Relationships Between Tridimensional Personality Questionnaire Dimensions and DSM-III-R Personality Traits in Italian Adolescents C. Maggini, P. Ampollini, C. Marchesi, S. Gariboldi, and C.R. Cloninger The predictions of Cloninger’s neurobiologic learning model on the relationships between novelty seeking (NS), harm avoidance (HA), reward dependence (RD), and persistence (P) and the traditional DSM-III-R per- sonality disorders (PDs) were tested on a sample of 2,889 (1,475 males and 1,414 females) Italian high school students aged 16 to 18 years, using the Struc- tured Clinical Interview for DSM-III-R Personality Disor- ders–self-report (SCID-II) and the Tridimensional Per- sonality Questionnaire (TPQ). All relationships were in the predicted direction for antisocial, narcissistic, avoidant, and obsessive-compulsive PD alone, and at least two were in the predicted direction for schizoid, histrionic, borderline-explosive, dependent, and pas- sive-aggressive PD. Eight of nine relationshipswere in the predicted direction for NS, but only seven of nine for HA and RD. This study provides substantial sup- port for Cloninger’s neurobiologic learning model as a useful tool to describe and classify personality vari- ants and, because of the supposed neurochemical implications, to link personality traits to the underly- ing neurochemical and neuroanatomic substrate. Copyright  2000 by W.B. Saunders Company A CCORDING TO Cloninger’s neurobiologic learning model, 1-4 temperament can be de- scribed by four genetically inherited independent dimensions: novelty seeking (NS), harm avoidance (HA), reward dependence (RD), and persistence (P). These dimensions are conceptualized as heri- table biases in the activation, inhibition, mainte- nance, and persistence of behaviors, respectively, and reflect the individual differences in percept- based habits and skills. 5 In his early reports, Cloninger 1,2 described only three dimensions (NS, HA, and RD), with the fourth dimension (P) derived later from RD. 3,4 He postulated that the combinations of the extremes in the first three temperament factors correspond to traditional per- sonality disorders (PDs) as described in the DSM- III-R axis II categories. 6 This model provides a useful framework to study the relationships be- tween the traditional personality categories and the underlying neuroanatomic and neurochemical sub- strate. Cloninger also developed a self-report instru- ment, the Tridimensional Personality Question- naire (TPQ), 2 to assess each dimension in a given population. To date, only one study by Goldman et al. 7 has tested the relationships between the three TPQ dimensions and the DSM-III-R axis II categories. The authors found good agreement with Clon- inger’s predictions, particularly regarding obsessive- compulsive, avoidant, dependent, antisocial, and schizoid PDs, in a sample of 99 psychiatric pa- tients. Another study by Svrakic et al. 8 was per- formed in a population of 136 psychiatric inpatients using a more recently developed questionnaire, the Temperament and Character Inventory, 4 which in- cludes adjunctive items exploring for the fourth temperament dimension (P) and for three character dimensions (self-directedness, cooperativeness, and self-transcendence). The authors concluded that temperament dimensions are useful to differentiate among DSM-III-R PD clusters. The patients in clusters A, B, and C were differentiated by low RD, high NS, and high HA, respectively. However, these investigations were performed in psychiatric patients, and thus the results cannot be extended to a general population because of the possible con- founding effects of axis I disorders. 4 No study has investigated the relationships be- tween temperament dimensions and DSM-III-R personality categories in a wide population of healthy subjects. This study is part of a broader epidemiologic investigation to assess the preva- lence of symptoms of eating disorders, obsessive- compulsive disorder, social phobia, and PDs and temperament features in a sample of 2,889 high- school students. In this report, we examine whether, in a normal population of adolescents, the personal- ity features are associated with NS, RD, and HA, as suggested in Cloninger’s model 2 (predictions listed From the Institute of Psychiatry, University of Parma, Parma; Department of Mental Health, Azienda Unita ` Sanitaria Locale Reggio Emilia, District of Castelnovo ne ` Monti, Italy; and Department of Psychiatry, Washington University, St Louis, MO. Address reprint requests to C. Maggini, M.D., Institute of Psychiatry, University of Parma, P.le Matteotti, 19, 43100- Parma, Italy. Copyright  2000 by W.B. Saunders Company 0010-440X/00/4106-0003$10.00/0 doi:10.1053/comp.2000.16559 426 Comprehensive Psychiatry, Vol. 41, No. 6 (November/December), 2000: pp 426-431