Event-related potential paradigms for delay discounting in schizophrenia: Five paradigms and a methodological roadmap

Abstract

Delay discounting is a core behavioral measure of impulsivity that reflects an individual’s preference for an immediate small reward over a delayed larger reward; however, there is a gap in research on the underlying neural mechanisms of delay discounting abnormalities in individuals with schizophrenia. Therefore, this paper reviews the characteristics of delay discounting in individuals with schizophrenia, identifies five event-related potential paradigms suitable for studying delay discounting in this population, and proposes a feasible research roadmap. This paper lays the foundation for elucidating the mechanisms underlying impulsive decision-making in individuals with schizophrenia and provides a basis for clinical assessment.

Key Words: Schizophrenia; Delay discounting; Event-related potentials; Research roadmap; Methodological framework

Core Tip: Direct neurophysiological evidence on delay discounting in patients with schizophrenia is extremely scarce. This article identifies five event-related potential paradigms suitable for delay discounting studies, providing a critical methodological foundation for future research. We propose a practical roadmap including case-control studies using these paradigms, multicomponent event-related potential recording with time-frequency analysis, rigorous confounder control, longitudinal and cross-diagnostic designs, and integration with computational models. This framework aims to accelerate the understanding of the neural mechanisms underlying impulsive decision-making in patients with schizophrenia.

INTRODUCTION

Patients with schizophrenia often have broad cognitive deficits. Among these flaws, impaired decision-making is a key factor that contributes to social dysfunction, poor quality of life, and loss of work capacity. These individuals tend to prefer immediate gratification, a tendency referred to as delay discounting. The delay discounting task is concise, has transdiagnostic applicability, and has become a classic paradigm for studying impulsive decision-making. In a typical task of this kind, a participant must choose between a smaller reward available now and a larger reward available later. A higher discount rate indicates a stronger preference for immediate gratification, which reflects greater impulsivity[1]. Therefore, understanding the behavioral performance of individuals with schizophrenia in this task and its neural basis is crucial for revealing the nature of their decision-making deficits.

Delay discounting is a consistent behavioral manifestation of schizophrenia[2-5]. Heerey et al[2] were the first to report that individuals with schizophrenia exhibited steeper discounting curves on monetary choice tasks, and Ahn et al[6] reported the same conclusion after controlling for factors such as age and sex. Furthermore, a meta-analysis encompassing multiple mental disorders[1] confirmed that increased delay discounting is a cross-diagnostic phenomenon, with an effect size of approximately 0.46 for schizophrenia. Although behavioral evidence is robust, behavioral abnormalities alone do not directly reveal the underlying neural mechanisms involved. Event-related potential (ERP) techniques provide a unique tool for tracking neural activity across different stages of the decision-making process in real time. In addition, researchers have successfully used delay discounting tasks to elicit multiple ERP components with clear functional significance[7-10]. However, there are currently no systematic reports in the open literature that integrate original research involving individuals with schizophrenia, delay discounting tasks, and ERP recordings.

In April 2026, we conducted a literature search of PubMed and Web of Science using the keywords “schizophrenia”, “delay discounting”, and “event-related potentials”, as well as synonyms. Using single keywords and paired combinations, we identified 211 records; after deduplication, 127 unique records remained. Two researchers independently screened the titles, abstracts, and full texts based on the following inclusion and exclusion criteria (Table 1). The screening process proceeded in the following order: Removal of duplicate publications, exclusion of nonoriginal articles, exclusion of studies involving nonhuman subjects, exclusion of studies that did not use a delay discounting task, and exclusion of studies involving non-schizophrenic participants who did not employ ERP during the delay discounting task.

Table 1 Literature inclusion and exclusion criteria.

No.	Inclusion criteria	Exclusion criteria
1	Human participants	Non-human subjects (e.g., animal experiments)
2	Research involving delay discounting or delay discounting tasks	Absence of a delay discounting task
3	Use of ERP techniques	Studies in non-schizophrenia populations that did not employ ERP during a delay discounting task
4	Original empirical research	Non-original studies (e.g., reviews, commentaries, editorials, book chapters, conference abstracts)
5	Fulltext availability	Fulltext unavailable

ERP: Event-related potential.

Therefore, this paper first reviews the behavioral evidence of delay discounting in patients with schizophrenia, summarizes the ERP components and their functions elicited by delay discounting tasks in healthy individuals, and, on this basis, identifies five applicable ERP paradigms. Finally, we propose a methodological roadmap comprising five priority research directions and provide a comprehensive evaluation of the paper’s contributions and limitations in the discussion. We believe that this methodological framework may accelerate the elucidation of the neural mechanisms underlying impulsive decision-making in patients with schizophrenia and provide objective neurophysiological indicators for clinical assessment and intervention.

