Occipital neural dynamics in cannabis and alcohol use: independent effects of addiction.

MEG sensor-level spectrograms and source images. (A) Sensor-level analysis revealed distinct oscillatory responses within the theta (4–8 Hz), alpha (8–16 Hz), and gamma (64–74 Hz) bands during visual processing in the occipital cortices. Each spectrogram reflects activity of a representative occipital MEG sensor that has been grand-averaged across all participants. The color scale legend shown on the right displays percent change from baseline (red indicating an increase in power relative to baseline and blue reflecting a decrease relative to baseline). (B) Significant sensor-level time–frequency components were imaged using a beamforming approach in each participant individually, and the resulting maps were grand-averaged across all participants per oscillatory response. The resulting functional maps reveal increases (i.e., synchronizations) in the theta and gamma range in the bilateral medial occipital cortices, as well as decreases in the strength (i.e., desynchronization) of alpha activity in more lateral occipital cortices. Color scale legends to the right of each image indicate average baseline-normalized power (pseudo-t) thresholds.

Voxel time series of alpha activity during visual-spatial processing. Grand averaged alpha source images revealed peak voxels in bilateral visual cortices. Time series were then extracted from these voxels in each participant, averaged across hemisphere, and were plotted with time (ms) on the x-axis and relative amplitude (%) on the y-axis. For display purposes, the time series has been averaged across those with AUD (teal) and those without AUD (purple). The area highlighted in gray (300 to 550 ms) reflects the time period that was used for the source analysis. Factorial ANCOVA showed a main effect of AUD, such that individuals diagnosed with AUD (teal) exhibited a significantly weaker alpha response compared to those without AUD (purple). Error bars indicate \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\pm 1$$\end{document}±1 standard error of the mean. To the right, source images of the alpha response have been split according to AUD diagnosis with the color scale legend shown on the right. These images clearly show the same effect observed in the time series data. Note that these images are shown for display purposes only and statistical comparisons were made using the average over the highlighted active window from the extracted time series in an ANCOVA model.

Spearman correlation between alpha power and number of AUD symptoms. Post-hoc testing of alpha power showed a significant correlation between number of AUD symptoms and alpha power such that participants that had more severe AUD exhibited weaker alpha responses. This relationship remained significant in a larger regression model that included number of CUD symptoms, and metrics of cannabis and alcohol consumption, showing that decreased alpha responses are associated with more severe AUD, above and beyond amount of alcohol consumption and cannabis consumption/use disorder. Note that more negative alpha power indicates a stronger alpha response (larger change from baseline). Shaded area represents 95% confidence interval.

Visual-spatial processing task. Every trial began with a fixation period lasting 1900 to 2100 ms, which was then followed by the stimulus grid (duration = 800 ms). The stimulus grid was presented in one of four locations; off-center by 75% to either the right or the left of the fixation crosshair in the top or lower quadrant. Participants were instructed to use the response pad to indicate whether the grid was presented to the left (index finger) or right (middle finger).