You have read the articles. Set firm limits. Don't reward the behaviour. Lower the emotional temperature. Learn everything you can. Most of that advice has never been tested in a single study, and at least two pieces of it point the opposite way when someone finally does test them.

The short answer: the burden on people who love someone with BPD is real and measurable — across 465 carers, scores on objective and subjective burden ran about half a standard deviation above those of carers of inpatients with other serious mental illnesses. There is a free twelve-week programme built specifically for relatives, delivered online, and a free national family course besides. The evidence for both is more modest than their marketing suggests, and worth knowing accurately. And a good deal of the standard family advice — set limits, don't be overinvolved, learn all you can — has either no supporting study or a study pointing the other way. What follows separates the three.¹ ² ³

What the research says about you

  • Burden and grief run higher than in comparable conditions. In a systematic review of six studies covering 465 carers: "Scores on objective and subjective burden were half a standard deviation above the mean compared to carers of inpatients with other serious mental illnesses."¹ In the largest single sample, 287 carers, "Burden and grief were significantly higher than that reported by carers of persons with other serious mental illnesses."⁴
  • Grief is the word the qualitative work keeps finding, not stress. One study's four themes: "a life tiptoeing; powerlessness, guilt, and lifelong grief; feeling left out and abandoned; and lost trust."⁵
  • But not hopelessness. In 75 relatives assessed before a family programme, 82 percent scored in the minimal or mild range on a hopelessness measure, with a mean of 4.61 — mild — and the programme did not change it.⁶ Worth saying because the literature about carers is often written as though despair were universal. It is not what the measures find.
  • What predicted doing better in a large carer survey: perceived social support, and positive reframing. What predicted doing worse: self-blame, behavioural disengagement, and substance use.⁷

The three pieces of standard advice that do not survive contact with the evidence

"Learn everything you can about BPD." The one study that tested this found the opposite of what everyone assumes: "greater knowledge about BPD was associated with higher levels of family members' burden, distress, depression, and greater hostility toward patients." The authors' own reading was that this "raises concerns about the value of information family members receive about BPD and the importance of the source and accuracy of the information they receive."⁸ Thirty-two participants, cross-sectional, so the direction of causation is unknown — but it is a real finding and it cuts against reading forums at 2 a.m. Where you get the information matters more than how much of it you accumulate.

"Lower the expressed emotion — stop being critical, stop being overinvolved." That advice was imported wholesale from the schizophrenia literature, where it is well supported. In the one prospective study in BPD, following 35 patients for a year after discharge: "Contrary to prediction, relatives' criticism and hostility did not predict how well patients did in the year after discharge. Neither did they predict rates of rehospitalization." And more surprising still: "Patients whose families scored higher on emotional overinvolvement had better clinical outcomes."⁹ Australia's national guideline reaches the same conclusion — "This limited evidence suggests a lack of association between family hostility and criticism and re-admission rates for family members with BPD."³

One study, thirty-five people, never prospectively replicated. It does not license the reverse advice. It does mean the confident version of the standard advice is not supported, and that the schizophrenia model does not simply transfer.

There is a genuine tension here worth naming. Emotional overinvolvement was associated with better patient outcome in that study, and in a separate survey of 280 carers, "Elevated emotional overinvolvement was correlated with higher burden and mental health problems."¹⁰ Both can be true. Involvement that helps the person you love may cost you, and a family plan that ignores your side of that is not a plan.

"Set firm limits." Limit-setting appears in essentially every set of family materials, including the curriculum of the main family programme. We could not find a single study measuring its effect on the outcomes of either the person with BPD or the relative. It is a plausible clinical convention. It is not a finding, and it should not be delivered to exhausted parents as though it were.

