It is illogical-and unsupported by the evidence-to attribute these features to malnutrition alone. They are, very likely, genetically determined. Unaffected first-degree relatives often exhibit the very same characteristics and research suggests that, in sufferers, they predate the onset of the illness and many persist despite the restoration of good physical health. 

How these features occur is a mystery, a mystery compounded by the fascinating paradoxes inherent in the illness. Patients feel fat when they are thin. They feel well when they are unwell. They behave with great strength despite their obvious fragility. They feel full when they are empty. They starve themselves but some also binge. They have extraordinary control over themselves and their environment but feel they have no control. Before being ill they have been conscientious, compliant children, but when ill they are defiant and rebellious. They experience the illness as a comfort and a refuge, yet are clearly tormented by it. Finally, they experience our attempts to help as coercive: at times they have insight and wish to recover, at others they are oblivious to their illness and desperately resist treatment.

Many attempts have been made to explain the mystery and paradoxes of anorexia nervosa. However, none so far has fulfilled the criteria required for a complete explanation: empirically derived, refutable, specific to the illness, necessary and sufficient to account for the development of the illness and its various manifestations. Research conducted by our team and others over several years has led us to propose a model based on neuroscience that purports to fulfil all these criteria. 

In the development of anorexia nervosa we fully accept the contribution of pregnancy complications that may affect foetal brain development, environmental factors such as sociocultural pressures to be thin, the onset of puberty with its associated body changes and other external stressors. However, these are insufficient to explain the illness in its entirety. A series of studies by our colleagues, and others, have produced convincing evidence of a significant genetic contribution.