Anxious or Avoidant? Two Distinct Routes Link Attachment to Eating Disorders

Why do some people who struggle with close relationships go on to develop troubled relationships with food? A new study published in the Journal of Eating Disorders offers one of the most detailed maps yet of the psychological terrain that connects the two, and its central finding is striking: the two main flavors of attachment insecurity appear to travel along entirely different routes toward eating disorder symptoms. Researchers Hwee Jen Lau, Isabel Krug, and Adela-Maria Isvoranu, based at the National University of Singapore and the University of Melbourne, used a statistical technique called network analysis to chart how attachment styles, deeply held negative core beliefs, difficulties regulating emotion, and eating disorder symptoms interconnect within a single system. Their results suggest that the fear of rejection and the discomfort with closeness, long treated as a single fuzzy construct called insecure attachment, may feed into disordered eating through distinguishable psychological machinery.

Attachment theory, first developed by John Bowlby in the mid-twentieth century, holds that the bonds formed with early caregivers create internal working models of relationships that persist into adulthood. When those bonds are inconsistent, cold, or frightening, adults tend to fall into one of two insecure patterns. Attachment anxiety is characterized by a chronic worry that others will abandon or reject you, coupled with an intense craving for reassurance. Attachment avoidance, by contrast, involves a defensive suppression of attachment needs, a preference for emotional distance, and discomfort with depending on anyone. Decades of research have linked both patterns to eating disorders, but the field has struggled to explain how a relationship style translates into restriction, binge eating, or preoccupation with body shape. The new study set out to trace that translation step by step.

The researchers recruited 760 adults, of whom 81.7 percent were women. Within the sample, 120 participants carried a lifetime diagnosis of an eating disorder, while 640 served as controls without such a diagnosis. Each participant completed a battery of self-report questionnaires covering four domains: adult attachment, early maladaptive schemas, emotion dysregulation, and eating disorder symptomatology. Early maladaptive schemas are rigid, self-perpetuating patterns of belief that typically originate in childhood experiences and fall into domains such as disconnection and rejection, impaired autonomy, impaired limits, other-directedness, and overvigilance. Emotion dysregulation, measured with instruments capturing difficulties such as non-acceptance of emotional responses, lack of emotional awareness, lack of emotional clarity, and impulse-control difficulties when upset, represents the failure to manage affective states in a flexible, goal-directed way.

The analytical approach is what sets this study apart from the usual correlational designs. Rather than asking whether one latent variable predicts another, the team estimated an undirected network model in which every measured variable, from attachment anxiety to binge eating to a specific schema, appears as a node, and the statistical associations between pairs of nodes appear as edges. This produces something like a weather map of psychopathology: you can see which variables cluster together, which connections are strongest, and which nodes act as hubs bridging otherwise separate regions of the network. The method also allows researchers to test the accuracy and stability of the estimated network, and the authors report that supplementary analyses covering centrality indices, network accuracy, stability checks, and group difference tests were conducted to verify that the picture was robust.

The headline result concerns the divergent pathways. Attachment anxiety, the anxious preoccupation with rejection, was most closely connected to the early maladaptive schema nodes, and particularly to schemas within the disconnection and rejection domain. These are the beliefs, forged early in life, that one is fundamentally flawed, unlovable, or destined to be abandoned. In other words, people who anxiously fear rejection in their relationships tend also to carry core convictions of defectiveness and abandonment, and in the network these two regions sit tightly adjacent to each other. This alignment makes theoretical sense: schema theory proposes that anxious attachment and disconnection schemas are two expressions of the same early relational wounds, one manifesting in how you relate to others and the other in what you believe about yourself.

Attachment avoidance told a different story. Rather than clustering with the schemas, avoidant attachment showed its strongest associations with the emotion dysregulation nodes, and specifically with lack of emotional awareness and lack of emotional clarity. This pattern fits a well-established picture of avoidant individuals as deactivating strategists who cope with distress by tuning out their inner states. If you learned in childhood that expressing needs drives people away, the adaptive solution is to stop noticing that you have needs at all. The network analysis captures the downstream cost of that strategy: when emotions go unrecognized and unlabeled, they become harder to manage, and the study suggests this deficit in awareness and clarity is the point where avoidance plugs into the wider system.