What has actually been tested

Programme Design What it found
MBT-FACTS — five evening meetings, delivered by trained family members Randomised, delayed-treatment control, 56 family members The strongest result in this field: a significantly steeper decline in adverse incidents between the relative and the person with BPD (β = −1.07, 95% CI −1.40 to −0.74, p < .000). No difference on depression, anxiety or total burden — both groups improved on those¹¹
Group psychoeducation for carers — 10 weeks Randomised against a waitlist, 68 carers Improvements in relationship adjustment, family empowerment and reduced expressed emotion, sustained at 12 months. Waitlist control, so attention effects are not ruled out¹²
Multi-family group therapy Randomised against usual care, 57 caregivers Improvements in social functioning, perceived support from clinicians, and coping. Notably absent from that list: burden, depression, anxiety¹³
A three-session multi-family group added to online modules Randomised, 79 family and friends of young people Negative. No significant differences on the primary or secondary outcomes; everyone improved on knowledge regardless of arm. Adding the group added nothing measurable over the online material alone¹⁴
Family Connections — 12 weeks, peer-led One randomised trial (121 family members) plus many uncontrolled studies See below

Family Connections, honestly

This is the programme you will be pointed to, and it is worth doing. It is also worth knowing what its evidence actually is.

What it is: a free programme for relatives and partners, run by the National Education Alliance for Borderline Personality Disorder, now operating as BPD Alliance. Their own description: "a free, evidence-based program… for family members and loved ones supporting someone with borderline personality disorder (BPD) traits or chronic emotion dysregulation," with "support from trained leaders and peers with shared lived experience."¹⁵

Format and cost, from their page: classes are primarily online by video call, with a limited number of in-person sessions in some locations. Formats include 12 weeks of two-hour meetings; asynchronous learning followed by 8 weeks of two-hour meetings; or asynchronous learning followed by two six-hour days. "The program is completely free."¹⁵

Who can join: family members and loved ones, not the person with BPD; participants must be 18 or over; the person they support must be 13 or over; a formal diagnosis is not required; and each participant applies separately, even family members attending together.¹⁵

How a Californian signs up: through the application form on the programme page. Because classes are primarily delivered online nationally, geography is not much of an obstacle. Applicants join a notification list and are emailed when it is their turn to enrol.¹⁵

The evidence, accurately. There is one randomised controlled trial of Family Connections in relatives of people with BPD: 121 family members, randomised against a dose-matched active comparator — also twelve two-hour sessions, also including validation and crisis skills. On its primary outcome, burden, the result was: "Although no statistically significant differences were found between conditions. However, the adjusted posttest means for FC were systematically better than for TAU, and the effect sizes were larger in burden, stress, depression, family functioning, and quality of life in the FC intervention." At six months, the significant between-group differences were confined to family empowerment.¹⁶

Everything else is uncontrolled pre-post work — from the US, Ireland, Italy, France, Switzerland, Sweden and Canada — and it is consistent: burden down, grief down, mastery up.¹⁷ ¹⁸ ¹⁹ The one controlled-but-not-randomised study compared 24 hours of Family Connections with a three-hour psychoeducation course, which its own authors said "does not constitute a direct comparison group."¹⁸

So the accurate sentence is: one randomised trial, negative on its primary outcome against an equally long active comparator, plus consistent uncontrolled evidence from at least six countries that people feel less burdened afterwards. That is a reasonable basis for spending twelve free evenings. It is not the same as "evidence-based" in the sense that word carries for a drug trial.

NAMI Family-to-Family is the other free option and easier to find locally: eight sessions, taught by trained family members, run through local affiliates — in California, through NAMI California and its affiliate network. Its randomised trial, in 318 family members, found "significantly greater improvements in problem-focused coping as measured by empowerment and illness knowledge," and reduced distress — but "Subjective illness burden did not differ between groups."²⁰ Two caveats a BPD family should know: none of the trials were in relatives of people with BPD specifically, and the published evidence is for the 12-week version rather than the 8-session course now offered.²⁰

What to actually do

Sorted by how much evidence sits behind it.

