Skip to main content
PMC home page
Social Cognitive and Affective Neuroscience logoLink to Social Cognitive and Affective Neuroscience
. 2012 Mar 5;8(4):404–409. doi: 10.1093/scan/nss011

Rationality and emotionality: serotonin transporter genotype influences reasoning bias

Melanie Stollstorff 1,2,✉, Stephanie E Bean 1, Lindsay M Anderson 1, Joseph M Devaney 3, Chandan J Vaidya 1,4
PMCID: PMC3624950  PMID: 22275169

Abstract

Reasoning often occurs under emotionally charged, opinion-laden circumstances. The belief-bias effect indexes the extent to which reasoning is based upon beliefs rather than logical structure. We examined whether emotional content increases this effect, particularly for adults genetically predisposed to be more emotionally reactive. SS/SLG carriers of the serotonin transporter genotype (5-HTTLPR) were less accurate selectively for evaluating emotional relational reasoning problems with belief-logic conflict relative to LALA carriers. Trait anxiety was positively associated with emotional belief-bias, and the 5-HTTLPR genotype significantly accounted for the variance in this association. Thus, deductive reasoning, a higher cognitive ability, is sensitive to differences in emotionality rooted in serotonin neurotransmitter function.

INTRODUCTION

Humans are not perfectly rational. We display biases leading to errors in reasoning and decision making (Tversky and Kahneman, 1974). One common type of bias in deductive reasoning is when one accepts or rejects a conclusion based on one’s knowledge about the world (hereafter termed beliefs) rather than logical validity (Evans et al., 1983). In relational reasoning tasks, one must evaluate whether a conclusion follows logically from the premises, regardless of the believability of the content. For example, given the premises, ‘Cars are bigger than motorcycles’ and ‘Motorcycles are bigger than airplanes’, the conclusion ‘Cars are bigger than airplanes’ is logically valid but unbelievable, as cars are not bigger than airplanes in the real world. Participants tend to be slower or less accurate in accepting unbelievable conclusions as logically valid and also in accepting believable conclusions that are logically invalid (Roberts and Sykes, 2003). Thus, accepting conclusions that are incongruent with our beliefs interferes with logical reasoning. Conversely, when the validity and belief of the conclusion are congruent (valid, believable or invalid, unbelievable), participants are faster or more accurate in evaluating the conclusion. Reduced performance for incongruent than congruent conclusions, termed the belief-bias effect, indexes how much our beliefs interfere with reasoning and has been demonstrated using categorical syllogisms and conditional and relational reasoning problems (Byrne and Tasso, 1999; Goel and Dolan, 2003; Roberts and Sykes, 2003).

In addition to our beliefs, emotion also influences deductive reasoning. Two lines of evidence indicate that manipulation of emotional processing hinders the reasoning process. First, affective state (pre-existing or induced) reduces logical reasoning performance. Patients with anxiety disorders made more errors on a conditional reasoning task (Wason Selection Task) than healthy controls when reasoning with anxiety-provoking material (de Jong et al., 1997). In healthy participants, experimentally inducing negative or positive mood resulted in more errors on the Wason Selection Task than without mood manipulation (Oaksford et al., 1996). Second, affective content (pre-existing or conditioned) reduces reasoning performance relative to neutral content. Given logically identical reasoning problems, participants made more errors when problems comprised emotionally charged statements (e.g. ‘If there is danger, then one feels nervous’) than neutral statements (e.g. ‘If one is in a library, then one sees books’) for conditional (Blanchette and Richards, 2004; Blanchette, 2006) and categorical (Lefford, 1946) reasoning problems. Reduced reasoning accuracy was also observed for neutral words associated with negative or positive emotional pictures relative to those associated with neutral pictures (Blanchette and Richards, 2004; Blanchette, 2006). Thus, neutral content with experimentally acquired emotional valence also reduced deductive reasoning performance. Together, these findings indicate that emotional state and content influence logical reasoning.

If emotional processing influences logical reasoning, then individual differences in emotional reactivity ought to influence reasoning performance. One source of individual differences in emotional reactivity is a polymorphism in the promoter region of the serotonin transporter gene (5-HTTLPR) that results in short (S) and long (L) variants. The S allele is linked to lower expression of serotonin transporter mRNA (Hu et al., 2006). Further, the L allele contains an A to G single-nucleotide polymorphism (SNP, rs25531) that influences transcriptional efficiency, rendering the LG allele functionally similar to the S allele (Hu et al., 2006). Findings of studies comparing S carriers (SS alone or with SLG) with homozygous L carriers (e.g. LL or LALA) suggest that the S allele is associated with higher emotional reactivity. Specifically, S allele carriers were overrepresented in patients with affective disorders (Caspi et al., 2003), exhibited more depressive and anxiety-related behaviors despite being healthy (Lesch et al., 1996; Gonda et al., 2009) and showed a stronger bias towards emotional content in spatial attention (Beevers et al., 2009) and interference (Koizumi et al., 2010) tasks. Further, the amygdala, a critical brain region underlying emotional behavior, is more responsive to negative stimuli in healthy S carriers (Munafo et al., 2008). Together, these findings indicate that S (and LG) carriers differ in emotional reactivity from L carriers (and LA alone). No study has examined whether these allelic differences influence emotional processing in the context of logical reasoning.

