Search

Recent developments suggest that autistic individuals require dynamic angry expressions to have a higher speed in order for them to be successfully identified. Therefore, it is plausible that autistic individuals do not have a 'deficit' in angry expression recognition, but rather their internal representation of these expressions is characterised by very high-speed movement. In this study, matched groups of autistic and non-autistic adults completed a novel emotion-based task which employed dynamic displays of happy, angry and sad point light facial (PLF) expressions. On each trial, participants moved a slider to manipulate the speed of a PLF stimulus until it moved at a speed that, in their 'mind's eye', was typical of happy, angry or sad expressions. Participants were shown three different types of PLFs-those showing the full-face, only the eye region, and only the mouth region, wherein the latter two were included to test whether differences in facial information sampling underpinned any dissimilarities in speed attributions. Across both groups, participants attributed the highest speeds to angry, then happy, then sad, facial motion. Participants increased the speed of angry and happy expressions by 41% and 27% respectively and decreased the speed of sad expressions by 18%. This suggests that participants have 'caricatured' internal representations of emotion, wherein emotion-related kinematic cues are over-emphasised. There were no differences between autistic and non-autistic individuals in the speeds attributed to full-face and partial-face angry, happy and sad expressions respectively. Consequently, we find no evidence that autistic adults possess atypically fast internal representations of anger.

The current study investigates whether autistic and non-autistic adults differ with respect to the precision and/or differentiation of their emotional experiences, and second, assess whether differences therein are related to challenges in accurately recognising emotional expressions. Importantly, in our analyses, we will assess the relative contributions of autistic and alexithymic traits to these abilities in order to identify which characteristics are responsible for any differences found between autistic and non-autistic participants.

The current study investigates whether autistic and non-autistic adults differ with respect to the precision and/or differentiation of their emotional experiences, and second, assess whether differences therein are related to challenges in accurately recognising emotional expressions. Importantly, in our analyses, we will assess the relative contributions of autistic and alexithymic traits to these abilities in order to identify which characteristics are responsible for any differences found between autistic and non-autistic participants.

The autistic population is at an elevated risk of developing co-occurring mental health difficulties, such as depression, anxiety and self-harm. Recent studies estimate the prevalence of self-harm, defined as the intentional injuring or damaging of bodily tissue, to be as high as 24% - 50% among autistic adults. Despite this self-harm is considerably under-researched among autistic individuals. Where research on self-harm has been conducted in autism, the majority of studies focus on samples with co-occurring intellectual disability, despite the fact that 60% of autistic people have borderline, average and above average IQ. Previous studies have also failed to investigate self-harm in specifically autistic young people, despite the onset and peak of self-harm behaviour typically occurring during adolescence and young adulthood. Additionally, the identification of potential risk-markers for the development of self-harm among autistic young people is of high importance for appropriate prevention, intervention and treatment. Therefore, this project aims to be the first to describe the prevalence, profile and risk-markers of self-harm in autistic young people without intellectual disability. Finally, Research in neurotypical populations shows that there is a bi-directional link between young person self-harm and family functioning. However, there is a paucity of research investigating this within autistic samples. This study will therefore also be the first to explore family risk and protective factors for autistic young people who self-harm.

