CS08.01 - Gene-environment interaction in anorexia nervosa

Abstract

Purpose:

To analyse GxE interactions assess non-shared environmental (E) risk factors for the development of AN specific for sisters discordant for an ED, polymorphisms in the serotonin transporter (G),.

Methods:

We interviewed 128 sister pairs discordant for an eating disorder using the Oxford-RFI as part of the European "Healthy Eating" multicenter study at 3 university centres (Vienna, London, Barcelona) (AN-R: 58; AN-BP: 70; 128 sisters without ED). To examine association between AN, G and E, and G x E-interaction, conditional logistic regression was used with a Cox proportional hazards regression model using the exact method.

Results:

Genotype (GT) distributions did not differ between the sister groups. Significant main effects were found for disruptive events, interpersonal problems and family dieting behaviour. The risk for AN increased with higher levels in these variables independently of the genotype. Significant interactions were found for G x parental problems and G x burden by parental psychiatric disorder. The increase of risk for AN with increasing number of problems with parents is larger for the S/S genotype than for L/L. However, a higher burden by parental psychiatric illness (subjective E according to Turkheimer 2000) increased the risk for AN-this was larger for the L/L than for the S/S GT.

Conclusions:

This study suggests that there is an interaction between stress (problems with parents) and the ss GT which increases the risk of developing AN.