Reward Decision-making Research Articles

Short-Term Memory Capacity Predicts Willingness to Expend Cognitive Effort for Reward.

We must often decide whether the effort required for a task is worth the reward. Past rodent work suggests that willingness to deploy cognitive effort can be driven by individual differences in perceived reward value, depression, or chronic stress. However, many factors driving cognitive effort deployment-such as short-term memory ability-cannot easily be captured in rodents. Furthermore, we do not fully understand how individual differences in short-term memory ability, depression, chronic stress, and reward anticipation impact cognitive effort deployment for reward. Here, we examined whether these factors predict cognitive effort deployment for higher reward in an online visual short-term memory task. Undergraduate participants were grouped into high and low effort groups (n HighEffort = 348, n LowEffort = 81; n Female = 332, n Male = 92, M Age = 20.37, Range Age = 16-42) based on decisions in this task. After completing a monetary incentive task to measure reward anticipation, participants completed short-term memory task trials where they could choose to encode either fewer (low effort/reward) or more (high effort/reward) squares before reporting whether or not the color of a target square matched the square previously in that location. We found that only greater short-term memory ability predicted whether participants chose a much higher proportion of high versus low effort trials. Drift diffusion modeling showed that high effort group participants were more biased than low effort group participants toward selecting high effort trials. Our findings highlight the role of individual differences in cognitive effort ability in explaining cognitive effort deployment choices.

Carnivore guild utilization of hunter‐provided food sources in boreal forest

Scavenging is an alternative foraging strategy to predation for many carnivore species, as they shift between predation and scavenging in response to changes in resource availability. The use of carrion may lead to interspecific competition and is thus influenced by a risk–reward trade‐off to balance coexistence with guild‐members, where smaller species are expected to be more vigilant due to their vulnerability to larger competitors. We used cameras to investigate the utilization of viscera from the annual moose Alces alces hunt by four sympatric carnivore species: wolf Canis lupus, wolverine Gulo gulo, red fox Vulpes vulpes and pine marten Martes martes in south–central Scandinavia, in relation to body size and habitat. Red foxes had highest probability of visiting viscera sites in both open and forested habitats. Visits by both red foxes and pine martens were longer in open habitats, while number of visits or activity did not differ between habitats. For pine martens, the probability of visiting viscera sites was twice as high in forest compared to open habitat; consequently, red foxes showed the highest overall use of viscera. Red foxes were most vigilant, especially in open habitat, whereas wolverines and pine martens spent a higher proportion of time feeding. Increased vigilance of red foxes facilitates extended resources use in open habitat, while for pine martens, the risk–reward decision occurs before leaving forest cover and entering open habitats. Viscera were not used to a large extent by wolves or wolverines. Wolves are generally less prone to scavenging and wolverines probably use other food resources, more suitable for caching. Overall, competition did not prevent use of viscera, probably due to small‐scale temporal segregation and limited use by the larger carnivores. Consequently, this pulse of human‐subsidized food resources before winter may have important implications for the smaller facultative scavengers.Keywords: carrion use, interspecific competition, pine marten, red fox, wolf, wolverine

Multi-dimensional resource allocation strategy for LEO satellite communication uplinks based on deep reinforcement learning

In the LEO satellite communication system, the resource utilization rate is very low due to the constrained resources on satellites and the non-uniform distribution of traffics. In addition, the rapid movement of LEO satellites leads to complicated and changeable networks, which makes it difficult for traditional resource allocation strategies to improve the resource utilization rate. To solve the above problem, this paper proposes a resource allocation strategy based on deep reinforcement learning. The strategy takes the weighted sum of spectral efficiency, energy efficiency and blocking rate as the optimization objective, and constructs a joint power and channel allocation model. The strategy allocates channels and power according to the number of channels, the number of users and the type of business. In the reward decision mechanism, the maximum reward is obtained by maximizing the increment of the optimization target. However, during the optimization process, the decision always focuses on the optimal allocation for current users, and ignores QoS for new users. To avoid the situation, current service beams are integrated with high- traffic beams, and states of beams are refactored to maximize long-term benefits to improve system performance.Simulation experiments show that in scenarios with a high number of users, the proposed resource allocation strategy reduces the blocking rate by at least 5% compared to reinforcement learning methods, effectively enhancing resource utilization.

