Dopamine transporter polymorphisms affect social approach–avoidance tendencies


  • D. Enter,

    Corresponding author
    1. Leiden University, Institute of Psychology, Leiden, The Netherlands
    2. Radboud University Nijmegen, Behavioural Science Institute (BSI), Nijmegen, The Netherlands
      D. Enter, Radboud University Nijmegen, Behavioural Science Institute, PO Box 9104, 6500 HE Nijmegen, The Netherlands. E-mail:
    Search for more papers by this author
  • L. S. Colzato,

    1. Leiden University, Institute of Psychology, Leiden, The Netherlands
    2. Leiden Institute of Brain and Cognition (LIBC), Leiden, The Netherlands
    Search for more papers by this author
  • K. Roelofs

    1. Radboud University Nijmegen, Behavioural Science Institute (BSI), Nijmegen, The Netherlands
    2. Donders Institute for Brain, Cognition and Behaviour, Nijmegen, The Netherlands
    Search for more papers by this author

D. Enter, Radboud University Nijmegen, Behavioural Science Institute, PO Box 9104, 6500 HE Nijmegen, The Netherlands. E-mail:


There is increasing interest in the role of striatal dopaminergic activity in social approach–avoidance motivation. The 9-repeat allele of the dopamine transporter (DAT) gene, associated with increased striatal dopamine levels, has been found to be related to increased sensitivity to reward. However, it remains unexplored whether this polymorphism influences automatic action tendencies in the social domain. We set out to test experimentally whether human carriers of the 9-repeat allele show increased approach–avoidance tendencies compared to non-9-repeat carriers. One hundred and one healthy adults, genotyped for the DAT gene, performed the social Approach–Avoidance Task, a reaction time task requiring participants to approach or avoid visually presented emotional (happy and angry) faces, by pulling a joystick towards them or pushing the joystick away from themselves, respectively. In accordance with expectations, 9-repeat carriers showed stronger approach–avoidance effects compared to non-9-repeat carriers. These results suggest a role for striatal dopaminergic polymorphisms in motivational responses to social-emotional cues. Our findings may be relevant in the selection of candidate genes in future studies involving social behavior.

The dopamine transporter (DAT) is responsible for dopamine (DA) reuptake in the striatum (Sesack et al. 1998). Genetic variations in the dopamine transporter gene (DAT1, SLC6A3) relate to deviations in expression. The chromosome 5p15.3 (Giros et al. 1992; Vandenbergh et al. 1992) contains a 40 base pair variable number of tandem repeats in the 30-untranslated region (3′UTR-VNTR). Alleles with a number of repeats ranging from 3 to 13 have been described, but the alleles with 9 and 10 repeats are the most frequently reported (Kang et al. 1999; Mitchell et al. 2000). The 9-repeat allele is associated with a reduced expression of the transporter, resulting in higher DA concentrations in the striatum as compared to the 10-repeat allele (Heinz et al. 2000; VanNess et al. 2005).

Typically, striatal dopaminergic activity is associated with motivational processes, and there is emerging evidence for a significant role of repeat polymorphisms in DAT genes. For example, DAT1 9-repeat carriers showed more striatal activity during processing of a monetary reward than 10-repeat carriers (Dreher et al. 2009; Forbes et al. 2009), suggesting increased reward sensitivity in 9-repeat carriers (see also Aarts et al. 2010). Recent work has also provided evidence for striatal involvement in the processing of social rewards. For instance, Spreckelmeyer and colleagues (2009) found increased striatal responding during the anticipation of social reward, signaled by emotional facial stimuli in healthy individuals. Interestingly, activity in overlapping brain regions was also related to the motivation to avoid social punishment (signaled by angry faces) in patients with social phobia (Cremers et al. 2011). These findings suggest that striatal functioning is not restricted to the sensitivity to obtaining social reward but also to avoiding punishment (see also Beninger et al. 1980; Darvas et al. 2011; Delgado et al. 2009). Indeed, also in healthy individuals changing a picture of an angry facial expression to one of a neutral look, elicits reward-related activity in the ventral striatum (Mühlberger et al. 2011).

