Maternal SARDs Linked to Childhood Mental Disorders

Key Takeaways

• In a Taiwanese cohort of 1,949,968 maternal-child dyads, childhood mental disorders were identified in 19.8% of offspring exposed to maternal SARDs versus 15.8% of comparison offspring.
• Higher childhood mental disorder risk was significantly associated with maternal SLE, PSS, IIM, and SSc.
• Associated childhood outcomes included anxiety disorders and several neurodevelopmental disorders, including intellectual disability, ADHD, specific learning disorders, and motor disorders, while autism spectrum disorder was not linked in this cohort.

In Taiwan, Huang and colleagues' nationwide Taiwan cohort study of maternal SARDs and childhood mental disorders linked the Taiwan Maternal and Child Health Database with birth, death, catastrophic illness, and National Health Insurance claims data. The cohort included 1,949,968 singleton live births from 2004 through 2015, including 3,907 offspring born to mothers with SARDs and 1,946,061 offspring born to mothers without any autoimmune rheumatic disease, who served as the comparison group, and childhood MD diagnoses were followed through the end of 2021. Maternal SARD exposure was defined before delivery by approved catastrophic illness certificates, and childhood MDs were identified from claims by at least one inpatient code or two outpatient codes. Adjusted Cox models accounted for parental psychiatric history and key maternal, pregnancy, socioeconomic, cohort-year, and child-sex factors within Taiwan’s health system.

In adjusted analyses, maternal SARDs were associated with childhood MDs overall (HR 1.30, 95% CI 1.19-1.41). Anxiety disorders were also associated with maternal SARD exposure (HR 1.55, 95% CI 1.17-2.06). Other associated childhood outcomes spanned both psychiatric diagnoses and developmental conditions in this Taiwanese cohort.

Among maternal disease subtypes, primary Sjögren’s syndrome (PSS) was associated with higher offspring MD risk (HR 1.59, 95% CI 1.29-1.97), and systemic sclerosis (SSc) showed a similar higher-magnitude estimate (HR 1.94, 95% CI 1.15-3.28). Elevated associations were also reported for systemic lupus erythematosus (SLE) and idiopathic inflammatory myositis (IIM). Neurodevelopmental disorders (NDDs) overall were associated with maternal SARDs, and the competing-risk analysis yielded similar results.

Because this retrospective cohort supports associations rather than causation, several limits affect interpretation. Maternal disease activity during pregnancy was unavailable, exposure and outcome definitions came from administrative and claims data, residual confounding remained possible because obesity, alcohol use, and direct smoking measures were unavailable, and some later-incident childhood MDs may have been undercaptured despite long follow-up. Low birth weight showed a partial mediating signal, with 18.7% of the association mediated, and estimates were attenuated after additional adjustment for antirheumatic medication prescriptions during pregnancy.

The authors concluded that maternal SARDs were associated with higher risk of several childhood MDs in this Taiwanese cohort. The signal was most evident across anxiety and selected NDD outcomes, while autism spectrum disorder was not associated in this dataset.

Clinician Questions

How were childhood mental disorders defined in the Taiwan maternal SARD cohort?

Childhood MDs in the Taiwan maternal SARD cohort were identified from ICD-9-CM and ICD-10-CM claims, requiring at least one inpatient record or at least two outpatient records with a relevant diagnosis code, and follow-up for diagnoses continued through the end of 2021.

Which maternal autoimmune conditions were linked to higher childhood mental disorder risk in this cohort?

The subtype signal centered on maternal SLE, PSS, IIM, and SSc, with PSS and SSc carrying the clearest higher-magnitude estimates in the adjusted analyses.

What limits how far the Taiwan maternal SARD findings can be generalized?

Generalizability is limited because the cohort was drawn from Taiwan, maternal disease activity during pregnancy was unavailable, exposure and outcome definitions relied on administrative and claims data, some confounders such as obesity, alcohol use, and direct smoking measures were unavailable, and later-incident childhood MDs may have been undercaptured despite long follow-up.

Recommended Reading

• For more on systemic sclerosis in SARDs: Spotlight on Systemic Sclerosis: Pathophysiology, Presentation, & Emerging Evidence
• For more on SLE among maternal SARDs: Lupus Nexus Builds a 3,500-Patient SLE Research Platform
• For more on Sjögren’s disease among maternal SARDs: Sjögren’s Myth Busters: Can You Detect Fact from Fiction on Emerging Therapies?