Research Articles, Behavioral/Cognitive

Dopaminergic Modulation of Dynamic Emotion Perception

B. A. Schuster, S. Sowden, A. J. Rybicki, D. S. Fraser, C. Press, P. Holland and J. L. Cook
Journal of Neuroscience 25 May 2022, 42 (21) 4394-4400; DOI: https://doi.org/10.1523/JNEUROSCI.2364-21.2022

Abstract

Emotion recognition abilities are fundamental to our everyday social interaction. A large number of clinical populations show impairments in this domain, with emotion recognition atypicalities being particularly prevalent among disorders exhibiting a dopamine system disruption (e.g., Parkinson's disease). Although this suggests a role for dopamine in emotion recognition, studies employing dopamine manipulation in healthy volunteers have exhibited mixed neural findings and no behavioral modulation. Interestingly, while a dependence of dopaminergic drug effects on individual baseline dopamine function has been well established in other cognitive domains, the emotion recognition literature so far has failed to account for these possible interindividual differences. The present within-subjects study therefore tested the effects of the dopamine D2 antagonist haloperidol on emotion recognition from dynamic, whole-body stimuli while accounting for interindividual differences in baseline dopamine. A total of 33 healthy male and female adults rated emotional point-light walkers (PLWs) once after ingestion of 2.5 mg haloperidol and once after placebo. To evaluate potential mechanistic pathways of the dopaminergic modulation of emotion recognition, participants also performed motoric and counting-based indices of temporal processing. Confirming our hypotheses, effects of haloperidol on emotion recognition depended on baseline dopamine function, where individuals with low baseline dopamine showed enhanced, and those with high baseline dopamine decreased emotion recognition. Drug effects on emotion recognition were related to drug effects on movement-based and explicit timing mechanisms, indicating possible mediating effects of temporal processing. Results highlight the need for future studies to account for baseline dopamine and suggest putative mechanisms underlying the dopaminergic modulation of emotion recognition.

SIGNIFICANCE STATEMENT A high prevalence of emotion recognition difficulties among clinical conditions where the dopamine system is affected suggests an involvement of dopamine in emotion recognition processes. However, previous psychopharmacological studies seeking to confirm this role in healthy volunteers thus far have failed to establish whether dopamine affects emotion recognition and lack mechanistic insights. The present study uncovered effects of dopamine on emotion recognition in healthy individuals by controlling for interindividual differences in baseline dopamine function and investigated potential mechanistic pathways via which dopamine may modulate emotion recognition. Our findings suggest that dopamine may influence emotion recognition via its effects on temporal processing, providing new directions for future research on typical and atypical emotion recognition.

Introduction

The ability to recognize others' emotions from facial and bodily cues is an important skill that facilitates the development of meaningful social relationships (Izard et al., 2001; Wang et al., 2019). However, the cues toward genuine expressions of emotions are often subtle. When we are sad, we do not pull a face that bears much resemblance to the Ekman example, but rather subtly alter the particular spatiotemporal dynamics of the way we move our body and face (Michalak et al., 2009; Roether et al., 2009; Sowden et al., 2021). Therefore, it is perhaps unsurprising that difficulties labeling emotions are widespread throughout a wide range of clinical conditions, but most notably, within those featuring a disruption of the dopaminergic neurotransmitter system (e.g., Parkinson's disease, Argaud et al., 2018; Huntington's disease, Henley et al., 2012; schizophrenia, Edwards et al., 2002). These observations warrant a closer examination of the role played by dopamine in socio-cognitive skills such as emotion recognition.

An incisive way to establish a causal role is to observe the influence of dopaminergic drugs on emotion recognition in the healthy population. Interestingly, while studies have found a range of mixed influences on neural responses (e.g., amygdala activation; Hariri et al., 2002; Takahashi et al., 2005; Delaveau et al., 2007) during emotion processing, they have typically not found behavioral influences. One explanation for this mixed picture is that there is an optimal level of dopamine for such tasks, and that, dependent on one's baseline levels, dopaminergic modulation brings individuals closer to, or further away from, that optimum (Delaveau et al., 2005, 2007). This theory has received widespread attention in other domains of cognition (Kimberg et al., 1997; Mattay et al., 2000; Gibbs and D'Esposito, 2005; Frank and O'Reilly, 2006; Cools and D'Esposito, 2011). Direct examinations of dopamine synthesis capacity and/or receptor binding are possible via positron emission tomography (PET) but are expensive and difficult to implement. Consequently, a large number of cognitive studies approximate striatal dopamine synthesis capacity via measures of working memory (WM) span, which are found to comprise a good proxy (Cools et al., 2008; Landau et al., 2009). Such studies find that individuals with low WM span, signaling low dopamine synthesis capacity, exhibit different behavioral responses to a dopaminergic modulation to those with higher span (capacity).

It is also notable that while a role for dopamine in emotion perception is suggested by studies illustrating aberrations in clinical conditions with dopamine dysfunctions, we do not have good mechanistic models of the nature of that role. Some plausible contenders could relate to the influence of dopamine on temporal and motor encoding. Many of the cues signaling emotional state are dynamic, and will therefore depend on one's ability to encode temporal features. For instance, whereas rapid, accelerated movements are associated with anger, slower and sluggish movements tend to be interpreted as sadness (Gross et al., 2012; Edey et al., 2017; Halovic and Kroos 2018). It has also been hypothesized, outside of the dopamine literature, that recognition of such temporal features relies on yoking to one's own movements and the emotional state experienced when performing such movements (Edey et al., 2017, 2020). Given the strong link between dopamine and temporal encoding (Coull et al., 2012; Tomassini et al., 2016), as well as motor performance (Niv et al., 2007; Tomassini et al., 2016), it is plausible that dopamine mediates emotion recognition via its influence on temporal processing.

To examine the role played by dopamine in emotion recognition, this study presented participants with a dynamic whole-body emotion recognition task under the dopamine D2 receptor antagonist haloperidol, and a placebo. We separated our analyses according to baseline WM span and examined whether influences of the drug were modulated by performance in movement-based and counting-based indices of temporal processing.

