Anxiety, hunger, and eating behavior.

In this study we examined the effects of anxiety and food deprivation on the amount of food con-sumed ad lib by dieters and nondieters. Eighty female college students served as subjects in an osten-sible market research study in which an anxiety manipulation was embedded. Reassignment of the subjects to anxiety condition on the basis of self-reported anxiety produced a significant (p <.02) three-way interaction among level of anxiety, food deprivation, and dieting status. The results suggest that (a) for nondieters, anxiety suppresses hunger but has no effect when subjects are not initially hungry, and (b) for dieters, anxiety increases eating only when the subject is initially hungry. These results are interpreted in terms of Herman and Polivy's (1984) boundary model of eating. Researchers have long viewed the effects of anxiety on eating as depending primarily on the type of person involved. Normal (nonobese, nondieting) people, who are presumably responsive to their physiological state, are expected to react to anxiety or stress by eating less. The physiological effects of stress include the inhibition of gastric contractions and the elevation of blood sugar, both of which ought to suppress hunger. Thus the normal individual, who uses such internal cues (or their correlates) as the basis for the regulation of eating, ought to eat less when stressed. Expectations regarding the effect of stress on overweight peo-ple are more variable. Early psychodynamic views (summa-rized by Kaplan & Kaplan, 1957) held that eating may reduce anxiety for some people; these people, owing to their reliance on eating to assuage distress, will in all likelihood become over-weight. By the same token, we may expect obese people to re-spond to distress by overeating. Bruch (1961) developed an in-teresting variation on this theme. She argued that early misla-beling may lead some people to confuse emotional distress with hunger and to respond to such distress with eating; naturally, responding to an inappropriately wide range of internal cues might well promote obesity.