Autism spectrum disorder is a developmental disorder characterised by persistent atypicalities in communication and social interaction, and restricted and repetitive interests (American Psychiatric Association, 2013). While not a diagnostic criterion for autism, differences in mentalising, that is the ability to correctly identify the mental states (e.g. seducing, mocking) of others, are wide-spread and have been argued to lie at the heart of many of the social challenges that autistic people face (e.g., Frith, 2001). Difficulties inferring mental states have been shown in experiments using the Frith-Happé animation task, where participants are asked to make judgements about actions and mental states shown in animated interactions between triangles (Abell et al., 2000; Heider & Simmel, 1944). Abell et al. (2000), for example, found that high-functioning children with autism used mentalising descriptions less frequently than neurotypical eight-year olds when describing animations and that these were often less accurate. A similar pattern has been found using the task with adults, where fewer and less appropriate descriptions of mental state animations are contrasted with descriptions of action words comparable in accuracy to a control group (Castelli et al., 2002). While social difficulties comprise a primary focus of autism research, atypical movement has also been shown to be an important component of autism (Cook, 2016). There is a marked difference between autistic and neurotypical people in terms of movement kinematics, with increased jerkiness being a characteristic feature of autistic kinematics (Cook et al., 2013; Fukui et al., 2018). The presence of atypical movement in autism is important because of the role that movement kinematics play in social cognition. For example, (Edey et al., 2017) found that neurotypical fast walkers rate slower emotions (e.g. sadness) more intensely than faster movements (e.g. anger), due to the fact that slower movements are further away from their own kinematic profile. This has important implications for understanding communication difficulties in autism because it raises the possibility that communicative difficulties between autistic and neurotypical people may not be wholly explained by limited mentalising abilities on the part of the autistic individual. Instead, it may be that the different kinematic profiles of autistic and neurotypical individuals cause difficulties inferring mental states in both directions (Cook, 2016; Edey et al., 2016). In a previous study, we used an adapted version of the Frith-Happé animation task (Abell et al., 2000; Heider & Simmel, 1944) to investigate neurotypical participants’ accuracy in attributing mental states to autistic movements and vice versa. In a modified version of the task, neurotypical and autistic participants were asked to manually direct triangles to generate their own animations depicting mental state and action words. Participants were then asked to judge animations that had been generated by other autistic and non-autistic participants. As predicted, it was found that neurotypical participants were more accurate in attributing mental states to animations created by other neurotypical participants compared to those generated by autistic participants. However, autistic participants did not show a same group advantage, scoring comparatively for both groups (Edey et al., 2016). This finding provides mixed results. On the one hand, it demonstrates that neurotypical individuals are not adept at inferring mental states of people with autism, a finding that is likely due to the atypically jerky kinematics in the autistic animations. On the other hand, if autistic people are no better at inferring mental states in animations made by autistic individuals than by neurotypical individuals, it is not possible to say that atypicalities in mentalising in autism are due to their atypical kinematic profile since, if this were the case, autistic participants should be able to correctly attribute mental states to animations which move ‘like them’. One could argue that our previous results provide strong evidence of mentalising difficulties in adults with autism. That is, autistic participants were less accurate in attributing mental states even when animations moved in a way that was similar to their own movement kinematics. However, an alternative explanation is that there were greater between-individual differences in kinematics in our autism, compared to neurotypical, group; meaning that autistic participants were less likely than neurotypical participants to see animations that truly matched their own kinematic profile. The present study will probe this alternative explanation by employing a completely digital version of the animations task. Our digital version efficiently indexes the jerkiness of animations generated by a participant and intelligently selects stimuli, from a pre-existing database, such that participants are presented with a tailored selection comprising animations that are highly similar, and highly dissimilar, to their own movements. Our digital version therefore enables us to test whether autistic participants are less accurate in attributing mental states even when animations move in a way that is similar to their own movement kinematics.