Materials and Methods

Participants

Forty-three healthy volunteers (19 females; mean (M) [SD] age = 26.36 [6.3]) took part on at least one of two study days after passing an initial health screening. Participants were recruited via convenience sampling from University of Birmingham campus and Birmingham city center, gave written informed consent and received either money (£10/h) or course credit for participation. Five participants (two placebo, three haloperidol) dropped out of the study after completing the first day, a further five could not complete the second test day because of COVID-19-related closures, consequently all analyses are based on 33 full datasets. Because of the novelty of the present design, no prior study exists to provide the basis for a meaningful a priori sample size estimation. All experimental procedures were approved by the University of Birmingham Research Ethics Committee (ERN 18–1588) and performed in accordance with the World Medical Association Declaration of Helsinki (1975).

Experimental design and statistical analyses

Pharmacological manipulation and general procedure

Participants' eligibility for the study was evaluated by a clinician via review of their medical history, electrocardiogram assessment and blood-pressure check. The main study took place on two separate test days, one to four weeks apart, where participants first completed an initial blood-pressure and blood-oxygenation check with the medic. Subsequently, in a double-blind, placebo controlled within-subjects design, each participant took part on two study days, wherein all participants received tablets containing either 2.5mg haloperidol or lactose (placebo) on the first day, and the respective other treatment on the second day (order of drug day counterbalanced). For this, participants were handed a preprepared envelope, instructed that this would contain either placebo or haloperidol tablets, informed that none of the experimenters knew the contents of the envelope, and asked to close their eyes before swallowing the tablets. Haloperidol is a dopamine D2 receptor antagonist, which affects dopamine transmission via binding either to postsynaptic D2 receptors (blocking the effects of phasic dopamine bursts), or to presynaptic autoreceptors (which has downstream effects on the release and reuptake of dopamine and thus modulates bursting itself; Benoit-Marand et al., 2001; Schmitz et al., 2003). Effects of dopaminergic agents can vary depending on an individual's baseline dopamine synthesis capacity, potentially because of increased drug sensitivity in those with low synthesis capacity resulting from upregulated receptor density and/or sensitivity (Cools et al., 2008; Landau et al., 2009).

Reported mean values for peak concentration and elimination half-life of oral haloperidol lie between 1.7 and 6.1 h and 14.5 and 36.7 h, respectively (Kudo and Ishizaki, 1999). After drug administration, participants rested for 1.5 h to allow for drug metabolization. Subsequently participants began the task battery, which included the emotion recognition task, a visual WM task and indices of drug effects on movement-based and counting-based temporal processing (see below, Tasks and procedure). Throughout the day, participants' blood pressure and oxygenation were checked hourly between tasks. All data were collected at the Center for Human Brain Health (CHBH) at the University of Birmingham, United Kingdom.

Tasks and procedure

Participants completed a task battery including tasks not described in this study. All relevant tasks are described below in order of presentation. Task order was the same on both study days.

Visual WM task

This task was implemented to establish a proxy for baseline dopamine synthesis capacity. Specifically, PET studies have shown that low WM scores are associated with low dopamine synthesis capacity, and thus have been suggested to reflect higher susceptibility to the effects of dopaminergic drugs (Cools et al., 2008). Correspondingly, for many cognitive tasks, behavioral effects of dopaminergic drugs are found to be different in individuals with low and high baseline WM span (typically determined based on median-split, Kimberg et al., 1997; Mattay et al., 2000; Gibbs and D'Esposito 2005; Frank and O'Reilly 2006). For example, on a number of tests of executive function, performance is boosted by dopaminergic drugs in low-span participants, and impaired in high-span participants (Kimberg et al., 1997; Mattay et al., 2000; Gibbs and D'Esposito, 2005).

Participants completed an adaptation of the Sternberg (Sternberg, 1966) visual WM task. Participants completed 60 trials across five blocks. On each trial, a fixation cross was displayed at the center of the screen (500–1000 ms), followed by a list of consonants (five to nine characters in length, depending on the block; 1000 ms), followed by a blue fixation cross (3000 ms). A single test letter was then displayed (4000 ms), and participants were asked whether the letter was taken from the previously displayed list (Fig. 1B). Participants responded by pressing 1–3 on the keyboard (1, Yes; 2, No; 3, Unsure). Responses (accuracy) and response time (time from test letter displayed until participant response) were recorded for each trial. Each block varied in length from five to nine consonants, with letters randomly selected from the alphabet on each trial. The total task duration was ∼10 min, and test trials were preceded by 10 practice trials.

Figure 1.
Figure 1.

Schematic depiction of main tasks. A, Depiction of one trial of PLW perception task. A fixation cross was presented for 1000 ms and followed by a PLW stimulus (on average 2000 ms). Subsequently, participants rated on three separately presented scales (each ranging from “Not at all” to “Very”) in pseudorandom order how angry, happy, and sad they perceived the PLW stimulus to be. B, Depiction of one trial of visual WM task. After presentation of a fixation cross (duration varied between 500 and 1000 ms), a list of five to nine characters was presented for 1000 ms, followed by a blue fixation cross (3000 ms).

Time estimation task

In the time estimation task, participants were asked to estimate temporal intervals of varying lengths by counting the number of seconds that had passed between two auditory signals. Four time intervals of varying lengths (between 22 and 103 s) were presented in a pseudorandom order.

Emotion recognition task

Stimuli were whole-body point light displays of male and female actors modeling angry, happy, and sad emotional walks i.e., point-light walkers PLWs; adopted from Edey et al., 2017. For each of the three affective states, the stimulus set contained 100% stimuli, which displayed the walkers at the speed they originally modeled. In line with the literature demonstrating that sadness is conveyed via slow, sluggish movements, anger with fast, jerky kinematics, and happiness intermediate to the two (Michalak et al., 2009; Roether et al., 2009; Gross et al., 2012; Halovic and Kroos, 2018), sad 100% PLWs exhibited the slowest mean speed, followed by happy, and then angry PLWs (Nackaerts et al., 2012). In addition, for each emotion, the stimulus set included three levels of velocity adapted stimuli, consisting of morphs between the speed of neutral walkers and the corresponding 100% stimuli. The resulting velocity adapted stimuli thus contained 0%, 33%, and 67% of emotion-specific speed information with full postural information. A total of 48 velocity adapted and 100% stimuli (four trials of angry, happy, and sad PLWs at four levels of speed information) were presented in pseudorandom order for an average of 2000 ms. On each trial, participants first viewed a fixation cross for 1000 ms, followed by a PLW stimulus. Subsequently three separate visual-analog scales ranging from 1 not at all to 10 very were presented one after another, in pseudorandom order, asking participants to rate how intensely they felt the stimulus was expressing an angry, happy, or sad emotional state (Fig. 1A).

