Maternal Spontaneous Abortion and the Risk of Attention-deficit/Hyperactivity Disorder in Offspring

A Population-based Cohort Study

Hui Wang; Fei Li; Maohua Miao; Yongfu Yu; Honglei Ji; Hui Liu; Rong Huang; Carsten Obel; Jun Zhang; Jiong Li


Hum Reprod. 2020;35(5):1211-1221. 

In This Article


There were 130 206 (12.2%) children born to mothers who had at least one SA. Among those, 112 995 (10.6%) children were born to mothers with one SA before childbirth, and 17 211 (1.6%) children were born to mothers with at least two SAs before childbirth (Figure 1). Table I presents the baseline characteristics of mothers and children in the exposed and unexposed groups. Compared with unexposed children, exposed children were more often born preterm and had lower birth weight and older parents. Mothers of the exposed children tended to smoke more during pregnancy, had a lower level of education and had higher comorbid diabetes or psychiatric disorders.

A total of 25 747 children were identified as ADHD individuals (554 of mothers having at least two SAs, 3087 of mothers having one SA and 22 106 of mothers without SA). The mean age at first ADHD diagnosis was 10.4 years (SD 3.4). The incidence rates of ADHD were 3.17, 2.78 and 2.54 per 1000 person years in offspring whose mothers had at least two SAs, one SA and no SA, respectively. Overall, children of mothers with a history of SA had an increased rate of ADHD (HR, 1.11; 95% CI, 1.07 to 1.15). Compared to unexposed children, children of mothers with one SA had an increased rate of ADHD (HR, 1.09; 95% CI, 1.05 to 1.13) and those children of mothers with at least two SAs had a higher rate (HR, 1.22; 95% CI, 1.12 to 1.33) (as shown in Table II). A statistically significant trend for the association of number of maternal SA and ADHD risk in offspring (HR0,1,2, 1.09, 95% CI, 1.06 to 1.12, P < 0.001) was observed.

The interaction between maternal history of SA and birth order was significant at the multiplicative scale (P for interaction term: <0.01). In the firstborn child, the risk of ADHD was consistently increased with the number of maternal SA, with a HR of 1.14 (95% CI, 1.07 to 1.23) in children of mothers with 1 SA before the childbirth and with a HR of 1.48 (95% CI, 1.23 to 1.78) in children of mothers with at least two SAs before the childbirth (as shown in Table III).

Adverse birth outcomes accounted for a very small proportion (0.1–3.6%) of the association between maternal SA and risk of ADHD, although almost all the natural indirect association estimates were marginally statistically significant (as shown in Table IV). Preterm birth was the strongest mediator that accounted for ~3.6% of the association.

Specifically, when maternal SA was dichotomized, maternal history of SA was associated with an increased risk of ADHD in offspring (OR, 1.22; 95% CI, 1.17 to 1.26). After accounting for misclassification of maternal SA by probabilistic bias analysis, the bias-corrected OR was 1.23 (95% CI, 1.18 to 1.27).

The population attributable fraction was 1.3%, indicating that maternal history of SA explained 1.3% of ADHD in the study population, if assuming an unconfounded causal association. For the bias analysis, the E-value for the observed overall estimate was 1.46.

The analyses stratified by offspring sex showed similar results (Supplementary Table SII). When we excluded induced abortion before the pregnancy, the estimates for the association were almost unchanged (Supplementary Table SIII). We observed a higher HR in children of parents with psychiatric disorders than that in children of parents without psychiatric disorders, although the increased rates in children whose mothers with at least two SAs and comorbid parental psychiatric disorders were not statistically significant (Supplementary Table SIV). Stratification analyses by adverse neonatal outcomes yielded similar results (Supplementary Table SV), so did analyses by different subtypes of SA (Supplementary Table SVI and Supplementary Table SVII) and analyses excluding the children with assisted reproductive technology (Supplementary Table SVIII). No significant association was observed between maternal SA that occurred after the childbirth and the risk of ADHD in offspring (Supplementary Table SIX). Results from sub-analyses restricted to offspring born before 2010 and to all offspring born from 1995 to 2016 were similar to those in primary analyses (Supplementary Table SX and Supplementary Table SXI). Analyses using multiple imputation also yielded similar results (Supplementary Table SXII).