Alexithymia : studies of alexithymia in relation to somatic complaints, cognitive bias, emotion, and affect intensity

Sammanfattning: This thesis presents four studies on alexithymia, measured by the twenty-item Toronto Alexithymia Scale (TAS-20), and explores its relation to somatization, cognitive bias, emotion, and affect intensity. Alexithymia, by definition, involves difficulties in identifying and describing emotions, and has been assumed to be associated with somatization. Study I investigated the association of alexithymia with emotion and somatic complaints in a community sample when trait anxiety and depression was controlled. Alexithymia showed no association with somatization but correlated negatively with positive affect, and positively with negative affect - the former association was more robust. Study II explored the relationship between alexithymia and cognitive bias for illness words and negative emotion words (masked and unmasked) in a community sample. No significant correlation was found between alexithymia and any measure of cognitive bias. Group comparison between high-TAS and low-TAS scorers, however, showed that the high-TAS scorers were significantly slower on colour-naming unmasked illness words than unmasked negative emotion words. The purpose of Study III was to develop a new Swedish translation of the twenty-item Toronto Alexithymia Scale, and to examine if the theoretical structure that underlies the factor structure of the English version of the TAS-20 could be recovered in this Swedish translation of the instrument. Using confirmatory factor analysis, the previously established three-factor TAS-20 model was found to be replicable in a sample of undergraduate psychology students. Study IV investigated the dimensional structure of the Swedish translation of the Affect Intensity Measure (AIM), with the aim of finding the best structural model for the Swedish AIM data, and to study its validity. The results showed that all of the multidimensional AIM models were superior to the original uni-dimensional model. The best-fitting model was a newly derived three-factor model, based on the factors Positive Affectivity, Negative Intensity, and Negative reactivity. Validation of this model in a community sample showed different correlational patterns for negative intensity and negative reactivity, on the one hand, and positive affectivity, on the other, which demonstrates the value of treating affect intensity as a multidimensional construct.