Walking task

Following the emotion recognition task (task order was fixed to avoid priming effects of own speed on emotion judgements from PLWs' speeds; for more details, see Edey et al., 2017), participants performed the walking task. Individuals were asked to walk continuously between two sets of cones (placed 10 meters apart) for 120 s at their preferred walking speed. Each participant completed two walks (of 120 s) ∼30 min apart. Acceleration data were recorded, using the SensorLog app (https://apps.apple.com/us/app/sensorlog/id388014573), with an iPhone 6s attached to the outer side of the participants' left ankle.

Statistical analyses

All data were processed in MATLAB R2021a (The MathWorks Inc., 2021) and analyzed with Bayesian linear mixed effects models using the brms package (Bürkner, 2017) in R (version 4.2, R Core Team, 2021). Before model building, any continuous predictors were normalized to allow comparisons between individual estimates. For all analyses, model building was guided by our experimental design and the final model selected based on leave-one-out cross validation (using the LOO subfunction of brms; for details on all models compared, see below, Code and data accessibility). The contribution of individual predictors to the model was evaluated based on the posterior probabilities of their expected values and confirmed by comparing the model to one excluding the predictor in question. For all relevant model parameters we report expected values under the posterior distribution and their 95% credible intervals (CrIs; the term credible interval comprises the Bayesian analogue of a classical confidence interval, except that probability statements can be made based upon it; see Nalborczyk et al., 2019), as well as the posterior probability that a certain effect (Eμ) is different from zero P(Eμ < 0)/P(Eμ > 0). In line with Franke and Roettger (2019), if a hypothesis states that an effect is not equal to zero (Eμ≠ 0), we conclude there is compelling evidence for this effect if zero is not included in the 95% CrI of Eμ and if the posterior probability P(Eμ≠ 0) is close to 1. We used weakly informative priors, following a normal distribution for the intercept and all regression coefficients and a half-Cauchy distribution for residual and random effect variances (all prior distributions were centered at 0 and had a standard deviation of 10; see Nalborczyk et al., 2019). For all models, the maximal possible random effects structure allowed by the design was defined (Barr et al., 2013). Each model was run for four sampling chains with 5000 iterations each (1000 warm-up iterations). There were no indications of nonconvergence (all Rhat values = 1, no divergent transitions).

Code and data accessibility

All data and code required to reproduce the analyses within this article is publicly available under https://osf.io/gcvyj/.

Results

Based on previous evidence indicating that WM span reliably predicts individual dopamine synthesis capacity (Cools et al., 2008; Landau et al., 2009) and drug effects on performance in other cognitive domains (Kimberg et al., 1997; Mattay et al., 2000; Gibbs and D'Esposito, 2005; Frank and O'Reilly, 2006; Rostami Kandroodi et al., 2021), we used individual baseline WM span as a proxy for baseline dopamine function and thus interindividual differences in drug responsivity. For this, WM task accuracy was calculated as the percentage of correct responses of all placebo trials. Groups of low and high baseline WM span (indexing low and high dopamine synthesis capacity, respectively) were then defined by performing a median split on WM span scores from the placebo day only (resulting in 18 and 20 subjects in the low and high WM groups, respectively).

Emotion recognition task

As in Edey et al. (2017), emotion recognition scores were calculated for each emotion and speed level by subtracting the mean of the ratings for the two non-modeled emotions from the rating for the modeled emotion. For example, for a sad PLW stimulus, we subtracted the mean of the ratings on the angry and happy scales from the rating given on the sad scale. Possible values thereby ranged from −9 to 9, with high emotion recognition scores reflecting judgements of the PLW intensely expressing the modeled emotion and successful discrimination between the three emotion scales, and low or negative emotion recognition scores indicating that participants felt the PLW was weakly expressing the modeled emotion or a lack of discrimination between the three emotion scales.

Haloperidol increased emotion recognition scores in low WM span and decreased emotion recognition in high WM span individuals

To confirm that individuals made use of the emotion-specific speed information when rating PLW stimuli, and to ascertain that the drug did not affect particiants' sensitivity to the speed manipulation, an initial control model was conducted: a Bayesian linear mixed effects model with a random intercept for subject ID was fitted to the factors drug placebo PLA, haloperidol HAL; dummy coded, emotion (sad, happy, angry; effects coded), speed level (i.e., emotion-specific speed information; 0%, 33%, 67%, 100%; orthogonal polynomial coded), and WM group (low, high; effects coded), as well as all possible two-way and three-way interactions, predicting emotion recognition scores. Because of the dummy-coding of the factor drug all main effects refer to the placebo level, which are compared with effects under haloperidol via individual contrasts. The control model revealed a strong positive linear trend for the variable speed level (EμPLA,speedLevel.L = 1.29, CrI = [0.94, 1.64]), confirming that, overall, participants gave increasing emotional intensity ratings with increasing speed levels. There were no interactions between drug and speed level, indicating that haloperidol did not affect participants' sensitivity to the speed manipulation. Consequently, all following results are reported based on emotion recognition scores collapsed across the four speed levels.

In the subsequent model (Bayesian linear mixed effects model with random intercept for subject ID, factors drug, emotion and WM group, dependent variable emotion recognition scores collapsed across speed level), there was a main effect of emotion for PLA trials, with contrasts revealing that overall, sad PLWs were rated with higher intensity (EμPLA,sad = 0.65, CrI = [0.36, 0.93]), while angry PLWs were rated lower than average in terms of intensity (EμPLA,angry = −0.59, CrI = [−0.88, −0.31]). There was no main effect of drug, with the contrast of PLA and HAL emotion recognition scores being close to zero (EμPLA−HAL=−0.06, CrI = [−0.37, 0.24]).

