Emotional Processing and Social Cognition in Alcohol Use Disorder

Abstract

Objective:

To review deficits in emotional processing and social cognition potentially contributing to the dysfunctional emotion regulation and difficulties with interpersonal relationships observed in individuals with alcohol use disorder (AUD) and to provide directions for future research.

Method:

First is presented a review of emotional and social cognitive impairments in recently detoxified AUD individuals that include alexithymia, difficulties in decoding others’ emotions, and reduced theory of mind and empathy skills. Social cognition disorders in AUD pose different issues discussed, such as whether (1) these deficits are consequences of excessive alcohol consumption or premorbid risk factors for addiction, (2) emotional and social impairments impede positive treatment outcome, (3) recovery of social abilities is possible with sustained abstinence, and (4) AUD patients are unaware of their emotional and social dysfunctions. Finally, current knowledge on structural and functional brain correlates of these deficits in AUD are reviewed.

Results:

Emotional and social cognitive functions affected in AUD can potentially compromise efforts to initiate and maintain abstinence by hampering efficacy of clinical treatment. Such dysfunction can obstruct efforts to enable or re-instate higher-order abilities such as emotional self-regulation, motivation to change, success in interpersonal/social interactions, and emotional insight and awareness of social dysfunctions (i.e., accurate metacognition).

Conclusions:

The present review highlights the need to account for emotional processing and social cognition in the evaluation and rehabilitation of alcohol-related neurocognitive disorders and to consider psychotherapeutic treatment involving remediation of emotional and social skills as implemented in psychiatric and neurological disorders.

I. Introduction

Alcohol Use Disorder (AUD) is a highly prevalent worldwide public health issue (Esser et al., 2014; Grant et al., 2015; Rehm et al., 2015). In the United States, while 14 percent of adults (approximately 33 million persons) suffer from an AUD, only about 20 percent of those with an alcohol-related problem have ever sought treatment (Grant et al., 2015), raising questions about the lack of insight regarding emotional and social awareness in those with addiction.

Indeed, AUD individuals have a tendency to underestimate the detrimental consequences related to their excessive alcohol consumption in life’s physical, psychological, and social spheres, reflecting a lack of recognition of their disease. To initiate behavioral change toward sobriety, individuals with AUD have to acknowledge the negative consequences associated with on alcohol on their social environment (i.e., family, community, work), as well as on the body (i.e., liver cirrhosis, brain damage). Accurate perception and understanding of the repercussions of alcohol on an individual’s social sphere require intact abilities in emotional self-assessment, decoding other’s emotions, and social cognition.

Chronic excessive alcohol consumption is associated with structural and functional brain abnormalities leading to mild to moderate cognitive impairment, particularly in attention, executive functions, memory, visuospatial and motor skills, as well as in metacognitive abilities (i.e., metamemory) (Le Berre, Fama, & Sullivan, 2017; Oscar-Berman et al., 2014). In the newest edition of The Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (American Psychiatric Association, 2013), the diagnosis of mild neurocognitive disorders requires the evidence of mild cognitive decline in one or more cognitive domains initially considered in the DSM-IV (i.e., complex attention, executive function, learning and memory, language, perceptual motor) to which social cognition has been added as a new core functional domain (Simpson, 2014). Recently, the exploration of emotional processing and social cognition, referring to processes contributing to the perception and understanding of social environments and social interactions (Frith, 2008), has received a growing interest in AUD. Emotional decoding and social cognition impairment need be considered in a clinical context as these deficits may challenge emotional and interpersonal interactions (Kornreich et al., 2002; Maurage, Grynberg, Noel, Joassin, Philippot, et al., 2011), disturb self-control in social contexts (Maurage et al., 2009), and, as a consequence, accelerate time to relapse in AUD.

Herein, first, alcohol-related emotional processing and social cognition impairments in recently detoxified AUD patients will be reviewed and will include alexithymia (i.e., difficulties in identifying and describing one’s own emotions) as well as impairment in decoding emotions in others, theory of mind (ToM) deficits, and reduced empathy skills. Second, different points of interest in these recently investigated cognitive constructs will be discussed, such as the question of whether (1) these deficits are a premorbid risk factor for addiction or the untoward consequences of excessive alcohol consumption, (2) emotional and social impairments impede a positive treatment outcome, (3) potential recovery of emotional and social abilities is possible with sustained abstinence from alcohol, and (4) AUD patients are unaware of their emotional and social dysfunctions, suggesting a mild form of anosognosia as previously observed for memory disorders (Le Berre et al., 2016; Le Berre & Sullivan, 2016). Finally, current knowledge on structural and functional brain correlates of these deficits in AUD will be reviewed. This review is not a meta-analysis but rather is a synthetic narrative review based mainly on papers identified through PubMed and reference lists. This paper includes studies on emotional processing and social cognition dysfunction in individuals suffering from alcohol dependence as previously defined with DSM-IV criteria. After the DSM-5 was introduced in 2013, studies have typically used the diagnosis of AUD to supplant the earlier diagnostic scheme. It is likely that moderate to severe AUD determined with DSM-5 criteria is equivalent to alcohol dependence diagnosed with DSM-IV. For convention, we use the term AUD to refer to alcohol dependence and alcohol use disorder throughout this review paper.

