Overlapping And Differentiating Clinical Features of Autism and Borderline Personality Disorder in Women and People Assigned Female at Birth: A Cross-Sectional Study.

Whether someone was autistic or met criteria for BPD was decided from what participants reported about themselves and from screening questionnaires, not from diagnostic interviews. The authors say interviews would have been more robust. People who looked likely to meet criteria for both conditions were screened out, so these two groups are cleaner than most clinic populations. If those screened out did not in fact have the other diagnosis, screening also made the groups more different from each other than they really are. 102 people took part, most of them educated to degree level or above and most of them White British, so the findings may not carry over to other groups. Most participants were also already diagnosed and under care for one condition only. The combination of questionnaires that sorted 95% of participants into the right group was built and tested on the same 102 people. It has not been tried on a fresh group, so that figure is likely to be better than it would be in practice. No one checked whether the questionnaires are usable by autistic people, where ambiguous, imprecise or metaphorical wording can cause problems. The comparisons with general population scores come from other studies with different samples and designs, and the authors describe those comparisons as tentative. Some possible differences were not measured at all: intense interests and differences in non-verbal communication for autism; emptiness, idealisation and devaluation, and dissociation for BPD.

Funded by a grant from Words That Carry On through the McPin Foundation. The authors declared no conflicts of interest. The study was co-led by an autistic researcher with experience of a BPD diagnosis, co-designed with another autistic researcher and advised by lived-experience and clinician panels, and the authors state a neurodiversity-affirming approach to autistic traits and a critical stance on the BPD diagnostic construct.

Do not read the 95% correct classification as a diagnostic test you can run in clinic. That combination of questionnaires was built from the same 102 people it was tested on, everyone's diagnosis and every questionnaire answer was self-reported, and the study deliberately left out people who looked likely to meet criteria for both conditions. The paper's own comparison with general population samples suggests many of these traits are present to some degree in both groups, so what separates them is how severe and how widespread the difficulties are, not whether they are there at all.