Most interestingly and confirming our primary hypothesis, there was an interaction between drug and WM group, with a 0.94 point difference between drug effects on emotion recognition scores in the low and high WM groups (Eμ(PLA−HAL,lowWM)−(PLA−HAL,highWM) = −0.94, CrI = [−1.56, −0.32]). To evaluate drug effects in the two WM groups, two separate post hoc models were run for low and high WM groups. These models confirmed the predicted nature of differences, revealing superior performance under haloperidol versus placebo in the low WM group (EμPLA−HAL,lowWM= 0.40, CrI = [−0.06, 0.87], probability that this is a true effect: P(EμPLA−HAL,highWM > 0) = 0.96), alongside poorer performance under haloperidol in the high WM group (EμPLA−HAL,highWM= −0.53, CrI = [−0.95, −0.12], probability that this is a true effect: P(EμPLA−HAL,highWM < 0) = 0.99; see Fig. 2). These improvements under haloperidol in the low WM group were generated via increased ratings on the modeled scales and decreased ratings on the nonmodeled scales, demonstrating improved discrimination abilities under haloperidol. Note that the same pattern emerged when using a continuous variable for WM span, hence for illustrative purposes (Fig. 2), we proceeded with the binary split.

Figure 2.
Figure 2.

A, Mean emotion recognition scores for placebo and haloperidol trials by working memory (WM) group. Boxes represent 1 SEM above and below the mean (i.e., horizontal lines within boxes), shaded areas surrounding boxes represent 1 SD above and below mean values. B, C, Probability density function (PDF) of emotion recognition difference scores for low (B) and high (C) WM groups. Emotion recognition difference scores represent the difference between emotion recognition scores in haloperidol and placebo trials, where positive difference scores indicate enhanced emotion recognition performance under haloperidol. The central mark of each of the box plots below PDFs represents the median of each group, edges represent 25th (Q1) and 75th (Q3) percentiles. Whiskers denote ranges of Q3 + 1.5 × (Q3 – Q1) above and Q1 + 1.5 × (Q3 – Q1) below box edges.

Effects of haloperidol on movement-based and counting-based indices of temporal processing

Movements

A Bayesian mixed effects model of drug (PLA, HAL; dummy coded) and WM group (low, high; deviation coded) predicting walking speed indicated a trend toward a negative main effect of drug (EμPLAvsHAL = −0.04, CrI = [−0.08, 0.01], P(EμPLAvsHAL) < 0 = 0.94), indicating that, overall, haloperidol tended to reduce walking speed. In addition, there was a trend toward a main effect of WM group (EμWMgroup = 0.08, CrI = [−0.02, 0.19], P(EμWMgroup > 0) = 0.94), demonstrating that under placebo, the low WM group tended to exhibit a slower walking pace relative to high WM individuals (low WM: M = 1.05 m/s, high WM: M = 1.13 m/s). There further was an interaction between drug and WM group (EμPLAvsHAL,WMgroup = 0.09, CrI = [0.00, 0.18], P(EμPLAvsHAL,WMgroup > 0) = 0.98). Separate post hoc models for low and high WM groups indicated that, whereas the drug slowed movement speed in the low WM group, there were no drug effects on movement in the high WM group (EμPLAvsHAL,lowWM = −0.08, CrI = [−0.16, −0.01]; EμPLAvsHAL,highWM = 0.01, CrI = [−0.4, 0.6]; Fig. 3A).

Figure 3.
Figure 3.

A, Drug effects on walking speed by working memory (WM) group. Boxes represent 1 SEM above and below the mean (i.e., horizontal lines within boxes), shaded areas surrounding boxes represent 1 SD above and below mean values. B, Relationship between drug effects on walking speed and drug effects on emotion recognition scores by WM group.

To investigate whether influences of the drug on emotion recognition were modulated by performance on our movement-based index of temporal processing, we first created an index of individual drug effects on emotion recognition by subtracting emotion recognition scores of PLA trials from emotion recognition scores of HAL trials (i.e., emotion recognition difference scores). Positive emotion recognition difference scores thus indicate enhanced emotion recognition under haloperidol. Second, we estimated drug effects on walking speed by subtracting mean speed values from HAL trials from mean speed values from PLA trials for each of the two walks (i.e., speed difference values), where negative speed difference values reflect decreased walking speed in HAL relative to PLA trials. Finally, we added speed difference as a covariate to a Bayesian mixed effects model (random effects for subject ID) fitted to emotion and WM group as well as all two-way and three-way interactions, predicting emotion recognition difference scores. As noted in the Introduction, previous studies (Edey et al., 2017, 2020) have suggested that emotion recognition relies on yoking to one's own movements (whereby slow walkers perceive fast movements as intensely angry). This suggests a possible interaction between emotion and speed difference such that individuals whose movements are slowed under haloperidol exhibit increased emotion recognition difference scores for angry (i.e., fast angry stimuli appear intensely angry), and decreased emotion recognition difference scores for sad PLWs. This exploratory hypothesis was not supported. The first model revealed no interactions with emotion; therefore, all following results are based on a model excluding this factor. There was a main effect of WM group, confirming the dependency of drug effects on WM group as reported above (EμWMgroup = −0.79, CrI = [−1.59, −0.00], P(EμWMgroup < 0) = 0.98). Furthermore, there was a main effect of speed difference, indicating that drug effects on walking speed were negatively related to drug effects on emotion recognition (EμspeedDiff = −0.66, CrI = [−1.19, −0.12], P(EμspeedDiff < 0) = 0.99). A lack of interaction between WM group and speed difference scores (EμspeedDiff,WMgroup = 0.44, CrI = [−0.30, 1.21]) suggests that the relationship between drug effects on movement and drug effects on emotion perception did not depend on WM span. Thus, in both the high and low WM groups, slower movement speed under the drug was associated with increased emotion recognition; however, haloperidol-induced slowing was observed only in the low WM group (Fig. 3B).

