ADHD Medication Safety in Pregnancy: Precise Estimates, Uncertain Causal Inference

See article by Babinski and Tuan

Attention-deficit/hyperactivity disorder (ADHD) medication use is on the rise in many countries, particularly in adolescent and adult females.^1,2 Consequently, more people face the difficult decision of whether to use ADHD medication when they are pregnant or planning a pregnancy. Shared decision-making on ADHD medication use during pregnancy should ideally start before conception. Early discussions may allow for thorough consideration of the risks and benefits; trialing discontinuation or changes in dose, medication, or formulation; and avoidance of unintentional fetal exposure. However, many pregnancies are unplanned, and when pregnancy is discovered, exposure may have already occurred during a critical window for the development of congenital anomalies. Fortunately, studies do not suggest that the most commonly used ADHD medications are major teratogens.^3,4 However, there is accumulating but inconsistent evidence that these medications may increase the risk of certain pregnancy complications.⁵

The study by Babinski and Tuan⁶ adds to the expanding evidence on the potential risks of ADHD medication use in pregnancy by examining obstetric outcomes using TriNetX data. They found that compared to no ADHD medication use, both stimulant and nonstimulant drugs were associated with increased risks of most of the adverse pregnancy outcomes examined, including gestational hypertension, preeclampsia, preterm delivery, intrauterine growth restriction (IUGR), and spontaneous abortion, after controlling for selected confounders. Head-to-head comparison of stimulant versus nonstimulant medications suggested that stimulants were associated with lower risks of placental abruption (odds ratio 0.40, 95% confidence interval 0.29–0.57), IUGR (0.66, 0.59–0.74), and preterm delivery (0.47, 0.42–0.54) and a higher risk of placenta previa (2.30, 1.82–2.91).

A key concern in studies assessing potential risks associated with ADHD medication in pregnancy is addressing confounding by indication. In the case of ADHD medication safety, it is not only the presence of ADHD that is relevant but also its severity and the presence of commonly occurring comorbid conditions that might confound any observed associations. In addition, those who use ADHD medication in pregnancy often use other psychotropic medications,⁷ which could introduce additional confounding. Health care databases lack clinical details like symptom severity, and the presence of other psychiatric and somatic comorbidities and use of other medications may serve as proxies for ADHD severity.

In the study by Babinski and Tuan, the authors identified the presence of ADHD recorded in electronic health records and adjusted for it by including a variable for ADHD in the propensity score models. However, ADHD was only recorded in the records of 47% of stimulant users and 15% of nonstimulant users included in the study. This suggests that there may be residual confounding by ADHD. In this context, restricting the study population to those with ADHD recorded and comparing exposed and unexposed, adjusting for other confounding factors, may have been more effective at controlling for confounding by indication. Adjusting for confounding by employing propensity score methods means that many confounders and proxies for unmeasured confounders can be adjusted for in the analysis, and concerns about model overfitting are less relevant.⁸ Adjusting for other psychotropic medication use at baseline might have captured aspects of ADHD severity that may contribute to the associations identified and is common practice in the existing literature.

Large databases such at TriNetX can result in the identification of large numbers of exposed pregnancies and sufficient sample size to give very precise effect estimates. However, the risk of bias remains. P-values should not be used to guide the interpretation of the importance of associations but rather how well the study addressed confounding, the strength of the estimate, and consistency across analyses and with existing literature should be considered. This study identified increased risks of preterm birth, gestational hypertension, and preeclampsia that appear across different settings, though most have described more modest associations.^9–13 Other findings may reflect true risks or residual confounding. In any case, these results add to a growing body of literature that finds ADHD predictive of adverse pregnancy outcomes. It remains difficult to disentangle the association between the disorder and associated characteristics from ADHD medication use. However, close pregnancy supervision, patient education on the symptoms of preeclampsia and early labor, and proactive discussions of whether patients may wish to use or avoid medication in pregnancy may help reduce perinatal risks.

Article Information

Published Online: August 26, 2026. https://doi.org/10.4088/JCP.26com16606
© 2026 Physicians Postgraduate Press, Inc.
J Clin Psychiatry 2026;87(3):26com16606
Submitted: July 14, 2026; accepted July 15, 2026.
To Cite: Cohen JM. ADHD medication safety in pregnancy: precise estimates, uncertain causal inference. J Clin Psychiatry 2026;87(3):26com16606.
Author Affiliations: Department of Chronic Diseases & Centre for Fertility and Health, Norwegian Institute of Public Health, Oslo, Norway.
Corresponding Author: Jacqueline M. Cohen, PhD, Norwegian Institute of Public Health, PO Box 222 Skøyen, Oslo, Norway ([email protected]).
Financial Disclosure: Dr Cohen reports participation in research projects funded by pharmaceutical companies mandated by regulatory authorities, with all funds paid to her institution (no personal fees) and with no relation to the work reported in this commentary.
Funding/Support: This work was partly supported by the Research Council of Norway through its Centres of Excellence funding scheme, project number 262700.