Bipolar Disorder and the Brain: What Neuroscience Can and Cannot Explain
Author: Ukrainian Psychological Hub · Published: October 9, 2026 · Editorial Policy
Bipolar disorder involves changes in mood, energy, activity, sleep, and behavior that occur in clinically significant episodes. Neuroscience has identified reproducible differences in brain structure, activation, connectivity, and sleep–wake regulation among groups of people with bipolar disorder. These findings strengthen the scientific understanding of the condition. They do not yet provide a brain scan, neurotransmitter measurement, or other stand-alone biological test capable of diagnosing bipolar disorder in an individual. The National Institute of Mental Health makes this distinction between population research and current clinical assessment.
The most useful scientific question is not which single part of the brain “causes” bipolar disorder. It is how multiple biological processes contribute to vulnerability and the expression of manic, hypomanic, depressive, and mixed episodes; how those processes interact with sleep, medication, environment, and life history; and which conclusions research can support. Large international imaging collaborations have helped answer parts of that question while demonstrating why simplistic explanations fail. See the ENIGMA Bipolar Disorder neuroimaging review and the 2026 ENIGMA medication–brain volume study.
For additional clinical context, see Bipolar Disorder Recovery and Remission: Stability, Recurrence, and Outlook.
What Does “Bipolar Disorder in the Brain” Mean?
Bipolar disorder is a clinical diagnosis describing a pattern of mood episodes over time. Bipolar I disorder requires a manic episode; bipolar II disorder involves hypomanic and major depressive episodes without a history of full mania; cyclothymic disorder has its own enduring pattern of subthreshold symptoms. Mania may involve substantially increased activation, a decreased need for sleep, expansive or irritable mood, impaired judgment, and sometimes psychosis. Hypomania is a distinct episode type with a different severity and impairment threshold. Depressive episodes bring another symptom profile. For detailed clinical classification and diagnostic distinctions, see the NIMH overview and NICE guideline CG185.
When scientists speak of bipolar disorder and the brain, they may be investigating anatomical features, patterns of blood flow during a task, synchronized activity between networks, electrical rhythms, genetic associations, molecular pathways, or the brain's response to treatment. These measurements address different questions. An MRI finding associated with a diagnosis does not reveal what a person is thinking, determine whether the person is currently manic, or establish what caused an episode. ENIGMA review
The brain is also dynamic. Sleep deprivation, stress, physical illness, substance use, medication exposure, mood state, and normal aging can influence measured brain features. A person may have pronounced symptoms with an ordinary clinical scan, while someone without bipolar disorder may show a particular imaging feature reported in a bipolar research sample. The goal is to understand probabilities and mechanisms without treating group averages as individual destinies.
What Brain-Scanning Methods Actually Measure
Structural MRI: anatomy rather than mental state
Structural magnetic resonance imaging measures features such as cortical thickness, cortical surface area, and the volume of specific brain regions. Cortical thickness is a measurement of the brain's outer gray matter, typically measured in millimeters. A finding that one research group has a slightly thinner cortex on average does not establish that any particular participant has injured, diseased, or deteriorating brain tissue. The interpretation depends on age, medication, clinical history, scanner methods, and overlap between groups. Hibar et al., ENIGMA cortical analysis
Functional MRI: task-related and resting signals
Functional MRI generally uses a blood-oxygen-level-dependent signal as an indirect marker related to local neural activity. Researchers compare responses to emotional faces, rewards, cognitive demands, or conditions of rest. The signal depends on how the task is constructed and how data are processed. Finding a mean difference in amygdala activation during an emotion task is not evidence that every person with bipolar disorder has an overactive amygdala at all times. Schumer et al. functional neuroimaging meta-analysis
Diffusion MRI: white matter microstructure
Diffusion-based imaging estimates properties of water movement within brain tissue. A commonly reported metric, fractional anisotropy, reflects the directionality of diffusion and is influenced by several tissue characteristics. It is an indirect imaging measure; it is not a direct count of functioning connections or proof that a tract is broken. Large studies find statistical differences in some white matter measurements in bipolar disorder, but these remain research findings. Favre et al., ENIGMA white matter analysis
EEG, sleep recordings, and molecular research
Electroencephalography records electrical activity detectable at the scalp. Sleep studies and actigraphy evaluate aspects of sleep physiology and behavioral timing. Blood, genetic, and other molecular measurements investigate biological processes associated with vulnerability or treatment. None of these techniques independently establishes a bipolar diagnosis. In clinical practice, tests may be useful for evaluating a different medical problem, medication safety, or a coexisting sleep condition. NIMH diagnostic overview.