Autism spectrum disorder (ASD) is clinically characterised by difficulties in social communication and restricted and repetitive interests (American Psychiatric Association, 2013). Despite mixed findings, a large body of research suggests that autistic individuals have difficulties recognising the emotion of other individuals from their facial expressions (see Keating & Cook, 2020 , Uljarevic & Hamilton, 2013 and Harms, Martin & Wallace, 2010 for reviews). Importantly, many of these empirical studies indicate that autistic individuals exhibit emotion-specific difficulties in static emotion recognition, with some studies claiming that the recognition of anger is particularly challenging for autistic individuals (Ashwin, Chapman, Colle & Baron-Cohen, 2006; Bal et al., 2009; Brewer et all., 2016; Song & Hakoda, 2017; Leung et al., 2019). Indeed, more support for this idea comes from meta-analytic evidence, which suggests that there are greater ASD-control group differences in the recognition of angry than happy and sad expressions (Lozier et al., 2014). This is interesting since recent evidence indicates that autistic individuals (relative to non-autistic individuals) require angry, but not happy or sad, expressions to have higher emotional intensity in order for them to be correctly identified (Song & Hakoda, 2018). In this study, static photographic stimuli at varying expressive intensities (constructed by repeatedly morphing a full expression with a neutral expression to result in 9 intensity levels for each emotion) were used to identify the ‘discrimination threshold’ (the intensity at which an expression is identified correctly on two consecutive trials) for autistic and control participants. This study identified that, compared to control participants, autistic individuals had a significantly higher discrimination threshold for angry expressions, meaning that a higher intensity was necessary before an expression could be correctly identified as angry. Importantly, this study also found no significant group-differences in discrimination thresholds for happiness or sadness (Song & Hakoda, 2018). Since these findings suggest that autistic individuals have a different discrimination threshold for static angry expressions, perhaps the same could be true for dynamic angry expressions; however to date, no studies have attempted to test this idea. In our recent study, we used point-light displays (a series of dots that convey biological motion) of the face (point light faces; PLFs) to compare the emotion recognition of autistic and neurotypical individiuals. In the study, the PLFs had been manipulated to achieve three spatial levels, ranging from reduced to increased spatial movement (S1 – 50% spatial; S2 – 100%; S3 – 150%), and three kinematic levels, ranging from reduced to increased speed (K1 – 50% speed; K2 – 100%; K3 – 150%; as in Sowden et al., under review). There were two key findings of our recent study (Keating et al., in prep). Firstly, the autistic participants (relative to controls who were matched on age, gender, non-verbal reasoning and alexithymia) had significantly lower emotion recognition accuracy for dynamic angry (but not happy or sad) expressions moving at a normal (K2 – 100%) speed and with normal (S2- 100%) level of spatial movement. Secondly, whilst for controls, recognition accuracy increased when angry expressions were sped up from 50% to 100% speed and from 100% to 150% speed, the recognition accuracy of autistic participants only increased from 100% to 150% (and not from 50% to 100%). Hence, these findings suggest that just as autistic participants may have difficulties with static angry expressions, they may also have difficulties recognising dynamic angry expressions. Moreover, these findings highlight the possibility that autistic individuals may have a different ‘speed threshold’ for dynamic angry expressions (i.e. an angry expression has to be moving relatively more quickly before it is labelled angry for ASD participants) in addition to differing thresholds for static angry expressions. Therefore, the present study will utilise dynamic PLF stimuli to identify whether autistic and neurotypical individuals possess different speed thresholds for dynamic emotional expressions. One reason that autistic individuals, relative to controls, may have particular difficulty with angry expressions is due to differences in preferential looking. There is some evidence to suggest that autistic individuals tend to avoid looking at the eye region of the face, and instead preferentially look at the mouth region (Klin, Jonres, Schultz, Volkmar & Cohen, 2002; Riby, Doherty-Sneddon & Bruce, 2009; Rutherford, Clements, & Sekluer, 2007; Wolf et al., 2008). Since the majority of the expressive information for angry