طبیعت ترجیحات اجتماعی و تعاملات در سندرم اسمیت

Abstract This natural observation study was designed to evaluate hypothesized elevated ‘attention-seeking’ and preference for adult attention in Smith–Magenis syndrome. Ten children with Smith–Magenis syndrome were observed across one school day, together with an age matched sample of 10 children with Down syndrome. Levels of attention given to, and vigilance for, adults and peers were recorded and compared. Sequences of behaviour were analyzed to evaluate the temporal relationships between giving and receiving attention during adult-child interactions. Compared to children with Down syndrome, children with Smith–Magenis syndrome gave preferential attention to adults and looked towards adults significantly more than they looked towards peers. Sequential analyses revealed that while children with Smith–Magenis syndrome did not initiate interactions with adults more than children with Down syndrome did, reciprocity between child and adult social behaviours in Smith–Magenis syndrome within interactions was compromised. This less synchronous sequence of child and adult interactions in Smith–Magenis syndrome may be the result of children with Smith–Magenis syndrome attempting to initiate interaction at times when it is unavailable. The marked preference for interacting with adults over peers in Smith–Magenis syndrome indicates atypicality of social interaction in this syndrome.

Introduction Smith–Magenis syndrome (SMS) is a rare genetic syndrome with a prevalence rate of approximately 1/25,000 to 1/15,000 births (Greenberg et al., 1991 and Laje et al., 2010). It is caused by a deletion on chromosome 17 p11.2 (Greenberg et al., 1991 and Smith et al., 1986) or, more rarely, by a mutation of the retinoic acid-induced 1 (RAI1) gene located on this chromosome (Slager, Newton, Vlangos, Finucane, & Elsea, 2003). SMS is characterized by distinctive facial appearance, a range of health problems and increased likelihood of behavioural problems including sleep disturbance, stereotyped behaviours, challenging behaviour, impulsivity and attention-seeking (De Leersnyder et al., 2001, Dykens and Smith, 1998, Martin et al., 2006 and Smith et al., 1998). Sleep disorder and challenging behaviour have been the focus of behavioural phenotype research in SMS (e.g. De Leersnyder et al., 2001, De Leersnyder et al., 2003, Finucane et al., 2001 and Martin et al., 2006). Other areas, including attention-seeking and impulsivity, are less well researched despite their potentially critical role in influencing challenging behaviours (Oliver et al., 2013, Sloneem et al., 2011 and Taylor and Oliver, 2008). Research reporting high levels of attention-seeking behaviour has primarily employed caregiver report behavioural checklists, for example the Reiss Screen for Maladaptive Behavior and the Child Behavior Checklist (Dykens et al., 1997 and Dykens and Smith, 1998). While the findings of this research are striking (attention-seeking reported in over 80% of individuals), use of behavioural checklists limits both the scope of behaviours examined and detail. Other descriptions of attention-seeking come from anecdotal accounts. For example Haas-Givler (1994) describes individuals with SMS as attention-seeking at school and ‘very adult-oriented’ with little interest in interacting with peers. They are further described as demanding an ‘inordinate amount of’ and having a ‘sometimes insatiable’ need for individualized attention from adults, with aggression resulting if availability of attention is restricted. While this description gives valuable insight into the nature of attention-seeking, it is limited by its anecdotal nature. High rates of attention-seeking behaviour are likely to have implications both for those with the syndrome and those who care for and work with them. In addition to being reported by caregivers to be problematic in and of itself, links have also been made between attention-seeking and challenging behaviour in SMS. In addition to Haas-Givler's description (1994), suggesting aggression results from lack of adult attention, Sarimski (2004) found aggressive behaviours were reported by caregivers to be motivated by desire for social attention. Two subsequent studies have carried out indirect assessments of functions of challenging behaviour using the Questions About Behavioral Function scale (Matson & Vollmer, 1995) and found that attention maintained challenging behaviour is a feature of SMS. Sloneem et al. (2011) found that for aggressive behaviour the attention sub-scale of the QABF was scored highest by caregivers (above escape, tangible, pain and discomfort and self-stimulation). Subsequent analyses revealed that this function was significantly more common than self-stimulation or pain and discomfort. Langthorne and McGill (2012) found that more children with SMS met the criteria for attention maintained challenging behaviour than children with Fragile X syndrome. Although the authors did not find a within groups difference in function, suggesting that gaining attention is one of a number of functions of challenging behaviour, this finding does indicate that attention is likely to be one function of the difficult behaviours shown by those with SMS. The only study employing direct observations of social behaviour in SMS also supports this association. Taylor and Oliver (2008) observed children at school and found that challenging behaviour was