(182s) Metabolites from Blood Samples of Pregnant Mothers Predict Autism Risk

Authors: 
Hahn, J., Rensselaer Polytechnic Institute
Hollowood, K., Rensse
Kruger, U., Rensselaer Polytechnic Institute
James, J., University of Arkansas for Medical Sciences
Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by difficulties in social interaction, verbal and nonverbal communication, and repetitive behaviors [1]. ASD has a high prevalence as approximately one in every 68 children in the United States are estimated to be diagnosed with ASD [1]. This prevalence has significant economic impacts and a recent study found an annual national economic burden of $268 billion resulting from ASD [2].

It is generally acknowledged that the earlier ASD is diagnosed in life, the sooner support services can start, and the better the general outcome will be. However, ASD is currently a clinical diagnosis as no lab test exists. The current average age of diagnosis is 4 years old, but the desired diagnosis would be 18-24 months [3]. One promising approach has found that there are significant differences in measurements of the metabolites of the Folate One-Carbon Metabolism (FOCM) and Transsulfuration (TS) pathways between children with ASD and their typically developing (TD) peers [4]. These differences have the potential for the development of a test to support ASD diagnosis. However, the key question is how early differences could be detected that point towards an ASD diagnosis in the future. Ideally a diagnosis could be made even prior to birth, however, it is unclear if this is possible. This work investigates this point and focuses on metabolite measurements of the FOCM/TS pathways of the mothers during pregnancy as the earliest possible time point of detection.

For this investigation, blood samples were taken at three different time points throughout the pregnancy, during the 1st trimester, 2nd trimester, and 3rd trimester. The samples came from two groups of pregnant mothers: mothers who have already had a child with ASD (High Risk) and those who have not (Control). The data from these mothers are labeled as High Risk and Control because the probability of having a child with ASD across the general population is 1.5% [1], but the probability increases to 18.7% once the mother has already had a child diagnosed with ASD [5]. Furthermore, the High Risk group is divided into two subgroups based upon an evaluation of the children at three years of age. One subgroup consists of the High Risk mothers whose children have been diagnosed with ASD at age three, while the other subgroup contains mothers from the High Risk group whose children do not have an ASD diagnosis by age three. It should be noted that implementation of this study involved tracking of the mothers for the entire pregnancy as well as following up with the mothers at age three of the child to determine if an ASD diagnosis was made, which resulted in a very long and non-trivial to conduct clinical study.

The data collected from this clinical trial were analyzed using multivariate statistical analysis. Specifically, Fisher Discriminant Analysis (FDA) [6] was used to separate the different groups of data. The results of this study show that it is not possible to predict if a child will be diagnosed with ASD from the FOCM/TS metabolites of the mother during pregnancy. However, the results indicate that it is possible to predict with reasonably high probability (misclassification errors of approximately 10%) if a mother falls into the High Risk or the Control group. These results are important insofar as no test for predicting ASD risk exists so far and our results indicate if the ASD risk is the same as the one found in the general population (Control group) or is increased by more than an order of magnitude (High Risk group). While more studies will be needed to replicate these results, the initial findings are promising that the probability of having a child with ASD can be predicted during pregnancy based on the multivariate analysis of FOCM/TS metabolites.

[1] “Facts About ASDs,” CDC - Facts about Autism Spectrum Disorders - NCBDDD. [Online]. Available: http://www.cdc.gov/ncbddd/autism/facts.html. [Accessed: 22-May-2017].

[2] J. P. Leigh and J. Du, “Brief Report: Forecasting the Economic Burden of Autism in 2015 and 2025 in the United States,” J. Autism Dev. Disord., vol. 45, no. 12, pp. 4135–4139, Dec. 2015.

[3] Autism and Developmental Disabilities Monitoring Network Surveillance Year 2008 Principal Investigators and Centers for Disease Control and Prevention, “Prevalence of autism spectrum disorders--Autism and Developmental Disabilities Monitoring Network, 14 sites, United States, 2008,” Morb. Mortal. Wkly. Rep. Surveill. Summ. Wash. DC 2002, vol. 61, no. 3, pp. 1–19, Mar. 2012.

[4] D. P. Howsmon, U. Kruger, S. Melnyk, S. J. James, and J. Hahn, “Classification and adaptive behavior prediction of children with autism spectrum disorder based upon multivariate data analysis of markers of oxidative stress and DNA methylation,” PLOS Comput. Biol., vol. 13, no. 3, p. e1005385, Mar. 2017.

[5] S. Ozonoff et al., “Recurrence Risk for Autism Spectrum Disorders: A Baby Siblings Research Consortium Study,” Pediatrics, vol. 128, no. 3, pp. e488–e495, Sep. 2011.

[6] R. Fisher, “The Use of Multiple Measurements in Taxonomic Problems,” Ann. Eugen., vol. 7, no. 2, pp. 179–188, 1936.

Checkout

This paper has an Extended Abstract file available; you must purchase the conference proceedings to access it.

Checkout

Do you already own this?

Pricing


Individuals

AIChE Members $150.00
AIChE Graduate Student Members Free
AIChE Undergraduate Student Members Free
Non-Members $225.00