Supporting someone with borderline personality disorder: what the latest research says

A US study began with 290 hospital inpatients with BPD and assessed them every two years for 24 years. It reported estimated remission rates of 77% - 100%, depending on whether remission had to last 2, 4, 6, 8, 10 or 12 years [1]. Each figure describes reaching a particular length of remission during follow-up. Remission means no longer meeting the diagnostic criteria; some symptoms may remain.

A review of 11 studies involving 837 people found long-term remission rates of 50% - 70% [2]. Another US study reported 85% remission over ten years [3]. Populations, definitions and follow-up differ. Taken together, the findings offer hope; they do not promise a particular outcome or mean that support is no longer needed.

The 24-year study reported recovery rates of 37% - 60%, again depending on the required duration [1]. Its research definition included good social and vocational functioning as well as symptom remission [4]. That is a study measure, not a rule that someone must work full time to count as recovering.

Depending on the duration measured, 11% - 40% experienced recurrence after remission, and 29% - 59% lost a period of recovery [1]. Losing “recovery” could concern everyday functioning as well as symptoms. The separate ten-year study found modest improvements in social functioning despite substantial symptom improvement [3]. Help with relationships, meaningful activity, work or study may remain useful after crises ease.

In the 24-year study, researchers compared five groups of other mental health conditions. Absence of a substance use disorder had the strongest association with recovery among those five groups [5]. The comparison was limited to those five groups and cannot show cause.

A 2025 review argues that loneliness can persist when other symptoms improve, and that treatment should also help people build connections beyond a partner or one main supporter [6]. The authors propose broader social support; they did not test particular hobbies as treatments. If they want company, you could help them find a regular group or activity they enjoy. Alcohol or drug difficulties also deserve support alongside their mental health care.

In a US study of 35 people followed for a year after leaving hospital, relatives' criticism didn't predict how they got on or whether they went back into hospital. Relatives who were highly emotionally involved, meaning very close, anxious and protective, were linked with people doing better. The authors noted this was the opposite of the pattern in schizophrenia and mood disorders [7]. It's one small, old observational study: it did not test whether becoming more involved improves someone’s health. It is not a reason to monitor them more closely or set aside your own boundaries. Closeness also has a cost. In a 2015 survey of 280 carers of someone with BPD, 83% described criticism at home and 70% over-involvement, and over-involvement went with more burden and poorer mental health for the carer [8]. These studies do not establish an ideal level of involvement. Support for you matters in its own right.

In a Canadian trial of 240 people with BPD and suicidal or self-harming behaviour, six months of comprehensive DBT met the trial's threshold for doing no worse than twelve months on total self-harm through 24 months [9]. Both included individual therapy, skills groups and other support. The finding did not cover every outcome equally.

A Danish trial of 166 adults with diagnosed or subthreshold BPD found no clear difference in BPD symptoms at 24 months between 5 and 14 months of MBT. Longer treatment had a better quality-of-life result, but 40% of those data were missing [10]. No clear difference is not proof that the courses are interchangeable.

In a UK follow-up, 98 of the original 134 participants took part. More of the MBT group continued to meet the trial's recovery criteria over eight years than the structured-management group: 74% versus 51% [11]. The measure concerned self-harm, suicide attempts and hospital use, not every aspect of recovery. Ask what treatment fits their needs and what support continues afterwards.

NICE's guideline says that in a crisis, staff should check the person's crisis plan, stay calm and try to understand things from their point of view. Hospital admission should only be considered when the risk can't be managed by other services, or under the Mental Health Act, with its length and purpose agreed in advance [12]. The trial evidence on particular crisis plans and self-referred brief admissions is limited: a 2022 Cochrane review found only two trials, with 213 people, and neither joint crisis plans nor the offer of brief admission made a clear difference, though the evidence was very uncertain [13]. This does not establish that all crisis planning or hospital care is ineffective. These admission decisions need professional assessment. With their agreement, it's worth knowing what their crisis plan says, what your part is and who to call out of hours. NICE also says that when someone is discharged back to their GP, this should be discussed with them and, where possible, their family, with a plan for future crises and for getting back into services [12].

A 2023 study of 167 parents of people with BPD linked heavier burden with more distress and lower quality of life [14]. It was a snapshot and cannot establish what caused what.

Family Connections is a 12-session course built on DBT skills. A Spanish trial randomly allocated 121 relatives to it or another structured course. At the end, there was no clear difference between groups on the measured outcomes [15]. That does not prove equal effectiveness. Some later empowerment measures favoured Family Connections, but around 42% of follow-up data were missing.

Ask what the available course covers, who leads it and whether it fits your needs. Support for you is worthwhile in its own right; it is not another responsibility to make someone else recover.

Humber runs it free and online for adults supporting someone, where either of them lives in Hull or the East Riding (01482 689154) [16]. Northamptonshire runs it on Zoom twice a year for relatives and close friends of someone living in the county (03000 271012) [17]. Cambridgeshire and Peterborough runs it online for families and carers of people using its Personality Disorder Community Service or Relational Emotional Difficulties Service (01733 746800) [18]. Elsewhere, ask the care team or the trust's carers lead what family courses exist.

Next check due: 3 November 2026. Details to check: Local eligibility, course availability and phone numbers.

A 2009 trial assigned 180 people with BPD and a history of repeated self-harm to a year of DBT or structured general psychiatric management. Both groups improved, with no clear difference between them [19]. General management was a specified treatment delivered by psychiatrists experienced in BPD; it was not simply occasional appointments or unstructured support.

If specialist therapy is unavailable, ask what organised treatment can be offered, who will coordinate it and how progress will be reviewed. NICE's published BPD recommendations remain a useful reference [12]; its guidance-review page lists work on future updates [20].

Next check due: 3 November 2026. Details to check: NICE guidance and review status.

Schema therapy works on longstanding patterns in how someone sees themselves and relates to others. A 2022 trial of 495 adults in five countries compared two formats with usual treatment. The combination of individual and group schema therapy reduced BPD severity more than either usual treatment or predominantly group schema therapy, and more people stayed in treatment. Predominantly group therapy alone did not clearly outperform usual treatment on the main outcome [21].

This does not establish local NHS availability. Ask the team which therapies they offer and why a particular format might suit the person.

A Swedish trial of 125 adults with self-harm or suicidality and at least three BPD criteria tested access to short, self-referred hospital stays alongside usual care. It found no clear reduction in overall inpatient use compared with usual care alone [22].

A later study followed 62 participants and described how they used this brief-admission option over four years. Use varied; average days spent in brief admission fell [23]. This descriptive follow-up cannot show that the scheme reduced overall hospital use. Whether it would work in NHS services is also uncertain.

A 2024 review included 40 publications covering 6,611 participants, many without a BPD diagnosis. It found no clear improvement in overall BPD symptom severity, but some benefit for suicidal thoughts and paranoia across the relevant studies [24]. Results from this mixed group of studies need care when choosing an app for someone with BPD.

A later German trial of 580 adults found a small improvement in BPD symptoms at three months when the digital programme priovi was added to usual care, compared with usual care alone [25]. The trial was funded by its developer. It adds evidence about one programme, without establishing UK availability or making it a substitute for crisis support. Discuss any programme's purpose, privacy and suitability with the care team.