Personality disorders

A personality-disorder diagnosis describes persistent difficulties in how someone experiences themselves, manages emotions or relates to other people. Those difficulties can cause considerable distress and disruption. They are not a judgement about a person's worth, and “personality” does not mean that change is impossible.

The best-supported explanation involves several developmental pathways. Inherited dispositions, early experiences, continuing relationships and present circumstances can all contribute. Research is much stronger on these broad influences than on the precise biological steps connecting them. No single brain or chemical explanation covers all personality disorders, and no biological test provides a clinically validated diagnosis. WHO diagnostic guidance, Comprehensive review of borderline personality disorder

Treatment research offers a more practical message: organised psychological care can help. The strongest evidence concerns borderline personality disorder, usually shortened to BPD. Other presentations need their own assessment of what is known. Findings about BPD should not be presented as if they explain or treat every personality disorder.

What the diagnosis describes

Different people can have markedly different difficulties. Intense and rapidly changing emotions may dominate one person's experience. Another may withdraw because relationships feel threatening, struggle with rigid perfectionism, act impulsively, or repeatedly disregard other people's needs. A useful assessment asks how these patterns affect identity, relationships and everyday functioning, while considering the person's culture and circumstances.

The World Health Organization's ICD-11 places particular weight on the severity of difficulties with oneself and other people. It also allows descriptions of recurring traits: frequent distress, detachment, disregard for others, impulsivity, or rigid perfectionism and control. A borderline pattern can be specified as well. This approach avoids assuming that every traditional label describes a separate biological disease. It also does not imply that all the differences between people can be reduced to one scale. WHO clinical descriptions

Researchers continue to debate how best to organise these diagnoses. Some see the older categories as imperfect divisions within broader dimensions of personality and mental-health difficulties. There is no final agreement about the best replacement, or about how difficulties in personality functioning relate to particular traits. That debate concerns classification; it does not make the distress or impairment unreal. Review of current personality-disorder concepts

One further distinction matters when reading population statistics. The Global Burden of Disease “other mental disorders” category uses a restricted personality-disorder group, excluding people with other mental or substance-use disorders separately counted in its model. That is a statistical accounting choice. It does not describe the broader population encountered in clinical services, establish a separate biological subtype or allow treatment-trial results to be converted directly into a percentage reduction in that residual burden. GBD methods appendix, GBD 2023 mental-disorder analysis

Inherited vulnerability matters without fixing a person's future

Twin and family research supports inherited influences across personality-disorder features. A study of 2,794 Norwegian twins found that some genetic influences crossed the boundaries of the ten traditional diagnoses. Because full diagnoses were relatively uncommon, much of the analysis concerned the number of diagnostic features people showed. It therefore informs shared vulnerability more directly than it establishes a genetic explanation for each separate diagnosis. Norwegian twin study

A Swedish study followed nearly 1.9 million people, including 11,665 diagnosed with BPD. It estimated heritability at about 46%, with a statistical uncertainty range of 39–53%. This means that inherited differences accounted for a substantial share of variation in liability in that population under the study's assumptions. It does not mean that 46% of one person's difficulties are genetic, and it says nothing about whether they can improve. Family resemblance generally became weaker as genetic relatedness decreased. Swedish family study of diagnosed BPD

A large genetic study published in July 2026 compared 12,339 BPD cases with over a million controls before further replication. Six regions of the genome met its discovery threshold; the combined analysis identified eleven. The contribution captured by common genetic variants was estimated at 17.3%, with a range of 15.2–19.4%. This is a different quantity from the family-based estimate. The study also found shared genetic risk with other disorders. Its participants had European ancestry and individual prediction remained modest, so these findings do not provide a diagnostic test or establish equal predictive value across populations. 2026 BPD genetic study

Inherited dispositions may affect how people respond to circumstances and which situations they encounter. That broad possibility is compatible with development and change. It does not demonstrate a particular gene interacting with a particular childhood experience in every person with BPD.

