Some earlier studies have reported a positive association between alexithymia and delinquency. We studied this association in a sample of Finnish prison inmates. A questionnaire including the 20-item Toronto Alexithymia Scale (TAS-20), the 13-item Beck Depression Inventory and questions on socio-demographic variables as well as current and previous convictions, was delivered to 209 male prisoners. Of these, 113 individuals (54.1%) aged 17-65 years (mean 33.5) returned the questionnaire acceptably filled in. From a general population study, 1300 men aged 30-50 years (mean 40.3) were drawn as a control group.

The prevalence of alexithymia (TAS-20 cut-off point 60/61) was 7.5% in the population sample and 26.5% in the prisoner sample (p<.001). In a logistic regression analysis, controlling for age, marital status, basic education and depression, being a prisoner was still highly significantly associated with dichotomous alexithymia (OR 2.60, p=.003). Moreover, the mean TAS-20 score differed significantly between the samples (45.9 vs. 50.6 points, p<.001).

Of the prisoners, 18 (15.9%) reported having committed homicide. When they alone were compared with the population sample, no significant difference in the prevalence (7.5% vs. 11.1%) or level (mean TAS-20 score 45.9 vs. 46.8) of alexithymia was found. In a logistic regression analysis with confounders, being a convict confessing to homicide was not associated with alexithymia.

Male prisoners are more alexithymic than men in general population. There are, however, differences between different types of crimes. Those who confessed to homicide were, surprisingly, not more alexithymic than controls. Studies with larger samples are needed.