Observational studies have shown that children who develop type 1 diabetes (T1D) have larger size at birth — a marker of foetal growth — or rapid postnatal weight gains. Recent studies have explored the contribution of the T1D susceptibility human leukocyte antigen (HLA) alleles, which explain half of disease heritability, to perinatal growth but findings are inconsistent.

In the Cambridge Baby Growth Study (CBGS), a birth cohort which follows children over the first 2 years of life, I conducted genotype-phenotype correlation analyses with the objective of identifying T1D susceptibility single nucleotide polymorphisms (SNPs) and candidate genes that influence the archetypal phenotypes of the disease prodrome, which starts early in life, as a means of shedding light on the heterogeneity of data and sign-posting mechanisms. Based on prior ‘biological’ knowledge I selected T1D SNPs conjectured to affect growth mediated by hormones, vitamin D, or the microbiome. Genotypes were determined in ~600 children and used to derive high-risk T1D HLA haplotypes and genetic risk scores. Examination of data in the CBGS defined the contributors to physiological growth and set the framework for analyses.

My findings concurred with most prior studies on the lack of an effect of the SNP rs689 in INS on early growth. A genetic contribution to size at birth was detected by variants in the imprinted DLK1 gene, the vitamin D metabolism CYP2R1 gene and the gene encoding TYK2 known for its susceptibility to infections. Examination of longitudinal phenotypes identified that infancy growth is influenced by genes with a key role in innate immunity, e.g. IFIH1. The strongest associations were found between skinfold thickness, a proxy for adiposity, and variants in genes whose products play a role in immune and inflammatory pathways, exemplified by the rs653178 in SH2B3, previously known for its association with cardiovascular disease. Taking on the lead from my findings with anthropometry, I explored genetic-hormonal correlates and found evidence in support of IGF-1 mediating the association that I identified, more strongly in boys than in girls, between the SNP tagging the high-risk HLA-DR3 and rapid linear growth. I made the novel finding that the rs12785878 in DHCR7, known to strongly associate with vitamin D levels, inversely correlated with leptin levels in girls but not in boys.

This study elucidated the effects of T1D susceptibility SNPs on infancy adiposity, which might serve as a better disease marker than weight. It identified a link between leptin and a common polymorphism in DHCR7, opening up the possibility that vitamin D suppresses leptin production in a sex-specific manner with protective effects against autoimmunity.