reliably preceded by reduced availability of adult attention. Whilst this study was an important step in examining attention-seeking behaviour in SMS, lack of a contrast group limits whether this behaviour can be considered to be related to SMS. Furthermore, the role of peers, a key aspect of school environments, was not examined. In summary, problematic attention-seeking and difficult behaviour in conditions of reduced attention together with preference for adult attention, are purported to be characteristic of SMS. Alongside causal models of self-injury and aggression in SMS (Oliver et al., 2013, Sloneem et al., 2011 and Taylor and Oliver, 2008) such findings indicate a need to further investigate this behaviour, using objective and reliable methods. While excessive social drive in genetic syndromes is studied infrequently (in contrast to widely studied social deficits, for example in Autism Spectrum Disorder and Fragile X syndrome, Carter et al., 2005, Constantino et al., 2003, Cornish et al., 2007 and Turk and Graham, 1997), there is an increasing body of research specifically examining strong social drive in genetic syndromes which may inform the study of attention-seeking in SMS. A similar strong drive for attention is evident in Angleman syndrome (AS) caused by the loss of genetic information on the maternal chromosome 15 (Brown and Consedine, 2004, Oliver et al., 2002, Oliver et al., 2007 and Strachan et al., 2009). Those with this syndrome are described as highly sociable, actively seeking and maintaining adult interaction more successfully than those without the syndrome (Oliver et al., 2007). They also demonstrate a range of prosocial behaviours including laughing and smiling (Horsler & Oliver, 2006). Oliver et al. (2007) used natural observations of children with AS at school in order to examine social drive in AS, recording child and adult social behaviours and the amount of attention children received. Analyses indicated children with AS smiled more when receiving adult attention and initiated interaction more than a contrast group. Sequential analyses revealed that the child smiled first and adult attention, smiling and eye-contact were more likely to occur after a child with AS smiled than by chance, a pattern not found for the contrast group. This supported assertions that children with AS are more likely to initiate social interactions with ‘prosocial’ behaviours to solicit attentional resources from caregivers (Brown & Consedine, 2004). This methodology is applicable to the study of attention-seeking in SMS as it enables examination of sequences of behaviour and identification of who initiates interaction. In SMS, children reportedly seek attention, thus in sequential analyses it would be anticipated that they would initiate interactions more than a contrast group. In the current study, natural observation methods will be used to examine behaviour of children with SMS at school. This setting was chosen as it is the context in which both adults and peers are present, it has social validity and there is competition for social resources. Levels of attention (verbal and non-verbal social initiations) and looking (which can be considered an index of allocation of attentional resources, e.g. Riby and Hancock (2009) report prolonged face looking in Williams syndrome suggesting “atypical allocation of attention” in this syndrome) directed towards adults and peers, can then be examined to evaluate preference for adult attention. Examination of sequences of behaviour will also be performed, using lag sequential analyses, to evaluate the nature of reciprocal interactions between adults and children and also the association between adult attention and child behaviours. Results will be contrasted to a Down syndrome (DS) sample, selected because of the similar profile of intellectual disability and expressive language deficits (Greenberg et al., 1996 and Chapman et al., 1998). The following hypotheses were tested: (1) Social preference: Children with SMS will have greater preference than children with DS for interacting with and attending to adults versus peers. Differences between child to adult attention (verbal and non-verbal initiations from the child towards the adult) and child to peer attention (verbal and non-verbal initiations from the child towards the peer) and between child to adult looking and child to peer looking, will be greater in SMS than in DS. (2) Sequences of adult–child interactions (i) Children with SMS will initiate and maintain interaction to a greater extent than those with DS. This will be reflected by different patterns of reciprocal social interaction such that child to adult attention is shown earlier in sequences of interaction, i.e. before adult to child attention (verbal and non-verbal initiations from the adult towards the child) occurs, and child to adult attention is shown for longer after adult to child attention has ceased. Additionally, sequence of interactions between adults and children will be examined to explore how adults behave prior to, during and after child to adult attention occurring. (ii) Attentional bias: Children with SMS will show greater vigilance for adult attention as evidenced by looking at adults earlier in an interaction (i.e. before adult to child attention occurs) than those with DS and will continue to look at the adult for longer after adult to child attention has occurred (thus maintaining interaction).

