April 30, 2026 Eating Disorders

The Obesity Clinic That Discovered Trauma: Vincent Felitti, the ACE Study, and What Your Binge Is Holding

By Brian Nuckols, MA, LPC-A

LPC-A · Center for Discovery · Gottman Trained · EFT · DBT · Depth Psychology · Pittsburgh, PA

TL;DR: In the late 1980s, Vincent Felitti, running the obesity clinic at Kaiser Permanente in San Diego, could not understand why his highest-outcome patients were the ones who dropped out of the program. The follow-up interviews he conducted across the next two years produced the clinical observations that became, with Robert Anda and the Centers for Disease Control and Prevention, the Adverse Childhood Experiences (ACE) Study. The 1998 paper in the American Journal of Preventive Medicine reorganized the epidemiology of trauma and health in the United States. For binge eating specifically, the study proposed what Felitti’s clinical work had already suggested: the behavior was often functioning as an attempted solution to a regulatory problem whose origin preceded the behavior by decades.

San Diego, 1985

The obesity clinic at Kaiser Permanente in San Diego, housed on the fifth floor of a building whose ground floor was an outpatient internal medicine clinic, ran a medically supervised liquid-diet protocol under Vincent Felitti’s direction beginning in the late 1970s. By the mid-1980s the clinic had accumulated enough patients and enough follow-up time to produce the kind of outcome data that clinic directors report at internal medicine conferences. The outcome data was anomalous. The patients who lost the most weight, fastest, were not the patients who maintained the loss. The patients who lost the most weight, fastest, were disproportionately the ones who left the program before the maintenance phase began, and when Felitti and his staff attempted to track the dropouts, the dropouts were hard to reach, and the ones they did reach could not, or would not, say in ordinary terms why they had quit.

Felitti has told the story of what happened next in several places, most accessibly in his 2002 essay “The Origins of Addiction,” and the version below follows his own account. He began conducting longer clinical interviews with the patients who remained in the program, slowed down the intake protocol, and added a sequence of questions about the patient’s history that the medical-surgical architecture of a weight-management clinic had not previously invited. One of those questions, asked of a female patient in the mid-1980s, produced a disclosure Felitti had not anticipated. He had asked what her weight was when she was sexually active, meaning what weight she had reached before her current presenting weight. The patient, at the word sexually, disclosed a childhood rape.

The disclosure was not unique. As Felitti continued the interviewing protocol, he accumulated enough disclosures of childhood sexual abuse, physical abuse, and household dysfunction that the pattern could no longer be treated as incidental to the clinical presentation. The clinic was not treating the presenting problem. The clinic was intervening, without knowing it was intervening, on a regulatory pattern that a portion of its patients had constructed in childhood and had maintained through a large adult body that the clinic was now successfully disassembling. When the body shrank, the pattern re-emerged without its defensive infrastructure, and the patient quit the program before the reconstruction could be completed in less defended form.

The Insight Felitti Stated Plainly

Felitti stated the insight plainly in his subsequent writing. The clinic had been operating under the assumption that obesity was the problem. What the interviews were indicating was that the behavior the clinic was treating was not the problem but the attempted solution, and that the underlying problem was a regulatory task the patient had been performing since childhood, through a combination of behavioral and physiological means, to manage an affective and somatic load that the earliest years of the patient’s life had deposited and that the subsequent decades had not yet metabolized.

The clinical implication reorganized the internal logic of the clinic. A patient who had used compulsive eating and a larger body as a regulatory infrastructure for thirty years did not arrive at the clinic as a behaviorally compliant candidate for a caloric-reduction protocol. The patient arrived with a working regulatory infrastructure that was, at the level of subjective experience, the only functional thing a difficult history had managed to produce, and the clinic was now proposing to disassemble it without substituting anything in its place. The disassembly worked, behaviorally, for a time. Then it stopped working, because the regulatory task the infrastructure had been performing was still there and now had no performer.