On the basis of these studies, it can be suggested that striatal functioning, and the specific role of DAT1 polymorphisms, is of crucial importance for the adaptive regulation of social behavior (e.g. Caldú & Dreher 2007; Yacubian & Büchel 2009). However, direct evidence for such relation is lacking. The aim of this study was therefore to test the relationship between DAT1 polymorphisms and alterations in social motivational behavior directly, by using an implicit social approach–avoidance task (AAT). The AAT is a valid and reliable measure of social approach–avoidance tendencies (Heuer et al. 2007; Roelofs et al. 2009a,b, 2010). This reaction time task requires participants to approach and avoid socially appetitive and aversive visually presented stimuli (happy and angry faces, respectively) by pulling (approach) or pushing away a joystick (avoidance). Happy and angry faces elicit automatic approach and avoidance tendencies, respectively (Chen & Bargh 1999; Roelofs et al. 2010; Seidel et al. 2010).

On the basis of the rewarding nature of approaching happy faces and avoiding social threat, we hypothesized that healthy individuals carrying the DAT1 9-repeat polymorphism would show increased social approach–avoidance tendencies on the AAT.



One hundred and one young Caucasian healthy adults (26 male/75 female), with a mean age of 23 years (SD = 2.4, range 19–31), served as participants in exchange for partial fulfillment of course credits or a financial reward. The sample was drawn from adults in the Leiden and Rotterdam metropolitan area (The Netherlands) who volunteered to participate in studies of behavioral genetics. Exclusion criteria were any major medical illness that could affect brain function, current substance abuse, neurological conditions, a history of head injury, or a personal history of psychiatric treatment. Participants were selected by a phone interview on the basis of the Mini International Neuropsychiatric Interview script (M.I.N.I.; Lecrubier et al. 1997). Written informed consent was obtained from all participants after the nature of the study had been explained to them; the protocol was approved by the local ethical committee.

Approach–Avoidance Task (AAT)

During this reaction time (RT) task, participants were asked to classify stimuli on the basis of an aspect (color) that was orthogonal to the aspect of interest (facial emotion). Participants responded to emotional face pictures presented on a computer screen, by pulling a joystick either towards their body (approach movement) or pushing it away from their body (avoidance movement) (task adapted from Heuer et al. 2007). Pulling or pushing the joystick increased or decreased the size of the picture respectively. The speed of the size change was proportional to the amplitude of the joystick movement. As soon as the joystick reached its target position (i.e. the required direction; full movement involved a 30° rotation from the upright position) the picture disappeared from the screen. The time between the onset of the stimulus and its disappearance from the screen was recorded with >1 ms accuracy. After each completed trial, the participant moved the joystick back to its central position and initiated a new trial by pressing the fire button near the top of the joystick. Face stimuli were selected from the Karolinska Directed Emotional Faces database (Lundqvist et al. 1998). Happy, Angry, Neutral, and Disgusted1 facial expressions were taken from the same model (five male and five female models in all) and each picture was presented either with a yellowish or a greyish filter. In addition, checkerboards (10 yellow, 10 grey) were included as non-facial control stimuli. This resulted in a total of 100 different stimuli, which were presented in random order. All participants were instructed to push away yellow stimuli and to pull grey stimuli towards them, and to respond as fast and as accurately as possible. Usually, response latencies are shorter for affect-congruent (e.g. happy-approach; angry-avoid) as compared to affect incongruent response conditions (e.g. angry-approach; happy-avoid). Before the real test started, participants were presented with 18 practice trials, which were similar to the test trials except for the fact that the pictures showed different models.

DNA laboratory analysis

Genomic DNA was extracted from saliva samples by means of the Oragene™ DNA self-collection kit, and following the manufacturer's instructions (DNA Genotek, Inc., Kanata, Ontario, Canada; 2006). The DAT1 polymorphism was amplified on an MJ DNA engine thermal cycler (MJ Research), with an initial denaturation at 94°C for 4 min, followed by 32 cycles of 45 seconds at 94°C, 45 seconds at 68°C, 60 seconds at 72°C, and a final elongation of 5 min at 72°C. The 25 ml reaction mixture consisted of 50 mM Tris (pH 9.0), 20 mM NH4SO4, 3 mM MgCl2, 200 mM dNTPs, 0.5 mM primers, and 1U Taq polymerase (Invitrogen, Carlsbad, CA, USA). Products were electrophoresed on 2% agarose gel and visualized by means of ethidium bromide. The oligo primer sequences used to amplify the VNTR are DAT1-F: 5′-TGT GGT GTA GGG AAC GGC CTG AG-3′ DAT1-R: 5′-CTT CCT GGA GGT CAC GGC TCA AGG, as originally described in Waldman et al. (1998). Each individual was genotyped twice.


