TL;DR: In 1985, Janet Polivy and C. Peter Herman published Breaking the Diet Habit, consolidating a decade of experimental work on cognitive dietary restraint and its behavioral consequences. The research demolished the willpower model of binge eating and proposed restraint itself, rather than a failure of willpower, as the proximal cause of disinhibited eating. The finding has been substantiated across four decades by Eric Stice’s dual-pathway longitudinal work and by the treatment-outcome data from Christopher Fairburn’s CBT-E protocol, which drops dietary restraint early in treatment because the restraint is the mechanism that produces the binge.

The 1985 Paper

In 1985, Janet Polivy and C. Peter Herman, working together at the University of Toronto, published Breaking the Diet Habit: The Natural Weight Alternative, a book-length consolidation of the experimental program they had been running since the mid-1970s. The publication was the culmination of a decade of laboratory work at Toronto. It drew together the counterregulation studies Herman and Mack had begun in 1975, the dietary restraint scale Herman and Polivy had developed in 1980, and the growing behavioral literature on what happened in the laboratory when chronic dieters were given an experimental preload and then invited to eat freely from a taste-test spread. The book made an argument the field had been circling and had not yet named: the attempt to govern eating through cognitive rules is not a partial solution to compulsive eating that more discipline could complete. The attempt to govern eating through cognitive rules is a cause of compulsive eating in its own right.

The claim was a departure. The dominant model of binge eating in the early 1980s was still a willpower model inherited from mid-century weight-control medicine and reinforced by the commercial diet industry. Under the willpower model, a binge episode was evidence that the dieter had insufficient resolve and that the solution was to renew the dieting effort with more structure, more accountability, and more rigor. Polivy and Herman’s experimental data broke the model in one direction. The more cognitive restraint the dieter was imposing, the worse the disinhibited eating became when the restraint collapsed, and the collapse was not an occasional failure of an otherwise sound strategy. The collapse was the predictable behavioral consequence of the strategy itself.

Counterregulation

The experimental finding that carried the argument was counterregulation. In the standard paradigm, Herman and his collaborators would bring participants into the laboratory under a cover story about taste testing, administer a preload (typically a milkshake of known caloric content, or occasionally two milkshakes, or no preload at all), and then invite the participant to sample freely from an array of ice creams and other palatable foods, ostensibly to rate them. The critical measure was how much the participant consumed during the taste test.

Non-dieters behaved as the homeostatic hunger model predicts. The larger the preload, the less they ate from the taste-test spread. Their consumption was regulated downward by the caloric load already in their stomachs.

Chronic dieters behaved in the opposite direction. The larger the preload, the more they ate from the taste-test spread. The preload, rather than suppressing subsequent consumption, appeared to release it. Polivy and Herman proposed that the preload had breached a cognitive dietary rule (the caloric budget for the day, the prohibition on milkshakes, the category of food that was supposed to be off-limits until the weekend), and the breach had collapsed the restraint scaffolding that had been suppressing eating all day. Once the rule had been broken, the rule was gone, and the behavior that the rule had been holding back proceeded without the rule to hold it back.

The mechanism had a name in the cognitive literature: the what-the-hell effect, rendered more formally as counterregulation and later as disinhibition. The everyday phrasing is accurate. Once the day has been ruined by a 400-calorie lapse, the dieter’s reasoning is that an additional 1,200 calories do not, in a meaningful sense, worsen the situation that is already lost, and the behavioral program that had been running all day (do not eat the bread, do not eat the cheese, do not eat until dinner) is replaced by a program whose target is the indulgence the restraint had been promising and denying in equal measure.

The Saturday Night Pattern

The experimental finding predicts a clinical pattern that shows up in consulting rooms across the decades since the 1985 book appeared. A composite patient lives it in the following form.

A man in his late twenties, a software engineer, eats clean from Monday morning through Saturday afternoon. He tracks macros in an app. He keeps categories of food out of his apartment. He considers the program sustainable. By late Saturday afternoon, six days of restraint have accumulated a state his cognitive system is working hard to maintain, and a small breach at 6 pm (two slices of pizza at a colleague’s birthday) triggers a cognitive shift. The day is ruined. The week is ruined. The restraint program that had been running for six days is replaced by a program whose target is the food categories the program had been denying. He leaves the birthday party at 7:30. He stops at a taco truck at 8:10. He stops at a 7-Eleven at 9:45. He is in bed by midnight with a quantity of food consumed that he cannot, from the inside, reconstruct the decision-making for.

On Sunday morning he renews the restraint with more structure, more accountability, and more rigor. The loop will run again by the following Saturday. The clinical error the patient is making is not a failure of willpower. The clinical error is the restraint program itself, which is generating the disinhibited episode it is supposedly preventing.

The Substantiating Literature

The counterregulation findings were not a one-laboratory result. Across the 1990s and 2000s the paradigm was replicated and elaborated in multiple research groups, and restraint theory as a whole was integrated into longitudinal developmental work on the etiology of binge eating and bulimia nervosa.

