do all people with personality disorders, or any psychopathology, struggle with mentalizing?
“It turns out that people with personality disorder, and that is across the board, all personality disorder, and in fact all people with psychotherology have difficulties in mentalizing.”
a guest, a clinical psychologist said on The Carlat Psychiatry Podcast, 2026-10-05, at 03:52. hear it in context →
the episode: Treating BPD Series, Episode 4: Mentalization-Based Treatment—From Reacting to Reflecting with Carla Sharp, PhD (at 03:52). we link the show's own page; we do not re-host audio.
first, the part that holds
the core of this holds up. people diagnosed with borderline personality disorder show measurable mentalizing problems, and pooled studies find group-level deficits in many other diagnoses too, including schizophrenia, depression, obsessive-compulsive disorder, bipolar disorder and social anxiety.
what we found when we went and looked
we read the meta-analyses. for borderline personality disorder, one pooled 34 studies and found a very large gap in reflective functioning (d = 1.68) but only a modest gap on theory-of-mind tasks (d = 0.36), with no gap at all in reading emotions from faces and eyes. a 2025 network meta-analysis of 89 studies found deficits in 10 of 11 psychiatric groups compared with healthy controls. the effects were mostly moderate (g about -0.3 to -0.6, with schizophrenia near -1.0), and the one group that showed no deficit was people at familial risk for bipolar disorder. what the record does not support is "all": the personality disorder evidence is almost entirely about borderline personality disorder, plus psychopathic traits, and group averages of this size mean many individual patients score in the normal range.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| people with personality disorder (as studied, mainly borderline personality disorder) have difficulties in mentalizing | confirmed | Bora 2021 (34 studies, 1,448 patients, 2,006 controls): reflective functioning impairment d = 1.68, theory-of-mind impairment d = 0.36. Németh 2018 (17 studies) also found deficits in overall, reasoning and cognitive theory of mind. limitation: both meta-analyses cover only borderline personality disorder, and the size of the deficit depends heavily on how mentalizing is measured. [2][3] |
| this is 'across the board, all personality disorder' | overstated | the pooled evidence we read covers borderline personality disorder and psychopathic traits. Song 2023 (42 studies, 7,463 participants) found psychopathic traits linked to worse theory-of-mind performance, though this is about traits, not a diagnosis. the 11-condition network meta-analysis included no personality disorder except borderline. we found no meta-analysis showing deficits in avoidant, dependent, schizoid or obsessive-compulsive personality disorder. 'all' goes beyond what has been tested. [2][4][5] |
| all people with any psychopathology have difficulties in mentalizing | overstated | Tsui 2025 (89 studies, 9,038 participants) found deficits in 10 of 11 groups, mostly moderate (bipolar g = -0.33, depression -0.43, obsessive-compulsive disorder -0.61, schizophrenia -0.96). the familial-risk-for-bipolar group showed no deficit. Baez 2023 found a theory-of-mind deficit in social anxiety (SMD -0.44) but called generalized anxiety 'inconclusive' because only two studies existed. these are differences in group averages. at g of about -0.4 to -0.6, the patient and control distributions overlap heavily (our own estimate is roughly 75-85%), so this does not show that every individual is impaired, and many disorders have never been tested. [4][6] |
| in borderline personality disorder the difficulty is uniform across kinds of mentalizing | misattributed | the deficit is uneven. Bora 2021 and Németh 2018 both found no deficit in mental-state decoding (reading emotions from eyes and faces). Bora found more hypermentalizing errors (over-reading intentions) and concluded the difficulty reflects a maladaptive 'meta-social-cognitive' style more than a basic loss of social-cognitive ability. calling this simply 'difficulties' hides a profile that includes some intact abilities. [2][3] |
| mentalizing as a transdiagnostic explanation of psychopathology | confirmed | the leading review from the mentalizing research group proposes this directly: mentalizing is 'a helpful transtheoretical and transdiagnostic concept to explain vulnerability to psychopathology.' it also says disorders involve 'different imbalances' in mentalizing, not one shared deficit. limitation: this is the authors' theoretical position, not a measured result. [1] |
both directions
for the speaker: the strongest support is for borderline personality disorder, where a 34-study meta-analysis found a very large reflective functioning deficit (d = 1.68). a 2025 network meta-analysis found significant group-level mentalizing deficits compared with healthy controls in 10 of 11 psychiatric conditions, including schizophrenia, autism, obsessive-compulsive disorder, depression, bipolar disorder and anorexia nervosa. social anxiety disorder shows a theory-of-mind deficit too. psychopathic traits are linked to worse theory-of-mind performance across 42 studies. the review from the mentalizing research group treats mentalizing as a transdiagnostic vulnerability, so the guest's framing matches a mainstream research program.
