Volitional eyes opening perturbs brain dynamics and functional connectivity regardless of light input

The act of opening (or closing) one's eyes has long been demonstrated to impact on brain function. However, the eyes open condition is usually accompanied by visual input, and this effect may have been a significant confounding factor in previous studies. To clarify this situation, we extended the traditional eyes open/closed study to a two-factor balanced, repeated measures resting state fMRI (rs-fMRI) experiment, in which light on/off was also included as a factor. In 16 healthy participants, we estimated the univariate properties of the BOLD signal, as well as a bivariate measure of functional connectivity and multivariate network topology measures. Across all these measures, we demonstrate that human brain adopts a distinctive configuration when eyes are open (compared to when eyes are closed) independently of exogenous light input: (i) the eyes open states were associated with decreased BOLD signal variance (P-value=0.0004), decreased fractional amplitude of low frequency fluctuation (fALFF. P-value=0.0061), and decreased Hurst exponent (H. P-value=0.0321) mainly in the primary and secondary sensory cortical areas, the insula, and the thalamus. (ii) The strength of functional connectivity (FC) between the posterior cingulate cortex (PCC), a major component of the default mode network (DMN), and the bilateral perisylvian and perirolandic regions was also significantly decreased during eyes open states. (iii) On the other hand, the average network connection distance increased during eyes open states (P-value=0.0139). Additionally, the metrics of univariate, bivariate, and multivariate analyses in this study are significantly correlated. In short, we have shown that the marked effects on the dynamics and connectivity of fMRI time series brought by volitional eyes open or closed are simply endogenous and irrespective of exogenous visual stimulus. The state of eyes open (or closed) may thus be an important factor to control in design of rs-fMRI and even other cognitive block or event-related experiments.