Which Brain Regions Have Researchers Studied?
Prefrontal cortex and cognitive control
The prefrontal cortex contributes to planning, response selection, evaluating consequences, attention, and flexible regulation of emotion. It operates through networks rather than in isolation. Structural imaging collaborations have observed modest average cortical differences across parts of frontal, temporal, and parietal cortex in adults with bipolar disorder. Functional research also reports differences in inferior frontal and orbitofrontal responses under particular experimental conditions. These patterns are relevant to scientific models of regulation and decision-making, but the imaging data cannot identify the cause of one person's impulsivity or explain every manic decision. Hibar et al.; Schumer et al..
Amygdala and emotional salience
The amygdala participates in detecting and learning about emotionally significant events. It responds to a wide range of experiences, including rewarding, threatening, novel, and personally important stimuli. A systematic review and meta-analysis of functional imaging reported differences in emotion-processing circuits, including the amygdala and frontal regions, among groups with bipolar disorder. Another large meta-analysis found an amygdala-related signal in emotional-task studies while emphasizing variation by experimental condition. The expression “the amygdala is switched on” compresses complex, task-dependent findings into a misleading metaphor. Frontolimbic network meta-analysis; Schumer et al..
Anterior cingulate, insula, and salience processing
The anterior cingulate cortex and insula participate in processing internal bodily signals, salience, conflict, and aspects of emotion and cognitive control. They interact with many other areas. A 2022 systematic review and meta-analysis reported lower cortical thickness in selected frontoinsular and cingulate areas among bipolar samples. Its authors also noted that included imaging studies were cross-sectional, limiting conclusions about why those differences developed. Cortical thickness systematic review
Hippocampus, memory, and contextual learning
The hippocampus supports memory formation and contextual learning. Its volume has been investigated repeatedly in bipolar disorder. Findings depend importantly on treatment history, particularly lithium. An earlier meta-analysis found different hippocampal-volume patterns in lithium-exposed versus non-lithium groups, while acknowledging uncontrolled clinical differences. A large 2026 ENIGMA analysis again showed that medication exposure complicates comparisons of subcortical volumes. It is therefore inaccurate to conclude that bipolar disorder always shrinks the hippocampus, or that an individual MRI can document how much memory capacity a person has lost. Hajek et al.; King et al. 2026.
Striatum and reward-related learning
The striatum participates in motivation, learning, selection of actions, and reward processing, in cooperation with cortical and other subcortical systems. Functional imaging meta-analyses report differences in reward-related activation in some bipolar samples. These results suggest plausible routes for investigating changes in goal-directed activity during episodes. They do not imply that bipolar disorder can be explained as “too much dopamine,” excessive pleasure seeking, or a single reward-center defect. Frontolimbic network meta-analysis
What the Largest Structural MRI Studies Found
The ENIGMA Bipolar Disorder Working Group combines data from multiple countries and research centers using standardized analysis procedures. Its cortical study examined thousands of MRI scans and reported thinner cortical gray matter on average across parts of the frontal, temporal, and parietal lobes among adults with bipolar disorder. Some reported standardized effects were approximately 0.3, illustrating measurable but overlapping group distributions. It found no significant case–control differences in the cortical surface-area regions it examined and no significant cortical thickness or surface-area differences between bipolar I and bipolar II subgroups in that sample. Hibar et al., Molecular Psychiatry
Three aspects of this result matter. First, cortical thickness and surface area are different quantities, and they need not change together. Second, even robust average differences permit substantial overlap between people with and without the diagnosis. Third, associations involving disease duration, symptoms, or medication cannot automatically establish causation. Longer illness duration was associated with lower cortical thickness in some regions in the same study, but cross-sectional associations cannot tell whether illness caused the difference, whether the difference preceded illness, or what other factors contributed.