We investigated the effect of 5-HTTLPR genotype on belief-bias in relational reasoning problems with and without emotional content. In light of evidence indicating functional similarity between the S and LG alleles (Hu et al., 2006), we included LG carriers in the S group as done in past work (Armbruster et al., 2009). Other heterozygous carriers (SLA, LALG) were excluded in order to maximize observed allelic differences (Roiser et al., 2009). We predicted that carriers of the 5-HTTLPR S or LG alleles (SS, SLG, LGLG) would demonstrate increased belief-bias relative to homozygous carriers of the LA allele (LALA) during reasoning with emotional, but not non-emotional, content. This prediction is based upon findings of increased sensitivity to negative affective stimuli in S (and LG) carriers, which ought to make suppression of beliefs with emotional content more difficult relative to LALA carriers. Such a finding would indicate that serotonin-related differences in emotional reactivity influence cognitive control during deductive reasoning. Furthermore, in light of past findings relating the S allele to increased anxiety, we also examined the association between the 5-HTTLPR genotype, trait anxiety and belief-bias for emotional material.

METHODS

Participants

One hundred and sixty-nine Georgetown University undergraduates (65 males) of primarily European descent without history of psychiatric diagnosis or medication, who were native English speakers or fluent by age 10 years, participated for course-credit or payment. Consent was acquired according to Institutional Review Board guidelines. Participants provided a saliva sample that was analysed for 5-HTTLPR and the rs25531 SNP in the serotonin transporter gene (SLC6A4). Genotype frequencies were in Hardy–Weinberg equilibrium (χ2 = 2.45, df = 1, P > 0.1). For SLC6A4, after excluding LAS and LALG genotypes, our final sample included two groups, LALA and SS/SLG. The LALA group (N = 41; 44% males; 85% European descent; age: M = 19.4, s.d. = 1.2) did not differ from the SS/SLG group (N = 34; 41% males; 74% European descent; age: M = 19.1, s.d. = 1.1) in age (P > 0.4), gender (P > 0.8) or ethnicity (P > 0.2). The SS/SLG group was composed of SS (n = 22) and SLG (n = 12) carriers. No participants had the rare LGLG genotype.

Genotyping

Saliva samples (Oragene, Canada) were analyzed for 5-HTTLPR using a two-step process. First, the long (L) and short (S) variants were determined. The repeat polymorphism in the promoter region of the 5-HTT gene was genotyped by PCR as previously described (Lesch et al., 1996) using the following primers at concentrations of 10 µM; forward: 5'-GGCGTTGCCGCTCTGAATGC-3'; reverse: 5'-GAGGGACTGAGCTG-GACAACCAC-3'. PCR was performed using the AccuPrime™ GC-Rich DNA polymerase system (Invitrogen) with the following PCR program: 95°C for 10 min, followed by 35 cycles of 95°C for 30 s, 65°C for 30 s and 72°C for 1 min. A final extension time of 72°C for 10 min was performed after the 35 cycles were complete. The PCR products were then run out on a 2% agarose gel stained with ethidium bromide. The amplification yielded distinct bands at 484 bp (S allele = 14 copies of repeat) and 528 bp (L allele = 16 copies of repeat), which were distinguished by a 100 bp DNA ladder run on the same gel. Second, the LA and LG variants were determined for the rs25531 SNP, present only on the long allele. Genotyping for rs25531 was performed by digesting the PCR products generated from the 5-HTTLPR PCR reactions with the restriction enzyme MspI (New England BioLabs). Specifically, 10 µl restriction digestion reactions were performed by combining 8 µl of the 5-HTTLPR PCR product, 1 µl of 10× NEBuffer 4 and 1 µl of MspI (concentration = 100 000 U/ml) and incubating the reactions for 2 h at 37°C followed by heat inactivation of the enzyme at 80°C for 20 min. The substitution of the G for A in the SNP produces an additional MspI recognition site (CCGG) on the long allele of the 5-HTTLPR PCR product. Genotypes were determined by running the digested PCR products out on a 2% agarose gel stained with ethidium bromide. Samples with two copies of the A allele for rs25531 showed a band at 340 bp (as well as bands at 127 and 62 bp due to multiple MspI recognition sites on the 5-HTTLPR PCR product), while samples with two copies of the G allele for rs25531 had additional digestion of the 340 bp product, yielding bands at 166 and 174 bp (as well as bands at 127 and 62 bp). Samples that were heterozygous for rs25531 showed a combination of these two band patterns.

Stimulus materials

Stimuli consisted of 96 three-term relational reasoning problems (e.g. A > B, B > C, therefore A > C) that varied by emotion and congruency (Table 1). Words for the reasoning problems were selected from the Affective Norms for English Words database [ANEW; (Bradley and Lang, 1999)], which provides ratings for arousal and valence on a 10-point scale. Emotional problems contained primarily negative (<4 = negative, 84.7%, >7 = positive valence, 15.3%) and highly arousing (>3.5; mean: 5.9) words, whereas non-emotional problems contained words that were neither positive nor negative (4–7 valence) and low in arousal (<5.5; mean: 4.2).

