Drug effects on responses to emotional facial expressions: recent findings

Abstract

Many psychoactive drugs increase social behavior and enhance social interactions, which may, in turn, increase their attractiveness to users. Although the psychological mechanisms by which drugs affect social behavior are not fully understood, there is some evidence that drugs alter the perception of emotions in others. Drugs can affect the ability to detect, attend to, and respond to emotional facial expressions, which in turn may influence their use in social settings. Either increased reactivity to positive expressions or decreased response to negative expressions may facilitate social interaction. This article reviews evidence that psychoactive drugs alter the processing of emotional facial expressions using subjective, behavioral, and physiological measures. The findings lay the groundwork for better understanding how drugs alter social processing and social behavior more generally.

Introduction

For reasons that are not fully understood, most recreational drug use takes place in social settings, in the presence of other people (Single and Wortley, 1993; Halkitis et al., 2005; Rodgers et al., 2006; Acosta et al., 2008). The presence of others appears to facilitate drug use, and drugs, in turn, can enhance social interactions. Both laboratory-based and naturalistic studies support the idea that social contexts facilitate both drug use (i.e. initiation, maintenance, and relapse) and the quality and magnitude of direct drug effects (Carlin et al., 1972; Kelly et al., 1994; Doty and de Wit, 1995; Kirkpatrick and de Wit, 2013). Conversely, many abused drugs also enhance social interactions by increasing the desire to socialize, social bonding, verbal behavior, and the perception of attractiveness in others (Sayette et al., 2012).

The ability to recognize and respond to the emotional states of others is a critical component of daily social interaction. Facial expressions are potent social cues that signal how others are feeling and thus guide appropriate social responses. Early investigations concerning facial processing examined basic facial recognition and identification of emotion. Initial work by Ekman (1972) and Izard (1971, 1977) indicated that humans identify basic emotions that are universally recognized across cultures: surprise, fear, anger, sadness, disgust, and happiness. Since then, studies on facial processing have expanded our understanding of how differences and biases in the perception of these emotions are directly correlated with mood and behavior. Even modest difficulties with emotional perception can influence mood states and underlie several psychiatric conditions such as depression, anxiety, and schizophrenia (Feinberg et al., 1986). The direction of this relation is not always clear: mood states can also influence the perception of emotional expressions (Bouhuys et al., 1995). Beyond emotional states, difficulties in facial processing contribute significantly to interpersonal problems (Carton et al., 1999; Bourke et al., 2010), including social anxiety and rejection from peers (Edwards et al., 1984; Philippot and Feldman, 1990; Feldman et al., 1991).

Clinical studies suggest that behavioral therapy and psychotropic drugs, such as antidepressants and anxiolytics, can positively affect the processing of social cues both in individuals with mood disorders (Harmer et al., 2009; Tranter et al., 2009) and in healthy adults (Harmer et al., 2003, 2004). Indeed, it has been suggested that changes in perception of affective stimuli precede and even predict improvements in mood and social functioning. Recreational drugs of abuse also alter the processing of facial expressions in ways that may make the drugs more attractive to users. Here, we examine the evidence that drugs of abuse alter the interpretation of emotions in others and speculate how these effects might influence social behavior and further consumption of the drug. We review studies examining emotional processing in response to five drugs of abuse: alcohol, nicotine, cannabinoids, stimulants, and opiates. For each drug type, we review evidence of acute effects of the drug on behavioral, psychophysiological, and neural responses to emotional faces, and then briefly discuss indirect evidence that chronic use of these drugs affects emotional processing, by comparing nonusers with long-term drug users.

Drug effects on facial processing

Alcohol

Alcohol is typically initially used and regularly consumed in the presence of others (Wechsler et al., 1995; Cashin et al., 1998; Kairouz and Greenfield, 2007). Peers strongly influence the initiation of drinking (Simons-Morton et al., 2001), and social drinkers consume greater amounts when they are in the presence of others (Doty and de Wit, 1995). Many people claim to drink to lower inhibitions and increase their sociability (Smith et al., 1993), and there is evidence that alcohol increases the attractiveness of social settings and interactions, as well as the perceptions of others. Here, we review evidence that alcohol enhances social interactions by altering the way drinkers perceive others. We review behavioral and imaging studies of the acute effects of alcohol on the perception and recognition of facial cues. In addition, we discuss findings suggesting that prolonged, heavy consumption impairs emotional processing and discuss the social implications of these effects.

