Skip to main content
NCBI home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2018 Sep 6;54(1):30–37. doi: 10.1093/alcalc/agy065

Interaction Effects between the Cumulative Genetic Score and Psychosocial Stressor on Self-Reported Drinking Urge and Implicit Attentional Bias for Alcohol: A Human Laboratory Study

Jueun Kim 1, Michael A Marciano 2, Shyanne Ninham 3, Michelle J Zaso 3, Aesoon Park 3,✉
PMCID: PMC6324655  PMID: 30192917

Individuals with high cumulative genetic risk score of the five monoamergic genotypes showed greater attentional bias toward alcohol cues when exposed to a psychosocial stressor than when not exposed.

Abstract

Aims

The current candidate gene and environment interaction (cGxE) study examined whether the effects of an experimentally manipulated psychosocial stressor on self-reported drinking urge and implicit attentional bias for alcohol cues differ as a function of a cumulative genetic score of 5-HTTLPR, MAO-A, DRD4, DAT1 and DRD2 genotypes. The current study also examined whether salivary alpha-amylase level or self-reported anxiety state mediate these cGxE effects.

Short Summary

Individuals with high cumulative genetic risk score of the five monoamergic genotypes showed greater attentional bias toward alcohol cues when exposed to a psychosocial stressor than when not exposed.

Methods

Frequent binge-drinking Caucasian young adults (N = 105; mean age = 19; 61% male) completed both the control condition and stress condition (using the Trier Social Stress Test) in order.

Results

Regarding attentional bias, individuals with high and medium cumulative genetic risk scores showed greater attentional bias toward alcohol stimuli in the stress condition than in the control condition, whereas, those with low genetic risk scores showed greater attentional bias toward alcohol stimuli in the control condition than in the stress condition. No mediating roles of salivary alpha-amylase and anxiety state in the cGxE effect were found. Regarding self-reported drinking urge, individuals with high cumulative genetic score reported greater drinking urge than those with low genetic score regardless of experimental conditions.

Conclusions

Although replication is necessary, the findings suggest that the association of a psychosocial stressor on implicit (but not explicit, self-reported) alcohol outcomes may differ as a function of the collective effects of five monoamine genes.

Exposure to stressors has been consistently associated with a greater risk of alcohol misuse (Dawson et al., 2005) and higher rate of alcoholism (Fox et al., 2007). Stress response dampening theory (Sher, 1987) maintains that individuals in stressful situations drink alcohol to reduce their stress, and their drinking behavior is reinforced by its short-term stress-reducing effects after repeated exposures to stressful situations. Among young adults, drinking to cope with stress has been associated with greater drinking problems as compared to other reasons to drink such as mood enhancement and social facilitation (for a review, see Kuntsche et al., 2005). However, the role of stress on young adult drinking was shown to differ as a function of individual and environmental factors (Rutledge and Sher, 2001), and the potential moderating role of genetic factors has been suggested (Enoch, 2006).

The diathesis stress model (Zuckerman, 1999) suggests that some individuals have a predispositional vulnerability (e.g. temperament, genotypes) to stressful environments, and exposure to stressors triggers such underlying vulnerability leading to engagement in maladaptive behaviors. Recently, the differential susceptibility hypothesis (Belsky and Pluess, 2009) has suggested that individuals with certain genotypes may not only be more vulnerable to adverse environments but also more susceptible to the beneficial effects of protective environments (or abstinence/lack of adverse environments). That is, some genes may be ‘plasticity genes’ that are associated with sensitivity to both positive and negative environmental exposures rather than ‘risky genes’ as pathogens of alcohol misuse. Gene and environment interaction studies (GxE) have reported that associations of the environment with drinking behavior can depend upon an individual’s genotype, with an individual’s genotype strengthening or weakening their susceptibility to environmental influences (Caspi and Moffitt, 2006).

A growing literature highlights the important role of monoamine genes in the production, secretion, and regulation of dopamine, serotonin, and norepinephrine in the brain and peripheral nervous system. Monoamine neurotransmitters have been shown to play a role in alcohol appetite, alcohol withdrawal symptoms, and development of tolerance in animal and human studies (for a review, see Nutt and Glue, 1986). However, previous candidate GxE (cGxE) studies on associations of a single monoamine gene variant (e.g. 5-HTTLPR, DRD4, DAT1, DRD2 and MAO-A genotypes) with young adult alcohol outcomes have shown mixed findings (for a review of cGxE studies across development, see Kim et al., under review). For example, 5-Hydroxy Tryptamine Transporter Linked Promoter Region (5-HTTLPR) moderated the effect of stressful life events on binge-drinking among Caucasian college students (Covault et al., 2007), although another study of African-American college students found such moderation only in women (Kranzler et al., 2012). There is initial evidence for significant moderation by Dopamine D2 receptor (DRD2, Madrid et al., 2001) and Dopamine Transporter gene (DAT1, Stogner, 2015) in the effect of stressful life events on drinking outcomes, but it was not on young adult drinking and has not been replicated in other samples. Two studies (Nilsson et al., 2007, 2008) found that Monoamine Oxidase A (MAO-A) moderated the effects of adverse environments on negative drinking consequences or alcohol use disorder symptoms in young adulthood, but both studies had small sample sizes (n = 66 and n = 114, respectively). Two prospective studies from college to young adulthood (Park et al., 2011) reported inconsistent findings regarding the moderating role of Dopamine D4 receptor (DRD4) in the effect of childhood adversity on alcohol dependence symptoms.