II. Emotional and Social Cognition Impairment in Alcohol Use Disorder

Alexithymia

Alexithymia is a multidimensional personality construct that includes four core characteristics: (1) difficulty identifying feelings in the self and differentiating feelings from the physical sensation of emotional arousal, (2) difficulty describing feelings to others, (3) restricted imaginative processes featured by limited fantasy life, (4) an externally-oriented style of thinking (Sifneos, 1973; Taylor, Bagby, & Parker, 1997). This impairment in emotion recognition and regulation is often associated with low levels of positive well-being, including low levels of happiness and life satisfaction (Timoney & Holder, 2013) and may play a critical role in the development and maintenance of AUD (de Timary, Luts, Hers, & Luminet, 2008; Lyvers, Onuoha, Thorberg, & Samios, 2012). Although an initial critical review reported a high prevalence rate (45–67%) of alexithymia in AUD (Thorberg, Young, Sullivan, & Lyvers, 2009), it emphasized that studies conducted from 1973 to 2008 provided limited evidence for an association among alexithymia, alcohol consumption, and severity of alcohol dependence or for alexithymia as a risk factor in the development of AUD. A recent review for studies since 2009 reported lower prevalence rates (30–49%) of alexithymia in AUD individuals than previously suggested but revealed evidence to support alexithymia as an independent risk factor for alcohol related problems in clinical groups of AUD inpatients and outpatients (Cruise & Becerra, 2018). This latter review also emphasized that the relation between alexithymia and alcohol problem severity was mediated by psychological drinking constructs including alcohol expectancy, drinking motives, craving, and alcohol related intrusive thoughts and by psychological risk factors for the development of alcohol related problems including mood and emotion dysregulation, attachment, trauma, and cognitive function.

Decoding emotions in others

Emotion decoding skills are crucial when assessing one’s immediate social environment, providing valuable information regarding others’ internal affective state, enabling behavioral adaptation according to others’ thoughts and intentions, and facilitating social interactions in daily life. Impairment in decoding basic and complex emotional facial expressions of others have been consistently reported in individuals with AUD compared with healthy controls, in terms of misinterpretation or overestimation (i.e., emotional intensity) of the emotional signals conveyed by facial expressions (Castellano et al., 2015; D’Hondt, Campanella, Kornreich, Philippot, & Maurage, 2014; Donadon & Osorio Fde, 2014; Erol, Akyalcin Kirdok, Zorlu, Polat, & Mete, 2017; Marinkovic et al., 2009; Maurage, Campanella, Philippot, Martin, & de Timary, 2008; Maurage, Campanella, Philippot, Vermeulen, et al., 2008; Maurage, Grynberg, Noel, Joassin, Hanak, et al., 2011; Oscar-Berman, Hancock, Mildworf, Hutner, & Weber, 1990; Philippot et al., 1999; Townshend & Duka, 2003; Uekermann & Daum, 2008). Supporting this observation, two recent meta-analyses showed a large effect size associated with facial emotion recognition deficit in AUD, with a particular challenge in decoding negative affects such as disgust and anger (Bora & Zorlu, 2017; Castellano et al., 2015). In addition to the behavioral level, disturbance in anger processing has been demonstrated at a electrophysiological level (Maurage, Campanella, Philippot, Vermeulen, et al., 2008). Erroneous detection of emotional content in neutral facial expressions has also been observed in AUD patients early in abstinence (Kornreich et al., 2013; Kornreich, Petit, et al., 2016; Philippot et al., 1999), while recognition of happiness appeared relatively preserved (Bora & Zorlu, 2017). This emotional decoding deficit has also been generalized to positive but complex emotions with the “Reading the Mind in the Eyes” test (Baron-Cohen, Wheelwright, Hill, Raste, & Plumb, 2001), which assesses more subtle affective states than basic emotions (Maurage, Grynberg, Noel, Joassin, Hanak, et al., 2011). Such facial emotion recognition compromise could constitute a risk factor to relapse by contributing to interpersonal problems (Kornreich et al., 2002) and by potentially impeding clinical treatment efficiency (Kornreich, 2017; Rupp, Derntl, Osthaus, Kemmler, & Fleischhacker, 2017).

Emotional decoding impairments have also been observed for affective prosody, body postures, and music in AUD (Kornreich et al., 2013; Maurage et al., 2009; Monnot, Lovallo, Nixon, & Ross, 2002; Monnot, Nixon, Lovallo, & Ross, 2001). Affective prosody deficits are particularly evident when prosody does not match the semantic content of the sentence or when trying to match prosody to facial expressions (Uekermann, Daum, Schlebusch, & Trenckmann, 2005). Further, individuals with AUD may not benefit from crossmodal processing facilitation effect when attempting to identify emotions, that is, when affective information is conveyed through multiple sensory modalities [e.g., simultaneous auditory (voices) – visual (faces) processing] (Brion et al., 2017; Maurage, Campanella, Philippot, Pham, & Joassin, 2007; Maurage et al., 2013). This decoding impairment is specific for emotional features and does not result from a more general impairment in visual or facial processing or an ‘affect labelling’ impairment (Foisy, Kornreich, Petiau, et al., 2007; Maurage et al., 2009; Maurage, Campanella, Philippot, Martin, et al., 2008).

Theory of mind and empathy

Theory of mind (ToM), the ability to attribute mental states to oneself and others, and to understand that others’ mental states might differ from our own (Premack & Woodruff, 1978), enables individuals to predict, anticipate, and interpret the behavior of others and facilitates appropriate social interactions (Frith & Frith, 1999). ToM can be divided into (1) cognitive theory of mind, referring to thinking about cognitive states, beliefs, thoughts or intentions of other people, and (2) affective theory of mind, referring to thinking about affective states, feelings and emotions of other people (Shamay-Tsoory et al., 2007).

Empathy, the ability to understand and experience what others feel without confusion between oneself and others (Decety & Lamm, 2006), is a multidimensional construct encompassing (1) a cognitive component, the ability to understand another’s perspective and feelings with preserved distinction between self and other, and (2) an emotional component (i.e., affective empathy), the ability to experience and sharing emotional states of another person (Shamay-Tsoory, Tomer, Goldsher, Berger, & Aharon-Peretz, 2004). Affective ToM is often considered as an overlapping construct with cognitive empathy (Dvash & Shamay-Tsoory, 2014; Thoma, Friedmann, & Suchan, 2013).