Best evidenced: validation, with a caveat. In an experimental study of 126 participants, validation's effects were emotion-specific: people higher in emotion dysregulation showed greater increases in positive emotion when shame or sadness was validated, and less when fear was.²¹ Relatives who have been through a family programme nominate validation and radical acceptance as the most useful skills they learned.²² And in a study of what language helps, "Both consumers and carers reported helpful language as being connecting, validating and accepting."²³ Validating is not agreeing. It is saying the feeling makes sense given how things look from where they are standing.

Well evidenced in the sense that a trial moved it: reducing adverse incidents. That was the MBT-FACTS primary outcome, and a five-session peer-led programme moved it substantially.¹¹ It is the most concrete thing in this literature: the number of blow-ups in your house is a thing a short course can change.

Consensus opinion, honestly labelled. Australia's guideline lists what families say helps and what hurts, and states plainly beforehand that "there was insufficient evidence to make specific evidence-based recommendations."³ Its list of things that may make matters worse is still worth reading, because it is specific:³

  • Denial that the person has BPD, which delays help.
  • The belief that "if they can only find the right treatment they will be completely cured," which drives a cycle of new clinicians and disrupts continuing treatment.
  • Demands to stop self-harm: "demands to stop this behaviour can be counter-productive and increase the person's distress… An empathic response may be more helpful."
  • Giving in to avoid a confrontation, agreeing to things you do not believe will help.
  • Repeatedly rescuing, which can prevent the person becoming independent.
  • Accessing "highly stigmatising, blaming information about BPD."

And what the same guideline says supports: understanding the condition; empathy; a non-judgemental attitude — including understanding that "during an episode of difficult emotions the person with BPD may say or do things that they would not say or do at other times, and that do not express how the person normally feels" about you; allowing independence; balancing your own needs against theirs; and listening to what they want even when you disagree.³

On crises, the guidance is procedural rather than behavioural. Guidelines ask that carers be included in developing crisis plans with the person's consent, be given information about how to respond to self-harm and suicide attempts, and have a separate crisis plan of their own; and that after any crisis, the plan be reviewed with the family.³ ²⁴ Australia's guideline adds a line worth asking a clinician to explain to you: "Health professionals should explain the difference between suicidal behaviour and other kinds of self-harm."³ Crisis routes that don't run through the police →

One thing to stop carrying

You will encounter the idea that BPD is caused by an invalidating family environment. It is a model, and — as the paper that proposed it says in its own opening sentence — "no prospective epidemiological studies have evaluated the relationship between family interactions and the development of borderline personality disorder."²⁵ Australia's guideline puts the practical version bluntly: "it is incorrect for health professionals to assume that all family environments are 'toxic' and have 'caused' the person's BPD," adding that "not all people with BPD have a history of abuse or neglect, and that the condition is partly due to genetic and biological factors."³

If a clinician has implied otherwise to you without evidence, that was not a finding either.

Looking after your own side of it

Q&A

Q: Is there a support programme for families of someone with BPD? A: Yes — Family Connections, a free twelve-week programme for relatives run by the National Education Alliance for Borderline Personality Disorder, delivered primarily online, open to anyone 18 or over supporting a person 13 or over, with no diagnosis required.¹⁵ NAMI Family-to-Family is a free eight-session alternative available through California affiliates.²⁰

Q: Does it work? A: Its one randomised trial was negative on its primary outcome against an equally long active comparator, though effect sizes favoured it; a large body of uncontrolled studies across six countries consistently finds burden and grief lower afterwards.¹⁶ ¹⁷ It is free and low-risk. Go in with accurate expectations.

Q: Did I cause this? A: The model that puts family interaction at the centre has never been tested prospectively, and the guideline position is that it is incorrect to assume family environments caused a person's BPD.³ ²⁵

Q: Should I set boundaries or is that abandoning them? A: Nobody has measured this, in either direction. What the guidelines describe as unhelpful is giving in to avoid confrontation and repeatedly rescuing; what they describe as helpful is allowing independence while giving support when it is needed.³ That is consensus, and it is the honest status of the advice.

