Journal of Anxiety Disorders 24 (2010) 433–439 Contents lists available at ScienceDirect Journal of Anxiety Disorders The role of genes and environment in shaping co-occurrence of DSM-IV defined anxiety dimensions among Italian twins aged 8–17 Anna Ogliari a,b,∗ , Chiara A. Spatola a , Paola Pesenti-Gritti a , Emanuela Medda c , Luana Penna c , Maria Antonietta Stazi c , Marco Battaglia a,b,d , Corrado Fagnani c a The Academic Centre for the Study of Behavioural Plasticity, Vita-Salute San Raffaele University, 20 via S. D’Ancona, 20127 Milan, Italy b The Department of Clinical Neuroscience, Istituto Scientifico San Raffaele, 20 via S. D’Ancona, 20127 Milan, Italy c The Italian Twin Registry, National Centre for Epidemiology, Surveillance and Health Promotion, Istituto Superiore di Sanità, 299 viale Regina Elena, 00161 Rome, Italy d The Department of Child Psychiatry, Istituto Scientifico Eugenio Medea, via Don Luigi Monza 20, 23842 Bosisio Parini (LC), Italy article info Article history: Received 14 July 2009 Received in revised form 2 February 2010 Accepted 15 February 2010 Keywords: Childhood Anxiety disorders Comorbidity Twin studies abstract This study investigated the ultimate causes of co-variation between symptoms of four common DSM- IV anxiety dimensions – Generalized Anxiety, Panic, Social Phobia and Separation Anxiety disorder – assessed with the Italian version of the Screen for Child Anxiety-Related Emotional Disorders question- naire in a sample of 378 twin pairs aged 8–17 from the population-based Italian Twin Register. Genetic and environmental proportions of covariance between the targeted anxiety dimensions were estimated by multivariate twin analyses. Genetic influences (explaining from 58% to 99% of covariance) and unique environmental factors were the sole sources of co-variation for all phenotypes under study. Genetic influences associated with different anxiety dimensions coincide remarkably, as indicated by genetic correlations ranging from 0.40 to 0.61, while unique environmental overlap is less substantial. Thus, while additive genetic effects are important in explaining why children report symptoms from multi- ple anxiety disorders, environmental idiosyncratic factors seem to play a marginal role in shaping the co-occurrence of different anxiety dimensions in childhood. © 2010 Elsevier Ltd. All rights reserved. 1. Introduction Childhood and adolescence are critical for the development of anxiety disorders (AD), the most frequent form of psychopathology in the developmental years (Beesdo, Knappe, & Pine, 2009). Lifetime prevalence of AD in children and adolescents is about 15% to 20%, with 1-year or 6 month rates that do not differ greatly from life- time estimates (Beesdo et al., 2009; Kessler et al., 2005). The most frequent diagnosis is Separation Anxiety Disorder which typically manifests before age 12 with estimates that vary from 2.8% in child- hood to 8% in adolescence, while Agoraphobia and Panic Disorder have low prevalence (1% and 3–4% respectively, see Beesdo et al., 2009 for Review). Girls usually report more symptoms than boys, with a ratio that varies from 2:1 to 3:1 depending on data collection methodology and age (Beesdo et al., 2009; Kessler et al., 2005). ∗ Corresponding author at: The Academic Centre for the Study of Behavioural Plasticity, Vita-Salute San Raffaele University, 20 via S. D’Ancona, 20127 Milan, Italy. Tel.: +39 02 2643 5390; fax: +39 02 2643 3408. E-mail addresses: ogliari.anna@hsr.it (A. Ogliari), spatola.chiara@hsr.it (C.A. Spatola), pesentigritti.paola@hsr.it (P. Pesenti-Gritti), emanuela.medda@iss.it (E. Medda), luana.penna@iss.it (L. Penna), stazi@iss.it (M.A. Stazi), battaglia.marco@hsr.it (M. Battaglia), corrado.fagnani@iss.it (C. Fagnani). Several studies have already suggested that anxiety disorders (AD) are highly comorbid with each other, both in the developmen- tal age (Beesdo et al., 2009; Essau, Conradt, & Petermann, 2000; Last, Strauss, & Francis, 1987; Otto et al., 2001) and in adulthood (Merikangas & Angst, 1995). This could be due to symptoms’ over- lapping among different anxiety disorders and/or to the artificial nature of the categorical classification (Angold & Costello, 1993; Maser & Cloninger, 1990), so that high rates of comorbidity could be ultimately connected to poor discriminant validity (Brown & Barlow, 1992). Comorbidity can also originate from the fact that one disorder is part of, or is caused by, another disorder (Perrin & Last, 1995). More sophisticated concepts and approaches to comorbid- ity encompass the hypothesis of shared etiological factors (genetic and/or environmental) and shared underlying disease processes. In contrast to a large body of literature that examined the causes of comorbidity between anxiety disorders and depression in adult- hood (Hettema, Prescott, Myers, Neale, & Kendler, 2005; Kendler et al., 1995), the origin of the co-occurrence between different AD in youth has rarely been addressed. Behavioral genetic strategies can clarify the causes of comorbidity. While univariate approaches estimate the role of genes and environment in the expression of a single phenotype, multivariate designs address how genetic and environmental risk factors influence the covariation of different disorders. Information derived from multivariate studies is then 0887-6185/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.janxdis.2010.02.008