Figure 3.

The effect of folic acid (FA) randomization on the primary clinical outcome depends on the catechol‐O‐methyltransferase (COMT) genotype. There was an interaction between COMT and FA, such that Val/Val homozygotes (FA placebo, n=13; FA, n=15) showed a greater improvement than Met carriers (FA placebo, n=43; FA, n=49). Planned comparisons indicated that the negative effect of FA on lamotrigine response was limited to Met carriers, since there was an effect of FA randomization in the Met carrier group randomized to lamotrigine (indicated with an asterisk) that was absent from all other groups (numbers in parentheses on the x‐axis indicate subgroup numbers). Lam, lamotrigine; Pbo, placebo; QIDS, Quick Inventory for Depressive Symptoms