Decomposing the role of alpha oscillations during brain maturation

Abstract

Childhood and adolescence are critical stages of the human lifespan, in which fundamental neural reorganizational processes take place. A substantial body of literature investigated accompanying neurophysiological changes, focusing on the most dominant feature of the human EEG signal: the alpha oscillation. Recent developments in EEG signal-processing show that conventional measures of alpha power are confounded by various factors and need to be decomposed into periodic and aperiodic components, which represent distinct underlying brain mechanisms. It is therefore unclear how each part of the signal changes during brain maturation. Using multivariate Bayesian generalized linear models, we examined aperiodic and periodic parameters of alpha activity in the largest openly available pediatric dataset (N=2529, age 5-22 years) and replicated these findings in a preregistered analysis of an independent validation sample (N=369, age 6-22 years). First, the welldocumented age-related decrease in total alpha power was replicated. However, when controlling for the aperiodic signal component, our findings provided strong evidence for an age-related increase in the aperiodic-adjusted alpha power. As reported in previous studies, also relative alpha power revealed a maturational increase, yet indicating an underestimation of the underlying relationship between periodic alpha power and brain maturation. The aperiodic intercept and slope decreased with increasing age and were highly correlated with total alpha power. Consequently, earlier interpretations on age-related changes of total alpha power need to be reconsidered, as elimination of active synapses rather links to decreases in the aperiodic intercept. Instead, analyses of diffusion tensor imaging data indicate that the maturational increase in aperiodic-adjusted alpha power is related to increased thalamocortical connectivity. Functionally, our results suggest that increased thalamic control of cortical alpha power is linked to improved attentional performance during brain maturation.

Keywords: 1/f; DTI; EEG; alpha; brain development; human; neuroscience; thalamocortical connectivity.

Figure 1.. Visualization of data of the main HBN sample used in the Bayesian regression model.

Solid lines represent fitted regression lines. The schematic head on the right indicates the location of the electrode cluster from which data was aggregated. The results revealed a decrease in total individualized alpha power with increasing age. Importantly, this relationship inverts when individualized alpha power is adjusted for the aperiodic signal, which then shows an age-related increase in power. Furthermore, relative individualized alpha power exhibits a positive relationship to brain maturation. An age-related increase of the IAF and a decrease of the aperiodic intercept and slope are also indicated (bottom row).

Figure 1—figure supplement 1.. Distribution of age and gender in the final included sample plotted in Figure 1 and used for the statistical analyses.

Figure 1—figure supplement 2.. Visualization of data of the HBN subsample without any given diagnosis used in the Bayesian regression models.

Solid lines represent fitted regression lines. The schematic head (right) indicates the location of the electrode cluster from which data was aggregated.

Figure 2.. Visualization of age-related changes during brain maturation in (A) the measured power spectrum (i.e., total power spectrum), (B) the periodic (i.e. aperiodic-adjusted) power spectrum, and (C) the aperiodic signal.

Younger children represent the 20% youngest children in the sample, young adults the 20% oldest participants. This split of the sample was only done for visualization purposes and not used in any statistical analysis.

Figure 3.. Visualization of data of the validation sample used in the Bayesian regression model.

Solid lines represent fitted regression lines. The schematic head on the right indicates the location of the electrode cluster from which data was aggregated.

Figure 3—figure supplement 1.. Distribution of age and gender in the final included validation sample plotted in Figure 3 and used for the statistical analyses.

Figure 3—figure supplement 2.. Visualization of data of the healthy validation subsample used in the Bayesian regression models.

Solid lines represent fitted regression lines. The schematic head (right) indicates the location of the electrode cluster from which data was aggregated.

Figure 4.. Illustration of the two components (left) superimposed in the measured neural power spectrum (right).

The dark blue bar (right) indicates how total power is assessed relative to the absolute zero. The light blue bar represents aperiodic-adjusted power, which is assessed relative to the aperiodic signal.

Appendix 1—figure 1.. Visualizations of possible fallacies in total and relative power measures in simulated data.

Bar plots on the right indicate the difference in alpha power between simulated data 2 and simulated data 1 in A, and simulated data 4 to simulated data 3 in B. (A) Two simulated power spectra with identical true alpha oscillatory power. The high amplitude oscillation in the theta range (~5 Hz) in simulated data 1 conflates results in relative power differences in the alpha band. (B) Two simulated power spectra with identical alpha oscillatory power. Here, differences in the aperiodic intercept and slope between the two signals conflate results in total and relative power differences in the alpha band.

Appendix 2—figure 1.. Flow chart of exclusion criteria applied to the main HBN dataset, yielding 1770 subjects plotted in Figure 1 and used for the statistical analyses.