The ENIGMA review places structural findings alongside evidence for subcortical and white matter differences, highlighting how multisite replication improves credibility. A cross-disorder review found overlap among patterns reported in bipolar disorder, schizophrenia, and major depressive disorder. Brain features associated with bipolar disorder are therefore not necessarily specific to it. ENIGMA review; Cross-disorder ENIGMA comparison.
White matter: communication pathways in research
In a 2019 ENIGMA mega-analysis of 3,033 people, including 1,482 with bipolar disorder, investigators found lower fractional anisotropy in numerous white matter regions, with prominent average differences in the corpus callosum and cingulum. The findings indicate a reproducible group-level association with white matter microstructural measures. They do not establish disconnection in a particular brain or permit clinical diagnosis based on one person's white matter scan. Favre et al., Neuropsychopharmacology
Brain Networks and Mood Regulation
Emotion involves coordinated activity among neural networks implicated in salience, self-referential processing, reward, memory, attention, and flexible action. Functional studies often focus on interactions between frontally mediated regulation and limbic responses. Yet these networks are not anatomical switches for “mania” or “depression.” Their activation varies across tasks, people, and phases of illness.
A large 2023 meta-analysis synthesized 205 studies containing 506 experiments, with 5,745 bipolar participants and 8,023 controls across included experiments. It identified reproducible functional differences in limited regions depending on whether participants were resting, processing emotion, or performing cognitive tasks, including posterior cingulate, amygdala, parietal, and medial orbitofrontal areas. The dependence on task type is part of the result: different experiments test different mechanisms. Schumer et al., Molecular Psychiatry
A separate systematic review and meta-analysis identified patterns involving amygdala, hippocampus, inferior frontal gyrus, and orbitofrontal areas across emotion, reward, and working-memory experiments. It suggested that some differences may be detected even during euthymia, a period of relative mood stability, but such group results cannot establish a permanent biological state for every individual. Both reviews emphasized heterogeneity. Frontolimbic network systematic review
Does the brain function differently in mania, depression, and euthymia?
Mood state matters. Mania, bipolar depression, and euthymia differ clinically; research tasks may capture different activation profiles within these states. A person with bipolar disorder is not necessarily showing the same patterns during all phases. Studies frequently combine different diagnoses, medications, episode histories, and states, which can obscure meaningful distinctions. The state-specific brain mechanisms of mania and bipolar depression remain active research questions, and neuroscientific associations should not substitute for clinical episode criteria. Schumer et al.; NICE CG185.
Can research distinguish bipolar I from bipolar II in the brain?
The diagnoses are distinguished clinically by lifetime mood-episode history and severity criteria. The large ENIGMA cortical study did not find significant subtype differences in the cortical measures it tested. This finding does not mean the diagnoses are neurobiologically identical; it means that those particular MRI measures did not reliably separate the groups in that analysis. Bipolar II also cannot be described as simply a mild form of bipolar I: its burden of depression and impairment can be substantial. Hibar et al.; NIMH.
A 2026 systematic review examining 38 neuroimaging studies reported candidate structural and functional differences between bipolar I and bipolar II, particularly in emotion-regulation and reward-related circuits. The findings are heterogeneous across methods, samples, and mood states, and the authors describe potential subtype-related neural signatures rather than a validated individual diagnostic classifier. This newer synthesis complements the earlier ENIGMA result: different studies assess different brain features, and emerging subgroup associations do not overturn the established clinical criteria. Miola et al., 2026 systematic review.
Circadian Rhythms, Sleep, and the Bipolar Brain
The brain regulates sleep and wakefulness through interacting circadian and homeostatic processes. Circadian timing coordinates daily patterns of alertness, hormonal activity, body temperature, and behavior. Sleep changes are particularly relevant to bipolar disorder because reduced need for sleep is a characteristic symptom of mania or hypomania, and changes in sleep may precede mood episodes. This relationship is important without reducing bipolar disorder to a sleep problem. NIMH; Ulrichsen et al. systematic review.
A 2025 systematic review examining whether sleep variables predict mood synthesized 17 publications from 13 studies. Sleep changes sometimes preceded shifts toward manic or depressive mood, but variation across study designs prevented quantitative meta-analysis. Another 2025 systematic review and meta-analysis found differences in sleep features across euthymic, depressive, and manic/mixed phases and reported poor sleep quality during euthymia in a substantial share of participants. Neither analysis proves that every insomnia episode causes mania or that an individual's future episode can be predicted precisely from a wearable device. Ulrichsen et al.; Marchetti et al..
