Associations between birth outcomes and development of adult cardiometabolic disease: revisiting the Walker Cohort

Abstract

Cardiometabolic diseases have been ranking at the top of the leading causes of death globally for over thirty years, and their burden is only expected to continue rising due to sedentarism, obesogenic lifestyles, and increased life expectancy. Disorders such as coronary heart disease (CHD), type 2 diabetes (T2D) or cancer are a global concern and have a profound impact on national economies and individual welfare.

During the latter half of the 20th century, research investigating the impact of adverse conditions such as famines over multiple independent European populations identified associations between low birth weight (BW) and adult cardiovascular disease (CVD) and glucose intolerance, suggesting the existence of a mechanism for early determination of cardiometabolic disorders, and a possible avenue for intervention and disease prevention.

Multiple hypotheses have been proposed to explain the biological intricacies of the mechanism of adult disease programming, but two are the most prominent. The “fetal origins of adult disease” hypothesis by Barker and colleagues proposed that environmental influences during fetal development, such as maternal undernutrition, led to permanent physiological alterations which predisposed the individual for later development of metabolic disorders. The “fetal insulin hypothesis” by Hattersley and colleagues proposed that the observed associations were driven through insulin and glucose metabolism alterations, primarily determined by genetics, and causing adult metabolic disease. Both hypotheses have been revisited and refined, and current research focuses on elucidating the mechanism of programming through novel methodologies and finding new associations studying additional birth outcomes and adult traits. Through familial cohorts, maternal and paternal CVD and T2D have been associated with low offspring BW, finding intergenerational associations even for the extended family. In addition, genome-wide association analyses have identified 243 loci associated with BW, many of which are also involved in metabolic networks. The biologic processes behind the association between fetal growth and later development of cardiometabolic disease remain largely unknown, a mechanism undoubtedly complex governed by environmental, genetic, and other hereditary influences.

The goal of this PhD thesis was to investigate the associations described using a novel dataset, the Walker Cohort, particularly focusing on the association between parental health and offspring birth outcomes. Being a fairly unused source of data, the original Walker datasets are described, detailing the cleaning process performed to prepare the data for the analyses, including the merging of additional health records and demography data. The potential of Walker resided in the number of identifiable offspring, mothers, and fathers available, allowing for lifetime follow-up, and the collection of additional birth outcomes such as placental weight (PW).

The first set of analyses presented investigated the association of BW and PW with the development of T2D, CHD, and cancer as a preliminary investigation of the correlations reported previously. Logistic regression analyses paralleled the literature, finding negative associations between BW and both T2D and CHD. PW to BW ratio was positively associated with T2D, and a similar but not significant association was found for PW, providing novel associations and suggesting a link between T2D development and placental inefficiency.

The following set of analyses aimed to identify associations between offspring BW and PW and the post-natal development of maternal and paternal T2D, CHD, and cancer. Offspring from fathers who developed T2D were born lighter and had lighter placentas, associations which also were significant through Fine-Gray survival regression analyses. The correlation of low PW and increased paternal T2D risk suggests a genetic effect lowering PW and increasing T2D susceptibility, similar to BW. Mothers who developed T2D gave birth to heavier babies with heavier placentas, but these associations were not seen through survival regression, most likely due to the independent and BW-increasing effect of maternal hyperglycaemia.

Offspring from mothers who later developed CHD were born lighter, and presented higher PW to BW ratios, associations which also were significant through Cox survival regression. These analyses also identified negative associations between paternal CHD risk and offspring BW.

No significant associations were found between maternal or paternal cancer and offspring BW, PW, or PW to BW ratio through Fine-Gray survival regression. When analysing individual cancer types, offspring BW was negatively associated with maternal and paternal lung cancer risk, an association most likely confounded by parental smoking. Paternal bowel cancer risk was positively associated with offspring BW.

The final set of analyses presented investigated the association of maternal and paternal polygenic risk of T2D and coronary artery disease (CAD) and their offspring’s BW and PW, aiming to identify the genetic component of the associations presented in the previous section. The project summary and data analysis plan were sent for consideration by the Early Growth Genetics (EGG) consortium, to identify suitable cohorts including genotyped mothers and fathers and offspring BW and PW measurements that could contribute data to meta-analyse. Pooling the results from 7 birth cohorts, paternal CAD polygenic score (PGS) was negatively associated with offspring BW, with a nearly significant negative association with offspring PW. A similar negative association trend was found between paternal T2D PGS and offspring BW. This is the first time the association of parental cardiometabolic disease PGS and offspring birth outcomes has been considered, and these analyses provide evidence for direct associations between genetically common variant-determined paternal CAD and T2D risk and offspring BW, also explaining the association between low BW and susceptibility to these disorders when the same variants are inherited.

In conclusion, the work performed in this PhD thesis has repurposed a fairly unused birth cohort and replicated the associations between low BW and increased T2D and CHD incidence described in the literature. The association of parental health with offspring outcomes was also considered, identifying novel associations with offspring PW, and suggesting a link between genetic cardiometabolic disease predisposition and offspring BW. Finally, novel evidence of such a link was provided through the paternal T2D and CAD PGS association meta-analyses. This thesis has provided additional evidence of associations between birth outcomes and adult health, and provides new epidemiological evidence supporting the genetic determination of fetal growth and later cardiometabolic disease development.