faces is thought to be conveyed in the upper half of the face (Calder, Young, Keane, & Dean, 2000; Smith & Cottrell, 2005), this could explain why autistic individuals may have particular difficulty with angry expressions. In order to explore whether autistic and neurotypical individuals possess different speed thresholds for the eye and mouth regions respectively, we will also include stimuli displaying just the eye and just the mouth regions (in addition to full face PLF stimuli). Alexithymia When examining the emotion recognition ability of autistic individuals, it is crucial to also consider the role of alexithymia. Alexithymia is a subclinical condition characterised by difficulties identifying, differentiating and expressing emotions (Kooiman, Spinhoven & Trijsburg, 2002). The incidence of alexithymia is high in the autistic population, with 40-65% of autistic adults reaching criteria (Berthoz & Hill, 2005; Berthoz & Frith, 2004) compared to just 13% of the general population (Salminen et al., 1999). The “alexithymia hypothesis” (Bird and Cook, 2013) proposes that difficulties in emotion-processing (including emotion recognition) in ASD are caused by co-occurring alexithymia, rather than ASD itself (Bird & Cook, 2013). There are mixed findings relating to this hypothesis, with some studies finding that there are no group-differences in static facial expression recognition (Cook, Brewer, Shah & Bird, 2013), and others finding that group-differences remain for dynamic expressions (Keating et al., in prep) when autistic and neurotypical participants are matched on alexithymia. Nevertheless, in this study we will aim to match participants groups on alexithymia to ensure that any group-differences that we discover are related to autistic (and not alexithymic) characteristics. The current study Participants will first complete various questionnaires including the Autism Quotient (AQ; Baron-Cohen, et al., 2006), and the Toronto Alexithymia Scale (TAS; Bagby, et al., 1994). Next, participants will complete the PLF slider task. This involves participants moving a slider to manipulate the speed of a PLF video (which either displays the full face, the eye region or the mouth region) until it matches what they perceieve as the speed of a typical angry, happy or sad expression. This task will allow us to identify whether a ‘speed threshold’ exists for dynamic facial expression and compare the thresholds of autistic and control individuals. Finally, participants will complete the Matrix Reasoning Item Bank (MaRs-IB; Chierchia et al. 2019), which assesses non-verbal reasoning.

Parkinson’s Disease (PD) is a neurodegenerative disease of the basal ganglia primarily involving the degeneration of dopamine neurons. Studies have shown that a number of disease processes including neuroinflammation, the clustering and aggregation of the protein -synuclein, as well as the dysfunction of certain organelles such as the mitochondria combine to cause the death of mainly dopaminergic neurons (Bloem et al., 2021) resulting in the classic symptoms of Parkinson’s Disease. There are multiple symptoms of PD and symptoms can be motor or cognitive in nature. Studies have linked dopaminergic neuron death with motor symptoms including bradykinesia, tremor and muscle rigidity (Grosch et al., 2016; Maiti et al., 2017; Surmeier, 2018). Bradykinesia refers to slowed movements, tremor refers to shaking and muscle rigidity refers to the stiffness of muscles (Beradelli et al., 2001; del Rosario Ferreira-Sánchez et al., 2020; Heusinkveld, et al., 2018). In addition to this, a body of evidence has also linked dopamine neuron degeneration in the context of PD to non-motor symptoms such as cognitive impairment (Bloem et al., 2021; Decourt et al., 2021; Wu et al., 2017). Recent reports suggest that in addition to motor and cognitive difficulties people with Parkinson's (PwP) may experience socio-cognitive challenges (Eddy & Cook, 2018). Social cognition difficulties in PD can manifest in a wide range of ways from differences with the ability to recognise the thoughts and feelings of other people to challenges with face perception (Foley et al., 2019; Ricciardi et al., 2017). Over the last 60 years, there has been increasing interest in the recognition of emotions from facial expressions (Bidet-Ildei et al., 2020; Melzer et al., 2019; Tarnowski et al., 2017). Studies show that difficulties with recognising others’ emotions can have a significant and negative impact on PwP’s social relationships, and ultimately on their quality of life (Coundouris et al., 2022; Prenger et al., 2020). PD is diagnosed on the basis of motor symptoms. Currently there are a number of rating scales that can assess what stage of PD an individual is in (Goetz et al., 2008; Hoehn and Yahr, 