130 Chapter 6 Figure 4. Alpha and beta power averaged over the four regions showed a general pattern of lowering at T2, which reached nominal significance for posterior L alpha (p=0.034) and anterior / posterior L (p=0.034 and p=0.013, respectively). Dotted line: average over patients. Dashed line: average of HC. The comparison of AN at T1 to HC revealed significant differences for beta power all four regions (FDR-p=0.043). Power differences of AN versus HC No significant effects were observed for alpha power when comparing the four regions in a simple t-test (p>0.18). Beta power showed significant effects (uncorrected-p between 0.012 and 0.022; FDR-p=0.043 in all regions). To specify in more detail which channel by frequency combination may have been responsible for these significant effect, we calculated a linear effect comparing the 4 AN patients to the 8 HC for each of the 62 channels and each of the 1 Hz frequency bins from 1 to 30 Hz. Supplementary Figures S1-3 show the topographic distribution of power and significant electrodes after FDR correction. Supplementary figure S1 compared T1 against HC, figure S2 T2 against HC, and figure S3 T4 against HC. All times showed some significant differences compared to HC. For T1, the main differences were observed at lower (15 to 18 Hz) and upper (23 to 30 Hz) beta frequencies. These effects were localized at fronto-central electrodes. Alpha power did not show any significant change, except for a small right-central cluster at 9 Hz. T2 showed only few channels with significant increases (FDR q=0.05), consistent with the observed reduced power observed at T2 in the spectrum. T4 showed more widespread differences with strong increases in power compared to HC and many channel by frequency combinations, reaching corrected significance at FDR q=0.01. Interestingly, here, too, upper alpha power did not show significant differences. This may indicate that the observed power increases in the alpha band are confounded with the lower alpha peak
RkJQdWJsaXNoZXIy MjY0ODMw