PATIENTS WITH SCHIZOPHRENIA EXHIBIT ELEVATED AND ROBUST DELAY DISCOUNTING RATES

Over the past fifteen years, multiple independent studies have consistently reported that patients with schizophrenia exhibit higher discount rates than healthy controls do. Using the Kirby Monetary Choice Questionnaire, Heerey et al[2] reported that among 42 patients with schizophrenia and 29 healthy controls, patients had significantly higher discount rates, and this difference was associated with patients’ working memory capacity. Ahn et al[6] replicated these findings in a sample of 21 patients with schizophrenia, 22 patients with bipolar disorder, and 30 healthy controls; even after controlling for working memory, intelligence, employment status, and substance use, the schizophrenia group (compared with both the bipolar disorder group and healthy controls) remained a significant predictor of discounting rates. Weller et al[11] reported that in 35 patients with schizophrenia and 21 healthy controls, the discount rates were significantly higher than those in the control group only in the subgroup of patients with consistent task performance, whereas inconsistent patients (approximately 40%) exhibited abnormal behavioral patterns, suggesting that some patients may have difficulty understanding or adhering to task requirements. Therefore, consistency in task performance must be considered when interpreting discount rate data in patient populations.

This phenomenon uses hierarchical Bayesian modeling. Yao et al[5] found that patients with schizophrenia not only exhibited higher discount rates but also demonstrated significantly greater differences in the perception of time between the near and distant future than healthy controls did, suggesting that patients may have distorted perceptions of time intervals. Bartolomeo et al[4] not only confirmed elevated discounting rates in chronic patients but also observed a similar trend in clinically high-risk adolescents. Although the between-group differences in the latter group did not reach statistical significance, correlations between discounting rates and negative symptoms were present in both groups. These findings further strengthen the evidence that delay discounting abnormalities constitute a stable behavioral phenotype in patients with schizophrenia.

However, a few studies have reported no significant differences between the two groups or even opposite results. In an functional magnetic resonance imaging study, Avsar et al[12] found that when only patients with consistent task performance were included (approximately 60%), there was no significant difference in discount rates between the patient and control groups. MacKillop and Tidey[13] reported no difference in discount rates between the schizophrenia group and the control group in a sample of nicotine-dependent smokers. Fan et al[14] reported that the discount rates of patients with schizophrenia were significantly lower than those of healthy controls, indicating that patients exhibited “excessive patience.” These inconsistencies may stem from the following factors.

First, sample characteristics are critical. Differences in patients’ disease duration, cognitive levels, symptom severity, and medication regimens have been reported across studies. For example, the sample in Fan et al[14] had low negative symptom scores (only 14.00 ± 1.00 for female patients), whereas in other studies, negative symptoms are often associated with high discount rates[2,4]. Second, different task paradigms yield different results. For example, individuals with schizophrenia exhibit elevated discount rates in experiential (real-wait) tasks, but this effect is not significant in hypothetical tasks[12]. Third, the statistical power is insufficient to detect moderate effect sizes, and confounding variables are often not adequately controlled for (e.g., working memory capacity, severity of clinical symptoms, nicotine use, and medication dosage).

ERP COMPONENTS ELICITED BY DELAY DISCOUNTING TASKS

Although there is a gap in ERP research on delay discounting in individuals with schizophrenia, there is a wealth of evidence from behavioral experimental studies on delay discounting in this population. More importantly, these behavioral studies have been validated by ERP research on delay discounting in other populations and have established a comprehensive system of neural markers ranging from early attention, conflict monitoring, and reward encoding to late evaluation. Therefore, we have summarized the ERP components elicited by delay discounting tasks and their implications (Table 2).

Table 2 Summary of major event-related potentials components in delay discounting tasks.

ERP	Time window	Typical electrode sites	Functional significance	Typical findings in delay discounting tasks
CNV	Cue-to-response (sustained slow potential)	Frontocentral (Fz, Cz)	Anticipatory preparation, motivational engagement	Larger CNV for delay-escape vs delay-inescapable conditions
P200	150-250 millisecond	Frontocentral (Fz, FCz)	Early attentional allocation	Greater amplitude in loss vs gain context; larger for delayed vs immediate options
N2	200-350 millisecond	Frontocentral (Fz, FCz)	Conflict monitoring, cognitive control	Increased amplitude under high conflict; negatively correlated with immediate choice rate
RewP/FRN	250-350 millisecond	Frontocentral (FCz)	Reward value encoding, outcome evaluation	Larger for immediate than delayed rewards; high discounters show larger RewP for immediate rewards
P300	300-500 millisecond	Parietal (Pz, P3, P4)	Attentional resource allocation, motivational salience, subjective value representation	Amplitude decreases with longer delays; larger for delayed than immediate choices
LPP	400-900 millisecond	Parietal (Pz, CPz)	Sustained emotional and motivational processing	Sensitive to probability; larger in loss context; larger for immediate vs delayed rewards
Theta	4-8 Hz (time-frequency)	Frontocentral	Decision conflict, cognitive control	Increased theta power after immediate choices in low discounters; higher theta during impulsive choices
Beta	13-30 Hz (time-frequency)	Frontal	Cognitive resource allocation, “cool” system processing	Lower beta power in the time-overestimation group; increased beta during non-impulsive choices

ERP: Event-related potential; CNV: Contingent negative variation; Fz: Frontal zero; Cz: Central zero; RewP: Reward positivity; FRN: Feedback-related negativity; FCz: Fronto-central zero; Pz: Parietal zero; CPz: Centro-parietal zero.

