Restrained Eating

What it is

The term comes from the work of Peter Herman and his colleagues in the 1970s. In Herman and Mack's 1975 study, participants drank a milkshake preload and then tasted ice cream. For low-restraint participants, intake varied inversely with the size of the preload. For high-restraint participants, it varied directly: the more milkshake they had drunk, the more ice cream they ate.

Polivy and Herman's 2020 review suggests one reason. Restrained eaters may be more likely to judge a given amount of food as overeating, especially foods they have forbidden themselves, so a preload is more likely to register as the diet already broken.

In effect

Restrained eating is the population in which the what-the-hell effect was first found, and it links eating research to the abstinence violation effect in relapse prevention. Both describe the same turn: a goal kept as a clean record, a slip that breaks the record, and a conclusion that the rest no longer counts.

The work has practical implications for how diets are framed. Claire Adams and Mark Leary found in 2007 that among highly restrictive eaters, a brief self-compassion exercise after a diet-breaking snack reduced later eating.

What it does not say

Restrained eating describes a pattern of dieting, not a clinical diagnosis, and this page gives no advice about eating disorders. The laboratory findings also have limits: when the milkshake preload was doubled to 30 ounces, restrained and unrestrained eaters alike ate very little afterwards, so the pattern does not grow without bound.

---