1967). One of these scales is designed by the International Parkinson and Movement Disorder Society (MDS). These diagnostic criteria can be referred to as the MDS-Unified Parkinson’s Disease Rating Scale (UPDRS) (Goetz et al., 2008; Skorvanek et al., 2015). To be diagnosed with PD, bradykinesia must be present as well as rest tremor and/or muscle rigidity (Marsili et al., 2018). Although motor symptoms are crucial for a diagnosis of PD, non-motor symptoms have also been added into the diagnostic criteria, as there is more awareness surrounding the presence of non-motor symptoms in PD, and how debilitating these can be (Goetz et al., 2008). Non-motor symptoms may include emotion recognition difficulties. The literature suggests that the recognition of negative emotions such as anger and sadness is more difficult for PwP than positive emotions such as happiness. Nevertheless, findings supporting selective difficulties with the recognition of negative emotions in PD are mixed. A review by Argaud et al., collated evidence from 59 papers and made 97 comparisons between typical controls and PwP to surmise whether facial emotion recognition difficulties were seen in PwP, and assessed whether there was evidence of a specific difficulty in each of the six basic emotions. Argaud and colleagues found that the percentage of confirmed difficulties for negative emotions such as anger (44%) and sadness (51%) appear to be higher than the percentage of confirmed deficits for positive emotions such as happiness (27%) and surprise (30%). These percentages refer to the percentage of experiments reviewed that found a facial emotion recognition difficulty in PwP. In line with this, many papers argue that the recognition of angry expressions is particularly impaired in PwP (Clark et al., 2008, Lawrence et al., 2007; Lin et al., 2016, Mioni et al., 2015,). For example, in a study by Lawrence et al., they showed photographs of 10 models displaying six facial expressions (anger, disgust, fear, sadness, happiness and surprise) to 38 participants (17 PwP) and asked them to choose which label (anger, disgust, fear, sadness, happiness and surprise) best related to the photograph they saw displayed. They found that the identification of an angry expression was particularly difficult for PwP who had been temporarily withdrawn from their dopaminergic medication. Thus, illustrating that the recognition of angry expressions was significantly reduced in PwP relative to all other emotions. Surprise and disgust are more controversial, with different papers stating conflicting findings (Clark et al., 2008, Sprengelmeyer et al., 2003). One issue for this field is that studies have typically over-relied on static stimuli (Clark et al.,2008; Lawrence et al., 2007; Lin et al., 2016; Mioni et al., 2015; Sprengelmeyer et al., 2003; Waldthaler et al., 2019). Recent work from our lab has shown that the accurate labelling of others’ facial emotional expression depends on processing both spatial and temporal cues (Sowden et al., 2021). Spatial cues are based on positioning and location of features and temporal cues refer to timing. In addition to this, temporal information has also been found to be important in emotion recognition (Delis et al., 2016). One paper demonstrated that not only are particular movements of the features of the face important, but also the temporal order of these movements (Jack, Garrod and Schyns, 2014). Sowden et al., using point light display stimuli (white dots on a black background in the spatial configuration of the human face), manipulated spatial extent of facial expressions (meaning they changed the coordinates/positioning of the features of the face). This created the impression of spatially reduced and spatially exaggerated facial expressions. They also manipulated the speed of the point light displays of faces that depicted angry, sad and happy emotions. They then investigated if manipulating the speed and spatial extent affected how participants ‘rated’ these emotions on visual analogue scales. They found that increasing the speed of facial expressions increases ratings for happy and angry facial expressions, whereas slowing down the speed of facial expressions inclines people to choose a sad rating (Sowden et al., 2021). Thus, illustrating that in addition to static cues, dynamic cues (such as speed) causally impact on emotion recognition. One study using dynamic stimuli found no significant differences between general controls and PwP in terms of the recognition of the six basic emotions. In this study, which is one of the only so-called dynamic expression recognition in PwP, Coundouris and colleagues used one second videos which began with a neutral stare before progressing into either a high intensity emotion, a low intensity emotion