Object: Increases in low-frequency fluctuations (LFF, 0.01-0.1 Hz) of BOLD fMRI signal were observed during sedation and sleep. We tested the hypothesis that sedation-induced LFF increases may be mediated by increased head motion likely associated with these brain states.
Materials and methods: Twenty healthy volunteers were scanned in resting-state baseline and (15 thereof) in midazolam sedation condition. After standard preprocessing, LFF power was computed from average time series from 14 ROIs representing 5 functional networks (visual, motor, auditory, default mode, dorsal attentional). We assessed the effects of sedation on LFF and indirect effects of head motion (mean relative displacement) using mediation analysis.
Results: Sedation increased LFF power with the indirect effect of motion on sedation-induced LFF significant for all networks. Global mean corrected data showed similar but weaker results: significant total effect for auditory, motor and visual; significant mediation for auditory and motor networks. Motion was strongly correlated with LFF even at baseline. Reported results were significant at p < 0.05.
Conclusion: The study confirms that conscious sedation induces LFF increase, but demonstrates that this effect is mediated by increased head motion. Hence, caution in design and interpretation is recommended for studies comparing subjects or conditions differently predisposing to head motion.