Your relationships with the people around you – friends, family, colleagues – have a strong influence on your overall life happiness. Even so, many people struggle to engage with the people around them. Social interactions can be stressful and many people choose to avoid them, even at a cost.

Being able to measure these tendencies experimentally is a first useful step for assessing social avoidance without relying on people’s, often biased, recollections of their actions and behaviours. But how can a tendency to avoid social situations be quantified? And what can an experiment to measure this tendency reveal about the neural underpinnings of social avoidance?

Schultz et al. asked volunteers to play a social game. If they played, the volunteers had the chance to win three euros, but they could choose not to play and receive a fixed amount of money, which varied across trials between zero and three euros. This approach allowed Schultz et al. to quantify how much the volunteers valued playing the game. The game involved playing with other virtual human partners, who gave either positive or negative social feedback depending on the outcome of the game in the form of videos of facial expressions. In a non-social control experiment, a computer gave abstract feedback in the form of symbols.

Schultz et al. found that the value people placed on playing the social game varied with their level of social anxiety (established using a standard questionnaire). The more anxious people attributed less value to engaging in the game. Neuroimaging experiments revealed that the activity and connectivity between the amygdala and ventral striatum, two parts of the brain involved in processing emotions and reward-related stimuli, varied according to people’s levels of social anxiety.

Social interactions have a major impact on the quality of life of both healthy people and those with mental disorders. Developing new ways to measure and understand the differences in the brain linked to social traits could help to characterise certain conditions and document therapy progress. Methods to quantify social anxiety and avoidance are also in line with efforts to explore the neuroscience behind the full range of human behaviour.

Pursuing interpersonal relationships is a powerful human drive. Social relationships contribute to the feeling that life has meaning (Baumeister and Leary, 1995) and social isolation is a major health risk factor (House et al., 1988; Cacioppo et al., 2015) with an influence on mortality risk comparable with smoking or alcohol consumption (Holt-Lunstad et al., 2010). Several social stimuli and situations can act as a reward (Krach et al., 2010), including beautiful faces (Aharon et al., 2001; O'Doherty et al., 2003), praise and attention (Izuma et al., 2008), anticipation of positive social feedback (Spreckelmeyer et al., 2009) and interactions in the game Cyberball (Kawamichi et al., 2016). However, the value of engaging in social interactions varies across individuals (Cheek and Buss, 1981). Social situations are a source of pleasure for sociable people for which they invest time and money, such as when inviting friends for celebrating one’s birthday (Dunn et al., 2008). In contrast, socially anxious people tend to avoid social situations in order to avoid receiving negative social feedback and experiencing feelings such as fear and shame (Clark and Wells, 1995).

The neural mechanisms underlying the decision to engage in a social situation and their variations across individuals are unclear. To better understand these mechanisms, we first aimed to quantify how much individuals value engaging in a simple social situation. We measured the choice frequency of engaging in the social situation as opposed to obtaining a monetary amount, from which we inferred the value that engagement has for a given individual. A reduced frequency of choosing the social situation would indicate a reduced value of the situation, and a tendency to avoid the social situation. We hypothesised that considering whether to engage in a social situation and experiencing its outcome would evoke avoidance-related neural signals proportional to the individual level of social anxiety, specifically increased BOLD response in the amygdala, a neural structure strongly associated with fear and avoidance (Shackman and Fox, 2016; Terburg et al., 2018), decision-making (Jenison et al., 2011; Grabenhorst et al., 2012) and stimulus valuation (Paton et al., 2006). Further, we hypothesised that the reduced value of the social situation associated with higher levels of social anxiety would be reflected in reduced reward-related activations in the ventral striatum (Schultz et al., 1997; O'Doherty et al., 2004) in response to positive outcomes of the situation. Finally, we hypothesised that social anxiety may affect functional connectivity between amygdala and ventral striatum, a pathway implicated in active avoidance behaviour (Ramirez et al., 2015).

We conducted two studies on separate groups of healthy participants. We quantified social anxiety and related personality traits used to assess the specificity of our results with standard questionnaires. In Study 1, we evaluated whether the value of engaging in a social situation varied with a widely accepted measure of trait social anxiety, the Liebowitz Social Anxiety Score, LSAS (Rytwinski et al., 2009). Because high autistic traits are associated with deficits in social communication (American Psychiatric Association, 2013) and social anxiety (Kuusikko et al., 2008) yet described by partly distinct diagnostic criteria from social anxiety, we included autism quotient [AQ (Baron-Cohen et al., 2001)] as a covariate. Further, we included participant gender as predictor because gender is known to influence social anxiety (Xu et al., 2012). In Study two we replicated and extended these findings by evaluating effects of LSAS, AQ, gender, as well as two scores measuring traits frequently associated with social anxiety (Acarturk et al., 2008), namely depression [Beck’s Depression Index, BDI (Beck et al., 1996)] and trait general anxiety [STAI-T (Spielberger and Gorsuch, 1970)]. Descriptive statistics, internal consistency and reliability measures of these questionnaire data are provided in Supplementary file 1A and 1B. We then investigated the neural mechanisms involved in our task using a slightly modified version of the task.

On every trial (Figure 1A–B), participants chose between a risky option consisting of a monetary gamble with two potential known outcomes (0 and 3 Euros), presented as a game of dice against a virtual partner (one of six human characters, Figure 1C, or a computer as control) and a safe option consisting of a certain amount of money. Critically, choosing the risky option exposed participants to a display of the partner’s reaction to the game outcome (Figure 1D). The human partner’s reaction consisted of a video displaying a facial expression of admiration (i.e., positive social feedback) or condescension (i.e., negative social feedback) depending on whether the participant won or did not win (respectively), while the computer’s reactions were abstract symbols. We measured the frequency of choosing the risky or the safe option for different monetary amounts offered in the safe option. These choice frequencies allowed us to quantify the subjective value of engaging in the game against both partners and thereby estimate participants’ social avoidance or approach tendencies.