A large effect size has been reported for deficits in ToM in AUD in two recent meta-analyses (Bora & Zorlu, 2016; Onuoha, Quintana, Lyvers, & Guastella, 2016). Inconsistencies, however, have been observed with some studies reporting ToM impairment in recently detoxified AUD patients compared to control participants (Maurage, de Timary, Tecco, Lechantre, & Samson, 2015; Maurage, Grynberg, Noel, Joassin, Hanak, et al., 2011; Nandrino et al., 2014; Thoma, Winter, Juckel, & Roser, 2013; Uekermann, Channon, Winkel, Schlebusch, & Daum, 2007), while other studies did not report ToM deficit in AUD (Amenta, Noel, Verbanck, & Campanella, 2013; Kornreich et al., 2011). Using the Reading the Mind in the Eyes Test [RMET, (Baron-Cohen et al., 2001)] to assess affective ToM and a more ecological ToM movie task [the Versailles-Situational Intention Reading, V-SIR, (Bazin et al., 2009)] to assess cognitive ToM, Nandrino and their colleagues (2014) showed poorer performance on the RMET in AUD patients compared with controls, whereas no significant between-group difference was observed on the V-SIR suggesting a specific affective ToM deficit. This dissociation between impaired affective ToM and preserved cognitive ToM has also been postulated by Maurage et al. (2016) with another ecological task, the Movie for Assessment of Social Cognition [MASC, (Dziobek et al., 2006)], which explores both ToM components. This dissociation is consistent with the assumption of a specific “affect processing system” compromise in substance dependence (D’Hondt et al., 2014; Kornreich et al., 2002; Marinkovic et al., 2009; Maurage, Grynberg, Noel, Joassin, Philippot, et al., 2011). Nevertheless, while affective ToM deficits were again confirmed using the RMET, Thoma et al. (2013) reported social cognition impairment for both cognitive ToM with reduced faux pas detection and affective ToM with lower faux pas understanding and faux pas empathy scores evaluating by the Faux Pas Story Test (Stone, Baron-Cohen, & Knight, 1998).

Using self-report questionnaires to measure empathy [e.g., empathy quotient (EQ, Baron-Cohen & Wheelwright, 2004)], the Interpersonal Reactivity Index (IRI, Davis, 1983)], AUD has been associated with difficulties in experiencing empathy in recently detoxified AUD patients (Amenta et al., 2013; Erol et al., 2017; Martinotti, Di Nicola, Tedeschi, Cundari, & Janiri, 2009; Massey, Newmark, & Wakschlag, 2018; Maurage, Grynberg, Noel, Joassin, Philippot, et al., 2011; Schmidt et al., 2017; Thoma, Friedmann, et al., 2013). Nevertheless, this impairment has not been consistently observed in all studies (Thoma, Friedmann, et al., 2013; Thoma, Winter, et al., 2013). In the same vein as ToM investigations, rather than a general empathy deficit in AUD, Maurage et al. (2011) showed a specific compromise in the emotional component, while the cognitive component of empathy is relatively preserved. It is worthy to note that this emotional empathy deficit was negatively correlated with alexithymia and severity of interpersonal problems (Maurage, Grynberg, Noel, Joassin, Philippot, et al., 2011). Therefore, emotional empathy impairment, or the difficulty to respond appropriately to another’s emotions, could compromise interpersonal relationships, thereby promoting social isolation for which heavy drinking could be a maladaptive coping strategy. The dissociation of impaired emotional empathy vs. preserved cognitive empathy has been also replicated by Ferrari et al. (2014), strengthening the idea of a specific “affect processing system” impairment in substance dependence (D’Hondt et al., 2014; Marinkovic et al., 2009). An inverse dissociated pattern combining preserved emotional and impaired cognitive empathy, however, has been observed in a small group of AUD individuals (Schmidt et al., 2017). It is important to acknowledge that self-report empathy questionnaires require a certain level of emotional insight and should be supplemented with objective measures of empathy, especially as alexithymia has been frequently observed in AUD individuals (Cruise & Becerra, 2018; Thorberg et al., 2009). Using an ecologically valid and objective measure of empathy, the Condensed and Revised Multifaceted Empathy Test [MET-Core, (Edele, Dziobek, & Keller, 2013)], a dissociation between impaired ability to decode complex emotional states associated with a specific context (i.e. cognitive empathy) and preserved emotion-sharing abilities (i.e. affective empathy) has been observed in AUD individuals early in abstinence (Grynberg, Maurage, & Nandrino, 2017).

III. Risk Factor for Addiction or Consequences of Excessive Alcohol Consumption

The observation of social cognition disorders in AUD individuals raises the question of whether these deficits are a premorbid risk factor for addiction or the untoward consequences of excessive alcohol consumption or both. This question can only be settled by conducting a longitudinal study. To our knowledge, the literature using emotional decoding and social cognition measures is scarce regarding exploration of these different hypotheses in the context of alcohol dependence. Indeed, only cross-sectional studies have been conducted on emotional and social cognition in AUD, offering only a glimpse of an answer through inference.

Emotional and social cognition disorders as premorbid risk factors for addiction

A study conducted in pathological gamblers provided evidence in favor of the premorbid vulnerability hypothesis. Specifically, an impairment in emotion decoding (i.e., poorer accuracy in vocal and facial emotion reading, overestimation of emotional intensity in neutral voices and faces) was identified in this clinical population (Kornreich, Saeremans, et al., 2016). Thus, emotional processing impairments might be a core feature of addiction, whether the addictive agent is itself neurotoxic like alcohol, or is not, like gambling.

