Aims To determine whether adverse childhood experiences (ACE) influence child behaviour and correlation with final diagnosis of ASD in a cohort of children referred for a diagnosis of autism spectrum disorder (ASD).

Methods We did a retrospective analysis of 156 children referred to a district level child development centre for assessment of ASD. Correlation by means of univiarate chi squared tests using SPSS of behavioural features (sensory difficulties, poor sleep, loses temper, physical aggression and sad or self-harm) in relation to potential predisposing factors including: age, sex, gestation, epilepsy, chromosomal defect, presence of developmental delay, final diagnosis of ASD and ACE in the form of maternal depression, domestic violence and child protection plan (past or present). Source of information was electronic records (SystmOne), where the evidence for domestic violence and the child protection plan was derived directly from police and social care.

Results The average age of the children at the time of referral was 6.5 (median 5.4, range 1–15) years, 40 (26%) were girls, only 3 had been born pre-term (<36 weeks). 115 had been through the complete ASD diagnostic process and 63 (55%) had received a diagnosis of ASD. In relation to ACE; 12 (8%) had or had had a protection plan, 19 (12%) had experienced domestic violence and 9 (6%) maternal depression.

Significant associations were: domestic violence with temper (p=0.037) and aggression (p=0.030), maternal depression with aggression (p=0.049), child self-harm with temper (p=0.001) and aggression (p=0.04), sleep difficulties with aggression (p=0.002), developmental delay with temper (p=0.012). The diagnosis of ASD correlated with sensory difficulties (p<0.001) and developmental delay (p=0.014). Many aggressive children received medication (p=0.003).

Conclusion The results show exposure to domestic violence and maternal depression had an impact on behaviour but not on the diagnosis of ASD which was strongly correlated with developmental delay. Computerised medical records systems deriving information from a variety of sources can reveal the impact of ACE on child behaviour. It is important two of the three ACE were recorded independently of parental history. These relationships, using these types of information sources, deserve further systematic prospective study with larger cohorts.