Timing

Time perception scores were calculated by subtracting the estimate provided by the participant from the actual duration of a given interval. Time perception difference scores were calculated as the difference between time perception scores in HAL and PLA trials, where negative time perception difference scores reflect slowed time perception under haloperidol. A model with time perception difference scores added as a covariate revealed a marginal negative effect for time perception difference scores for happy PLWs only (EμtimeDiff,happy = −0.37, CrI = [−0.73, 0.00]).

Discussion

The current study tested whether the dopamine D2 receptor antagonist haloperidol modulated emotion recognition from dynamic, whole-body, motion cues. As predicted, the influence of haloperidol on emotion recognition was dependent on WM stratification. In our low WM group emotion recognition improved under haloperidol, whereas performance deteriorated in the high WM group. The low WM group also demonstrated slower own movements under the drug, with no impact of haloperidol on walking pace in the high WM group.

To the best of our knowledge, our study is the first to illustrate a clear behavioral impact of dopaminergic manipulation on the recognition of numerous emotions and our results thereby highlight the critical importance of accounting for individual differences in measures thought to reflect baseline dopamine function. Such results are consistent with an effect of a dopamine antagonist on emotion recognition previously reported in a sample of 14 males (Lawrence et al., 2002). However, whereas Lawrence and colleagues' results were restricted to anger recognition we demonstrate effects across emotions, likely because of accounting for individual differences in baseline dopamine levels. Indeed, our analyses revealed only an interaction between drug and WM span, and no main effect of drug. Thus, previous mixed neural findings and the absence of behavioral effects likely reflect such individual differences in drug response.

The observation that the low WM group exhibiting an improvement in emotion recognition also slowed their own walking pace is potentially informative with respect to the underlying mechanism. Our results illustrated a negative relationship between drug effects on movement and drug effects on recognition of all three emotions. The crucial role for the motor system in time perception has received widespread recent attention (De Kock et al., 2021), such that a temporal influence of haloperidol on movement performance is likely to reflect wider influences on temporal encoding. Consequently, haloperidol induced movement slowing may signify a slowing of internal timing processes. Furthermore, we observed a similar relationship between individual drug effects on supra-second time perception and drug effects on recognition of happy PLWs, where slower time perception under haloperidol was associated with increased emotion recognition. Thus, effects of haloperidol on both movement-based and counting-based timing tasks suggest that slowing of temporal processing is coupled with enhanced emotion recognition. Speculatively, slowed time perception may have led to increased emotion discrimination by enhancing individuals' sensitivity to temporal cues conveyed in the PLWs.

Notably we did not observe that haloperidol-related movement slowing had emotion-specific effects on recognition (slowing simply predicted improved recognition across all emotions). Such emotion-specific effects would have been interesting given our previous work (Edey et al., 2017, 2020) which indicated that we recognize emotions according to comparisons between observed kinematic features and one's own baseline kinematics, e.g., if the kinematics are faster than the observer's baseline movement kinematics the model must be angry because this is the speed at which the observer themself feels anger. To be consistent with this, haloperidol-induced slowing should have improved recognition of fast emotions (e.g., anger) yet impaired recognition of slow emotions (e.g., sadness). Our results, however, did not reveal such an interaction between emotion (depicted in the PLW) and drug effects on movement speed. Nevertheless, given the likelihood that one builds models for emotion recognition across a lifetime of experience (Hunnius and Bekkering, 2014; Edey et al., 2020), artificially slowing one's movement pace in a particular setting (e.g., via haloperidol administration) would be unlikely to re-anchor all models. Given these concerns, we did not feel confident to make strong predictions about emotion-specific effects, and we are, indeed, unsurprised to see that this pattern was not reflected in the data.

An important question concerns why we would see such dramatically different results in individuals with high WM versus low WM. Interestingly, despite the absence of an effect of haloperidol on movement speed in the high WM span group, we nevertheless observed that the drug impaired emotion recognition in this group. Thus, suggesting that timing/movement-based effects are not the only mechanism by which haloperidol can affect emotion recognition. One additional mechanism concerns haloperidol's effects on the maintenance of mental representations. Biologically-inspired models (Frank et al., 2001; Frank, 2005; Frank and Claus, 2006; O'Reilly and Frank, 2006) categorize the effects of haloperidol on mental representations in terms of putative presynaptic (i.e., primary blocking of autoreceptors leading to increased dopamine transmission) and postsynaptic (i.e., blocking of postsynaptic heteroreceptors resulting in decreased dopamine transmission) drug effects. Presynaptic effects should correspond to enhanced updating of mental representations linked to dopamine bursts (e.g., representations that are rewarded or highly salient; Bromberg-Martin et al., 2010; Diederen and Fletcher, 2021). Postsynaptic effects should result in stable representations that are robust against interference from nontarget information. Frank and O'Reilly (2006) have argued that low-span subjects exhibit significantly greater responses to haloperidol (indexed by prolactin, an indirect measure of dopamine levels; Nilsson et al., 1996) than high-span subjects and that higher doses are more likely to result in both presynaptic and postsynaptic effects. Since we used a slightly higher dose than Frank and O'Reilly (2.5 vs 2 mg) it is feasible that our low-span subjects obtained a high enough dose of haloperidol that they experienced both presynaptic and postsynaptic effects, whereas our high-span subjects experienced only mild presynaptic effects. It would follow from this that our low-span subjects should exhibit enhanced updating of rewarded/salient mental representations (the presynaptic effect) and more stable representations in general that are robust against interference from nontarget information (the postsynaptic effect). In contrast, our high-span participants should have only experienced the former (presynaptic) effect.

