No Increased Neurodevelopmental Risk in Children of Fathers Treated with Valproate: JAMA

The initial concerns arose from a study conducted by IQVIA, which reported a possible signal of heightened risk of NDDs among children whose fathers were taking valproate. Following this, researchers from Denmark performed an extensive cohort study based on nationwide register data, with the report initially published in March 2024. Access to the full IQVIA report was restricted at that time. Following the publication of the IQVIA data in May 2024, the Danish research group went back to their dataset, using the same definitions and methods to determine whether their results would correspond.

The Danish cohort study consisted of 1,180,308 singleton live-born children in Denmark from 1997 to 2017. Exclusion criteria were children born from in vitro fertilization, those with implausible or missing gestational age, and those with no known paternal identity. Exposure was defined as children of fathers with at least one filled prescription for valproate, lamotrigine, or levetiracetam during the period 120 days before through 14 days after the LMP of the pregnancy.

A further sensitivity analysis with a more restricted exposure window (60 days pre-LMP to 14 days post-LMP) was also conducted. Neurodevelopmental outcomes were followed from age 1 through to the end of 2018. The main analysis used Cox proportional hazards regression to estimate adjusted hazard ratios (aHRs) after adjustment for parental epilepsy, education, and other possible confounding factors.

Key Findings

• 961 children were exposed to paternal valproate monotherapy and 501 (52.1%) of them were males. Their median age at end of follow-up was 10.6 years.

• 1,401 children were exposed to levetiracetam or lamotrigine, of whom 725 (51.8%) were male with a median follow-up age of 5.5 years.

• There was no substantial association between paternal valproate exposure and risk of NDDs (aHR: 1.02, 95% CI: 0.67–1.54).

• Results were similar across various sensitivity analyses, such as stratification by valproate dose, time period, inclusion of polytherapy, wider NDD definitions, and fathers with epilepsy.

• In a subgroup analysis limited to fathers with epilepsy of unknown cause, an aHR of 2.48 (95% CI, 1.13–5.44) was seen but became nonsignificant after matching by birth year (aHR: 2.33, 95% CI: 1.00–5.44).

• Analyses that imitated the methods employed in the IQVIA study again found no elevated risk of NDD.

This large-scale Danish cohort study did not find evidence to substantiate an increased risk of neurodevelopmental disorder in children fathered by men treated with valproate during the time of conception. These results are contrary to previous observational evidence that signaled regulatory prudence and underscore the importance of transparency and peer-reviewed scrutiny of influential studies such as IQVIA's. Scholars demand that the entire analytical code and data of the IQVIA study be made available to the public for independent validation due to the severe consequences of EMA's precautionary action.

Reference:

Christensen J, Trabjerg BB, Dreier JW. Risk of Neurodevelopmental Disorders and Paternal Use of Valproate During Spermatogenesis. JAMA Netw Open. 2025;8(5):e2512139. doi:10.1001/jamanetworkopen.2025.12139