Effects of individual and collective decision rule on cooperation in public goods game

Research on public goods game has traditionally focused on studying the effects of punishment and reward in experimental settings, commonly known as standard games. However, these standard games often examine isolated interactions among individuals, leading to varying results based on factors such as the population studied, number of participants, and geographical context. Consequently, the understanding of cooperation evolution remains imprecise.To address this limitation, we are trying to generalize these experimental studies, we employ an agent-based model in this study to simulate interactions in public goods game and explore the influence of punishment and reward on the evolution of cooperation. Unlike standard games, evolutionary games offer a more comprehensive framework to capture and analyze a broader range of strategic interactions and behaviors.In our approach, we introduce two methods for applying punishment or reward: Individual Punishment or Reward and Collective Punishment or Reward. Under the first method, actors can individually reward or punish each other based on their contributions or free-riding behavior towards the collective good. In the second method, decisions for punishment or reward are made collectively after achieving a consensus or agreement among the members of the group. We provide a model for comparing Individual Decision Rule and Collective Decision Rule under punishment and reward, to study their effects on the evolution of contributions.Consistent with recent experimental findings, our results indicate that these mechanisms lead to a significant increase in contributions compared to the outcomes observed in standard public goods game. Additionally, we demonstrate the significant role of rewards in promoting contribution within public goods game.

Strategic use of fare-reward schemes in a ride-sourcing market: An equilibrium analysis

Reward schemes are common in ride-sourcing markets. Each reward scheme incentivizes users with economic benefits as rewards for rides from the ride-sourcing platform. In practice, reward schemes differ in terms of their operational features. Operational features are the ways in which a ride-sourcing platform manages a reward scheme and users perceive and redeem their rewards. Therefore, different reward schemes involve different decision-making contexts for the platforms. As a result, it poses a challenge to understand the behavior of platforms with different types of reward schemes in the market. To address this challenge, this study analyzes the impacts of point-based and crypto-based fare-reward schemes on a ride-sourcing market. The market is defined as a mathematical model in which a ride-sourcing platform determines trip fares, commission fees, and rewards for passengers and drivers to maximize matching profits. A points-based reward scheme offers loyalty points with fixed redemption values. Issuance of point rewards incurs a liability for the reward scheme to accept users' point redemptions in the future. The crypto-based approach involves the cost of securing cryptocurrencies to reward users. Crypto rewards have various redemption values. Therefore, users perceive the value of crypto rewards based on their beliefs regarding their future value. This study performed an equilibrium analysis to elucidate the relationships between platform decisions and endogenous variables in the market using two fare-reward schemes. Subsequently, the properties of the optimal market solutions are further explored. These properties provide insights into the platform's reward and pricing decisions as follows: First, it is shown that a ride-sourcing platform is likely to adopt a single-sided reward strategy; that is, the platform's reward decisions with reward schemes are made exclusively between passengers and drivers. Second, the heterogeneity of user redemption behavior polarizes the platform's pricing decision on trip fares; that is, a ride-sourcing platform has incentives to set lower and higher trip fares when deciding to reward drivers and passengers, respectively. Third, such distinctive pricing decisions balance passenger demand with driver supply. Therefore, the ride-sourcing market can remain stable and efficient, even in the presence of the platform's reward decisions. A series of numerical experiments were performed to verify the findings of the current study.

Embodied decision biases: individually stable across different tasks?