Figure 1

Download asset Open asset

Behaviour (study 1)

As expected, participants’ likelihood of choosing the safe option and thus avoid the gamble increased as a function of the amount offered as an alternative to the gamble, with both human and computer partners (Figure 2A). Based on the choice frequencies, we calculated the certainty equivalent (CE₅₀) of the gamble against each partner (CE_{50 Human}; CE_{50 Computer}), which represents a measure of the subjective value of engaging in the situation. We expected that the value of engaging in the gamble against the human, but not against the computer, would be lower in participants with more pronounced social anxiety traits (=higher LSAS score). We thus computed the difference between the estimated values of the two gambles (CE_{50 Human} minus CE_{50 Computer}), which yields a measure of the value of social engagement controlled for variations in risk preference. A multiple regression model with LSAS, AQ and gender as predictor variables could explain individual differences in the value of social engagement (global F(3,64) = 3.2, p=0.028, R²=0.13). Only LSAS contributed significantly to the prediction of the interindividual variations in the value of social engagement (value decreased with LSAS: t(64) = −2.99, p=0.004, P_corr = 0.012 Bonferroni-corrected for the use of three predictor variables, unstandardized coefficient = −0.01 Euro/LSAS point, adjusted R² = 0.09, Figure 2C, Supplementary file 1C), while variations in AQ and gender did not (respectively: t(64) = 1.75, p=0.084; t(64) = 0.24, p=0.81).

Figure 2

Download asset Open asset

To test if differences in avoidance as a function of social anxiety were due to differences in the perception of the social feedback, we analysed valence ratings of the feedback stimuli. As expected, ratings of positive and negative feedback differed (F(1,66) = 355.8, p<0.001, η_p²=0.84, higher ratings for positive feedback), ratings for human and computer partners did not (F(1,66) = 0.4, p=0.43, η_p²=0.005), and ratings differed more between negative and positive human feedback than computer feedback (interaction between partner and positive vs. negative feedback: F(1,66) = 234.8, p<0.001, η_p²=0.78). Importantly, none of these ratings varied with social anxiety (all F(1,66) < 0.24, all p>0.6, all R²<0.01).

Neuroimaging results

Decision can be decoded from amygdala activation

To examine the involvement of the amygdala in the decision-making process, we tested whether neural activity in the amygdala during the decision phase could predict trial-by-trial decisions. To this end, we trained a linear support vector machine classifier with default parameters on patterns of BOLD response in both amygdalae to discriminate between safe and risky choices against both kinds of partners (searchlight MVPA with leave-one-run-out cross-validation; see Materials and methods). The data from 39 participants were included in this analysis, one was excluded because of insufficient BOLD signal in the left amygdala. Decisions could indeed be decoded slightly above chance from activation in the basolateral subregion of the right amygdala (mean accuracy = 56.4%; MNI coordinates of peak: x = 28, y = −4, z = −16, Z = 4.05, 51 voxels, FWE-corrected at p<0.05 for multiple comparisons at the cluster level within anatomically-defined amygdalae, based on a voxelwise threshold of p<0.001 uncorrected; Figure 3A). Decoding accuracy did not vary with social anxiety (F(1,36) = 0.1, p=0.76, R²=0.00). Although decoding accuracy was relatively low, these findings are compatible with the involvement of a part of the amygdala in the decision-making process engaged by our task.

Figure 3

Download asset Open asset

Activation during decision and outcome

We focused on investigating the contributions of the amygdala and ventral striatum to the neural substrates underlying the tendencies to seek or avoid social situations. We used the BOLD responses obtained during the outcome phase of the trial to identify regions of interest (ROIs) sensitive to the human partner and to wins vs. no wins. Activation in bilateral amygdalae was higher in response to feedback from the human partner compared to feedback from the computer partner (Left: MNI coordinates x = −20, y = −6, z = −16, Z = inf., 143 voxels, p<0.001; right: MNI x = 20, y = −8, z = −16, Z = inf., 163 voxels, p<0.001; P values FWE-corrected for multiple comparisons at the voxel level across the whole brain; Figure 3B). 9% of the right amygdala cluster lay within the cluster identified in the decoding analysis reported above, 29% of the decoding analysis cluster lay within the right amygdala cluster described here, and 7.5% of the voxels of these clusters were common to both. Activation in bilateral ventral striatum was higher in response to wins vs. no wins (Left: MNI coordinates x = −10, y = 10, z = −8, Z = 3.84, p=0.014; right: MNI x = 12, y = 10, z = −8, Z = 4.03, p=0.009; P values FWE-corrected for multiple comparisons within ventral striatum; Figure 3C). We extracted parameter estimates from these ROIs to assess whether BOLD responses varied as a function of trait social anxiety (=LSAS score) or value of social engagement (CE_{50 Human} – CE_{50 Computer}) at both the decision and outcome stages of the task. These tests were independent of the contrast used to define the regions of interest, and data from each individual participant were extracted from ROIs defined in a group analysis run on the data of all the other participants (Esterman et al., 2010).

Activation during the decision: Amygdala

Participants’ amygdala BOLD activation was higher when deciding whether to engage in the game against a human rather than a computer (main effect of human vs. computer opponent: left amygdala: F(1,38) = 6.24, p=0.017, η_p²=0.141, n = 39 participants as one had insufficient BOLD signal in the left amygdala ROI; right amygdala: F(1,39) = 5.07, p=0.030, η_p²=0.12). While there was no significant difference between risky and safe decisions (left amygdala: F(1,38) = 1.85, p=0.18, η_p²=0.05; right amygdala: F(1,39) = 2.71, p=0.11, η_p²=0.07), importantly there was a significant interaction between partner and decision (left amygdala: F(1,38) = 11.61, p=0.002, ηp²=0.234; right amygdala: F(1,39) = 13.83, p<0.001, η_p²=0.262). Post-hoc tests revealed a stronger response to humans than computers when participants chose the risky option (left amygdala: t(38) = 3.35, p=0.002, d = 0.54; right amygdala: t(39) = 3.57, p=0.001, d = 0.56), and no significant difference between partners when participants chose the safe option (left amygdala: t(38) = 1.11, p=0.28, d = 0.17; right amygdala: t(39) = 0.65, p=0.52, d = 0.10, see Figure 4A and B).

Figure 4

Download asset Open asset

To relate the amygdala response to our behavioural measures, we assessed whether the interaction just described [(Risky _Human - Risky _Computer) - (Safe _Human - Safe _Computer)] varied with LSAS or our experimental estimate of the value of social engagement (CE_{50 Human} – CE_{50 Computer}). The left amygdala response showed influences of both variables (LSAS: F(1,37) = 4.67, p=0.037, R²=0.12; social engagement: F(1,37) = 5.49, p=0.025, R²=0.13, partial regression plot see Figure 4C), but the right amygdala response did not (F(1,38) < 2.6, p>0.12, R²<0.07).

