Prolonged exertion of self-control causes increased sleep-like frontal brain activity and changes in aggressivity and punishment

The study investigates whether prolonged exertion of self-control (a central component of executive functions) induces local sleep-like brain activity in frontal regions and alters socially relevant behavior, particularly increasing aggressive choices. Ego depletion paradigms have aimed to manipulate the balance between intuition and deliberation, but findings have been inconsistent, possibly due to short task durations. Building on evidence that extended task practice and sleep loss can elicit local, sleep-like slow waves in task-related areas, the authors hypothesize that longer, cognitively demanding self-control tasks will produce frontal sleep-like activity and measurable changes in economic game behavior (e.g., increased aggression and altered punishment). The purpose is to clarify the neural basis of ego depletion effects and their impact on prosocial versus aggressive decision-making.

Prior work has linked intuition and deliberation to cooperative or antisocial choices, often via economic games. Ego depletion—resource reduction after self-control exertion—has been widely used to study social behavior, but large-scale replications have produced mixed evidence. Some analyses suggest certain manipulations are more effective than others and that the effect may lie on a spectrum dependent on manipulation intensity/duration. Separately, EEG studies show extended task practice and sleep restriction/deprivation produce local slow waves (delta/theta) during wakefulness, especially in frontal executive areas, associated with performance lapses. Animal and human data indicate these slow waves reflect neuronal off-periods akin to NREM sleep and correspond to cognitive fatigue. This literature motivates the hypothesis that ego depletion effects may reflect task-dependent neural fatigue expressed as local sleep-like activity in frontal regions, potentially explaining variability in prior findings.

Design: Preregistered two-study design with matched experimental groups: Frontal Fatigue (FF) vs. No Fatigue (NF). Study 1 collected both behavioral and high-density EEG data; Study 2 replicated behavioral findings in a larger sample. Participants: Total N=447; Study 1 N=44 (22 females; mean age 30.3, SD 7.8); Study 2 N=403 (251 females; mean age 23.3, SD 3.4). Prescreened for medical/neurological/psychiatric conditions. Ethical approvals obtained (Study 1: No. 1485/2017; Study 2: No. 03/2022). Informed consent obtained. Procedures (Study 1 timeline): - Baseline (T0): hd-EEG (64 electrodes), resting-state EEG, brief questionnaires, Go/NoGo task (~20 min). - Fatigue manipulation (~45 min total): Three 15-min tasks. FF group performed tasks requiring self-control (emotion suppression; false response; Stroop). NF group performed partially different versions not requiring self-regulation and relying on more automatic processes. - Post-manipulation (T1): Repeat resting-state EEG and tests similar to T0. - Behavioral assessment: One-shot, anonymous, incentivized economic games (~15 min): Dictator Game (DG), Ultimatum Game (UG; proposer offer and minimum acceptance threshold), Public Goods Game (PGG) without punishment, PGG with punishment (including modality classification: prosocial, antisocial, spiteful vs. second-order free riding), Hawk and Dove (HD) game. - Final EEG (T2): Resting-state EEG. - Follow-up: Next-day online self-report questionnaires (personality traits, empathy, aggressivity, impulsiveness). Measures and analyses: - EEG: Power spectral density changes from T0 to T1 (and T2), focusing on delta (1–4 Hz) and theta (4–8 Hz). Topography and source localization; cluster-based corrections. - Behavior: Nonparametric tests (Pearson’s chi-squared), probit regressions, multinomial logit for punishment modalities; diff-in-diff for Go/NoGo accuracy/RT; belief measures. Individual-level clustered SEs for regressions linking EEG changes to behavior.

