| dc.description.abstract | Background: Increased plasma glucose levels in pregnancy or gestational diabetes may have implications for future maternal health, such as increased risk of type 2 diabetes mellitus (T2DM) developing within 10 years. Increasing evidence indicate that epigenetics, such as DNA methylation, may play an important role in development of gestational diabetes. Epigenome-wide association studies (EWAS) explore association between traits and DNA methylation across the genome. The EPIPREG study (role of epigenetics in diabetes development during pregnancy) provided EWAS using the Illumina Methylation EPIC BeadChip (EPIC), and identified two CpG (cytosine – phosphate – guanine) sites associated with fasting plasma glucose (FPG) (cg08098128 and cg14120215) and one CpG site associated with 2 hour plasma glucose (2 h PG) (cg19327414) in Europeans. Aims: The first aim of this project was to technically validate the EPIC using bisulfite pyrosequencing (BSP). Further aims were to explore how DNA methylation of top tree candidate CpG sites found from the EWAS in EPIPREG study was associated with different factors, and how DNA methylation of candidate CpG sites found in studies associated with T2DM was associated with FPG and 2-h PG. Methods: The EPIPREG study provided EWAS in healthy pregnant Europeans (EU) (n=314) and South Asians (SA) (n=166) women, using the data collected from the STORK Groruddalen study. The BSP method was used for EPIC validation. The DNA methylation levels found in EPIPREG study and two other studies associated with T2DM, and other measurements collected at 28±2 gestational week were sorted according in tertiles according to low, medium and high methylation level at each CpG site. Results: The strong correlation was observed (r = 0.64, p < 0.001) between EPIC and BSP methods. Tertiles of DNA methylation indicated a positive association between cg08098128 methylation and BMI (p = 0.046), and a negative association between cg14120515 methylation and age (p < 0.001) in Europeans, indicating that BMI and age may be confounding factors. Four significant associations between CpG sites found in other studies and plasma glucose levels was observed (cg06500161 (SA, FPG), cg11024682 (SA, FPG), cg00574958 (EU, FPG), cg04999691 (EU, 2-h PG)), using ANOVA. The applying of Person correlation indicated larger number of significant associations between CpG methylation and plasma glucose levels: cg11024682 (SA, FPG), cg07988171 (EU, FPG), cg07988171 (EU, 2-h PG), cg11024682 (SA, 2-h PG), cg18181703 (EU, 2-h PG), cg00574958 (EU, 2-h PG), cg19693031 (EU, FPG), cg06500161 (SA, FPG). Conclusion: High technical validity of the EPIC was suggested due to strong correlation between EPIC and BSP. The association between BMI and age found in tertiles, probably identifies that BMI and age are confounding factors. Some of CpG sites found in other studies confirmed the association with plasma glucose levels in STORK Groruddalen study. Applying the Person correlation detected more associations. | eng |