S0033291721003172jra 1..11

Background. To understand the emergence of symptoms in autism spectrum disorder (ASD), we need to identify the mechanisms that underpin the development of core social skills. Mounting evidence indicates that young children with later ASD attend less to other people, which could compromise learning opportunities with cascading effects. Passive looking behaviour does not tell us about engagement with visual information, but measures of physiological arousal can provide information on the depth of engagement. In the current study, we use heart rate (HR) and heart rate variability (HRV) to measure engagement with social dynamic stimuli in ASD. Methods. Sixty-seven preschoolers with ASD and 65 typical developing preschoolers between 2 and 4 years of age participated in a study where HR was measured during viewing of social and non-social videos. Using latent profile analyses, more homogeneous subgroups of children were created based on phenotype and physiology. Results. Preschool-aged children with ASD, regardless of their non-verbal, verbal and social competencies, do not differ in overall HR or HRV compared to TD children. However, the ASD group showed a larger increase in HR (more disengagement) than the TD group to later-presented social stimuli. Phenotypic and physiological profiles showed this was primarily the case for children with below average verbal and non-verbal skills, but not necessarily those with more ASD symptoms. Conclusion. Children with ASD, especially a subgroup showing moderate cognitive delays, show an increase in HR to social stimuli over time; this may reflect difficulties re-engaging with social information when attention is waning. Autism spectrum disorder (ASD) is characterised by difficulties in social interaction and communication and the presence of restricted or repetitive behaviours (American Psychiatric Association, 2013) and often co-occurs with language difficulties (Kwok, Brown, Smyth, & Oram Cardy, 2015). Reliable diagnosis of ASD is possible from 24 months (Ozonoff et al., 2015), but often does not occur in the community until childhood or later (Brett, Warnell, McConachie, & Parr, 2016). Identifying robust objective markers of early-emerging social or communication atypicalities may improve this picture. A promising domain is social attention, for which there is mounting evidence of disruption in early ASD (Chita-Tegmark, 2016; Klin, Shultz, & Jones, 2015). Key observations include reduced social orienting, joint attention (Dawson et al., 2004) and attention to eyes (Klin, Jones, Schultz, Volkmar, & Cohen, 2002). Alterations in early attention patterns may have consequences for developing communication skills: for example, social orienting and joint attention are related to language in 1−4-year olds with and without ASD (Dawson et al., 2004) and attention to a person over an object is associated with parent report of both social and communication skills in 2−6-year olds with ASD (Murias et al., 2018). Measurements of looking behaviour can provide insight into what children could potentially take in but provides limited insight into how deeply that information is processed (e.g. Richards, 1997a, b). Previous studies have shown that active engagement ends before visual attention is terminated (Richards, 1997b) and the type of (physiological) attention has an effect on how stimuli are processed (Richards, 1997a). Disruptions in social engagement in autism are already observed in infancy, both on a cognitive and a neural level (Jones et al., 2016). Thus, the depth with which a stimulus is processed and whether it is prioritised for later learning and memory is dependent on the engagement state of the child during viewing. Measurements of the autonomic nervous system like heart rate (HR) or heart rate variability (HRV) can provide critical information about engagement states during visual attention https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291721003172 Downloaded from https://www.cambridge.org/core. IP address: 35.160.27.221, on 23 Apr 2022 at 21:48:05, subject to the Cambridge Core terms of use, available at (Richards & Casey, 1992; Richards & Cronise, 2000; Thayer, Ahs, Fredrikson, Sollers, & Wager, 2012). The autonomic nervous system modulates responses to information and events (Thayer & Lane, 2000), decreasing output in periods of relative calm attention and increasing physiological output in periods of perceived stress. Infants are less easily distracted during periods of slowing HR (Lansink & Richards, 1997), moments of sustained attention when information processing takes place (Frick & Richards, 2001; Richards, 1997a). Calm, attentive states signalled by low HR and high HRV may indicate greater receptiveness to social interaction, with less distress or distraction (e.g. de Barbaro, Clackson, & Wass, 2017; Porges, 2009). Thus, physiological responses can provide information about a child’s internal state and the degree to which they are likely to learn and remember information they experience. A recent meta-analysis showed that in addition to lower HRV at baseline, individuals with ASD also show lower HRV and less HRV reactivity to social stress and cognitive tasks compared to controls (Cheng, Huang, & Huang, 2020). In children, studies comparing responses to social (e.g. child-directed speech) and non-social (e.g. toys) dynamic stimuli have reported similar ‘chronic’ patterns of hyper-arousal that may be apparent across a range of contexts (Vaughan Van Hecke et al., 2009; Watson, Roberts, Baranek, Mandulak, & Dalton, 2012). For example, Watson et al. (2012) found that 2−3-year olds with ASD had increased HR, but no differences in HRV, compared to a typical developing (TD) group during social/non-social videos. Vaughan Van Hecke et al. (2009) reported decreased HRV in 8−12-year olds in ASD compared to TD children during social/ non-social videos. Additionally, children with ASD showed a larger deceleration in HRV to a video of an unfamiliar person, compared to the TD group. Thus, to date there is evidence for increased arousal (which may reflect increased stress and/or decreased attention engagement) more generally in young children with ASD. Such findings can be contrasted with recent evidence of typical physiological activity in a large group of preschoolers with ASD during a less content-rich presentation (watching wildlife videos; Bazelmans et al., 2019). Subtyping of children with ASD may be beneficial to understand conflicting findings. Variability in ASD, both phenotypically and biologically, is high and stratification may help in finding significant effects for subtypes of ASD (Loth et al., 2016). Differences in physiological responses are dependent on context and may vary in directionality and magnitude between young children with ASD. Data-driven approaches can help identify whether there are groups of children that respond in a different way to social and non-social stimuli, e.g. based on intellectual or cognitive ability (see Patriquin, Hartwig, Friedman, Porges, & Scarpa, 2019). Thus, looking at the ASD group as a whole might be masking subgroups of children who do show either chronic physiological differences or differences in response to social and non-social information.