For a person living with bipolar disorder, tracking a change from their own usual sleep pattern may be clinically helpful when integrated into an agreed relapse-prevention plan. A reduced need for sleep accompanied by rising energy, accelerated speech, agitation, or unusual confidence deserves timely discussion with a clinician. For severe activation, psychosis, or unsafe behavior, urgent assessment may be necessary. Sleep routines are one component of care; they do not replace evidence-based treatment. NICE relapse-planning recommendations.
Neurotransmitters: Dopamine, Serotonin, Glutamate, and GABA
Neurotransmitters allow nerve cells to communicate through multiple receptor types and regulatory pathways. Dopamine is involved in learning, salience, motivation, and movement; serotonin participates in diverse signaling systems; glutamate is a major excitatory transmitter; and GABA is a major inhibitory transmitter. These systems interact with each other and with sleep, endocrine regulation, cellular signaling, and neuroplasticity. Their relevance makes them plausible subjects of bipolar research, but a specific “chemical imbalance” formula is not an established diagnostic explanation for the condition.
Claims such as “bipolar mania means high dopamine” or “bipolar depression means low serotonin” exceed what current clinical neuroscience can verify for a particular patient. Medication response alone does not demonstrate an underlying neurotransmitter deficiency or excess. A medication can affect multiple receptors and pathways, and its benefits may depend on a person's illness phase, treatment history, and other conditions. Contemporary clinical guidelines recommend treatments according to clinical outcomes and safety evidence, rather than neurotransmitter test results. VA/DoD guideline (2023); CANMAT/ISBD guideline update.
What about inflammation and oxidative stress?
Researchers also investigate immune signaling, inflammatory markers, energy metabolism, oxidative stress, mitochondrial biology, and neuroplasticity. These candidate mechanisms may eventually clarify differences among bipolar presentations, but inflammatory markers are not presently validated as stand-alone diagnostic tests for bipolar disorder. A raised laboratory value may reflect many causes, including infection, metabolic disease, medication effects, or other conditions. Mechanistic hypotheses should be distinguished from an established clinical test or proven cause.
Does Bipolar Disorder Damage the Brain?
The phrase “brain damage” usually implies injury to tissue or a clinically established neurological lesion. Bipolar disorder research identifies statistical structural and functional associations; it does not justify telling every person with the diagnosis that their brain is being irreversibly damaged by each episode. Some imaging studies associate longer illness duration or repeated episodes with structural measures, and longitudinal research has examined whether changes emerge over time. Such observations deserve attention, but the causal story is considerably more complicated than a claim of inevitable deterioration. Hibar et al.; Longitudinal imaging systematic review.
A 2023 systematic review of longitudinal imaging included 11 studies, 329 people with bipolar disorder and 277 controls, and reported heterogeneous changes across studies. Another review evaluating mania-associated structural changes described evidence linking episodes with some prefrontal cortical changes while also discussing stability or increases in some brain measures during periods of euthymia. Taken together, the evidence supports further study of illness burden and brain development, not deterministic predictions of irreversible neuronal loss for an individual. Longitudinal systematic review; Mania-related imaging review.
Neuroprogression is a research hypothesis, not a personal forecast
Neuroprogression refers to hypotheses about changes in brain and clinical functioning across illness course. Researchers must distinguish an effect of illness from preexisting differences, aging, cumulative medication exposure, coexisting medical conditions, substance use, and selection effects. People also differ in treatment access, number and severity of episodes, physical health, and recovery conditions. Cross-sectional snapshots cannot establish the sequence of events. Longitudinal studies improve inference, but existing samples are often small or clinically heterogeneous. The individual course remains a clinical and personal question requiring longitudinal evaluation.
Can stability and effective treatment matter for the brain?
Reducing recurrence, improving sleep regularity, addressing coexisting conditions, and treating episodes promptly are important clinical goals. Evidence-based care may improve symptoms, functioning, and safety. These benefits do not require promising to “reverse brain shrinkage.” Long-term follow-up should examine the person's mood, functioning, treatment effects, physical health, and priorities rather than treating MRI volume as a substitute for quality of life. NICE CG185; VA/DoD guideline.
