) project. BMJ, 394, e100216. https://doi.org/10.1136/bmj-2026-100216
- Fornaro, M., Di Lorenzo, C., Oliva, V., De Prisco, M., & Vieta, E. (2026). Efficacy, acceptability, and related outcomes of pharmacological interventions for acute bipolar depression: A systematic review and dose-related network meta-analysis across different age groups. Journal of Affective Disorders, 412, 122160. https://doi.org/10.1016/j.jad.2026.122160
- Frank, E., Kupfer, D. J., Thase, M. E., et al. (2005). Two-year outcomes for interpersonal and social rhythm therapy in individuals with bipolar I disorder. Archives of General Psychiatry, 62(9), 996–1004. https://doi.org/10.1001/archpsyc.62.9.996
- Frank, E., Swartz, H. A., & Kupfer, D. J. (2000). Interpersonal and social rhythm therapy: Managing the chaos of bipolar disorder. Biological Psychiatry, 48(6), 593–604. )00969-0 https://doi.org/10.1016/S0006-3223(00
- Geddes, J. R., Burgess, S., Hawton, K., Jamison, K., & Goodwin, G. M. (2004). Long-term lithium therapy for bipolar disorder: Systematic review and meta-analysis of randomized controlled trials. American Journal of Psychiatry, 161(2), 217–222. https://doi.org/10.1176/appi.ajp.161.2.217
- Hirschfeld, R. M. A., Williams, J. B. W., Spitzer, R. L., et al. (2000). Development and validation of a screening instrument for bipolar spectrum disorder: The Mood Disorder Questionnaire. American Journal of Psychiatry, 157(11), 1873–1875. https://doi.org/10.1176/appi.ajp.157.11.1873
- Judd, L. L., Akiskal, H. S., Schettler, P. J., et al. (2002). The long-term natural history of the weekly symptomatic status of bipolar I disorder. Archives of General Psychiatry, 59(6), 530–537. https://doi.org/10.1001/archpsyc.59.6.530
- Judd, L. L., Akiskal, H. S., Schettler, P. J., et al. (2003). A prospective investigation of the natural history of the long-term weekly symptomatic status of bipolar II disorder. Archives of General Psychiatry, 60(3), 261–269. https://doi.org/10.1001/archpsyc.60.3.261
- Keramatian, K., Chithra, N. K., & Yatham, L. N. (2023). The CANMAT and ISBD guidelines for the treatment of bipolar disorder: Summary and a 2023 update of evidence. Focus, 21(4), 344–353. https://doi.org/10.1176/appi.focus.20230009
- Medicines and Healthcare products Regulatory Agency. (2025). Valproate—Reproductive risks. UK Government. https://www.gov.uk/guidance/valproate-reproductive-risks
- Merikangas, K. R., Jin, R., He, J.-P., et al. (2011). Prevalence and correlates of bipolar spectrum disorder in the world mental health survey initiative. Archives of General Psychiatry, 68(3), 241–251. https://doi.org/10.1001/archgenpsychiatry.2011.12
- Miklowitz, D. J., Axelson, D. A., Birmaher, B., et al. (2008). Family-focused treatment for adolescents with bipolar disorder: Results of a 2-year randomized trial. Archives of General Psychiatry, 65(9), 1053–1061. https://doi.org/10.1001/archpsyc.65.9.1053
- Miklowitz, D. J., Efthimiou, O., Furukawa, T. A., et al. (2021). Adjunctive psychotherapy for bipolar disorder: A systematic review and component network meta-analysis. JAMA Psychiatry, 78(2), 141–150. https://doi.org/10.1001/jamapsychiatry.2020.2993
- Miklowitz, D. J., George, E. L., Richards, J. A., Simoneau, T. L., & Suddath, R. L. (2003). A randomized study of family-focused psychoeducation and pharmacotherapy in the outpatient management of bipolar disorder. Archives of General Psychiatry, 60(9), 904–912. https://doi.org/10.1001/archpsyc.60.9.904
- Miklowitz, D. J., Otto, M. W., Frank, E., et al. (2007). Psychosocial treatments for bipolar depression: A 1-year randomized trial from the Systematic Treatment Enhancement Program. Archives of General Psychiatry, 64(4), 419–426. https://doi.org/10.1001/archpsyc.64.4.419
- Miura, T., Noma, H., Furukawa, T. A., Mitsuyasu, H., Tanaka, S., Stockton, S., Salanti, G., Motomura, K., Shimano-Katsuki, S., Leucht, S., Cipriani, A., Geddes, J. R., & Kanba, S. (2014). Comparative efficacy and tolerability of pharmacological treatments in the maintenance treatment of bipolar disorder: A systematic review and network meta-analysis. The Lancet Psychiatry, 1(5), 351–359. )70314-1 https://doi.org/10.1016/S2215-0366(14
