Still Wanting to Win: Reward System Stability in Healthy Aging

Figures

• 

  FIGURE 1. A scheme of the monetary incentive delay task (MID). The MID consisted of 450 trials (300 trials for the older group) separated into three blocks. Each trial randomly started with one of the three following reward incentive cues: a circle with one line indicated a possible gain of 3 cents; a circle with two lines, indicating a possible gain of 30 cents; and a triangle, indicating a possible gain of 0 cent. The cues were shown for 250 ms each. The trials cued with a triangle (0 cent-gain) served as the neutral control condition. Between the cue and the target, a fixation cross with randomized duration (750-1,250 ms) was presented. Target duration was adapted to individual reaction times, resulting in a predestined hit rate of 75%. Performance feedback was presented by a green laughing smiley if the time limit was not exceeded. A red sad smiley appeared otherwise. Afterward, pictures of coins visualized monetary rewards. Both feedback screens had duration of 1,000 ms each.

• 

  FIGURE 2. EEG channel groups (A), an overview of EEG channels and their assignment to regional channel groups. (B) Simplified scheme of EEG channel groups defined in 2 (A).

• 

  FIGURE 3. Behavioral data: (A), Boxplot: single mean reaction time data, lines/points correspond to different participants. P-values from the paired Mann-Whitney U-Test show significant reward-related reaction time speeding within both groups. Violin plots are based on single reaction time data. (B) and (E), the GEE model results for effects of the groups (young, old) and the cue (0 ct, 3 ct, 30 ct) on reaction times. Reaction time differences between cues did not significantly differ between groups. (D), a hit rate (accuracy) in percent for each block and condition. Older adults show higher hit rates than younger adults during the second and third blocks across all conditions. (C) and (F), analyzing the effect of the blocks on reaction times showed that the older adults significantly improved performance from Blocks B1 to B3 and B2 to B3. These improvements are significantly higher in comparison to the younger adults. The younger adults showed only a significant improvement between Blocks B1 and B2, not significantly differing from the older group; color code: green = old, pink = young, gray = contrast.

• 

  FIGURE 4. Analysis of reward effects on frontal intra- and internetwork connectivity and of group differences in reward effect on interregional connectivity; (A), a schematic overview over the analyzing procedure: Within-group coherence was analyzed for the reward vs. neutral cue contrast between frontal and non-frontal EEG channel groups. (B), Results show a higher frontal network and higher frontal-nonfrontal alpha-band coherence during the reward cue. There were no significant differences in the older group. (C), A schematic overview of the analyzing procedure: Between group comparison of interregional coherence for the reward vs. neutral cue contrast. (D), Results: significant differences appeared in the delta band; reward-based modulation was higher in the older group for frontocentral right and parietal right with parietooccipital and occipital channel groups. In detail, this applied to FCR with POR and O; and PR with POC and O (green arrows). Modulation was higher in the younger group for FCR with POL and POC (red arrows). Coherence was significantly different between PR and POL; both groups modulated it in the same absolute value, but in different directions (a black arrow).…

• 

  FIGURE 5. The HAROLD and PASA effect during reward processing: (A), a schematic overview of the analyzing procedure: Testing HAROLD (left side), group differences in coherence between frontal left and frontal right channel groups were assessed for the reward vs. neutral cue contrast. Testing PASA (right side), group differences in coherence between frontal and occipital channel groups were assessed for the reward vs. neutral cue contrast. (B), Results: a) HAROLD in the delta band, the older group exhibited lower reward-based coherence modulation within the left frontal and the right frontal channel groups. In the high-beta band, the older group showed lower differences in coherence within the right frontal channel group than the younger group (red arrows). PASA - the difference between conditions was greater in the older compared to the younger group for coherence among occipital channel groups in the delta band and in the theta band. In the high-beta band, the older group exhibited a significantly higher reward-based modulation in coherence between frontal and occipital groups (green arrows). b) HAROLD - to investigate hemispheric asymmetry, the difference in reward-based coherence…

• 

  FIGURE 6. Aging effects on the frontoparietal control network during reward processing. (A) Analyzing the procedure scheme: group differences in coherence between frontal and parietal channel groups were assessed for the reward vs. neutral cue contrast. (B) Results: Older adults exhibited lower reward-related frontoparietal left and frontal left-parietal right coherence modulation in the alpha band (red arrows). Asterisks: one asterisk: p ≤ 0.05.