Table 1.

Experimental conditions and example stimuli

Trial type Emotional Non-emotional
Congruent Cockroaches are smaller than snakes. Adults are younger than children.
Cockroaches are bigger than maggots. Adults are older than infants.
Snakes are bigger than maggots? Children are older than infants?
(Valid, Believable); 24 trials (Valid, Believable); 24 trials
Incongruent Tobacco is more poisonous than venom. Trees are taller than flowers.
Tobacco is less poisonous than mucus. Trees are shorter than grass.
Venom is less poisonous than mucus? Flowers are shorter than grass?
(Valid, Unbelievable); 24 trials (Valid, Unbelievable); 24 trials
Open in a new tab

Problems also varied on belief-logic congruency. For congruent problems, the validity of the conclusion was in accordance with semantic beliefs (valid, believable and invalid, unbelievable). For incongruent problems, the validity of the conclusion was in conflict with semantic beliefs (valid, unbelievable and invalid, believable). The 96 logic problems varied by Emotional content (emotional, non-emotional) and Congruency (congruent, incongruent), creating four conditions, 24 problems each: Emotional Congruent, Emotional Incongruent, Non-emotional Congruent and Non-emotional Incongruent. Conditions were equated for conclusion believability (12 believable, 12 unbelievable), validity (12 valid, 12 invalid), determinacy (18 determinate, 6 indeterminate) and content type (13 non-living, 6 living, 5 abstract).

Procedure

Practice

Following task description and explanation of the task (what constitutes a logical conclusion), participants completed 14 practice problems where they determined if the conclusion followed logically from the premises by basing the decision on logical form and not on the factual truth or falsity of the conclusion. Participants were given unlimited time and were asked to re-think the problem if they made errors and to correct the error until all problems were correctly solved.

Reasoning task

Problems were presented on a computer screen in pre-determined random order held constant across participants. Participants were instructed to press the ‘F’ key if a problem was ‘logical’ and ‘J’ key if it was ‘not logical’, as quickly as possible. Premise 1 appeared on the screen for 3 s, followed by Premise 2 below it for 3 s, and then the conclusion below it, after which all three remained on the screen for 6 s. Participants had 6 s to respond, and the next problem appeared immediately after their response or after the 6 s. No feedback was provided.

Belief questionnaire

The questionnaire measured whether the participant’s beliefs matched those of the experimenters. Forty-eight conclusions (half believable, half unbelievable) were selected randomly, including 12 problems from each of the four experimental conditions. Participants were asked to mark each conclusion as ‘True’, ‘False’ or ‘Don’t Know’, based on their own knowledge.

Working memory

Participants completed a verbal N-back task, consisting of six alternating 1.2-min blocks of two- and three-back conditions (‘low’ and ‘high’ working memory load, respectively). Each block comprised 24 trials preceded by an instruction screen stating the type of trial in the block, for example, ‘2-back’ or ‘3-back’. For all conditions, one letter was presented on the screen at a time (for 0.5 s followed by a 2.5-s inter-trial interval), and the participant was instructed to press a button with their right index finger on the keyboard when the letter on the screen was the same as the one presented n trials previously. In the two-back condition, participants were instructed to press the button if the letter was the same as 2 before it (e.g. ‘R’ then ‘L’ then ‘R’); in the three-back condition, participants were instructed to press the button if the letter was the same as 3 before it (e.g. ‘M’ then ‘K’ then ‘P’ then ‘M’). The number of target responses was identical across trial conditions. Stimuli comprised consonants only; vowels were omitted to prevent encoding series of letters as pronounceable strings.

Trait anxiety

Participants completed the STAI (Spielberger et al., 1983), a self-report measure of state and trait anxiety. Scores of trait but not state anxiety were used in further analysis in order to examine the influence of a stable rather than situational characteristic of emotionality.

RESULTS

Belief questionnaire

For each participant, responses were coded based upon the match with the experimenter as agree (true/false by both), disagree (true/false mismatch between the two) or uncertain (‘don’t know’). Mean agreement was high, disagreement and uncertainly low and independent t-tests revealed no genotype differences. Mean percentage of ‘agree’ responses was emotional: congruent (LALA: M = 94.1%, s.d. = 7.2; SS/SLG: M = 92.1%, s.d. = 8.6; P > 0.2) incongruent (LALA: M = 92.9%, s.d. = 6.7; SS/SLG: M = 92.8%, s.d. = 5.0; P > 0.9); non-emotional: congruent (LALA: M = 94.3%, s.d. = 6.9; SS/SLG: M = 93.0%, s.d. = 6.4; P > 0.3) incongruent (LALA: M = 92.9%, s.d. = 6.3; SS/SLG: M = 91.4%, s.d. = 7.3; P > 0.3). Since average disagreement (2.4%) and uncertainty (4.5%) across conditions was extremely low, mean percentage of disagreement (LALA: M = 2.4%, s.d. = 1.6; SS/SLG: M = 2.6%, s.d. = 1.8; P > 0.5) and uncertainty (LALA: M = 4.2%, s.d. = 2.8; SS/SLG: M = 4.8%, s.d. = 2.9; P > 0.3) were collapsed across conditions.