Contradictory findings in favor of the interpretation of a premorbid risk factor for addiction has been reported by studies conducted in individuals at high-risk for AUD. An fMRI study including individuals at risk of developing AUD, that is, those with a positive family history for alcohol dependence (FH+), revealed differences in brain activation in FH+ during the RMET compared with individuals without this history (FH−). FH+ group exhibited lower activation in the right middle temporal gyrus and the left inferior frontal gyrus, providing evidence for a deterioration in the ToM cerebral network in FH+ (Hill et al., 2007). In the same vein, an abnormal pattern of cortical activation during a facial emotion recognition matching-task has been reported in youth at high-risk for substance use who showed greater activation in the medial prefrontal, precuneus, and occipital cortices compared to low-risk participants. This difference suggests impaired affective processing and regulation in youth at high familial risk for substance use disorders even before the onset of substance or alcohol use disorders (Hulvershorn et al., 2013). Further, attenuated amygdala activation to faces expressing fear has been observed in offsprings of alcoholics with behavioral desinhibition compared to youth adults without family history of alcoholism (Glahn, Lovallo, & Fox, 2007). ToM impairment, however, was not confirmed in a large sample of young adult offspring of alcohol-dependent parents using the RMET, in accordance with the idea that social cognition deficits could be the direct consequence of excessive alcohol consumption (Kopera et al., 2014). Similarly, healthy individuals with familial predisposition for AUD showed preserved cognitive and affective empathy on a self-report empathy questionnaire (IRI). Additionally, no reduced cortical thickness in empathy-related brain regions (i.e., the medial prefrontal cortex, temporoparietal junction, precuneus, insula, anterior cingulate cortex, and inferior frontal gyrus) was reported compared to healthy individuals without familial predisposition for AUD (Schmidt et al., 2017).

Emotional and social cognition disorders as consequences of alcohol toxicity

Recent cross-sectional studies suggested that emotional and social cognition impairments could be a result of neurotoxicity linked to the chronic and excessive consumption of alcohol. A recent meta-analysis reported larger effect sizes for deficits in recognition of facial emotions in studies including AUD individuals with longer duration of alcohol abuse (Bora & Zorlu, 2017). Although no group differences were observed in another study, greater lifetime alcohol consumption was associated with poorer performance on a facial emotion recognition-matching task (Charlet et al., 2014). Longer durations of heavy drinking periods were related to poorer facial emotions identification and larger amounts of alcohol consumed per day were associated with poorer ability to match emotional tone to facial expressions in long-term abstinent AUD individuals (Valmas, Mosher Ruiz, Gansler, Sawyer, & Oscar-Berman, 2014). Furthermore, AUD individuals with ToM deficits showed longer duration of alcohol dependence, greater daily alcohol consumption, and stronger current craving for alcohol compared with those without ToM impairment (Maurage et al., 2015). The demonstration of links among clinical alcoholic features (e.g., length of alcoholism, usual daily alcohol intake, or craving) and emotional and social cognitive compromise comports with the contention that alcohol toxicity can adversely affect social skills in AUD.

IV. Prediction of relapse

Longitudinal studies that examined how alexithymia at treatment entry is predictive of treatment outcomes in terms of abstinence or alcohol reduction provided conflicting results. Ziolkowski et al. (1995) suggested a negative relationship between alexithymia and maintenance of abstinence in outpatient AUD individuals (Ziolkowski, Gruss, & Rybakowski, 1995). Another investigation found that higher level of alexithymia early in abstinence predicted poorer outcome (i.e., relapse) at 15 months in inpatient AUD individuals (Loas, Fremaux, Otmani, Lecercle, & Delahousse, 1997). De Haan et al. (2012), however, did not find any link between level of alexithymia at baseline (i.e., tested after 1 month of controlled abstinence) and time in treatment, abstinence (i.e., during the previous 30 days at the 1-year follow-up) and changes in alcohol-related problems severity at 1-year follow-up in AUD individuals after an inpatient cognitive behavioral therapy (CBT) treatment. Rather, alexithymia was not a negative prognostic factor in the CBT treatment for AUD (de Haan et al., 2012). Pretreatment levels of alexithymia in treatment seeking AUD individuals were not related to attrition or to level of alcohol consumption at post-treatment (i.e., after a 12-week cognitive-behavioral intervention for AUD) (Stasiewicz et al., 2012).

Evidence suggests that poor emotional facial expressions decoding likely impedes a positive treatment outcome in AUD. Foisy et al. (2007) reported that, at the baseline assessment (i.e., 3^rd or 4^th weeks in their detoxification process), AUD individuals who dropped out from treatment had worse emotion recognition performance than abstinent AUD individuals and control participants. Similarly, poorer facial emotion recognition performances at treatment onset, especially for disgust, anger, and neutral faces, were observed in AUD individuals who relapsed or/and dropped out from treatment compared with AUD individuals who completed the treatment (Rupp et al., 2017). The AUD individuals with relapse/dropout had a tendency to misinterpret disgust as anger, anger as fear, and neutral expression as sad. The authors confirmed that poor emotion decoding abilities was a significant predictor for relapse, dropout, or both outcomes. The same study, however, did not identify group differences on perspective taking (i.e., inferences on other’s emotion, cognitive empathy or affective ToM measure) and affective responsiveness (i.e., how one’s feel if experiencing real-life situations, affective empathy measure) performance, potentially explained by lack of sensitivity of these empathy tasks. Therefore, further prospective studies investigating the influence of ToM and empathy abilities on treatment outcome in AUD remain necessary, especially as empathic processing and family social support would be negatively associated with drinking behavior at least in treatment-seeking male individuals with AUD (Robinson, Fokas, & Witkiewitz, 2018).