Q: What do I do when they say they will hurt themselves? A: Ask now, while things are calm, for the clinician to build a crisis plan that includes you with the person's consent, and to give you a separate plan of your own — both are in the guidelines.³ ²⁴ In the moment: 988 is free and 24/7, and demanding that the behaviour stop is specifically named in the guidance as likely to increase distress rather than reduce it.³

Q: They are wonderful to everyone else and terrible to me. Why? A: The guideline names this directly: people with BPD "often direct their anger or difficult behaviour towards those closest," and what is said in an episode may not express how they normally feel about you.³ That is an explanation, not a requirement that you absorb it.

Q: How do I know if their treatment is any good? A: Ask which structured treatment it is and what its components are. What a full DBT programme includes → · The printable check →

Ready to find support for yourself, not only for them? Filter therapists by approach, schedule and payment route → · The full BPD map → · Caregivers →

In crisis? Call or text 988 — free, 24/7.

Sources

  1. Bailey RC, Grenyer BFS, "Burden and support needs of carers of persons with borderline personality disorder: a systematic review," Harvard Review of Psychiatry 21(5), 2013, 248–258 — 465 carers; "Scores on objective and subjective burden were half a standard deviation above the mean compared to carers of inpatients with other serious mental illnesses" — doi.org.
  2. Guillén V, Díaz-García A, Mira A, et al., "Interventions for family members and carers of patients with borderline personality disorder: a systematic review," Family Process 60(1), 2021, 134–144 — "The quality of the included studies varies, and the empirical support for these programs is still preliminary" — doi.org.
  3. National Health and Medical Research Council, "Clinical practice guideline for the management of borderline personality disorder," Melbourne, 2012 — the lists of family behaviours that may support or worsen outcomes, prefaced by "The Committee determined that there was insufficient evidence to make specific evidence-based recommendations on the potential influences of family, partners and/or carers on health outcomes for people with BPD"; "This limited evidence suggests a lack of association between family hostility and criticism and re-admission rates"; "it is incorrect for health professionals to assume that all family environments are 'toxic' and have 'caused' the person's BPD"; recommendations 53 and 54 and Table 7.1 on crisis plans, and "Health professionals should explain the difference between suicidal behaviour and other kinds of self-harm" — nhmrc.gov.au.
  4. Bailey RC, Grenyer BFS, "Supporting a person with personality disorder: a study of carer burden and well-being," Journal of Personality Disorders 28(6), 2014, 796–809 — 287 carers; "Burden and grief were significantly higher than that reported by carers of persons with other serious mental illnesses" — doi.org.
  5. Ekdahl S, Idvall E, Samuelsson M, Perseius KI, "A life tiptoeing: being a significant other to persons with borderline personality disorder," Archives of Psychiatric Nursing 25(6), 2011, e69–e76 — "a life tiptoeing; powerlessness, guilt, and lifelong grief; feeling left out and abandoned; and lost trust" — doi.org.
  6. Joyce M, Kells M, Boylan E, et al., "Hopelessness for family members of individuals with borderline personality disorder," Family Process 63(4), 2024, 2135–2150 — 75 participants; "The majority of participants (82%) reported scores within the 'minimal' or 'mild' ranges of hopelessness before the FC program… There was no significant difference in hopelessness scores after program completion" — pmc.ncbi.nlm.nih.gov.
  7. Hayes A, Dempsey M, Kells M, Murphy M, "Carers of individuals with borderline personality disorder: well-being, distress and coping," Borderline Personality Disorder and Emotion Dysregulation 10, 2023, 31 — 863 carers analysed; "Perceived social support and positive reframing were the strongest predictors of higher positive mental well-being and lower psychological distress. Self-blame, behavioural disengagement and substance use were the strongest predictors of adverse outcomes" — pmc.ncbi.nlm.nih.gov.
