Quantitative Electroencephalographic Correlates of Post-Stroke Depression

Abstract

Post stroke depression (PSD) is widely recognized as a complication in the rehabilitation process, and numerous studies have shown a relationship between depression and compromised outcome following stroke (e.g., Parikh, et al., I 990). Several studies have examined specific anatomic and physiological correlates of PSD. One of the more recent of these is the work of Astrom, Adolfsson, and Asplund (1993), in which the researchers found left anterior lesions to be among the most significant predictors of PSD in the acute stage of recovery. Quantitative electroencephalography (QEEG) has emerged has a useful, non-invasive, functional method of evaluating both neurological and psychiatric disorders. Using QEEG, Davidson and associates (e.g., Davidson & Tomarken, 1989) have consistently found greater left frontal alpha power to be associated with depression, and with tendency toward depression. Given the robustness of the relationship between QEEG patterns and depressive symptomatology, and given that the approach has not been used among PSD patients for such a purpose, the present study was conducted to compare qEEG patterns between CVA patients with and without depression. Twenty one patients with cortical or subcortical CVAs were recruited as volunteer participants, and upon proper consent were administered the Beck Depression Inventory (BDI) and the Yesavage Geriatric Depression Scale (GDS). Participants underwent QEEG testing with the Cadwell Spectrum 32 digital EEG apparatus, and analyses were based on the NYU database (John, 1994). Interhemispheric power asymmetries, averaged across traditional frequency bandwidths and measured in homologous pairs from prefrontal to occipital sites, were correlated with global depression scores on both indices. Significant correlations were found between frontal QEEG asymmetries and both the BDI and GDS, and in every case the asymmetry is indicative of greater left frontal power. Correlations of the highest magnitude were found in the prefrontal and superior frontal areas, which is consistent with Davidson and associates' findings with non-neurologically involved depressives. None of the posterior asymmetry measures showed significant correlations with subjective depression indices. These results, taken from a larger, ongoing study, are consistent with previous research implicating dysfunction of the left anterior quadrant in depressivesymptoms, but go beyond those findings by extending them to the PSD population.