Contingent negative variation (CNV) reflects an individual’s state of readiness for an anticipated event and occurs between the cue stimulus and the response. A study using a delayed escape incentive task revealed that the amplitude of CNV increases significantly when a rapid response can prevent delay[15]. In a modified monetary incentive delay task, Rong et al[16] also observed that the anticipation of an immediate reward elicited a larger CNV amplitude than the anticipation of a delayed reward did. Furthermore, whereas CNV abnormalities have been widely reported in patients with schizophrenia, they have not yet been systematically studied in delay discounting tasks. CNV abnormalities have been consistently documented in patients with schizophrenia. A systematic review and meta-analysis by Osborne et al[17] which pooled data from 30 studies (685 patients and 714 healthy controls), reported robust CNV blunting in individuals with schizophrenia relative to healthy controls, with a large overall effect size (ES = -0.79). In particular, CNV reductions were more pronounced at the central site (ES = -0.87) than at the frontal site (ES = -0.48) and were comparable in magnitude to reaction time slowing. Extending these findings, a more recent study by Akgül et al[18] described early and late CNV in patients with schizophrenia using a modified monetary incentive delay task. These findings indicate that when patients anticipate receiving a reward, their CNV amplitude decreases, and the duration of their late CNV response to incentive cues is shorter than that of their early response. In other words, patients with schizophrenia exhibit deficits in the processing of motivational and reward-related information and display weaker CNV signals while waiting for a reward. Thus, CNV represents a promising neurophysiological marker for indexing both psychomotor and motivational dysfunctions in this clinical population. Therefore, incorporating the CNV into delay discounting paradigms may provide a new perspective on patients’ motivational attitudes toward delayed waiting.

The P200 is an indicator of early attentional allocation, typically peaking 150-250 millisecond after stimulus presentation. In delay discounting tasks, the P200 is more sensitive to time delays in loss-inducing conditions. Pei et al[19] found that the P200 amplitude induced by loss-inducing conditions was greater than that induced by gain-inducing conditions. Zhao et al[20] also reported that the delay option in loss conditions elicited a larger P200 than the immediate option did. Furthermore, individuals with a high propensity for delay exhibited larger P200 amplitudes, reflecting their abnormal attentional capture of temporal information. Additionally, high-delay individuals showed larger P2 amplitudes and longer latencies when processing temporal delay information, indicating greater sensitivity to delay duration[21]. P200 abnormalities have been consistently reported in individuals with schizophrenia. For instance, Şahin et al[22] reported attenuated P200 amplitudes and impaired sensory gating in patients with schizophrenia spectrum disorders (n = 30) compared with matched healthy controls (n = 31), with the attenuation being more pronounced for N100 and P200 than for P50. In addition, the patient group exhibited more pronounced temporal irregularity in the P200 component at the central zero (Cz) electrode, as reflected by changes in latency and reduced trial-to-trial consistency[23]. Taken together, the P200 may serve as a neurophysiological marker of early attentional and sensory processing deficits in patients with schizophrenia.

The N2 reflects an individual’s monitoring of conflict and typically peaks 200-350 millisecond after stimulus presentation. In delay discounting tasks, the N2 changes when participants face a conflict (e.g., an immediate vs a delayed option or a low-probability vs a high-probability option)[24]. Furthermore, stronger conflict monitoring helps inhibit impulsive responses; for example, in healthy individuals, a larger N2 amplitude is associated with fewer immediate choices. When individuals with schizophrenia experience impairments in their cognitive control and reasoning abilities, these deficits manifest as N2 abnormalities - specifically, a marked reduction in N2 amplitude and a longer latency[25]. For example, during conditional reasoning tasks, individuals with schizophrenia exhibit significantly smaller N2 amplitudes when they are processing abstract material[26]. Therefore, the N2 is an ideal indicator for examining whether conflict monitoring is impaired in intertemporal choice in patients with schizophrenia.

Reward positivity (RewP) (also known as feedback-related negativity) is a negative component observed in the middle frontal cortex 250-350 millisecond after reward presentation, reflecting the temporal sensitivity of reward processing. Cherniawsky and Holroyd[8] found that compared with low discounters, high discounters exhibited significantly greater RewP amplitudes for immediate rewards, whereas the RewP for future rewards was not significant in either group, suggesting that high discounters overvalue immediate rewards rather than undervalue future rewards. Schmidt et al[27] also reported that the difference in RewP between immediate and future rewards was significantly correlated with impulsivity as a personality trait, and this difference showed high test-retest reliability (r = 0.72) one week later. Guan et al[28] found that a large reward delayed by six months no longer elicited a magnitude effect in the RewP among individuals with anhedonia, suggesting that anhedonia stems from impaired neural representations of future rewards. Because anhedonia frequently co-occurs with schizophrenia, the RewP may serve as a powerful tool for exploring the neural mechanisms underlying its negative symptoms.

Furthermore, studies on reward-induced cortical electrical activity have shown that participants with schizophrenia or schizoaffective disorder exhibit increased delta wave activity upon winning and blunted theta wave activity upon losing, and that these patterns are associated with the severity of their psychiatric symptoms. These findings underscore the value of a dimensional framework for examining reward processing deficits in patients with schizophrenia. While the time-domain RewP showed a significant dimensional association with psychosis (especially when accounting for comorbid negative emotionality), the time-frequency measures - delta activity to wins and theta activity to losses - exhibited stronger and more consistent relationships across both lower-order detachment and higher-order general psychopathology dimensions. This pattern suggests that time-frequency neural responses to reward, rather than the RewP alone, may more accurately reflect the neural dysfunction associated with psychosis.

