Explanations for Anorexia
Genes for specific behaviours have not yet been identified, but anecdotally there is a tendency for the disorders to run in families (like noses!). The American Psychological Association (1994) found an increased incidence in family members if first-degree relatives (parents and siblings) had the disorder.
Twin studies (compare concordance rates between MZ (identical twins) and DZ (fraternal twins). If the concordance rate is higher for MZ than DZ it is evidence for a genetic component.
Holland et al (1984):
MZ twins: concordance rate 56%
DZ twins: concordance rate 7%
However there isn't a 100% concordance rate, so other factors must be having an effect.
The major problem is in ruling out environmental factors. Not only do MZ twins share the same genes they also share very similar environments, far more so than DZ twins. MZ twins are often dressed similarly, have the same friends, same interests, same teachers and so on. And of course they are always the same sex, unlike DZ twins that can be brother and sister. This last factor is particularly relevant when considering eating disorders because of their much greater prevalence in girls of the female gender!
When considering genetic causation it is also worth mentioning that the effect may not be direct. Perhaps genes are influencing a characteristic which in turn is increasing the risk…