Effect of Maternal Diagnostics Received During Pregnancy on Autism Spectrum Disorder of the Child

Introduction: Autism spectrum disorder (ASD) is an early on-set neurological disorder that affects 1 in every 59 eight-year-old children in the United States [1]. Current etiological understanding is a combination of genetic and environmental effects that contribute to the development of ASD. Maternal events that occur during pregnancy have been found to influence the risk of ASD of the child [2]. Many studies have found that maternal infections during pregnancy increase the risk of ASD [3] though there are contradictions about the type of infection. An increased risk has also been found for autoimmune diseases or disorders [4], polycystic ovary syndrome [5], rheumatoid arthritis [6], but also with delivery related events such as preterm delivery [7] and cesarean section [8].

Methods: A retrospective analysis was done using medical claims data from the OptumLabs® Data Warehouse (OLDW). Children having ASD and children without ASD (general population, or POP) were previously identified [9] and linked to the mothers in this study through OLDW family identifiers and delivery claims. 478 diagnostic conditions were identified by their International Classification of Disease coding, version 9 (ICD-9). Diagnoses were included in analysis if at least two percent of the mother cohorts contained a claim for that diagnosis. The difference in diagnostic claims between mother cohorts were investigated to assess the risk of ASD in the child. Logistic regression was used to evaluate the unadjusted and adjusted odds ratios using a 95% confidence interval.

Results: 164,003 mothers were included in the analysis, 1,807 (1.1%) of which had children later diagnosed with ASD. After screening, 82 diagnostic claims were included in the analysis. While the POP cohort is 1,000-fold larger in size, both cohorts exhibit similarities in their distributions, including a positive-skew. The three most common non-pregnancy related diagnostics were special investigation and examination, special screening for malignant neoplasms, and carrier or suspected carrier of infectious diseases (ICD-9 V72, V76, V02 respectively). 21 of the 84 diagnostics included in the analysis were found to have a significant effect on the risk of the child later being diagnosed with ASD. A diagnosis of depression (ICD-9 311) during pregnancy was found for 3,218 POP mothers and 65 ASD mothers (2.0% and 3.6%, respectively) and showed a significant increase in risk for ASD (adjusted odds ratio 1.72; 95% confidence interval 1.28, 2.28). 18,851 POP and 156 ASD mothers (11.6% and 8.6%, respectively) received a claim for vaccination or inoculation against certain viral diseases (ICD-9 V04) which was associated with an approximate 30% decrease in ASD risk (adjusted odds ratio 0.71; 95% confidence interval 0.59, 0.84). Maternal age above 35 along with other indications for care or intervention related to labor and delivery (ICD-9 659) was diagnosed for 66,792 POP mothers and 885 mothers of children with ASD (41.2% and 49.0%, respectively) and increased the risk of the child having ASD (adjusted odds ratios 1.27; 95% confidence interval 1.15, 1.41). Complications of labor and delivery (ICD-9 669), including an unplanned cesarean section, was diagnosed for 25,828 POP mothers and 374 ASD mothers (15.9% and 20.7%, respectively). This diagnostic was associated with an 17% increase in risk (1.17 adjusted odds ratio; 95% confidence interval 1.03, 1.33). 43,306 POP mothers and 582 mothers of children with ASD (26.7% and 32.2%, respectively) were diagnosed with early or threatened labor (ICD-9 644), a diagnostic associated with preterm infants, which was associated with an 11% increased risk of ASD (adjusted odds ratio 1.11; 95% confidence interval 0.99, 1.24).

Conclusions: Multiple factors in this highly inclusive study were found to have a significant risk on the child subsequently being diagnosed with ASD. Future studies may be used to uncover the biological effect these conditions have on the development of ASD in prenatal children. It is also of interest to segment medical claims by trimester to evaluate the effect of timing for each diagnostic event. Also, the analysis of pharmacy prescriptions within the medical claims data and their effect on the outcome of ASD in the child should be investigated in the future.

References:

1. Baio J. MMWR Surveill Summ. 2018;67:1-23.
2. Kim JY. Lancet Psychiatry. 2019;6(7):590-600.
3. Jiang H. Brain Behav Immun. 2016;58:165-172.
4. Hughes HK. Front.Cell Neurosci. 2018;12.
5. Katsigianni M. Mol Psychiatry. 2019.
6. Rom AL. J Am Acad Child Psy. 2018;57(1):28-31.e1.
7. Chen L. Mol Autism. 2019;10(1):32.
8. Zhang T. JAMA Netw Open. 2019; 2(8):e1910236.
9. Vargason T. J Autism Dev Disord. 2019;49(2):647-659.

Acknowledgements: The authors gratefully acknowledge John Rodakis, the founder and CEO of N of One, for supporting the interactions with OptumLabs, and to the staff at OptumLabs for supporting this study design.