Perhaps most clinically consequential is where the eating disorder symptoms themselves sit. Across the network, ED symptoms were most strongly connected to emotion dysregulation rather than to attachment or schemas directly. Two edges stood out. Eating concerns, worry and guilt around food and eating, were most tightly linked to non-acceptance of emotional responses, the tendency to judge one’s own feelings as unacceptable and to resist them. Objective binge eating, defined as consuming unusually large amounts of food with a sense of lost control, was most strongly tied to difficulties with impulse control when experiencing negative emotions. These connections sketch a plausible mechanism: for many individuals, disordered eating behaviors may function as maladaptive attempts to regulate or escape emotional states that other strategies fail to contain.

Putting the pieces together, the authors propose that disconnection and rejection schemas on one side, and emotion dysregulation on the other, may be the key features carrying the association between insecure attachment and eating disorder symptoms. The anxious route appears to run through beliefs about rejection and defectiveness, while the avoidant route runs through an inability to identify and understand one’s own emotions, and both routes converge on the emotion regulation problems that sit closest to the symptoms themselves. The team is careful about causality. Because the data are cross-sectional, the network shows associations, not directions, and the authors explicitly call for future longitudinal and experimental studies to test whether these pathways unfold over time and whether they carry clinical significance in treatment contexts.

Even with that caution, the practical implications are considerable. If attachment anxiety and avoidance reach eating disorder symptoms through different intermediate mechanisms, then assessment and case formulation should not treat insecure attachment as a monolith. A patient whose profile is dominated by anxious attachment might benefit most from work targeting abandonment and defectiveness schemas, of the kind central to schema therapy. A patient with avoidant features might first need help building basic emotional awareness and labeling skills, the foundation of many emotion-focused and dialectical behavior therapy approaches. And because ED symptoms cluster most tightly with emotion dysregulation, interventions that directly strengthen emotion regulation skills may be valuable regardless of a patient’s attachment style, potentially cutting the final common pathway that feeds the symptoms themselves.

The study also carries a broader message for how psychology understands mental health. Network analysis is part of a shift away from viewing disorders as single underlying diseases and toward seeing them as systems of interacting symptoms and vulnerabilities, where problems reinforce one another in loops. In this framework, an eating disorder is not simply caused by attachment insecurity; it emerges from a web in which relationship fears, core beliefs, and emotional chaos feed each other. Mapping that web, as this team has done with a sample large enough to include both diagnosed cases and controls, gives clinicians a hypothesis-rich picture of where interventions might break the loops. The authors note that this is the fifth manuscript drawn from the same dataset, but the first to apply network analysis to this particular combination of attachment, schemas, emotion regulation, and eating pathology, and they frame the work as a foundation for the longitudinal and experimental studies that must now follow.

Subject of Research: Network analysis of the relationships between insecure attachment, early maladaptive schemas, emotion dysregulation, and eating disorder symptoms

Article Title: Pathways between insecure attachment and eating disorder symptoms: a network analysis study exploring the roles of early maladaptive schemas and emotion dysregulation

Article References: Lau, H. J., Krug, I., & Isvoranu, A.-M. (2026). Pathways between insecure attachment and eating disorder symptoms: a network analysis study exploring the roles of early maladaptive schemas and emotion dysregulation. Journal of Eating Disorders. https://doi.org/10.1186/s40337-026-01727-2

Image Credits: AI Generated

DOI: 10.1186/s40337-026-01727-2

Keywords: insecure attachment, attachment anxiety, attachment avoidance, early maladaptive schemas, emotion dysregulation, eating disorders, network analysis, binge eating, schema therapy, emotion regulation, psychology, mental health