Cytokine Levels In Amniotic Fluid : a Marker of Maternal Immune Activation In Autism?

Background:   

Converging evidence sheds the light on the important role of immunologic dysfunction in Autism Spectrum Disorders (ASD), and many studies have repeatedly reported abnormal cytokines profiles in ASD patients. To our knowledge, no investigations have been carried out using amniotic fluid samples combined with clinical data regarding maternal immune activation (MIA) during pregnancy.

Objectives:

To assess differences between children with ASD and controls in their amniotic fluid levels of seven MIA associated inflammatory cytokines (Interleukins 1ß, 6, 8 and 18, Interleukin 6 receptor alpha, Tumor Necrosis Factor alpha and Triggering receptor expressed on myeloid cells 1) and the potential role of MIA in the development of ASD. 

Methods:  

We adopted a case-control study design including all singleton offspring born in Denmark from January 1, 1982 through December 31, 2000 with a reported diagnosis of ASD (ICD-8 codes 299, ICD-10 codes F84) in the Danish Psychiatric Central Register and with a corresponding amniotic fluid sample in a historic birth cohort (HBC) kept and maintained at Statens Serum Institute (SSI) in Copenhagen . Controls were randomly selected from the HBC and frequency matched to cases by sex and year of birth. Perinatal data were retrieved from the Medical Birth Register and The Danish National Hospital Register (DNHR). Amniotic fluid samples from 414 singleton cases and 820 singleton controls were analyzed for seven cytokines using Luminex xMAP technology. Case-control differences in biomarker levels were assessed as continuous measures (Tobit censored regression model) or dichotomized at below the 10^th percentile or above the 90^th percentile cutpoints derived from control biomarker distributions (logistic regression). MIA was assessed through utilizing the DNHR for different intrauterine maternal infection/inflammation diagnoses.

Results:

There was a significantly increased risk for ASD overall with elevated TNFα [elevated 90th percentile OR=1.63 (95% CI 1.13 - 2.36)]. This pattern was mainly seen for TNFα in girls [2.85(1.28-6.35)]. TNFα was still significantly different in cases compared to controls when controlling for any intrauterine maternal infections diagnoses. We found no significant association between ASD and maternal intrauterine infections/inflammation retrieved from the DNHR in our study population.

Conclusions:  

Children later diagnosed with ASD are more likely to have levels of TNFα falling in the upper centile of the distribution of this biomarker in children without ASD. This pattern of TNFα may reflect either a response to an adverse environmental stimulus, specifically MIA, which is possibly contributing to the development of ASD or an inherent dysregulation of this immune biomarker later diagnosed with ASD.