  8. Hoffman PD, Buteau E, Hooley JM, Fruzzetti AE, Bruce ML, "Family members' knowledge about borderline personality disorder: correspondence with their levels of depression, burden, distress, and expressed emotion," Family Process 42(4), 2003, 469–478 — "Contrary to expectation, greater knowledge about BPD was associated with higher levels of family members' burden, distress, depression, and greater hostility toward patients" — doi.org.
  9. Hooley JM, Hoffman PD, "Expressed emotion and clinical outcome in borderline personality disorder," American Journal of Psychiatry 156(10), 1999, 1557–1562 — 35 patients followed one year after discharge; "Contrary to prediction, relatives' criticism and hostility did not predict how well patients did in the year after discharge… Patients whose families scored higher on emotional overinvolvement had better clinical outcomes" — doi.org.
  10. Bailey RC, Grenyer BFS, "The relationship between expressed emotion and wellbeing for families and carers of a relative with borderline personality disorder," Personality and Mental Health 9(1), 2015, 21–32 — 280 carers; criticism in 82.9% and emotional overinvolvement in 69.6%; "Elevated emotional overinvolvement was correlated with higher burden and mental health problems" — doi.org.
  11. Bateman A, Fonagy P, "A randomized controlled trial of a mentalization-based intervention (MBT-FACTS) for families of people with borderline personality disorder," Personality Disorders: Theory, Research, and Treatment 10(1), 2019, 70–79 — 56 family members, five evening meetings delivered by trained family members; "a significantly steeper decline for the immediate-treatment group compared with the delayed-intervention group (β = −1.07, 95% CI [−1.40, −0.74], z = −6.3, p < .000)"; "There were no differences in depression, total anxiety, and total burden" — doi.org.
  12. Grenyer BFS, Bailey RC, Lewis KL, et al., "A randomized controlled trial of group psychoeducation for carers of persons with borderline personality disorder," Journal of Personality Disorders 33(2), 2019, 214–228 — 68 carers; improvements in dyadic adjustment, family empowerment and reduced expressed emotion, sustained at 12 months — doi.org.
  13. Tempia Valenta S, Bortolotti B, Martino F, et al., "Psychoeducation for caregivers of individuals with borderline personality disorder: a randomized controlled trial of multiple family group therapy," Personality and Mental Health 19(3), 2025, e70029 — 57 caregivers — pmc.ncbi.nlm.nih.gov.
  14. Betts JK, Seigerman MR, Hulbert C, et al., "A randomised controlled trial of a psychoeducational group intervention for family and friends of young people with borderline personality disorder features," Australian and New Zealand Journal of Psychiatry 57(11), 2023, 1453–1464 — 79 randomised; "did not find any significant differences between the groups on the primary (d = −0.32…) or secondary outcomes" — pmc.ncbi.nlm.nih.gov.
  15. BPD Alliance (National Education Alliance for Borderline Personality Disorder), "Family Connections" — programme description, formats of 12 weeks of two-hour meetings or asynchronous learning followed by 8 weeks or two six-hour days, "The program is completely free," the participation requirements, and the application and notification-list process — bpdalliance.org.
  16. Guillén V, Fernández-Felipe I, Marco JH, Grau A, Botella C, García-Palacios A, "'Family Connections', a program for relatives of people with borderline personality disorder: a randomized controlled trial," Family Process 63(4), 2024, 2195–2214 — 121 family members, dose-matched active comparator; "Although no statistically significant differences were found between conditions. However, the adjusted posttest means for FC were systematically better than for TAU" — pmc.ncbi.nlm.nih.gov.