against: "all personality disorder" is not what has been studied. the meta-analytic evidence is almost entirely about borderline personality disorder, and the 11-condition network meta-analysis included no other personality disorder. even in borderline personality disorder the picture is mixed: the theory-of-mind deficit is modest (d = 0.36), decoding of mental states is intact, and patients make more hypermentalizing errors, which points to a distorted style more than a missing ability. one at-risk group (familial risk for bipolar disorder) showed no deficit, and generalized anxiety had too few studies to say. effect sizes of g about -0.3 to -0.6 describe differences between group averages with heavy overlap, so they cannot support "all people with psychopathology." many diagnoses, such as specific phobias and most substance and sleep disorders, are absent from these syntheses.
what this cannot say
mentalizing is measured in very different ways: interview-rated reflective functioning, self-report questionnaires, cartoon and false-belief tasks, and eye-reading tests. effect sizes swing widely depending on the measure, as the gap between d = 1.68 and d = 0.36 in borderline personality disorder shows. the network meta-analysis covered only tasks with static pictures, and its authors report ceiling effects, substantial heterogeneity, publication bias and poorly reported comorbidity. the studies are cross-sectional, so they cannot show whether mentalizing problems come before a disorder or result from it. our overlap figure is our own approximation from the reported effect sizes, not a number from the papers. we read abstracts and summaries, not every included study. we did not find or read a primary meta-analysis on avoidant, dependent, schizoid, paranoid or obsessive-compulsive personality disorder, so the absence we describe is an absence in what we found, not proof that no such data exist. the podcast quote is a spoken remark, and the guest may have meant the transdiagnostic model, not a literal claim about every individual.
sources, each one fetched and read
- Luyten P, Campbell C, Allison E, Fonagy P. The mentalizing approach to psychopathology: state of the art and future directions. Annual Review of Clinical Psychology, 2020;16:297-325. doi:10.1146/annurev-clinpsy-071919-015355 What we read there: "Research suggests that mental disorders are associated with different imbalances in these dimensions... We suggest that mentalizing is a helpful transtheoretical and transdiagnostic concept to explain vulnerability to psychopathology and its treatment." Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/32023093/
- Bora E. A meta-analysis of theory of mind and 'mentalization' in borderline personality disorder: a true neuro-social-cognitive or meta-social-cognitive impairment? Psychological Medicine, 2021;51(15):2541-2551. PMID 34530941. doi:10.1017/S0033291721003718 What we read there: 34 studies, 1,448 people with borderline personality disorder and 2,006 controls. a very large reflective functioning impairment (d = 1.68), a modest theory-of-mind impairment (d = 0.36), worse performance on theory-of-mind reasoning but not decoding, and more hypermentalizing errors. Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/?term=Bora+meta-analysis+theory+of+mind+mentalization+borderline+personality+disorder+metasocialcognitive
- Németh N, Mátrai P, Hegyi P, et al. Theory of mind disturbances in borderline personality disorder: a meta-analysis. Psychiatry Research, 2018;270:143-153. doi:10.1016/j.psychres.2018.08.049 What we read there: 17 studies, 585 patients and 501 controls. patients underperformed controls in overall theory of mind, mental-state reasoning and cognitive theory of mind, with no deficits in mental-state decoding. comorbid anxiety disorders had a positive moderating effect. Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/30248485/
- Tsui HKH, Lo JCC, Chan SKW. Mentalizing impairments across 11 psychiatric conditions: a transdiagnostic systematic review and network meta-analysis of tasks with static illustrations. European Psychiatry, 2025;69(1):e2. doi:10.1192/j.eurpsy.2025.10146 What we read there: 89 studies, 9,038 participants: "Significant mentalizing deficits were observed across all conditions versus HCs, except for the familial risk for bipolar disorder group." schizophrenia g = -0.960, borderline personality disorder g = -0.612, obsessive-compulsive disorder g = -0.613, depression g = -0.426, bipolar disorder g = -0.326. borderline is the only personality disorder included. Fetched 2026-10-30 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC12816936/
- Song Z, Jones A, Corcoran R, Daly N, Abu-Akel A, Gillespie SM. Psychopathic traits and theory of mind task performance: a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 2023;151:105231. PMID 37172923. doi:10.1016/j.neubiorev.2023.105231 What we read there: 142 effect sizes from 42 studies (n = 7,463): "psychopathic traits are associated with impaired ToM task performance." interpersonal/affective traits were linked to a larger impairment than lifestyle/antisocial traits. Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/?term=psychopathic+traits+theory+of+mind+task+performance+systematic+review+meta-analysis
- Baez S, Tangarife MA, Davila-Mejia G, Trujillo-Güiza M, Forero DA. Performance in emotion recognition and theory of mind tasks in social anxiety and generalized anxiety disorders: a systematic review and meta-analysis. Frontiers in Psychiatry, 2023;14:1192683. doi:10.3389/fpsyt.2023.1192683 What we read there: in social anxiety disorder, theory of mind SMD = -0.44 (95% CI -0.83 to -0.04). "results for GAD were inconclusive due to the limited number of studies meeting the inclusion criteria (two for each domain)." Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/37275989/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam