Restricted and repetitive behaviors (RRBs) are part of the core criteria for autism spectrum disorders (ASD). They form a heterogeneous class of behaviors that are characterised by invariant repetition and desire for sameness in the environment (Kanner, 1943). Factor analytic studies using the Autism Diagnostic Interview-Revised (ADI-R) consistently indicate a division into two subclasses: (a) repetitive motor and sensory (RSM) behaviors such as repetitive hand or finger movements and (b) insistence on sameness (IS), including narrow interests, rigid routines, and rituals ( Cuccaro et al., 2003, Honey et al., 2012b, Richler et al., 2007 and Szatmari et al., 2006). Subgroups of RRB have been proposed to represent different neural pathways ( Langen, Durston, Kas, Van Engeland, & Staal, 2011) and show different presentations in early typical development ( Arnott et al., 2010).
Early theoretical accounts considered the use of RRBs to be a coping mechanism for maintaining a homeostatic state of arousal, with RRBs helping to increase sensory stimulation when an individual is under-aroused and reduce stimulation or soothe when over-aroused (Kinsbourne, 1980 and Ornitz and Rivto, 1968; Zentall & Zentall, 1983). Consistent with this explanation, RRBs are often considered to be a marker for anxiety, forming a buffer to alleviate anxiety and distress in a similar way to the role of RRBs in young children (Evans et al., 1997) and compulsions in obsessive compulsive disorder (Zandt, Prior, & Kyrios, 2007).
To date, evidence supports the claim that RRBs are associated with anxiety in individuals with ASD (Joosten et al., 2009, Kamp-Becker et al., 2009, Rodgers et al., 2012, Spiker et al., 2012, Sukhodolsky et al., 2008 and Tantam, 2003). RRBs are also associated with sensory features (Boyd et al., 2009, Chen et al., 2009 and Gabriels et al., 2008), even after controlling for age and IQ (Boyd et al., 2010 and Gabriels et al., 2008). There is also evidence of an association between atypical sensory features and anxiety in ASD (Ben-Sasson et al., 2009, Green and Ben-Sasson, 2010, Green et al., 2012 and Pfeiffer et al., 2005). However, little is known about the particular way in which atypical reactions to sensory stimuli contribute to the relation between anxiety and RRBs in children with ASD. Evidence from toddlers with non-specific ASD (PPD-NOS) suggests that the onset of sensory features developmentally precedes the onset of symptoms of anxiety (Green et al., 2012), but as yet the nature of the three way relation between sensory features, anxiety and RRBs remains to be characterized.
One proposal is that different subclasses of RRBs may function in different ways to either increase or reduce sensory stimulation and anxiety (Leekam, Prior, & Uljarević, 2011). To date only one published study has examined the relation between anxiety and each of the RSM and IS subclasses of RRB. Using the 38-item Spence Children's Anxiety Scale-Parent Version (SCAS-P; Spence, 1998) and the 33-item Repetitive Behavior Questionnaire (RBQ; Turner, 1996) with 8- to 16-year-olds with ASD, Rodgers et al. (2012) found a significant association between total anxiety score and IS. This association was especially strong among children meeting the cut-off score indicating clinical levels of anxiety. However RSM behaviors did not significantly relate to anxiety. More recently, Gotham et al. (2013) also explored the relationship between anxiety and IS in a very large sample of 5- to 18-year-olds, using the anxiety problems score of the Child Behavior Checklist, derived from six items based on DSM-IV criteria for anxiety disorders (Achenbach & Rescorla, 2001), and an IS score based on six items from the ADI-R (Rutter, Le Couteur, & Lord, 2003). Although this study found a weaker relation between IS and anxiety than shown in the Rodgers et al. study, nevertheless a modest and statistically significant relation existed. In this study, the relation between anxiety and RSM was not reported. Similarly, while a study of repetitive motor behaviors in ASD found that an elevated RRB score is associated with both sensory under-responsiveness and over-responsiveness (Gal, Dyck, & Passmore, 2002), this study did not include measures of either anxiety or IS.
The purpose of the current study was to explore the potential contribution made by atypical sensory features to the anxiety-RRB relation in children with ASD. Standard RRB questionnaire measures traditionally include a number of sensory items as these are part of the category of RRBs within diagnostic criteria for DSM-IV-TR (American Psychiatric Association, 1994). To enable comparison with previous research, in Study 1 we tested the structure of a RRB questionnaire, the Repetitive Behavior Questionnaire-2 (RBQ-2; Leekam et al., 2007), which includes a set of sensory items. We then removed these items in a second factor analysis. Study 1 also analyzed associations with age given longitudinal evidence of age changes in childhood using the ADI-R (Richler, Huerta, Bishop, & Lord, 2010). In Study 2, to avoid artificially inflating relations with other sensory measures, we studied the relation between these non-sensory RRB items and anxiety, and also an independent measure of sensory atypicality taken from the Sensory Profile (Dunn, 1999).