Reasoning task

A response was scored as ‘correct’ if it was consistent with the logical validity of the problem and ‘incorrect’ if it was not consistent with logical validity or if there was no response within 6 s (‘timed-out’; M = 7% of problems). For each participant, mean accuracy (% correct) and mean reaction time (ms) for correct responses were computed for congruent and incongruent problems with and without emotional content (Table 2). Mixed ANOVAs with genotype (SS/SLG vs LALA) as a between-subject factor and congruency (congruent vs incongruent) and emotion (emotional vs non-emotional content) as within-subjects factors were computed separately for accuracy and reaction time. Controlling for ethnicity and for working memory performance (3-back accuracy) did not change the significance of any reported results.

Table 2.

Mean (s.d.) accuracy and reaction time for relational reasoning problems with emotional and non-emotional content in SS/SLG and LALA carriers

Short (SS/SLG; N = 34) Long (LALA; N = 41)
Accuracy
    Emotional
        Congruent 85.57% (10.31) 85.88% (10.19)
        Incongruent 73.93% (14.01) 80.74% (10.82)
    Non-emotional
        Congruent 85.06% (11.54) 88.27% (8.19)
        Incongruent 77.94% (12.59) 80.67% (10.75)
Reaction time
    Emotional
        Congruent 2482 ms (718) 2411 ms (572)
        Incongruent 2850 ms (584) 2700 ms (468)
    Non-emotional
        Congruent 2586 ms (546) 2447 ms (528)
        Incongruent 2785 ms (588) 2676 ms (521)
Open in a new tab

Accuracy

A main effect of congruency [F(1, 73) = 65.99, P < 0.001, η2 = 0.47] indicated a significant belief-bias effect as participants were more accurate for belief-logic congruent (M = 86.19%, s.d. = 8.77) than incongruent (M = 78.32%, s.d. = 10.83) problems. While no other main effects or two-way interactions reached significance (Ps > 0.10), the genotype × congruency × emotion interaction was significant [F(1, 73) = 6.28, P = 0.014, η2 = 0.08]. Planned comparisons testing for group differences indicated that accuracy was lower in SS/SLG (M = 73.93%, s.d. = 14.01) relative to LALA (M = 80.74%, s.d. = 10.82) participants, only for problems in the emotional incongruent condition [t(73) = 2.37, P = 0.020]; genotype groups did not differ in other conditions (P’s > 0.20). Further, planned comparisons showed that each genotype group exhibited belief-bias (congruent > incongruent) for both emotional (SS/SLG: t(33) = 6.31, P < 0.001; LALA: t(40) = 3.240, P < 0.001) and non-emotional (SS/SLG: t(33) = 3.71, P = 0.001; LALA: t(40) = 5.30, P = 0.002) conditions.

As expected based on the three-way interaction, a genotype × emotion ANOVA on the amount of belief bias (Congruent minus Incongruent) revealed an interaction [F(1, 73) = 6.28, P = 0.014, η2 = 0.08], such that SS/SLG carriers (M = 11.64%, s.d. = 10.75) had higher belief-bias relative to LALA carriers (M = 5.14%, s.d. = 10.16) for emotional problems [t(73) = 2.68, P = 0.009; Figure 1]. Amount of belief-bias did not differ by genotype for non-emotional problems (SS/SLG: M = 7.12%, s.d. = 11.19; LALA: M = 7.60%, s.d. = 9.18; P > 0.8). Further, SS/SLG carriers had higher belief-bias for emotional (M = 11.64%, s.d. = 10.75) relative to non-emotional (M = 7.12%, s.d. = 11.19) problems [t(33) = 2.234, P = 0.032]; belief-bias in LA/LA carriers did not differ by emotional content (P > 0.2). No main effects reached significance (P’s > 0.1).

Fig. 1.

Fig. 1

Open in a new tab

Mean belief-bias (percentage of correct congruent – incongruent) for relational reasoning problems with emotional and non-emotional content in LALA and SS/SLG carriers (*P < 0.05).

Reaction time

A main effect of congruency [F(1, 73) = 64.86, P < 0.001, η2 = 0.47] showed that participants were faster to evaluate conclusions of congruent (M = 2477 ms, s.d. = 555) than incongruent (M = 2747 ms, s.d. = 505) problems. Thus, participants’ response latencies exhibited a belief-bias effect. No other main effects or interactions reached significance (P’s > 0.1, Table 2).

Working memory

A between-subject ANOVA showed that mean accuracy for two-back (LALA: M = 92.78%, s.d. = 7.89; SS/SLG: M = 91.84%, s.d. = 9.82) and three-back (LALA: M = 88.03%, s.d. = 14.11; SS/SLG: M = 82.17%, s.d. = 17.89) working memory conditions did not differ between groups (Ps > 0.1). Furthermore, neither two-back (P = 0.51) nor three-back (P = 0.20) accuracy correlated with emotional belief-bias scores. Including three-back accuracy as a covariate did not change any significance value in our belief-bias ANOVAs or correlational analyses.