To our knowledge, no study has examined the influence of emotional and social impairment on motivation to change in AUD. Emotional and social cognition deficits could indirectly influence clinical outcomes (abstinence or relapse) by affecting readiness to change, as several studies have highlighted a strong relationship between motivation to change and drinking outcome (Penberthy et al., 2007; Project MATCH Research Group, 1997, 1998). To progress through the first stages of change according to the Transtheoretical Model of intentional behavior change (DiClemente, 2007; Prochaska & DiClemente, 1983), patients need to apply experiential ‘processes of change,’ i.e., cognitive-affective activities (Prochaska & DiClemente, 1983; Prochaska, Velicer, DiClemente, & Fava, 1988), changing the way one thinks and feels about alcohol abuse. These ‘processes of change’ include, but are not limited to, 1) a dramatic relief process, which entails experiencing and expressing feelings about the problem behavior and solutions and 2) an environmental reevaluation process, which involves assessing how the problem behavior affects the physical and social environment. These processes could be disturbed in AUD to the extent that they require efficient emotional and social skills. Identification of compromised emotional and social cognitive processes could provide insight into why therapy or aspects of therapy do or do not work. A structural MRI study showed that the low level of motivation to modify drinking behavior observed in some alcoholic patients at treatment entry may be related to macrostructural brain abnormalities in regions subtending emotional and social cognition skills, emphasizing the need to conduct further studies of the relationship between social cognition and readiness to change drinking behavior (Le Berre et al., 2013).

V. Recovery from social cognition impairment with abstinence in AUD

Using cross-sectional design studies, accuracy in decoding emotional facial expressions such as anger, disgust, and sadness can remain impaired with mid- to long-term abstinence, that is, at least two months and beyond (Carmona-Perera, Clark, Young, Perez-Garcia, & Verdejo-Garcia, 2014; Foisy, Kornreich, Fobe, et al., 2007; Kornreich et al., 2001; Valmas et al., 2014), even if some improvement occurred with abstinence such as absence of emotional intensity overestimation (Kornreich et al., 2001). Nevertheless, to our knowledge, only one longitudinal study demonstrated improvement in facial expression recognition and discrimination after a 3 month-interval of abstinence, with AUD individuals reaching similar performance level as control participants (Erol et al., 2017).

Regarding ToM processes, conflicting findings have been reported with some studies showing persistence of deficits with prolonged abstinence (Bosco, Capozzi, Colle, Marostica, & Tirassa, 2014; Gizewski et al., 2013; Schmidt et al., 2016), whereas other studies showed no impairment in long-term abstinent AUD patients in recognition of complex social emotions and mental states (Matyassy, Kelemen, Sarkozi, Janka, & Keri, 2006). For empathy processes, using self-report measures, no general empathy deficit was observed in AUD individuals with years of abstinence in a cross-sectional study (Schmidt et al., 2016). A longitudinal study reported that the empathy ability of AUD patients improved after 3 months of abstinence and reached control levels (Erol et al., 2017). Further investigations using longitudinal designs remain essential especially for assessing within-subject change by retesting the same group of alcoholic patients and control participants at variable time intervals.

VI. Awareness of social cognition impairment in AUD

Recently, new considerations to expand the understanding of functional impairment in alcoholism have included the investigation of metacognition (the ability to accurately assess one’s own cognitive abilities) (Le Berre & Sullivan, 2016). The hypothesis of a mild form of anosognosia for memory impairment in AUD has been proposed as individuals with AUD have a tendency to overestimate their memory skills (Le Berre et al., 2016; Le Berre et al., 2010). A potential lack of awareness of social cognitive deficits in AUD has also grown as a source of interest.

Despite observation of overestimation of emotional facial expressions intensity and decoding inaccuracy in recently detoxified and mid- to long-term abstinent AUD individuals, some studies have suggested that patients did not perceive their deficits, estimating the emotional facial recognition task as difficult as controls’ evaluation level (Kornreich et al., 2001; Kornreich et al., 2002; Philippot et al., 1999). Moreover, despite poorer facial emotion recognition performance in individuals with relapse/dropout during clinical treatment, “relapse/dropout” AUD individuals provided similar subjective self-estimation of task difficulty and performance than AUD individuals who did not relapse and/or prematurely terminate their treatment. “Relapse/dropout” AUD patients had a tendency to overestimate their emotional decoding abilities, judging that they made fewer errors than patients who did not relapse (Rupp et al., 2017). In the same spirit, despite reduced empathy abilities, recently detoxified AUD individuals did not differ in terms of self-reported trait empathy compared with control participants (Thoma, Winter, et al., 2013).

Emotional and social cognition deficits can impede interpersonal interactions leading AUD individuals to experience an emotional imbalance without real insight (i.e., accurate metacognition) regarding its origins and reinforced by an unawareness of their difficulties in decoding their social environment. Alcohol consumption could be used as a coping mechanism to relieve the emotional and social burden, feeding the vicious cycle of addiction (Kornreich et al., 2002).