or staying neutral. Following this, participants had to select which emotion each video represented when given a choice between neutral, one of the basic emotions (happiness, surprise, fear, sadness, anger and disgust) or a ‘self-conscious’ emotion (contempt, embarrassment and pride). In contrast to other studies however, they found no significant difference between PwP and general controls in recognising any of the six basic emotions (Coundouris et al., 2022). Dopamine has been implicated in the recognition of facial expressions. In one paper by Sprengelmeyer et al., they found that deficits in facial expression recognition were more pronounced in unmedicated, compared to medicated, PwP. In this study participants were asked to label static facial expression stimuli presented on a computer screen. An Emotion Hexagon was used, which consisted of a series of morphed images. The images were made by ‘blending’ two emotions together in various proportions: 90:10, 70:30, 50:50, 30:70 and 10:90. For example, 90:10 could represent 90% happy and 10% sad (Sprengelmeyer, 1996). The images of the facial expressions were adapted from Ekman and Friesan (1976) and were ordered as follows: happiness – surprise – fear – sadness – disgust – anger (Sprengelmeyer et al., 2003). This order was selected because it groups emotions together which are likely to be confused with one another. It is called an Emotion Hexagon because each of the six emotions, which are grouped in an order as described above, are joined together like a hexagon. These morphed images were individually presented, and participants had to label these images using one of the six basic emotions they thought it represented best. Facial expression recognition for disgust and anger in unmedicated PwP was significantly different from medicated PwP. In a separate study 38 participants were asked to label 60, randomly presented static facial expression stimuli (from Ekman and Friesen, 1976) using one of the six basic emotions they thought it represented best (Lawrence et al., 2007). Lawrence and colleagues found that when PwP were withdrawn from dopamine replacement therapy, the recognition of anger was significantly lower than age-matched controls (Lawrence et al., 2007). These two studies together suggest that dopamine plays a role in the recognition of facial expressions. One point that must be noted is that although we see differences in emotion recognition in PwP, it does not necessarily mean that these differences are due to dopaminergic signalling changes. Currently, there are no mechanisms that explain how dopaminergic signalling changes may influence emotional perception. One recent paper by Schuster and colleagues speculated whether dopamine may affect emotion recognition by influencing temporal perception (Schuster et al., 2022). Therefore, the first study is going to investigate the role of spatial and temporal cues in emotion recognition when PwP are both on and off their dopaminergic medications. To summarise: previous work suggests that PwP experience difficulties with accurately recognising facial expressions of emotion. Preliminary evidence, wherein emotion recognition has been assessed ON versus OFF dopaminergic medication (Lawrence et al., 2007; Sprengelmeyer et al., 2003), suggests that such difficulties may be directly associated with dopaminergic signalling. A second line of work also suggests that such difficulties may be particularly pronounced for negative, as opposed to positive, emotions. To date, however, this literature has largely relied upon static expression stimuli and therefore there is little evidence suggesting that emotion recognition also typically relies upon the processing of dynamic cues, such as movement kinematics (Sowden et al., 2021). To give a brief overview: this study will use a well-validated emotion recognition task (Keating et al., 2021; Sowden et al., 2021). This is the Point Light Faces task which we will run in PwP who are ON and OFF their dopaminergic medication. This will help us to answer important questions such as: what is the role of dopamine in expression recognition? We will also look at positive and negative emotions to further validate whether it is really the case that it is just the negative emotions that are a problem, even when we use dynamic stimuli since this has predominantly been validated in static images (Waldthaler et al., 2019; Lin et al., 2016; Mioni et al., 2015; Clark et al.,2008; Lawrence et al., 2007; Sprengelmeyer et al., 2003). Within this task, we will manipulate speed and spatial extent to investigate which one of those are really feeding into the difficulties seen in PwP. We have a working hypothesis that speed will be especially affected because speed and timing are known to be affected in PwP.