Effects of acute administration

Alcohol is often considered a ‘social lubricant’, and it may have this effect by increasing the positive perceptions of contextual stimuli, including other people. This phenomenon is based on both anecdotal reports (‘rose colored glasses’, ‘beer goggles’) and controlled studies indicating that alcohol increases the attractiveness of others. In one quasiexperimental study, intoxicated drinkers in a bar were asked to rate the attractiveness of faces of strangers, and higher blood alcohol concentrations were related to higher ratings of attractiveness only for opposite-sex faces (Jones et al., 2003). This suggests that the effect may be related to increases in sexual interest or arousal. In contrast, controlled studies indicate that moderate doses (0.4 g/kg) of alcohol increase drinkers’ ratings of attractiveness of faces of strangers regardless of their sex, indicating that the effect may be independent of sexual attraction (Parker et al., 2008; Attwood et al., 2012). Thus, the use of alcohol in social settings such as bars may facilitate both sexual and nonsexual social interactions, mediated in part by alcohol-induced alterations in the perception of others.

Another way in which alcohol may enhance social interactions is by positively biasing the way drinkers perceive emotions in others. Several studies have addressed the effects of alcohol on the ability to identify facial expressions varying in emotional intensity. In three studies, moderate doses of alcohol (0.4 g/kg) disrupted the processing of sad expressions (Craig et al., 2008; Attwood et al., 2009a; Kamboj et al., 2013). Similarly, using the same dose of alcohol, Kamboj et al. (2013) showed that participants frequently misidentified sad faces as neutral. Interestingly, the effects on emotion processing in each of these studies were unrelated to alcohol-induced changes in mood or alcohol expectancies. In contrast, two other studies failed to find an effect of alcohol on emotion recognition compared with placebo (Kano et al., 2003; Walter et al., 2011). Thus, there is some evidence that alcohol disrupts the recognition of sadness, which may have the effect of facilitating positive social interactions.

Some work has also examined the effect of alcohol on social processing in individuals with social phobias. Individuals with social anxiety or impaired social function are at a greater risk for alcohol-related problems (Himle and Hill, 1991; Lépine and Pélissolo, 1998), perhaps because alcohol alleviates their social anxieties (Burke and Stephens, 1999; Randall et al., 2001). Using a dot probe task, Stevens et al. (2009) examined the effects of alcohol on attentional biases to facial expressions in adults with social phobia compared with healthy controls. Before alcohol administration, individuals with social phobia displayed a greater attentional bias to angry faces compared with controls, and a moderate dose of alcohol significantly reduced this bias in the clinical group, without affecting controls. Thus, the ability of alcohol to reduce attention to negative emotions may be especially pronounced in individuals with pre-existing negative biases in social processing.

Another ‘social’ effect of alcohol is its tendency to increase aggressive and violent behavior, an effect that may also be mediated by changes in the perception of emotions in others – that is, alcohol-induced misidentification of emotions (i.e. misidentifying a neutral expression for anger) could increase interpersonal conflict and aggression. To date, there is no strong evidence for this. Typically, alcohol-induced aggression is observed with high doses of alcohol, probably higher than the doses used in most laboratory studies (Ito et al., 1996). Alcohol doses of 0.56 and 0.8 g/kg failed to affect detection of angry faces (Kano et al., 2003; Kamboj et al., 2013), but these doses are still lower than those used in natural settings (Monahan and Lanutti, 2000). The possibility that alcohol-induced aggression results from misattributions of emotions in others remains to be studied.

Taken together, these studies suggest that moderate doses of alcohol increase perceptions of facial attractiveness and reduce sensitivity to negative emotions, independent of the drug’s effects on mood. These effects may facilitate positive social interactions and thus increase the likelihood of continued drinking. Alcohol-induced decrease in detection of negative emotions may be especially pronounced in individuals with high levels of social anxiety. It remains to be determined whether the effects of alcohol on aggression are also influenced by changes in social perceptions.