VII: Brain correlates of emotional processing and social cognitive impairment in AUD

Although emotional decoding deficits in AUD have been consistently observed in multiple behavioral investigations, few studies have explored the associated brain substrates of these impairments. Those that have generally report brain structural, functional activation and connectivity differences related to emotional facial expression recognition in the prefrontal cortex, limbic brain regions including the hippocampus and amygdala, and the insula (Charlet et al., 2014; Marinkovic et al., 2009; O’Daly et al., 2012; Salloum et al., 2007; Trick, Kempton, Williams, & Duka, 2014; Wilcox, Pommy, & Adinoff, 2016). Structural and functional brain differences in the fronto-limbic circuitry are one of the critical neural signatures in alcohol dependence (Beresford et al., 2006; Cardenas, Studholme, Gazdzinski, Durazzo, & Meyerhoff, 2007; Chanraud et al., 2007; Demirakca et al., 2011; Jung et al., 2007; Laakso et al., 2000; Le Berre, Pitel, et al., 2014; Makris et al., 2008; Momenan et al., 2012; Pfefferbaum, Sullivan, Mathalon, & Lim, 1997; Pitel et al., 2012; Rosenbloom, Sullivan, & Pfefferbaum, 2003; Senatorov et al., 2015; Sullivan, Marsh, Mathalon, Lim, & Pfefferbaum, 1995; Sullivan et al., 2013; Sullivan & Pfefferbaum, 2005). To our knowledge, the only cross-sectional brain structural neuroimaging study showed smaller gray matter volume in frontal brain regions implicated in emotional processing, including the inferior frontal cortex (IFC) and the insula, in AUD individuals who were less accurate in fearful expressions recognition (Trick et al., 2014). While poorer emotional decoding for fear was associated with smaller gray matter volume in the IFC in the whole group of AUD patients and controls, AUD individuals with more than two medically supervised detoxifications had smaller IFC gray matter volume relative to AUD patients with two or fewer detoxifications and controls. Nevertheless, it remains unclear why, in the same study, difficulties in decoding disgust and anger were only observed in AUD with two or fewer detoxifications compared to controls (Trick et al., 2014). Lower functional activation in the rostral anterior cingulate cortex (rACC) has been consistently reported during decoding of negative facial emotions such as fear and disgust in AUD individuals abstinent at least 2 weeks (O’Daly et al., 2012; Salloum et al., 2007; Wilcox et al., 2016). Only one study has investigated whether abnormal brain functional activation pattern during negative emotional face expressions (i.e., angry and fearful) processing could be predictive of treatment outcome in recently detoxified AUD (Charlet et al., 2014). Greater functional activation in the rACC was observed in AUD patients compared to control participants, even after controlling for gray matter volume differences, and was associated with lower total amount of alcohol consumed over a lifetime, longer sobriety, and less subsequent binge drinking in AUD individuals (i.e., better treatment outcome). Charlet et al. (2014) speculated that higher activation in ACC, a brain region known to be involved in emotion regulation and error monitoring, might participate in the cognitive control of affective responses toward aversive stimuli and thus constitute a key resilience factor against relapse.

Distinct neural networks are hypothesized to be associated with ToM and empathy abilities: (1) Cognitive ToM - mediated by the medial prefrontal cortex (more particularly the dorsomedial prefrontal cortex), dorsolateral prefrontal cortex, dorsal anterior cingulate cortex, superior temporal sulcus, temporoparietal junction, and temporal poles (Dvash & Shamay-Tsoory, 2014), (2) affective ToM (i.e., cognitive empathy) - mediated by the medial prefrontal cortex (more particularly the ventromedial prefrontal cortex) (Dvash & Shamay-Tsoory, 2014; Schnell, Bluschke, Konradt, & Walter, 2011), temporoparietal junction (Schulte-Ruther, Markowitsch, Fink, & Piefke, 2007), and precuneus (Nummenmaa, Hirvonen, Parkkola, & Hietanen, 2008), (3) affective empathy - mediated by the inferior frontal gyrus, anterior cingulate cortex, amygdala, and insula (Bernhardt & Singer, 2012; Dvash & Shamay-Tsoory, 2014; Shamay-Tsoory, Aharon-Peretz, & Perry, 2009). Brain differences in alcohol dependent individuals have been identified notably in the frontal lobes including the anterior cingulate cortex and the dorsolateral, dorsomedial and ventromedial prefrontal cortices (Dupuy & Chanraud, 2016; Grodin, Lin, Durkee, Hommer, & Momenan, 2013; Le Berre, Rauchs, et al., 2014; Sullivan et al., 2018) and the limbic system including hippocampal and parahippocampal gyri, amygdala, and insula (Le Berre, Pitel, et al., 2014; Le Berre, Rauchs, et al., 2014; Makris et al., 2008; Pfefferbaum et al., 2018; Pitel, Chanraud, Muller-Oehring, Pfefferbaum, & Sullivan, 2013; Senatorov et al., 2015; Sullivan et al., 2013), as well as in the parietal and lateral temporal cortices (Le Berre et al., 2016; Le Berre et al., 2013). To date, the literature on brain correlates of ToM and empathy in AUD is scarce. Preliminary evidence of thinner cortex in empathy-associated brain regions has been reported in recently detoxified patients with AUD relative to healthy participants with and without familial predisposition for AUD (Schmidt et al., 2017). Using a self-report empathy questionnaire [IRI, (Davis, 1983)], affective empathy (i.e., IRI personal distress subscale) was related to thinner cortex in the inferior frontal gyrus as expected, whereas cognitive empathy (i.e., IRI perspective taking subscale) was unexpectedly associated with thinner cortex in inferior, middle, and superior frontal gyri. Schmidt et al. (2017) suggested that structural brain alteration in these frontal regions, known to be involved in the “cognitive control” network, could contribute to poorer insight and thus overestimation of one’s perspective taking skills, in accordance with the assumption of a metacognition impairment in AUD (Le Berre & Sullivan, 2016). Due to small sample size and potential confounding factors such as withdrawal, the findings of this study should be interpreted with caution. Only one fMRI study suggested that alcohol-related impairment in the ability to infer other’s mental states is limited to insular dysfunction and thus deficits in affective empathy in long-term abstinent AUD individuals, but this interpretation is limited again by the small sample sizes (Gizewski et al., 2013).

