Interaction Between Childhood Adversity, Brain- Derived Neurotrophic Factor val/met and Serotonin Transporter Promoter Polymorphism on Depression: The TRAILS Study Esther Nederhof, Esther M.C. Bouma, Albertine J. Oldehinkel, and Johan Ormel Background: The three-way interaction between the functional polymorphism in the serotonin transporter gene linked promoter region, the val66met polymorphism in the brain-derived neurotrophic factor gene, and childhood adversity in the prediction of depression in children, reported by Kaufman and colleagues in 2006, has only been confirmed in adult samples. This study examines the gene-by-gene- by-environment interaction in an adolescent sample. Methods: In a longitudinal population-based study, depression scores were assessed with the Youth Self Report at ages 11, 13.5, and 16. Pre- and perinatal adversities and childhood events were assessed in a parent interview at age 11. Long-term difficulties until age 11 were assessed with a parent questionnaire at age 13.5. Blood or buccal cells were collected for genotyping at age 16. The study included 1096 complete data sets. Results: Depression score over the three measurements was not significantly predicted by any interaction between genotypes and childhood adversities. Conclusions: We were unable to confirm the three-way interaction in a representative, population-based sample of adolescents. The large sample resulted in adequate power, which in combination with the reliability of our measures gives confidence in our findings. Key Words: Gene– environment interaction, gene– gene interac- tion, stressful life events T he goal of this study was to replicate the finding that childhood adversity predicts higher depression scores in individuals carrying the brain-derived neurotrophic factor (BNDF) gene met allele and the serotonin transporter gene linked promoter region (5-HTTLPR) s/s genotype. In 2006, Kauf- man and colleagues (1) were the first to show a three-way interaction between childhood maltreatment history, brain-de- rived neurotrophic factor (BDNF) val66met, and HTTLPR in the prediction of depression scores during adolescence. A sample of 109 children with a severe maltreatment history who had just been removed from the parental home was compared with 87 control subjects. Kaufman et al. found that depression scores were highest in maltreated children carrying a BDNF met allele and the 5-HTTLPR s/s genotype compared with control subjects and to maltreated children with other genotypes (1). This finding was replicated in a female adult twin sample using a continuous childhood adversity measure (2) but not in a college-aged sample (3). A similar three-way interaction with recent life stress has been reported in an elderly sample (4). Because it is important to confirm genetic associations (5), we investigated whether this three-way interaction was present in a large repre- sentative population cohort of Dutch adolescents. Methods and Materials Sample Data from the first, second, and third wave of the TRacking Adolescents’ Individual Lives Survey (TRAILS) were used. TRAILS is a prospective cohort study of Dutch adolescents. At baseline, 2230 children were enrolled in the study (response rate 76%, mean age 11.09 [SD .59], 51% girls) (6), and at the second wave 96.4% of the baseline sample participated again (n = 2149, mean age 13.55 [SD .54], 51% girls) (7). At Wave 3, the response rate was 81.4% (n = 1838, mean age 16.13 [SD .59], 52% girls) of whom 1599 gave blood (or buccal cells) for DNA analysis. Procedures At the first assessment wave, well-trained interviewers visited one of the parents or guardians (preferably the mother, 95.6%) at their homes to administer an interview covering a wide range of topics, including prenatal and perinatal risks and the occurrence of stressful events. At the second wave, the parents filled out a questionnaire including questions about long-term difficulties affecting the child. At all three assessment waves, children were evaluated at school, where they filled out questionnaires under the supervision of TRAILS assistants. At the third wave, blood or buccal cells were collected for DNA analysis. All procedures were approved by the Central Committee for Research involving Human Subjects. All participants and their parents gave written informed consent. Pregnancy and Delivery Adversities The variable pregnancy and delivery adversities (PDadv) was based on questions about maternal smoking and alcohol use during pregnancy, birth weight, gestational age, and pregnancy and delivery complications (8). Childhood Events The variable childhood events (CE) was a sum score of stressful events that occurred before the first wave and included From the Interdisciplinary Center for Psychiatric Epidemiology, Department of Psychiatry, University Medical Center Groningen, University of Gro- ningen, Kingdom of the Netherlands. Address correspondence to Esther Nederhof, Ph.D., Interdisciplinary Center for Psychiatric Epidemiology, Department of Psychiatry, CC72, Univer- sity Medical Center Groningen, P.O. Box 30 001, 9700 RB Groningen, the Netherlands; E-mail: e.nederhof@med.umcg.nl. Received Nov 9, 2009; revised Mar 23, 2010; accepted Apr 1, 2010. BIOL PSYCHIATRY 2010;68:209 –212 0006-3223/$36.00 doi:10.1016/j.biopsych.2010.04.006 © 2010 Society of Biological Psychiatry