Original article New evidence of heterogeneity in social anxiety disorder: Defining two qualitatively different personality profiles taking into account clinical, environmental and genetic factors C. Binelli a,b , A. Mun ˜iz a,c , S. Sanches a,d , A. Ortiz a,c , R. Navines a,e , E. Egmond a,b , M. Udina a , A. Batalla a,c , C. Lo ´ pez-Sola b , J.A. Crippa d , S. Subira ` b , R. Martı ´n-Santos a,c, * a Service of Psychiatry and Psychology, Hospital Clı´nic, IDIBAPS, CIBERSAM, Barcelona, Spain b Department of Clinical and Health Psychology, Autonomous University of Barcelona, Bellaterra, Spain c Department of Psychiatry and Clinical Psychobiology, University of Barcelona, Barcelona, Spain d Department of Neuroscience and Cognitive Behavior, Hospital das Clinicas, Ribeirao Preto, University of Sao Paulo, Sao Paulo, Brazil e Human Pharmacology and Clinical Neurosciences Research Group, IMIM-Hospital del Mar Medical Research Institute, Barcelona, Spain 1. Introduction Social anxiety disorder (SAD) is a highly prevalent [1] psychiatric disorder characterized by fear and avoidance of interpersonal situations, which exceeds the adaptive threshold and interferes with daily life [2]. SAD is associated with significant psychosocial impairment and high rates of comorbidity [3], early life events [4] and has been reported to be a risk factor for subsequent depressive and substance abuse disorders [5]. The majority of people with SAD is shy, avoid meeting new people, and are withdrawn in unfamiliar social settings, with frequent avoidant behaviors associated [2,5]. Available evidence suggest that a great number of patients with SAD show dysfunctional personality traits and high rates of comorbidity with certain personality disorders, in particular with avoidant and dependent personality disorders [6,7]. Of particular interest in the study of SAD have been two temperamental dimensions from Cloninger’s psychobiological model of personality: harm avoid- ance (HA) and novelty seeking (NS) [8]. It should be noted that both dimensions assess differences in automatic emotional responses to stimuli defining personality style, and have their corresponding neurobiological substrate [9]. HA is defined as a heritable bias in the inhibition of behaviors and the tendency to respond fearfully to European Psychiatry 30 (2015) 160–165 A R T I C L E I N F O Article history: Received 10 February 2014 Received in revised form 22 August 2014 Accepted 21 September 2014 Available online 8 December 2014 Keywords: Social anxiety disorder Impulsive subtype Harm avoidance Novelty seeking Temperament and character inventory, STin2.VNTR A B S T R A C T Purpose: To study qualitatively different subgroups of social anxiety disorder (SAD) based on harm avoidance (HA) and novelty seeking (NS) dimensions. Method: One hundred and forty-two university students with SAD (SCID-DSM-IV) were included in the study. The temperament dimensions HA and NS from the Cloninger’s Temperament and Character Inventory were subjected to cluster analysis to identify meaningful subgroups. The identified subgroups were compared for sociodemographics, SAD severity, substance use, history of suicide and self-harm attempts, early life events, and two serotonin transporter gene polymorphisms (5-HTTLPR and STin2.VNTR). Results: Two subgroups of SAD were identified by cluster analysis: a larger (61% of the sample) inhibited subgroup of subjects with ‘‘high-HA/low-NS’’, and a smaller (39%) atypical impulsive subgroup with high–moderate HA and NS. The two groups did not differ in social anxiety severity, but did differ in history of lifetime impulsive-related-problems. History of suicide attempts and self-harm were as twice as frequent in the impulsive subgroup. Significant differences were observed in the pattern of substance misuse. Whereas subjects in the inhibited subgroup showed a greater use of alcohol (P = 0.002), subjects in the impulsive subgroup showed a greater use of substances with a high-sensation-seeking profile (P < 0.001). The STin2.VNTR genotype frequency showed an inverse distribution between subgroups (P = 0.005). Conclusions: Our study provides further evidence for the presence of qualitatively different SAD subgroups and the propensity of a subset of people with SAD to exhibit impulsive, high-risk behaviors. ß 2014 Elsevier Masson SAS. All rights reserved. * Corresponding author. Service of Psychiatry and Psychology, Hospital Clı ´nic, IDIBAPS, University of Barcelona, CIBERSAM, Villarroel, 150, 08036 Barcelona, Spain. E-mail addresses: rmsantos@clinic.ub.es, rmsantos@imim.es (R. Martı ´n-Santos). Contents lists available at ScienceDirect European Psychiatry jo u rn al h om epag e: h ttp ://ww w.eu ro p s y- jo ur n al.co m http://dx.doi.org/10.1016/j.eurpsy.2014.09.418 0924-9338/ß 2014 Elsevier Masson SAS. All rights reserved.