Recognition of emotion from subtle and non-stereotypical dynamic facial expressions in Huntington's disease

Abstract

Emotion recognition deficits in Huntington's disease (HD) are well-established. However, most previous studies have measured emotion recognition using stereotypical and intense facial expressions, which are easily recognized and artificial in their appearance. By contrast, everyday expressions are often more challenging to recognize, as they are subtle and non-stereotypical. Therefore, previous studies may have inflated the performance of HD patients and it is difficult to generalize their results to facial expressions encountered in everyday social interactions. In the present study, we tested 21 symptomatic HD patients and 28 healthy controls with a traditional facial expression set, as well as a novel stimulus set which exhibits subtle and non-stereotypical facial expressions. While HD patients demonstrated poor emotion recognition in both sets, when tested with the novel, ecologically looking facial expressions, patients' performance declined to chance level. Intriguingly, patients' emotion recognition deficit was predicted only by the severity of their motor symptoms, not by their cognitive status. This suggests a possible mechanism for emotion recognition impairments in HD, in line with embodiment theories. From this point of view, poor motor control may affect patients' ability to subtly produce and simulate a perceived facial expression, which in turn may contribute to their impaired recognition.

Introduction

Huntington's Disease (HD) is an inherited neurodegenerative disorder, caused by an expanded CAG trinucleotide repetition on chromosome 4, and is characterized by progressive motor dysfunction, cognitive decline, and affective disturbance (Walker, 2007). HD is associated with neuronal loss within corticostriatal circuits (Lawrence, Sahakian, & Robbins, 1998), as neurodegeneration originates in the striatum, a component of the basal ganglia (Aylward et al., 2000, Paulsen et al., 2010). Dysfunction and death of neurons in other brain regions, including the cerebral cortex, subcortical white matter, amygdala, insula, thalamus and hypothalamus, are also evident, though atrophy in these regions is less severe than in the striatum (Dogan et al., 2013, Dumas et al., 2012, Ross and Tabrizi, 2011).

Patients are diagnosed with HD based on the presence of unequivocal motor signs, particularly chorea, in conjunction with a positive genetic testing (McColgan & Tabrizi, 2018). This diagnosis, based on the manifestation of motor symptoms, represents the onset of the disease. The manifest period is preceded by a pre-manifest period, which can be further subdivided into two clinical stages: at the first stage, pre-symptomatic HD gene carriers are clinically similar to typical controls (Ross et al., 2014). However, around 10–20 years before disease onset, subtle cognitive and behavioral alterations are frequently demonstrated (Paulsen et al., 2008, Scahill et al., 2013). This second stage of the pre-manifest period is called the prodromal stage of HD (Ross & Tabrizi, 2011).

Amongst the cognitive-affective symptoms of HD, patients sometimes experience deficits in social cognition and emotion processing (Bora, Velakoulis, & Walterfang, 2016). These deficits include social-neuropsychiatric symptoms such as apathy and disinhibition (Kempnich et al., 2018), alterations in the way emotions are experienced (Ille, Holl, Kapfhammer, Reisinger, Schäfer & Schienle, 2011), emotional blunting (Craufurd, Thompson, & Snowden, 2001), poor theory of mind (Eddy & Rickards, 2015), impaired emotion regulation (Zarotti, Fletcher, & Simpson, 2018), and a tendency to draw faulty inferences from social situations (Snowden et al., 2003). As a result, individuals with HD commonly experience reduced social quality of life, including social isolation, social withdrawal, disrupted relationships and disturbances in social interactions (Kempnich et al., 2018).

Specifically, HD patients are impaired in their ability to recognize negative facial expressions (Bora et al., 2016). This well-established reduction in negative emotion recognition in symptomatic HD patients is already detectable at the prodromal stage of the disease, in which performance is slightly impaired (Bora et al., 2016, Johnson et al., 2007, Stout et al., 2011). Early at the prodromal stage of the disease this finding is less consistently demonstrated, and is sometimes limited to some of the negative emotions (Henley et al., 2012, Zarotti et al., 2018). With disease progression performance worsens, and emotion recognition deficits become salience and robust among manifest patients (Johnson et al., 2007, Tabrizi et al., 2009).

The reduced negative emotion recognition in HD patients is correlated with reduced grey matter volumes in the insula, orbitofrontal cortex (OFC), hippocampus (Ille et al., 2011), and different regions of the occipital lobe (Scahill et al., 2013). The insula and the OFC are associated with emotion processing (Sprengelmeyer, Rausch, Eysel, & Przuntek, 1998), while the occipital lobe is involved in tasks requiring the processing of visual stimuli (Scahill et al., 2013). However, there is conflicting evidence for association between negative emotion recognition performance and striatal atrophy (Scahill et al., 2013, Johnson et al., 2007, but see; Harrington et al., 2014, Henley et al., 2008).

Understanding emotional disturbances in HD is of particular interest. First, since the ability to recognize emotions is a key social skill, there is a strong impact of an impairment in this domain on patients' and their relatives' quality of life. Second, early identification of symptom onset can facilitate symptomatic management and other interventions that may improve patients' quality of life at the prodromal stage and at early stages of manifest HD. Medical care may include, among others, diagnostic evaluation, management of symptoms (for example, addressing early signs of chorea, sleep disturbance, depression and anxiety), and an emphasis on good nutrition and exercise (Nance, 2007). Furthermore, identification of subtle emotional and cognitive symptoms, before considerable motor signs, increases the predictability of the disease onset. This suggests that a treatment that may delay or even prevent the onset of manifest disease could be possible in the future (Ross & Tabrizi, 2011). Therefore, the interest in emotion recognition deficits as an early sign of the prodromal stage in HD gene carriers has grown increasingly over the last years. Third, from a neurobiological point of view, the associations between the behavioral deficits and the affected brain regions in HD may contribute to emotion recognition models by elucidating the underlying mechanisms and brain structures involved in the processing of facial expressions (Henley et al., 2012).