These mixed findings in cGxE studies may be due in part to the fact that the effect of a single genetic variant on complex behavior such as alcohol use and misuse is most likely small. Examining the cumulative genetic effect of multiple genetic variants can account for a number of genetic variants with small effect sizes that co-exist in individuals and jointly influence alcohol outcomes. Thus, a cumulative genetic score approach affords increased power and allows us to identify a multi-locus genetic profile that interacts with environmental factors to influence behaviors (Dick et al., 2015). To our best knowledge, only one observational study has examined and found a significant interaction effect of a cumulative monoaminergic genetic score (consisting of 5-HTTLPR, DRD4, DAT1, DRD2 and MAO-A) with stressful environments on adolescent drinking (Stogner and Gibson, 2016).

Results of observational cGxE studies also are possibly confounded by gene and environment correlation, because individuals carrying certain genotypes may evoke or seek certain environments that are compatible with their genetic propensity (Plomin et al., 1977). Experimental study designs can better resolve these potential confounding effects of gene and environment correlation by assigning environmental conditions independent of a participant’s genotype. Also, experimental studies are relatively free from self-report bias and environmental confounds, because the stress experiment is operationally defined and manipulated and the alcohol-related outcome is directly observed when other factors are controlled. However, no prior research with a cumulative genetic score of 5-HTTLPR, DRD4, DAT1, DRD2 and MAO-A on alcohol outcomes has used an experimental study design.

Once the interaction between a cumulative genetic score of monoaminergic genes and stressful environmental exposure on alcohol outcomes is identified, potential biological and psychological mechanisms need to be examined. The Sympathetic Adrenal Medullary axis reactivity measured by salivary alpha-amylase level is a promising biological mediator of the monoaminergic genes and stressor interaction effects. Monoaminergic genes, including serotonin genetic variants have been found to influence salivary alpha-amylase response to stressors (Mueller et al., 2012), and elevated salivary alpha-amylase has been associated with increases in substance seeking behaviors (Sinha et al., 2003; Duskova et al., 2010). In addition, anxiety state may serve as a psychological mediator of the monoaminergic genes and stressor interaction effects. Positive associations of anxiety with alcohol use disorder have been reported (for a review, see Zuckerman, 1999), and several studies have found significant associations of monoaminergic genes with anxiety state in response to stressors (Brummett et al., 2008).

Using a within-subject experimental study design, this current study aimed to (a) examine whether a cumulative genetic score of 5-HTTLPR, DRD4, DAT1, DRD2 and MAO-A genotypes moderates the effects of psychosocial stressors on self-reported drinking urge and implicit attentional bias for alcohol-related stimuli and (b) investigate the mediating roles of salivary alpha-amylase and anxiety state responses in any cGxE effects. It was hypothesized that the effects of stressors on drinking urge and attentional bias for alcohol would increase as individuals carried more risk conferring alleles of the monoaminergic genes (i.e. as their cumulative genetic score increases). It was also hypothesized that individuals with a higher cumulative genetic score would show higher levels of salivary alpha-amylase and anxiety in response to stressors, which in turn would be associated with elevated drinking urge and attentional bias for alcohol-related stimuli.

METHOD

Participants

Participants were 105 Caucasian frequent binge drinkers (mean age = 19 [SD = 1.54]; 61% men) recruited from a mid-sized northeastern community. Only Caucasians were recruited to minimize confounding effects of population stratification (i.e. differences in allele or genotype frequencies across racial groups (Hu et al., 2006)), which may confound the results of genetic studies (called ‘population stratification’, Pritchard and Rosenberg, 1999). Frequent binge-drinking was defined as drinking five or more alcoholic drinks for men and four or more alcoholic drinks for women on three or more occasions within the past two weeks, which has been used to screen for high-risk drinking among a young adult population (Knight et al., 2002). Participants were recruited through an undergraduate research participation pool, flyers, classroom/email solicitations and community online advertisements. Exclusion criteria is presented in supplementary document.

Procedures

Interested participants were pre-screened for eligibility and all experimental sessions were scheduled at 5 pm and lasted until 8:30 pm to approximate the time of natural drinking episodes. Upon arrival at the lab, participants completed a baseline questionnaire and donated a saliva sample for genotyping. Then, all participants completed a control condition followed by a stress condition using the Trier Social Stress Test (Birkett, 2011); all participants completed the control condition before the stress condition, weighing the benefits of minimizing carryover effect over the limitations of order effect of experimental conditions. Detailed experimental procedures are described in supplementary document.

