Child temperament as a developmental or an epidemiological concept: a methodological point of view.

Epidemiological and developmental methodologies are 2 complementary approaches to the study of the relationship between adverse temperament and the development of personality and psychopathology in childhood. The association between adverse temperament and clinical disorders in childhood is now well documented in the general population. This paper argues that at least 5 epidemiological questions about child temperament need answers before investing time and energy in limited analyses of molecular interactions between temperament and specific environmental characteristics. (1) We need to know to what extent child adverse temperament really increases risk of clinical disorders; this must be expressed quantitatively and compared with the other well-documented psychosocial risk factors. (2) We need to test, in random samples, different general models of the temperament-environment interplay through the 'additive, synergistic, control of exposure to other environmental risks' paradigm. (3) Do we have evidence that adverse temperament interacts with some of the well-known psychosocial risk factors or with special areas of family functioning? (4) Does adverse temperament predispose, in the general population, to particular types of childhood disorders or is there a non-specific risk effect? (5) Are there sex differences in the risk associated with adverse temperament in the population?