Childhood adversity is important, but it is not a universal explanation

Abuse, neglect and other adverse childhood experiences are strongly associated with BPD. However, the size of the association depends greatly on how a study is designed. A review of 97 studies found that people with BPD had roughly fourteen times the odds of reporting adversity compared with non-clinical controls, about three times the odds compared with other psychiatric groups, and a smaller estimate of about 2.6 in population studies. The respective statistical uncertainty ranges were approximately 11–17, 2.6–3.8 and 1.2–5.3. These are comparisons of reported exposure across groups. The largest figure is not proof that trauma multiplies an individual's future BPD risk by fourteen. Review of childhood adversity and BPD

Studies following people over time add evidence about development. A review of 39 reports from 24 distinct samples identified earlier maltreatment, family conditions, frequent negative emotions, impulsivity and other psychological difficulties as predictors of later BPD features. Most predictors were not specific to BPD. Evidence about attachment—the expectations and security people develop in close relationships—was mixed, rather than identifying one history shared by everyone. Review of prospective risk studies

Studies designed to separate family influences give a more complicated picture. In 2,808 young adult twins, links between reported childhood trauma and BPD traits weakened considerably after accounting for shared family and genetic influences. A different British study of 1,116 twin pairs found evidence consistent with harsh treatment contributing environmentally to borderline-related characteristics at age twelve. The latter outcome was childhood characteristics, not adult BPD. Together these findings leave room for real environmental effects while showing why ordinary associations cannot determine their exact size or timing. Young-adult twin study, Prospective childhood twin study

An overview of five research reviews, covering 56 associations involving borderline, antisocial and dependent personality disorders, found the strongest environmental evidence for childhood adversity and BPD. None of the associations reached its highest credibility category. Retrospective recall, the choice of comparison group and shared family influences all matter. A careful clinical account can take adversity seriously without assuming that every person has experienced it or that one experience explains the whole condition. Umbrella review of risk factors

BPD and complex post-traumatic stress disorder can overlap, and someone can meet criteria for both. Yet a 2024 review found that most studies identified distinguishable symptom patterns. Overlap does not establish that the conditions are identical or have a single cause. Review comparing complex PTSD and BPD

How emotions, expectations and relationships may interact

Researchers study how people react emotionally, interpret social signals, understand themselves and respond to distress. These processes can influence one another. An expectation of rejection might increase distress; a defensive response might then strain a relationship and strengthen the expectation. This is a possible reinforcing pattern, not a story that can be assumed from a diagnosis. Researchers would need repeated measurements and tests of alternative explanations to establish the sequence.

One example concerns interpreting other people's thoughts and intentions, sometimes called mentalising. People can occasionally infer more about another person's motives than the available evidence justifies. A review of 36 studies with 4,188 participants found an association between this tendency and BPD, but the association was about as strong in other mental-health conditions. The correlations were approximately 0.26 and 0.24, respectively—modest group associations, not a diagnostic divide. Review of overinterpreting mental states

Such a difficulty may be relevant for a particular person without being present in everyone, unique to BPD or responsible for the original onset. Researchers must also avoid offering a description as though it were an explanation. Saying that unstable emotions cause a disorder partly defined by unstable emotions adds little unless the underlying process and its development are demonstrated.

What biological research has found

Brain findings depend on what is measured

A familiar explanation of BPD contrasts powerful emotional responses with insufficient regulation. Brain systems involved in emotion, memory and flexible control are reasonable subjects for study, but a simple “emotional brain overpowering the thinking brain” account has not been established as a universal mechanism. A comprehensive 2024 review concluded that an integrated, well-supported biological model remains unavailable. Comprehensive BPD review

An analysis of 52 brain-imaging studies illustrates the uncertainty. Across all the different tasks, it did not find a consistent location where patient and comparison groups differed. When researchers restricted attention to emotional-processing tasks, differences appeared in regions involved in emotion, memory and monitoring. Task selection changed the result. Neither an overall unclear finding nor a positive subgroup result tells us what happens in every individual. Review of variation in BPD imaging studies

The stress response is more complex than “too much cortisol”

Cortisol is part of the body's hormonal response to challenge. A review of 37 studies, including 803 people with BPD and 1,092 controls, found higher cortisol output over continuous measurement periods but weaker responses to social stress challenges. Single measurements and responses to drug challenges did not clearly differ. There is therefore no simple uniform “high cortisol” signature. Review of stress-hormone functioning

These differences might reflect vulnerability, adversity, current distress, other illnesses, treatment or consequences of living with the condition. The measurements alone cannot decide between those explanations.