- Nabi, Z., Stansfeld, J., Plöderl, M., Wood, L., & Moncrieff, J. (2022). Effects of lithium on suicide and suicidal behaviour: A systematic review and meta-analysis of randomised trials. Epidemiology and Psychiatric Sciences, 31, e65. https://doi.org/10.1017/S204579602200049X
- National Institute for Health and Care Excellence. (2025). Bipolar disorder: Assessment and management (CG185; updated September 2, 2025). https://www.nice.org.uk/guidance/cg185/chapter/Recommendations
- National Institute of Mental Health. (2024). Bipolar disorder. https://www.nimh.nih.gov/health/topics/bipolar-disorder
- National Institute of Mental Health. (n.d.). Bipolar disorder: Statistics. https://www.nimh.nih.gov/health/statistics/bipolar-disorder
- O'Connell, K. S., Koromina, M., van der Veen, T., et al. (2025). Genomics yields biological and phenotypic insights into bipolar disorder. Nature, 639, 968–975. https://doi.org/10.1038/s41586-024-08468-9
- Rea, M. M., Tompson, M. C., Miklowitz, D. J., Goldstein, M. J., Hwang, S., & Mintz, J. (2003). Family-focused treatment versus individual treatment for bipolar disorder: Results of a randomized clinical trial. Journal of Consulting and Clinical Psychology, 71(3), 482–492. https://doi.org/10.1037/0022-006X.71.3.482
- Sachs, G. S., Nierenberg, A. A., Calabrese, J. R., et al. (2007). Effectiveness of adjunctive antidepressant treatment for bipolar depression. New England Journal of Medicine, 356(17), 1711–1722. https://doi.org/10.1056/NEJMoa064135
- Song, J., Sjölander, A., Joas, E., Bergen, S. E., Runeson, B., Larsson, H., Landén, M., & Lichtenstein, P. (2017). Suicidal behavior during lithium and valproate treatment: A within-individual 8-year prospective study of 50,000 patients with bipolar disorder. American Journal of Psychiatry, 174(8), 795–802. https://doi.org/10.1176/appi.ajp.2017.16050542
- The BALANCE investigators and collaborators. (2010). Lithium plus valproate combination therapy versus monotherapy for relapse prevention in bipolar I disorder (BALANCE): A randomised open-label trial. The Lancet, 375(9712), 385–395. )61828-6 https://doi.org/10.1016/S0140-6736(09
- U.S. Department of Veterans Affairs & U.S. Department of Defense. (2023). VA/DoD clinical practice guideline for the management of bipolar disorder. https://healthquality.va.gov/guidelines/MH/bd/
- Wang, J. X., Le, G. H., Wong, S., Teopiz, K. M., Kwan, A. T. H., Rosenblat, J. D., Rhee, T. G., Ho, R., Lo, H. K. Y., Goldberg, J. F., Vinberg, M., Grande, I., Mansur, R., Meyer, J. M., & McIntyre, R. S. (2025). The efficacy of lithium in the treatment of suicidal ideation, behavior and suicide: An updated systematic review and meta-analysis of randomized controlled trials. Journal of Affective Disorders, 387, 119487. https://doi.org/10.1016/j.jad.2025.119487
- World Health Organization. (2024). Clinical descriptions and diagnostic requirements for ICD-11 mental, behavioural and neurodevelopmental disorders. https://www.who.int/publications/i/item/9789240077263
- World Health Organization. (2026). Bipolar disorder (Fact sheet, September 11, 2026). https://www.who.int/news-room/fact-sheets/detail/bipolar-disorder/
- Yalin, N., Yildiz, A., Siafis, S., Vieta, E., & Leucht, S. (2026). Treatment of bipolar depression: Results from a comprehensive network meta-analysis and updated systematic review. European Neuropsychopharmacology, 108, 112818. https://doi.org/10.1016/j.euroneuro.2026.112818
- Yatham, L. N., Kennedy, S. H., Parikh, S. V., et al. (2018). Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) 2018 guidelines for the management of patients with bipolar disorder. Bipolar Disorders, 20(2), 97–170. https://doi.org/10.1111/bdi.12609
- Young, R. C., Biggs, J. T., Ziegler, V. E., & Meyer, D. A. (1978). A rating scale for mania: Reliability, validity and sensitivity. British Journal of Psychiatry, 133, 429–435. https://doi.org/10.1192/bjp.133.5.429