Who Gets Labeled 'Borderline'? What Research Shows
There's a pattern I see constantly in consultation, supervision, and case review: a client gets labeled "borderline," and from that point forward, the diagnosis does more work than the clinical picture does. It shapes how the next provider reads their chart before that provider ever meets them. It shapes how a crisis is interpreted. It shapes whether a client's anger gets treated as valid feedback or as a symptom to manage. Few clinicians seem to question this diagnosis after arrival…
I'm not writing this to argue that borderline personality disorder isn't real, or that no one meets criteria for it. I'm writing this because the research on who actually receives this diagnosis tells a much more troubling story than most of us were trained to expect.
Who actually gets labeled "borderline"
Look at the pattern in the research and you start to see who ends up with this diagnosis attached to their chart, and it isn't simply a matter of who meets criteria.
A 2024 study gave psychiatrists identical case vignettes and varied only the gender listed on the intake. Female cases presenting with antisocial personality traits were misdiagnosed with BPD at roughly five times the rate of male cases with the same presentation, which were far more likely to be correctly identified as antisocial personality disorder instead (Özel et al., 2024). Same behavior. Different label, depending on what gender is on the paperwork.
That bias isn't limited to gender. A 2024 study using vignette-based experiments with 426 mental health practitioners across the US and Canada found that practitioners were significantly more likely to diagnose BPD when a case vignette described a transgender client than when the identical case described a cisgender heterosexual client (odds ratio = 1.99) (Rodriguez-Seijas et al., 2024).
None of this is a personality disorder. It's a client responding reasonably to an unreasonable amount of pain, evaluated through a diagnostic system that reads gender and identity as much as it reads symptoms. But once the label goes in the chart, that context disappears. What's left is a word that every future provider will read before they read the person.
The label as documentation of clinician frustration
This is the part that doesn't get said out loud often enough: personality disorder diagnoses, especially BPD, frequently function as documentation of the clinician's experience of the client rather than an accurate clinical formulation. A client is hard to work with. A client doesn't respond to the intervention as expected. A client's presentation doesn't resolve neatly. The chart needs to explain why treatment isn't progressing, and "borderline" becomes the explanation that puts the difficulty back on the client instead of on the fit between the client and the treatment approach.
This matters because diagnosis isn't neutral. It's a document that follows someone through a system that often treats providers' prior notes as more reliable than the person sitting in front of them.
What this looks like for autistic and neurodivergent clients
Neurodivergent clients, particularly autistic women and people assigned female at birth, get swept into this pattern constantly, and the research on this is specific and damning. A 2024 phenomenological study interviewed autistic adults who had previously been diagnosed with BPD and found that participants consistently described unnoticed autistic traits from childhood and lasting harm from a misdiagnosis that led to inappropriate, sometimes actively harmful treatment (Tamilson et al., 2024).
What makes this especially hard to justify is that the misdiagnosis isn't inevitable. A 2026 cross-sectional study comparing autistic women and women diagnosed with BPD found that the two conditions can be distinguished with roughly 95% accuracy using the right combination of measures, with sensory processing differences representing the single biggest distinguishing factor (Barnicot et al., 2026). The tools to tell these apart exist. They're just not being used consistently.
This is the same mechanism I write about in the double empathy problem: the assumption that the clinician's read of a client's behavior is objective, when it's actually shaped by the clinician's own frame of reference. Emotional intensity that's a nervous system response gets read as manipulation. A meltdown or shutdown that's a regulation response gets read as a personality trait. A "borderline" label often means the clinician couldn't make sense of the client, not that the client meets diagnostic criteria.
Race breaks the pattern in the opposite direction
Here's where the research gets more complicated, and where I think the most important point actually lives. Race doesn't follow the same "over-labeled" pattern as gender, identity, and neurotype. A systematic review and meta-analysis of personality disorder research found that Black patients are diagnosed with BPD at significantly lower rates than white patients, and the review's authors note this likely reflects personality disorder diagnoses being overlooked in Black patients' care rather than an actual lower prevalence of the underlying presentation (McGilloway et al., 2010).
That underdiagnosis doesn't mean Black patients' distress goes unlabeled. A 2023 study analyzing national ambulatory care data found that Black patients had nearly twice the odds of receiving a schizophrenia diagnosis compared to white patients with similar presentations (Bazargan-Hejazi et al., 2023).
Put those two findings together and a pattern emerges that's different from, but just as troubling as, the gender and identity bias above: the same underlying emotional dysregulation and distress that gets labeled "borderline" in a white client more often gets funneled toward a more severe, more stigmatizing, more coercive diagnostic category when the client is Black. Same bias. Different mechanism, depending on who's sitting across from you.
What to do with this in practice
You don't need to stop using the diagnosis where it's accurate. What you can do is get curious before you write it down:
Would I read this same presentation differently if the client's gender, race, or gender identity were different?
Have I ruled out autism, especially in a client who's spent years unknowingly masking?
Am I documenting a pattern of behavior across contexts and time, or a handful of hard sessions?
If this client were less able to advocate for themselves, would this diagnosis be more likely to stick, or would their distress more likely get redirected somewhere else entirely?
That last question is something to chew on for a minute. A diagnosis is supposed to help the next clinician understand the client. Too often, it does the opposite. It tells the next clinician what to expect before the client has said a word, and the research now shows that what gets predicted depends heavily on who's in the room.
If you want a deeper structure for thinking through documentation, power, and client autonomy, that's exactly what the Responsive Boundaries course walks through, including a framework for telling the difference between clinical necessity and clinician self-protection in your notes.
[Explore Responsive Boundaries →]
References
Barnicot, K., Thompson, E., Turner, S., Mandy, W., McCabe, R., Stark, E., & Parker, J. (2026). Overlapping and differentiating clinical features of autism and borderline personality disorder in women and people assigned female at birth: A cross-sectional study. Autism. Advance online publication. https://doi.org/10.1177/13623613261431309
Bazargan-Hejazi, S., Shirazi, A., Hampton, D., Pan, D., Askharinam, D., Shaheen, M., Ebrahim, G., & Shervington, D. (2023). Examining racial disparity in psychotic disorders related ambulatory care visits: An observational study using national ambulatory medical care survey 2010–2015. BMC Psychiatry, 23, Article 601. https://doi.org/10.1186/s12888-023-05095-y
McGilloway, A., Hall, R. E., Lee, T., & Bhui, K. S. (2010). A systematic review of personality disorder, race and ethnicity: Prevalence, aetiology and treatment. BMC Psychiatry, 10, Article 33. https://doi.org/10.1186/1471-244X-10-33
Özel, B., Karakaya, E., Köksal, F., Altinoz, A. E., & Yilmaz-Karaman, I. G. (2024). Gender bias of antisocial and borderline personality disorders among psychiatrists. Archives of Women's Mental Health, 28(3), 563–571. https://doi.org/10.1007/s00737-024-01519-0
Rodriguez-Seijas, C., Warren, M., Vupputuri, P., & Hawthorne, S. (2024). Bias in the diagnosis of borderline personality disorder among sexual- and gender-minority persons: Results from a vignette-based experiment. Clinical Psychological Science, 13(2), 388–406. https://doi.org/10.1177/21677026241267954
Tamilson, B., Eccles, J. A., & Shaw, S. C. K. (2024). The experiences of autistic adults who were previously diagnosed with borderline or emotionally unstable personality disorder: A phenomenological study. Autism, 29(2), 502–514. https://doi.org/10.1177/13623613241276073