(PS1-93) Investigating the Association Between High-Power Transient Spectral Events and Depression

Abstract Text: Neural oscillations within the theta (3-7 Hz), alpha (8-12 Hz), and beta (13-30 Hz) electroencephalography (EEG) bands are well-established indices of brain activity and are associated with various psychopathologies. Emerging research suggests that beta band activity at rest is associated with depression alluding to its potential utility as a neural index. Furthermore, novel approaches now examine beta band activity using non-averaged trial-by-trial analyses that capture transient, high-power, spike-like events (i.e., beta bursts). These bursts have been characterized by their power, duration, number, and frequency span (f-span). Thus, the objective of the present study was to investigate whether spectral event features, specifically beta burst features, are associated with self-reported depression severity. Young adults (N = 29, M age = 23.9 years, 78.4% female) had their brain activity recorded at rest and completed questionnaires regrading depression, anxiety, stress (Depression Anxiety Stress Scale – 21 Item; DASS-21), and demographics. EEG was acquired using 13-channel BIOPAC. The rate, power, duration, and f-span of transient spectral events were calculated for all electrodes (Fz, F3, F4, Cz, C3, C4, Pz, P3, P4, T5, T6, O1, O2). The associations between depression severity and spectral event features were calculated while controlling for age, sex, treatment history, anxiety severity, and stress severity. Correction for multiple comparisons was applied to the initial correlations, as event features across all electrodes were calculated (i.e., p < .05/13 electrodes = .003). After correction for multiple comparisons elevated depression severity was positively associated with the length of parietal (i.e., P4, p < .001, r = .72) and occipital (i.e., O2, p < .002, r = .56) beta burst duration. A hierarchical linear regression model was conducted to determine the unique variance P4 and O2 beta duration account for in explaining depression severity while controlling for relevant demographic and other emotional distress variables (i.e., age, sex, treatment history, anxiety, and stress). This model was significant (p = <.001, R2 = .69, ΔR2 = .32) with P4 beta duration as a significant predictor (p = .03, β = .57). However, O2 beta duration was not a significant predictor in the model (p = .62, β = .11). As a secondary/follow-up analysis, depression severity scores were re-classified into the clinical cutoffs for clinically significant depression (DASS-D severity > 14). Receiver Operating Characteristic (ROC) curve was conducted, and P4 beta duration showed a statistically significant area under the curve (p < .001, AUC = .80, “acceptable” classification) with a burst duration longer than 17 ms providing optimal sensitivity (66.7%) and specificity (75.0%) in classifying positive cases. Higher severity is DASS-Depression is associated with longer parietal beta burst duration. Beta burst duration may have potential utility as a biomarker to understand the underlying pathophysiology of depression. This is congruent with emerging research showing the association between beta burst characteristics with depression.