Activation during the outcome: Amygdala

At the time of gamble outcome, the left amygdala response difference to outcomes of games (winning and losing) against human compared to computer partners increased with social anxiety (F(1,37) = 5.89, p=0.02, R²=0.14; Figure 4D) and decreased with the value of social engagement (F(1,37) = 6.03, p=0.02, R²=0.14; Figure 4E). These effects did not quite reach significance in the right amygdala data (LSAS: F(1,38) = 1.74, p=0.20, R²=0.05; social engagement: F(1,38) = 3.88, p=0.056, R²=0.10). The main effect of partner was not tested as the response difference between human and computer was the contrast used to define these ROIs. There was no significant difference between amygdala responses to wins or losses, nor interaction between outcome and partner, nor variations in these effects as a function of anxiety or value of social engagement (all F < 3.2, p>0.08).

Activation during the outcome: Nucleus accumbens

Right nucleus accumbens response to wins varied as a function of the partner and participants’ anxiety level: the response was higher when participants won against the human rather than the computer (F(1,36) = 5.26, p=0.028, η_p²=0.13; n = 38 participants, two were excluded because of insufficient BOLD signal) and this difference decreased with increasing social anxiety traits (F(1,36) = 5.59, p=0.024, R²=0.13, Figure 4F). This effect was not significant in the left NAcc response (F(1,37) = 2.29, p=0.14; n = 39 participants, one was excluded because of insufficient BOLD signal). Responses to ‘no win’ outcomes, and differences between not winning against a human vs. a computer did not vary with social anxiety; none of these variables varied with the value of social engagement (all F < 1.2, p>0.2, R²<0.04). The main effect of outcome (win vs. no win) was not tested as the response difference between wins and no wins was the contrast used to define these ROIs.

Functional connectivity during the decision

To investigate whether social anxiety traits influenced the functional connectivity between amygdala and nucleus accumbens during social decision-making, we performed a psychophysiological interaction analysis on the data obtained at the time of the decision (Friston et al., 1997; McLaren et al., 2012). Using our ROIs as seeds, we first examined functional connectivity between these ROIs. We extracted parameter estimates for the psychophysiological regressors based on the nucleus accumbens signal at the time of the decision from the amygdala ROIs (data were obtained from n = 38 participants). We investigated connectivity in the critical interaction [difference between human vs computer in the risky decision, minus the difference between human vs computer in the safe decision: (Risky _Human - Risky _Computer) - (Safe _Human - Safe _Computer)]. We found that connectivity between right nucleus accumbens and both the left and right amygdalae increased with social anxiety (right NAcc - left amygdala: F(1,36) = 5.5, p=0.025, R²=0.13, Figure 5A; right NAcc - right amygdala: F(1,36) = 12.56, p=0.001, R²=0.26, Figure 5B). These effects were not found in the connectivity between left nucleus accumbens and the amygdalae (all F(1,36) < 1.7, all p>0.2, all R²<0.06). None of these connectivity measures varied with the value of social engagement (all F < 3.3, all p>0.07, all R²<0.08).

Figure 5

Download asset Open asset

We then examined functional connectivity between our ROIs and the rest of the brain. We searched for regions showing connectivity increases when participants decided to engage in the social situation (interaction contrast described above), and identified three clusters with significant effects (p<0.05 FWE-corrected across the whole brain at the voxel level): dorsal anterior cingulate connected to left amygdala (Peak: MNI coordinates x = −2, y = 16, z = 40, Brodmann Area (BA) 24a/b, Z = 5.28, p=0.01); cuneus/medial occipital cortex connected to left nucleus accumbens (x = −2, y = −62, z = 8, BA 17, Z = 5.32, p=0.007), and the perigenual part of the anterior cingulate cortex connected to right nucleus accumbens (henceforth called pACC; x = 0, y = 36, z = 14; BA 24a/b, Z = 5.36, p=0.005, Figure 5C). We then examined whether connectivity in any of these clusters varied with social anxiety traits, and found significant effects in the pACC only (F(1,36) = 6.42, p=0.016, R²=0.15, Figure 5D).

This study shows that trait social anxiety is associated with reduced subjective valuation of engaging in a social situation, and amygdala and ventral striatum activation and functional connectivity differences related to social anxiety during social decision-making, both at the decision stage and when experiencing the outcome of a social situation. Interestingly, both relatively more anxious and relatively more sociable participants deviated from economically optimal decisions in order to either seek (more sociable participants) or avoid (more anxious participants) the social situation. This mirrors real-life behaviour, where sociable people spend money to interact with other people whereas socially anxious participants may instead spend money in order to avoid social interactions. Crucially, the observed behavioural effects were (i) specific to interactions with a human partner and thus not due to general risk aversion, which is increased in anxiety (Hartley and Phelps, 2012), (ii) not related to differences in the subjective perception of the human feedback stimuli as a function of anxiety, and (iii) specific for social anxiety (measured with the LSAS) and not with measures of related but non-specific traits such as general anxiety (STAI-T), depression (BDI-II) or autistic traits (AQ).

BOLD signal data acquired during this task revealed that amygdala activation may be related to the decision-making process: participants’ safe or risky choices could be decoded from activation at the time of decision-making in the right amygdala. These findings are compatible with the previously reported involvement of the primate amygdala in social decision-making (Chang et al., 2015; Grabenhorst et al., 2013). While decoding accuracy did not vary with social anxiety, amygdala activation, both at the time of social decision-making and during the outcome phase of the social situation (i.e. when participants experienced social feedback), varied with social anxiety and with the value of social engagement estimated prior to the scan. These findings are compatible with findings of increased amygdala responses to threat stimuli in people with higher trait anxiety (e.g. Etkin et al., 2004) and extend this association to the domain of social decision-making.

Nucleus accumbens response to receiving positive social feedback, that is winning the game of dice against the human compared to winning against a computer partner, decreased with social anxiety. In contrast, no effects of social anxiety were found on the response to ‘no win’ outcomes. These results are compatible with previous findings indicating that social anxiety is associated with reduced striatal response to different kinds of social rewards (Richey et al., 2014; Sripada et al., 2013).

