ON THE DEVELOPMENT OF VOLUNTARY AND REFLEXIVE COMPONENTS IN HUMAN SACCADE GENERATION

The saccadic performance of a large number (n=281) of subjects of diff erent ages (8-70 years) was studied applying two saccade tasks: the pr osaccade overlap (PO) task and the antisaccade gap (AG) task. From the PO task, the mean reaction times and the percentage of express saccad es were determined for each subject. From the AG task, the mean reacti on time of the correct antisaccades and of the erratic prosaccades wer e measured. In addition, we determined the error rate and the mean cor rection time, i.e. the time between the end of the first erratic prosa ccade and the following corrective antisaccade. These variables were m easured separately for stimuli presented (in random order) at the righ t or left side. While strong correlations were seen between variables for the right and left sides, considerable side asymmetries were obtai ned from many subjects. A factor analysis revealed that the seven vari ables (six eye movement variables plus age) were mainly determined by only two factors, V and F. The V factor was dominated by the variables from the AG task (reaction time, correction time, error rate) the F f actor by variables from the PO task (reaction time, percentage express saccades) and the reaction time of the errors (prosaccades!) from the AG task. The relationship between the percentage number of express sa ccades and the percentage number of errors was completely asymmetric: high numbers of express saccades were accompanied by high numbers of e rrors but not vice versa, Only the variables in the V factor covaried with age. A fast decrease of the antisaccade reaction time (by 50 ms), of the correction times (by 70 ms) and of the error rate (from 60 to 22%) was observed between age 9 and 15 years, followed by a further pe riod of slower decrease until age 25 years. The mean time a subject ne eded to reach the side opposite to the stimulus as required by the ant isaccade task decreased from approximate to 350 to 250 ms until age 15 years and decreased further by 20 ms before it increased again to app roximate to 280 ms. At higher ages, there was a slight indication for a return development. Subjects with high error rates had long antisacc ade latencies and needed a long time to reach the opposite side on err or trials. The variables obtained from the PO task varied also signifi cantly with age but by smaller amounts. The results are discussed in r elation to the subsystems controlling saccade generation: a voluntary and a reflex component the latter being suppressed by active fixation. Both systems seem to develop differentially. The data offer a detaile d baseline for clinical studies using the pro- and antisaccade tasks a s an indication of functional impairments, circumscribed brain lesions , neurological and psychiatric diseases and cognitive deficits. (C) 19 97 Elsevier Science B.V.