Objectives: The objectives were to assess life-span alterations in intrasubject variability of EEG spectral amplitudes and in amplitude and mean frequency computed from spectral amplitude profiles.
Methods: EEG was recorded from the central and occipital scalp of 222 healthy males aged 4 to 90 years. Amplitude spectral profiles derived from FFT procedures provided data for the computation of amplitude variability (ampCV), absolute and relative amplitude and mean frequency for each of five spectral bands. Analysis of variance, coefficient of correlation and t test were employed in data analyses.
Results: The ampCV measure did not provide direct evidence of age-related intrasubject changes in EEG frequency but was useful for indexing the presence of rhythmic EEG activity. Marked decreases in absolute amplitude occurred during childhood with little change thereafter. Age changes in relative amplitude were more complex and differed for slow and fast EEG activity. We also report that an algorithm frequently used to estimate mean frequency can introduce consistent bias into mean frequency computations under some conditions.
Conclusions: Life-span changes in measured EEG characteristics were generally consistent with those from earlier studies. It is important that investigators validate and describe procedures used to determine mean frequency of EEG spectral data.