Social anxiety influenced the functional connectivity between the left nucleus accumbens and both amygdalae, and between right nucleus accumbens and the perigenual anterior cingulate cortex (pACC). Connectivity between amygdala and nucleus accumbens during the decision to engage in the social interaction increased with social anxiety. These findings are consistent with the importance of this connection for avoidance behaviour: Disruption of the (basolateral) amygdala - nucleus accumbens (shell) projection in rats has been shown to impair avoidance behaviour (O'Doherty et al., 2004). Our findings suggest that in humans, functional connectivity between these regions when engaging in a social interaction is greater in people with more pronounced social anxiety. In contrast, nucleus accumbens – pACC connectivity decreased proportionally with increasing social anxiety. Several previous studies have reported a reduced functional connectivity in social anxiety between parietal, limbic and executive network regions during resting state or in response to presentation of emotional faces (Brühl et al., 2014). The pACC is functionally dissociable from subgenual anterior cingulate (Pezawas et al., 2005; Apps et al., 2016), connected to ventral striatum (Kunishio and Haber, 1994; Margulies et al., 2007) and reduced in volume in people with a genotype associated with increased anxiety-related temperamental traits and risks for depression (Pezawas et al., 2005). pACC has also been associated with the response to social stress and negative social feedback (Dedovic et al., 2009; Lederbogen et al., 2011) as well as with specific social cognitive functions such as the tracking of other people’s motivation (Apps et al., 2016). A dysfunction of pACC and its regulatory influence on amygdala and ventral striatum has even been associated with poor response to chronic social defeat, a possible pathophysiological mechanism of schizophrenia (Selten et al., 2017). Our findings are compatible with the idea that these differences in regulatory influence of pACC on subcortical circuits are also observable during social decision-making.

Our neuroimaging findings described above reveal a multifaceted variation in the subcortical neural correlates of both decision-making and processing of social feedback associated with social anxiety and social avoidance. These response variations are compatible with increases of threat-related responses, reduced social reward signals, increased avoidance-related network activity and reduced top-down feedback on these networks. While causal relationships between these findings cannot be assessed, all may play a role in social avoidance and social anxiety. Interestingly, the combination of higher threat-related amygdala and lower reward-related ventral striatum activity have recently been associated with increased risk of developing alcohol use disorder in response to stress (Nikolova et al., 2016). As engaging in social decisions can be stressful for many people, paradigms such as ours may allow to investigate the link between social decisions, social stress and mental disorders. Further studies specifically designed to investigate connectivity between these structures will be required to elucidate the details of the neural mechanism underlying social avoidance.

Avoidance behaviour was not reflected in explicit valence judgments of the feedback stimuli. Our effects thus seem not to have been driven by differences in the conscious interpretation of the feedback but rather by differences in the desire to expose oneself to such feedback. While reduced explicit approachability ratings for positive facial expressions have been reported in social anxiety (Campbell et al., 2009), socially anxious individuals in another study have shown avoidance behaviour without reduced valence ratings (Heuer et al., 2007). It thus appears that while implicit measures often show avoidance, effects on explicit measures may be task-dependent (Staugaard, 2010).

If our task can be taken as an example of behaviour in real-life social interactions, it may be used to quantify one aspect of the costs resulting from the avoidance behaviour observed in social anxiety. Our linear regression results suggest that persons likely to suffer from generalised social phobia would earn about 22% less in this game than people with an LSAS score of 0. It may be interesting to combine our experimental approach with estimations based on patient’s costs related to medical care or productivity loss (Wittchen et al., 2000; Patel et al., 2002; Stein et al., 2005; Acarturk et al., 2009) in future studies. It may also be interesting to run our study on patients with diagnosed social anxiety to evaluate whether our task could be useful as an experimental rather than a self-report-based measure to identify socially anxious individuals, an advance aligned with the NIMH’s RDoC proposal (Insel, 2014).

There are several limitations to the work presented here. First, the number of participants in both studies is relatively low for the assessment of interindividual differences. While the sample size of Study two roughly corresponds to the number of participants recommended by a power analysis based on the results of Study 1 (see Materials and methods), effect size estimates for inter-individual differences based on such a relatively low number of participants are bound to be imprecise (see size of the confidence bounds in Figures 2, 4 and 5) (Schönbrodt and Perugini, 2013). One should thus be careful when interpreting our results, including our estimates of the costs of social avoidance. Another issue concerns our study sample: our participants were young German participants, most of them students at the local university, who were willing to subject themselves to behavioural and neuroimaging experiments performed by unknown experimenters; such situations are likely to be quite distressing for individuals with severe levels of social anxiety. This restriction is reflected in the limited range of LSAS scores of our sample: only 11% of participants showed LSAS levels found in people suffering from generalised social phobia (>60). We must thus exercise caution when drawing inferences about clinical populations based on our results. Next, while we could significantly decode participants’ choice from activation in the right amygdala, the average accuracy was quite low. Therefore, caution must be used in interpreting this finding: the amygdala cluster identified in our analysis is unlikely to be the major contributor to participants’ choices. Further studies specifically designed for a decoding analysis and investigating additional brain regions are required to better understand the neural mechanism underlying the decision-making process in our task. Despite these considerations, the fact that we could replicate the link between social anxiety and our experimental measure of the value of social engagement across two studies, with almost identical slope estimates, suggests that our approach has identified an interesting link between a relevant but subjectively reported, real-world behavioural trait and a controlled experiment.

Data are freely available on Dryad, http://doi.org/10.5061/dryad.jq44b1r.

1. 1. Schultz J

   (2019) Dryad Digital Repository

   Data from: A human subcortical network underlying social avoidance revealed by an econometric task.

   https://doi.org/10.5061/dryad.jq44b1r

1. 1. Acarturk C
   2. de Graaf R
   3. van Straten A
   4. Have MT
   5. Cuijpers P

   (2008) Social phobia and number of social fears, and their association with comorbidity, health-related quality of life and help seeking: a population-based study

   Social Psychiatry and Psychiatric Epidemiology 43:273–279.

   https://doi.org/10.1007/s00127-008-0309-1

   • PubMed
   • Google Scholar
2. 1. Acarturk C
   2. Smit F
   3. de Graaf R
   4. van Straten A
   5. Ten Have M
   6. Cuijpers P

   (2009) Economic costs of social phobia: a population-based study

   Journal of Affective Disorders 115:421–429.

