Risk Factors for Self-Injurious Behavior Among Infants at-Risk for Autism Spectrum Disorder

Background: Lifetime prevalence of self-injurious behavior (SIB) is approximately 50% among individuals with autism spectrum disorder (ASD). Once entrenched, SIB can be difficult and costly to treat. Commonly identified risk factors for this behavior include a diagnosis of ASD, lower intellectual functioning, communication impairment, and presence of motor stereotypy (Baghdadli et al., 2003; Oliver & Richards, 2015). To date, there is little empirical research utilizing a prospective research design to identify risk factors for SIB during the first years of life.

Objectives: Our aims were two-fold: 1) investigate the relationship of early SIB to a diagnosis of ASD; and 2) identify early risk factors for SIB. 

Methods: Participants were from a longitudinal study of infants at familial risk for ASD. The present sample included 237 high-risk infants who completed the following assessments at ages 12 and 24 months: MSEL, Vineland-II, and Repetitive Behavior Scales-Revised (RBS-R). The RBS-R was used to characterize SIB and motor stereotypy. The study sample was 62.9% male. Mean ages at 12 and 24 month assessment dates were 12.54 (SD = 0.62) and 24.82 (SD =1.47), respectively. Logistic regression was used to evaluate potential 12 month predictors of SIB at age 2.

Results: SIB was reported for 32% of participants at age 24 months and was significantly higher among children also meeting diagnostic criteria for ASD (χ² = 4.30, p =.04). The risk of engaging in SIB at 24 months was 1.85 times higher among children who were later diagnosed with ASD compared to children with no diagnosis. The first logistic regression model included sex, total composite scores from the MSEL and Vineland-II, and SIB and stereotypy from the RBS-R at 12 months. The overall model significantly predicted 24 month SIB (χ² = 31.2, p < .001, R²_pseudo = 0.22). Of individual predictors, Mullen composite score and presence of SIB at 12 months were significantly associated with SIB at age 24 months. A second model excluding Vineland composite and RBS-R stereotypy score was fit. The overall model was significant (χ² = 34.8, p < .001, R²_pseudo = 0.19). For participants who exhibited SIB at 12 months, the odds of SIB at 24 months increased by 93%. Odds of SIB increased by 3% for each point decrease in MSEL composite score. 

Conclusions: The purpose of this study was to evaluate behavioral characteristics predicting early SIB among children at high familial risk for ASD. SIB was more prevalent among those children receiving a diagnosis of ASD but was not exclusive to this subset of high-risk infants. Logistic regression results indicated that presence of SIB and lower intellectual functioning at age 12 months significantly predicted later SIB. In contrast with previous studies, motor stereotypy was not associated with SIB in our sample. This may indicate a developmental relationship between stereotypy and SIB which, early in life, qualitatively differs from that observed in older individuals. Identifying risk factors for the early development of SIB has the potential to inform prevention and early intervention programming.