For accurate emotion recognition, in the context of our paradigm, one must maintain a stable and robust representation of the target PLW (e.g., angry PLW), and resist replacing it with a nontarget representation (for example, an imagined PLW prompted by a sad or happy rating scale). Thus, postsynaptic effects, which promote stable and robust mental representations would benefit emotion recognition, resulting in the pattern (high target ratings and low nontarget ratings) we observed in our low-span group. In contrast, since presynaptic effects favor flexible, rapidly updated representations, they are more likely to result in the pattern we observed in the high-span group where the target and nontarget ratings are confused. Consequently, models of the role of dopamine in the updating of mental representations (Frank et al., 2001; Frank, 2005; Frank and Claus, 2006; O'Reilly and Frank, 2006) offer a potential explanation for the differing effects we observe in the high and low-span group, and a potential pathway to explain drug effects on emotion recognition in the absence of effects on temporal processing.

Although the importance of accounting for individual differences in baseline dopamine levels has received widespread attention in other domains of cognition (Williams and Goldman-Rakic, 1995; Kimberg et al., 1997; Mattay et al., 2000; Cools et al., 2008), this study comprises the first illustration within the domain of emotion recognition. We observed that slowed temporal processing under haloperidol was associated with increased emotion recognition, indicating that drug effects on emotion perception could, at least in part, be mediated by effects on movement/timing mechanisms. However, a decline in emotion recognition performance in the absence of drug effects on movement speed in high WM individuals suggests that other mechanisms must also be at play. This work paves the way for future studies to examine how such effects play out with different types of emotion stimuli including static emotion snapshots wherein timing-based mechanisms are less relevant.

Footnotes

  • This work was supported by the European Union's Horizon 2020 Research and Innovation Program under ERC-2017-STG Grant Agreement No. 757583 (Brain2Bee; to J.L.C.).

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Bianca Schuster at biancaschuster05{at}gmail.com

