Meta-analysis of genome-wide association studies for personality
Marleen HM De Moor, Paul T Costa, Antonio Terracciano, Robert F Krueger, Eco JC De Geus, Tanaka Toshiko, Brenda WJH Penninx, Tanu Esko, Pamela AF Madden, Jaime Derringer, Najaf Amin, Gonneke Willemsen, Jouke-Jan Hottenga, Marijn A Distel, Manuela Uda, Serena Sanna, Philip Spinhoven, Catharina A Hartman, Patrick Sullivan, Anu Realo, Jüri Allik, Andrew C Heath, Michele L Pergadia, Arpana Agrawal, Peng Lin, Richard Grucza, Teresa Nutile, Marina Ciullo, Dan Rujescu, Ina Giegling, Bettina Konte, Elisabeth Widen, Diana L Cousminer, Johan G Eriksson, Aarno Palotie, Leena Peltonen, Michelle Luciano, Albert Tenesa, Gail Davies, Lorna M Lopez, Narelle K Hansell, Sarah E Medland, Luigi Ferrucci, David Schlessinger, Grant W Montgomery, Margaret J Wright, Yurii S Aulchenko, A Cecile JW Janssens, Ben A Oostra, Andres Metspalu, Gonçalo R Abecasis, Ian J Deary, Katri Räikkönen, Laura J Bierut, Nicholas G Martin, Cornelia M van Duijn, Dorret I Boomsma
2012-03-01
Abstract:Personality can be thought of as a set of characteristics that influence people's thoughts, feelings and behavior across a variety of settings. Variation in personality is predictive of many outcomes in life, including mental health. Here we report on a meta-analysis of genome-wide association (GWA) data for personality in 10 discovery samples (17 375 adults) and five in silico replication samples (3294 adults). All participants were of European ancestry. Personality scores for Neuroticism, Extraversion, Openness to Experience, Agreeableness and Conscientiousness were based on the NEO Five-Factor Inventory. Genotype data of∼ 2.4 M single-nucleotide polymorphisms (SNPs; directly typed and imputed using HapMap data) were available. In the discovery samples, classical association analyses were performed under an additive model followed by meta-analysis using the weighted inverse variance method. Results …