Neglected Dimension: Regulation of Affect and Attention in Toddlers with ASD

Background: Temperament, biologically-based behavior style reflecting regulation of affect, attention, and activity, has been studied relatively infrequently in children with autism, particularly in toddlers (Garon et al., 2009; Zwaigenbaum et al., 2005). The assessment of temperament may provide novel dimensions on which to characterize young children with ASD, enhancing our understanding of the heterogeneity of the autism spectrum.

Objectives: To examine temperament in toddlers with autism spectrum disorders (ASD) including autism (AUT) and pervasive developmental disorder-not otherwise specified (PDD-NOS), and in two control groups: toddlers with non-autistic developmental delays (DD) and typical development (TD).

In addition, we explored associations among temperament traits, verbal/nonverbal ability, as well as severity of autism symptoms.

Methods: 180 participants (AUT, n = 60; PDD-NOS, n = 33; DD, n = 26; TD, n = 61, mean age 26 months) were assessed by a multidisciplinary team.  Diagnostic classification was based on outcome diagnosis at 36-48 months.  Measures included the Toddler Behavior Assessment Questionnaire–Supplemental (TBAQ-S; Goldsmith, 1996; Becken Jones, 1999), the Mullen Scales of Early Learning (Mullen, 1992), and the ADOS-G (Lord et al., 2000).

Results: The temperament scales that differentiated toddlers with AUT and PDD-NOS from the other groups were Inhibitory Control, Positive Affectivity, Low Pleasure, and Perceptual Sensitivity, with lower scores for the children with ASD. Scales that differentiated AUT from PDD-NOS, DD, and TD were Attention Shifting and Soothability, whereby the children with AUT received the lowest ratings.  Attention Focusing differentiated all three clinical groups from the TD toddlers, who were rated highest.  Social Fearfulness differentiated only the AUT group from the TD group, who were rated with highest levels of social fearfulness (all p <. 05, using ANOVA with post-hoc comparisons, with Bonferroni correction).    Associations between temperament features and measures of symptom severity and cognition in the ASD group were negligible. Only the Attention Shifting scale was correlated with severity of social-affective impairment on the ADOS (r=-.31, p<.01) as well as restricted-repetitive behaviors (r=-.26, p<.01).   Scores on Soothability were correlated with restricted-repetitive behaviors (r=-.25, p<.05).

Conclusions: Parents rated toddlers with ASD (autism and PDD-NOS combined) as exhibiting less excitement and enthusiasm for positive events, and as less able to regulate their behavior in response to adult instruction compared to non-ASD controls.  Toddlers with autism had greater difficulties regulating their arousal and shifting attention between tasks flexibly compared to toddlers with PDD-NOS and the non-ASD groups. Difficulties in sustaining focused attention and higher social fearfulness were not specific to ASD, as the toddlers with DD were rated similarly. Taken together, these results suggest that toddlers with ASD show a decreased ability to regulate their arousal, attention, and emotional responses.  However, the vast majority of parent-reported temperament features were not associated with symptom severity or DQ.  Thus, our findings suggest that temperament may constitute a dimension that is relatively independent from the other phenotypic features and is likely to contribute to the heterogeneity of early syndrome expression and affect the child’s amenability to treatment and outcomes.