  17. Hoffman PD, Fruzzetti AE, Buteau E, et al., "Family Connections: a program for relatives of persons with borderline personality disorder," Family Process 44(2), 2005, 217–225 — uncontrolled pre-post, 44 participants; "significant reductions in grief and burden, and a significant increase in mastery" — doi.org; Lanfredi M, Meloni S, Ferrari C, et al., "Family Connections: the impact of an education program for carers of individuals with borderline personality disorder in Italian mental health services," Family Process 64(1), 2025, e13098 — uncontrolled, 202 analysed — pmc.ncbi.nlm.nih.gov.
  18. Flynn D, Kells M, Joyce M, et al., "Family Connections versus optimised treatment-as-usual for family members of individuals with borderline personality disorder: non-randomised controlled study," Borderline Personality Disorder and Emotion Dysregulation 4, 2017, 18 — the authors' own caveat that the comparison programme "does not constitute a direct comparison group… due to the significant variation in duration between the programmes" — pmc.ncbi.nlm.nih.gov.
  19. Cohen S, Salamin V, Perroud N, et al., "Family Connections in France and Switzerland," Borderline Personality Disorder and Emotion Dysregulation 11, 2024, 16 — 149 participants across five centres, uncontrolled; burden d = −0.48, depression d = −0.36, coping d = 0.53 — pmc.ncbi.nlm.nih.gov.
  20. National Alliance on Mental Illness, "NAMI Family-to-Family" — "a free, 8-session educational program for family, significant others and friends of people with mental health conditions," taught by trained family members — nami.org; NAMI California affiliate finder — namica.org; Dixon LB, Lucksted A, Medoff DR, et al., "Outcomes of a randomized study of a peer-taught Family-to-Family Education Program for mental illness," Psychiatric Services 62(6), 2011, 591–597 — 318 participants; "FTF participants had significantly greater improvements in problem-focused coping as measured by empowerment and illness knowledge… Subjective illness burden did not differ between groups" — pmc.ncbi.nlm.nih.gov.
  21. Kuo JR, Fitzpatrick S, Ip J, Uliaszek A, "The who and what of validation: an experimental examination of validation and invalidation of specific emotions and the moderating effect of emotion dysregulation," Borderline Personality Disorder and Emotion Dysregulation 9, 2022, 15 — 126 participants; "Higher emotion dysregulation was associated with greater increases in self-reported positive emotion when shame or sadness was validated and lesser increases when fear was validated" — pmc.ncbi.nlm.nih.gov.
  22. Fernández-Felipe I, Díaz-García A, Marco JH, García-Palacios A, Guillén V, "Family Connections in relatives of people with borderline personality disorder: a qualitative study," International Journal of Environmental Research and Public Health 19(1), 2021, 79 — "validation and radical acceptance were determined to be the most useful skills" — pmc.ncbi.nlm.nih.gov.
  23. van Schie CC, Lewis K, Barr KR, et al., "Borderline personality disorder and stigma: lived experience perspectives on helpful and hurtful language," Personality and Mental Health 18(3), 2024, 216–226 — 33 consumers and 30 carers; "Both consumers and carers reported helpful language as being connecting, validating and accepting" — doi.org.
  24. National Institute for Health and Care Excellence, "Borderline personality disorder: recognition and management," clinical guideline CG78, recommendation 1.3.7.4 on reviewing crisis plans with families after a crisis — nice.org.uk.
  25. Fruzzetti AE, Shenk C, Hoffman PD, "Family interaction and the development of borderline personality disorder: a transactional model," Development and Psychopathology 17(4), 2005, 1007–1030 — "Although no prospective epidemiological studies have evaluated the relationship between family interactions and the development of borderline personality disorder…"; the model is described as "neither blaming of the family member with BPD nor of her or his parents and caregivers" — doi.org.
  26. Scheirs JGM, Bok S, "Psychological distress in caretakers or relatives of patients with borderline personality disorder," International Journal of Social Psychiatry 53(3), 2007, 195–203 — caregivers "scored higher on all symptom dimensions of the SCL-90 than the general population" — doi.org.

Paid for by participating therapists. Inclusion is computed from availability data — never purchased. No ads, no data sold.