Examining the Structure of Temperament Traits in Toddlers with and without ASD

Background: Temperament is defined as constitutionally based individual differences in reactivity and regulation (Rothbart & Bates, 2006). Research has highlighted the importance of this construct to social and behavioral outcomes in typically developing (TD) children.  One important measurement of temperament is the Toddler Behavior Questionnaire-Supplemental (TBAQ-S; Goldsmith, 1996). The TBAQ-S includes 13 subscales subsumed by three composites: Attention (including Perceptual Sensitivity, Attention Focusing, Attention Shifting, Inhibitory Control, and Low Intensity Pleasure), Negative Emotionality (consisting of Discomfort, Sadness, Soothability (negatively), Anger, and Social Fear), and Surgency (comprising Social Fear (negatively), High Intensity Pleasure, Activity Level, and Approach/Positive Anticipation; Becken-Jones, Gartstein, Rothbart, & Chasman, 1999). Now scholars have turned to examining temperament in children with autism spectrum disorder (ASD). The TBAQ-S was only validated for use in TD populations, and no study to date has explored the composite trait structure of this measure in toddlers with ASD. Improving our understanding of temperament in special populations may elucidate the relationship between temperamental characteristics and later outcomes and help us identify subgroups of ASD for diagnostic and treatment purposes.

Objectives: To examine the structure of temperament traits as measured by the TBAQ-S in toddlers with ASD compared to their TD peers.

Methods: Participants included 146 toddlers with a diagnosis of ASD, derived from a clinically-referred sample (Mean age = 26.01 months) and a control group of TD toddlers (n = 118; Mean age = 24.16 months). All participants underwent a comprehensive developmental and clinical assessment to confirm diagnostic status. Measures included the Mullen Scales of Early Learning and ADOS-G. Parents completed the TBAQ-S prior to developmental testing. An exploratory factor analysis (EFA) was performed for both groups using principal axis factoring with direct oblimin rotation. Factor scores in ASD were correlated with Mullen and ADOS scores.

Results: EFA revealed three-factor models for both groups. The TD model appeared similar to Becken-Jones and colleagues’ (1999). Loadings were also similar in the ASD model, but some notable exceptions emerged. Perceptual Sensitivity loaded on Negative Emotionality, and Social Fear only loaded on this factor. In addition, Activity Level, High Intensity Pleasure, and Approach/Positive Anticipation loaded together negatively, resulting in an Introversion factor. The Introversion composite was positively correlated with ADOS Severity (r = .289, p < .001) and negatively correlated with nonverbal DQ (r = -.208, p < .05).

Conclusions: The traits measured by the TBAQ-S in toddlers with ASD appear slightly different from those in TD toddlers. Perceptual Sensitivity and Social Fear are both more associated with Negative Emotionality in ASD. Additionally, the most socially relevant scales loaded negatively, suggesting that they represent a more withdrawn pattern of behavior in ASD. This Introversion factor is associated with greater ASD severity and poorer nonverbal functioning. These results suggest that temperament may be organized differently in toddlers with ASD, and these differences may be related to autism symptomatology. Future work should explore how this trait structure may improve our characterization of children with ASD, identify possible subpopulations, and identify other ASD outcomes that may be influenced by temperament.