Effects of chronic use

Whereas acute use of alcohol appears to facilitate social behavior, long-term, heavy use of the drug impairs social functioning. Some of the interpersonal problems related to excessive consumption, including violence, aggression, and damaged social/familial relationships (Nixon et al., 1992; Duberstein et al., 1993; Kornreich et al., 2002), may be related to impaired or distorted perceptions of emotions in others. Alcoholics show deficits in social processing (Philippot et al., 1999; Kornreich et al., 2001a, 2002; Foisy et al., 2007) that persist even after long periods of abstinence (Kornreich et al., 2001b). They are less able to recognize negative emotions such as anger and disgust compared with healthy controls (Townshend and Duka, 2003), they tend to misidentify expressions such as anger (Philippot et al., 1999; Frigerio et al., 2002), and they show less brain activation in response to expressions of negative emotion compared with controls (Salloum et al., 2007). However, alcoholics also overestimate the intensity of facial expressions (Philippot et al., 1999; Frigerio et al., 2002; Townshend and Duka, 2003). These combined impairments, including misclassifications and exaggerated perceptions of emotional intensity, may contribute to interpersonal difficulties characteristic to alcohol use disorders. Indeed, Kornreich et al. (2002) observed a direct relationship between alcoholics’ deficits in emotion recognition and self-reported interpersonal problems. However, without detailed longitudinal studies, we cannot determine whether these deficits in social perception precede the development of a drinking problem, or whether they result from years of alcohol use.

Cigarette smoking and nicotine

Social factors influence cigarette smoking at several stages, including at initiation when use begins with peers, at maintenance when social settings increase smoking (e.g. ‘social smokers’, who smoke only in social situations, Moran et al., 2004), and during quit attempts, when social factors increase relapse (Cengelli et al., 2012). Some of these effects may be attributable to the direct effects of nicotine on perception of social stimuli. Interestingly, in laboratory animals, nicotine also increases the rewarding properties of environmental stimuli, including food, other drugs, and conditioned incentive stimuli that have been paired with rewards (Popke et al., 2000; Clark et al., 2001; Bechtholt and Mark, 2002; Donny et al., 2011). Thus, it is possible that nicotine enhances the value of other, nonpharmacological stimuli, and that perception of emotions in humans is another instance of this more general effect. That is, nicotine may enhance the value of social cues, including the value of faces of strangers.

Effects of acute administration

Two studies investigated the effects of nicotine on ratings of facial attractiveness. In one study, nondependent smokers who smoked a nicotinized cigarette rated facial stimuli as significantly more attractive compared with those who smoked a denicotinized cigarette (Attwood et al., 2009a, 2009b). Moreover, this effect was not related to the sex of the participant or the images, indicating a global increase in the inherent value of faces, probably unrelated to sexual attraction. Another study examined the effects of alcohol and nicotine, alone and together, on facial attractiveness ratings (Attwood et al., 2012). Both drugs increased the ratings of attractiveness, and the combination produced higher ratings of attractiveness than either drug alone. The effects of nicotine on ratings of attractiveness were unrelated to the subjective or mood effects of the drug, suggesting that this reflects a distinct cognitive process.

These initial findings suggest that nicotine enhances the value of social cues just as it enhances other rewarding stimuli in laboratory animals. Thus, the drug may enhance the influence of peers and social context on smoking behavior, thereby escalating further use. Although the two studies reviewed here examined the rating of attractiveness of faces, little is known about whether nicotine specifically alters processing of positive or negative emotional facial expressions, which may be a fruitful direction for future research.

Effects of chronic use

To our knowledge, no studies have examined the longterm effects of smoking on social processing. A recent study reported that nondependent smokers exhibited impaired recognition of happy and sad faces compared with adults who had never smoked (Meyers et al., 2015), suggesting that a history of smoking might contribute to changes in emotional recognition. It remains to be determined whether chronic exposure to nicotine and nicotine dependence affect appraisal and processing of social cues.

Cannabinoids

Like alcohol and cigarettes, cannabis is commonly used in social contexts. Some cannabis users report feelings of closeness, empathy, and interpersonal warmth under the effects of the drug (Tart, 1971), and some report taking the drug to alleviate social stress and anxiety (Reilly et al., 1998; Ogborne et al., 2000; Buckner et al., 2006; Bonn-Miller et al., 2007). Further, the two primary constituents of Cannabis sativa, Δ⁹-tetrahydrocannabinol (THC) and cannabidiol (CBD), appear to alter core emotional processes including anxiety (Sethi et al., 1986; Wachtel et al., 2002; De Souza Crippa et al., 2004). Thus, THC or CBD may facilitate social behavior, perhaps by reducing the effect of socially threatening stimuli. There is evidence that both acute and long-term cannabis use specifically alter processing of threatening facial stimuli.