Is Bipolar Disorder Genetic or Caused by Brain Differences?
Family and genetic studies support substantial inherited contributions to bipolar disorder, but susceptibility is distributed across many genetic variants, and genetic risk does not determine an individual's outcome. Brain differences may reflect combinations of inherited vulnerability, development, life events, current mood state, treatment exposure, and other factors. It would be a mistake to treat a structural difference as proof of a particular genetic cause. The NIMH overview describes a multifactorial picture of genetic and environmental influences.
The scientific problem has several directions of causation: an early biological difference might increase vulnerability; episodes or chronic stress might influence a measured feature; medication might modify the same measure; or a third variable might affect both symptoms and imaging. Many studies cannot discriminate among these explanations. Detailed discussion of risk factors and family history belongs to the distinct English Hub articles What Causes Bipolar Disorder? and Is Bipolar Disorder Genetic?.
Do Medications Change the Brain?
Psychiatric medications can influence neural signaling, behavior, sleep, and sometimes brain measures used in research. This fact complicates interpretation of MRI studies: treated and untreated populations may differ in illness severity, episode history, health status, and treatment selection. An association between a medicine and a structural measurement cannot by itself establish that the medication caused tissue harm or protection.
Lithium and MRI findings
Studies have reported associations between lithium exposure and greater cortical thickness or hippocampal volume in certain samples. The ENIGMA cortical study reported higher cortical thickness in some regions among lithium-treated participants. A meta-analysis focusing on hippocampus volume also found lithium-associated differences. These findings support research into possible neurobiological effects, but the designs cannot definitively establish neuroprotection for each individual or replace clinical outcome evidence. Hibar et al.; Hajek et al..
Lithium remains an evidence-based medication for selected phases of bipolar disorder, with particularly important long-term maintenance uses in guidelines. It has a narrow therapeutic range and requires professionally managed monitoring, including blood levels and attention to kidney and thyroid function, interactions, dehydration, and toxicity symptoms. Decisions about initiation, continuation, or dose changes belong with the prescriber. No one should adjust lithium because of a generalized MRI claim. NICE lithium and monitoring recommendations.
Antipsychotics, antiseizure medications, and medication confounding
A 2026 ENIGMA mega-analysis pooled 2,664 bipolar participants and 4,065 controls from 34 research groups. It investigated subcortical brain volumes and medication associations, including combined exposure. Some medication categories correlated with smaller measured volumes in selected regions, while lithium modified certain associations. The results were observational, with potential confounding by illness and prescribing patterns, and the investigators called for longitudinal research. The study cannot be used to conclude that an antipsychotic or valproate causes brain damage in a particular individual. King et al., Molecular Psychiatry (2026).
The safety profile of each medication must be evaluated separately. Antipsychotic treatment can carry metabolic, movement-related, and other risks requiring appropriate monitoring. Valproate carries serious pregnancy-related reproductive risks and additional adverse effects. Lamotrigine carries a warning concerning potentially serious rash. Antidepressants require bipolar-specific evaluation because of possible switching and other clinical considerations. These distinctions are covered by phase-specific treatment guidelines; neuroimaging studies do not determine medication suitability. CANMAT/ISBD update; NICE CG185.
Can an MRI or Brain Scan Diagnose Bipolar Disorder?
No validated clinical brain scan can confirm or exclude bipolar disorder on its own. Imaging researchers can classify groups above chance under some experimental conditions, but individual clinical diagnosis demands sufficiently accurate, reproducible, and specific performance in real-world patients, including those with overlapping conditions. A result that separates research volunteers with bipolar disorder from healthy controls may perform differently when comparing bipolar depression with unipolar depression, ADHD, psychotic disorders, neurological illness, or medication effects.
A 2025 systematic review and meta-analysis of T1-weighted MRI diagnostic performance included 15 studies and 1,645 participants. It characterized gray-matter-volume classification as having moderate accuracy, with important variation by age, sex, analytical method, and diagnostic comparison group. Overlap with other psychiatric conditions and methodological heterogeneity limited immediate stand-alone clinical utility. This independent synthesis reinforces the distinction between promising statistical classification and a clinically validated diagnostic test. González-Ojeda et al., 2025 meta-analysis.