This is the finding the clinical literature had not previously organized. The Heatherton and Baumeister escape-theory work of 1991 and the Haedt-Matt and Keel EMA meta-analysis of 2011 substantiated the affect-regulation mechanism at the level of proximate behavior. Felitti’s clinic work substantiated a developmental-epidemiological version of the same finding at the level of childhood origin. The binge was doing regulatory work. The regulatory work had been assigned to the binge, in many patients, decades before the clinic saw them.

The Study That Followed

Felitti carried the clinical observations to Robert Anda, then an epidemiologist at the Centers for Disease Control and Prevention, and the collaboration that began in 1995 produced what has since been called the largest and most important epidemiological study of childhood adversity and adult health ever undertaken. The ACE Study enrolled 17,421 adult members of Kaiser Permanente in San Diego who were undergoing routine physical examinations between 1995 and 1997. The study administered a ten-item questionnaire about adverse childhood experiences across three categories (abuse, neglect, household dysfunction) and linked the questionnaire data to the participants’ detailed medical records.

The findings, published in 1998 in the American Journal of Preventive Medicine under the title “Relationship of Childhood Abuse and Household Dysfunction to Many of the Leading Causes of Death in Adults,” established a dose-response relationship between the number of adverse childhood experience categories endorsed and the prevalence of a wide range of adult health outcomes. The outcomes included the ones the medical system had been tracking for decades (heart disease, liver disease, cancer, chronic pulmonary disease, early mortality) and the ones the medical system had been tracking separately in different clinics (depression, suicide attempts, substance use disorders, sexually transmitted infections, disordered eating). The relationship between ACE count and outcome was not merely statistically significant. It was clinically striking across categories, consistent in direction, and large in effect size for an epidemiological study of adult chronic disease.

For binge eating and body size specifically, the relationship was pronounced. Patients who endorsed four or more ACE categories had substantially elevated prevalence of binge eating patterns and significantly elevated prevalence of body sizes that the medical system at the time was categorizing as obese. The prevalence did not distribute itself as an isolated behavioral risk factor. It distributed itself as part of a broader developmental pattern in which a range of regulatory behaviors (binge eating, substance use, risky sexual behavior, self-harm) appeared in populations whose childhood adversity burden was dose-responsively elevated, which is the pattern Felitti’s clinic had already been tracking at small scale.

What the Study Did Not Say

The ACE Study is often read, in the popular press and in trauma-informed training programs, as if it said that adverse childhood experience causes adult pathology. The study did not say that, and the distinction matters clinically. What the study established was correlational and dose-responsive at the population level, which is to say that a substantial number of adults with high ACE scores do not develop the outcomes in question and a substantial number of adults with low ACE scores develop them, even as the aggregate correlation remains large enough to anchor an epidemiological claim. The study’s authors have been careful to state this, and the popular reading that treats ACE scores as determinative of adult pathology overstates a finding whose clinical value lies in the identification of an elevated-risk population, not in the prediction of an individual trajectory.

The clinical application of the ACE framework to binge eating follows the same caution. A patient with a high ACE score and a binge eating pattern has, at the level of population epidemiology, a meaningful probability that the two are related through a regulatory-functional link of the sort Felitti described. The link is not certain. The clinical work of establishing whether the link is present in a specific patient is the work of the intake interview, the formulation, and the ongoing case conceptualization, not the work of a score interpreted on its own.

What the Binge Is Holding

The phrase “what your binge is holding” is a clinical shorthand rather than a poetic flourish, and the shorthand compresses what Felitti’s clinic work and the subsequent ACE literature and the affect-regulation research have collectively substantiated. A binge eating pattern in a patient with a trauma history is frequently, though not universally, performing regulatory work whose origin predates the pattern. The work includes the attentional downregulation Heatherton and Baumeister described in 1991 (a shift from aversive self-evaluation to narrow sensory focus) and the affective modulation the Haedt-Matt and Keel meta-analysis confirmed in 2011 (negative affect rising before a binge and partially attenuating after). In trauma-shaped presentations, the regulatory work extends further. The binge is modulating not only the proximal affective state but a nervous-system configuration whose set points were established in childhood and whose activation patterns have been rehearsed for years. Peter Levine’s somatic-experiencing framework, elaborated across his clinical writing beginning with Waking the Tiger in 1997, describes the somatic and autonomic dimension of this pattern in terms that the cognitive-behavioral literature tends to leave implicit. The nervous system has a history. The binge, in a patient with adverse childhood experience, is operating on a nervous system whose history the binge has been helping to manage for a long time.