Measures

Genotypes

Participants’ DNA for polymerase chain reaction (PCR) was extracted from saliva samples and genotyped through allele-specific fluorescence using the Thermo Fisher TaqMan genotyping assay in an Applied Biosystems 7500 real time PCR instrument. All 105 participants’ genotypes were analyzed except for two participants’ 5-HTTLPR and MAO-A genotypes, which could not be genotyped due to low DNA concentration (2% indeterminate genotypes). Hardy–Weinberg equilibrium was investigated using Fisher’s exact test (Wigginton et al., 2005). Genotype frequencies of 5-HTTLPR, DAT1, MAO-A, DRD4 and DRD2 were in Hardy–Weinberg equilibrium (P’s > 0.05). Detailed genotype frequencies and the method of generating cumulative genetic risk score are described in supplementary document.

Alcohol outcomes

Alcohol urge questionnaire

Drinking urge was assessed after the control and stress conditions using the eight-item Alcohol Urge Questionnaire (Bohn et al., 1995). High test–retest reliability and high convergent validity with the Severity of Alcohol Dependence Questionnaire (Bohn et al., 1995; Drummond and Phillips, 2002) have been reported. Participants will respond to each item based on a seven-point response scale, with responses from 1 (Strongly Disagree) to 7 (Strongly Agree). For the current analyses, mean scores after the control and stress conditions were used as dependent variables.

Visual probe task. The visual probe task is a well-established protocol to investigate participants’ attentional bias towards a substance related stimuli (Ehrman et al., 2002). Eighty-four pairs of alcohol-related pictures (e.g. glass of beer) and neutral pictures (e.g. a chair) were shown. A fixation point (x) was shown for 500 ms in the center of the screen, followed by a pair of alcohol-related and neutral pictures in which one is shown on the left side and another shown on the right side for 1000 ms. When pictures disappeared from the screen, the fixation point (x) appeared on the left or right side, and participants were asked to answer which picture was located on that side. For the current analyses, participants’ reaction time ratios of alcohol-related pictures to neutral pictures after the control and stress conditions were used as dependent variables.

Manipulation check, mediator and confounding variable measures are described in supplementary document.

Data analyses

Interactions between cumulative genetic score and stressor

In order to examine the interaction effect of cumulative genetic score and psychosocial stressor on drinking urge, a two-way mixed ANOVA was conducted in SPSS. Non-normally distributed Alcohol Urge Questionnaire residuals were transformed to a normal distribution using a rank transformation method (Solomon and Sawilowsky, 2009). Two sets of mixed ANOVA models were estimated: the main effect of cumulative genetic score (cG), main effect of experimental stress versus control condition (E) and interaction effect of the grand-mean centered cumulative genetic score with experimental condition (cGxE) were examined on drinking urge and the visual probe task score in separate models after controlling for the confounding effect of sex. When cGxE effects were found to be significant, a post hoc paired samples t-test was separately conducted across cumulative genetic score groups.

Mediating effects of anxiety state and salivary alpha-amylase

Two models were estimated to separately examine potential mediators: changes in salivary alpha-amylase and anxiety state between the control and stress conditions. The SPSS macro PROCESS was used to test for significant mediated moderation (Hayes, 2013). Estimates of the mediated effects and their 95% confidence intervals were obtained by bootstrap analysis with 5000 bootstrap samples. The 95% confidence interval (CI) of the mediated effect that does not include zero indicates a significant mediation effect.

Information on descriptive analysis, manipulation check and power analysis methods is provided in supplementary document.

RESULTS

Descriptive statistics

Baseline differences as a function of the cumulative genetic score (CGS) are presented in Table 1. There were no baseline differences on sociodemographic, stressful life events, and binge-drinking frequency as a function of CGS except for social desirability, F(2,102) = 4.22, P = 0.02, np2= 0.08. Regarding social desirability, Tukey’s HSD indicated that participants with high CGS (i.e. four or five risk alleles) reported significantly higher social desirability than those with low CGS (i.e. one or two risk alleles), P = 0.04.

Table 1.

Means (and standard deviations) of all participants and baseline differences as in three CGS groups

Variable (possible range) Low CGS group (n = 45) Middle CGS group (n = 41) High CGS group (n = 19) Test statistics comparing three CGS groups
Male sex 64% 61% 53% χ2(2) = 0.78
Age 19.78 (1.33) 19.95 (1.89) 19.68 (1.20) F(2,101) = 0.23
Stressful life events in the last year (0–36) 3.04 (2.26) 2.68 (2.14) 2.74 (2.23) F(2,101) = 0.33
Social desirability (0–33) 15.06 (4.69) 17.22 (4.23) 18.00 (3.37) F(2,102) = 4.22*
Past-90-days binge-drinking frequency (0–90) 20.42 (11.00) 19.90 (10.48) 15.79 (8.13) F(2,101) = 1.43
Open in a new tab

Note. CGS, cumulative genetic risk score *P < 0.05.

Descriptive statistics of all participants and bivariate correlation coefficients of study variables are presented in Table 2. CGS was positively associated with social desirability, r = 0.27, P = 0.01, which was in line with above mentioned social desirability difference as a function of CGS.

Table 2.