One ENIGMA machine-learning study used structural MRI data from 853 bipolar participants and 2,167 controls across 13 sites. Its pooled classification accuracy was about 65.2%, with performance varying between sites. That is meaningful research but insufficient for a dependable stand-alone clinical diagnostic test. Even higher accuracies in selected datasets require external validation and attention to class balance, false positives, false negatives, and the population in which a test will be used. Nunes et al., Molecular Psychiatry.
Clinicians diagnose bipolar disorder through careful assessment of distinct episodes across a person's lifetime, including duration, severity, changes in energy and sleep, functional consequences, family history, medications, substance use, medical conditions, and alternative explanations. Laboratory tests or imaging may be indicated to investigate suspected medical or neurological conditions; they are not substitutes for the longitudinal psychiatric assessment. NICE CG185 assessment recommendations; NIMH.
Can an AI algorithm read bipolar disorder from a scan?
Machine-learning tools can recognize statistical patterns in data, but a published model is not the same thing as a clinically validated diagnostic instrument. Algorithms may learn patterns associated with scanner site, age, medication use, recruitment procedures, or diagnosis labels. Independent external validation, transparent thresholds, prospective studies, and evidence that use improves patient outcomes are needed before clinical implementation. A consumer-facing “AI brain test” claiming to diagnose bipolar disorder from an image should not be treated as established care. Nunes et al..
When might a clinician order an MRI?
Imaging can be appropriate when a clinician is concerned about an underlying neurological condition, unusual neurological findings, a new or atypical presentation, or another medical indication. The decision is guided by history and examination, not by a general requirement that everyone with bipolar disorder obtain a scan. A normal MRI does not rule out bipolar disorder, and an incidental imaging finding does not confirm it. NIMH diagnostic guidance.
How to Read Bipolar Brain Research Without Being Misled
Group averages are different from individual predictions
A group difference tells us that two measured distributions differ on average. It does not mean the distributions are separate. Many people with bipolar disorder may fall within the range of measured values found in controls. A small effect can be scientifically reproducible while remaining unhelpful for individual classification. Strong headlines about “the bipolar brain” often leave out the overlap and uncertainty.
Correlation is not a causal explanation
A finding that an MRI measure correlates with years since diagnosis or with taking a medication may be influenced by many other variables. Identifying cause requires stronger designs, repeated measurements, careful adjustment, and, where feasible, experimental evidence. For medication studies, illness severity may affect prescribing and also relate to imaging differences. The 2026 ENIGMA analysis explicitly underscores the complexity of medication associations. King et al.
Clinical heterogeneity matters
The label bipolar disorder includes people with different episode histories, age at onset, medication exposures, coexisting conditions, and patterns of recovery. Some studies combine bipolar I and II, current and remitted symptoms, or younger and older adults. Results from one subgroup should not silently become claims about all individuals or about children, pregnant patients, or older adults. Specialized clinical evidence is needed for these populations. ENIGMA review; NICE CG185.
Replication, sample size, and independent validation matter
Single small studies can generate useful hypotheses, but results are more credible when repeated using independent populations and sufficiently similar measures. Even very large cross-sectional collaborations remain limited in causal inference. Prospective within-person studies may be especially informative because they can compare someone with their own baseline across mood states. Clinical translation then requires proof that a measurement changes a meaningful decision or outcome.
Can Brain Research Explain Symptoms People Notice?
Neuroscience provides plausible frameworks for understanding shifts in emotion, cognition, motivation, and sleep. It cannot assign a unique region or neurotransmitter to an individual's experience. A period of unusually high energy may reflect a manic or hypomanic episode, but could also have other causes; persistent concentration problems might relate to symptoms, a coexisting condition, medication effects, insufficient sleep, or something else. Clinical interpretation requires the history of timing and change from baseline.
For many patients and families, a more useful practical question is how to distinguish an individual's recognizable early warning pattern from ordinary variation. Changes in sleep, speech, activity, spending, irritability, or energy can be tracked over time and reviewed collaboratively with a clinician. This is consistent with recommended relapse planning. A personalized clinical history often yields more actionable information than a speculative consumer neuroimaging interpretation. NICE relapse prevention.
