Rhythmic activity in resting-state EEG predicts trait anxiety

Fig. 1.. (Left) Experimental design: subjects received verbal and written instructions to begin with their eyes open and to close their eyes gently upon hearing a beep. The beep was presented every 1 minute (instructing participants to open or close their eyes) with four beep presentations in total, constituting 4 minutes (2 minutes eyes-open and 2 minutes eyes-closed) EEG recording. (Right) 21 electrodes of the International 10–20 system for EEG with highlighted electrodes were used in our analysis. Namely, red represents Fz and Pz electrodes used in alpha and theta power and oscillatory multiple regressions and correlations; gray represents F3 and F4 for mid-frontal alpha asymmetry, and orange represents F7 and F8 for lateral-frontal alpha asymmetry analyzes. Note: The EGI Geodesic-Sensor-Net used in the experiment consists of 256 electrodes; a 10–20 system is provided as an easier reference.

Fig. 2.. These figures demonstrate the detection of oscillatory activity (A) and its predictive relationship with BIS scores (B). (A) Example of the raw EEG traces recorded during eyes open rest at electrode Fz, with oscillations detected by BOSC in theta (in red) and alpha (in yellow) ranges in one high-BIS (top, score 48) and one low-BIS (bottom, score 23) participant. (B) The representation of the multiple regression model predicting BIS scores with theta and alpha oscillatory activity (Pepisode) measures recorded at frontal midline (Fz) electrode during eyes open and eyes closed resting states. Red bars represent beta coefficients for theta oscillations predicting BIS, and yellow bars represent beta coefficients for alpha oscillations predicting BIS.

Fig. 3.. Reliability analysis was performed using bootstrapping with sub-sampled EEG data segments ranging from 10% to 100% of the total recording duration (48 seconds for oscillation and 60 seconds for power analysis). For each percentage window of data used, we performed 100 bootstrapped linear regressions to predict BIS scores using mid-frontal (Fz) and mid-parietal (Pz) theta and alpha oscillations (A and C) and power (B and D) recorded during eyes-open (EO) and eyes-closed (EC) conditions. The error bars represent the 95% confidence intervals of the standardized beta coefficients.

Fig. 4.. Scatter plots for Pearson correlations (df=45) between eyes-open alpha (A) and theta (B) oscillations recorded at Fz electrode and BIS scores. Red dots represent individuals with self-reported anxiety diagnosis.

Fig. 5.. Pearson correlation (df=45) for BIS anxiety scores correlating with resting-state neural oscillations (Pepisode) recorded at Fz and Pz electrodes during eyes-open (A) and eyes-closed (B) conditions plotted for all frequencies. The dashed lines denote the significance thresholds (p<0.05).

Fig. 6.. The representation of the multiple regression model predicting BIS scores with lateral-frontal (F8–F7) and mid-frontal (F4–F3) alpha oscillatory (Pepisode) asymmetry measured during eyes-open and eyes-closed resting states. Orange bars represent beta coefficients for lateral-frontal (F8–F7) alpha oscillatory asymmetry measures predicting BIS, and gray bars represent beta coefficients of mid-frontal (F4–F3) alpha oscillatory asymmetry measures predicting BIS scores.