A composite patient illustrates the clinical face of the finding. A forty-four-year-old high-school English teacher presents with a fifteen-year history of binge eating, a completed course of trauma-focused therapy (two years, weekly, with a specialist), and the specific clinical fact that the binge pattern did not meaningfully shift during or after the trauma treatment. Her ACE score is seven. She has read the trauma literature widely and has not yet encountered a clinical frame that connects what her trauma therapist treated to what her eating pattern has been doing, and the disconnection has organized a self-reading in which she has failed trauma therapy rather than been offered a treatment whose design matches her presentation.

The frame she has not encountered is the frame Felitti stated plainly at Kaiser in the late 1980s. The binge has been performing regulatory work the trauma therapy did not redistribute, because the trauma therapy was not designed to redistribute it, which means the clinical implication is not that her trauma therapy failed but that the binge and the trauma are clinically continuous and a treatment targeting one without targeting the other will produce partial results whose partiality the patient is then left to interpret in the worst possible light.

Integrated Treatment, Structurally

The treatment model that matches the finding is integrated care rather than parallel care. Parallel care is the current default, in which a trauma clinician and an eating-disorder clinician see the same patient without shared formulation and without coordinated intervention. Integrated care, whether delivered by one clinician trained in both modalities or by two clinicians in active case conferencing, treats the presenting pattern as a single preserved regulatory architecture with two faces rather than as two separate disorders whose treatments can be scheduled on alternate Tuesdays.

The specific structural features of integrated BED-and-trauma treatment include a shared formulation produced collaboratively at intake, coordinated pacing (trauma processing work is sequenced with eating-pattern stabilization rather than run in parallel), explicit attention to the nervous-system dimension of the binge cue (window of tolerance, activation patterns, the specific somatic signature of the pre-binge state), and a treatment plan that names the affect-regulation function of the binge as a clinical target rather than as a symptom to be eliminated. The dialectical behavior therapy adaptation for binge eating disorder developed by Safer, Telch, and Chen at Stanford in the early 2000s is one of the protocols that integrates naturally with a trauma-informed case conceptualization, because its theoretical target (the affect-regulation function of the behavior) is continuous with the trauma-regulation framework the ACE literature substantiates.

For the English teacher, the integrated case conceptualization would name the fifteen-year binge pattern as a preserved regulatory infrastructure whose function her trauma therapy had not yet redistributed, and the treatment plan would sequence the redistribution across months of careful clinical work whose pacing she would participate in designing. The clinical target would not be weight loss. The clinical target would be the capacity to meet the affect the binge has been managing, with sufficient interpersonal and skills-based support that the meeting did not require the regulatory function the binge has been performing for her since she was twelve.

The Clinic That Never Called Back

Most of Felitti’s original clinic dropouts never called back. A smaller number did. The ones who did, after the clinical interview protocol had added the relevant questions, were among the patients whose cases Felitti describes in his subsequent writing as the ones who taught him what he had not previously understood about the clinic he was running. The patients had not been noncompliant. They had been responding accurately to a clinical intervention whose design had not contemplated the function their body size was performing.

A version of that misreading is still in circulation. Patients with high ACE burdens and binge eating patterns arrive at weight-management clinics, bariatric surgery consultations, and pharmacological-weight-loss programs that are behaviorally effective at disassembling the regulatory infrastructure and that have, in aggregate, no coherent plan for what the patient is supposed to do when the infrastructure is gone and the underlying regulatory task is still there. Some proportion of those patients will quit the program for the reasons Felitti’s patients quit. Some proportion will complete the intervention and rebuild the infrastructure in some other form, because the regulatory work has to be done somewhere and the clinic has not offered a different place to do it. Some proportion will enter a trauma treatment afterward and do the slower work the clinic did not know it was asking them to do.