References

  1. ↵
    2. Argaud S,
    3. Vérin M,
    4. Sauleau P,
    5. Grandjean D
    (2018) Facial emotion recognition in Parkinson's disease: a review and new hypotheses. Mov Disord 33:554–567. doi:10.1002/mds.27305 pmid:29473661
    OpenUrlCrossRefPubMed
  2. ↵
    2. Barr DJ,
    3. Levy R,
    4. Scheepers C,
    5. Tily HJ
    (2013) Random effects structure for confirmatory hypothesis testing: keep it maximal. J Mem Lang 68:255–278. doi:10.1016/j.jml.2012.11.001
    OpenUrlCrossRefPubMed
  3. ↵
    2. Benoit-Marand M,
    3. Borrelli E,
    4. Gonon F
    (2001) Inhibition of dopamine release via presynaptic D2 receptors: time course and functional characteristics in vivo. J Neurosci 21:9134–9141.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    2. Bromberg-Martin ES,
    3. Matsumoto M,
    4. Hikosaka O
    (2010) Dopamine in motivational control: rewarding, aversive, and alerting. Neuron 68:815–834. doi:10.1016/j.neuron.2010.11.022 pmid:21144997
    OpenUrlCrossRefPubMed
  5. ↵
    2. Bürkner PC
    (2017) brms: an R package for Bayesian multilevel models using Stan. J Stat Softw 80:28.
    OpenUrlCrossRef
  6. ↵
    2. Cools R,
    3. D'Esposito M
    (2011) Inverted-U-shaped dopamine actions on human working memory and cognitive control. Biol Psychiatry 69:e113–e125. doi:10.1016/j.biopsych.2011.03.028 pmid:21531388
    OpenUrlCrossRefPubMed
  7. ↵
    2. Cools R,
    3. Gibbs SE,
    4. Miyakawa A,
    5. Jagust W,
    6. D'Esposito M
    (2008) Working memory capacity predicts dopamine synthesis capacity in the human striatum. J Neurosci 28:1208–1212. doi:10.1523/JNEUROSCI.4475-07.2008 pmid:18234898
    OpenUrlAbstract/FREE Full Text
  8. ↵
    2. Coull JT,
    3. Hwang HJ,
    4. Leyton M,
    5. Dagher A
    (2012) Dopamine precursor depletion impairs timing in healthy volunteers by attenuating activity in putamen and supplementary motor area. J Neurosci 32:16704–16715. doi:10.1523/JNEUROSCI.1258-12.2012 pmid:23175824
    OpenUrlAbstract/FREE Full Text
  9. ↵
    2. De Kock R,
    3. Gladhill KA,
    4. Ali MN,
    5. Joiner WM,
    6. Wiener M
    (2021) How movements shape the perception of time. Trends Cogn Sci 25:950–963. doi:10.1016/j.tics.2021.08.002
    OpenUrlCrossRef
  10. ↵
    2. Delaveau P,
    3. Salgado-Pineda P,
    4. Wicker B,
    5. Micallef-Roll J,
    6. Blin O
    (2005) Effect of levodopa on healthy volunteers' facial emotion perception: an FMRI study. Clin Neuropharmacol 28:255–261. doi:10.1097/01.wnf.0000186651.96351.2e
    OpenUrlCrossRefPubMed
  11. ↵
    2. Delaveau P,
    3. Salgado-Pineda P,
    4. Micallef-Roll J,
    5. Blin O
    (2007) Amygdala activation modulated by levodopa during emotional recognition processing in healthy volunteers: a double-blind, placebo-controlled study. J Clin Psychopharmacol 27:692–697. doi:10.1097/jcp.0b013e31815a444d pmid:18004139
    OpenUrlCrossRefPubMed
  12. ↵
    2. Diederen KMJ,
    3. Fletcher PC
    (2021) Dopamine, prediction error and beyond. Neuroscientist 27:30–46. doi:10.1177/1073858420907591 pmid:32338128
    OpenUrlCrossRefPubMed
  13. ↵
    2. Edey R,
    3. Yon D,
    4. Cook J,
    5. Dumontheil I,
    6. Press C
    (2017) Our own action kinematics predict the perceived affective states of others. J Exp Psychol Hum Percept Perform 43:1263–1268. doi:10.1037/xhp0000423 pmid:28639823
    OpenUrlCrossRefPubMed
  14. ↵
    2. Edey R,
    3. Yon D,
    4. Dumontheil I,
    5. Press C
    (2020) Association between action kinematics and emotion perception across adolescence. J Exp Psychol Hum Percept Perform 46:657–666. doi:10.1037/xhp0000737 pmid:32584128
    OpenUrlCrossRefPubMed
  15. ↵
    2. Edwards J,
    3. Jackson HJ,
    4. Pattison PE
    (2002) Emotion recognition via facial expression and affective prosody in schizophrenia: a methodological review. Clin Psychol Rev 22:789–832. doi:10.1016/s0272-7358(02)00130-7 pmid:12214327
    OpenUrlCrossRefPubMed
  16. ↵
    2. Frank MJ
    (2005) Dynamic dopamine modulation in the basal ganglia: a neurocomputational account of cognitive deficits in medicated and nonmedicated Parkinsonism. J Cogn Neurosci 17:51–72. doi:10.1162/0898929052880093 pmid:15701239
    OpenUrlCrossRefPubMed
  17. ↵
    2. Frank MJ,
    3. Claus ED
    (2006) Anatomy of a decision: striato-orbitofrontal interactions in reinforcement learning, decision making, and reversal. Psychol Rev 113:300–326. doi:10.1037/0033-295X.113.2.300 pmid:16637763
    OpenUrlCrossRefPubMed
  18. ↵
    2. Frank MJ,
    3. O'Reilly RC
    (2006) A mechanistic account of striatal dopamine function in human cognition: psychopharmacological studies with cabergoline and haloperidol. Behav Neurosci 120:497–517. doi:10.1037/0735-7044.120.3.497 pmid:16768602
    OpenUrlCrossRefPubMed
  19. ↵
    2. Frank MJ,
    3. Loughry B,
    4. O'Reilly RC
    (2001) Interactions between frontal cortex and basal ganglia in working memory: a computational model. Cogn Affect Behav Neurosci 1:137–160. doi:10.3758/cabn.1.2.137 pmid:12467110
    OpenUrlCrossRefPubMed
  20. ↵
    2. Franke M,
    3. Roettger T
    (2019) Bayesian regression modeling (for factorial designs): a tutorial. Available at https://psyarxiv.com/cdxv3.
  21. ↵
    2. Gibbs SE,
    3. D'Esposito M
    (2005) Individual capacity differences predict working memory performance and prefrontal activity following dopamine receptor stimulation. Cogn Affect Behav Neurosci 5:212–221. doi:10.3758/cabn.5.2.212 pmid:16180627
    OpenUrlCrossRefPubMed
  22. ↵
    2. Gross MM,
    3. Crane EA,
    4. Fredrickson BL
    (2012) Effort-shape and kinematic assessment of bodily expression of emotion during gait. Hum Mov Sci 31:202–221. doi:10.1016/j.humov.2011.05.001 pmid:21835480
    OpenUrlCrossRefPubMed
  23. ↵
    2. Halovic S,
    3. Kroos C
    (2018) Not all is noticed: kinematic cues of emotion-specific gait. Hum Mov Sci 57:478–488. doi:10.1016/j.humov.2017.11.008 pmid:29174557
    OpenUrlCrossRefPubMed
  24. ↵
    2. Hariri AR,
    3. Mattay VS,
    4. Tessitore A,
    5. Fera F,
    6. Smith WG,
    7. Weinberger DR
    (2002) Dextroamphetamine modulates the response of the human amygdala. Neuropsychopharmacology 27:1036–1040. doi:10.1016/S0893-133X(02)00373-1 pmid:12464460
    OpenUrlCrossRefPubMed
  25. ↵
    2. Henley SMD,
    3. Novak MJU,
    4. Frost C,
    5. King J,
    6. Tabrizi SJ,
    7. Warren JD
    (2012) Emotion recognition in Huntington's disease: a systematic review. Neurosci Biobehav Rev 36:237–253. doi:10.1016/j.neubiorev.2011.06.002 pmid:21699916
    OpenUrlCrossRefPubMed
  26. ↵
    2. Hunnius S,
    3. Bekkering H
