We investigated the parameters of colored noise in EEG data of 17 722 professional drivers aged 18-70. The whole study is based upon experiments showing that biological neural networks may operate in the vicinity of the critical point and that the balance between excitation and inhibition in the human brain is important for the transfer of information. This paper is devoted to the study of EEG power spectrum which can be described best by a power function with 1/f(λ) distribution and colored noise corresponding to the critical point in the EEG signal has the value of λ = 1 (purple noise). The slow accumulation of energy and its quick release is a universal property of the 1/f distribution. The physiological mechanism causing energy dissipation in the brain seems to depend on the number and strength of the connections between clusters of neurons. With ageing, the number of connections between the neurons decreases. Learning ability and intellectual performance also decrease. Therefore, age-related changes in the λ coefficient can be anticipated. We found that absolute values of λ coefficients decrease significantly with increasing age. Deviations from this rule are related to age-dependent slowing of the dominant frequency in the alpha band. Age-dependent change in the parameter and colored noise may be indicative of age-related changes in the self-organization of brain activity. Results obtained include (i) the age-dependent decrease of the absolute values of the average λ coefficient with the regression coefficient 0.005 1/year, (ii) distribution of λ value changes related to EEG frequency bands and to localization of electrodes on the scalp, and (iii) relation of age-dependent changes of colored noise and EEG energy in separate frequency bands.