ERP correlates of performance monitoring in elderly.

Previous studies on performance monitoring repeatedly found attenuated error-related negativities (Ne/ERN) in elderly, while findings for the correct-related negativity (Nc/CRN) are inconsistent. The present study aimed at clarifying inconsistent Nc/CRN results in elderly. Therefore, a refined design was employed to control for potential influences on the Nc/CRN, namely decision uncertainty and partial error processing. Further, we intended to study Nc/CRN variations with trial compatibility that were found in previous studies for younger but not for older adults. Results revealed increased Nc/CRN and decreased Ne/ERN amplitudes in older compared to younger adults. While the Ne/ERN was larger than the Nc/CRN in younger adults, both components were similar-sized in older adults. Further, a modulation of Nc/CRN amplitudes between compatible and incompatible trials was observed in younger adults, but was absent in older adults. Reduced differentiation of response-related negativities with response accuracy or stimulus compatibility in elderly suggests a reduced adaptation of associated processes to changing demands. Further, this might also point to different processes reflected by Nc/CRN and Ne/ERN and to reduced error-specific monitoring but increased general or strategic monitoring in elderly.

Age-related alterations in performance monitoring during and after learning.

Changes of performance monitoring during and after learning of stimulus-response (S-R) associations were examined in younger and older adults using event-related brain potentials (ERPs). To determine whether age-related differences in performance monitoring are attenuated when learning performance is equalized between age groups, participants learned S-R associations until they reached a learning criterion. After completed learning, time pressure was introduced to assess whether time pressure pronounces age-related differences in performance monitoring. In the initial learning phase, performance monitoring was unspecific, indicated by similar-sized ERN/Ne (error-related/error negativity) and CRN (correct response-related negativity) amplitudes. With advanced learning, ERN/Ne and CRN amplitudes dissociated in younger participants, suggesting error-specific performance monitoring, whereas in older participants, performance monitoring remained unspecific. However, after completed learning and under time pressure, larger ERN/Ne than CRN amplitudes were observed in older participants, too. This finding suggests that error-specific performance monitoring develops with learning in younger, but not necessarily in older adults. The administration of time pressure seemed to force older adults to monitor error responses more specifically.

Aging, probabilistic learning and performance monitoring.

This study examined age-related alterations in electrocortical indices of performance monitoring in a probabilistic learning situation. The results showed differences between younger and older adults in the modulation of response- and feedback-related negativities by feedback validity. In younger adults, the error-related negativity/error negativity (ERN/Ne) was larger than the correct response-related negativity (CRN) in the high and medium feedback validity condition, whereas the feedback-related negativity (FRN) differentiated between positive and negative feedback in the low feedback validity condition. In older adults, amplitudes of the ERN/Ne and the CRN were dissociated only in the high feedback validity condition and the FRN did not differ between positive and negative feedback. This finding might suggest that performance monitoring is more prone to interference by probabilistic information in older adults. Further, the salience of negative feedback seems reduced with age.

Changes of performance monitoring with learning in older and younger adults.

This study examined age differences in performance monitoring during learning of a stimulus-response association task. At the beginning of the learning process, amplitudes of the response-locked error-related negativity (ERN or Ne) and correct response-related negativity (CRN) did not differ in both age groups. With advanced learning the response ERN/Ne increased and the CRN decreased in younger adults, but did not dissociate in older adults. Feedback ERN amplitudes decreased with learning in both age groups and were reduced in older relative to younger adults. Results indicate that performance monitoring became error specific with advanced learning in younger adults, but not in older adults. This might be due to weak representations of stimulus-response mappings in older adults as they learned worse. The decrease of the feedback ERN with learning and aging might suggest a decline of attention paid to negative feedback.