Longitudinal Predictors of Co-Occurring Anxiety in Children at Increased Familial Risk for ASD

Background:  

Co-occurring anxiety is highly prevalent in individuals with ASD and their family members. Yet the shared neurocognitive correlates and early childhood predictors of these two conditions remain unclear. Research on the predictors of anxiety in children without ASD suggests that early difficulties with attentional flexibility and regulation of temperament may be markers of later development of anxiety.  However, there is presently a scarcity in longitudinal research exploring the pathways leading to the development of anxiety in individuals with ASD.

Objectives:  

This prospective longitudinal study aimed to explore the symptoms and longitudinal predictors of anxiety in children at increased familial risk for ASD. The aims of the study are to examine if 1) increased anxiety is observed among children at high-risk for ASD; 2) there are differences in anxiety symptoms among high-risk children who meet diagnostic criteria for ASD and those who do not; 3) difficulties with attentional disengagement and temperament in early life are associated with the development of anxiety within high-risk children. 

Methods:  

Symptoms of anxiety were measured at age 6-8 in a cohort of children at high-risk (HR) for ASD, who have been studied prospectively since infancy. Anxiety symptoms were measured using the Spence Children’s Anxiety Scale parent-report (SCAS-P). Anxiety symptoms were compared between high-risk children who met DSM-5 criteria for ASD (HR-ASD; n=15), high-risk children without ASD (HR-no ASD; n=27) and low-risk controls (LR; n=37). Disengagement time on the Gap/Overlap task measured at 7m, 14m and 36m was used to explore the relationship between early attention and later anxiety. Finally, early temperament was measured using parent report at 7m, 14m, 24m and 36m with the Infant Behaviour Questionnaire (IBQ), Early Childhood Behaviour Questionnaire (ECBQ) and Child Behaviour Questionnaire (CBQ).  

Results:  

The HR participants exhibited higher anxiety than LR across SCAS-P subscales. Furthermore, the HR-ASD group had significantly higher anxiety compared to LR, and there was a trend for higher anxious symptoms in HR-no ASD than LR (Figure 1). Within the high-risk group, disengagement times on the Gap/Overlap task at 14m and higher levels of Negative Affect (NA) in toddlerhood were associated with increased anxiety at 6-8 years (Table 1).

Table 1 Associations between anxiety, attentional disengagement and temperament

SCAS-P subscale	Negative Affect 7m	Negative Affect 14m	Negative Affect 24m	Negative Affect 36m	Disengagement 14m
Separation Anxiety	.13	.14	.19	.20	.22
OCD	-.05	.24	.40*	.38*	.22
Panic/Agoraphobia	.04	.22	.54**	.41*	.06
Physical Injury Fears	.10	.13	.29	.18	.36*
Generalised anxiety	.14	.32*	.46*	.42*	.24
Social phobia	.13	.39*	.54**	.37*	.22
Total anxiety	.11	.31*	.51**	.41*	.28

*p<.05 **p<.01

Conclusions:  

Children at risk for ASD, particularly those who develop the condition themselves, exhibit increased anxiety. Among HR children, difficulty regulating attention and negative emotionality in infancy and toddlerhood are associated with the development of anxiety. This suggests that the pathways to developing anxiety are similar within children with ASD and their siblings as they are in non-ASD populations.