3. Results 3.1. Differences in environmental context As noted more children with DS attended mainstream schools than those with SMS (χ2(2) = 12.80, p < .001), potentially undermining internal validity. Consequently, to ensure that the time the children spent in each social context within the school did not differ, this and the ratios of adults to children throughout the school day were compared across syndromes. Mean times spent in each social context are presented in Table 3. Table 3. Percentage time spent by children with SMS and DS in one-to-one attention, shared attention and free-play social contexts. One-to-one attention Shared attention Free play SMS Mean (SD) 24.42 (16.33) 49.64 (20.03) 27.21 (14.41) DS Mean (SD) 24.57 (10.98) 40.83 (11.04) 34.82 (13.77) Table options There were no between-group differences in amount of time spent in one-to-one attention (t(18) = −1.21, p = .24), shared attention (t(18) = −.18, p = .86) and free-play (t(18) = 1.22, p = .24), suggesting groups were exposed to social contexts that were comparable on critical variables despite differences in schooling. Analyses of ratios of adults to children throughout the school day suggested significantly smaller overall ratio of adults to children for SMS (mean adult child ratio .54) compared to DS (mean adult child ratio .23) (t(18) = 5.53, p = .002). Adult:child ratio was therefore entered into comparisons of child attention and looking to adult and peers as a covariate in order to control for variability in these ratios. 3.2. Between-groups differences in child attention and looking 3.2.1. Attention SMS and DS were compared on differences between child to adult attention and looking and child to peer attention and looking in each social context (one-to-one attention, shared attention and free play). Fig. 1 demonstrates the similarities of patterns of results for attention and looking preferences. Adult–peer difference for adult and peer directed attention (left panel) and ... Fig. 1. Adult–peer difference for adult and peer directed attention (left panel) and looking (right panel) across social contexts for SMS and DS (positive values represent adult preference, negative values indicate peer preference) (±1 standard error). Figure options For attention a main effect of social context was evident (F(2, 16) = 25.62, p < .001). Contrasts indicated that, as expected, preferential attention directed towards adults was greater in one-to-one attention than shared attention (F(1, 17) = 24.71, p < .001) and greater in shared attention than in free-play (F(1, 17) = 5.84, p = .03). There was also a main effect of syndrome (F(1, 16) = 8.87, p = .008), whereby SMS values were consistency higher, denoting greater preferential attention directed towards adults than peers compared to DS, when overall adult child ratio is controlled for. There was no interaction between syndrome and social context (F(2, 17) = .004, p = .996), thus pattern of differences between SMS and DS was independent of setting. 3.2.2. Vigilance for adults and peers For child to adult and child to peer looking a main effect of social context was found (F(2, 16) = 23.68, p < .001), again contrasts showed preferential looking towards adults was greater in one-to-one attention than shared attention (F(1, 17) = 21.32, p < .001) and greater in shared attention than in free play (F(1, 17) = 7.07, p = .017). There was also a main effect of syndrome (F(1, 16) = 7.58, p = .014), with SMS having consistently higher scores, denoting greater preferential looking to adults, compared to DS. Lack of interaction between syndrome and social context (F(2, 17) = 12.98, p = .7) suggests pattern of differences between SMS and DS did not differ across the settings. These analyses indicate that the SMS group showed higher levels of child to adult attention and looking (over peers) across different social contexts than the DS group. For both syndromes preference towards adults was greatest in one to one attention and least in shared attention. 