In everyday life, action and decision-making often run in parallel. Action-based models argue that action and decision-making strongly interact and, more specifically, that action can bias decision-making. This embodied decision bias is thought to originate from changes in motor costs and/or cognitive crosstalk. Recent research confirmed embodied decision biases for different tasks including walking and manual movements. Yet, whether such biases generalize within individuals across different tasks remains to be determined. To test this, we used two different decision-making tasks that have independently been shown to reliably produce embodied decision biases. In a within-participant design, participants performed two tasks in a counterbalanced fashion: (i) a walking paradigm for which it is known that motor costs systematically influence reward decisions, and (ii) a manual movement task in which motor costs and cognitive crosstalk have been shown to impact reward decisions. In both tasks, we successfully replicated the predicted embodied decision biases. However, there was no evidence that the strength of the biases correlated between tasks. Hence, our findings do not confirm that embodied decision biases transfer between tasks. Future research is needed to examine whether this lack of transfer may be due to different causes underlying the impact of motor costs on decisions and the impact of cognitive crosstalk or task-specific differences.

On-demand ride-sourcing markets with cryptocurrency-based fare-reward scheme

Cryptocurrency-based fare-reward scheme has been emerging from on-demand ride-sourcing platforms. The reward scheme provides each passenger with some units of cryptocurrency as rewards for use of ride service. The passengers can redeem the rewards for various redemption options. The redemption options offer the passengers economic benefits to offset a portion of costs of using ride services. In this way, the economic benefits serve as incentives for the passengers to repeatedly use ride services, which leads to increase in passenger demand for the platforms. From a passenger perspective, the reward scheme has a distinctive feature in that economic benefits can fluctuate according to changes in price of cryptocurrency in the future. Therefore, the passengers perceive the benefits differently based on their own risk preference and expectation of future price changes. It creates a challenge for the platforms because the optimal decisions for ride services vary by the value that the passengers place for their cryptocurrency rewards. To tackle this challenge, this paper analyzes the impacts of the reward scheme on on-demand ride-sourcing markets. The analysis is done with a mathematical model where a ride-sourcing platform makes decisions for trip fare, vehicle fleet size and cryptocurrency reward size. The reward decision is available under the assumption that the platform secures as many units of cryptocurrency as it rewards to passengers at their equilibrium price level. The model defines passengers’ perceived value of cryptocurrency rewards in terms of risk aversion level and expectation of future price growth of cryptocurrency. Afterwards, we investigate equilibrium and optimal properties of monopoly, social-optimum and second-best optimum ride-sourcing markets. Our analysis gives several key insights on the markets. First, the markets always locate at normal regime where interactions between passenger demand and vehicle supply for ride services converges to a stable equilibrium state. Second, this study shows that a ride-sourcing platform would make a decision to reward if passengers expect larger economic benefits from future redemption of cryptocurrency rewards than their current value. Third, it is shown that the quality of ride services would get worse because average pick-up time for passengers and drivers gets longer with more passenger retention. Fourth, this paper finds that the markets have more room to improve profit or social welfare as passengers expect larger speculative gains from reward redemption.

Large-scale neural network computations and multivariate representations during approach-avoidance conflict decision-making

Many real-world situations require navigating decisions for both reward and threat. While there has been significant progress in understanding mechanisms of decision-making and mediating neurocircuitry separately for reward and threat, there is limited understanding of situations where reward and threat contingencies compete to create approach-avoidance conflict (AAC). Here, we leverage computational learning models, independent component analysis (ICA), and multivariate pattern analysis (MVPA) approaches to understand decision-making during a novel task that embeds concurrent reward and threat learning and manipulates congruency between reward and threat probabilities. Computational modeling supported a modified reinforcement learning model where participants integrated reward and threat value into a combined total value according to an individually varying policy parameter, which was highly predictive of decisions to approach reward vs avoid threat during trials where the highest reward option was also the highest threat option (i.e., approach-avoidance conflict). ICA analyses demonstrated unique roles for salience, frontoparietal, medial prefrontal, and inferior frontal networks in differential encoding of reward vs threat prediction error and value signals. The left frontoparietal network uniquely encoded degree of conflict between reward and threat value at the time of choice. MVPA demonstrated that delivery of reward and threat could accurately be decoded within salience and inferior frontal networks, respectively, and that decisions to approach reward vs avoid threat were predicted by the relative degree to which these reward vs threat representations were active at the time of choice. This latter result suggests that navigating AAC decisions involves generating mental representations for possible decision outcomes, and relative activation of these representations may bias subsequent decision-making towards approaching reward or avoiding threat accordingly.

