Performance - Based Clustering Enables the Study of Physiological Features Supporting Sustained Attention Capacities

Abstract

AbstractIn the present study, we investigated different sustained attention performance profiles in the general adult population and if age is a determinant factor among these profiles. We also reported some specificities in terms of brain oscillations. The sustained attention to response task was applied combined with electrophysiological recordings in 59 adults, aged from 19 to 86 years. We used a hierarchical cluster analysis to group individuals based on similarity of levels and patterns of performance across variables of interest. We focused on the most common attentional–related variables, which allowed us to identify four distinct clusters showing specific sustained attention profiles. Furthermore, our analysis clearly revealed that age was not the determinant factor of these profiles. Finally, we highlighted specificities in terms of brain oscillations. Subjects from cluster 1– the “high-performers” – demonstrated a globally increased theta-gamma coupling maintained over time and an increase of theta activity at the FC6 and F8 electrodes. Subjects from cluster 2, who adopted a more cautious response strategy, presented a higher level of modulation exerted by the electrode FT7 over Oz and a higher theta-gamma coupling at the Oz. This result suggests that cluster-based method to useful in understanding specific mechanisms underlying sustained attention features.