Self‐concept and depressive symptoms three years after stroke: An evaluation of predictive value, the role of subdomains and individual importance

TL;DR

This study explores how self-concept and its subdomains predict depressive symptoms in stroke survivors three years after the event.

Contribution

The study introduces the individual importance of self-concept subdomains in predicting depressive symptoms after stroke.

Findings

Self-discrepancy significantly interacts with depressive symptoms (F (1, 118) = 32.69, p < .001).

Present self-concept is more strongly associated with depressive symptoms than past self-concept or self-discrepancy.

Age, history of depression, social support, and present self-concept are independent predictors of depressive symptoms.

Abstract

Depressive symptoms (DS) after stroke are associated with marked negative consequences for rehabilitation. Identifying determinants of DS is needed to enable prediction and develop psychological interventions. A promising candidate may be self‐concept and changes thereof, so‐called self‐discrepancy. Consulting recent self‐concept models, we investigated the role of self‐concept subdomains and their individual importance. Within a prospective longitudinal study, 120 stroke survivors were interviewed via telephone 3 years post‐ictus to assess present and past self‐concept, self‐discrepancy, self‐concept subdomains and DS. The association of self‐concept measures and DS was investigated using an ANCOVA. Controlling for established determinants (age, sex, history of depression, functional independence, social support), multiple regression analyses were used to examine the independent influence of self‐concept measures and the role of subdomains and importance‐weightings. Self‐discrepancy showed a significant interaction with DS (F (1, 118) = 32.69, p < .001, η
2 = .22). DS showed a stronger association with present (r = −.72) than with past self‐concept (r = −.34) and self‐discrepancy (r = −.47; all p < .001). Age, history of depression, social support and present self‐concept were independent predictors of DS while functional independence was not (∆F (1, 113) = 48.04, p < .001). Importance‐weighting of subdomains did not affect explained variance, though the number of self‐concept subdomains showing significant association with DS increased. Findings propose appraisals of self‐concept as independent predictors of DS after stroke. Considering individual importance of subdomains reveals their differential influence. The results suggest investigating the use of general self‐concept for prediction and considering the individual relevance of subdomains in psychological interventions after stroke.

Figures