Researchers also study chemical marks that affect how genes are used without changing the underlying DNA sequence. A 2025 review of nineteen studies of one such process, DNA methylation, found mixed results and substantial limitations. This remains a possible bridge between experience and biology, not an established explanation or clinical test. Review of epigenetic research in BPD

Findings from one presentation cannot simply be transferred to another

Obsessive–compulsive personality disorder involves enduring rigidity and perfectionism; it is distinct from obsessive–compulsive disorder. A review of 49 relevant articles found recurring evidence concerning inflexible thinking and difficulties with mental control, but no comprehensive biological explanation. Many articles studied related traits rather than people with the diagnosis itself. Review of obsessive–compulsive personality-disorder biology

Psychopathy also overlaps with, but is not interchangeable with, antisocial personality disorder. A review of 101 psychopathy imaging studies found that about 85% of 1,573 tests concerning medial frontal brain regions did not detect the proposed association. This challenges confident claims that psychopathy consistently resides in one prefrontal abnormality. It leaves open other networks and differences between subgroups, rather than proving that biology is irrelevant. Review of psychopathy imaging findings

Why confidence in precise causes remains limited

Many studies compare people after difficulties have already developed. A brain or hormone difference might have preceded the condition or followed it. Comparing patients with exceptionally healthy controls may pick up general distress, adversity or additional illnesses rather than something unique to personality disorder. Recollections of childhood and records collected at the time can differ. Adult diagnoses, adolescent symptoms, childhood characteristics and ordinary traits also answer different questions.

Small imaging samples and numerous analytical choices make findings unstable. Several reports can describe the same underlying participants: 39 prospective publications did not mean 39 independent samples. European-ancestry genetic studies and selected specialist-service samples cannot automatically represent other populations.

Inconsistent biology may mean that studies are too small or their methods too variable. It may also mean that a single diagnosis contains several different mechanisms, so one average biological signature is the wrong target. Both possibilities call for better research; neither supports dismissing the person's difficulties.

What effective psychological treatment involves

For BPD, the strongest treatment evidence supports structured psychotherapy: a coherent approach delivered by trained clinicians, with agreed goals, active work on difficulties and regular review. Several programmes can help, and no approach has emerged as best for everyone. American Psychiatric Association treatment guidance

A major Cochrane review included 75 trials and 4,507 participants. Compared with usual care, psychotherapy reduced overall BPD severity by about half a standard deviation—a way of comparing differences across research scales. The estimate was 0.52, with a statistical uncertainty range of 0.33–0.70, favouring psychotherapy. This is a moderate average difference, not a percentage cure rate. Evidence for self-harm, suicide-related outcomes and social functioning was less certain. Only overall severity met the review's chosen threshold for a clinically important average improvement. Individuals may nevertheless experience substantial benefit in other areas. Cochrane review of psychological therapies

The names of therapies reflect different emphases. Dialectical behaviour therapy, or DBT, teaches skills for handling intense emotions, tolerating distress and managing relationships, while directly addressing dangerous behaviours. Mentalisation-based treatment, or MBT, helps people consider their own and others' thoughts and feelings when emotions make interpretation difficult. Schema therapy works on longstanding expectations, emotional patterns and ways of coping. Transference-focused psychotherapy examines recurring experiences of oneself and relationships as they arise in therapy. Each is a treatment model; a programme helping people does not automatically prove its proposed mechanism.

Comprehensive DBT is particularly well studied when recurrent self-harm is a central problem. That makes it a strong option to discuss, not proof that it always outperforms every credible alternative. Treatment should fit the person's main difficulties, preferences, ability to attend and access to skilled care. NICE BPD recommendations

Comparing established therapies

A 2024 review of 25 studies and 2,545 participants found support for several psychological treatments but insufficient comparative evidence to identify a consistent winner. Review of BPD psychotherapies

The 2026 BOOTS trial directly compared DBT and schema therapy in 204 adults at nine Dutch centres. Both combined individual and group work for two years, followed by another year of observation. Both groups improved markedly; there was no statistically clear difference in overall BPD severity or secondary outcomes. The estimated difference at three years was small: 0.15 of a standard deviation, with an uncertainty range from −0.17 to 0.47 that included an advantage in either direction. Because this was designed to test superiority, an unclear difference did not prove exact equivalence. By two years, 52% had stopped DBT treatment and 46% schema therapy, highlighting the demands of engagement and the importance of missing follow-up information. BOOTS trial

In another trial of 180 people, DBT did not clearly outperform an organised general psychiatric management programme delivered by experienced clinicians. Both groups improved. That supports the specific management approach studied, not unstructured follow-up or medication checks alone. DBT and general-management comparison

Shared features of good treatment—a dependable relationship, an understandable plan, active work and coordinated care—may account for some benefit. However, a 2026 review found the evidence too weak to conclude that simpler adapted programmes are generally equivalent to specialist therapies. The content, staff expertise and organisation of the comparison treatment matter. Review of simpler and specialist models

Treatment for other personality-disorder presentations

Avoidance, dependence and rigid perfectionism

Avoidant, dependent and obsessive–compulsive personality disorders are often grouped under “cluster C”. Schema therapy has growing evidence in this group, and CBT has direct support for avoidant personality disorder.

