People Diagnosed with Personality Disorder Face Smaller, More Isolated Social Networks, Review Finds

One of the most persistent and quietly devastating features of living with a diagnosis of ‘personality disorder’ may not be the emotional turmoil that defines the clinical picture, but the profound social emptiness that often surrounds it. A new systematic review published in BMC Psychiatry has pulled together decades of quantitative research to answer a deceptively simple question: how socially isolated are people who have received this diagnosis, and what do their social networks actually look like? The answer, drawn from 29 studies involving 6,225 participants, is that people with a ‘personality disorder’ diagnosis tend to have less social contact, smaller networks, and fewer friendships than the general population, although their isolation looks strikingly similar to that seen in other mental health conditions. The finding challenges both the stereotype that social difficulties are unique to this diagnostic group and the assumption that they are well understood.

The review, led by Léonie Richardson of the Division of Psychiatry at University College London and the University of Glasgow, together with colleagues including Sarah Ikhtabi, Eiluned Pearce, Alexandra Pitman and Sonia Johnson, set out to map a landscape that has remained surprisingly under-described. Despite decades of clinical observation that people diagnosed with ‘personality disorder’ struggle to form and maintain social connections, the precise structural characteristics of their social networks—their size, diversity, composition, and the frequency of contact within them—had never been systematically synthesised. The team searched four major electronic databases, PsycINFO, Embase, Web of Science and Medline, from their inception to April 2025, capturing quantitative studies that compared people with the diagnosis against other clinical groups and against the general population.

Methodologically, the review followed established standards for evidence synthesis. Each study was appraised using the Joanna Briggs Institute critical appraisal tools, which examine sources of bias in study design, sampling and measurement. The certainty of the body of evidence was then graded using the GRADE approach—Grading of Recommendation, Assessment, Development and Evaluation—which rates confidence in findings from ‘high’ down to ‘very low’ depending on factors such as risk of bias, inconsistency, indirectness and imprecision. That rigour matters here, because the headline certainty rating the authors assigned to their findings was ‘very low’, a caveat that shapes how the results should be interpreted and underscores how much high-quality research remains to be done.

The 29 included studies were dominated by cross-sectional designs, the kind of snapshot research that measures a group at a single point in time, and most were conducted in the United States. This matters for interpretation. Cross-sectional studies can reveal associations but cannot establish whether social isolation precedes the diagnosis, follows from it, or arises from shared underlying causes. The geographic concentration in the US raises further questions about how findings might generalise to countries with different healthcare systems, welfare arrangements and cultural expectations around friendship and family obligation. Nevertheless, the consistency of certain patterns across studies gives the synthesis its weight.

The clearest contrast emerged between people diagnosed with ‘personality disorder’ and the general population. Eleven studies, encompassing 1,348 participants, found that the diagnosed group had reduced social contact and smaller, less diverse social networks than people without mental health conditions. Network diversity is a particularly telling metric in social neuroscience: a healthy network typically includes a mix of family, friends, colleagues and acquaintances, each serving different emotional and practical functions. A network shrunk to a few contacts, or dominated by a single category such as immediate family or professional carers, is more fragile and offers less buffering against stress. The review suggests this kind of structural thinning is a real feature of life for many people carrying the diagnosis.

Yet there was a striking exception buried in the data. Three studies, including 1,349 participants, found that people diagnosed specifically with ‘avoidant personality disorder’ showed no differences from the general population in these network measures. The finding is counterintuitive on its face—avoidance is the defining feature of that diagnosis—but it may reflect how social networks are measured, or the possibility that people with avoidant patterns maintain a small but stable set of connections rather than the broader, more volatile networks seen in other presentations. It is also possible that the pooled samples in those three studies were drawn differently, and the authors themselves note this subgroup finding as an area requiring careful further investigation rather than a settled conclusion.

When the comparison shifted to other clinical groups, the picture changed again. Eight studies, covering 892 participants, indicated that people diagnosed with ‘personality disorder’ show similar degrees of social isolation and comparable network characteristics to people with other mental health conditions. This is arguably the review’s most consequential finding for clinical practice and public discourse. Social isolation is not a signature of ‘personality disorder’ alone; it is a common companion of serious mental ill health more broadly. That framing matters because the ‘personality disorder’ label has historically carried heavy stigma, with social difficulties often attributed to the individual’s character or interpersonal style rather than recognised as a shared consequence of mental illness, poverty, exclusion and the disruptions that conditions impose on education, employment and relationships.

A further thread of evidence came from four studies, with 695 participants and mixed quality ratings, which found that people with the diagnosis reported fewer friendships and fewer supportive relationships than both general population samples and other clinical groups. Supportive relationships—the people one can turn to in crisis, confide in, or rely on for practical help—are among the strongest known predictors of recovery and physical health. Their scarcity in this group suggests that even where the sheer number of social contacts resembles that of other patient populations, the felt quality of those connections may be diminished. The review’s authors argue that future quantitative and qualitative research should focus on identifying the barriers that prevent people with the diagnosis from developing and maintaining meaningful, satisfying relationships, including friendships and intimate partnerships, and on how those barriers can be overcome.

Why does this matter beyond the clinic? Loneliness and social isolation have emerged over the past decade as major public health concerns, with associations to cardiovascular disease, cognitive decline and premature mortality that rival well-known behavioural risk factors. If people diagnosed with ‘personality disorder’ experience structural network deficits—fewer contacts, less diversity, less support—then any treatment plan that addresses symptoms while ignoring the social architecture of the person’s life is treating only half the problem. The review was explicitly motivated by this gap: the authors aimed to synthesise the quantitative literature in order to inform intervention strategies targeting the specific social needs of this group. Social network analysis, once the preserve of sociology, is increasingly seen as a clinical instrument, and this review supplies the baseline measurements that intervention designers have lacked.

The limitations are as instructive as the findings. With GRADE certainty rated ‘very low’, the synthesis is best understood as a well-mapped starting point rather than a definitive verdict. The reliance on cross-sectional data, the dominance of US samples, and the heterogeneity of measures used to define isolation and network size all constrain what can be concluded. The review also excluded qualitative work, meaning the lived experience of isolation—how it feels, how it develops, how people cope—remains outside its scope, though the authors call for qualitative research to complement the numbers. Funded in part through the UK’s Loneliness and Social Isolation in Mental Health Research Network, supported by UK Research and Innovation and the National Institute for Health and Care Research, the study signals a growing institutional commitment to treating social connection as a core outcome in mental health care. What the review establishes beyond reasonable doubt is that the deficit is real, measurable, and shared across diagnostic boundaries—and that understanding why, and how to reverse it, is now one of the most pressing questions in social psychiatry.

Subject of Research: Social isolation and social network characteristics in people diagnosed with personality disorder

Article Title: Objective social isolation and social network characteristics of people who have received a diagnosis of ‘personality disorder’: a systematic review of quantitative studies

Article References: Richardson, L., Ikhtabi, S., Liguori, A., Wang, Y., Pearce, E., Lee, C. Y., Pitman, A., & Johnson, S. (2026). Objective social isolation and social network characteristics of people who have received a diagnosis of ‘personality disorder’: a systematic review of quantitative studies. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08662-1

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08662-1

Keywords: personality disorder, social isolation, social networks, systematic review, loneliness, mental health, psychiatry, friendships, GRADE, BMC Psychiatry, social support, avoidant personality disorder