    (2014) What are you doing? How active and observational experience shape infants' action understanding. Philos Trans R Soc Lond B Biol Sci 369:20130490. doi:10.1098/rstb.2013.0490 pmid:24778386
    OpenUrlCrossRefPubMed
  27. ↵
    2. Izard C,
    3. Fine S,
    4. Schultz D,
    5. Mostow A,
    6. Ackerman B,
    7. Youngstrom E
    (2001) Emotion knowledge as a predictor of social behavior and academic competence in children at risk. Psychol Sci 12:18–23. doi:10.1111/1467-9280.00304 pmid:11294223
    OpenUrlCrossRefPubMed
  28. ↵
    2. Kimberg DY,
    3. D'Esposito M,
    4. Farah MJ
    (1997) Effects of bromocriptine on human subjects depend on working memory capacity. Neuroreport 8:3581–3585.
    OpenUrlCrossRefPubMed
  29. ↵
    2. Kudo S,
    3. Ishizaki T
    (1999) Pharmacokinetics of haloperidol. Clin Pharmacokinet 37:435–456. doi:10.2165/00003088-199937060-00001 pmid:10628896
    OpenUrlCrossRefPubMed
  30. ↵
    2. Landau SM,
    3. Lal R,
    4. O'Neil JP,
    5. Baker S,
    6. Jagust WJ
    (2009) Striatal dopamine and working memory. Cereb Cortex 19:445–454. doi:10.1093/cercor/bhn095 pmid:18550595
    OpenUrlCrossRefPubMed
  31. ↵
    2. Lawrence AD,
    3. Calder AJ,
    4. McGowan SW,
    5. Grasby PM
    (2002) Selective disruption of the recognition of facial expressions of anger. Neuroreport 13:881–884.
    OpenUrlCrossRefPubMed
  32. ↵
    2. Mattay VS,
    3. Callicott JH,
    4. Bertolino A,
    5. Heaton I,
    6. Frank JA,
    7. Coppola R,
    8. Berman KF,
    9. Goldberg TE,
    10. Weinberger DR
    (2000) Effects of dextroamphetamine on cognitive performance and cortical activation. Neuroimage 12:268–275. doi:10.1006/nimg.2000.0610
    OpenUrlCrossRefPubMed
  33. ↵
    2. Michalak J,
    3. Troje NF,
    4. Fischer J,
    5. Vollmar P,
    6. Heidenreich T,
    7. Schulte D
    (2009) Embodiment of sadness and depression—gait patterns associated with dysphoric mood. Psychosom Med 71:580–587. doi:10.1097/PSY.0b013e3181a2515c
    OpenUrlAbstract/FREE Full Text
  34. ↵
    2. Nackaerts E,
    3. Wagemans J,
    4. Helsen W,
    5. Swinnen SP,
    6. Wenderoth N,
    7. Alaerts K
    (2012) Recognizing biological motion and emotions from point-light displays in autism spectrum disorders. PLoS One 7:e44473. doi:10.1371/journal.pone.0044473 pmid:22970227
    OpenUrlCrossRefPubMed
  35. ↵
    2. Nalborczyk L,
    3. Batailler C,
    4. Lœvenbruck H,
    5. Vilain A,
    6. Bürkner PC
    (2019) An introduction to Bayesian multilevel models using brms: a case study of gender effects on vowel variability in standard Indonesian. J Speech Lang Hear Res 62:1225–1242. doi:10.1044/2018_JSLHR-S-18-0006 pmid:31082309
    OpenUrlCrossRefPubMed
  36. ↵
    2. Nilsson CL,
    3. Ekman A,
    4. Hellstrand M,
    5. Eriksson E
    (1996) Inverse agonism at dopamine D2 receptors. haloperidol-induced prolactin release from GH4C1 cells transfected with the human D2 receptor is antagonized by R(-)-n-propylnorapomorphine, raclopride, and phenoxybenzamine. Neuropsychopharmacology 15:53–61. doi:10.1016/0893-133X(95)00156-8
    OpenUrlCrossRefPubMed
  37. ↵
    2. Niv Y,
    3. Daw ND,
    4. Joel D,
    5. Dayan P
    (2007) Tonic dopamine: opportunity costs and the control of response vigor. Psychopharmacology (Berl) 191:507–520. doi:10.1007/s00213-006-0502-4 pmid:17031711
    OpenUrlCrossRefPubMed
  38. ↵
    2. O'Reilly RC,
    3. Frank MJ
    (2006) Making working memory work: a computational model of learning in the prefrontal cortex and basal ganglia. Neural Comput 18:283–328. doi:10.1162/089976606775093909 pmid:16378516
    OpenUrlCrossRefPubMed
  39. ↵
    2. Roether CL,
    3. Omlor L,
    4. Christensen A,
    5. Giese MA
    (2009) Critical features for the perception of emotion from gait. J Vis. 9:15. doi:10.1167/9.6.15
    OpenUrlAbstract
  40. ↵
    2. Rostami Kandroodi M,
    3. Cook JL,
    4. Swart JC,
    5. Froböse MI,
    6. Geurts DEM,
    7. Vahabie A-H,
    8. Nili Ahmadabadi M,
    9. Cools R,
    10. den Ouden HEM
    (2021) Effects of methylphenidate on reinforcement learning depend on working memory capacity. Psychopharmacology 238:3569–3584. doi:10.1007/s00213-021-05974-w
    OpenUrlCrossRef
  41. ↵
    2. Schmitz Y,
    3. Benoit-Marand M,
    4. Gonon F,
    5. Sulzer D
    (2003) Presynaptic regulation of dopaminergic neurotransmission. J Neurochem 87:273–289. doi:10.1046/j.1471-4159.2003.02050.x pmid:14511105
    OpenUrlCrossRefPubMed
  42. ↵
    2. Sowden S,
    3. Schuster BA,
    4. Keating CT,
    5. Fraser DS,
    6. Cook JL
    (2021) The role of movement kinematics in facial emotion expression production and recognition. Emotion 21:1041–1061. doi:10.1037/emo0000835
    OpenUrlCrossRef
  43. ↵
    2. Sternberg S
    (1966) High-speed scanning in human memory. Science 153:652–654. doi:10.1126/science.153.3736.652 pmid:5939936
    OpenUrlAbstract/FREE Full Text
  44. ↵
    2. Takahashi H,
    3. Yahata N,
    4. Koeda M,
    5. Takano A,
    6. Asai K,
    7. Suhara T,
    8. Okubo Y
    (2005) Effects of dopaminergic and serotonergic manipulation on emotional processing: a pharmacological fMRI study. Neuroimage 27:991–1001. doi:10.1016/j.neuroimage.2005.05.039 pmid:15978846
    OpenUrlCrossRefPubMed
  45. ↵
    R Core Team (2021) R: a language and environment for statistical computing, Ed 4.0.2. Vienna, Austria. Available at https://www.R-project.org/.
  46. ↵
    The MathWorks Inc. (2021) MATLAB, version R2021a. Natick: The MathWorks Inc.
  47. ↵
    2. Tomassini A,
    3. Ruge D,
    4. Galea JM,
    5. Penny W,
    6. Bestmann S
    (2016) The role of dopamine in temporal uncertainty. J Cogn Neurosci 28:96–110. doi:10.1162/jocn_a_00880 pmid:26401816
    OpenUrlCrossRefPubMed
  48. ↵
    2. Wang Y,
    3. Hawk ST,
    4. Tang Y,
    5. Schlegel K,
    6. Zou H
    (2019) Characteristics of emotion recognition ability among primary school children: relationships with peer status and friendship quality. Child Ind Res 12:1369–1388. doi:10.1007/s12187-018-9590-z
    OpenUrlCrossRef
  49. ↵
    2. Williams GV,
    3. Goldman-Rakic PS
    (1995) Modulation of memory fields by dopamine D1 receptors in prefrontal cortex. Nature 376:572–575. doi:10.1038/376572a0 pmid:7637804
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Dopaminergic Modulation of Dynamic Emotion Perception
B. A. Schuster, S. Sowden, A. J. Rybicki, D. S. Fraser, C. Press, P. Holland, J. L. Cook
Journal of Neuroscience 25 May 2022, 42 (21) 4394-4400; DOI: 10.1523/JNEUROSCI.2364-21.2022
del.icio.us logo Digg logo Reddit logo Twitter logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.