A 2024 trial compared group schema therapy with group CBT in 154 people who had both avoidant personality disorder and social anxiety disorder. Both groups improved substantially; symptom outcomes did not clearly differ at follow-up, although more people completed schema therapy. The result applies to that combination of difficulties, not automatically to avoidant personality disorder without social anxiety. Avoidant personality-disorder trial

The QUEST trial, published in October 2026, randomised 375 adults and analysed 369 who began treatment. Group and individual schema therapy did better than specialised usual care at one year, with small-to-moderate average effects of 0.33 and 0.41 of a standard deviation, respectively. The differences were no longer statistically clear at two years. Among those analysed, 283 had avoidant, 76 obsessive–compulsive and only ten dependent personality disorder—just 2.7% of the analysed group. The group-therapy finding was also less robust after unusually influential observations were excluded. Schema therapy is a reasonable option, while long-term superiority and results specifically for dependent personality disorder remain less certain. QUEST trial

Antisocial personality disorder

Care can address impulsivity, relationships, antisocial behaviour and offending through cognitive and behavioural approaches. Routine medication is not recommended for antisocial personality disorder or its associated aggression and impulsivity. NICE antisocial personality-disorder guidance

The 2025 MOAM trial offers important evidence for adapted MBT. It involved 313 men on probation in England and Wales, comparing MBT plus probation with probation alone. Aggression improved more with MBT at twelve months, with a moderate-to-large average effect of about three-quarters of a standard deviation. The advantage remained detectable at eighteen months but was no longer statistically clear at twenty-four months. The comparison did not match the extra therapeutic attention. These results support treatment potential while leaving uncertainty about durability, other settings and people outside this selected male population. MOAM trial

Presentations with thinner evidence

For paranoid, schizoid and schizotypal personality disorders, a 2025 review found only nineteen treatment studies with 468 participants. Small samples and varied treatments prevented a reliable ranking. A separate 2026 review of schizotypal disorder found preliminary benefits from psychological, medication and combined approaches, but judged the evidence low or very low certainty. Limited medication evidence in schizotypal disorder cannot be transferred to personality disorders collectively. Cluster A treatment review, Schizotypal-disorder review

For narcissistic personality disorder, it is no longer accurate to say that no randomised trials exist. A 2026 trial of 114 people compared schema-focused therapy with another structured approach, the Unified Protocol. Both groups improved, with different dimensions favouring different treatments. Later criticism and the authors' response raised unresolved issues including retrospective registration and heavy reliance on self-report. This is preliminary evidence, not an established preferred therapy. Narcissism treatment trial, Authors' response to methodological criticism

Evidence specific to histrionic and other specified or unspecified presentations is especially limited. Trials combining several diagnoses can support offering care, but a small subgroup cannot establish which programme is best for that diagnosis. Mixed-diagnosis schema-therapy trial

Where medication fits

Medication is not an established lasting treatment for the whole pattern of BPD. A 2022 Cochrane review of 46 trials and 2,769 participants found no dependable advantage over placebo for overall severity, self-harm, suicide-related outcomes or social functioning. Much of the evidence was very uncertain and reporting of adverse effects was inadequate. That is a limit on what has been established, rather than proof that no medicine can ever help any symptom. Cochrane medication review

A 2026 comparison of 35 trials with 2,551 participants reported short-term anger or hostility benefits from topiramate, lamotrigine and aripiprazole, with weaker evidence for some other symptom–medicine pairings. Positive treatment periods were generally six to twelve weeks. The authors judged some anger findings to have moderate or high certainty; those ratings concern particular short-term symptoms. This qualifies a blanket claim that medication never helps BPD symptoms, but it does not establish durable benefit for the whole disorder, everyday functioning or suicide mortality. 2026 medication comparison

Lamotrigine illustrates the distinction. In the larger LABILE trial of 276 people, it produced no clinically useful improvement in overall BPD severity at fifty-two weeks. The adjusted difference was just 0.1 points, with an uncertainty range from −1.8 to 2.0. A short-term effect on anger and a lasting effect on the whole condition are different claims. LABILE trial