Eric Stice, working at the University of Texas at Austin and later at the Oregon Research Institute, published the dual-pathway model in the late 1990s and substantiated it across a series of prospective longitudinal studies in adolescent girls. The model proposes two partially independent pathways to binge eating: one through negative affect (which connects to Heatherton and Baumeister’s escape theory) and one through dietary restraint (which operationalizes Polivy and Herman’s claim in a longitudinal behavioral-prediction framework). Stice’s 2001 paper in the Journal of Abnormal Psychology, tracking adolescent girls across an extended follow-up, demonstrated that baseline dietary restraint predicted subsequent onset of binge eating behavior with effect sizes that permitted the restraint pathway to be treated as an established etiological contributor rather than as one hypothesis among many.

Christopher Fairburn’s clinical research program at Oxford produced the treatment-side evidence. Enhanced cognitive-behavioral therapy (CBT-E), the transdiagnostic protocol Fairburn developed across the 1990s and 2000s and published in its definitive form in 2008, drops dietary restraint in the opening phase of treatment. The sequencing is deliberate. Regular structured eating (three meals and two to three planned snacks distributed across the day) is established in the first four sessions, before the body-image work, before the cognitive restructuring, before the relapse-prevention module, because the restraint is the behavioral mechanism that the subsequent clinical work cannot function over the top of. Patients who maintain dietary restraint through the remainder of the protocol do not respond as well as patients who accept the regular-eating structure in the opening phase. The treatment-outcome data is consistent with Polivy and Herman’s mechanism and consistent with Stice’s longitudinal findings.

Why the Willpower Frame Persists Anyway

The experimental and clinical literature has been consistent for four decades. The willpower frame persists, in the popular press and in the commercial diet industry and in some fraction of primary-care consulting rooms, for reasons that are economic rather than evidentiary. The diet industry is a revenue category whose product survives only if the product’s failure is reinterpreted as the consumer’s failure. If the consumer accepts the experimental and clinical finding that the restraint is the mechanism that produces the binge, the consumer has no further reason to buy the next program. The reframing of the restraint-binge loop as a willpower problem preserves the revenue category by relocating the failure from the method to the person.

Patients who have spent a decade or more inside that reframing often arrive at intake with a clinical history the intake interview cannot organize without considerable work, because the patient has been given the vocabulary of discipline and moral accountability for a pattern whose underlying mechanism is behavioral and cognitive and, at the level of treatment, straightforward to name. The restraint is the mechanism. The restraint is what treatment drops first.

What Replaces the Diet

Regular structured eating is distinct from its apparent opposite, which, in the cognitive-behavioral sense, is chaos: unstructured grazing, long gaps followed by large meals, eating organized entirely around impulse and availability. CBT-E does not advocate chaos. It advocates a different kind of structure whose organizing logic is regularity rather than restriction, specifying three meals and two to three planned snacks distributed across the day, no stretches longer than four hours between eating occasions, no categorical bans on food groups, no caloric budgeting, and no moralization of food choices as good or bad.

The Tribole and Resch intuitive eating framework, in its sequenced principles, makes a parallel clinical move. Principle one, rejecting the diet mentality, is a precondition for the remaining work rather than one of ten equally weighted suggestions. Honoring hunger and respecting fullness, principles two and eight, require the dieter to have re-established enough interoceptive contact with hunger and fullness to feel them accurately, which chronic restraint has typically disabled. The ten principles are a sequence. They have to be worked in order.

For the engineer living inside the Saturday-night loop, the clinical implication is that the treatment that would resolve the pattern is not an upgrade of his current restraint program. The treatment is the drop of the restraint program and the substitution of a regular-eating structure in its place. The transition is not comfortable. Patients who have spent a decade restraining often experience the first two to three weeks of regular eating as permission to eat dangerously, and the cognitive work that happens in the first phase of CBT-E is largely the work of tolerating that experience while the behavioral infrastructure establishes itself. The restraint relaxes. The binge frequency declines. The loop closes.

The Clinical Referral

If the pattern above describes something you recognize, the eating disorders assessment provides a structured starting point. The binge eating topic page collects the related clinical writing. The companion piece on escape from self-awareness traces the Heatherton and Baumeister mechanism that runs alongside the restraint pathway in the affect-dysregulation presentation, and the developmental companion on what your eating disorder is trying to become holds the longer-range question the present piece has left implicit.

The engineer closed the tracking app on a Thursday. The first regular-eating week was the worst week of the year. The second was worse than the second week of his last liquid-diet attempt, and the third was the first week in six years in which the Saturday night had not ended at the 7-Eleven.

Sources

  • Herman, C. P., and Mack, D. (1975). Restrained and unrestrained eating. Journal of Personality, 43(4), 647 to 660.
  • Herman, C. P., and Polivy, J. (1980). Restrained eating. In A. J. Stunkard (Ed.), Obesity (pp. 208 to 225). W. B. Saunders.
  • Polivy, J., and Herman, C. P. (1985). Breaking the Diet Habit: The Natural Weight Alternative. Basic Books.
  • Stice, E. (2001). A prospective test of the dual-pathway model of bulimic pathology: Mediating effects of dieting and negative affect. Journal of Abnormal Psychology, 110(1), 124 to 135.
  • Fairburn, C. G. (2008). Cognitive Behavior Therapy and Eating Disorders. Guilford Press.
  • Tribole, E., and Resch, E. (2020). Intuitive Eating: A Revolutionary Anti-Diet Approach (4th ed.). St. Martin’s Essentials.