Developmental risk sensitivity theory: the effects of socio-economic status on children's risky gain and loss decisions.

Evolutionary developmental theories propose that early environments shape human risk preferences. Developmental risk sensitivity theory (D-RST) focuses on the plasticity of risk preferences during childhood and makes predictions about the effect of reward size based on a child's social environment. By contrast, prospect theory predicts risk aversion for gains and risk seeking for losses regardless of environment or status. We presented 4 to 10-year-olds (n = 194) with a set of trials in which they chose between a certain amount and a chance to receive more or nothing. Two trials were equal expected value choices that differed by stake size and two were unequal expected value choices. Children either received gain trials or loss trials. Social environment was assessed using socio-economic status (SES) and subjective social status. Results confirmed the predictions of D-RST for gains based on SES. Children from lower-SES families differentiated between the high- and low-value trials and made more risky decisions for the high-value reward compared with higher-SES children. Children from higher-SES families were more risk averse for both trial types. Decisions for loss trials did not conform completely to either theory. We discuss the results in relation to evolutionary developmental theories.

Implementation Of Additive Ratio Assessment (Aras) Method For Online Reward Driver Provision

Giving rewards to online drivers is one of the annual agendas carried out at Grab Bengkulu Branch to encourage driver motivation and professionalism in improving the quality of work. The obstacles faced in giving these rewards are due to the diversity of educational backgrounds, experiences, competencies and portfolios of the drivers, so we need a system that can assist in the process of giving rewards to drivers so that they do not become sluggish and experience difficulties. In the process of designing the application of this reward decision support system using the ARAS approach. The ARAS method is a method based on the intuitive principle that alternatives must have the largest ratio to produce an optimal solution. The ARAS method performs ranking by comparing the value of each criterion on each alternative by looking at the weights of each to obtain the ideal alternative. The implementation of the system uses the Visual Basic 2010 programming language. From the results of the tests carried out, it can be concluded that the best employee with a value of 0.90, the lowest value of 0.81

Designing internal innovation contests

Innovation contests within firms offer the potential to tap latent employee expertise across the organization and expand innovation beyond the core R&D teams. Still, firms undertaking an internal contest, unlike the more commonly studied external contests, incur significant opportunity costs when employees divert their efforts toward the contest (and away from their routine work). Managing this key challenge forms the focus of the current study, with the overarching question being How should firms design internal contests? Motivated by fieldwork in electronics companies, we develop an application‐driven model that explicitly captures the fundamental trade‐off between incurring greater opportunity cost (of lost routine work) and increasing participation and contest effort. Analysis of our model demonstrates that for problem types that do not need significant effort for solution generation, the optimal contest reward is non‐monotonic in both the problem uncertainty and the wage rate of the participating employees. In case of more complex problems that require contestants to exert some minimal effort, the firm uses the reward decision to restrict participation and to ensure that contestants exert the required effort. In this case, the optimal reward increases with problem uncertainty, thereby incentivizing more contestants to enter. While the firm can attempt to mitigate the loss due to opportunity costs by organizing internal contests during employee leisure time, we identify conditions under which it is still optimal to organize these contests during working hours.

Optimal pricing and offering reward decisions in a competitive closed-loop dual-channel supply chain with recycling and remanufacturing

Recycling of materials has two significant perspectives: it may reduce the waste, and also, it can save raw materials. This study deals with the returned-obsolete products and the fresh items in a closed-loop dual-channel supply chain, where the manufacturer operates the whole production department and sells a percentage of products directly through his online channel, and delivers the rest of them to the retailer at a wholesale price. Additionally, the retailer collects unused items from customers with an appropriate reward to determine whether the customers intend to return the items. Both players screen the condition of the collected materials and then transfer the qualified materials for further use. Our paper formulates a mathematical model to evaluate scenarios such as scenarios manufacturer Stackelberg, Retailer Stackelberg, vertical Nash under the decentralized system, and a centralized system. The study is primarily concerned with finding optimal pricing plans and rewarding the customer analytically under various scenarios. Numerical explorations signify that the manufacturer Stackelberg’s scenario is more economical than the retailer Stackelberg and vertical Nash frameworks. The findings illustrate that the higher acceptance ratios of the returned materials benefit all the members and increase the keenness to return. Also, it is important for members to control the price-sensitive parameters within the demand function in order to save their markets. Further, the study suggests that an increase in production cost forces us to collect more returned materials, regardless of whether the increased remanufacturing cost suppresses that collection.