The P300 is one of the most extensively studied ERP components, typically peaking in the parietal region 300-500 millisecond after stimulus presentation, reflecting the allocation of attentional resources and motivational salience. In delay discounting tasks, the P300 amplitude decreases with increasing delay duration and is positively correlated with the discounting rate. Yu et al[29] found that food exhibits a steeper discounting curve than money does and that the amplitude changes in N2 and P3 differ between food tasks and money tasks. In overweight adults, Liu et al[9] reported that P3 amplitude was positively correlated with body mass index and negatively correlated with the area under the curve. In intertemporal choice tasks, the P300 amplitude elicited by selecting a delayed option is typically larger than that elicited by selecting an immediate option[10], reflecting a motivational evaluation of delayed rewards; furthermore, a stronger stimulus-preceding negative and a larger P3 are associated with a preference for delayed large rewards. P300 amplitude reduction is one of the most well-replicated electrophysiological findings in patients with schizophrenia. An early meta-analysis by Jeon and Polich[30] demonstrated significantly smaller P300 amplitudes and longer latency in patients than in healthy controls, with the largest effect sizes observed for auditory oddball paradigms; the paranoid subtype showed greater amplitude effects than the other subtypes did, whereas latency effect sizes decreased with disease duration. More recently, a large-scale network meta-analysis by Lu et al[25], encompassing 687 studies across multiple psychiatric disorders, confirmed significant P300 amplitude reductions and latency prolongations in patients with schizophrenia. The composite ERP score indicated that schizophrenia was associated with the most pronounced ERP abnormalities among the disorders examined, with P300 and P50 identified as the most prominently affected components. P300 abnormalities are present even before the onset of schizophrenia and can serve as neurophysiological biomarkers for the early detection, cognitive assessment, and risk stratification of schizophrenia spectrum disorders, manifesting as a consistently reduced amplitude and prolonged latency.

The LPP reflects ongoing emotional and motivational processing and typically peaks in the parietal regions 400-900 millisecond after stimulus presentation. In delay discounting tasks, the LPP is sensitive to a variety of factors and exhibits greater amplitudes in response to factors such as immediacy[31], probability[24], and loss[19]. In both patients with psychotic disorders and healthy controls, the LPP amplitude in response to pleasant and unpleasant stimuli was greater than that in response to neutral images; however, the overall LPP amplitude was lower in patients with schizophrenia, and the greater the deficit in social pleasure was, the smaller the LPP response to pleasant images[25]. Overall, the LPP is sensitive to the motivational value of specific emotional content and to the absence of social pleasure, making it useful for studying emotional processing dysfunction.

Time-frequency analysis reveals oscillatory activity that ERP components fail to capture. Stam et al[32] found that theta power significantly increased after low-discounters chose an immediate reward, whereas high-discounters did not, suggesting that theta reflects post decision conflict. Yi et al[33] reported that time over estimators exhibited more negative N2 and smaller P3 on the delayed option, along with lower beta power, indicating that time over estimators employ a “hot” system (emotion-driven), whereas time under estimators employ a “cold” system (cognitive control). In intracranial electroencephalography recordings, Gui et al[34] reported that theta power increased in the left lateral prefrontal cortex during impulsive choices, whereas beta power increased during self-controlled choices. Therefore, combining spatiotemporal analysis with ERP component analysis allows for a more comprehensive characterization of the neurodynamic features of individuals with schizophrenia during delay discounting tasks.

APPLICABLE DELAY DISCOUNTING ERP PARADIGMS

Based on the functional characteristics of the aforementioned ERP components and a systematic evaluation of the available literature, the selection of the five paradigms was guided by evidence at three levels: (1) Behavioral evidence of delay discounting abnormalities in schizophrenia; (2) Direct ERP evidence from schizophrenia populations - which is currently absent for all paradigms reviewed; and (3) ERP evidence from healthy populations or other psychiatric disorders that supports the transferability of these paradigms. Paradigm selection further adhered to the following criteria: (1) Clear task logic and moderate cognitive load, rendering them suitable for individuals with potentially impaired cognitive function; (2) The presence of well-established ERP component markers; and (3) Preliminary behavioral evidence supporting their applicability to patients with schizophrenia. Notably, although behavioral evidence for these paradigms in schizophrenia is relatively well documented, direct ERP evidence in this population remains entirely absent. The key characteristics of the five paradigms are summarized in Table 3.

Table 3 Summary of event-related potentials paradigms for delay discounting studies.

Paradigm	Key parameters	Primary ERP components	Remarks for schizophrenia applicability
Classical hypothetical delay discounting task	Binary choice between SS vs LL; delays ranging from 1 day to 6 months; parametric variation of reward amounts; Kirby titration or McClure orthogonal design	P300, RewP, CNV, N2, P2	Clear behavioral phenotype (elevated discounting) and well-documented P300 abnormalities; no direct ERP study yet; recommended as firstchoice paradigm
Delay discounting with risk	Simultaneous manipulation of delay (short/Long) and probability (high/Low); typically a 2 × 2 design	N1, P2, N2, N400, LPC, P300	Allows dissociation of risk and time discounting; suitable for examining probabilistic reasoning biases
Emotionprimed intertemporal choice	Presentation of emotional pictures/faces (e.g., disgust, anger, gratitude) before each choice; pre-rating of valence and arousal recommended (e.g., IAPS/CAPS)	N1, P2, P3, LPP, SPN, EPN	Useful for studying interaction between affective processing and decisionmaking; can explore negative symptom mechanisms (anhedonia)
Loss-framed intertemporal choice	Monetary losses instead of gains; symmetric design with gain frame to compare gain-loss asymmetry	P2, N2, P3	Reveals asymmetry of loss aversion; suitable for testing emotion regulation deficits
Social distance/social comparison intertemporal choice	Decisions for self, friend, or stranger; may include social comparison feedback	N1, P2, P3 (P3a/P3b), LPP	Explores the neural basis of social cognition deficits; recommended for patients with relatively preserved cognitive function

