Alexithymia and Executive Function in Younger and Older Adults

Abstract

Objective: Alexithymia is associated with adverse mental and physical health outcomes. Determining neuropsychological factors associated with alexithymia may aid in elucidating its underlying mechanisms and identifying treatment targets. Accumulating evidence indicates that executive dysfunction may co-occur with alexithymia in younger adults (YA). However, research on this link in older adults (OA), who may be at greater risk for alexithymia, is scant. This study determined associations between alexithymia and executive function (EF) in healthy younger and OA. Alexithymia was predicted to be associated with poorer EF in both age-groups. Method: Younger (n = 65, aged 18–30; 46% female) and OA (n = 44, aged 61–92; 73% female) completed the 20-item Toronto Alexithymia Scale, three EF tasks (Verbal Fluency, Design Fluency, and Trail Making), measures of memory and verbal IQ, and a self-report measure of depressive symptoms. Three EF composites were created to assess verbal EF, visuospatial EF, and global EF. Results: Greater alexithymia and difficulty describing feelings were associated with poorer verbal EF in OA (p = .02 and p = .005, respectively) but not in YA (ps > .05). The other neuropsychological measures were not significantly associated with alexithymia in regression analyses. Conclusions: Findings are consistent with previous research identifying links between EF and alexithymia. The association between alexithymia and verbal EF may be due to shared prefrontal circuitry involved in emotion regulation. Results provide insight into possible origins of emotion self-awareness deficits in OA.

Keywords:

• Alexithymia
• Aging
• Executive function
• Emotion.

Notes

¹ Levene’s test of homogeneity of variance was significant (p < .05) for the following variables in OA vs. YA comparisons: education, CES-D, verbal EF composite, and visual memory composite. This violation of homogeneity of variance was corrected by not using the pooled estimate for the error term of the t-statistic and by adjusting the degrees of freedom using the Welch–Satterthwaite method for these variables.