   https://doi.org/10.1016/j.jad.2008.10.008

   • PubMed
   • Google Scholar
3. 1. Aharon I
   2. Etcoff N
   3. Ariely D
   4. Chabris CF
   5. O'Connor E
   6. Breiter HC

   (2001) Beautiful faces have variable reward value: fMRI and behavioral evidence

   Neuron 32:537–551.

   https://doi.org/10.1016/s0896-6273(01)00491-3

   • PubMed
   • Google Scholar
4. 1. Ambadar Z
   2. Schooler JW
   3. Cohn JF

   (2005) Deciphering the enigmatic face: the importance of facial dynamics in interpreting subtle facial expressions

   Psychological Science 16:403–410.

   https://doi.org/10.1111/j.0956-7976.2005.01548.x

   • PubMed
   • Google Scholar
5. Book

   1. American Psychiatric Association

   (2013) Diagnostic and Statistical Manual of Mental Disorders

   American Psychiatric Pub.

   https://doi.org/10.1176/appi.books.9780890425596

   • Google Scholar
6. 1. Apps MA
   2. Rushworth MF
   3. Chang SW

   (2016) The anterior cingulate gyrus and social cognition: tracking the motivation of others

   Neuron 90:692–707.

   https://doi.org/10.1016/j.neuron.2016.04.018

   • PubMed
   • Google Scholar
7. 1. Arsalidou M
   2. Morris D
   3. Taylor MJ

   (2011) Converging evidence for the advantage of dynamic facial expressions

   Brain Topography 24:149–163.

   https://doi.org/10.1007/s10548-011-0171-4

   • PubMed
   • Google Scholar
8. 1. Baron-Cohen S
   2. Wheelwright S
   3. Skinner R
   4. Martin J
   5. Clubley E

   (2001) The autism-spectrum quotient (AQ): evidence from asperger syndrome/high-functioning autism, males and females, scientists and mathematicians

   Journal of Autism and Developmental Disorders 31:5–17.

   https://doi.org/10.1023/a:1005653411471

   • PubMed
   • Google Scholar
9. 1. Baumeister RF
   2. Leary MR

   (1995) The need to belong: desire for interpersonal attachments as a fundamental human motivation

   Psychological Bulletin 117:497–529.

   https://doi.org/10.1037/0033-2909.117.3.497

   • PubMed
   • Google Scholar
10. 1. Beck AT
    2. Steer RA
    3. Ball R
    4. Ranieri W

    (1996) Comparison of beck depression inventories -IA and -II in psychiatric outpatients

    Journal of Personality Assessment 67:588–597.

    https://doi.org/10.1207/s15327752jpa6703_13

    • PubMed
    • Google Scholar
11. 1. Biele C
    2. Grabowska A

    (2006) Sex differences in perception of emotion intensity in dynamic and static facial expressions

    Experimental Brain Research 171:1–6.

    https://doi.org/10.1007/s00221-005-0254-0

    • PubMed
    • Google Scholar
12. 1. Brühl AB
    2. Delsignore A
    3. Komossa K
    4. Weidt S

    (2014) Neuroimaging in social anxiety disorder—a meta-analytic review resulting in a new neurofunctional model

    Neuroscience & Biobehavioral Reviews 47:260–280.

    https://doi.org/10.1016/j.neubiorev.2014.08.003

    • PubMed
    • Google Scholar
13. 1. Cacioppo JT
    2. Cacioppo S
    3. Capitanio JP
    4. Cole SW

    (2015) The neuroendocrinology of social isolation

    Annual Review of Psychology 66:733–767.

    https://doi.org/10.1146/annurev-psych-010814-015240

    • PubMed
    • Google Scholar
14. 1. Campbell DW
    2. Sareen J
    3. Stein MB
    4. Kravetsky LB
    5. Paulus MP
    6. Hassard ST
    7. Reiss JP

    (2009) Happy but not so approachable: the social judgments of individuals with generalized social phobia

    Depression and Anxiety 26:419–424.

    https://doi.org/10.1002/da.20474

    • PubMed
    • Google Scholar
15. 1. Chang SW
    2. Fagan NA
    3. Toda K
    4. Utevsky AV
    5. Pearson JM
    6. Platt ML

    (2015) Neural mechanisms of social decision-making in the primate amygdala

    PNAS 112:16012–16017.

    https://doi.org/10.1073/pnas.1514761112

    • PubMed
    • Google Scholar
16. 1. Cheek JM
    2. Buss AH

    (1981) Shyness and sociability

    Journal of Personality and Social Psychology 41:330–339.

    https://doi.org/10.1037/0022-3514.41.2.330

    • Google Scholar
17. Book

    1. Clark DM
    2. Wells A

    (1995)

    A Cognitive Model of Social Phobia

    In: Heimberg RG, Liebowitz MR, Hope DA, Schneier FR, editors.  Social Phobia: Diagnosis, Assessment, and Treatment. New York: The Guilford Press. pp. 69–94.

    • Google Scholar
18. 1. Cunningham DW
    2. Wallraven C

    (2009) Dynamic information for the recognition of conversational expressions

    Journal of Vision 9:7–17.

    https://doi.org/10.1167/9.13.7

    • PubMed
    • Google Scholar
19. 1. Deaner RO
    2. Khera AV
    3. Platt ML

    (2005) Monkeys pay per view: adaptive valuation of social images by rhesus macaques

    Current Biology 15:543–548.

    https://doi.org/10.1016/j.cub.2005.01.044

    • PubMed
    • Google Scholar
20. 1. Dedovic K
    2. D'Aguiar C
    3. Pruessner JC

    (2009) What stress does to your brain: a review of neuroimaging studies

    Canadian Journal of Psychiatry. Revue Canadienne De Psychiatrie 54:6–15.

    https://doi.org/10.1177/070674370905400104

    • Google Scholar
21. 1. Dunn EW
    2. Aknin LB
    3. Norton MI

    (2008) Spending money on others promotes happiness

    Science 319:1687–1688.

    https://doi.org/10.1126/science.1150952

    • PubMed
    • Google Scholar
22. 1. Esterman M
    2. Tamber-Rosenau BJ
    3. Chiu YC
    4. Yantis S

    (2010) Avoiding non-independence in fMRI data analysis: leave one subject out

    NeuroImage 50:572–576.

    https://doi.org/10.1016/j.neuroimage.2009.10.092

    • PubMed
    • Google Scholar
23. 1. Etkin A
    2. Klemenhagen KC
    3. Dudman JT
    4. Rogan MT
    5. Hen R
    6. Kandel ER
    7. Hirsch J