Effects of acute administration

One recent study examined the effects of oral THC (7.5 and 15 mg) on identification of emotional facial stimuli in healthy young adults (Ballard et al., 2012). The higher dose impaired recognition of anger, whereas both doses impaired recognition of fear and neither dose affected the recognition of happy or sad faces. These findings suggest that THC specifically disrupts processing of threatening stimuli. Because psychoactive drugs often have direct effects on mood states, the question arises whether the effects on face recognition can be accounted for by the drug’s effects on mood. In this study, 15 mg THC increased self-reported ratings of anxiety, but this mood effect did not account for the effects of the drug on the recognition of threatening stimuli.

Several studies have examined the effects of THC on brain activation induced by emotional faces (Phan et al., 2008; Fusar-Poli et al., 2009; Bhattacharyya et al., 2010). Social signals of threat are known to increase amygdalar activity (Morris et al., 1996; Whalen et al., 1998). Using functional MRI, Phan et al. (2008) reported that a relatively low dose of THC (7.5 mg oral) reduced amygdalar activity in response to angry and fearful faces, without affecting responses to other emotions. However, a higher dose (10 mg) did not produce this effect in two other studies (Fusar-Poli et al., 2009; Bhattacharyya et al., 2010). These differences across doses may be related to the mood-altering effects of THC. In the study by Fusar-Poli et al. (2009), THC (10 mg) increased subjective ratings of anxiety and skin conductance, consistent with other evidence that high doses increase anxiety and even induce paranoia (Berrendero and Maldonado, 2002; Manzanares et al., 2004; Viveros et al., 2005).

CBD, the other important constituent of cannabis, has also been studied in this regard. CBD appears to reduce anxiety and may counteract the anxiogenic effects of THC (Zuardi et al., 1982; Guimaraes et al., 1990; Zuardi et al., 1993; Williamson and Evans, 2000; De Souza Crippa et al., 2004; Moreira et al., 2006). Both Bhattacharyya et al. (2010) and Fusar-Poli et al. (2009) reported that CBD (600 mg) reduced neural responses to fearful faces, as well as lowered self-reported anxiety and reduced skin conductance. Thus CBD might exert its anxiolytic effects through disruptions in the processing of threatening stimuli occurring at the physiological level.

Together, these findings indicate that both THC and CBD can reduce neural responding to threatening stimuli. Dampened responses to threatening faces may be related to reduced social anxiety and feelings of warmth with others, especially in individuals with high baseline levels of social anxiety. This effect may lead to greater use. Studies on the effect of CBD on behavioral responses or recognition of facial expressions will be important to understand the changes in neural processing of threatening stimuli and the potential role of cannabis as a prosocial drug.

Effects of chronic use

There is limited evidence that long-term use of cannabis affects processing of social stimuli, such as detection of emotional faces. In one study, Platt et al. (2010) reported that cannabis users are significantly slower to identify facial expressions and that they require greater perceptual information compared with nonusers. However, the differences were not specific to particular emotional valence (i.e. happiness, sadness, anger). Another study used functional MRI to show that heavy marijuana users exhibited lower amygdalar activity activation during the presentation of angry faces compared with controls (Gruber et al., 2009). It is possible that these lasting changes in the processing of emotions, and in particular threatening emotions, might reinforce the habitual use of marijuana as a way to reduce social anxiety. The impaired ability to identify or process emotional cues may also adversely affect social functioning. These potential effects remain to be determined.

Opioids

Opioid drugs and the endogenous opioid system are integrally involved in social processing. Early studies in rodents and other mammals have shown that opioid agonists blunt responses to isolation distress (Herman and Panksepp, 1978; Kalin et al., 1988; Stein et al., 2007; Wilson and Junor, 2008) and may serve as a substitute for social reward (Machin and Dunbar, 2011). Opioid receptors are widely expressed in regions of the brain involved in emotion processing, including the bed nucleus of the stria terminalis and the amygdala (Mansour et al., 1995; Simonin et al., 1995; Peckys and Landwehrmeyer, 1999). Opioid users have a high prevalence of social phobia, and many individuals report using the drugs to reduce anxiety and ‘fit into’ social settings (Becker et al., 2008; McCabe et al., 2009; Rigg and Ibañez, 2010; Barth et al., 2013). Several neuroimaging and behavioral studies have recently extended our understanding of the effects of opioids on responses to emotional faces in both healthy volunteers and drug users.