3.3. Sequences of child–adult interactions 3.3.1. Association between adult to child attention and child to adult attention In order to evaluate whether children with SMS initiate and maintain interaction to a greater extent than those with DS the association between child to adult attention and adult to child attention was compared across syndromes. Time based lag sequential analysis was used to examine sequential relationships between adult to child attention and child to adult attention, before (−60 s to −20 s; lag −3 to −1), during (lag 0) and after (+20 s to +60 s; lag 1–3) adult to child attention occurred. Two mixed 3 × 2 ANOVAs were used to compare strength of association (denoted by mean Yule's Q values) between child and adult behaviour across syndrome groups and over time, one ANOVA leading up to when adult to child attention occurred and one after it occurred. A t-test was used to compare the strength of association between child and adult behaviours at the point at which adult to child attention occurred. Fig. 2 (left panel) shows that for both syndromes the association between adult to child attention and child to adult attention increased leading up to when the adult gave the child attention and then decreased following this point. The ANOVA performed for the time period before adult to child attention occurred indicated a significantly stronger association between adult to child attention and child to adult attention overall for DS than SMS, indicated by the main effect of syndrome (F(1, 18) = 5.46, p = .031) and across syndromes the association increased as occurrence of adult attention being given to the child became closer, demonstrated by the main effect of lag (F(1.25, 22.43) = 83.04, p < .001). Syndrome groups did not show different patterns of association over time, only magnitude of association, denoted by lack of interaction between syndrome and lag (F(1.23, 22.43) = .7, p = .44). Association between adult to child attention and child to adult attention (left ... Fig. 2. Association between adult to child attention and child to adult attention (left panel) and association between child to adult attention and adult to child attention (right panel) for SMS and DS (±1 standard error). Figure options At the point at which adult to child attention occurred, groups did not differ in strength of association, t(10.86) = −1.87, p = .09, equal variances not assumed). However, for the period after adult to child attention occurred the ANOVA again showed a stronger association overall for DS than SMS, indicated by the main effect of syndrome (F(1, 18) = 4.41, p = .05) and across syndromes the association between adult to child attention and child to adult attention decreased the further away from the occurrence of the adult giving the child attention, illustrated by the main effect of lag (F(1.24, 22.26) = 57.07, p < .001). There was no interaction between syndrome and lag (F(1.24, 22.26) = .04, p = .90). The lack of a significant interaction in the ANOVAs of the period prior to and after the adult giving the child attention indicates that that although there was a generally weaker association between adult to child attention and child to adult attention in SMS compared to DS there was not a difference in the timings at which the association between these child and adult behaviours changed. Therefore in the period prior to the adult giving the child attention the association between child and adult behaviour increased at a similar rate as the point at which the adult directed their attention to the child became closer, i.e. children with SMS did not start directing their attention to the adult earlier in the sequence, thus children with SMS did not initiate interaction earlier than those with DS. Similarly in the period after the adult gave the child attention the association between adult and child behaviour decreased at a similar rate after the adult gave attention to the child, i.e. children with SMS did not continue directing attention to the adult after attention from the adult has ceased to a greater extent than those with DS, thus children with SMS and DS appear to maintain social interaction to a similar extent. 3.3.2. Association between child to adult attention and adult to child attention As an exploratory analysis the association between child to adult attention and adult to child attention was examined, evaluating adult behaviour directed to the child before, during and after the child directing attention to the adult. Fig. 2 (right panel) shows that the association between child to adult attention and adult to child attention increased up to the point that the child gave attention to the adult and decreased after this point. This association was consistently stronger for DS compared to SMS. The ANOVA of the time period prior to the occurrence of child to adult attention indicated that the association between child to adult attention and adult to child attention was stronger for DS than SMS (F(1, 18) = 4.87, p = .04). Across groups this association increased as the point of child to adult attention became closer (F(1.53, 27.46) = 61.33, p < .001) and lack of interaction between