Background. The aim of this pilot study was to examine which unique factors (genetic and
environmental) increase the risk for developing anorexia nervosa by using a case–control design of
discordant sister pairs.
Methods. Forty-five sister-pairs, one of whom had anorexia nervosa and the other did not, were
recruited. Both sisters completed the Oxford Risk Factor Interview for Eating Disorders and
measures for eating disorder traits, and sib-pair differences. Blood or cheek cell samples were taken
for genetic analysis. Statistical power of the genetic analysis of discordant same-sex siblings was
calculated using a specially written program, DISCORD.
Results. The sisters with anorexia nervosa differed from their healthy sisters in terms of personal
vulnerability traits and exposure to high parental expectations and sexual abuse. Factors within the
dieting risk domain did not differ. However, there was evidence of poor feeding in childhood. No
difference in the distribution of genotypes or alleles of the DRD4, COMT, the 5HT2A and 5HT2C
receptor genes was detected. These results are preliminary because our calculations indicate that
there is insufficient power to detect the expected effect on risk with this sample size.
Conclusions. A combination of intrinsic and extrinsic factors increases the risk of developing
anorexia nervosa. It would, therefore, be informative to undertake a larger study to examine in
more detail the unique genetic and environmental factors that are associated with various forms of