All participants were tested individually. They completed a 30-min reasoning-based intelligence test (SPM; Raven et al. 1988), and subsequently performed the AAT, which took about 10 min.

Statistical analyses

Independent samples t-tests were performed for analyses of age and gender between 10/10 homozygous and 9-repeat carriers. RT outliers were filtered using a <150 and >1500 ms cut-off. For each participant, the median of the remaining RTs (97%) for the correct responses was calculated per cell (defined by: Emotion and Movement). Following previous studies (Heuer et al. 2007; Rotteveel & Phaf 2004; Roelofs et al. 2005, 2009a,b, 2010; van Peer et al. 2007, 2009; Volman et al. 2011), the analysis was based on RTs for happy and angry faces, both known to elicit reliable approach–avoidance effects. RTs for neutral faces were used for baseline correction. Median RTs were baseline corrected by subtracting the corresponding RTs for neutral faces (e.g. RT angry push–RT neutral push; RT angry pull–RT neutral pull; RT happy push–RT neutral push; RT happy pull–RT neutral pull).

Corrected RTs for angry and happy faces were entered in a three-way repeated-measures Analysis of Variance (rm anova), with as between-subject factor Group (DAT1 9-repeat carriers; DAT1 10/10 homozygotes) and within-subject factors Valence (angry; happy) and Movement (push; pull). Alpha was set at.05, and effect sizes are reported in partial eta squared (ηp2).

For display purposes (Fig. 1), AAT-effect scores were calculated for each participant and for each emotion separately by subtracting median pull RTs from the corresponding median push RTs (e.g. RT angry push–RT angry pull; RT happy push–RT happy pull). These AAT-effect scores for angry and happy faces were baseline-corrected by subtracting the AAT-effect scores for neutral faces (e.g. AAT-effect angry–AAT-effect neutral; AAT-effect happy–AAT-effect neutral). Corrected AAT-effect scores were entered in a two-way repeated-measures Analysis of Variance (rm anova), with as between-subject factor Group (DAT1 9-repeat carriers; DAT1 10/10 homozygotes) and within-subject factor Emotion (happy; angry). AAT-effect scores with a negative sign (push is faster than pull) reflect a relative avoidance tendency, and AAT-effect scores with a positive sign (pull is faster than push) reflect a relative approach tendency.

Figure 1.

Neutral-corrected median AAT-effect scores (ms) for approach and avoidance movements to happy and angry faces for 9-repeat carriers and 10/10 homozygotes. DAT1 9-repeat carriers show increased approach–avoidance tendencies, compared to 10-repeat homozygotes. AAT-effect scores with a negative sign (push is faster than pull) reflect a relative avoidance tendency, and AAT-effect scores with a positive sign (pull is faster than push) reflect a relative approach tendency. *P≤ 0.05.



The genotype distribution of the DAT1 polymorphism in our population was 61 (16 males; age ± SD: 23.6 ± 2.5) 10/10 homozygous subjects (62.9%) and 36 (9 males; age ± SD: 22.8 ± 2.0) 9-repeat carrier subjects (37.1%). The allelic distribution of the gene corresponded to the Hardy Weinberg equilibrium (χ2 = 2.77, P = 0.096). No significant differences among genotype frequencies were found with respect to age (t(93) = −1.57, P = 0.121) or gender (t(95) = 0.13, P = 0.895).

Approach–Avoidance Task (AAT)

Owing to technical problems, data were missing for 4 participants, leaving 97 participants for the analyses. Table 1 provides an overview of the outcomes for RTs. Error rates were low: 3.3% for 10/10-homozygotes, and 3.7% for 9-repeat carriers.

Table 1.  Genotype-specific mean reaction times for approach (pull) and avoidance (push) responses to angry, happy, and neutral faces
Valence9-repeat carriers10/10 homozygotes
Push (ms ± SEM)Pull (ms ± SEM)Push (ms ± SEM)Pull (ms ± SEM)
Angry515 (11)549 (12)526 (9)541 (10)
Happy534 (14)535 (12)538 (10)549 (10)
Neutral524 (11)553 (13)528 (10)555 (11)