Trait anxiety

A between-subject ANOVA showed that mean standard scores were higher in SS/SLG (M = 50.48, s.d. = 8.07) than LALA (M = 46.32, s.d. = 7.56) participants [F(1,71) = 5.25, P = 0.025, η2 = 0.07]. Further, anxiety scores correlated positively with the amount of belief-bias for problems with emotional (r = 0.24, P = 0.04) but not non-emotional (r = 0.03, P = 0.789) content (Figure 2). Thus, individuals with higher trait anxiety were more biased towards their beliefs during reasoning with emotional content alone.

Fig. 2.

Fig. 2

Open in a new tab

Correlation between trait anxiety standard scores and (A) emotional and (B) non-emotional belief-bias (percentage of correct congruent–incongruent). *P < 0.05.

OLS linear regression

In light of the association between anxiety and emotional belief-bias and higher anxiety in SS/SLG carriers, we examined the role of genotype in the relationship between anxiety and emotional belief-bias. A simple regression model with anxiety scores as the sole independent variable revealed that the estimated association between anxiety and emotional belief-bias was 0.32 with an associated standard error (s.e.) of 0.15 (P = 0.04; standardized β = 0.24, replicating the bivariate correlation). This association was reduced to non-significance (P > 0.1) upon adding 5-HTTLPR genotype to the model. This reduction suggests that the positive association between anxiety and emotional belief-bias is explained by the covariance between both behavioral variables (anxiety and emotional belief-bias) and genotype. Moreover, genotype was significantly associated with emotional belief-bias even with anxiety held constant (b = 5.61, s.e. = 2.46, P = 0.02). The unstandardized coefficient of 5.61 indicates that the SS/SLG group scored 5.61% points higher on emotional belief-bias than the LALA group, even with anxiety held constant. With 5-HTTLPR and anxiety included, the model accounted for 12% of variance in emotional belief-bias.

DISCUSSION

A polymorphism of the 5-HTTLPR genotype influenced the extent to which beliefs interfered with deductive reasoning, selectively for emotional content. Overall, participants exhibited belief-bias, defined by more errors and slower evaluation of conclusions of relational reasoning problems in which beliefs and logical structure were in conflict relative to congruent. However, the amount of belief-bias was larger in carriers of the S/LG alleles relative to the LA allele, for problems with highly arousing and emotionally valenced content but not for those with less arousing, less emotional content. Further, trait anxiety correlated positively with belief-bias for emotional problems, and the 5-HTTLPR genotype accounted significantly for the variance in this relationship. Thus, individual variation in deductive reasoning, a higher cognitive ability, depends upon an interaction between semantic content and serotonin-based genetic differences in emotional reactivity.

It has been argued that reasoning with meaningful content involves inhibiting one’s semantic knowledge in order to process the logical structure (Houde et al., 2000; Handley et al., 2004; Prado and Noveck, 2007; De Neys et al., 2008; Luo et al., 2008; De Neys and Franssens, 2009). Behaviorally, suppression of one’s beliefs is effortful as reflected in failures or slower speed of reasoning for problems where beliefs and logic conflict, as observed in this study. Support for the involvement of inhibitory processing during deductive reasoning comes from two sources. First, inhibitory control abilities are associated with the amount of belief-bias. Children with higher response inhibition (as measured by the stop signal task) had lower belief-bias in conditional and relational reasoning tasks (Handley et al., 2004). Second, functional brain imaging studies show that a brain region known to support inhibitory control, right inferior frontal cortex [IFC; (Aron et al., 2004)] was consistently activated during belief-bias [e.g. incongruent vs congruent and neutral problems (Goel et al., 2000; Stollstorff et al., 2012); logical vs belief-based responses for incongruent problems (Goel and Dolan, 2003)] and increased activity of this region but not its left-hemisphere homologue was associated with less belief-bias (Tsujii and Watanabe, 2010). Further, disruption of neural activity in the right, but not left, IFC with transcranial magnetic stimulation reduced accuracy for incongruent, but not congruent, syllogisms, thereby increasing belief-bias relative to a control group (Tsujii et al., 2010). Together, these findings suggest that successful logical reasoning requires the inhibition of interference from semantic knowledge.