ERP: Event-related potential; SS: Smaller-sooner; LL: Larger-later; RewP: Reward positivity; CNV: Contingent negative variation; LPC: Late positive component; IAPS: International affective picture system; CAPS:Chinese affective picture system; LPP: Late positive potential; SPN: Stimulus preceding negativity; EPN: Early Posterior negativity.

Paradigm 1: The classic hypothesis delay discounting task

The classic hypothesis delay discounting task is currently the most widely used delay discounting task. In this task, participants must choose between an immediate small reward and a delayed large reward, and it elicits P300, RewP, CNV, N2, and P2 waveforms. More importantly, although individuals with schizophrenia exhibit distinct behavioral phenotypes in delay discounting tasks[2,5,6,11] and abnormalities in P300 components have been widely documented[30], neurophysiological studies integrating these two aspects remain scarce. To date, there are no systematic reports in the open literature that simultaneously integrate patients with schizophrenia, delay discounting tasks, and ERP recordings. Accordingly, evidence supporting this paradigm is currently drawn from behavioral studies in schizophrenia and ERP investigations in other populations, as direct ERP evidence in schizophrenia remains absent. At the behavioral level, this paradigm has been the most extensively validated in schizophrenia: 14 of the 17 identified studies employed a Kirby-type hypothetical delay discounting task, and most reported significantly higher discounting rates in patients than in healthy controls[2,4,6]. In other populations, abundant ERP evidence has established that P300 and N2 components are reliably elicited during delay discounting[7,9,10]. In contrast, no study to date has directly recorded ERPs from individuals with schizophrenia while they performed a delay discounting task. Based on this behavioral evidence and the established ERP correlates in other populations, the following methodological proposals are advanced: The Kirby titration method may be used to control task difficulty; electrode recordings are recommended at frontal zero, fronto Cz, Cz, and parietal zero - the standard sites where N2 and P300 are most robustly elicited and reported; and analyses should focus on the P300 (parietal) and N2 (frontocentral) components, given their well-established functional significance and documented abnormalities in schizophrenia[25,30]. It should be noted that these recommendations represent proposed guidelines informed by ERP evidence from other populations and behavioral evidence in schizophrenia, rather than evidence-based conclusions drawn from direct ERP studies in schizophrenia, which are currently lacking.

Paradigm 2: Risk-involving delay discounting tasks

To date, no study has directly recorded ERPs associated with risk-involving delay discounting tasks in patients with schizophrenia, but ERP evidence from healthy participants supports the feasibility of this paradigm. This paradigm manipulates both delay and probability to examine the dissociation between risk discounting and time discounting. In a typical design, one option offers an immediate and certain reward (e.g., 50 yuan), whereas the other offers a delayed and risky reward (e.g., an 80% probability of receiving 100 yuan and a 20% probability of receiving 0 yuan after one month). Delay (short/Long) and probability (high/Low) can form a 2 × 2 within-subjects design. With respect to ERP components, Wang et al[24] reported that P200 and N2 were sensitive only to probability (not to delay), P300 was sensitive to both delay and probability, and LPP was sensitive only to probability. This suggests that probability information dominates processing and exerts a more enduring influence. With respect to applicability to schizophrenia, patients often exhibit biases in probabilistic reasoning and abnormalities in risk decision-making; for example, Duva et al[35] found that patients with psychotic disorders and a history of cocaine use actually preferred delayed large rewards in choice tasks, suggesting that the interaction between risk and temporal discounting may be moderated by substance use. To date, no studies have directly recorded risk-related delay discounting ERPs in this population; with respect to task comprehension, Weller et al[11] demonstrated that approximately 60% of patients with schizophrenia exhibited consistent performance (R² > 0.60) on delay discounting tasks, indicating that the majority of these individuals are able to understand and complete such paradigms. Nevertheless, a potential data exclusion rate of up to 40% should be anticipated, and sample sizes should be planned accordingly. For trial numbers, a minimum of 40 trials per condition is recommended, based on standard ERP reliability considerations and signal-to-noise ratio requirements, to ensure stable ERP waveforms at the individual level and adequate statistical power for component analysis. Notably, this paradigm is proposed only as an exploratory auxiliary approach, as ERP evidence is currently absent in both schizophrenia and other populations.

