Temperament in Individuals with Autism Spectrum Disorder (ASD): A Systematic Review

Background: Individuals with Autism Spectrum Disorder (ASD) vary considerably in the onset and course of their condition, for both core and co-morbid symptoms and, ultimately, in terms of life outcomes. Despite being well-recognized, heterogeneity in ASD is not fully captured by current diagnostic classification systems, nor adequately explained by disorder-specific features. Hence it remains difficult to predict longer-term outcomes on the basis of early presentation. A large body of research outside the context of ASD has demonstrated that particular dimensions of temperament can have both direct and indirect positive and negative influences on development across the lifespan, across a broad range of domains including social, emotional, and behavioural outcomes. Research on temperament in the context of ASD could therefore hold significant potential for furthering our understanding of heterogeneity inherent in this condition.

Objectives: We conducted a systematic review of the literature to integrate existing findings concerning temperament in the context of ASD within a proposed unified taxonomy of overarching temperament traits in order to identify consistencies in the evidence base and gaps which still require attention.

Methods: Medline, PsychInfo and Scopus databases were searched for published articles available through July 2017. Forty studies were identified, including 30 cross-sectional and 10 prospective longitudinal studies.

Results: Current research is complicated by different conceptualisations and measures of temperament across studies. However, when integrated within the proposed unified taxonomy, findings converge to suggest that individuals with ASD may be temperamentally different from both typically developing and other clinical non-ASD groups at a higher-order level. Consistent with research on typically developing children, correlational findings suggest that lower self-regulation and higher negative emotionality are associated with increased internalizing and externalizing problems in ASD samples. Longitudinal studies suggest the possibility that, from as early as 6-months, certain temperament traits can be useful in discriminating among infants who do and do not develop ASD.

Despite promising findings, current research is limited in that it remains largely descriptive, and the nature of the relationship between temperament and core ASD symptoms needs to be explored further. Methodological considerations and directions for future research are highlighted, including the need for more longitudinal investigations into temperament in ASD and the use of more refined statistical techniques. We also suggest that researchers exploring temperament in ASD adopt the proposed overarching temperament dimensions of negative emotionality, surgency and self-regulation in their work to enable generalization of findings across studies.

Conclusions: This field is still in its relative infancy, with many methodological issues to be considered and research gaps to be filled. Nonetheless, work in this area holds promise to further our knowledge of the early developmental pathway/s toward ASD diagnosis, and predictors of outcomes beyond this point.