Means (and standard deviations) or percentages and bivariate correlation coefficients of study variables

r
Variable (possible range) M (SD) 1 2 3 4 5 6 7 8 9
1. Male sex 61% –
2. Age 19.83 (1.54) 0.05 –
3. Cumulative Genetic Score (0–2) 0.75 (0.74) −0.08 −0.004 –
4. Stressful life events (0–36) 2.85 (2.19) 0.15 0.01 −0.07 –
5. Social desirability (0–33) 16.44 (4.43) −0.09 0.20* 0.27** −0.08 –
6. Binge-drinking frequency (0–90) 19.38 (10.38) 0.05 −0.11 −0.14 −0.03 −0.06 –
7. Alcohol urge – control condition (1–7) 3.41 (1.60) −0.13 0.09 0.19* −0.19 0.22* 0.02 –
8. Alcohol urge – stress condition (1–7) 3.44 (1.67) −0.16 0.12 0.15 −0.13 0.15 0.05 0.82** –
9. Attentional bias – control condition 1.00 (17.13) −0.02 0.03 −0.23* −0.05 −0.14 0.06 −0.11 −0.04 –
10. Attentional bias – stress condition −0.39 (18.43) 0.07 0.05 0.13 −0.03 −0.003 0.003 0.04 −0.01 −0.07
Open in a new tab

Note. N = 105. For correlation with sex, Spearman correlation coefficients are presented; for all other variables, Pearson correlation coefficients are presented.

*P < 0.05, **P < 0.01.

Manipulation check

As shown in Figure 1, changes in anxiety state, salivary alpha-amylase and heart rate responses between the control and stress conditions showed that stress was successfully manipulated. Regarding anxiety state, a significant effect of experimental conditions was found, F(1,98) = 28.93, P < 0.001. A post hoc paired samples t-test indicated that anxiety state was significantly higher after the stress condition than the control condition, t(100) = 7.67, P < 0.001. The same pattern was found in salivary alpha-amylase, F(1,92) = 33.84, P < 0.001, and heart rate, F(1,89) = 6.03, P = 0.02. Amylase level was significantly higher after the stress condition than the control condition, t(96) = 7.19, P < 0.001. Further, heart rate was significantly higher in the stress condition than in the control condition, t(96) = 5.80−9.78, P’s < 0.001. Finally, in the manipulation check questionnaire, 79% of participants endorsed that they were concerned/nervous in the stress condition, whereas only 8% of participants endorsed that they were concerned/nervous in the control condition.

Figure 1.

Figure 1.

Open in a new tab

Observed mean levels (and standard error bars) of anxiety state, salivary alpha-amylase and heart rate responses throughout the experimental procedures.

Cumulative genetic score x experimental condition on alcohol outcomes

For Alcohol Urge, as shown in Table 3, no significant interaction effects of CGS with experimental condition were found on self-reported drinking urge, F(2,99) = 0.42, P = 0.66, np2 = 0.00, after controlling for sex. However, a significant main effect of CGS was found on self-reported drinking urge regardless of experimental condition, F(2,99) = 3.76, P = 0.03, np2 = 0.07. Post hoc pairwise comparisons of estimated marginal means (adjusted for sex) demonstrated that participants with high CGS reported a higher drinking urge than those with medium CGS, P < 0.001, or low CGS, P = 0.02 (independent of experimental conditions).

Table 3.

Interaction effect of CGS and experimental condition on alcohol outcomes

Variables Low CGS group Middle CGS group High CGS group Main effect of CGS Main effect of experimental condition CGS x condition interaction effect
M (SD) M (SD) M (SD) F statistics np2 F statistics np2 F statistics np2
Alcohol urgea
 Control condition 3.25 (1.47) 3.19 (1.54) 4.26 (1.39) F(2,99) 0.07 F(1,99) 0.004 F(2,99) <0.001
 Stress condition 3.40 (1.74) 3.14 (1.51) 4.32 (1.46) =3.76* =0.40 =0.42
Attentional bias
 Control condition 5.19 (17.70) −1.07 (15.24) −4.73 (18.00) F(2,97) 0.001 F(1,97) <0.001 F(2,97) 0.08
 Stress condition −3.55 (14.79) 1.09 (18.68) 5.91 (22.24) =0.05 =0.00 =4.44*
Open in a new tab

Note. aA rank-based normal transformation was applied to alcohol urge, but not for attentional bias because attentional bias was normally distributed; sex and the interaction of sex with experimental condition were included as covariates in all analyses. *P < 0.05.

For attention bias for alcohol cues (as assessed by the Visual Probe Task), as presented in Figure 2, a significant interaction of CGS with experimental condition was found, F(2,97) = 4.44, P = 0.01, np2 = 0.08, after controlling for sex. Post hoc paired samples t-test showed that individuals with high CGS showed a significantly higher visual probe task score in the stress condition compared to the control condition, t(17) = 38.29, P < 0.001. Individuals with medium CGS also showed a significantly higher visual probe task score in the stress condition compared to the control condition, t(38) = 11.90, P < 0.001. On the contrary, individuals with low CGS showed a significantly lower visual probe task score in the stress condition than the control condition, t(43) = −51.70, P < 0.001. Also, the interaction effect size between CGS and experimental condition was larger (ηp2 = 0.08) than the effect sizes of separate cGxE interaction terms calculated for each variant (ηp2 = 0.00−0.06). This result suggested that the CGS provided added utility over examining any of the individual candidate genes in isolation.