Neural activity: - FF vs. NF showed a significantly stronger increase in delta (1–4 Hz) power from T0 to T1 in a frontal electrode cluster (cluster-based corrected P<0.05; Fig. 1B). - Source localization indicated delta power differences peaking in left inferior frontal gyrus (IFG; MNI −36, 16, 28) and right middle/superior temporal gyri (P<0.01, cluster-based correction; Fig. 2). - Theta (4–8 Hz) changes from T0 to T1 did not differ significantly between groups; however, after economic games (T2), FF showed stronger theta increases in temporoparietal/occipital electrodes (corrected P<0.05), with source peak in right temporal lobe (P<0.01). Behavior (Study 1, N=44): - Hawk and Dove: Dovish (peaceful) choices dropped from 86% (NF) to 41% (FF); P=0.0017 (Pearson’s chi-squared). Probit confirmed (P=0.0020). - Delta-power change (T0→T1) in frontal electrodes positively associated with likelihood of aggressive (hawkish) play; P=0.0408 (SE clustered at individual level). - No other significant differences across DG, UG, PGG tasks; results robust to multiple-testing considerations; no group differences in beliefs; diff-in-diff found no differences in Go/NoGo accuracy/RT. Behavior (Study 2, N=403): - Hawk and Dove: Significant differences between NF and FF (P<0.001, Pearson’s chi-squared); probit confirms (P<0.001). - PGG punishment modalities: Significant distribution differences (P<0.001). Pairwise tests show reduced prosocial punishment (P<0.001) and increased spiteful punishment (P=0.0019) in FF; total punishment unchanged (P=0.2961). Results confirmed by probit (P<0.001 and P=0.0022) and multinomial logit. - No other significant differences in remaining games; diff-in-diff regressions showed no differences in Go/NoGo accuracy/RT. Self-reports: No significant condition or condition×time effects for sleepiness, fatigue, mood, affects; only a non-corrected trend for motivation.

Extended exertion of self-control increased aggressive choices in social conflict scenarios (Hawk and Dove) and shifted punishment from prosocial to spiteful without changing overall punishment propensity. Concurrently, frontal sleep-like delta activity rose in regions implicated in executive control (e.g., left IFG), supporting a mechanism whereby local neural fatigue reduces top-down regulation and fosters impulsive, aggressive responses. These findings align with the conceptualization of ego depletion as a spectrum dependent on manipulation intensity/duration and suggest that variability in prior studies may stem from insufficient task lengths to induce detectable local sleep-like fatigue. The link between frontal sleep-like episodes and behavior provides a physiological account for diminished prosocial behaviors under cognitive fatigue, though direct causality between sleep-like activity and specific decisions remains to be established.

Prolonged self-control practice (~45 min) leads to local sleep-like changes in frontal cortical activity and increases aggressive choices in economic conflict scenarios, with altered punishment preferences favoring spiteful actions over prosocial ones. These results help reconcile inconsistencies in ego depletion research by emphasizing manipulation duration/efficacy and point to local sleep-like neural fatigue as a plausible physiological basis. Future research should establish causal links between sleep-like episodes and decision-making, explore effects in repeated interactions and varied game structures, and characterize individual/population differences in vulnerability to frontal local sleep and cognitive fatigue.

- Task specificity: Although FF and NF tasks were matched in structure, the study did not include a control condition with comparable cognitive effort but different functions, so a contribution of general (non-task-specific) mental fatigue cannot be fully excluded. - Self-reports: No significant differences in sleepiness/fatigue/mood/affects by condition or over time (only an uncorrected trend for motivation), making nonspecific fatigue explanations less likely but not definitively ruled out. - Behavioral scope: Clear significant effects emerged primarily in the Hawk and Dove game; other games (DG, UG) showed no significant changes, partially contrasting previous findings. Game complexity and decision processes may modulate susceptibility to fatigue. - One-shot design: The protocol used one-shot games and a longer depletion manipulation; effects in repeated interactions and potential compensatory mechanisms remain untested. - Go/NoGo performance: No differences observed; absence of a practice/learning period may have allowed learning and fatigue to counterbalance, masking performance changes. - Causality: The study design does not prove a direct causal link between sleep-like activity and specific decisions; experimental causation needs future testing.