Attributes of excellent clinician teachers and barriers to recognizing and rewarding clinician teachers' performances and achievements: a narrative review.

BackgroundOver the last 31 years, there have been several institutional efforts to better recognize and reward clinician teachers. However, the perception of inadequate recognition and rewards by clinician teachers for their clinical teaching performance and achievements remains. The objective of this narrative review is two-fold: deepen understanding of the attributes of excellent clinician teachers considered for recognition and reward decisions and identify the barriers clinician teachers face in receiving recognition and rewards.MethodsWe searched OVID Medline, Embase, Education Source and Web of Science to identify relevant papers published between 1990 and 2020. After screening for eligibility, we conducted a content analysis of the findings from 43 relevant papers to identify key trends and issues in the literature.ResultsWe found the majority of relevant papers from the US context, a paucity of relevant papers from the Canadian context, and a declining international focus on the attributes of excellent clinician teachers and barriers to the recognition and rewarding of clinician teachers since 2010. ‘Provides feedback’, ‘excellent communication skills’, ‘good supervision’, and ‘organizational skills’ were common cognitive attributes considered for recognition and rewards. ‘Stimulates’, ‘passionate and enthusiastic’, and ‘creates supportive environment’, were common non-cognitive attributes considered for recognition and rewards. The devaluation of teaching, unclear criteria, and unreliable metrics were the main barriers to the recognition and rewarding of clinician teachers.ConclusionsThe findings of our narrative review highlight a need for local empirical research on recognition and reward issues to better inform local, context-specific reforms to policies and practices.

PLGRIM: Hierarchical Value Learning for Large-scale Exploration in Unknown Environments

In order for an autonomous robot to efficiently explore an unknown environment, it must account for uncertainty in sensor measurements, hazard assessment, localization, and motion execution. Making decisions for maximal reward in a stochastic setting requires value learning and policy construction over a belief space, i.e., probability distribution over all possible robot-world states. However, belief space planning in a large spatial environment over long temporal horizons suffers from severe computational challenges. Moreover, constructed policies must safely adapt to unexpected changes in the belief at runtime. This work proposes a scalable value learning framework, PLGRIM (Probabilistic Local and Global Reasoning on Information roadMaps), that bridges the gap between (i) local, risk-aware resiliency and (ii) global, reward-seeking mission objectives. Leveraging hierarchical belief space planners with information-rich graph structures, PLGRIM addresses large-scale exploration problems while providing locally near-optimal coverage plans. We validate our proposed framework with high-fidelity dynamic simulations in diverse environments and on physical robots in Martian-analog lava tubes.

Joint Content Sharing and Incentive Mechanism for Cache-Enabled Device-to-Device Networks

Cache-enabled device-to-device (C-D2D) networks allow the constituent user devices to share their cached content with other user devices through D2D communication. As a result, the communication delay of participating users is minimized in C-D2D networks. Clearly, the success of C-D2D networks relies on the willingness of the participating users to share their cached content. In this paper, we analyze the interaction among participating cache-enabled D2D users and determine their caching, sharing, and reward decisions to minimize the users' total cost. In view of the fact that the participating users have heterogeneous content interest profile (CIP) and storage capacity, and are self-centric, there is a need to design a fair incentive mechanism to encourage cooperation among users. To this end, we model the interaction among the D2D users as a multi-person bargaining game and design a novel incentive mechanism using the Nash bargaining solution (NBS) approach. The proposed incentive mechanism is capable of minimizing users' total costs while ensuring fair reward transfer among participating users. Further, we proposed a distributed algorithm which allows the execution of the proposed mechanism without any involvement of base station (BS), which is much needed for autonomous D2D networks. The simulation results demonstrate that the proposed mechanism improves fairness by at least 74% and reduces the users' total cost by at least 13.83% compared to the benchmark schemes.