Figure 2.

Figure 2.

Open in a new tab

Predicted means (and standard error bars) of visual probe task scores for implicit attention bias toward alcohol stimuli among carriers of low (one or two risk alleles), middle (three risk alleles) and high (four or five alleles) cumulative genetic groups in the control and stress experimental conditions.

Mediating roles of alpha-amylase and anxiety state

No significant mediation effect of anxiety state was indicated, b = −0.70, β = −0.02, 95% bootstrapped CI [−2.67, 0.08], SE = 0.64, P = 0.39. Specifically, the indirect path from CGS to anxiety state change between control and stress conditions (b = −1.72, β = −0.16, P = 0.12), as well as the indirect path from anxiety state to Visual Probe Task change between control and stress conditions (b = 0.41, β = 0.12, P = 0.20) were not significant, after controlling for sex. However, the direct path of CGS on Visual Probe Task change was significant, b = 10.84, β = 0.31, P = 0.002, after accounting for the indirect and sex effects.

No significant mediation effect of alpha-amylase was indicated, b = −0.34, β = −0.01, 95% bootstrapped CI [−3.23, 0.45], SE = 0.73, P = 0.60. Specifically, the indirect path from CGS to the alpha-amylase change between control and stress conditions (b = −6.19, β = −0.13, P = 0.21), as well as the indirect path from alpha-amylase to Visual Probe Task change between control and stress conditions (b = 0.05, β = 0.07, P = 0.48) were not significant, after controlling for sex. However, the direct effect of CGS on Visual Probe Task change was significant, b = 11.02, β = 0.32, P = 0.003, after accounting for the indirect and sex effects.

Ancillary analyses

Ancillary analysis results about the effect of potential confounding factors including baseline binge-drinking frequency and life stress events and the role of social desirability are presented in supplementary document.

DISCUSSION

The current study extended cGxE literature on alcohol phenotypes by examining the interaction effects of a cumulative genetic score of five monoaminergic genotypes (5-HTTLPR, MAOA-A, DRD4, DAT1 and DRD2) with psychosocial stressors on self-reported drinking urge and attentional bias towards alcohol stimuli among frequent heavy drinkers. This study also extended previous observational cGxE studies by examining the effect of experimentally manipulated acute stressors in a cGxE context. The results indicate that individuals with high cumulative genetic risk score of the five monoamergic genotypes showed greater attentional bias toward alcohol cues when exposed to a psychosocial stressor than when not exposed. In contrast, those with low cumulative genetic risk rather showed lower attentional bias toward alcohol cues when exposed to a psychosocial stressor than when not exposed. However, this significant cGxE effect was not found on self-reported drinking urge. Instead, individuals with a high cumulative genetic score reported greater drinking urge regardless of experimental condition.

Our finding on attentional bias toward alcohol stimuli showed a cross-over interaction in which cumulative genetic score was associated with greater attentional bias for alcohol stimuli in the stress condition but lower attentional bias for alcohol stimuli in the control condition. These cross-over patterns were observed in two previous large cGxE studies (n = 1586, Belsky and Beaver, 2011; n = 1495, Stogner and Gibson, 2016) testing cumulative effects of the same five monoamineric genes examined in the current study. This cross-over pattern of interaction supports a differential susceptibility hypothesis. That is, the current study findings indicate that individuals with a high cumulative genetic score of the five monoaminergic genes may be genetically plastic individuals (rather than genetically vulnerable individuals) who show worse outcomes in adverse environments but better outcomes in the abstinence of adverse environments than non-plastic individuals. Emerging neurobiological studies (For a review, see Moore and Depue, 2016) have demonstrated that dopamine related genetic variants (including DRD4, DRD2, DAT1) have shown keen reactivity to rewarding cues through tonic enabling activity in postsynaptic neurons, but also keen reactivity to aversive cues through Corticotropin-releasing hormone and glucocorticoids. Also, serotonin related genetic variants (including 5-HTTLPR, MAO-A) have shown sensitivity reactivity to both rewarding and aversive environments by being involved in neural activity in limbic and cortical circuitries. However, cross-over patterns of interaction may be an artifact due to a low power and small sample size (Sher and Steinley, 2013; Boardman et al., 2014), and thus the current study findings await replication in large, independent samples.

Different from attentional bias, the cumulative genetic effect of five monoaminergic genotypes on self-reported drinking urge did not differ as a function of manipulated psychosocial stressors. This is, a greater level of cumulative genetic score was positively associated with drinking urge regardless of experimental condition. This finding is in part in line with prior association studies demonstrating significant main effects of monoaminergic genotypes such as DRD4, DRD2, DAT on drinking urge (Hutchison et al., 2002, 2003). However, non-significant moderating effects of the experimentally manipulated psychosocial stressor on drinking urge may have been due to the high social desirability reported among individuals with high cumulative genetic score (as shown in Descriptive Statistics and Ancillary Analyses). That is, individuals with high cumulative genetic risk may have reported elevated drinking urges across the control and stress conditions based on their belief that such responses would be viewed more favorably by the research team.

