Genetic Exploration of Valproate Induced Weight Gain

Background

Childhood obesity has both long- and short-term health consequences and is a growing public health problem in Ireland (O’Neill et al, 2007; Finucane et al, 2008; Riddoch et al, 1991). Medications used in the treatment of chronic medical conditions like epilepsy and psychiatric illness might be contributing to this epidemic (Ben-Menachem, 2007, Pickrell, 2013; Martine). Epilepsy affects about 50 million people worldwide with about 40,000 people living with the condition in Ireland (JEC, 2011, Linehan et al, 2010). Sodium valproate (VPA) is an effective anti-epileptic drug, but up to 71% of the patients exposed to the drug report weight gain as an adverse drug reaction (Biton, 2003). The clinical and genetic risk factors associated with this ADR are unknown.

Objective

To identify clinically relevant predictors of VPA-induced weight change in children with epilepsy within the first year of initiation of therapy.

Methods

Under a retrospective framework, we recruited and performed clinical phenotyping on paediatric epilepsy patients treated with VPA. We employed maximum percentage weight change, and change in weight-for-age z-score as weight change phenotypes. We extracted DNA from these subjects and carried out genotyping for 17 SNPs that had previously been robustly associated with changes in BMI and weight in healthy individuals. Meta-analysis of our results was conducted using a fixed-effects model.

Results

We recruited 251 children, of which 218 underwent genotyping. About 48% of the children experienced a gain in weight of ≥5% from that predicted by weight at initiation of therapy. None of the clinical factors tested significantly influenced weight change. Meta-analysis indicated that the genetic variants rs10938397, rs7498665 and rs1805081 showed nominally significant association with quantitative weight change due to VPA. However none survived correction for multiple testing.

Conclusions

None of the clinical phenotypes tested in this study was associated with VPA-induced weight change, nor did our results support clinical predictors reported in previous studies. Although three of the SNPs associated with changes in BMI showed nominally significant association with weight change associated with VPA, none of them survived Bonferroni correction. A large multi- centre research is required to provide sufficient power to identify any genetic factors associated with this ADR.