The three-way (Group (10/10 homozygotes, 9-repeat carriers) × Valence (happy, angry) × Movement (push, pull)) rm anova for the RTs showed a main effect of Valence (F(1,95) = 4.36, P = 0.039, ηp2 = 0.044), a main effect of Movement (F(1,95) = 5.63, P = 0.020, ηp2= 0.065), and a Valence × Movement interaction (F(1,95) = 6.80, P = 0.011, ηp2 = 0.067). Most critically, there was a significant Group × Valence × Movement interaction (F(1,95) = 4.09, P = 0.046, ηp2 = 0.041). In order to explore the nature of this group interaction, we conducted separate two-way (Valence (happy, angry) × Movement (push, pull)) rm anovas for each group, demonstrating a significant Valence × Movement interaction in the 9-repeat carriers group (F(1,35) = 5.92, P = 0.020, ηp2 = 0.145), but not in the 10/10 homozygous group (F(1,60) = 0.312, P = 0.578, ηp2 = 0.005). Further one-way rm anovas for the 9-repeat carriers group revealed a significant effect for movement for happy faces (F(1,35) = 7.54, P = 0.009, ηp2 = 0.177), but not for angry faces (F(1,35) = 0.21, P = 0.653, ηp2 = 0.006). However, no Group × Movement effects were found when testing effects for happy and angry faces separately (happy: F(1,95) = 1.12, P = 0.292, ηp2 = 0.012; angry F(1,95) = 1.49, P = 0.225, ηp2 = 0.015). Together, these findings indicate that 9-repeat carriers show increased social approach–avoidance tendencies, as compared to 10/10 homozygous subjects (Fig. 1).

In order to control whether the RTs were affected by individual differences related to physical characteristics of the joystick movement (unrelated to stimulus valence) we added the AAT-effect scores for the control stimuli (checkerboards) as covariate in the analyses and found that the Group × Valence × Movement interaction remained unaffected: F(1,94) = 4.11, P = 0.045, ηp2 = 0.042. When we checked for Gender we found that the Group × Valence × Movement interaction also remained significant (F(1,94) = 4.03, P = 0.048, ηp2 = 0.041), and no significant effects for Gender emerged (all P > 0.700).


The aim of this study was to explore the possible role of DAT1 polymorphisms in social-motivational behavior, as assessed by an objective measure of approach–avoidance tendencies.

Results indicated that DAT1 9-repeat carriers show increased approach-avoidance tendencies as compared to 10-repeat homozygotes. These increased approach–avoidance effects of the DAT1 9-repeat carriers were particularly reflected in a significantly stronger approach vs. avoidance tendency of happy faces. The effects remained significant after controlling for possible confounding factors such as gender and individual differences in joystick handling, and suggest that striatal dopaminergic polymorphisms do play a role in social motivation.

Increased synaptic DA availability is generally associated with increased reward-related activity, and usually linked to augmented motivational behavior (see for a review Cools 2008). In our study, the DAT1 9-repeat carriers showed increased emotion-driven action tendencies in reaction to stimuli that communicate a motivational drive to the observer, and with which their response was immediately rewarded in the sense that the speed by which the happy faces grew or the angry faces shrank and disappeared was a direct function of the speed with which participants pulled or pushed the joystick respectively. These findings correspond to a study by Talmi et al. (2008) showing a positive relation between the vigor of an action and motivation to obtain a reward, which was also related to an increased blood oxygen level-dependent signal in the Nucleus Accumbens, suggestive of increased DA signaling. In addition, Huys and colleagues (2011) have shown that the motivation of healthy individuals to acquire a reward equally influences both approaching of appetitive and avoiding of aversive stimuli. Note that the concept of avoidance used here does not imply a response that prevents exposure to the feared stimulus, but rather the tendency to withdraw as fast as possible when it appears, and relative to the tendency to approach the feared stimulus (e.g. Chen & Bargh 1999; Rinck & Becker 2007).

Interestingly, in our study the increased approach–avoidance effects of the DAT1 9-repeat carriers, compared to the 10-repeat homozygotes, were particularly reflected in a significantly stronger approach vs. avoidance of happy faces, whereas the relative avoidance tendency for angry faces did not reach significance. This relatively strong effect for happy faces in the DAT1 9-repeat carriers is in line with findings relating striatal dopamine, thought to be specifically affected by DAT1 polymorphisms (Heinz et al. 2000; VanNess et al. 2005), particularly to the approach of appetitive stimuli (Boureau & Dayan 2011; Huys et al. 2011). However, it should be noted that the group difference was only significant when the approach–avoidance tendencies for happy faces were contrasted with the approach–avoidance tendencies for angry faces, suggesting a general increase in motivational behavior in 9-repeat carriers.