This study did not find significant mediating roles of anxiety state or salivary alpha-amylase in the interaction effects of cumulative genetic score with psychosocial stressor on attentional bias for alcohol stimuli. The non-significant mediating role of anxiety state is surprising given considerable evidence of co-occurring problematic alcohol use and internalizing symptoms (especially depression and anxiety) among monoaminergic plasticity genotype carriers (for a review, see Saraceno et al., 2009). The non-significant mediation of salivary alpha-amylase may be because that amylase response has been confounded due to other non-stress related stimuli during experiment. Although salivary alpha-amylase is widely accepted as a physical stress measure (Nater et al., 2006; Het et al., 2009; Maruyama et al., 2012), there is also a concern that amylase is reflective of sympathetic nervous system response, which has a broader set of activating stimuli. Amylase is known to be more active than the hypothalamic-pituitary-adrenal axis response measured by cortisol which needs strong social evaluative stress to be activated (Buss et al., 2004). Unfortunately, cortisol response was not analyzed in this current study, and thus future studies need to examine a potential mediating role of cortisol levels.

Findings should be interpreted within the context of several limitations. First, although the genotypes included in this study were selected based on previous cGxE study findings, these genes are not the only risk conferring genes for alcohol outcomes. Thus, genome-wide GxE studies using a theoretical genetic variant finding approach need to determine whether these five monoaminergic genes remain significantly associated with alcohol outcomes in stressful environments within the context of other genes. Recent genome-wide association studies have not demonstrated significant associations of the monoaminergic genes examined in the current study with alcoholism (Edenberg et al., 2010; Wang et al., 2013; Gelernter et al., 2014). However, because these monoaminergic genes have not been rarely studied in genome GxE studies, future research is warranted. Second, there is a possibility that our small sample size may have resulted in false positive or negative findings. Particularly, there has been concerns about a high potential of false positives in cGxE studies with small samples (Hewitt, 2012). Thus, the current findings need to be replicated across independent and large samples. Third, potential limitations in generalization arise from the nature of this study’s experimental design in a controlled situation. In order to exclude potential confounding factors on stress response, young adults who were regularly using psychoactive drugs or smoking cigarettes were excluded at prescreening. Also, to reduce possible exacerbation of alcohol problems, individuals who have ever received treatment for alcohol-related problems were excluded from this study where alcohol cues are repeatedly presented. Thus, this current finding may not be generalized to binge drinkers who are also regular drug users/cigarette smokers and alcohol treatment seeking young adults. Finally, self-reported ancestry was used to determine participants’ race. Self-report assessment has been shown to be a reasonable measurement for researchers to use to control for the potential threat of population stratification (Hutchison et al., 2004). However, self-reported ancestry may not be as accurate as genetic ancestry inferred from Ancestry Informative Markers (AIMS), and thus future studies may benefit from using the precise characterization of individuals’ biological ancestry.

Despite the limitations, the current study findings have potential clinical implications for prevention and intervention efforts to curtail alcohol misuse among young adults. cGxE findings in general allow us to identify ‘high-risk’ groups of individuals who are more vulnerable to certain environmental effects based upon their genotypes. Although individuals’ genotypes cannot be changed, cGxE findings can help us to design targeted prevention or intervention strategies for a population at risk of drinking problems. Specifically, attentional bias modification training (which is designed to train individuals to disengage attention from alcohol-related stimuli) has been found to be effective in reducing alcohol misuse (Fadardi and Cox, 2009) and maintaining a longer abstinence period (Schoenmakers et al., 2010). Given the current study’s finding regarding attentional bias, individuals with a high cumulative genetic score of these five monoaminergic genes who are also exposed to stressful environments may benefit from attentional bias modification training.

Supplementary Material

Supplementary Data
Click here for additional data file. (40.2KB, docx)

FUNDING

This work was supported in part by an National Institutes of Health grant R15 AA022496 to Aesoon Park.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interests to disclose.