VIII. Conclusion

AUD is associated with difficulties in emotion processing and social cognition, including difficulties in expressing emotions (i.e., alexithymia), decoding other’s emotions, inferring other’s mental states or feelings (i.e., ToM deficit), and experiencing empathy. Any or all of these impairments can potentially contribute to the development and maintenance of alcohol-related problems and interfere with recovery. Negative affect (e.g., depression) or social pressure are responsible for approximately 70% of relapse after detoxification (Zywiak, Westerberg, Connors, & Maisto, 2003). This high rate highlights the importance of considering emotional and interpersonal difficulties in the clinical treatment of AUD individuals. Deficits in emotional processing and social cognition, along with lack of awareness of the deficits, are thought to impede emotional self-regulation and social interactions. Clearly, further studies are required to specify the role played by emotional and social impairment in negative affect and social pressure relapses (Figure 1).

Figure 1.

Emotional and social cognitive functions affected in chronic alcoholism compromising efforts to initiate and maintain abstinence by potentially hampering efficacy of clinical treatment and potentially obstructing efforts in enabling other higher-order abilities such as emotional self-regulation, motivation to change, success in interpersonal/social interactions, and emotional insight and awareness of social dysfunctions (i.e., accurate metacognition).

Blue arrows with question mark = Need of further investigation on how emotional and social cognition processes interact and potentially affect the higher-order abilities.

Arrow 1 = Need of further investigation on how component episodic memory and executive processes interact and potentially influence emotional and social cognition processes.

Arrow 2 = Need of further investigation on whether social cognition disorders are the untoward consequences of excessive alcohol consumption or a premorbid risk factor for addiction.

Arrow 3 = Need of further investigation on the dynamic course of recovery with abstinence or on the negative influence of multiple relapses/detoxifications on emotional and social cognitive impairment.

Arrow 4 = Need of further investigation on emotional social cognitive deficits as predictors for relapse.

Recovery from social cognition impairment in abstinent AUD individuals has been mainly investigated with cross-sectional designs, comparing alcoholic groups with different lengths of sobriety -varying from months to several years- with a control group of healthy participants (Bora & Zorlu, 2017; Onuoha et al., 2016). Future longitudinal investigations are still needed to determine whether the course of recovery of emotional and social cognitive deficits is associated with length of abstinence, and if so, whether these impairments completely recover with sobriety and how long does recovery take. Furthermore, it has yet to be determined whether the contribution of alcohol toxicity is relevant to social cognitive impairment in AUD or whether these deficits constitute a premorbid risk factor for addiction. Only longitudinal studies can impart new insight to these questions. A recent literature review reported significant association between early childhood social experience, including insecure attachment and adversity relating to neglect or abuse, and poorer social cognitive performance in individuals with major psychiatric disorders such as schizophrenia, bipolar disorder, borderline personality disorder, major depressive disorder, and posttraumatic stress disorder (Rokita, Dauvermann, & Donohoe, 2018). Childhood trauma including abuse, maltreatment, and family adversity are directly and indirectly associated with alcohol problems in adulthood (Fenton et al., 2013; Pilowsky, Keyes, & Hasin, 2009; Strine et al., 2012), and a more severe clinical profile of alcohol dependence, and poorer response to alcoholism treatment (Brady & Back, 2012). The different observations underline the necessity to investigate the relationship between childhood trauma, attachment styles, and social cognition impairment in AUD, as already suggested by Kornreich (2017). Early childhood adversity might favor the emergence of social cognition dysfunctions making this vulnerable population more prone to use and abuse of alcohol as a coping strategy.

Previous studies reported that between 50% and 80% of AUD individuals early in abstinence experience mild to moderate cognitive impairments marked by heterogeneity of pattern and severity (Ihara, Berrios, & London, 2000; Oscar-Berman et al., 2014; Petit et al., 2017). Similarly, the question of heterogeneity in emotional and social cognition deficits needs additional exploration in AUD individuals in order to identify those with potentially more interpersonal difficulties, less benefit from clinical treatment, and higher risk to relapse. To date, heterogeneity in social cognition was only addressed by one study reporting ToM impairment in less than 50% of the recently detoxified AUD individuals tested (Maurage et al., 2015).

AUD is associated with deficits in a number of cognitive domains including executive functioning and memory (for reviews, see Le Berre et al., 2017; Oscar-Berman et al., 2014). Executive functions (EF) refer to cognitive processes involved in the self-regulation of goal-directed and adaptive behaviors in response to novel or non-routine situations (Alvarez & Emory, 2006; Miyake et al., 2000). Specific executive processes have been reported as impaired in AUD, such as attention, working memory, response inhibition, problem-solving, deduction of rules, updating, cognitive flexibility and set-shifting, and impulsivity (for a review, see Le Berre et al., 2017). By investigating different assumptions regarding potential functional overlap between executive and ToM processes through a developmental cognitive neuroscience perspective, Wade et al. (2018) highlighted key findings including (1) interrelations between behavioral measures of ToM and EF during childhood would be preferentially in favor of the hypothesis that ToM relies on EF, (2) in ontogenetic development, coincident and rapid maturation of ToM and EF was observed over the first 5 years of life, and (3) ToM and EF rely on distinct but also shared neural circuits, particularly in the medial prefrontal cortex, inferior parietal lobule, temporo-parietal junction, and inferior frontal gyrus (For a review, see Wade et al., 2018).