Our findings suggest that inhibiting emotionally valenced semantic knowledge was selectively effortful for carriers of the 5-HTTLPR short allele. Overall, reasoning abilities were similar across genotype groups, as their accuracy did not differ on either incongruent or congruent problems without emotional content [consistent with Gong et al. (2011), who found no relation between analogical reasoning ability and 18 functional genetic variants influencing neurotransmitter function, including several that influence 5-HT function]. Thus, greater interference from beliefs in S carriers was specific to reasoning with emotional material. Logic problems in this study included primarily negatively valenced words, with only a few highly arousing positive words. Thus, our results cannot be examined by valence, but should be in future studies in light of past findings of the 5-HTTLPR polymorphism. Specifically, S/LG carriers showed a higher attentional bias towards negative faces, whereas L/LA carriers showed a higher attentional bias towards happy faces in an emotional dot-probe task (Beevers et al., 2009)]. These findings together with the observed interaction between emotional valence and 5-HTTLPR genotype serve to elucidate past findings showing reduced reasoning performance with both negative and positive content relative to non-emotional content (Blanchette and Richards, 2004; Blanchette, 2006). It is possible that positive emotional content was detrimental selectively for LALA carriers (due to their attentional bias towards positive stimuli), and negative emotional content was detrimental selectively for S/LG carriers (due to their attentional bias towards negative stimuli). This prediction should be tested in future work. Further, it would also be important to examine whether the 5-HTTLPR genotype influences belief-bias for other types of emotional content such as politically incorrect statements violating social norms. A recent study found reduced belief-bias for such material (Goel and Vartanian, 2011); perhaps such beliefs are easier to suppress as one is practiced at doing so in everyday life. Such content, comprising beliefs one should not have, may differ in its susceptibility to emotional reactivity than emotional content detached from social norms such as that used in our study.

Enhanced belief-bias from emotional content in S carriers who are known to be more emotionally reactive may result from two sources: Parallel to findings from emotional Stroop-like tasks [e.g. color-word and face-word (Koizumi et al., 2010)], S carriers’ increased attention to negative emotional content could have increased inhibitory demands making its suppression more difficult than that of non-emotional information. Another possibility is that the emotional content, which was primarily negative, could have temporarily evoked a negative mood state. Indeed, past findings indicate that S carriers have a higher propensity for negative mood (Lesch et al., 1996; Gonda et al., 2009). Thus, negative emotional material may draw more attention and induce a negative affective state, two factors known to impede reasoning (Oaksford et al., 1996; Blanchette and Richards, 2004). Together, they may serve to reduce reasoning accuracy for emotional problems with belief-logic conflict, only in participants with higher emotional reactivity.

Trait anxiety, one property of emotional reactivity, was higher in SS/SLG than LALA carriers, consistent with previous reports (Lesch et al., 1996; Lonsdorf et al., 2009). Past studies using the emotional Stroop task found greater interference from threat-related words in patients with anxiety disorder (Becker et al., 2001) as well as healthy participants with high anxiety (Dresler et al., 2009). Further, belief-bias during reasoning with social-anxiety-provoking statements was higher in healthy participants with higher levels of social anxiety (Vroling and de Jong, 2009). Similarly, belief-bias for emotional problems was higher in participants with higher trait anxiety in the present study. However, multiple regression analysis revealed that 5-HTTLPR genotype accounted for the relationship between anxiety and emotional belief-bias. Indeed, mediation analysis indicated that anxiety was not a significant mediator between 5-HTTLPR and emotional belief-bias. We suggest that the S-allele leads to a pre-disposition to negative emotional reactivity, which in turn leads to higher anxiety and emotional belief-bias. In contrast, N-back working memory performance did not differ between groups and did not correlate with emotional belief-bias, indicating 5-HTTLPR effects found in the present study have some specificity, acting through emotional rather than general cognitive processing mechanisms. Thus, a more anxious temperament due to 5-HTTLPR genotype was related to reduced inhibitory control during reasoning selectively with emotional material.

The ability to make rational decisions relates to success in various aspects of contemporary society. Superior deductive reasoning ability predicts higher academic achievement. Specifically, children showing less belief-biased reasoning errors had higher math and reading performance (Handley et al., 2004). Furthermore, relational reasoning is also important for social functioning (Maclean et al., 2008). For example, if Johnny knows that his older sibling Pat is stronger than him based on previous experience and that Mark, the new kid in school, is stronger than Pat, the ability to infer that Mark is therefore stronger than Johnny without having to directly test this hypothesis can be highly beneficial. Interestingly, 5-HTTLPR genotype relates to a variety of social functions, including establishment of social dominance and aggression in animals that live in social groups (Neumann et al., 2010). The present findings show that genotypic differences in the functioning of the serotonin transporter lead some individuals to be more vulnerable to the influence of emotion and its deleterious effects on reasoning, an important ability for academic and social success.

Acknowledgments

This work was supported by a Canadian Institutes for Health Research doctoral research award to M.S., and the National Institutes of Health (MH065395-01 to C.J.V. and NCMRR/NINDS 2R24HD050846-06 and NINDS 5R01NS029525 to C.N.M.C.). We thank Rebecca Ryan for statistical assistance.

