31840
Emotion Dysregulation As a Risk Factor for Suicidality: Comparison of Inpatient ASD, Community ASD, and US Census-Matched Youth Samples
Individuals with ASD experience increased rates of emotion regulation (ER) impairments and suicidality, and ER deficits are predictive of suicidal ideation (SI) in neurotypical adolescents. However, neither ER nor SI have been compared across large representative ASD and general samples to elucidate the magnitude of these problems. Further, the association between impaired ER and suicidality in youth with ASD has yet to be investigated.
Objectives:
This study aimed to: 1) compare the prevalence of ER impairment and parent-reported SI across ASD and general US samples; 2) investigate the association between ER and SI across groups.
Methods:
Participants were 6-to 18-year-olds recruited from three sources: 1) 387 psychiatric inpatients with ASD from Autism Inpatient Collection (AIC); 2) 1,209 with ASD recruited via the Interactive Autism Network (IAN) across the US; and 3) 1,000 US-census matched youth recruited through YouGov, a polling company. All parents completed the Child Behavior Checklist (CBCL) item regarding SI (‘talks about killing self’), Emotion Dysregulation Inventory (EDI), that produces theta scores (M= 0 and SD= 1) for Reactivity (EDI-R; poor ER and high emotional intensity) and Dysphoria (EDI-D), and Social Communication Questionnaire (SCQ). In addition to within-group analyses and descriptives, logistic regressions were run predicting SI based on age, gender, race, SCQ, group, EDI scales, and EDI-R/D by group interactions.
Results:
EDI-R scores in the non-ASD sample (YouGov; M= -1.20, SD= .78) were 1SD lower than the community ASD sample (IAN; M= -.30, SD= .85), and 2SD lower than the hospitalized ASD sample (AIC; M= .91, SD= .80; F(2, 2752) = 1029.95, p < .001). EDI-D scores in YouGov (M= -.68, SD= .78) were 0.5SD lower than IAN (M= -.18, SD= .87), and 1SD lower than AIC (M= .55, SD= .83; F(2, 2752)= 344.13, p< .001). Parent report of youth SI significantly differed across groups, (AIC= 27.4%; IAN= 14.3%; YouGov= 4.8%; F(2, 2593)= 44.74, p< .001). Participants with SI had significantly higher EDI-R scores in IAN (t = -.6.06, p<.001) and significantly higher EDI-D scores in IAN (t= -8.00, p < .001) and AIC (t=-2.02, p = .044) than participants from the same source without SI (See Figure 1).
In the model with EDI-R predicting SI, AIC and IAN (β= .24, p< .001; β= .09, p= .010), older age (β= .007, p= .009), and minority race (β= .01, p= .024) were related to a higher likelihood of SI, and there was a significant interaction suggesting a stronger association between EDI-R and SI in IAN (β= .12, p< .001). In the model with EDI-D predicting SI, there were main effects of AIC and IAN (ps< .001), as well as an interaction between both ASD groups and higher EDI-D (AIC: β= .13, p< .001; IAN: β= .12, p< .001).
Conclusions:
Results indicate markedly worse ER and higher SI in community and inpatient ASD samples compared to a non-ASD sample. Poor ER was significantly associated with higher reported SI across ASD groups, which supports poor ER as a risk factor and potential treatment target for suicidality in ASD.