There is extensive use of the 20-item Toronto Alexithymia Scale (TAS-20) in research and clinical practice in anorexia nervosa (AN), though it is not empirically established in this population. This study aims to examine the factorial validity of the TAS-20 in a Portuguese AN sample (N = 125), testing four different models (ranging from 1 to 4 factors) that were identified in critical examination of existing factor analytic studies. Results of confirmatory factor analysis (CFA) suggested that the three-factor solution, measuring difficulty identifying (DIF) and describing feelings (DDF), and externally oriented thinking (EOT), was the best fitting model. The quality of measurement improves if two EOT items (16 and 18) are eliminated. Internal consistency of EOT was low and decreased with age. The results provide support for the factorial validity of the TAS-20 in AN. Nevertheless, the measurement of EOT requires some caution and may be problematic in AN adolescents.
Keywords: Alexithymia; Anorexia nervosa; Clinical samples; Confirmatory factor analysis; Factor structure.