1. INTRODUCTION AIMS AND PREDICTIONS METHODS Participants: 18 children (9-10; M = 10). 38 young adults (20-30; M = 24) 26 older adults (65-85; M = 72) EEG recording: 62 sintered Ag/AgCl electrodes referred to nosetip continuous, DC-100 Hz 500 Hz sampling rate epoch: 450 ms prior to and 450 ms post reaction time (RT); -450 to -350 ms pre-RT baseline RESULTS SUMMARY AND DISCUSSION Cognitive control describes processes that work separately and in concert to select and coordinate the execution of willed actions, and are initiated in prefrontal cortex (1). The efficiency of control develops along with frontal lobe maturation throughout childhood (2) and declines, along with frontal lobe function, with aging (3). One such control process, the ability to monitor and detect response conflict, was assessed by measuring medial frontal negativity (MFN) amplitude in a variant of the flanker task (Figure 1) in children, young and older adults. MFN, a response-locked ERP component, emanates from the ACC and reflects the amount of response conflict present when a decision must be made (4). Figure 1 Congruent Incongruent Neutral Stimuli and Procedure Stimulus duration = 200ms 900-1200 ms jittered ISI Subjects respond in same and opposite direction to central arrow (3 blocks of each condition). 120 trials/block; 33 percent probability of congruent, incongruent and neutral flankers Only correct-response trials are analyzed in both the ERP and behavioral data. Neutral trials are not considered in the analyses; their inclusion does not change the basic results. Assess developmental changes in the monitoring, detection and resolution of response conflict engendered by an error, to competing response tendencies (incongruent responses) and irrelevant information (incongruent flankers). We predicted that children and older adults would perceive greater response conflict than young adults and would have difficulty implementing control when such conflict was heightened on post-error and both types of incongruent trials. Increases in the amount of response conflict would be demonstrated by increments in MFN magnitude. Control would be demonstrated by RT slowing on post-error and both types of incongruent trials and perhaps no decrement in accuracy on post-error trials. Post-error trials: Young and older adults, but not children, show post-error slowing (Fig. 2A). Importantly, children and older adults produced a greater percentage of errors on post-error than post-correct trials (Fig. 2B). Opposite-direction trials: Young and older adults, but not children, show slowing to opposite- compared to same-direction trials (Figure 2C); decreases in accuracy for opposite- than same- direction trials are reliable for older adults only (not shown). Incongruent flankers: Incongruent relative to congruent flankers lead to age-invariant RT slowing and reductions in accuracy (not shown). Behavioral Data Figure 4 depicts the current source density distributions of MFN activity (-50 to + 50 ms window) for the post-error ERPs depicted in Figure 3A. The topographies suggest that the neural processes underlying the MFN are qualitatively similar in the 3 groups. Dots represent the electrode sites. Figure 4 Young AdultsOlder Adults Children + -  V/cm 2 ±0.15 ±0.08 The MFN data suggest that, in general, older adults perceive greater response conflict than young adults, whereas children do not appear to detect differing levels of conflict appropriately. The behavioral data indicate that children and older adults do not exert sufficient cognitive control to overcome the heightened response conflict present on post-error (lack of slowing in children and decreases in accuracy in children and older adults) and opposite-direction (lack of slowing in children) trials. Implementing control to counteract interference from irrelevant information appears relatively intact in children and older adults. The results suggest that regions of the ACC that support the ability to monitor and detect response conflict function appropriately in older adults but are still maturing in 9-10 year olds. However, children and older adults show difficulty in exerting control to resolve these conflicts perhaps due, respectively, to incomplete maturation of and breakdown in the prefrontal circuits responsible for cognitive control (5,6). The Development of Cognitive Control: An ERP Investigation in Children, Young and Older Adults David Friedman 1, Yael M. Cycowicz 2, Cort Horton 1, Doreen Nessler 1, & Marianne de Chastelaine 1 1 Cognitive Electrophysiology Laboratory, and 2 Division of Brain Stimulation, New York State Psychiatric Institute, NY, NY RT in ms Figure 2A Post Error Post Correct 100 200 500 600 300 400 Percent Correct 20 60 40 100 80 0 1. E. K. Miller, Nature Reviews: Neuroscience 1, 59 (2000). 2. F. N. Dempster, Developmental Review 12, 45 (1992). 3. R. L. West, Psychological Bulletin 120, 272 (1996). 4. W. J. Gehring, A. R. Willoughby., Science 295, 2279 (2002). 5. S. A. Bunge, N. M. Dudukovic, M. E. Thomason, C. J. Vaidya, J. D. Gabrieli, Neuron 33, 301 (2002). 6. D. Nessler, D. Friedman, R. Johnson, Jr., M. Bersick, Neurobiology of Aging (in press). * * * * Figure 2B * P <0.05 Figure 3B Opposite Same Figure 3A Post Error Post Correct Incongruent Congruent Figure 3C Post-error trials: MFNs are larger only in young and older adults on post-error compared to post-correct trials; this effect is greater in older adults (Fig. 3A). Opposite-direction trials: Young and older adults show larger MFNs on opposite- than same-direction trials; children do not (Fig. 3B). Incongruent flankers: Unexpectedly, incongruent relative to congruent flankers do not elicit larger MFNs (Fig. 3C). Figure 2C * Same Opposite Young AdultsOlder AdultsChildren * RT in ms 100 200 500 600 300 400 -450450-450 RT 450 RT Young Adults Older Adults Children RT-Locked ERPs REFERENCES * MFN * * FPz AFz Fz * FPz AFz Fz FPz AFz Fz