The Persona That Eats for You: Rejection Sensitivity, Masking, and the Eating Disorder That Steps In When the Performance Collapses

Someone in a group said something about popcorn. She had eaten several cups at a memorial for a relative who loved popcorn, and while she recounted this, two things happened simultaneously: another member calculated whether that amount was too much, and a third began explaining that her compulsive rituals would have prevented her from eating the popcorn at all, because food obtained in certain ways becomes, in her word, “cursed.”

The popcorn was beside the point. What the group was negotiating, in the fifteen minutes that followed, was the question of who gets to eat without performing acceptability while doing it.

The Presentation That Confused Three Diagnoses

The member who described the cursed food had collected diagnoses the way some people collect credentials: one for emotional dysregulation, one for compulsive behavior, and a history of being called “too much” that preceded all of them. What she identified as the most clinically relevant pattern was none of these. It was rejection sensitivity, the intense, fast, disproportionate emotional response to perceived criticism that the ADHD literature has been documenting for two decades and that clinicians working with other neurodivergent populations are only now learning to recognize.

Dodell-Feder and Germine (2018) provided neurological grounding for this experience. The dopamine system differences that characterize ADHD alter how the brain encodes social feedback, making negative social signals register with greater intensity and longer duration than they would in a neurotypical system. This is not a cognitive distortion amenable to restructuring. It is a processing difference that changes the phenomenology of being criticized, evaluated, or observed. Barkley’s (2015) model of ADHD as a disorder of self-regulation rather than attention places rejection sensitivity within a broader framework of executive dysfunction: the same system that struggles to regulate attention also struggles to regulate the emotional response to social threat.

Her developmental history made the mechanism visible. She had been punished for emotional expression throughout childhood. The behaviors that adults identified as problems, the intensity, the persistence, the failure to modulate according to social context, were the behaviors her neurodivergent nervous system produced naturally. The punishment installed a template: authentic self-expression generates social rejection. The template then automated.

What emerged in its place was a set of compulsive avoidance behaviors that looked, on intake paperwork, like OCD. She developed rituals around food procurement designed to prevent the possibility of eating something “wrong” in front of others. The avoidance used OCD’s architecture, but the function was different. The rituals were not driven by intrusive thoughts about contamination. They were driven by the learned expectation that being herself in a social eating context would produce the punishment she had internalized as a child.

Winnicott (1960) would have recognized this immediately. The false self develops when the environment cannot tolerate the infant’s spontaneous gesture. The child learns to present a compliant version that earns approval, and the authentic self goes into hiding. What Winnicott could not have anticipated is the specific form this takes when the spontaneous gesture is neurodivergent: the false self becomes a neurotypical performance, and the authentic self retreats not just from relationships but from basic biological activities like eating. Hull et al. (2017) described this as “putting on my best normal,” capturing the performative quality of autistic camouflaging that Winnicott’s framework predicts but never named.

The Research That Almost Gets There

Cassidy and colleagues (2020) published a finding that should have reorganized how eating disorder programs think about neurodivergence: camouflaging autistic traits predicted eating disorder severity more than autism itself. The effort to mask, not the neurodivergence being masked, was the variable that tracked with clinical deterioration. Mandy and Tchanturia (2015) had prepared the ground for this finding by identifying the overlap between autistic traits and anorexia nervosa, proposing cognitive rigidity and systemizing as shared mechanisms, but the Cassidy finding shifted the explanatory weight from trait to strategy. The problem is not how the brain is organized. The problem is what the person does to hide it.

This maps onto what I presented to the group as the dual-task problem. Social eating requires simultaneous processing of four competing demands: reading social cues, managing conversation, monitoring body signals, and handling food mechanics. For someone whose neurodivergent nervous system already processes sensory and social information at higher cost, adding the fifth demand of masking leaves no bandwidth for the one thing eating disorder recovery depends on: noticing what is happening in the body while food is being consumed.

Lavender et al. (2015) found that difficulties in emotion regulation mediated the relationship between neurodivergence and eating pathology, suggesting that the regulatory burden of managing neurodivergent experience in neurotypical environments is itself a risk factor. The member who had eaten popcorn without shame had done so with family, people she described as safe, people who had already learned about her eating disorder and chose to stay. With family, the masking demand was zero. With the clinician adjusting her meal plan, the masking demand consumed everything.

What the Room Taught Me

Another group member offered something that landed harder than any of the research I had prepared. She said that rejection is not a feeling. It is a judgment she makes about other people’s behavior, a story she constructs about their motives and then mistakes for direct perception. She described years of reactive anger on the road that ended the day she witnessed something that made another driver’s behavior suddenly legible as desperation rather than aggression.

This is mentalizing in Fonagy and Bateman’s (2008) sense: the capacity to recognize one’s own affective state as a mental representation rather than a transparent window onto external reality. Luyten et al. (2020) have argued that mentalizing failures in eating disorders are not global deficits but context-dependent collapses, triggered specifically by attachment-relevant threats. What struck me is that this member arrived at mentalization capacity not through therapy but through accumulated experience and what she called being older. The therapeutic mechanism was time, consequences, and the slow accumulation of disconfirming evidence.

The Jungian frame here is shadow integration. The neurodivergent self that was designated unacceptable in childhood becomes shadow material, disowned and projected. The persona that develops to replace it maintains social connection at the cost of authentic contact. What I watched in the group was shadow material returning not as symptom but as assertion. One member’s insistence on her own worth, stated bluntly and without apology, was not grandiosity. It was the shadow claiming the space the persona had vacated.

A member connected rejection sensitivity to gendered research disparities: the conditions most commonly identified in males have received disproportionate research attention, while presentations more common in female-socialized populations remain under-studied. Lai et al. (2015) documented this gap empirically, finding that females with autism are diagnosed later and less frequently, not because they are less affected but because they mask more effectively. The instruments and diagnostic frameworks built on male presentations mean the masking behaviors that female-socialized neurodivergent people develop are not only more elaborate but less likely to be recognized as masking at all.

The Treatment Question This Raises

Process-based therapy would identify the keystone process here as experiential avoidance driven by rejection sensitivity. The avoidance drives masking. The masking depletes executive function. The depletion reduces recovery capacity. The eating disorder steps in as low-effort coping when everything else has been spent performing normalcy. Hayes, Hofmann, and Ciarrochi (2020) argued that targeting keystone processes rather than diagnostic categories produces more durable change because it interrupts the cascade at its origin rather than its endpoint.

But the member with the “cursed” food and the competing diagnoses raised a harder question. Her compulsive rituals developed to protect her from rejection. Her eating disorder developed inside the rituals. Her rejection sensitivity developed from childhood punishment for neurodivergent behavior. Which layer do you treat first when each one was built to protect against the failure of the one beneath it?

I do not have an answer to that question yet. What I have is a woman scraping yogurt off a foil lid with her finger, calling it a habit from eating alone too much, and the group laughing, and nobody in that room performing anything for the sixty seconds it took her to describe it.