Gisle Berg Helland, Brian Enriquez, Mona K. Beyer, Henning Stople Rise, Håkon Ihle‐Hansen, Stein Andersson, Ramunė Grambaitė, Helle Stangeland, Anja Vaskinn, Hege Ihle-Hansen, Hanne F. Harbo, Einar August Høgestøl, Guri Hagberg, Anne Hege Aamodt
BACKGROUND: Social cognition, with emotion recognition (ER) being a key component, is recognized as a core cognitive domain, yet research on ER post-stroke remains limited. This exploratory pilot study aimed to provide preliminary estimates of long-term ER impairment after stroke and to examine its association with lesion volume and location. METHODS: We invited patients <65 years, admitted for endovascular treatment (EVT) and included in a prospective cohort; the Oslo Acute Reperfusion Stroke Study, to participate in a 4-year follow-up study, and a healthy control (HC) group. ER was assessed using the "Pictures of Facial Affect" (PFA) test, impairment defined as performance ≥1.5 SD below HC mean. Linear regression analysis examined relationship between baseline predictors and ER at follow-up. Voxel-based lesion-symptom mapping (VLSM) were used to identify brain regions associated with ER, complemented by disconnectome mapping. RESULTS: Of 109 participants included, 40 completed a 4-year follow-up including the PFA test (24 (60%) women, mean age 51 (SD ± 10) years). ER impairment was detected in 9 (23%) patients. Stroke lesion volume was significantly associated with PFA scores in both univariable and multivariable analysis, adjusted for sex, education, and time to recanalization. VLSM revealed associations between right-hemisphere cortical regions and lower PFA scores, notably the inferior parietal lobule and and adjacent perisylvian cortex. CONCLUSIONS: Emotion recognition impairment was identified in more than one in five stroke patients four years after EVT. These preliminary findings identify stroke lesion burden and location as candidate group-level prognostic markers that warrant confirmation in larger studies of this under-researched cognitive domain.