Epidemiology of prenatal alcohol use and fetal alcohol spectrum disorder

Prenatal alcohol exposure (PAE) can lead to fetal alcohol spectrum disorder (FASD). FASD refers to a range of lifelong conditions caused by PAE, characterised by a distinctive facial phenotype, growth deficiencies and/or neurobehavioural impairments. This thesis presents four studies that I conducted to address knowledge gaps relevant to the epidemiology of PAE and FASD. First, objective measures of PAE are essential for identifying children at risk of adverse outcomes. Biomarkers have been advocated for use in universal PAE screening programs but their validity had not been comprehensively evaluated. I conducted a systematic review and found that biomarker test performance varied widely across studies. The quality of published studies was low, resulting in insufficient evidence to support the use of objective measures of PAE in practice. Second, the prevalence of FASD in the UK was unknown. Active case ascertainment studies have not been possible due to funding and ethical issues. To overcome these issues, I developed an algorithm to estimate FASD prevalence using existing data from a population-based birth cohort in England (ALSPAC). Up to 17% of children met criteria for FASD, indicating that it is a significant public health concern. Third, although PAE is the sole necessary cause of FASD, it is not always sufficient. Understanding risk factors for FASD is important for informing prevention strategies. However, existing studies have mostly been limited to discussion of association, rather than causation. I produced a causal diagram to depict hypothesised causal pathways to FASD. I used this diagram to guide analyses in a FASD risk factor study, reported below. Finally, I investigated FASD risk factors using multivariable logistic regression within the ALSPAC cohort. Prenatal stress, smoking and mental health problems increased the odds of FASD. Social support and folic acid supplementation were protective. These results indicate novel potential targets for FASD intervention