While emotion recognition deficits among gene carriers and HD patients are well documented, their underlying cognitive mechanisms remain unclear. A key question regarding these mechanisms is whether the impairment represents a specific emotion recognition deficit, or rather, if it occurs in the context of a more widespread cognitive impairment (Bäckman et al., 1997, Paulsen, 2011). Patients demonstrate deterioration of general cognition across a wide range of cognitive functions, including attention, working memory, executive functions and inhibitory control (e.g., Duff et al., 2010, Lawrence et al., 1998, Montoya et al., 2006). At the prodromal stage of the disease, a subtle cognitive decline is detectable around 10–20 years before the predicted disease onset, which is best detected by the Symbol Digit Modalities Test, a sensitive task requiring visual attention and psychomotor speed (Tabrizi et al., 2013). In general, tests with a substantial motor or psychomotor component tend to be more sensitive at the prodromal and early stages of HD, emphasizing the interaction between motor and cognitive aspects of the disease (Ross et al., 2014). Reduced performance in emotion recognition tasks may be a sequela of this cognitive deterioration.

Another possible mechanism that may underlie emotion recognition deficits in HD is related to the motor deficits in the disease. According to embodiment theories, which emphasize the link between action and perception, there is a reciprocal relationship between the bodily expression of emotion and the manner in which emotional information is interpreted (Niedenthal, 2007). Numerous studies have shown that while watching a facial expression, people subtly produce the perceived facial expression themselves (Hess & Fischer, 2013). This sensorimotor simulation of the perceived facial expression is believed to be an important source of information which facilitates emotion recognition (Wood, Rychlowska, Korb, & Niedenthal, 2016). Hence, if the motor system is impaired, one area that may be affected is the ability to perceive and decipher facial expressions (Kordsachia, Labuschagne, & Stout, 2017).

In the present study we aimed to examine emotion recognition in HD patients, using a novel approach to emotion recognition testing. The vast majority of previous studies examined emotion recognition in HD patients with static images of stereotypical and highly intense facial expressions from sets developed or inspired by the ‘Basic Emotion’ approach (Bora et al., 2016, Ekman and Friesen, 1976, Johnson et al., 2007). By contrast, our working assumption was that real-life emotional cues are dynamic, and often also subtle and non-stereotypical (Gaspar et al., 2014, Jack et al., 2014, Yitzhak et al., 2017).

Thus, in contrast to the mainstream method of examining emotion perception using static stereotypical facial expressions, our first goal was to examine emotion recognition using dynamic facial expressions that are also low in intensity and stereotypicality. To this end, we used two different facial expression datasets containing dynamic facial expressions, as facial configuration changes from neutral to emotional in a few seconds video clip. The first, consisting of intense and stereotypical emotional facial displays (ADFES: Van Der Schalk, Hawk, Fischer, & Doosje, 2011). The second, a novel dataset consisting of subtle emotional facial displays conveyed in an unconstrained manner (JeFEE: Yitzhak et al., 2017). Unlike sets in which the actors were strictly instructed and trained to activate specific facial muscles to convey stereotypical configurations (e.g., Ekman and Friesen, 1976, Van Der Schalk et al., 2011), in this set the actors had the liberty of using any facial muscles they wished. When validating this set (Yitzhak et al., 2017) it was shown that the JeFEE subtle expressions were not merely low intensity stereotypical expressions, but rather that they included different muscular cues, compared to the stereotypical expressions, when conveying the emotions. The JeFEE subtle stimuli were also rated as more ecologically looking and judged as better representations of facial expressions in every-day life, compared to the highly stereotypical stimuli (Yitzhak et al., 2017). Therefore, patients' performance in recognizing these facial expressions may be a more valid measure regarding their difficulties in real-life social interactions.

Aside from improved ecological validity, using subtle and non-stereotypical expressions may also carry the benefit of detecting slight deficits in emotion recognition. Specifically, while emotion recognition deficits are well established in symptomatic HD patients, among prodromal HD gene carriers this deficit appeared to be less pronounced and less consistently demonstrated (for a meta-analysis, see Bora et al., 2016). These findings indicate that at the prodromal period of HD, facial expression recognition is not a very sensitive measure for the detection of early symptoms, as it is at later stages of the disease. Since recognizing subtle non-stereotypical expressions is more challenging compared to the highly recognizable stereotypical stimuli, early impairments of HD may be easier to detect using such an emotion recognition task.

In line with previous studies, we predicted that HD patients' emotion recognition will be impaired compared to control participants. However, we expected this impairment to be more pronounced for the subtle, non-stereotypical stimuli in comparison to the highly intense and stereotypical stimuli.

In addition to characterizing emotion recognition in HD patients and controls, our second goal was to examine how different factors, such as cognitive status, motor symptoms, depression symptoms and an estimation of HD pathology progression, are correlated with emotion recognition decline in HD patients. Particularly, we aimed to compare two possible predictors of emotion recognition performance: general cognitive screening and severity of motor symptoms. The predictive power of these two factors on emotion recognition performance can shed light on its underling mechanism. On the one hand, if the cognitive screening scores are strongly associated with emotion recognition, it will reinforce the assumption that the reduced performance in emotion recognition task is one aspect of more general cognitive deterioration. On the other hand, if the motor symptom severity is strongly correlated with the accuracy of facial expression classification, it will reinforce the action-perception hypothesis, suggesting that emotion recognition deficits in HD patients are rooted in impaired motor control.