Paradigm 3: Emotion-primed intertemporal choice

In the “emotion-primed intertemporal choice” experiment, an emotional stimulus (e.g., disgust, fear, sadness, or gratitude) was presented before each intertemporal choice, and participants were asked to either passively view it or provide a simple rating. Subsequently, smaller-sooner and larger-later options are presented. By comparing choice behavior and ERP components across different emotion-priming conditions, the influence of emotion on delay discounting can be examined. The emotional priming phase elicits ERP components related to emotional processing, such as EPN, P300, and LPP; changes in P2, N2, P3, and LPP are observed during the decision-making phase. For example, Suo et al[36] reported that compared with sadness or neutral priming, anger priming elicited larger P1, P2, and P3 waves and that individuals under anger conditions more strongly preferred delayed rewards. Individuals with schizophrenia commonly exhibit deficits in emotional processing and anhedonia. The use of an emotional priming paradigm can explore whether negative symptoms drive high discount rates by weakening emotional expectations of future rewards. Evidence supporting this paradigm is drawn from behavioral findings in schizophrenia and ERP evidence from other populations, as direct ERP evidence in schizophrenia during delay discounting is currently lacking. At the behavioral level, Bartolomeo et al[4] reported that delay discounting rates in patients with schizophrenia were positively correlated with Brief Negative Symptom Scale negative symptom scores (r = 0.36, P = 0.04), suggesting that emotional-motivational deficits are closely linked to temporal discounting behavior and providing a rationale for the introduction of emotion-priming manipulations. Studies in other populations have demonstrated that at the ERP level, emotion primes modulate P2 (early attention), P3 (motivational evaluation), and LPP (sustained emotional processing) during intertemporal choice. Specifically, Suo et al[36] found that compared with sadness or neutral priming, anger priming induced greater P1, P2, and P3 amplitudes and a stronger preference for delayed rewards; furthermore, during the processing of negative emotions, the P300 amplitude was positively correlated with the proportion of impulsive choices. In schizophrenia, although emotional processing deficits have been examined using ERP components such as the LPP[37,38], no study has yet integrated emotion priming with a delay discounting paradigm. Therefore, an emotion-priming paradigm, simplified for clinical populations, can be employed to examine the neural dynamics by which emotional valence modulates temporal discounting in schizophrenia, particularly in relation to anhedonia. However, it should be noted that this recommendation is derived from ERP findings in other populations and behavioral associations in patients rather than from direct ERP evidence in patients with schizophrenia. Although no direct studies currently exist, there is a behavioral research foundation regarding the influence of emotions on decision-making among individuals with schizophrenia. Therefore, this paradigm is recommended as an exploratory supplementary paradigm, with a focus on LPP components; emotional stimuli should be drawn from the International Affective Picture System, and valence and arousal ratings should be pre-assessed in the patient population.

Paradigm 4: Intertemporal choice in a loss-framing context

This protocol uses monetary losses rather than gains, typically employing a task structure symmetrical to the gain-framing condition to compare loss-gain asymmetry. A typical choice is “lose 5 yuan today” vs “lose 20 yuan in one month”. In ERP studies, Zhao et al[20] reported that the amplitudes of both the P200 and N2 evoked by loss conditions were greater than those evoked by gain conditions and that the P200 was delayed in the loss condition compared with the immediate option; however, there was no significant difference in the P300 between loss and gain conditions, but brain network analysis indicated that loss decisions require greater information integration. Pei et al[19] reported that the P2 amplitude was greater in loss scenarios than in gain scenarios. Patients with schizophrenia may exhibit abnormalities in loss aversion; because decision-making in loss scenarios is more susceptible to emotional influence, this paradigm may reveal deficits in emotional regulation among patients. Preliminary behavioral evidence already exists (Yao et al[5] found that patients exhibit an expanded discrepancy in their perception of time between the near and distant future), but systematic research within a loss framework has not yet been conducted. Evidence supporting this paradigm comes from behavioral evidence in schizophrenia and ERP evidence from other populations, although direct ERP evidence in schizophrenia is currently lacking. Behaviorally, some studies have reported that patients with schizophrenia paradoxically preferred delayed larger rewards in certain contexts, suggesting an abnormal valuation of gain vs loss outcomes that warrants systematic investigation within a loss framework[39,40]. At the ERP level in other populations, studies have consistently shown that loss contexts elicit larger P200 and N2 amplitudes than gain contexts do, accompanied by stronger negative emotions and greater cognitive conflict[20]. At the ERP level in patients with schizophrenia, no study has yet examined loss-framed delay discounting. We propose that a loss-framed delay discounting task could serve as an independent control condition, conducted in parallel with the gain framework, with a recommended focus on the P2 and N2 components. We acknowledge that this recommendation represents a proposed guideline rather than an evidence-based conclusion derived from patient studies, as direct ERP evidence in schizophrenia is currently absent.

