This proposal exploits the opportunity created by the existence of a normative qEEG data base of 250 normal adults aged 17 to 90, as well as qEEG database of over 100 subjects with crack cocaine dependence, built by previous and ongoing NIDA grant support. Cocaine dependent subjects were tested at baseline 5 to 14 days after their last use and retested at 1, 3, 6, 9, 12 and 18 months of drug free abstinence. The serial observations already obtained on these cocaine dependent subjects represents a resource which greatly multiplies scientific return relative to the budget limits of the present proposal. In order to develop a measure set to study change in cocaine withdrawal, we propose to systematically factor analyze the qEEG on 250 normal subjects in the spatial domain defined by the standard 10/20 lead placements, and the temporal domain consisting of the frequency spectrum from 0.39 to 40 Hz utilizing spatial-temporal principal components analysis (STPCA). This will allow comparison to the 100 cocaine dependent subjects evaluated in our ongoing NIDA project. The major goals of this research are to: [1] Construct a set of age regressed normative values for qEEG factors; [2] Systematically map the normal factors to their neuroanatomical generators utilizing low resolution electromagnetic topography (LORETA), a source localization technique; [3] To create a qEEG atlas of the LORETA localization of significant EEG factors; and [4] Test specific hypothesis utilizing data we have previously obtained from 100 subjects with cocaine dependence regarding normalization and prediction of efficacy, in extended abstinence in drug free residential treatment. Gender differences on the qEEG factors will also be explored. This proposed research addresses two key issues in qEEG research: [1] The need to span the large univariate measure set with a smaller number of multivariate measures to achieve parsimony; [2] The need for hypothesis driven qEEG research based on a knowledge of the correspondence between the EEG and its underlying neuroanatomical generators.