    (2004) Individual differences in trait anxiety predict the response of the basolateral amygdala to unconsciously processed fearful faces

    Neuron 44:1043–1055.

    https://doi.org/10.1016/j.neuron.2004.12.006

    • PubMed
    • Google Scholar
24. 1. Faul F
    2. Erdfelder E
    3. Lang AG
    4. Buchner A

    (2007) G*power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences

    Behavior Research Methods 39:175–191.

    https://doi.org/10.3758/BF03193146

    • PubMed
    • Google Scholar
25. 1. Fox CJ
    2. Iaria G
    3. Barton JJ

    (2009) Defining the face processing network: optimization of the functional localizer in fMRI

    Human Brain Mapping 30:1637–1651.

    https://doi.org/10.1002/hbm.20630

    • PubMed
    • Google Scholar
26. 1. Friston KJ
    2. Buechel C
    3. Fink GR
    4. Morris J
    5. Rolls E
    6. Dolan RJ

    (1997) Psychophysiological and modulatory interactions in neuroimaging

    NeuroImage 6:218–229.

    https://doi.org/10.1006/nimg.1997.0291

    • PubMed
    • Google Scholar
27. 1. Furl N
    2. van Rijsbergen NJ
    3. Treves A
    4. Friston KJ
    5. Dolan RJ

    (2007) Experience-dependent coding of facial expression in superior temporal sulcus

    PNAS 104:13485–13489.

    https://doi.org/10.1073/pnas.0702548104

    • PubMed
    • Google Scholar
28. 1. Grabenhorst F
    2. Hernádi I
    3. Schultz W

    (2012) Prediction of economic choice by primate amygdala neurons

    PNAS 109:18950–18955.

    https://doi.org/10.1073/pnas.1212706109

    • PubMed
    • Google Scholar
29. 1. Grabenhorst F
    2. Schulte FP
    3. Maderwald S
    4. Brand M

    (2013) Food labels promote healthy choices by a decision Bias in the amygdala

    NeuroImage 74:152–163.

    https://doi.org/10.1016/j.neuroimage.2013.02.012

    • PubMed
    • Google Scholar
30. 1. Hartley CA
    2. Phelps EA

    (2012) Anxiety and decision-making

    Biological Psychiatry 72:113–118.

    https://doi.org/10.1016/j.biopsych.2011.12.027

    • PubMed
    • Google Scholar
31. 1. Heuer K
    2. Rinck M
    3. Becker ES

    (2007) Avoidance of emotional facial expressions in social anxiety: the Approach-Avoidance task

    Behaviour Research and Therapy 45:2990–3001.

    https://doi.org/10.1016/j.brat.2007.08.010

    • PubMed
    • Google Scholar
32. 1. Holt-Lunstad J
    2. Smith TB
    3. Layton JB

    (2010) Social relationships and mortality risk: a meta-analytic review

    PLOS Medicine 7:e1000316.

    https://doi.org/10.1371/journal.pmed.1000316

    • PubMed
    • Google Scholar
33. 1. House JS
    2. Landis KR
    3. Umberson D

    (1988) Social relationships and health

    Science 241:540–545.

    https://doi.org/10.1126/science.3399889

    • PubMed
    • Google Scholar
34. 1. Insel TR

    (2014) The NIMH research domain criteria (RDoC) Project: precision medicine for psychiatry

    American Journal of Psychiatry 171:395–397.

    https://doi.org/10.1176/appi.ajp.2014.14020138

    • PubMed
    • Google Scholar
35. 1. Izuma K
    2. Saito DN
    3. Sadato N

    (2008) Processing of social and monetary rewards in the human striatum

    Neuron 58:284–294.

    https://doi.org/10.1016/j.neuron.2008.03.020

    • PubMed
    • Google Scholar
36. 1. Jenison RL
    2. Rangel A
    3. Oya H
    4. Kawasaki H
    5. Howard MA

    (2011) Value encoding in single neurons in the human amygdala during decision making

    Journal of Neuroscience 31:331–338.

    https://doi.org/10.1523/JNEUROSCI.4461-10.2011

    • PubMed
    • Google Scholar
37. 1. Kaulard K
    2. Cunningham DW
    3. Bülthoff HH
    4. Wallraven C

    (2012) The MPI facial expression database--a validated database of emotional and conversational facial expressions

    PLOS ONE 7:e32321.

    https://doi.org/10.1371/journal.pone.0032321

    • PubMed
    • Google Scholar
38. 1. Kawamichi H
    2. Sugawara SK
    3. Hamano YH
    4. Makita K
    5. Kochiyama T
    6. Sadato N

    (2016) Increased frequency of social interaction is associated with enjoyment enhancement and reward system activation

    Scientific Reports 6:24561.

    https://doi.org/10.1038/srep24561

    • PubMed
    • Google Scholar
39. 1. Krach S
    2. Paulus FM
    3. Bodden M
    4. Kircher T

    (2010) The rewarding nature of social interactions

    Frontiers in Behavioral Neuroscience 4:22.

    https://doi.org/10.3389/fnbeh.2010.00022

    • PubMed
    • Google Scholar
40. 1. Kriegeskorte N
    2. Simmons WK
    3. Bellgowan PS
    4. Baker CI

    (2009) Circular analysis in systems neuroscience: the dangers of double dipping

    Nature Neuroscience 12:535–540.

    https://doi.org/10.1038/nn.2303

    • PubMed
    • Google Scholar
41. 1. Krumhuber EG
    2. Kappas A
    3. Manstead ASR

    (2013) Effects of dynamic aspects of facial expressions: a review

    Emotion Review 5:41–46.

    https://doi.org/10.1177/1754073912451349

    • Google Scholar
42. 1. Kunishio K
    2. Haber SN

    (1994) Primate cingulostriatal projection: limbic striatal versus sensorimotor striatal input

    The Journal of Comparative Neurology 350:337–356.

    https://doi.org/10.1002/cne.903500302

    • PubMed
    • Google Scholar
43. 1. Kuusikko S
    2. Pollock-Wurman R
    3. Jussila K
    4. Carter AS
    5. Mattila ML
    6. Ebeling H
    7. Pauls DL
    8. Moilanen I

    (2008) Social anxiety in high-functioning children and adolescents with autism and asperger syndrome