lag and syndrome indicates that the pattern of this increase did not differ between groups (F(1.53, 27.46) = .006, p = .995). This shows that while there was a generally weaker association between child to adult attention and adult to child attention in SMS compared to DS there was no difference in the timings at which this association increased approaching the point at which the child gave attention to the adult. When child to adult attention occurred no difference in strength of association was found between groups (t(18) = −1.65, p = .12). In the period after the occurrence of child to adult attention, the ANOVA showed that the association between child to adult attention and adult to child attention decreased across groups (F(1.33, 23.85) = 96.9, p < .001) and this association was stronger overall for DS than SMS (F(1, 18) = 5.24, p = .03). Rate of decrease was the same across syndromes, indicated by lack of interaction between syndrome and lag (F(1.33, 23.85) = 1.49, p = .24), again suggesting that groups showed similar patterns of results, with no difference in the timings at which the association between these behaviours decreased after the child gave attention to the adult. 3.3.3. Association between adult to child attention and child to adult looking To identify whether children with SMS show increased vigilance for adult attention (demonstrated by looking at adults earlier in an interaction) than children with DS, and whether they look at the adult for longer after the adult has given them attention (potentially maintaining interaction), the association between adult to child attention and child to adult looking was compared across groups before, during and after the adult gave attention to the child. The results of this analysis are shown in Fig. 3. Association between adult to child attention and onset of child to adult looking ... Fig. 3. Association between adult to child attention and onset of child to adult looking for SMS and DS (±1 standard error). Figure options Fig. 3 shows that whilst there was a stronger association between adult to child attention and child to adult looking in DS than in SMS, there is variability in this association. For the ANOVA of the period before adult to child attention occurred a main effect of lag demonstrated increases in the strength of association between onset of child to adult looking and adult to child attention across the groups as the point at which adult attention is given to the child became closer (F(2, 36) = 23.16, p < .001). Groups did not differ in the magnitude of this association, as indicated by the lack of main effect of syndrome (F(1, 18) = .85, p = .37) and a non-significant interaction suggests that the timings at which the association increased in the lead up to the adult giving attention to the child did not differ between syndromes (F(2, 36) = 2.37, p = .11), i.e. children with SMS did not look at the adult earlier in the sequence. At the point when adult to child attention occurred the t-test indicated that groups did not differ in strength of association between adult to child attention and child to adult looking (t(18) = −1.39, p = .18). The ANOVA of the period after adult to child attention occurred also found no overall between syndrome differences in the strength of association between adult to child attention and child to adult looking, demonstrated by the lack of a main effect of syndrome (F(1, 18) = .56, p = .46). A main effect of lag suggests that across groups the association decreased over time after the adult gave attention to the child (F(2, 36) = 5.53, p = 0.008). A significant interaction between syndrome and lag (F(2, 36) = 3.52, p = 0.04) suggests that the groups showed different changes in their looking behaviour over time after adult to child attention occurred. Post hoc analysis of simple effects found no significant difference between groups at any lags; lag 1 (t(18) = −1.33, p = .2), lag 2 (t(18) = −.29, p = .77) or lag 3 (t(18) = −.43, p = .67). Within-groups analysis showed a significant decrease in DS between lags 1 and 2 (t(9) = 4.21, p = .002), but no difference between lags 2–3 (t(9) = −.82, p = .44). For SMS no significant differences were found between lags 1–2 (t(9) = .5, p = .63) or 2–3 (t(9) = −.16, p = .88). These analyses therefore indicate that there were early decreases in strength of association between adult to child attention and child to adult looking in DS which were not evident in SMS, suggesting that those with SMS continued to look at the adult after adult to child attention occurs for longer than those with DS.