Paradigm 5: Social distance/social comparison intertemporal choice

This paradigm requires participants to make intertemporal choices on behalf of themselves, a friend, or a stranger. In the social distance/social comparison intertemporal choice, participants were asked to decide, on behalf of a specific individual, whether to accept an immediate small reward or a delayed large reward. With respect to ERP components, Tang et al[41] reported that the friend condition elicited a larger P3a (200-300 millisecond) than the stranger condition did, reflecting that social distance is encoded early on; for P3b (300-400 millisecond), the delayed option elicited a larger P3b than the immediate option did in the friend condition, whereas no difference was observed in the stranger condition; furthermore, the N1 was sensitive to time delay. Patients with schizophrenia exhibit significant social cognitive deficits, including impairments in theory of mind, empathy, and social perception. The social distance paradigm can test whether patients exhibit discounting patterns different from those observed in self-decision-making when they make decisions for others. Behaviorally, Wang et al[39] found that patients exhibited “excessive patience” when they made self-decisions, but scenarios involving decisions for others have not yet been studied. Evidence supporting this paradigm is limited to ERP findings from other populations; behavioral and ERP evidence in patients with schizophrenia is currently lacking. Social cognitive deficits - including impairments in theory of mind, empathy, and social perception - are core features of schizophrenia, yet no behavioral study has directly manipulated social distance within a delay discounting paradigm in this population. At the ERP level, Tang et al[41] reported that the friend condition elicited a larger P3a than the stranger condition did and that an interaction effect between social distance and intertemporal choice was observed on P3b. In schizophrenia, although Catalano et al[42] found that patients exhibited blunted stimulus-preceding negative during social reward anticipation alongside relatively intact consummatory P300 responses - suggesting a “wanting” rather than “liking” motivational deficit - this was not assessed within a delay discounting framework. Therefore, the social distance/social comparison intertemporal choice task was proposed as an advanced exploratory method. Furthermore, given the significant differences in social cognitive impairments among individuals with schizophrenia, prescreening of participants’ basic social cognitive abilities and simplifying social scenarios (e.g., distinguishing only between “self” and “stranger”) are recommended.

METHODOLOGICAL PATHWAYS FOR DELAY DISCOUNTING ERP RESEARCH IN PATIENTS WITH SCHIZOPHRENIA

Based on the existing evidence regarding delay discounting in individuals with schizophrenia, we propose the following five priority research directions for ERP studies on delay discounting in this population, as well as a methodological roadmap for such research that follows an intrinsic logical progression (Figure 1). First, researchers should conduct cross-sectional case-control studies on delay discounting in individuals with schizophrenia to obtain reliable ERP data. Next, they should validate the reliability of these ERP findings by strictly controlling for confounding variables (such as sex, age, working memory capacity, nicotine use, and medication dosage). Longitudinal studies on delay discounting ERP measures in individuals with schizophrenia are needed to verify the clinical value of these measures. Next, researchers should validate the ecological validity of the ERP findings on delay discounting in patients with schizophrenia by extending the study to cross-diagnostic comparisons to test the specificity of these delay discounting characteristics. Finally, researchers should use computational models to elicit a mechanistic explanation for the delay discounting characteristics observed in patients with schizophrenia.

Figure 1 Methodological roadmap for event-related potential studies of delay discounting in schizophrenia. This schematic diagram illustrates a proposed five-step roadmap for event-related potential studies of delay discounting in schizophrenia, progressing from case-control studies to confounder control, longitudinal studies, cross-diagnostic comparisons, and computational modeling, with the ultimate goal of clinical translation. DD: Delay discounting; ERP: Event-related potential; LPP: Late positive potential; FTND: Fagerström test for nicotine dependence; CPZ: Chlorpromazine; rTMS: Repetitive Transcranial magnetic stimulation; SCZ: Schizophrenia; BD: Bipolar disorder; SUD: Substance use disorder; AN: Anorexia nervosa; DDM: Drift-diffusion model.

Direction 1: Conducting case-control ERP studies of delay discounting in patients with schizophrenia

The primary task is for researchers to conduct a large-sample, rigorously designed case-control cross-sectional study using one or more of the delay discounting research paradigms described above to identify the characteristics of individuals with schizophrenia in various delay discounting decision-making tasks. Key considerations for the study design examining delay discounting characteristics in individuals with schizophrenia include the following: Given the inconsistent performance on delay discounting tasks and data exclusion within the schizophrenia group (Weller), as well as the need to ensure statistical power, a sample size of at least 50 participants per group is recommended, such as when investigating differences in P300 amplitude between patients with schizophrenia and control subjects in a classic delayed discounting ERP task.

Direction 2: Strictly controlling key confounding variables

Existing behavioral studies of the delay discounting task in patients with schizophrenia have shown that factors such as working memory capacity, medication dosage, and smoking status influence delay discounting performance in these patients; however, these key confounding factors are often not strictly controlled for. Future research on delay discounting in individuals with schizophrenia must strictly control for the following confounding factors: (1) Working memory capacity (which can be assessed using tools such as the Hopkins Word Learning Test-Revised[43]); (2) Smoking status (recorded using standardized nicotine dependence scales, such as the Fagerström Nicotine Dependence Scale[44]); and (3) Clinical characteristics, including medical history, severity of clinical symptoms, and medication dosage. In summary, this phase aims to verify, after strictly controlling for the above key confounding variables and demographic factors, whether the between-group ERP differences between patients with schizophrenia and healthy controls can be reliably attributed to schizophrenia itself.

Direction 3: Conducting longitudinal and intervention studies

Researchers should investigate whether delay discounting ERP measures in patients with schizophrenia respond to treatment with antipsychotic medications, cognitive training, or transcranial magnetic stimulation interventions to assess whether these ERP components have the potential to serve as state markers for patients with schizophrenia. For example, after 8 weeks of cognitive training, do patients with schizophrenia show a significant increase in P300 amplitude when performing delay discounting tasks?

Direction 4: Advancing cross-diagnostic comparative studies

Increased delay discounting has been proposed as a cross-diagnostic process (Amlung et al[1], 2019). By comparing the characteristics of delay discounting in patients with schizophrenia and those with other mental disorders (e.g., bipolar disorder[6], substance use disorders, and anorexia nervosa[45] within the same delay discounting ERP experimental framework, we aim to elucidate the unique and shared mechanisms underlying reward processing and impulse control across different mental disorders. For example, are there differences in P300 amplitude between individuals with schizophrenia and those with bipolar disorder when they perform delay discounting ERP tasks?