Frontotemporal Regulation of Subjective Value to Suppress Impulsivity in Intertemporal Choices.

Impulsive decisions arise from preferring smaller but sooner rewards compared with larger but later rewards. How neural activity and attention to choice alternatives contribute to reward decisions during temporal discounting is not clear. Here we probed (1) attention to and (2) neural representation of delay and reward information in humans (both sexes) engaged in choices. We studied behavioral and frequency-specific dynamics supporting impulsive decisions on a fine-grained temporal scale using eye tracking and MEG recordings. In one condition, participants had to decide for themselves but pretended to decide for their best friend in a second prosocial condition, which required perspective taking. Hence, conditions varied in the value for themselves versus that pretending to choose for another person. Stronger impulsivity was reliably found across three independent groups for prosocial decisions. Eye tracking revealed a systematic shift of attention from the delay to the reward information and differences in eye tracking between conditions predicted differences in discounting. High-frequency activity (175-250 Hz) distributed over right frontotemporal sensors correlated with delay and reward information in consecutive temporal intervals for high value decisions for oneself but not the friend. Collectively, the results imply that the high-frequency activity recorded over frontotemporal MEG sensors plays a critical role in choice option integration.SIGNIFICANCE STATEMENT Humans face decisions between sooner smaller rewards and larger later rewards daily. An objective benefit of losing weight over a longer time might be devalued in face of ice cream because they prefer currently available options because of insufficiently considering long-term alternatives. The degree of contribution of neural representation and attention to choice alternatives is not clear. We investigated correlates of such decisions in participants deciding for themselves or pretending to choose for a friend. Behaviorally participants discounted less in self-choices compared with the prosocial condition. Eye movement and MEG recordings revealed how participants represent choice options most evident for options with high subjective value. These results advance our understanding of neural mechanisms underlying decision-making in humans.

Reward activity in ventral pallidum tracks satiety-sensitive preference and drives choice behavior.

A key function of the nervous system is producing adaptive behavior across changing conditions, like physiological state. Although states like thirst and hunger are known to impact decision-making, the neurobiology of this phenomenon has been studied minimally. Here, we tracked evolving preference for sucrose and water as rats proceeded from a thirsty to sated state. As rats shifted from water choices to sucrose choices across the session, the activity of a majority of neurons in the ventral pallidum, a region crucial for reward-related behaviors, closely matched the evolving behavioral preference. The timing of this signal followed the pattern of a reward prediction error, occurring at the cue or the reward depending on when reward identity was revealed. Additionally, optogenetic stimulation of ventral pallidum neurons at the time of reward was able to reverse behavioral preference. Our results suggest that ventral pallidum neurons guide reward-related decisions across changing physiological states.

Immunological Disturbances and Neuroimaging Findings in Major Depressive Disorder (MDD) and Alcohol Use Disorder (AUD) Comorbid Patients.

Mood disorders and Major Depressive Disorder, in particular, appear to be some of the most common psychiatric disorders with a high rate of comorbidity most frequently of anxiety or substance abuse disorders (alcohol use disorder). In both cases - MDD and AUD, a number of immunological disturbances are observed, such as chronic mild inflammation response, increased level of cytokines, hypercortisolaemia, which lead to specific changes in brain neurotransmitter functions. Some of the contemporary brain imaging techniques are functional magnetic resonance imaging (fMRI) and magnetic spectroscopy which are most commonly used to assess the brain metabolism and functional connectivity changes such as altered responses to emotional stimuli in MDD or overactivation of ventromedial prefrontal areas during delayed and underactivation of dorsolateral prefrontal regions during impulsive reward decisions in AUD and dysfunction of gamma-aminobutyric acid (GABA) and/or glutamate neurotransmitter systems, low NAA and myo-Inositol in both MDD and AUD.

