Main findings
This study reports no substantial risk for ADHD with prenatal SSRI/SNRI antidepressant exposure at different timings during pregnancy, and no definite duration-response associations when the hazard of ADHD is averaged over the study’s follow-up. Misclassification of exposure could underestimate of about 10% the observed point estimates, leading to an unaltered inference. When splitting the follow-up time, we observed that at pre- and early school-age children prenatally exposed to SSRI/SNRI have lower risk for ADHD diagnosis and symptoms than unexposed. At age 7-9 years, prenatal SSRI/SNRI exposure was associated with greater ADHD risk in offspring, and this seemed to be mainly driven by longer duration of SSRI/SNRI exposure. After taking into account biases and confounding, our best estimate for the weighted HR was around 1.58-1.93 for ever in-utero exposure to SSRI/SNRI, and 2.22-2.76 for 9-20 weeks duration. Nevertheless, we also document that maternal psychiatric illness during and prior to during pregnancy are possibly key factors of joined confounding, yielding substantial risk attenuation to the effect estimates for SSRI/SNRI in-utero exposure.