Genome-Wide Association Study to Identify Common Variants Associated with Body Shape

Genome-wide association studies (GWAS) have identified a large number of loci related to body mass index (BMI) and obesity. For BMI, 941 near-independent genome-wide significant SNPs at 536 polygenic loci have been recently detected 1. A GWAS for whole-body lean body mass (LBM), adjusted for sex, age, and height with or without fat mass adjustments, revealed seven LBM loci 2. Another GWAS for body fat distribution in arms, legs, and trunk on 362,499 individuals from the UK Biobank identified 98 independent loci, 29 of which had not previously been associated with anthropometric traits including BMI 3. A GWAS for body fat distribution as measured by waist-to-hip ratio adjusted for BMI (WHRadjBMI) reported 463 independent signals (P < 5 × 10−9) in 346 loci 4. However, our understanding of the mechanisms of body weight control and of the etiology of obesity is still not clear. This could be attributed to an oversimplification of the phenotype by using non-sophisticated anthropometric traits. BMI, waist and hip circumference (WC, HC) and waist-to-hip ratio (WHR) have been widely used for decades, but are known to be inaccurate measure of body fat content 5. Simply defining obesity using BMI could be the reason for the “obesity paradox” 6. This refers to some evidence suggesting that obesity might be associated with decreased mortality while there is abundant data against this claim. 
In order to overcome this problem, efforts have been invested in developing composite phenotypes of body shape. Ried et al, calculated four different averaged principal components (AvPCs) capturing specific aspects of body shape using six anthropometric traits (BMI, WHR, height, weight, WC and HC).  This method identified six novel loci (two for each of AvPC1, AvPC2 and AvPC4) but failed to associate some known loci from previous single trait GWAS analyses. Lower sample size or the introduced noise from PC analysis are considered as the reason 7. At any rate, calculation of PCs is not an easy job for every-day clinical practice and limitations should be considered when using such sophisticated analyses 8. 
Krakauer et al have defined a new allometric tool to consider the impact and the risk profile associated with body ‘shape’ compared to body ‘mass’. The new tool, named A Body Shape Index (ABSI), is based on a power-law relationship between WC and BMI. It is almost independent of BMI itself but clearly outperforms both WC and BMI as predictors of mortality 9.  There are also other indices considering body shape, such as waist-to-height ratio 10, body adiposity index (BAI) 11, Tri-Ponderal Mass Index (TPI) 12, etc. The goal of designing the newer anthropometric tools where to consider the impact of body shape compared to body size and are shown to be independent of BMI 13.