Characterization of the genetic architecture of BMI in infancy and early childhood reveals age-specific effects and implicates pathways involved in Mendelian obesity

Abstract

To elucidate the role of common genetic variation on infant and child weight development, we performed genome-wide association studies across 12 time points from birth to eight years in 28,681 children and their parents (27,088 mothers and 26,239 fathers) in the Norwegian Mother, Father and Child Cohort Study (MoBa). We identify 46 distinct loci associated with early childhood BMI at specific ages, matching different child growth phases, and representing four major trajectory patterns. Among these loci, 30 are independent of known birth weight and adult BMI loci, and 21 show peak effect between six months and three years, making these discoverable only at early age. Several of the 21 variants reside in/near genes previously implicated in severe forms of early-onset obesity, and monogenic obesity genes are enriched in the vicinity of the 46 loci. Four loci demonstrate evidence of several independent association signals as key drivers for BMI development near LEPR, GLP1R, PCSK1, and KLF14, all central to appetite and energy balance. At the KLF14 locus, we detect significant associations for maternally inherited alleles only, consistent with imprinting effects. Finally, we demonstrate how the BMI distribution stratified by different polygenic risk scores transitions from birth to adult profile throughout early childhood, and how age-specific polygenic risk scores improve the prediction of childhood obesity, outperforming scores based on adult BMI. In conclusion, our results offer a fine-grained characterization of the rapidly changing genetic association landscape sustaining early growth.

Modeling assortative mating and genetic similarities between partners, siblings, and in-laws

“…There were valid genomic data on 26,681 complete couples, 2170 complete sibling pairs, 3905 in-law pairs, and 1763 co-in-law pairs (after removing 4 extended families where the co-in-laws were siblings). Information about the genotyping, imputation, and quality control is available in the Supplementary Methods 1 and Supplementary Table 1 and in previous reports 49,50 . Depression.…”

Age-specific effects of childhood Body Mass Index on multiple sclerosis risk

“…This cohort was used to perform GWAS of BMI at various ages using data from between 11,095 and 28,681 children across 12 time points from birth to 8 years (birth, 6 weeks, 3 months, 6 months, 8 months, 1 year, 1.5 years, 2 years, 3 years, 5 years, 7 years, and 8 years). 46 distinct genetic loci were associated with raised childhood BMI at various ages 10 . The outcome of this GWAS was standardized BMI, and therefore the units are not directly comparable with published estimates from MR studies using other BMI GWAS for genetic instruments.…”

“…A secondary analysis then used data at each individual time point. These epochs were defined according to the shared genetics of BMI at each time point, and loosely correspond to the four phases of BMI genetics suggested by the authors 11 . Data are available from the MoBa authors on request.…”

“…The International Multiple Sclerosis Genetics Consortium (IMSGC) 2019 discovery phase GWAS (47,429 MS cases, 68,374 controls) was used as the outcome dataset 11 . We used summary statistics from the discovery stage meta-analysis (14,802 persons with MS, 26,703 controls).…”

“…Longitudinal MR analysis of the effect of BMI during development is now possible due to the recent publication of longitudinal BMI GWAS data from the Norwegian Mother, Father, and Child Cohort Study (MoBa) cohort 10 . We set out to explore the causal effect of BMI changes during early life on MS risk and examine the dynamics of this effect over time.…”