Effects of acute administration

The acute effects of opioids on responses to emotional faces have been studied in one recent behavioral study and two imaging studies. Ipser et al. (2013) showed that the mu-receptor partial agonist, buprenorphine (0.2 mg), reduces the accuracy in recognizing fearful faces in healthy volunteers, without affecting the perception of other emotions. Interestingly, these effects were observed in the absence of any drug effect on mood. Schmidt et al. (2013) reported that acute administration of heroin reduced amygdalar responses to fearful faces in heroin addicts, but other emotional expressions were not included in this paradigm. These findings are consistent with other evidence that opioid drugs reduce anxiety and dampen responses to social evaluative threat in humans and potentially social threat in animals (Trezza et al., 2010; Bershad et al., 2015), and with evidence suggesting that activation of the endogenous opioid system is associated with decreased responses to social rejection (Zubieta et al., 2003). In a second imaging study using healthy volunteers, Wardle et al. (2014) reported that the mu-receptor agonist, oxycodone (20 mg), decreased right medial orbitofrontal responses to happy faces. This finding seems to contradict other evidence suggesting that mu-receptor agonists dampen responses to negative social stimuli, but as the authors suggest, the result may be attributable to oxycodone’s secondary actions on the kappa-opioid receptor, which may influence processing of positive emotional expressions. Overall, opioid analgesics appear to reduce behavioral and neural responses to negative emotional expressions.

There is also limited evidence that opioids are involved in the modulation of other emotions. The endogenous opioid system is involved in social reward in laboratory animals (Vanderschuren et al., 1995; Trezza et al., 2010, 2011), and endorphins are released during positive social encounters, such as social touch (Keverne et al., 1989; Odendaal and Meintjes, 2003). Morphine, a mu-receptor agonist, enhances ratings of facial attractiveness (Chelnokova et al., 2014), which is broadly consistent with preclinical literature suggesting that mu-opioid agonists may enhance pleasure in response to rewarding social stimuli (Guard et al., 2002; Depue and Morrone-Strupinsky, 2005).

Effects of chronic use

Few studies have examined the effects of chronic use of opioids on facial processing. Two studies have shown that opiate users currently maintained on methadone and those recently detoxed are more slower to recognize emotions overall than healthy nondependent individuals (Kornreich et al., 2003; Martin et al., 2006). Like many studies addressing the effects of chronic use on behavioral measures, these results may also be confounded by differences in emotion processing that existed before drug use.

Stimulants

Stimulants are typically used in social contexts, and users frequently report using such drugs to enhance social interactions (Díaz et al., 2005). Stimulants such as methamphetamine and d-amphetamine increase measures of social behavior, such as verbal behavior and choice of social over nonsocial rewards (Higgins et al., 1989; Marrone et al., 2010; Wardle et al., 2012), and in addition affect emotional processing. The effects of stimulants on processing of emotional faces have been discussed below.

Effects of acute administration

The evidence from laboratory studies investigating the influence of stimulants on social behavior is mixed. In animals, stimulants such as amphetamine and cocaine reduce some types of social behavior, including play (Beatty et al., 1982; Achterberg et al., 2014). However, several studies in humans have shown that amphetamine derivatives facilitate social interaction, as they increase verbal behavior, and other measures of sociability (Bedi et al., 2010; Ballard et al., 2012; Wardle et al., 2012), as well as reduce amygdalar responses to faces expressing negative emotions (Rasetti et al., 2010). However, some of the differences between human and nonhuman studies may be a result of cross-species dosing discrepancies (Segal and Kuczenski, 2005). A handful of studies have shown that stimulants such as d-amphetamine, caffeine, and modafinil increase the speed of identification of emotional facial expressions, especially for more complex emotions (Huck et al., 2008; Wardle et al., 2012). These findings may provide insight into the inconsistencies in the literature; acute administration of stimulants such as d-amphetamine may decrease the threshold for identifying emotions, which may manifest itself as enhanced positive or negative emotional reactivity depending on the context.