REFERENCES

  1. Belsky J, Beaver KM (2011) Cumulative-genetic plasticity, parenting and adolescent self-regulation. J Child Psychol Psychiatry 52:619–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Belsky J, Pluess M (2009) Beyond diathesis stress: differential susceptibility to environmental influences. Psychol Bull 135:885–908. [DOI] [PubMed] [Google Scholar]
  3. Birkett MA. (2011) The Trier Social Stress Test protocol for inducing psychological stress. J Vis Exp 19:3238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boardman JD, Domingue BW, Blalock CL, et al. (2014) Is the gene-environment interaction paradigm relevant to genome-wide studies? The case of education and body mass index. Demography 51:119–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bohn MJ, Krahn DD, Staehler BA (1995) Development and initial validation of a measure of drinking urges in abstinent alcoholics. Alcohol Clin Exp Res 19:600–6. [DOI] [PubMed] [Google Scholar]
  6. Brummett BH, Boyle SH, Siegler IC, et al. (2008) Effects of environmental stress and gender on associations among symptoms of depression and the serotonin transporter gene linked polymorphic region (5-HTTLPR). Behav Genet 38:34–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Buss KA, Davidson RJ, Kalin NH, et al. (2004) Context-specific freezing and associated physiological reactivity as a dysregulated fear response. Dev Psychol 40:583–94. [DOI] [PubMed] [Google Scholar]
  8. Caspi A, Moffitt TE (2006) Gene-environment interactions in psychiatry: joining forces with neuroscience. Nat Rev Neurosci 7:583–90. [DOI] [PubMed] [Google Scholar]
  9. Covault J, Tennen H, Armeli S, et al. (2007) Interactive effects of the serotonin transporter 5-HTTLPR polymorphism and stressful life events on college student drinking and drug use. Biol Psychiatry 61:609–16. [DOI] [PubMed] [Google Scholar]
  10. Dawson DA, Grant BF, Ruan WJ (2005) The association between stress and drinking: modifying effects of gender and vulnerability. Alcohol Alcohol 40:453–60. [DOI] [PubMed] [Google Scholar]
  11. Dick DM, Agrawal A, Keller MC, et al. (2015) Candidate gene-environment interaction research: reflections and recommendations. Perspect Psychol Sci 10:37–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Drummond DC, Phillips TS (2002) Alcohol urges in alcohol-dependent drinkers: further validation of the Alcohol Urge Questionnaire in an untreated community clinical population. Addiction 97:1465–72. [DOI] [PubMed] [Google Scholar]
  13. Duskova M, Simunkova K, Hill M, et al. (2010) Higher levels of salivary alpha-amylase predict failure of cessation efforts in male smokers. Physiol Res 59:765–71. [DOI] [PubMed] [Google Scholar]
  14. Edenberg HJ, Koller DL, Xuei X, et al. (2010) Genome-wide association study of alcohol dependence implicates a region on chromosome 11. Alcohol Clin Exp Res 34:840–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ehrman RN, Robbins SJ, Bromwell MA, et al. (2002) Comparing attentional bias to smoking cues in current smokers, former smokers, and non-smokers using a dot-probe task. Drug Alcohol Depend 67:185–91. [DOI] [PubMed] [Google Scholar]
  16. Enoch MA. (2006) Genetic and environmental influences on the development of alcoholism: resilience vs. risk. Ann N Y Acad Sci 1094:193–201. [DOI] [PubMed] [Google Scholar]
  17. Fadardi JS, Cox WM (2009) Reversing the sequence: reducing alcohol consumption by overcoming alcohol attentional bias. Drug Alcohol Depend 101:137–45. [DOI] [PubMed] [Google Scholar]
  18. Fox HC, Bergquist KL, Hong KI, et al. (2007) Stress-induced and alcohol cue-induced craving in recently abstinent alcohol-dependent individuals. Alcohol Clin Exp Res 31:395–403. [DOI] [PubMed] [Google Scholar]
  19. Gelernter J, Kranzler HR, Sherva R, et al. (2014) Genome-wide association study of alcohol dependence:significant findings in African- and European-Americans including novel risk loci. Mol Psychiatry 19:41–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hayes AF. (2013) ) Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York, NY: Guilford Press. [Google Scholar]
  21. Het S, Rohleder N, Schoofs D, et al. (2009) Neuroendocrine and psychometric evaluation of a placebo version of the ‘Trier Social Stress Test’. Psychoneuroendocrinology 34:1075–86. [DOI] [PubMed] [Google Scholar]
  22. Hewitt JK. (2012) Editorial policy on candidate gene association and candidate gene-by-environment interaction studies of complex traits. Behav Genet 42:1–2. [DOI] [PubMed] [Google Scholar]
  23. Hu XZ, Lipsky RH, Zhu G, et al. (2006) Serotonin transporter promoter gain-of-function genotypes are linked to obsessive-compulsive disorder. Am J Hum Genet 78:815–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hutchison KE, McGeary J, Smolen A, et al. (2002) The DRD4 VNTR polymorphism moderates craving after alcohol consumption. Health Psychol 21:139–46. [PubMed] [Google Scholar]
  25. Hutchison KE, Stallings M, McGeary J, et al. (2004) Population stratification in the candidate gene study: fatal threat or red herring? Psychol Bull 130:66–79. [DOI] [PubMed] [Google Scholar]
  26. Hutchison KE, Wooden A, Swift RM, et al. (2003) Olanzapine reduces craving for alcohol: a DRD4 VNTR polymorphism by pharmacotherapy interaction. Neuropsychopharmacology 28:1882–8. [DOI] [PubMed] [Google Scholar]