Direction 5: Integrating ERP data with computational models

Researchers can use computational models (such as the drift-diffusion model) to integrate choice reaction times and ERP components from delay discounting tasks to better infer abnormalities in evidence accumulation and response thresholds during the decision-making process. Furthermore, the drift-diffusion model has been successfully applied to delay discounting behavior data from individuals with schizophrenia, revealing that their time sensitivity parameters are greater than those of healthy controls[5]. Therefore, future ERP studies on delay discounting characteristics in patients with schizophrenia could further explore whether ERP components can serve as prior distributions or validation metrics for model parameters. For example, the drift-diffusion model could be used to validate the P300 amplitude as a neural indicator of the drift rate and the N2 amplitude as an indicator of conflict monitoring.

CLINICAL IMPLICATIONS AND METHODOLOGICAL PROSPECTS

The author of this paper reviews the characteristics of delay discounting in individuals with schizophrenia and then summarizes the ERP components elicited by delay discounting tasks in healthy individuals and their significance. Based on existing behavioral evidence from delay discounting tasks in individuals with schizophrenia and ERP evidence from delay discounting tasks in other populations, the author recommends five ERP paradigms for delay discounting and provides a detailed description of each paradigm’s task structure, evoked potential components, and applicability to individuals with schizophrenia. Finally, the author proposes a prospective methodological roadmap for ERP research on delay discounting in schizophrenia, including five research directions at varying levels of depth, to investigate the ERP characteristics of delay discounting in this population. In summary, we aim to address the critical bottleneck of a severe lack of direct evidence in the field of ERP research on delay discounting in individuals with schizophrenia and to provide researchers with clear guidelines for designing ERP experiments to investigate these characteristics.

Apart from methodological considerations, existing evidence from ERP studies on schizophrenia suggests that the ERP components induced by delay discounting identified in this study reflect clinically significant deficits in decision-making ability among individuals with schizophrenia. First, evidence from both cross-sectional and longitudinal studies indicates that auditory P300 amplitude is significantly correlated with social functioning in patients with schizophrenia, as well as with impairments in daily living caused by negative symptoms. Furthermore, the P300 amplitude at the initial examination can predict clinical outcomes in terms of social functioning after an average follow-up of 2.4 years, suggesting that the P300 may serve as a prognostic biomarker for real-world functional outcomes[46]. Furthermore, although the N2 component is robustly observed in healthy controls, the trial-by-trial modulation of the N2 component is absent in patients with schizophrenia and appears only when these patients perform conflict resolution tasks[47]. This finding suggests that the N2 reflects fundamental disruptions in the neural circuits underlying cognitive control, disruptions that are likely to manifest in everyday decision-making situations requiring conflict resolution.

Patients with schizophrenia exhibit reduced responses in RewP and its associated time-frequency measures during reward processing[48], and they display behavioral inconsistencies when processing certain types of feedback[49]. These findings suggest that schizophrenia involves a neural deficit in distinguishing between contexts, which impacts real-world decision-making across different levels of predictability. Individuals with schizophrenia exhibit higher scores on measures of anhedonia and lower LPP amplitudes in response to pleasant stimuli, and there is a significant association between these two factors[37], suggesting that the LPP may be linked to the negative symptoms of schizophrenia. Taken together, the ERP components discussed by the authors are not only laboratory phenomena but also neurophysiological indicators capable of capturing clinically relevant dimensions of schizophrenia, such as negative symptoms, social dysfunction, and decision-making impairments in the real world[37,48,49].

This study has the following limitations. First, there is an extreme scarcity of direct evidence from ERP studies on delay discounting in patients with schizophrenia. No original studies in the open literature have simultaneously integrated patients with schizophrenia, delay discounting tasks, and ERP recordings; therefore, the discussion of neural mechanisms in this paper is primarily based on behavioral evidence of delay discounting in individuals with schizophrenia and indirect inferences drawn from healthy individuals or those with other mental disorders. Second, there is insufficient control for key confounding factors affecting the results of delay discounting ERP studies. Because most studies have not adequately reported or strictly controlled for factors such as working memory capacity, smoking status, medication use, and comorbid substance abuse, it is not possible to conduct a quantitative comparison of the entire body of literature. Third, the ecological validity of the paradigms remains to be verified. The acceptability of the five delay discounting paradigms recommended in this paper among the schizophrenia population has not yet been systematically assessed.

In summary, ERP studies on the delay discounting profile in individuals with schizophrenia not only clarify this profile and its underlying neural mechanisms but also transform physiological markers into biomarkers with clinical utility. These biomarkers hold promise for assisting in the assessment of symptoms and motivational deficits in individuals with schizophrenia, monitoring treatment response, and evaluating the effectiveness of targeted interventions.

CONCLUSION

ERP research on the delay discounting characteristics of individuals with schizophrenia is a highly promising area of study. The authors propose an experimental approach and paradigm to fill the gap in ERP research on delay discounting in individuals with schizophrenia and provide researchers with clear guidelines for designing ERP experiments to investigate these characteristics.

ACKNOWLEDGEMENTS

The authors thank The Affiliated Mental Health Center of Jiangnan University for institutional support.