Treatment: What Brain Science Can and Cannot Tell Us
Evidence-based management is guided by clinical trials, systematic reviews, phase-specific guidelines, and shared decision-making. Acute mania, bipolar depression, and maintenance treatment require distinct efficacy and safety judgments. The existence of a plausible neural mechanism does not prove that a treatment works, nor does a change on an fMRI signal guarantee improved functioning or fewer episodes. VA/DoD 2023 guideline; CANMAT/ISBD update.
Medication selection depends on current episode, past response, coexisting conditions, pregnancy-related considerations where relevant, drug interactions, laboratory monitoring, tolerability, and personal preferences. Psychotherapies and structured psychosocial support may help with illness understanding, relapse prevention, routines, relationships, and recovery goals as part of comprehensive care. Monitoring physical health is also essential. The dedicated English Hub Bipolar Disorder Treatment guide covers the treatment evidence in greater detail.
Practical Implications for Someone Living With Bipolar Disorder
It is reasonable to ask a clinician what a test is expected to rule out, what information is needed to understand the pattern of mood episodes, and which treatment outcomes matter most. A brain scan is not a measure of character, intelligence, prognosis, or personal worth. A statistical finding about a research group should not be used to predict someone's capacity for work, relationships, parenting, learning, or recovery.
A longitudinal account is often useful: dates of sustained shifts in energy and mood, changes in need for sleep, consequences for functioning, medication exposure, significant stressors, and prior episodes. With the person's permission, observations from trusted relatives or friends may help clarify episodes that were difficult to recognize at the time. A clinician can combine this history with medical assessment and a plan for follow-up. NICE clinical assessment.
If severe mania, psychosis, inability to care for basic needs, suicidal thoughts with intent, or immediate danger develops, seek urgent local medical or emergency help. If a person is in the United States, they can call or text 988 for suicide and crisis support; emergency services are appropriate when immediate safety is at risk. This guidance addresses the clinical situation rather than relying on an imaging result. NICE crisis and risk recommendations.
What Neuroscience Has Established—and What Remains Uncertain
Large studies support the existence of reproducible group-level associations involving cortical thickness, white matter microstructural measures, selected subcortical volumes, and task-dependent functional signals. Sleep and circadian regulation are clinically relevant and under active investigation. Genetic susceptibility contributes substantially to risk. Medication exposure and clinical state shape interpretation of imaging findings. ENIGMA neuroimaging review; Schumer et al.; Marchetti et al..
The field has not established one anatomical signature, a neurotransmitter threshold, or an MRI-based biomarker that reliably diagnoses individual patients. The direction and clinical meaning of many associations remain unresolved. Better longitudinal designs, diverse cohorts, consistent definitions of mood state and medications, and independent replication will determine which findings become clinically useful. Current diagnostic and treatment decisions should continue to be grounded in validated clinical methods and evidence-based guidelines.
Frequently Asked Questions
What part of the brain is responsible for bipolar disorder?
Current evidence implicates interacting networks that support emotion, reward, attention, memory, and self-regulation, including frontal and limbic structures. No single brain area has been identified as the sole cause. Schumer et al.
Does bipolar disorder show up on an MRI?
Some structural MRI differences appear when researchers compare large groups. A routine MRI cannot confirm or exclude the diagnosis in one individual, and clinical practice relies on episode history and assessment. Hibar et al.; NIMH.
Is bipolar disorder caused by a chemical imbalance?
There is no validated single-chemical explanation or diagnostic neurotransmitter test. Multiple interacting biological and environmental factors contribute to risk and expression of symptoms. NIMH.
Does mania permanently damage brain cells?
Longitudinal studies report associations between illness course, manic episodes, and some imaging measures, but results vary and do not demonstrate inevitable irreversible neuronal loss in every person. Longitudinal review; Mania-related review.
Are bipolar I and bipolar II visible as different patterns on scans?
The large ENIGMA cortical analysis did not detect significant subtype differences in its examined cortical measures. Diagnosis distinguishes the conditions through clinical episode criteria. Hibar et al..
Can lithium protect the brain?