    Journal of Autism and Developmental Disorders 38:1697–1709.

    https://doi.org/10.1007/s10803-008-0555-9

    • PubMed
    • Google Scholar
44. 1. Lederbogen F
    2. Kirsch P
    3. Haddad L
    4. Streit F
    5. Tost H
    6. Schuch P
    7. Wüst S
    8. Pruessner JC
    9. Rietschel M
    10. Deuschle M
    11. Meyer-Lindenberg A

    (2011) City living and urban upbringing affect neural social stress processing in humans

    Nature 474:498–501.

    https://doi.org/10.1038/nature10190

    • PubMed
    • Google Scholar
45. 1. Maldjian JA
    2. Laurienti PJ
    3. Kraft RA
    4. Burdette JH

    (2003) An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets

    NeuroImage 19:1233–1239.

    https://doi.org/10.1016/S1053-8119(03)00169-1

    • PubMed
    • Google Scholar
46. 1. Margulies DS
    2. Kelly AM
    3. Uddin LQ
    4. Biswal BB
    5. Castellanos FX
    6. Milham MP

    (2007) Mapping the functional connectivity of anterior cingulate cortex

    NeuroImage 37:579–588.

    https://doi.org/10.1016/j.neuroimage.2007.05.019

    • PubMed
    • Google Scholar
47. 1. McLaren DG
    2. Ries ML
    3. Xu G
    4. Johnson SC

    (2012) A generalized form of context-dependent psychophysiological interactions (gPPI): a comparison to standard approaches

    NeuroImage 61:1277–1286.

    https://doi.org/10.1016/j.neuroimage.2012.03.068

    • PubMed
    • Google Scholar
48. 1. Nikolova YS
    2. Knodt AR
    3. Radtke SR
    4. Hariri AR

    (2016) Divergent responses of the amygdala and ventral striatum predict stress-related problem drinking in young adults: possible differential markers of affective and impulsive pathways of risk for alcohol use disorder

    Molecular Psychiatry 21:348–356.

    https://doi.org/10.1038/mp.2015.85

    • PubMed
    • Google Scholar
49. 1. O'Doherty J
    2. Winston J
    3. Critchley H
    4. Perrett D
    5. Burt DM
    6. Dolan RJ

    (2003) Beauty in a smile: the role of medial orbitofrontal cortex in facial attractiveness

    Neuropsychologia 41:147–155.

    https://doi.org/10.1016/S0028-3932(02)00145-8

    • PubMed
    • Google Scholar
50. 1. O'Doherty J
    2. Dayan P
    3. Schultz J
    4. Deichmann R
    5. Friston K
    6. Dolan RJ

    (2004) Dissociable roles of ventral and dorsal striatum in instrumental conditioning

    Science 304:452–454.

    https://doi.org/10.1126/science.1094285

    • PubMed
    • Google Scholar
51. 1. Patel A
    2. Knapp M
    3. Henderson J
    4. Baldwin D

    (2002) The economic consequences of social phobia

    Journal of Affective Disorders 68:221–233.

    https://doi.org/10.1016/S0165-0327(00)00323-2

    • PubMed
    • Google Scholar
52. 1. Paton JJ
    2. Belova MA
    3. Morrison SE
    4. Salzman CD

    (2006) The primate amygdala represents the positive and negative value of visual stimuli during learning

    Nature 439:865–870.

    https://doi.org/10.1038/nature04490

    • PubMed
    • Google Scholar
53. 1. Pezawas L
    2. Meyer-Lindenberg A
    3. Drabant EM
    4. Verchinski BA
    5. Munoz KE
    6. Kolachana BS
    7. Egan MF
    8. Mattay VS
    9. Hariri AR
    10. Weinberger DR

    (2005) 5-HTTLPR polymorphism impacts human cingulate-amygdala interactions: a genetic susceptibility mechanism for depression

    Nature Neuroscience 8:828–834.

    https://doi.org/10.1038/nn1463

    • PubMed
    • Google Scholar
54. 1. Puce A
    2. Allison T
    3. Bentin S
    4. Gore JC
    5. McCarthy G

    (1998) Temporal cortex activation in humans viewing eye and mouth movements

    The Journal of Neuroscience 18:2188–2199.

    https://doi.org/10.1523/JNEUROSCI.18-06-02188.1998

    • PubMed
    • Google Scholar
55. 1. Ramirez F
    2. Moscarello JM
    3. LeDoux JE
    4. Sears RM

    (2015) Active avoidance requires a serial basal amygdala to nucleus accumbens shell circuit

    Journal of Neuroscience 35:3470–3477.

    https://doi.org/10.1523/JNEUROSCI.1331-14.2015

    • PubMed
    • Google Scholar
56. 1. Richey JA
    2. Rittenberg A
    3. Hughes L
    4. Damiano CR
    5. Sabatino A
    6. Miller S
    7. Hanna E
    8. Bodfish JW
    9. Dichter GS

    (2014) Common and distinct neural features of social and non-social reward processing in autism and social anxiety disorder

    Social Cognitive and Affective Neuroscience 9:367–377.

    https://doi.org/10.1093/scan/nss146

    • PubMed
    • Google Scholar
57. 1. Rytwinski NK
    2. Fresco DM
    3. Heimberg RG
    4. Coles ME
    5. Liebowitz MR
    6. Cissell S
    7. Stein MB
    8. Hofmann SG

    (2009) Screening for social anxiety disorder with the self-report version of the liebowitz social anxiety scale

    Depression and Anxiety 26:34–38.

    https://doi.org/10.1002/da.20503

    • PubMed
    • Google Scholar
58. 1. Sato W
    2. Kochiyama T
    3. Yoshikawa S
    4. Naito E
    5. Matsumura M

    (2004) Enhanced neural activity in response to dynamic facial expressions of emotion: an fMRI study

    Cognitive Brain Research 20:81–91.

    https://doi.org/10.1016/j.cogbrainres.2004.01.008

    • PubMed
    • Google Scholar
59. 1. Sato W
    2. Yoshikawa S

    (2004) BRIEF REPORT the dynamic aspects of emotional facial expressions

    Cognition & Emotion 18:701–710.

    https://doi.org/10.1080/02699930341000176

    • Google Scholar
60. 1. Schönbrodt FD
    2. Perugini M

    (2013) At what sample size do correlations stabilize?