Guidelines differ in how they handle this uncertainty. NICE advises against medication specifically for BPD or its associated behaviours, while permitting treatment for separately diagnosed conditions. The American Psychiatric Association allows time-limited medication aimed at a specific measurable symptom alongside psychotherapy and recommends a medication review at least every six months. Neither endorses an accumulating collection of long-term prescriptions as the main treatment. NICE medication recommendations, APA medication guidance

Depression, anxiety, post-traumatic stress and other conditions can require their own care within a coordinated plan. A personality-disorder diagnosis should not be a reason to dismiss those difficulties or withhold treatment. Medicine changes should be discussed with the prescriber, including what each medicine is intended to improve and whether it is doing so.

Newer approaches and the importance of everyday recovery

A 2025 trial of eye movement desensitisation and reprocessing, or EMDR, enrolled 159 people with personality disorders. It found symptom improvement compared with waiting list after a brief course, with a moderate average effect of 0.46 of a standard deviation at three months. The short follow-up and absence of an active psychological-treatment comparison mean it cannot establish superiority over existing programmes. EMDR trial

A 2026 English trial tested up to ten sessions of Structured Psychological Support in 336 adults with probable personality disorder. It did not improve work and social functioning at twelve months compared with enhanced usual care. This is a finding about that programme and outcome, not proof that all brief interventions fail. Structured Psychological Support trial

Ketamine remains experimental as a treatment specifically targeting BPD. An October 2026 review found only five varied studies and insufficient evidence for an established clinical role. Nor does the wider literature establish that increasing the length or intensity of every psychological treatment necessarily improves results. Ketamine review, Review of treatment delivery

Recovery must include more than a symptom score. In follow-up of the DBT–general-management trial, around two-thirds no longer met the diagnostic threshold, yet 53% were neither employed nor in education at three years. Those numbers describe one cohort, but show why relationships, work, education and participation need explicit attention even after symptoms improve. Longer-term functioning after treatment

How to read claims that a treatment is “best”

The comparison matters. Doing better than a waiting list is different from doing better than an experienced clinician offering a credible alternative. Outcomes matter too: less anger, fewer self-harm episodes, reduced overall symptoms and improved employment are different achievements. A study may be too small or too short to detect changes in rare outcomes such as suicide deaths. Effect sizes from different studies cannot fairly be arranged as a treatment league table.

Patients and therapists generally know which psychotherapy they receive. Expectations, therapist preferences and different amounts of attention can affect results. Independent assessors help, but do not remove every limitation. Improvement within one group may reflect ordinary change over time or concurrent care; the comparison between randomly assigned groups estimates the additional benefit over the actual alternative offered.

Women predominate in many BPD trials, while MOAM involved only men. Specialist services may exclude some crises or combinations of conditions. Treatment dropout and missing outcomes matter, and studies often record symptoms more thoroughly than deterioration, adverse experiences, practical burden or long-term quality of life. These limits support careful choices and review rather than a universal ranking.

Change is possible, and research can become more useful

A review of forty follow-up studies involving 38,432 people found substantial change as well as stability. Across the varied follow-up periods, about 57% retained any personality-disorder diagnosis and about 45% retained BPD. These are not individual forecasts or standardised recovery rates: time periods and assessment methods differed. Losing a diagnosis also does not guarantee that everyday difficulties have disappeared. Review of diagnostic stability and change

Better research should follow development before serious difficulties emerge, measure family influences and adversity together, and include a wider range of diagnoses, cultures and ancestries. Biological studies need larger samples, repeated measurements, comparison with other mental-health conditions and independent confirmation. Researchers should test whether changing a proposed mechanism comes before and explains later improvement, rather than assuming that two things changing together proves the theory.

Treatment trials need direct comparisons of credible programmes, sustained functioning and quality-of-life outcomes, careful reporting of harms, and better representation of neglected diagnoses and settings. Promising findings about medicines, EMDR or newer approaches need longer follow-up and independent replication.

For someone seeking care now, a practical starting point is an understandable plan with trained clinicians, goals that reflect the difficulties that matter most, support to remain engaged and regular review of symptoms, safety and daily life. A crisis plan can identify warning signs, useful support and who to contact. Immediate danger or an inability to stay safe needs urgent local help. Treatment should respond to the person and their circumstances, with room to change course as their needs change. NICE care and crisis recommendations