Microbiome stewardship: definition and guiding principles for implementation.

Understanding changes in neural activity across the lifespan is necessary for distinguishing between healthy and pathological aging. Since its development over 100 years ago, electroencephalography (EEG) has emerged as a critical tool for neuroscientists and clinicians hoping to understand how, why, and in whom neural activity changes. Traditionally, EEG research has focused on oscillatory (periodic) activity within canonical bands. Notably, variation in the amplitude and peak frequencies of alpha (8-13 Hz) oscillations has been linked to attention, memory, and executive control (; ; ). However, emerging evidence highlights the importance of non-oscillatory (aperiodic) activity as a distinct and complementary index of neural functioning. Despite recent advances in spectral parameterization techniques () and the understanding of the physiological significance of aperiodic activity (), the temporal reliability and stability of these measures remain poorly characterized. One reason for this is because almost all studies use cross-sectional data or have examined reliability and change over short timeframes, reducing their utility as reliable markers of neural ageing or indicators of pathological deviation. To address this gap, we investigate within-person changes in both aperiodic and periodic components of the EEG power spectrum across a five-year period. We show that parametrized aperiodic and periodic activity are reliable over time through rank order stability across a five-year timespan. Additionally, we demonstrate that these measures change systematically as a function of time. The majority of the neural power spectrum is comprised of non-oscillatory (i.e., aperiodic) activity, which follows a 1/f^χ-like distribution, such that as frequency increases, power decreases. The most used method for parameterizing aperiodic activity is SpecParam (formerly FOOOF; ), which characterizes aperiodic activity into two parameters: the aperiodic exponent (χ) and the offset. The exponent reflects the rate at which power decreases across frequencies, while the offset indicates the overall magnitude of neural activity across the frequency spectrum. Computational and pharmacological studies suggest that the aperiodic exponent is a non-invasive index of circuit-level excitation:inhibition (E:I) balance (; ; ), with greater exponents (i.e., steeper slopes) associated with greater cortical inhibition. Conversely, lower exponents (i.e., flatter slopes) are linked to increased excitability, reduced inhibition, and greater cortical noise (). Alterations in the offset parameter reflect changes in neuronal firing rates and arousal states (). These aperiodic parameters show systematic variation in healthy aging. Cross-sectional studies report that the aperiodic exponent increases with age from infancy through early adolescence (; ), before gradually decreasing from early adulthood to older age (; ). Beyond age-related alterations, systematic variation in the aperiodic exponent has been linked to systematic variation in cognitive functioning. Individuals with lower (i.e., flatter) exponents tend to perform worse on cognitive tasks that assess response time, working memory capacity, and attentional control (; ; ). Moreover, consistent with the E:I balance model, the exponent is altered in multiple developmental and neurological disorders associated with altered excitatory and inhibitory neurotransmission, including Parkinson's disease (; ; ; ), ADHD (), and autism spectrum disorder (). Similar findings have been observed for the aperiodic offset, with lower offsets associated with poorer verbal fluency (), and reduced cognitive control ().