REFERENCES

  1. Armbruster D, Moser DA, Strobel A, et al. Serotonin transporter gene variation and stressful life events impact processing of fear and anxiety. International Journal of Neuropsychopharmacology. 2009;12:393–401. doi: 10.1017/S1461145708009565. [DOI] [PubMed] [Google Scholar]
  2. Aron AR, Robbins TW, Poldrack RA. Inhibition and the right inferior frontal cortex. Trends in Cognitive Sciences. 2004;8:170–7. doi: 10.1016/j.tics.2004.02.010. [DOI] [PubMed] [Google Scholar]
  3. Becker ES, Rinck M, Margraf J, Roth WT. The emotional Stroop effect in anxiety disorders: general emotional or disorder specificity? Journal of Anxiety Disorders. 2001;15:147–59. doi: 10.1016/s0887-6185(01)00055-x. [DOI] [PubMed] [Google Scholar]
  4. Beevers CG, Wells TT, Ellis AJ, McGeary JE. Association of the serotonin transporter gene promoter region (5-HTTLPR) polymorphism with biased attention for emotional stimuli. Journal of Abnormal Psychology. 2009;118:670–81. doi: 10.1037/a0016198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Blanchette I. The effect of emotion on interpretation and logic in a conditional reasoning task. Memory Cognition. 2006;34:1112–25. doi: 10.3758/bf03193257. [DOI] [PubMed] [Google Scholar]
  6. Blanchette I, Richards A. Reasoning about emotional and neutral materials. Is logic affected by emotion? Psychological Science. 2004;15:745–52. doi: 10.1111/j.0956-7976.2004.00751.x. [DOI] [PubMed] [Google Scholar]
  7. Bradley MM, Lang PJ. Technical report C-1. Gainsville, FL: The Center for Research in Psychophysiology, University of Florida; 1999. Affective norms for English words (ANEW): stimuli, instruction manual and affective ratings. [Google Scholar]
  8. Byrne RM, Tasso A. Deductive reasoning with factual, possible, and counterfactual conditionals. Memory Cognition. 1999;27:726–40. doi: 10.3758/bf03211565. [DOI] [PubMed] [Google Scholar]
  9. Caspi A, Sugden K, Moffitt TE, et al. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science. 2003;301:386–9. doi: 10.1126/science.1083968. [DOI] [PubMed] [Google Scholar]
  10. de Jong PJ, Mayer B, van den Hout M. Conditional reasoning and phobic fear: evidence for a fear-confirming reasoning pattern. Behaviour Research Therapy. 1997;35:507–16. doi: 10.1016/s0005-7967(96)00124-6. [DOI] [PubMed] [Google Scholar]
  11. De Neys W, Franssens S. Belief inhibition during thinking: not always winning but at least taking part. Cognition. 2009;113:45–61. doi: 10.1016/j.cognition.2009.07.009. [DOI] [PubMed] [Google Scholar]
  12. De Neys W, Vartanian O, Goel V. Smarter than we think: when our brains detect that we are biased. Psychological Science. 2008;19:483–9. doi: 10.1111/j.1467-9280.2008.02113.x. [DOI] [PubMed] [Google Scholar]
  13. Dresler T, Meriau K, Heekeren HR, van der Meer E. Emotional Stroop task: effect of word arousal and subject anxiety on emotional interference. Psychological Research. 2009;73:364–71. doi: 10.1007/s00426-008-0154-6. [DOI] [PubMed] [Google Scholar]
  14. Evans JS, Barston JL, Pollard P. On the conflict between logic and belief in syllogistic reasoning. Memory Cognition. 1983;11:295–306. doi: 10.3758/bf03196976. [DOI] [PubMed] [Google Scholar]
  15. Goel V, Buchel C, Frith C, Dolan RJ. Dissociation of mechanisms underlying syllogistic reasoning. Neuroimage. 2000;12:504–14. doi: 10.1006/nimg.2000.0636. [DOI] [PubMed] [Google Scholar]
  16. Goel V, Dolan RJ. Explaining modulation of reasoning by belief. Cognition. 2003;87:B11–22. doi: 10.1016/s0010-0277(02)00185-3. [DOI] [PubMed] [Google Scholar]
  17. Goel V, Vartanian O. Negative emotions can attenuate the influence of beliefs on logical reasoning. [Research Support, Non-U.S. Gov't] Cognition and Emotion. 2011;25:121–31. doi: 10.1080/02699931003593942. [DOI] [PubMed] [Google Scholar]
  18. Gonda X, Fountoulakis KN, Juhasz G, et al. Association of the s allele of the 5-HTTLPR with neuroticism-related traits and temperaments in a psychiatrically healthy population. European Archives of Psychiatry and Clinical Neuroscience. 2009;259:106–13. doi: 10.1007/s00406-008-0842-7. [DOI] [PubMed] [Google Scholar]
  19. Gong P, Zhang F, Lei X, et al. No observable relationship between the 12 genes of nervous system and reasoning skill in a young Chinese Han population. [Research Support, Non-U.S. Gov't] Cellular and Molecular Neurobiology. 2011;31:519–26. doi: 10.1007/s10571-010-9645-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Handley SJ, Capon A, Beveridge M, Dennis I, Evans JSB. Working memory, inhibitory control and the development of children's reasoning. Thinking and Reasoning. 2004;10:175–95. [Google Scholar]