The stimulant 3,4-methylenedioxymethamphetamine (MDMA, ‘ecstasy’) is widely recognized as an ‘empathogen’ and is used recreationally to enhance sociability. Evidence from human laboratory studies indicates that MDMA produces several prosocial effects, including increased desire to socialize, heightened feelings of empathy, and enhanced recognition of positive emotions (Ter Bogt and Engels, 2005; Sumnall et al., 2006; Hysek et al., 2012, 2013; Kirkpatrick et al., 2014; Schmid et al., 2014; Wardle and de Wit, 2014). In an effort to understand how MDMA affects social processing, several studies have examined its effects on neural and behavioral responses to faces. Functional neuroimaging has shown that MDMA reduces amygdalar responses to fearful faces (Bedi et al., 2010). Another study using facial electromyography to assess emotional reactivity to facial expressions found that MDMA enhances responses to happy expressions, in addition to slowing recognition of angry expressions (Wardle and de Wit, 2014). This is consistent with other studies showing that MDMA improves the detection of positive, but not negative emotions (Hysek et al., 2012, 2013; Schmid et al., 2014). However, the doses used in these studies all produced significant subjective reports of euphoria, which makes drug effects on emotion processing difficult to dissociate from effects on mood. Taken together, these studies suggest that acute administration of MDMA acts both to dampen responses to negative emotional expressions and to enhance responses to positive expressions. Such effects may contribute to its reputation as a prosocial drug.

Effects of chronic use

In contrast to the acute effects of psychostimulants on emotion processing, chronic use of drugs such as cocaine and methamphetamine may reduce sensitivity to emotional expressions and dysregulate emotional responses more generally (Kemmis et al., 2007; Kim et al., 2011). Methamphetamine-dependent individuals show alterations in neural responses to emotional faces, such as hyperreactivity in the dorsal anterior cingulate cortex, in addition to reduced activity in the ventrolateral prefrontal cortex (Payer et al., 2008). These individuals also report increased interpersonal sensitivity and hostility. Such changes may underlie disruptions in social behavior that occur during methamphetamine dependence, or may be a result of pre-existing differences in emotion processing in methamphetamine-using populations.

Conclusion

We have reviewed evidence that psychoactive drugs affect emotional processing, both following single doses and after chronic use. The literature reviewed here suggests that drugs may acutely enhance social experiences, but also that prolonged, heavy use can disrupt or negatively bias social processing, thus impairing social functioning. Initial drug effects may facilitate drug taking by enhancing the positive aspects of social stimuli, either lessening threat-related stimuli or enhancing positive stimuli. Among chronic drug users, the deficits in social perception may contribute to the interpersonal problems characteristic of addiction. The directionality of the relationships between social functioning and chronic use is difficult to determine. That is, drug users may have pre-existing social deficits that make them more prone to using drugs to attenuate social anxiety. However, prolonged exposure to the drug may lead to the problems observed in chronic users. More longitudinal work is needed to understand the causality of this relationship.

Most drugs of abuse also produce pleasurable subjective effects (euphoria), which may influence the perception of emotional stimuli, including faces expressing emotions. This raises the question of whether mood effects mediate the changes in perception, or possibly even whether the perceptual changes affect mood states. However, in many of the studies discussed in this review, drug-induced changes in emotional processing were not related to subjective drug effects or changes in mood. This was the case in studies examining alcohol, nicotine, THC, and some opioids. This dissociation suggests that the effects of drugs on responses to social cues are independent of the mood-altering effects of drugs. Whether the effects of drugs on responses to social cues indeed contribute to the reinforcing effects of drugs remains to be determined.

Finally, this review focused on the effects of psychoactive drugs in both healthy adult populations and long-term drug users, with little focus on individual differences that may influence responsiveness to such drug effects. Sex differences, genetic differences, and differences in personality, psychopathology, and social functioning all influence responses to facial cues (McClure, 2000; Hamann and Canli, 2004). Such individual differences may also contribute to the complexity of the effects of drugs on facial processing reviewed here. An important future direction for work in this area is to identify populations that may be particularly sensitive to these drug effects.