Relationship of age to power spectrum analysis of EEG during sleep

We describe a method of power spectrum analysis of all-night sleep EEGs in 20 normal subjects aged 18-43 years. The analysis calculates the energy of the whole sleep period, the various frequency bands, and selected sleep stages automatically. The numerical value of the energies was directly comparable between subjects, presented in numbers (microV2s) and displayed as a color density spectral array with fixed power values for each color. A significant relationship between age and power spectra was found, in that total energy/minute was higher in younger subjects than in older. The energy in delta sleep/minute also showed an inverse relationship with age.