IDENTIFYING PRENATAL, PERINATAL AND ENVIRONMENTAL RISK FACTORS FOR AUTISM SPECTRUM DISORDER (ASD): A CASECONTROL STUDY

Objectives: We undertook a case-control (Case: Control ratio=1:3) study to investigate any association between prenatal, perinatal and environmental factors with the risk of developing Autism Spectrum Disorder (ASD).

Methods: Children with ASD diagnosed in the Child Assessment Centre in Duchess of Kent Children's Hospital during 1999–2005 were recruited. Controls were normal children matched by age, sex and social class. The risk factors were grouped under 4 categories: 1) Demographic characteristics (sex, age, social class, hand dominance, family history of neurological disease; 2) Pregnancy/Perinatal Complications (neonatal jaundice, perinatal asphyxia); 3) Parental Characteristics (occupation and educational status, number of languages exposed); 4) Prenatal Environmental Hazards (maternal exposure to Environmental Tobacco Smoke (ETS), Chinese herbal medication or caffeinated drinks during pregnancy). We used conditional logistic regression to calculate the crude and adjusted odds ratio (OR), Pvalue and 95% confidence intervals (CI). Statistically significant variables were analysed using multivariate logistic regression model according to the likelihood ratios.

Results: 122 children (101 males, mean age=59 months; 21 females, mean age=62 months) and 347 controls (161 males, mean age=58 months; 186 females mean age=61months) were recruited. The risk of having ASD was associated with male (OR=4.4; Pvalue=0.00; CI=1.6–11.8), positive family history of neurological diseases (OR=9.5; Pvalue=0.00; CI=2.7–33.7), neonatal jaundice (OR=3.4; P-value=0.02; CI=1.2–9.5) and Second-hand smoke exposure at home±at work [borderline significance (OR=2.7; Pvalue=0.05; CI=1.0–7.3)]. However, the following factors were not associated with developing ASD: use of Chinese herbal medication or caffeinated drinks during pregnancy (OR=2.0; P-value=0.2; CI=0.7–5.8) or multilingual exposure. Conclusions: Our findings suggest that ASD might be due to interaction of genetic with prenatal/perinatal factors.