Our findings indicate that the responses of individuals carrying the 9-repeat polymorphism are affected by task-irrelevant social-emotional features, whereas the responses of 10-repeat homozygotes are not. It has been suggested that the 9-repeat allele is a vulnerability factor for psychopathologies such as PTSD (Segman et al. 2002). But alternatively, our findings might signal increased sensitivity to both positive and negative contextual conditions (Belsky et al. 2009). Belsky and Pluess (2009) argued that an individual with such increased sensitivity may not only be more vulnerable to the negative effect of an adverse environment, but may also be more susceptible to beneficial consequences of a positive context. For example, an aversive context may enhance social avoidance, which might explain (mixed) results from association studies of DAT1 with anxiety (Kennedy et al. 2001; Segman et al. 2002), and be in line with earlier findings of a relation between DAT alterations and the presence of social anxiety such as reduced density of DA uptake sites in the striatum of social phobic patients relative to healthy controls (Tiihonen et al. 1997), or low DAT binding in healthy participants, which was associated with the personality trait of detachment (Laakso et al. 2000; Schneier et al. 2001).

Although DAT1 polymorphisms are typically associated with deviations in striatal DA transmission and reward processing (Dreher et al. 2009; Sesack et al. 1998), we cannot rule out that our behavioral results may also have been affected by DA signaling in other relevant brain structures involved in social approach–avoidance behavior, such as the amygdala or frontal areas (Kienast et al. 2008; Volman et al. 2011).

Our findings may also benefit the quest for clarifying issues concerning DAT expression by the DAT1 repeat polymorphism. Whereas imaging studies show mixed results for the influence of the number of repeats in the DAT1 VNTR (see for a review Costa et al. 2011), several molecular genetic studies point to the possibility of an increase in DAT1 gene expression depending on the number of repeats (Fuke et al. 2001; Michelhaugh et al. 2001; Mill et al. 2002; VanNess et al. 2005; but see Greenwood & Kelsoe 2003). These last findings support the idea of the 9-repeat allele accounting for decreased DAT1 availability when compared to the 10-repeat allele and are in line with our study, although much more research is necessary to elucidate the functional effects of the DAT1 repeat polymorphism and possible modulating factors (e.g. Shumay et al. 2010, 2011).

We acknowledge that complex human social behavior can never be attributed to one single genetic polymorphism, and that our results should therefore be considered preliminary, and for the benefit of generating hypotheses only. However, in our attempt to measure one single aspect of social behavior (e.g. behavioral approach–avoidance), we selected a sensitive, validated task that objectively measures subtle differences in implicit approach–avoidance tendencies (Heuer et al. 2007; Rinck & Becker 2007; Roelofs et al. 2010; Wiers et al. 2009). The use of well-controlled sensitive paradigms such as fMRI or implicit RT tasks, has previously been shown to result in replicable results for serotonergic transporter gene polymorphisms associated with angry face processing, where earlier self-report studies failed to find relations (Bertolino et al. 2005; Hariri et al. 2002, 2005; Heinz et al. 2005; Pezawas et al. 2005; Smolka et al. 2007). For these reasons, we believe that our findings may be relevant in the selection of candidate genes in future studies involving social motivational behavior, especially since previous studies using self-report measures of social anxiety and avoidance failed to find such correlations (e.g. Schneier et al. 2009; van der Wee et al. 2008).

In summary, this is the first attempt to relate striatal dopaminergic polymorphisms to an objective implicit measure of social approach–avoidance behavior. The findings demonstrate that DAT1 9-repeat carriers show increased approach–avoidance tendencies to social-emotional cues and may help the selection of candidate genes in future studies concerning social behavior.


This work was supported by research grants from the Netherlands Organization for Scientific Research (NWO) awarded to Karin Roelofs (Vidi grant: #452-07-008 – also supporting DE) and to Lorenza Colzato (Veni grant: #51-07-028). We acknowledge Frank Leonhard and Mike Rinck as authors of the AAT software. DE, LSC, and KR have no conflict of interest to declare.


  • 1

    The disgusted face stimuli have not yet been validated for the AAT (see also Seidel et al. 2010), and were added for the benefit of a research question beyond the scope of this study. These stimuli were therefore not included in the analyses but treated as filler stimuli.