Discordant Sibling Pair Comparisons in Observational Studies: A Research Design Simply Explained

Area of Science:

• Perinatal Psychiatry
• Developmental Pediatrics
• Genetic Epidemiology

Background:

• Conventional observational studies struggle to establish causality for gestational antidepressant exposure and autism spectrum disorder (ASD) risk due to residual confounding.
• Confounding by indication, genetic, and environmental risk factors pose significant challenges in accurately assessing the relationship between antidepressant use during pregnancy and ASD.
• Propensity score matching can partially address confounding by indication, but cannot fully account for unmeasured or unknown genetic and environmental influences.

Purpose of the Study:

• To evaluate the utility of sibling comparison studies in disentangling the effects of gestational antidepressant exposure from shared genetic and environmental factors on ASD risk.
• To explore the causal inference of antidepressant exposure on offspring ASD risk using a methodology that accounts for familial confounding.

Main Methods:

• Sibling comparison studies were employed, comparing siblings discordant for antidepressant exposure (cohort design) or discordant for ASD outcome (case-control design).
• This approach assumes that siblings share similar genetic and environmental risk factors, which are thus controlled for in the analysis.
• The study also presents findings from a recent investigation into ASD risk following gestational or early infancy antibiotic exposure to illustrate the method's application.

Main Results:

• If antidepressant exposure remains significantly associated with ASD in sibling comparisons, it suggests a potential causal role for the medication.
• Conversely, if the association loses significance, it indicates that shared genetic and/or environmental factors are more likely responsible for the observed ASD risk.
• Interpretation of findings requires careful consideration of the strengths and limitations inherent in sibling comparison designs.

Conclusions:

• Sibling comparison studies offer a robust approach to address unmeasured confounding in the study of gestational exposures and offspring neurodevelopmental outcomes like ASD.
• This methodology can help differentiate between medication effects and familial predispositions in the etiology of ASD.
• The findings underscore the importance of considering familial factors when interpreting associations between prenatal exposures and developmental disorders.