A pharmacoepidemiological study of the risks and benefits to mothers and offspring of prescribing drugs during pregnancy

Abstract

Background
Pregnant women's exclusion from clinical trials, due to ethical and practical considerations, limits understanding of how intrauterine medication exposure impacts developing foetuses. This knowledge gap heightens vulnerability of both mother and child, complicating clinical decision making and pharmacological research. Nonetheless, maternal medication for chronic conditions like hypertension, and hypo- or hyperthyroidism during pregnancy is often necessary, as left untreated, these conditions are linked to adverse neonatal outcomes, including preterm birth, foetal loss, and maternal complications. Thus, there is a pressing need to establish systematic, robust evidence to elucidate the risks and benefits of maternal prescription drug use during pregnancy.

Methods
I used two different approaches to integrate results from data of differing nature, first an observational study using electronic healthcare record data from the Clinical Practice Research Datalink, and second Mendelian randomization using phenotypic and genetic data from The Norwegian Mother and Child Cohort Study. In my observational study, I estimated the association of discontinuing maternal treatment during pregnancy and maternal and foetal outcomes for the chronic conditions stated above. Genetic data offers an alternative insight to determine causal relationships. To establish a fundamental understanding of the genetics contribution to traits, I reviewed a range of methods to estimate heritability. Further, I performed Mendelian randomization, a genetic instrumental variables analysis, to infer potential causal relationships between maternal genetic drug target perturbation and foetal outcomes.


Results
Observational analysis of maternal prescriptive discontinuation and maternal and neonatal outcomes showed a reassuring lack of association between treatment discontinuation for many drug subclasses and differential risk of many adverse outcomes. For example, maternal discontinuation of beta adrenoreceptor blocking drugs was estimated to have an odds ratio (OR) of 0.80 (95% confidence interval (CI): 0.45, 1.42) for miscarriage. Mendelian randomization analysis of derived maternal drug target instruments did not indicate causal relationships between multiple antihypertensive drug subclasses and increased risk of adverse neonatal outcomes. Similar paternal and maternal effect estimates suggest perturbation of these drug targets minimally impact the offspring via intrauterine exposure.

Conclusion
Within this thesis I establish and triangulate evidence using multiple methods and data sources to address the deficit of systematic studies from which clinicians may decide to prescribe medication for chronic conditions during pregnancy. I implemented a novel approach to examine the impact of maternal genetic drug target perturbation on neonatal outcomes. By integrating observational studies of drug use in pregnancy, critical analysis of current literature and genetic studies it is feasible to generate much needed evidence for a vulnerable population.

Date of Award	4 Feb 2025
Original language	English
Awarding Institution	University of Bristol
Sponsors	Wellcome Trust
Supervisor	Venexia M Walker (Supervisor), Christy Burden (Supervisor), George Davey Smith (Supervisor) & Neil M Davies (Supervisor)