  21. Houde O, Zago L, Mellet E, et al. Shifting from the perceptual brain to the logical brain: the neural impact of cognitive inhibition training. Journal of Cognitive Neuroscience. 2000;12:721–8. doi: 10.1162/089892900562525. [DOI] [PubMed] [Google Scholar]
  22. Hu XZ, Lipsky RH, Zhu G, et al. Serotonin transporter promoter gain-of-function genotypes are linked to obsessive-compulsive disorder. American Journal of Human Genetics. 2006;78:815–26. doi: 10.1086/503850. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Koizumi A, Kitagawa N, Kitamura MS, Kondo HM, Sato T, Kashino M. Serotonin transporter gene and inhibition of conflicting emotional information. NeuroReport. 2010;21:422–6. doi: 10.1097/WNR.0b013e32833833f0. [DOI] [PubMed] [Google Scholar]
  24. Lefford A. The influence of emotional subject matter on logical reasoning. Journal of General Psychology. 1946;34:127–51. doi: 10.1080/00221309.1946.10544530. [DOI] [PubMed] [Google Scholar]
  25. Lesch KP, Bengel D, Heils A, et al. Association of anxiety-related traits with a polymorphism in the serotonin transporter gene regulatory region. Science. 1996;274:1527–31. doi: 10.1126/science.274.5292.1527. [DOI] [PubMed] [Google Scholar]
  26. Lonsdorf TB, Weike AI, Nikamo P, Schalling M, Hamm AO, Ohman A. Genetic gating of human fear learning and extinction: possible implications for gene-environment interaction in anxiety disorder. Psychological Science. 2009;20:198–206. doi: 10.1111/j.1467-9280.2009.02280.x. [DOI] [PubMed] [Google Scholar]
  27. Luo J, Yuan J, Qiu J, Zhang Q, Zhong J, Huai Z. Neural correlates of the belief-bias effect in syllogistic reasoning: an event-related potential study. Neuroreport. 2008;19:1073–8. doi: 10.1097/WNR.0b013e3283052fe1. [DOI] [PubMed] [Google Scholar]
  28. Maclean EL, Merritt DJ, Brannon EM. Social Complexity Predicts Transitive Reasoning in Prosimian Primates. Animal Behaviour. 2008;76:479–86. doi: 10.1016/j.anbehav.2008.01.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Munafo MR, Brown SM, Hariri AR. Serotonin transporter (5-HTTLPR) genotype and amygdala activation: a meta-analysis. Biological Psychiatry. 2008;63:852–7. doi: 10.1016/j.biopsych.2007.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Neumann ID, Veenema AH, Beiderbeck DI. Aggression and anxiety: social context and neurobiological links. Frontiers in Behavioral Neuroscience. 2010;4:12. doi: 10.3389/fnbeh.2010.00012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Oaksford M, Morris F, Grainger B, Williams JMG. Mood, reasoning, and central executive processes. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1996;22:476–92. [Google Scholar]
  32. Prado J, Noveck IA. Overcoming perceptual features in logical reasoning: a parametric functional magnetic resonance imaging study. Journal of Cognitive Neuroscience. 2007;19:642–57. doi: 10.1162/jocn.2007.19.4.642. [DOI] [PubMed] [Google Scholar]
  33. Roberts MJ, Sykes ED. Belief bias and relational reasoning. Quarterly Journal of Experimental Psychology A. 2003;56:131–53. doi: 10.1080/02724980244000233. [DOI] [PubMed] [Google Scholar]
  34. Roiser JP, de Martino B, Tan GC, et al. A genetically mediated bias in decision making driven by failure of amygdala control. Journal of Neuroscience. 2009;29:5985–91. doi: 10.1523/JNEUROSCI.0407-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State-Trait Anxiety Inventory for Adults. Consulting Psychologists Press, Palo Alto, CA; 1983. [Google Scholar]
  36. Stollstorff M, Vartanian O, Goel V. Levels of conflict in reasoning modulate right lateral prefrontal cortex. Brain Research. 2012;1428:24–32. doi: 10.1016/j.brainres.2011.05.045. [DOI] [PubMed] [Google Scholar]
  37. Tsujii T, Masuda S, Akiyama T, Watanabe S. The role of inferior frontal cortex in belief-bias reasoning: an rTMS study. Neuropsychologia. 2010;48:2005–8. doi: 10.1016/j.neuropsychologia.2010.03.021. [DOI] [PubMed] [Google Scholar]
  38. Tsujii T, Watanabe S. Neural correlates of belief-bias reasoning under time pressure: a near-infrared spectroscopy study. Neuroimage. 2010;50:1320–6. doi: 10.1016/j.neuroimage.2010.01.026. [DOI] [PubMed] [Google Scholar]
  39. Tversky A, Kahneman D. Judgment under uncertainty: heuristics and biases. Science. 1974;185:1124–31. doi: 10.1126/science.185.4157.1124. [DOI] [PubMed] [Google Scholar]
  40. Vroling MS, de Jong PJ. Deductive reasoning and social anxiety: evidence for a fear-confirming belief bias. Cognitive Therapy Research. 2009;33:633–44. doi: 10.1007/s10608-008-9220-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Social Cognitive and Affective Neuroscience are provided here courtesy of Oxford University Press

RESOURCES