    Journal of Research in Personality 47:609–612.

    https://doi.org/10.1016/j.jrp.2013.05.009

    • Google Scholar
61. 1. Schultz W
    2. Dayan P
    3. Montague PR

    (1997) A neural substrate of prediction and reward

    Science 275:1593–1599.

    https://doi.org/10.1126/science.275.5306.1593

    • PubMed
    • Google Scholar
62. 1. Schultz J
    2. Brockhaus M
    3. Bülthoff HH
    4. Pilz KS

    (2013) What the human brain likes about facial motion

    Cerebral Cortex 23:1167–1178.

    https://doi.org/10.1093/cercor/bhs106

    • PubMed
    • Google Scholar
63. 1. Schultz J
    2. Pilz KS

    (2009) Natural facial motion enhances cortical responses to faces

    Experimental Brain Research 194:465–475.

    https://doi.org/10.1007/s00221-009-1721-9

    • PubMed
    • Google Scholar
64. 1. Selten J-P
    2. Booij J
    3. Buwalda B
    4. Meyer-Lindenberg A

    (2017) Biological mechanisms whereby social exclusion may contribute to the etiology of psychosis: a narrative review

    Schizophrenia Bulletin 51:sbw180.

    https://doi.org/10.1093/schbul/sbw180

    • Google Scholar
65. 1. Shackman AJ
    2. Fox AS

    (2016) Contributions of the central extended amygdala to fear and anxiety

    Journal of Neuroscience 36:8050–8063.

    https://doi.org/10.1523/JNEUROSCI.0982-16.2016

    • PubMed
    • Google Scholar
66. Book

    1. Spielberger CD
    2. Gorsuch RL

    (1970)

    STAI Manual for the State-Trait Anxiety Inventory ("Self-Evaluation Questionnaire")

    Consulting Psychologists Press.

    • Google Scholar
67. 1. Spreckelmeyer KN
    2. Krach S
    3. Kohls G
    4. Rademacher L
    5. Irmak A
    6. Konrad K
    7. Kircher T
    8. Gründer G

    (2009) Anticipation of monetary and social reward differently activates mesolimbic brain structures in men and women

    Social Cognitive and Affective Neuroscience 4:158–165.

    https://doi.org/10.1093/scan/nsn051

    • PubMed
    • Google Scholar
68. 1. Sripada C
    2. Angstadt M
    3. Liberzon I
    4. McCabe K
    5. Phan KL

    (2013) Aberrant reward center response to partner reputation during a social exchange game in generalized social phobia

    Depression and Anxiety 30:353–361.

    https://doi.org/10.1002/da.22091

    • PubMed
    • Google Scholar
69. 1. Staugaard SR

    (2010) Threatening faces and social anxiety: a literature review

    Clinical Psychology Review 30:669–690.

    https://doi.org/10.1016/j.cpr.2010.05.001

    • PubMed
    • Google Scholar
70. 1. Stein MB
    2. Roy-Byrne PP
    3. Craske MG
    4. Bystritsky A
    5. Sullivan G
    6. Pyne JM
    7. Katon W
    8. Sherbourne CD

    (2005) Functional impact and health utility of anxiety disorders in primary care outpatients

    Medical Care 43:1164–1170.

    https://doi.org/10.1097/01.mlr.0000185750.18119.fd

    • PubMed
    • Google Scholar
71. 1. Terburg D
    2. Scheggia D
    3. Triana Del Rio R
    4. Klumpers F
    5. Ciobanu AC
    6. Morgan B
    7. Montoya ER
    8. Bos PA
    9. Giobellina G
    10. van den Burg EH
    11. de Gelder B
    12. Stein DJ
    13. Stoop R
    14. van Honk J

    (2018) The basolateral amygdala is essential for rapid escape: a human and rodent study

    Cell 175:723–735.

    https://doi.org/10.1016/j.cell.2018.09.028

    • PubMed
    • Google Scholar
72. 1. Wittchen HU
    2. Fuetsch M
    3. Sonntag H
    4. Müller N
    5. Liebowitz M

    (2000) Disability and quality of life in pure and comorbid social phobia. Findings from a controlled study

    European Psychiatry 15:46–58.

    https://doi.org/10.1016/S0924-9338(00)00211-X

    • PubMed
    • Google Scholar
73. 1. Xu Y
    2. Schneier F
    3. Heimberg RG
    4. Princisvalle K
    5. Liebowitz MR
    6. Wang S
    7. Blanco C

    (2012) Gender differences in social anxiety disorder: results from the national epidemiologic sample on alcohol and related conditions

    Journal of Anxiety Disorders 26:12–19.

    https://doi.org/10.1016/j.janxdis.2011.08.006

    • PubMed
    • Google Scholar

Author details

1. Johannes Schultz

   1. Division of Medical Psychology, University of Bonn, Bonn, Germany
   2. Center for Economics and Neuroscience, University of Bonn, Bonn, Germany
   3. Institute of Experimental Epileptology and Cognition Research, University of Bonn, Bonn, Germany

   Contribution

   Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   johannes.schultz@ukbonn.de

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4117-232X

2. Tom Willems

   Division of Medical Psychology, University of Bonn, Bonn, Germany

   Contribution

   Investigation, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

3. Maria Gädeke

   Division of Medical Psychology, University of Bonn, Bonn, Germany

   Contribution

   Investigation, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

4. Ghada Chakkour

   1. Division of Medical Psychology, University of Bonn, Bonn, Germany
   2. Medical School, University of Bonn, Bonn, Germany

   Contribution

   Investigation, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

5. Alexander Franke

   1. Division of Medical Psychology, University of Bonn, Bonn, Germany
   2. Medical School, University of Bonn, Bonn, Germany

   Contribution

   Investigation, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

6. Bernd Weber

   1. Center for Economics and Neuroscience, University of Bonn, Bonn, Germany
   2. Institute of Experimental Epileptology and Cognition Research, University of Bonn, Bonn, Germany

   Contribution

   Conceptualization, Writing—review and editing

   Competing interests

   No competing interests declared

7. Rene Hurlemann

   1. Division of Medical Psychology, University of Bonn, Bonn, Germany
   2. Department of Psychiatry and Psychotherapy, University of Bonn, Bonn, Germany

   Contribution

   Conceptualization, Funding acquisition, Writing—review and editing

   Competing interests

   No competing interests declared

• 2,345

  views

• 300

  downloads

• 14

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.