Imaging research has found associations between lithium and certain brain-volume or thickness measures, but those associations do not establish neuroprotection for every patient. Lithium's clinical use requires individualized benefits-and-risks assessment and safety monitoring. Hajek et al.; NICE.
Will a future brain test replace psychiatric interviews?
Reliable biomarkers remain a research goal. They would need robust prospective and external validation and evidence of clinical benefit. Current diagnosis depends on an evaluation of episodes and their course, rather than a single brain measurement. Nunes et al.; NICE.
Related Articles
References
Abé, C., et al. (2023). Mania-related effects on structural brain changes in bipolar disorder: A narrative review of the evidence. Molecular Psychiatry, 28, 2674–2682. DOI.
Cheon, E.-J., et al. (2022). Cross disorder comparisons of brain structure in schizophrenia, bipolar disorder, major depressive disorder, and 22q11.2 deletion syndrome: A review of ENIGMA findings. Psychiatry and Clinical Neurosciences, 76, 140–161. DOI.
Ching, C. R. K., et al. (2022). What we learn about bipolar disorder from large-scale neuroimaging: Findings and future directions from the ENIGMA Bipolar Disorder Working Group. Human Brain Mapping, 43, 56–82. DOI.
Favre, P., et al. (2019). Widespread white matter microstructural abnormalities in bipolar disorder: Evidence from mega- and meta-analyses across 3,033 individuals. Neuropsychopharmacology, 44, 2285–2293. DOI.
Förster, K., et al. (2023). Progressive grey matter alterations in bipolar disorder across the life span: A systematic review. Bipolar Disorders, 25, 443–456. DOI.
González-Ojeda, L. M., et al. (2025). Diagnostic performance of T1-weighted MRI in bipolar disorder: A systematic review and meta-analysis. Journal of Psychiatric Research, 191, 604–616. DOI.
Hajek, T., et al. (2012). Smaller hippocampal volumes in patients with bipolar disorder are masked by exposure to lithium: A meta-analysis. Journal of Psychiatry & Neuroscience. PubMed.
Hibar, D. P., et al. (2018). Cortical abnormalities in bipolar disorder: An MRI analysis of 6,503 individuals from the ENIGMA Bipolar Disorder Working Group. Molecular Psychiatry. PubMed.
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, 344–353. DOI.
King, S., et al. (2026). Psychotropic medications and their interactions with subcortical brain volume in bipolar disorder: An ENIGMA mega-analysis. Molecular Psychiatry, 31, 2941–2953. DOI.
Marchetti, M., et al. (2025). Sleep abnormalities in bipolar disorders across mood phases: A systematic review and meta-analysis. Sleep Medicine Reviews, 83, 102137. DOI.
Mesbah, R., et al. (2023). Association between the fronto-limbic network and cognitive and emotional functioning in individuals with bipolar disorder: A systematic review and meta-analysis. JAMA Psychiatry, 80, 432–440. DOI.
Miola, A., et al. (2026). Neural signatures of bipolar disorder subtypes: A comprehensive systematic review of neuroimaging studies. Journal of Affective Disorders, 394, 120605. DOI.
National Institute for Health and Care Excellence. (2025 update). Bipolar disorder: Assessment and management (CG185). Guideline.
National Institute of Mental Health. (n.d.). Bipolar disorder. Official resource.
Nunes, A., et al. (2020). Using structural MRI to identify bipolar disorders: 13-site machine learning study in 3,020 individuals from the ENIGMA Bipolar Disorders Working Group. Molecular Psychiatry, 25, 2130–2143. DOI.
Schumer, M. C., et al. (2023). Prefrontal, parietal, and limbic condition-dependent differences in bipolar disorder: A large-scale meta-analysis of functional neuroimaging studies. Molecular Psychiatry, 28, 2826–2838. DOI.
Ulrichsen, A., et al. (2025). Do sleep variables predict mood in bipolar disorder? A systematic review. Journal of Affective Disorders, 373, 364–373. DOI.
US Department of Veterans Affairs & Department of Defense. (2023). VA/DoD clinical practice guideline for the management of bipolar disorder. Guideline.
Zhu, Z., et al. (2022). Cortical thickness abnormalities in patients with bipolar disorder: A systematic review and meta-analysis. Journal of Affective Disorders, 300, 209–218. DOI.