Maternal Hospitalization for Infection During Pregnancy and Risk of Autism Spectrum Disorders

Background:  Prenatal infection is a potent environmental influence of early life neurobehavioral function. Animal model studies indicate that maternal immune activation during pregnancy can result in behavioral abnormalities and neuropathologies in offspring. In humans, multiple epidemiological studies point to maternal infections during pregnancy as a risk factor for neurological disorders such as schizophrenia, cerebral palsy, and autism spectrum disorders (ASD).  A large, recent study of over 10,000 ASD cases by Atladottir and colleagues using Danish health register data reported that maternal hospitalization for bacterial and viral infections were associated with increased ASD risk, and that the risk varied by trimester. To date however, no other epidemiological studies have addressed this topic.

Objectives:  To examine the association between maternal hospitalization for infection during pregnancy and overall risk of any ASD.

Methods:  The Stockholm Youth Cohort is a record-linkage study comprising all individuals aged 0-17 years, ever resident in Stockholm County in 2001-2007 (N=589,114).  A sample of 4,435 ASD cases and 43,534 birth year and sex-matched controls were extracted for analysis. Using the identification number assigned to all persons in Sweden, individuals were linked to national and regional data registers. ICD-8, 9, and 10 codes for maternal infections recorded during the pregnancy period in the Inpatient Register were linked with child ASD outcomes as of the December 2007, ascertained from a complete case-finding approach covering all pathways to ASD care in Stockholm County. Conditional logistic regression models estimated odds ratios adjusted for maternal and paternal age, family use of psychiatric services, maternal origin, birth parity, and parental occupation and income.

Results:  Of the mothers of the 4,435 ASD cases, 66 (1.5%) had a primary inpatient diagnosis for infection during pregnancy, while 199 (4.5%) had any inpatient diagnosis for infection during pregnancy. In comparison, mothers of non-ASD cases had a 1.3% prevalence of primary inpatient diagnosis and 3.1% prevalence of any inpatient diagnosis for infection during pregnancy. Adjusted odds ratios for primary inpatient diagnosis and any inpatient diagnosis of infection during pregnancy were 1.14 (95% CI: 0.88 to 1.49), and 1.37 (95% CI: 1.26 to 1.71), respectively.

Conclusions:  In a large, population-based Swedish cohort, we found evidence of an increased risk of ASD with maternal hospitalization for infection during pregnancy, although results were sensitive to whether primary or secondary diagnoses were used. Additional work will be performed to examine sensitivity of results to confounding, trimester-specific associations, as well as by subtype of infection and ASD. These findings provide additional support that maternal infections during pregnancy may increase risk of ASD.