This is from the full text of the study:
Substance abusers were more likely than normal adults and psychiatric outpatients to say they had trouble identifying and describing their emotions (on the TAS-20), but showed no actual deficits in identifying and describing emotions (on the LEAS). Substance abusers performed as well on the LEAS as a college student sample, after controlling for age, gender, and IQ, and performed as well as a community adult sample. The observed differences between drug treatment and control groups on the self-report measure were highly statistically reliable, whereas the observed differences on the performance measure did not approach statistical significance. There are at least two potential explanations for this disassociation between self-report and performance.
The negative mood explanation: Negative mood was associated with self-reported alexithymia, which is consistent with what has been found in a normal sample (Haviland et al. 1994). Haviland et al. (1994) have suggested that negative emotional states (i.e., anxiety) may lead to increased alexithymia. Given that our drug treatment sample was experiencing much more negative affect than others, it is possible that the primary reason substance abusers are more alexithymic is because of this increased negative affect. Inconsistent with this view, psychiatric groups that might be expected to experience substantial negative affect (e.g., obsessive compulsive and simple phobia patients) do not report particularly high levels of alexithymia (about 13%; Taylor 2000). Future research should measure self-reported alexithymia and negative mood in both a substance abuse and control group, and then examine whether differences in alexithymia still exist after controlling for the effects of mood.
The inaccurate belief / low-motivation explanation: Substance abusers’ self-reports were unrelated to their actual performance, which is inconsistent with a study of normal adults (Lane et al. 1996). This finding suggests that substance abusers’ beliefs may be inaccurate. Even so, such beliefs may have a crucial impact on behavior. If people believe that they are not able to deal effectively with their emotions, they may be less motivated to do so. Less motivation may, in turn, lead to more alexithymic behavior. For this explanation to account for the disassociation between self-report and performance alexithymia, we would have to assume that something about the LEAS task motivated substance abuse participants to perform adequately (relative to our comparison groups). Such motivational factors may include the well-structured nature of the task, or the presence of a researcher.
I find these findings fascinating, but I don’t find either of these explanations very satisfying. I wonder if the self-reported alexithymia is rooted in a perceived inability to make others adequately understand their emotional state–that others do not understand the intensity of their emotional state or an aspect of their emotional state that they believe to be unique. I’m able to tell you that I am mad or sad, but you still don’t understand my suffering. This would fit nicely recovering community humor about “terminal uniqueness” and lend itself to cognitive behavioral therapy or rational emotive therapy to reduce these cognitive distortions.
Another possibility is that the subjects felt confused by their own emotional state and unable to construct an adequate narrative to explain their emotional state and therefore unable to adequately communicate their emotional state.