Each year at the Flux Congress, attendees have the opportunity to vote for their favorite posters, highlighting research that stands out for its innovation, impact, and presentation. As one of the 2024 People's Choice winners, Dashiell Sacks, PhD captured the attention of the Flux community with their compelling poster. In this blog post, they delve into the inspiration behind their research, key findings, and what it means to be recognized by peers at the forefront of developmental cognitive neuroscience.

Background and relevance

EEG is a particularly appealing modality for investigating potential biomarkers of mental health, given that it is portable and affordable relative to many other neuroimaging modalities. Using EEG, we can measure the brain’s electrical activity with excellent temporal resolution, at the scale of milliseconds. Traditional approaches for analyzing EEG data typically focus on averaging EEG spectral power within predefined frequency bands, such as delta, theta, alpha, beta, and gamma. However, this approach can conflate periodic, oscillatory brain activity in the EEG signal with aperiodic ‘background noise.’ SpecParam, an algorithm originally designed by Bradley Voytek’s group (Donoghue et al., 2020), allows us to parameterize the EEG signal into both periodic and aperiodic components.

Recent hypotheses suggest that the aperiodic component of the EEG signal may reflect mechanistic underpinnings of brain activity. This aperiodic component is characterized by a 1/fx distribution in which power decreases as frequency increases, which can vary in ‘slope’ (i.e., rate at which power decreases as frequency increases) and ‘offset’ (i.e., uniform shift of power across frequencies). The aperiodic slope and offset have been suggested to index cortical excitatory-inhibitory (E/I) balance (flattened slope indicates increased excitation over inhibition) and broadband neuronal firing, respectively. This was the motivation for our current work at Boston Children’s Hospital/Harvard Medical School, where we are investigating early brain development and mental health (particularly anxiety) longitudinally in children, starting from infancy.

When investigating aperiodic activity across the brain in our longitudinal sample, we noticed an interesting trend from infancy to 7 years that corresponds with recent cross-sectional research: Aperiodic activity followed a non-linear developmental trajectory (Figure 1). Specifically, in early childhood we observed increases in offset and steepening slope, followed by subsequent decreases in offset and flattening slope. Consistent with the hypothesis that offset reflects broadband neuronal firing, this observed shift in trajectory during childhood could mark the transition that occurs in the brain from rapid neurodevelopmental growth in early childhood to synaptic pruning and ongoing systems optimization in later childhood. Furthermore, the shift from steepening to flattening slope may reflect unique maturational processes in inhibitory networks, including ongoing migration, maturation, and integration of GABAergic inhibitory neurons into the cortices, which may contribute to early steepening of the aperiodic slope.

When we analyzed aperiodic activity separately in frontal, temporal, central, and posterior electrode regions, we observed variations in the non-linear developmental trajectory, with a pattern of earlier transition from steepening to flattening slope seen in posterior regions compared to frontal regions (Figure 2). This pattern is similar to the pattern of development seen across studies investigating structural brain development during childhood, with earlier maturation seen in ‘simpler’ posterior brain regions, compared to more ‘complicated’ structures like the prefrontal cortex. We also observed sex differences, with female participants showing steeper slopes and reduced offsets compared to male participants at ages 3 years and 5 years – another pattern also observed in studies of brain structure.

Building on these findings, we were interested in investigating potential associations between aperiodic activity and maternal and child anxiety symptoms. We observed an interesting pattern, whereby the nature of the association between slope and maternal anxiety varied depending on age (Figure 3). At 3 years (when slope was steepening), maternal anxiety was associated with a flatter slope. At 7 years (when slope was flattening), maternal anxiety was associated with a steeper slope. We also observed some interesting associations between aperiodic activity and child symptoms. For example, the aperiodic offset was associated with anxiety symptoms, but only in female participants at 3 and 5 years, when we also observed significant sex differences in offset.

More research is required to validate the biological mechanisms aperiodic EEG activity is indexing and to better understand its associations with mental health. With ongoing research, we hope this area of study will help us better understand early childhood development and the neural mechanisms underpinning mental health.

Donoghue, T., Haller, M., Peterson, E. J., Varma, P., Sebastian, P., Gao, R., Noto, T., Lara, A. H., Wallis, J. D., Knight, R. T., Shestyuk, A., & Voytek, B. (2020). Parameterizing neural power spectra into periodic and aperiodic components. Nature Neuroscience, 23(12), 1655-1665. https://doi.org/10.1038/s41593-020-00744-x