Abstract
Chronic alcoholism is associated with widespread regional differences from controls in brain activity and connectivity dynamics measured by blood-oxygen-level-dependent (BOLD) signals. Identification of alcoholism-related neurofunctional power dynamics using functional magnetic resonance imaging (fMRI) that relate to cognition and behavior may serve as biomarkers of alcoholism. Previously, resting-state fMRI studies examined BOLD signals at a single low-frequency (LF) bandwidth. BOLD signals, however, oscillate systematically at different frequencies and are organized in a resting brain where LF oscillation facilitates long-distance communication between regions across cortical regions, whereas high-frequency (HF) oscillation occurs in closely localized, subcortical areas. Using a frequency power quantification approach, we investigated whether the organization of BOLD signal oscillations across all measured frequency bandwidths is altered in alcoholism and relates to cognitive performance. Frequency-dependent oscillation power differences between 56 sober alcoholics and 56 healthy controls occurred for all frequency bands. Alcoholics exhibited greater frequency oscillation power in the orbitofrontal cortex and less power in the posterior insula within the HF bandwidth than controls. Aberrant orbitofrontal HF power was associated with poorer memory performance and slower psychomotor speed in alcoholics. Middle-frequency and LF power proved sensitive in detecting altered frequency oscillation dynamics in parietal and postcentral cortical regions of alcoholics. This study is novel in identifying alcohol-related differences in BOLD oscillation power of the full fMRI frequency bandwidth. Specifically, HF power aberrations were associated with poorer cognitive functioning in alcoholism and may serve as a biomarker for identifying neural targets for repair.