Electronic Theses and Dissertations



Document Type


Degree Name

Doctor of Philosophy


Public Health

Committee Chair

Wilfried Karmaus

Committee Member

Hongmei Zhang

Committee Member

Yu Jiang

Committee Member

Ali H Jiyab


The impact of maternal body mass index (BMI) on offspring health outcomes such as obesity has been widely investigated, with evidence suggesting that in-utero conditions may influence DNA methylation (DNAm) and BMI developmental trajectories. However, it is unclear whether differential DNAm on sites with a cytosine followed by a guanine linked by phosphate dinucleotide (CpG) related to gestational BMI are linked to BMI changes in offspring. Using data from the Isle of Wight birth cohort, UK, this study aims to investigate BMI trajectories and address the role of DNAm. Using trajectory analysis four distinct BMI developmental trajectories - ‘normal’ (n= 1042), ‘early persistent obesity’ (EPO, n=61), ‘early transient overweight’ (ETO, n=185), and ‘delayed overweight’ (DOW, n=149) - that spanned first 26 year of life were identified. The trajectories were found to be influenced by gestational BMI. To identify CpGs related to gestational BMI academic database were explored and 1,090 CpGs were found in the IOW cohort data as candidate CpGs for specific aims 2 and 3 out of 1,773 differentially methylated CpGs reported in prior investigations. Fourteen of these candidate CpGs were found to be significantly associated with BMI trajectories, two survived multiple testing. Higher methylation of cg23089913 (NANOS1 gene) was associated with decreased odds of being in the EPO trajectory with an odds ratio of 0.84 (95%CI: 0.76-0.93). In contrast, increased methylation of cg13217064 (SOX14) was associated with a 1.4 times higher odds (95%CI: 1.13-1.67) of being in the DOW compared to the 'normal' trajectory. Finally, associations between candidate CpGs and repeated BMI measurements from infancy to 26 years of age were investigated.Five CpGs - cg00488692 (SP3), cg14434213 (RNF5P1), cg23089913 (NANOS1), cg26862527 (BAI3), and cg17812850 (TMEM184C) were found to be statistically significantly linked with BMI. Female participants exposed to prenatal paternal smoking and mixed feeding during infancy had a higher BMI, while male participants with lower birth weight had 0.4 kg/m2 higher BMI. The study identified candidate CpGs on genes critical to metabolic disorders and provides a basis for further investigations to understand the biological role of DNAm sites in BMI development.


Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest


Open Access