Reply to Dr Martin's Letter Regarding Differences in Cognitive Outcomes After ECT Depending on BDNF and COMT Polymorphisms.

To the Editor: B ennett and colleagues recently examined the effects of both the brain-derived neurotrophic factor (BDNF) Val66Met singlenucleotide polymorphism and catecholO-methyltransferase (COMT) Val158Met single-nucleotide polymorphism cognitive and mood outcomes after electroconvulsive therapy (ECT). Their study included 87 depressed patients, and cognitive functioning was tested at multiple time points (preECT, after 4 treatments, post-ECT, 1 month post-ECT, and 3 months post-ECT) with 2 tasks, the Mini-Mental State Examination and Spatial Recognition Memory from the CANTAB computerized battery. The authors suggested that assessing either of these polymorphisms will not be helpful for predicting cognitive outcomes after ECT in clinical practice. In a larger sample, we similarly had investigated the effects of both these polymorphisms, in addition to other potentially relevant genes, for predicting cognitive and mood outcomes after ECT. However, in contrast to Bennett et al, we administered a comprehensive neuropsychological battery, which included measures of both anterograde memory (learning and delayed recall) and retrograde autobiographical memory, the cognitive functions most affected by ECT. Our results showed effects ofCOMTVal158Met on change in retrograde autobiographical memory (P = 0.048, η = 0.055) and BDNF Val66Met on change in anterograde memory functioning (P = 0.026, η = 0.043) from preto post-ECT in unadjusted models. However, these effects were no longer statistically significant after adjustment for relevant covariates. Notwithstanding, it is possible that these effects may still hold true in a larger sample because our sample (N = 117) was modest for a gene association study, a limitation we acknowledged. Although the Spatial Recognition Memory test used by Bennett and colleagues also assessed anterograde memory (visuospatial), it did not assess delayed recall,