Unique nature of appraisal politics as a work stress: test of stress–strain model from appraisee's perspective

PurposeThis paper delineates the distinctive nature of appraisal politics perceptions (referenced to organizational politics) experienced by appraisees (APAP) as a form of hindrance work stressor that is more episodic than chronic, salient during the PA rating and reward decisions. The study argues and attempts to establish empirically that due to its distinct nature, it causes both short-term episodic strain and long-term chronic strain. Further, the study investigates the distinctive role played by appraisee's hard and soft influence behaviour as a coping mechanism moderating the influence of APAP as a stressor on strain variables in Indian organizational context that ferments politics.Design/methodology/approachThe data was collected using self-reports from 407 employees in Indian organizations using survey method. Multivariate analyses including moderating tests were used for testing the hypotheses.FindingsOnly the episodic components of the APAP-appraiser's rating politics and pay and promotion politics were significantly related to anxiety felt by appraisees during PA – an episodic measure of strain. All three APAP components were significantly related to the chronic strain measure of dissatisfaction. There was modest support for the role of influence tactics (IT) as a coping mechanism attenuating the negative relation of APAP with the dissatisfaction variables as chronic strain measures. Contrary to the hypothesis, softer tactics exacerbated the APAP–PA anxiety relation, indicating the episodic nature of stressor and strain.Originality/valueThe study contributes significantly to enhance the understanding about the nature of Appraisal politics.

Computational Mechanisms of Effort and Reward Decisions in Patients With Depression and Their Association With Relapse After Antidepressant Discontinuation

Nearly 1 in 3 patients with major depressive disorder who respond to antidepressants relapse within 6 months of treatment discontinuation. No predictors of relapse exist to guide clinical decision-making in this scenario. To establish whether the decision to invest effort for rewards represents a persistent depression process after remission, predicts relapse after remission, and is affected by antidepressant discontinuation. This longitudinal randomized observational prognostic study in a Swiss and German university setting collected data from July 1, 2015, to January 31, 2019, from 66 healthy controls and 123 patients in remission from major depressive disorder in response to antidepressants prior to and after discontinuation. Study recruitment took place until January 2018. Discontinuation of antidepressants. Relapse during the 6 months after discontinuation. Choice and decision times on a task requiring participants to choose how much effort to exert for various amounts of reward and the mechanisms identified through parameters of a computational model. A total of 123 patients (mean [SD] age, 34.5 [11.2] years; 94 women [76%]) and 66 healthy controls (mean [SD] age, 34.6 [11.0] years; 49 women [74%]) were recruited. In the main subsample, mean (SD) decision times were slower for patients (n = 74) compared with controls (n = 34) (1.77 [0.38] seconds vs 1.61 [0.37] seconds; Cohen d = 0.52; P = .02), particularly for those who later relapsed after discontinuation of antidepressants (n = 21) compared with those who did not relapse (n = 39) (1.95 [0.40] seconds vs 1.67 [0.34] seconds; Cohen d = 0.77; P < .001). This slower decision time predicted relapse (accuracy = 0.66; P = .007). Patients invested less effort than healthy controls for rewards (F1,98 = 33.970; P < .001). Computational modeling identified a mean (SD) deviation from standard drift-diffusion models that was more prominent for patients than controls (patients, 0.67 [1.56]; controls, -0.71 [1.93]; Cohen d = 0.82; P < .001). Patients also showed higher mean (SD) effort sensitivity than controls (patients, 0.31 [0.92]; controls, -0.08 [1.03]; Cohen d = 0.51; P = .05). Relapsers differed from nonrelapsers in terms of the evidence required to make a decision for the low-effort choice (mean [SD]: relapsers, 1.36 [0.35]; nonrelapsers, 1.17 [0.26]; Cohen d = 0.65; P = .02). Group differences generally did not reach significance in the smaller replication sample (27 patients and 21 controls), but decision time prediction models from the main sample generalized to the replication sample (validation accuracy = 0.71; P = .03). This study found that the decision to invest effort was associated with prospective relapse risk after antidepressant discontinuation and may represent a persistent disease process in asymptomatic remitted major depressive disorder. Markers based on effort-related decision-making could potentially inform clinical decisions associated with antidepressant discontinuation.