Temporal stability and differential relationships with neuroticism and extraversion of the three subscales of the 20-item Toronto Alexithymia Scale in clinical and nonclinical samples

Objectives: To examine the cross-sample and temporal stability of the three subscales of the 20-item Toronto Alexithymia Scale (TAS-20) and to study the pattern of associations between the TAS-20 scales, neuroticism, and alexithymia. Methods: Two clinical and three nonclinical samples were included in the cross-sectional part of the study. One clinical and one nonclinical sample also participated in the 6-month follow-up study. To test the replicability of the three-factor structure of the TAS-20 across samples, a principal component analysis was conducted, followed by a Procrustes rotation. Linear regression analyses were conducted to examine temporal stability and to determine the specificity of the associations among the TAS-20 scales, neuroticism, and extraversion. Results: The three-factor structure of the TAS-20 was confirmed across all five samples. With a few exceptions, neuroticism and extraversion attained a higher level of temporal stability than the alexithymia scales. The TAS-subscale measuring difficulty identifying feelings (DIF) was less stable in the clinical than in the nonclinical sample. Neuroticism was most strongly associated with DIF, whereas extraversion was mainly related to externally oriented thinking (EOT). A combination of both neuroticism and extraversion was the best predictor of difficulty describing feelings (DDF). Conclusions: Future studies on the psychometric properties and the clinical value of the TAS-20 should pay more attention to its subscales. Especially the EOT facet needs further examination. (C) 2004 Elsevier Inc. All rights reserved.