  27. Knight JR, Wechsler H, Kuo M, et al. (2002) Alcohol abuse and dependence among U.S. college students. J Stud Alcohol 63:263–70. [DOI] [PubMed] [Google Scholar]
  28. Kranzler HR, Scott D, Tennen H, et al. (2012) The 5-HTTLPR polymorphism moderates the effect of stressful life events on drinking behavior in college students of African descent. Am J Med Genet B Neuropsychiatr Genet 159B:484–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kuntsche E, Knibbe R, Gmel G, et al. (2005) Why do young people drink? A review of drinking motives. Clin Psychol Rev 25:841–61. [DOI] [PubMed] [Google Scholar]
  30. Madrid GA, MacMurray J, Lee JW, et al. (2001) Stress as a mediating factor in the association between the DRD2 TaqI polymorphism and alcoholism. Alcohol 23:117–22. [DOI] [PubMed] [Google Scholar]
  31. Maruyama Y, Kawano A, Okamoto S, et al. (2012) Differences in salivary alpha-amylase and cortisol responsiveness following exposure to electrical stimulation versus the Trier Social Stress Tests. PLoS One 7:e39375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Moore SR, Depue RA (2016) Neurobehavioral foundation of environmental reactivity. Psychol Bull 142:107–64. [DOI] [PubMed] [Google Scholar]
  33. Mueller A, Strahler J, Armbruster D, et al. (2012) Genetic contributions to acute autonomic stress responsiveness in children. Int J Psychophysiol 83:302–8. [DOI] [PubMed] [Google Scholar]
  34. Nater UM, La Marca R, Florin L, et al. (2006) Stress-induced changes in human salivary alpha-amylase activity -- associations with adrenergic activity. Psychoneuroendocrinology 31:49–58. [DOI] [PubMed] [Google Scholar]
  35. Nilsson KW, Sjoberg RL, Wargelius HL, et al. (2007) The monoamine oxidase A (MAO-A) gene, family function and maltreatment as predictors of destructive behaviour during male adolescent alcohol consumption. Addiction 102:389–98. [DOI] [PubMed] [Google Scholar]
  36. Nilsson KW, Wargelius HL, Sjoberg RL, et al. (2008) The MAO-A gene, platelet MAO-B activity and psychosocial environment in adolescent female alcohol-related problem behaviour. Drug Alcohol Depend 93:51–62. [DOI] [PubMed] [Google Scholar]
  37. Nutt D, Glue P (1986) Monoamines and alcohol. Br J Addict 81:327–38. [DOI] [PubMed] [Google Scholar]
  38. Park A, Sher KJ, Todorov AA, et al. (2011) Interaction between the DRD4 VNTR polymorphism and proximal and distal environments in alcohol dependence during emerging and young adulthood. J Abnorm Psychol 120:585–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Plomin R, DeFries JC, Loehlin JC (1977) Genotype-environment interaction and correlation in the analysis of human behavior. Psychol Bull 84:309–22. [PubMed] [Google Scholar]
  40. Pritchard JK, Rosenberg NA (1999) Use of unlinked genetic markers to detect population stratification in association studies. Am J Hum Genet 65:220–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rutledge PC, Sher KJ (2001) Heavy drinking from the freshman year into early young adulthood: the roles of stress, tension-reduction drinking motives, gender and personality. J Stud Alcohol 62:457–66. [DOI] [PubMed] [Google Scholar]
  42. Saraceno L, Munafo M, Heron J, et al. (2009) Genetic and non-genetic influences on the development of co-occurring alcohol problem use and internalizing symptomatology in adolescence: a review. Addiction 104:1100–21. [DOI] [PubMed] [Google Scholar]
  43. Schoenmakers TM, de Bruin M, Lux IF, et al. (2010) Clinical effectiveness of attentional bias modification training in abstinent alcoholic patients. Drug Alcohol Depend 109:30–6. [DOI] [PubMed] [Google Scholar]
  44. Sher KJ. (1987) Stress response dampening In Blane HT, Leonard KS (eds). Psychological theories of drinking and alcoholism. New York, NY: Guilford, 227–71. [Google Scholar]
  45. Sher KJ, Steinley DS (2013) Some issues surrounding interactions., Paper presented at the NIAAA workshop on gene-environment itneractions. Rockville, MD.
  46. Sinha R, Talih M, Malison R, et al. (2003) Hypothalamic–pituitary–adrenal axis and sympatho-adreno-medullary responses during stress-induced and drug cue-induced cocaine craving states. Psychopharmacology (Berl) 170:62–72. [DOI] [PubMed] [Google Scholar]
  47. Solomon SR, Sawilowsky SS (2009) Impact of rank-based normlizing transformations on the accuracy of test scores. Journal of Modern Applied Statistical Methods 8:448–62. [Google Scholar]
  48. Stogner JM. (2015) DAT1 and alcohol use: differential responses to life stress during adolescence. Crim Just Stud 28:18–38. [PMC free article] [PubMed] [Google Scholar]
  49. Stogner JM, Gibson CL (2016) Genetic modification of the relationship between parental rejection and adolescent alcohol use. Alcohol Alcohol 51:442–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Wang JC, Foroud T, Hinrichs AL, et al. (2013) A genome-wide association study of alcohol-dependence symptom counts in extended pedigrees identifies C15orf53. Mol Psychiatry 18:1218–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Wigginton JE, Cutler DJ, Abecasis GR (2005) A note on exact tests of Hardy–Weinberg equilibrium. Am J Hum Genet 76:887–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zuckerman M. (1999) Vulnerability to Psychopathology: A Biosocial Model. Washington, DC: American Psychological Association. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
Click here for additional data file. (40.2KB, docx)

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES