EEG gamma-band synchronization in visual coding from childhood to old age: Evidence from evoked power and inter-trial phase locking

Objective: To investigate lifespan age differences in neuronal mechanisms of visual coding in the context of perceptual discrimination. Methods: We recorded EEG from 17 children (10-12 years), 16 younger adults (20-26 years), and 17 older adults (70-76 years) during a simple choice-reaction task requiring discrimination of squares and circles of different sizes. We examined age-group differences in the effect of stimulus size on early ERP components, evoked gamma-band power, and inter-trial phase-stability in the gamma band as assessed by the phase-locking index (PLI). Results: In the absence of age differences in discrimination accuracy, we observed reliable age differences in patterns of ERP, evoked gamma power, and PLI. PI and NI peak amplitudes were larger and the peak latencies longer in children than in adults. Children also showed lower levels of evoked power and PLI than adults. Older adults showed smaller increments in evoked power with increasing stimulus size than younger adults, but similar amounts of phase locking for small-and medium-sized stimuli as younger adults. Conclusions: The relative importance of different coding mechanisms in early visual areas changes from childhood to old age. Due to synaptic overproduction and immature myelination, the visual system of children is less entrained by incoming information, resulting in less synchronized neuronal responses. Adults primarily rely on sparse representations formed through experience-dependent temporally synchronized neuronal interactions. In old age, senescent decline in neuronal density and neurotransmitter availability further increase the reliance on temporally synchronized processing. Significance: Findings from this study defy the notion that sensory aging consists in a reversal of sensory development in childhood, and point to a high degree of age specificity in mechanisms of visual coding. (C) 2009 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.