Subgroups of emotion dysregulation in adults: a latent profile analysis in a clinically heterogeneous population

Even if it was first described in borderline personality disorder, emotion dysregulation (ED) is increasingly recognized as a transdiagnostic and dimensional construct, best understood along a continuum of severity rather than as a discrete category. Prior cluster-analytic studies have consistently explored heterogeneity within this continuum by identifying patient subgroups, but most were conducted in populations defined by specific diagnostic categories, limiting their generalizability. To better capture ED's core components, studies conducted in clinically heterogeneous adult populations are needed.

In this context, we conducted a latent profile analysis on 349 adults referred for ED-related difficulties at a specialized outpatient clinic. Profiles were derived from the six subscales of the Difficulties in Emotion Regulation Scale. Patients were assessed on sociodemographic variables, emotion regulation difficulties and strategies, mentalizing abilities, comorbid psychopathology, and well-being.

A two-class solution was retained. Subgroups did not align with any specific diagnostic category but rather emerged along a severity continuum. Cluster 1 (n = 267) was characterized by more severe ED in every domain, higher use of non-adaptive regulation strategies, lower self-mentalizing abilities, greater comorbid psychopathology, and lower well-being. Cluster 2 (n = 82) displayed comparatively preserved emotion regulation abilities and strategies, higher self-mentalizing, fewer comorbidities, and greater well-being.

Our findings provide further evidence for the existence and characterization of subgroups in patients with ED, reflecting gradations along a severity continuum. While exploratory, they also suggest that core mechanisms targeted by DBT and MBT (i.e., maladaptive emotion regulation strategies and self-mentalizing abilities) may be relevant therapeutic targets across a broader transdiagnostic spectrum.