Memory disorders have been also consistently described in AUD, including impairments in episodic memory (Le Berre et al., 2010; Noel et al., 2012; Pitel et al., 2007) and autobiographical memory (D’Argembeau, Van Der Linden, Verbanck, & Noel, 2006; Nandrino et al., 2016). Episodic memory is the memory system enabling the encoding, storage, and retrieval of personally experienced events, associated with a precise temporal and spatial context (Tulving, 2001, 2002). Episodic memory is also the foundation for conscious recollection of specific personal events from one’s past and the mental projection of anticipated events into one’s subjective future (Wheeler, Stuss, & Tulving, 1997). Recollection of episodic events includes autonoetic awareness, which is the feeling of re-experiencing or reliving the past and mentally traveling back in subjective time (Tulving, 2001). Compared to controls, deficits in encoding and retrieval processes were observed in recently abstinent alcoholics (Pitel et al., 2007) as well as altered spatiotemporal context of encoding (Salmon, Butters, & Schuckit, 1986; Sullivan, Shear, Zipursky, Sagar, & Pfefferbaum, 1997) and autonoetic consciousness (Le Berre et al., 2010; Pitel et al., 2007). The autobiograpical memory involves real-life events and is the product of multiple representations forming the memory, including the specific event itself (episodic component), general knowledge about one-self (semantic component), and the spatial and temporal context of its occurrence (Conway, 2001; Tulving, Schacter, McLachlan, & Moscovitch, 1988). Assessment of autobiographical memory in AUD patients revealed that they recall specific memories less frequently and general memories more frequently than controls (D’Argembeau et al., 2006).

In ontogenetic development, episodic memory and ToM emerge close in time, around the age of 4 years, and ToM development is associated with the improvement of episodic memory (Naito, 2003; Perner, Kloo, & Gornick, 2007; Perner, Kloo, & Rohwer, 2010; Perner & Ruffman, 1995). Moreover, neuroimaging studies provide evidence that ToM and episodic memory abilities are supported by similar brain regions known to be an integral part of the default mode network (DMN) implicated in internally focused tasks such as autobiographical memory retrieval, self-reference activities, and imaging the future (Buckner & Carroll, 2007; Rabin, Gilboa, Stuss, Mar, & Rosenbaum, 2010; Spreng & Grady, 2010; Spreng & Mar, 2012; Spreng, Mar, & Kim, 2009). This brain network involves the medial temporal lobe (hippocampus, parahippocampus, amygdala), medial prefrontal cortex, frontal pole, inferior frontal gyrus, posterior cingulate cortex, posterior midline and lateral parietal cortex (angular gyrus), and lateral temporal cortex (Buckner, Andrews-Hanna, & Schacter, 2008). In AUD, neuroimaging studies have shown volume deficits in brain structures involved in the DMN (Agartz, Momenan, Rawlings, Kerich, & Hommer, 1999; Beresford et al., 2006; Cardenas et al., 2011; Cardenas et al., 2007; Fein, Shimotsu, Chu, & Barakos, 2009; Jang et al., 2007; Le Berre et al., 2016; Le Berre, Pitel, et al., 2014; Le Berre, Rauchs, et al., 2014; Makris et al., 2008; Moselhy, Georgiou, & Kahn, 2001; Pfefferbaum et al., 1997; Sullivan et al., 1995). Activation synchrony among nodes of this resting state brain network is also disrupted in chronic alcoholics (Chanraud, Pitel, Pfefferbaum, & Sullivan, 2011). During resting-state functional MRI, connectivity analyses identified weaker within-default mode network connectivity and expanded connectivity to regions outside the DMN in alcoholics relative to the controls (Muller-Oehring, Jung, Pfefferbaum, Sullivan, & Schulte, 2015). Buckner and Carroll (2007) suggested that episodic memory (remembering the past), prospection (envisioning the future), ToM, and navigation might be a larger cluster of functions enabling flexible forms of self-projection (i.e., the ability to shift perspective from the immediate present to alternative perspectives). Self-projection feeds our skills to make social inferences and anticipate others’ beliefs and actions. With the exception of an identified role of autobiographical memory for ToM abilities (Nandrino et al., 2014) and attention/working memory and executive functions for empathy skills (Erol et al., 2017; Thoma, Winter, et al., 2013), the exploration of interactions between cognitive deficits (executive functions and memory) and emotional and social impairments are lacking in AUD. Extended knowledge of cognitive processes underlying these emotional and social alterations will provide relevant information on the origin of social cognitive disorders observed in many individuals with AUD and directions to adapt clinical treatment.

Emotional and social cognitive functions affected in AUD might compromise efforts to initiate and maintain abstinence by potentially hampering efficacy of clinical treatment and by obstructing efforts in enabling other higher-order abilities such as emotional self-regulation, motivation to change, success in interpersonal/social interactions, and emotional insight and awareness of social dysfunctions (i.e., accurate metacognition) (Figure 1). On a clinical level, recognition of these dysfunctions as a central component of the addiction syndrome indicates the necessity to take into account emotional processing and social cognition in the evaluation and the rehabilitation of alcohol-related neurocognitive disorders and to consider therapeutic program including the remediation of emotional and social skills as already implemented in other psychiatric and neurological disorders (Bornhofen & McDonald, 2008; Peyroux & Franck, 2014; Wolwer & Frommann, 2011).

Public Significance Statement.

Impairments in emotional processing and social cognition have the potential to contribute to the dysfunctional emotion regulation and interpersonal relationship difficulties observed in many individuals with alcohol use disorder (AUD). As a consequence, these problems have the potential to precipitate relapse. Emotional and socio-cognitive abilities should be considered in the evaluation and the rehabilitation of alcohol-related neurocognitive disorders with remediation programs targeting these skills, as already implemented in other psychiatric and neurological disorders.