Paternal exposure to valproate and the risk of neurodevelopmental disorders in children

Study type

• Study topic:
  Human medicinal product
• Study type:
  Non-interventional study
• Scope of the study:
  Safety study (incl. comparative)
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Name of medicine, other:
  Valproic acid
• Anatomical Therapeutic Chemical (ATC) code:
  (N03AG01) valproic acid

  valproic acid

Population studied

• Short description of the study population:
  All pregnancies resulting in live-born singletons will be linked with paternal data in Norway and Taiwan. ; The paternal linkage to pregnancies is possible in 91.0% and 95.6% in the Norwegian cohort and in the Taiwanese cohort, respectively.

Study design details

• Study design:
  A multinational cohort study will be conducted using the Norwegian national health registries, which cover all residents in Norway, and Taiwan's population health insurance data, which cover more than 99% of the population in Taiwan.
• Main study objective:
  This study aims to investigate the association between paternal exposure to valproate and the risk of NDDs. ; The study will identify all pregnancies resulting in live-born singletons with birth years from 2010 to 2015, as recorded in the Norwegian Medical Birth Register and from 2010 to 2015 in Taiwan's National Birth Certificate Application. ; This approach ensures a minimum of six years of follow-up by the end of the study period, which extends to 2021 for both cohorts.; ; We will use both active-comparator and non-active comparator designs (compared to non-exposure). ; For the non-active-comparator design, we will restrict the cohort to individuals with indications for antiepileptic drugs (AEDs), including epilepsy, psychiatric disorders (bipolar disorder, depression, schizophrenia, anxiety, and other psychiatric conditions), as well as somatic conditions (migraine, neuropathic pain, and chronic pain).
• Setting:
  This cross national cohort study was conducted using population based data from Norway and Taiwan. ; The Norwegian cohort included data from the Medical Birth Registry of Norway, the Norwegian Prescription Database, the Norwegian Patient Registry, and the Norwegian control and payment of health reimbursements. ; The Taiwanese cohort used information from the National Birth Certificate Application database, the National Health Insurance database, and the Maternal and Child Health Database.
• Comparators:
  We will evaluate the risk of NDDs in offspring of fathers exposed to valproate, compared to the risk in the offspring of fathers exposed to a comparator or to no exposure during the sperm development period (SDev), which is defined as the 3 months (90 days) prior to conception. ; Exposure refers to the period when there is an overlap in the days' supply of the dispensed medication during SDev. ; ; We will compare the following two groups (active-comparator and non-active comparator designs) under both monotherapy and combination therapy conditions (resulting in a total of four analysis):; Monotherapy:; 1. Valproate vs lamotrigine or levetiracetam (grouped as composite exposure); 2. Valproate vs no exposure, restricted to those with indications for AEDs; Combination therapy:; 3. Valproate vs lamotrigine or levetiracetam (grouped as composite exposure); 4. Valproate vs no exposure, restricted to those with indications for AEDs
• Outcomes:
  Primary outcome: Overall NDDs, including autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), specific learning disorders, developmental speech/language disorder, developmental coordination disorder, intellectual disability, behavioral disorder, grouped as one composite variable.; ; Secondary outcome: Individuals NDDs.
• Data analysis plan:
  Descriptive statistics will be compared between groups using standardized differences; covariates with standardized differences less than 15% will be considered balanced. ; ; The propensity score (PS), representing the probability of receiving valproate, will be derived using a multivariable logistic regression model that included all paternal and maternal covariates. ; When selecting the appropriate PS method, we considered the prevalence of paternal valproate use in the study population, the method’s precision, and its ability to reduce bias. ; Based on the initial exploration of sample size and the prevalence of paternal valproate use, we chose PS fine stratification weighting (FSW), with stratification based on the PS in the exposed group.; ; A pooled logistic regression model will be used to estimate the hazard ratio (HR), which approximates the odds ratio from the pooled logistic regression, along with 95% confidence intervals (CI). We chose pooled logistic regression, because it produces more reliable and robust estimates and to minimize the inherent bias associated with Cox proportional hazards regression. Moreover, we then avoid pitfalls of proportional hazards assumptions in Cox regression when they may not hold true.; ; Standardized incidence risk curves will also be plotted by fitting weighted pooled logistic regression models and calculating weighted risk differences. Robust standard errors will be applied to account for both weighting and data clustering, given the potential for multiple offspring per father.