If you are reading this and recognizing the pattern in your own history, the eating disorders assessment provides a structured starting point and does not ask about body size. The mood screener gives a structured snapshot of where the affective symptoms are sitting. The binge eating topic page collects what has been written about the condition across voices and angles, and the companion piece on escape from self-awareness traces the proximate affective mechanism that runs underneath the developmental pattern this piece has named.

Felitti is now in his late eighties. The clinic he ran on the fifth floor closed years ago. The finding it produced is still most of the way through the work of being absorbed by the treatment systems the finding would reorganize if they absorbed it. The English teacher has an intake scheduled for the eleventh of the month. She has not yet told the intake clinician about the cereal box on the kitchen floor. She will, eventually, on the Tuesday of the third session.

Sources

Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., Koss, M. P., and Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4), 245 to 258.
Felitti, V. J. (2002). The origins of addiction: Evidence from the Adverse Childhood Experiences Study. Praxis der Kinderpsychologie und Kinderpsychiatrie, 52, 547 to 559.
Heatherton, T. F., and Baumeister, R. F. (1991). Binge eating as escape from self-awareness. Psychological Bulletin, 110(1), 86 to 108.
Haedt-Matt, A. A., and Keel, P. K. (2011). Revisiting the affect regulation model of binge eating: A meta-analysis of studies using ecological momentary assessment. Psychological Bulletin, 137(4), 660 to 681.
Levine, P. A. (1997). Waking the Tiger: Healing Trauma. North Atlantic Books.
Safer, D. L., Telch, C. F., and Chen, E. Y. (2009). Dialectical Behavior Therapy for Binge Eating and Bulimia. Guilford Press.

Frequently Asked Questions

What is an ACE score, and what does it measure?

An ACE score is a count, from zero to ten, of the number of distinct categories of adverse childhood experience a person endorses on the ACE questionnaire developed by Felitti and Anda in the mid-1990s at Kaiser Permanente and the Centers for Disease Control and Prevention. The ten categories cluster into three groups: abuse (emotional, physical, sexual), neglect (emotional, physical), and household dysfunction (parental substance use, parental mental illness, domestic violence witnessed, parental separation or divorce, incarcerated household member). The score is a rough epidemiological instrument rather than a clinical diagnosis. It tells you how many distinct categories were present, not how severe any category was, and its predictive value is at the population level. An ACE score of four is associated with substantially elevated risk for a range of adult health outcomes, including depression, substance use disorders, cardiovascular disease, and disordered eating. The association is correlational, not deterministic.

Why did obesity clinics historically miss the trauma layer?

Because the trauma layer was not the question the clinics were asked to answer. The medical-surgical architecture of weight-management medicine, from the liquid-diet programs of the 1970s and 1980s through the bariatric programs that replaced them, organized its clinical work around the behavioral-mechanical question of how to reduce body size rather than the epidemiological question of what the body was doing. Felitti's Kaiser clinic was built on the same assumption. What disrupted the assumption, in Felitti's case, was a specific clinical anomaly (the highest-outcome patients dropping out of the program at disproportionate rates) that he pursued through follow-up interviews rather than dismissing as noncompliance. Most clinics did not have the follow-up infrastructure, the administrative permission, or the clinical orientation to pursue the anomaly the way Felitti did. The missing was structural, not individual.

What does trauma-informed treatment for binge eating actually look like?

It looks like integrated care rather than parallel care. A patient presenting with a trauma history and a binge eating pattern is frequently referred to two clinicians who do not coordinate: a trauma therapist working on the historical material and an eating-disorder clinician working on the behavioral pattern. The integration model, advocated by clinicians working at the intersection of complex trauma and eating disorders, holds that the historical and behavioral tracks are not two problems requiring two treatments but one preserved regulatory pattern whose two faces cannot be clinically separated without losing therapeutic purchase on either. In practice, integrated treatment usually involves one clinician trained in both modalities, or two clinicians in active case conferencing with shared formulation. The somatic-experiencing lineage of Peter Levine, which treats the nervous-system dimension of trauma rather than its narrative content alone, is one of several frameworks that integrate naturally with the affect-regulation model of binge eating that Safer, Telch, and Chen developed in their DBT-BED protocol.

Talk to someone who gets it →

Questions about insurance or cost? See fees and insurance info