One of the most important advancements in the understanding of eating disorders is the recognition that many of the symptoms once thought to be primary features of anorexia nervosa are actually symptoms of starvation. An experimental study, conducted and published 50 years ago by Ancel Keys and his colleagues at the University of Minnesota (Keys, Brozek, Henschel, Mickelsen & Taylor, 1950) is the best example of the wide-ranging physical, cognitive, social and behavioural effects of starvation, initially thought to be the symptoms of anorexia nervosa. The subjects of the study were 36 healthy young men who volunteered as an alternative to military service.

During the first three months of the study, the volunteers ate normally while their personality, behaviour and eating patterns were studied. Over the next six months, the men decreased their food intake by about half, losing on average about a quarter of their former weight. This was followed by three months of rehabilitation where gradual refeeding occurred.

Each volunteer responded differently but there were dramatic physical, psychological and social changes, many of which persisted even during the refeeding stage. The significance of this study (commonly known as the “Starvation Study”) is that the experiences displayed by the volunteers due to starvation are the same symptoms experienced by people with an eating disorder.

The volunteers became overwhelmingly preoccupied with food, plagued by incessant thoughts of food and eating, resulting in poor concentration in other areas. Food became the topic of conversation, reading and daydreams. They played with their food, and there was an over consumption of salt, spices, coffee, tea, and gum.

As starvation progressed, the volunteers increasingly hoarded food and food-related items, such as cookbooks and menus. This tendency to hoard is often observed in persons with anorexia nervosa. During refeeding, most of the abnormal attitudes and behaviours persisted. 40% of volunteers mentioned cooking as an interest post-experiment and some even changed occupations, three becoming chefs and one went into agriculture.

All volunteers reported increased hunger in the semi-starvation phase. Some were able to tolerate this while others were could not control themselves, partaking in episodes of binge eating followed by self-deprecation and disgust, and in some cases vomiting. No factors could be identified to distinguish reasons for these differences. After five months of refeeding, overcompensation of eating persisted, but by eight months of refeeding most had returned to normal eating habits.

The previously psychologically healthy volunteers suffered significant emotional deterioration as a result of semi-starvation. These symptoms included severe depression and extreme mood swings, anxiety symptoms such as biting nails and smoking, apathy, and significant neglect of personal hygiene tasks. Two men were disturbed enough to be admitted to a psychiatric ward.

The emotional disturbances of the volunteers persisted during the refeeding period, with some men becoming more depressed, irritable, argumentative and negative than during semi-starvation. Previously outgoing and friendly, the men became withdrawn and isolated, unwilling to be involved in planning, decision-making or participating in group activities.

The men’s social interaction with women significantly declined, with those relationships that did continue becoming strained. Libido was also reported to have decreased and was slow to return. Similar to people with eating disorders during weight gain, the subjects reported that they were “feeling fat” despite body weight and body fat percentages being at pre-experimental levels after nine months of rehabilitation.

Many of the symptoms outlined above are similar to those displayed by individuals with anorexia or bulimia nervosa. Therefore, it is recognised that these symptoms are due to starvation rather than being specific to these disorders. Furthermore, symptoms are not limited to food and weight but affect nearly all areas of psychological and social functioning. The study challenges the popular idea that body weight is altered with a bit of “willpower” and demonstrates that the body does not simply adjust to a lower weight during severe dieting.