Large scale genome wide association analyses of food intake uncover associations with health and behavior

The molecular genetics of food intake knows two substantial challenges. First, only two genetic loci have been discovered in the published GWAS literature. Second, there exist no reliable measurement methods of habitual food... [ view full abstract ]

The molecular genetics of food intake knows two substantial challenges. First, only two genetic loci have been discovered in the published GWAS literature. Second, there exist no reliable measurement methods of habitual food intake in the first place: systematic underreporting of total food intake is the norm, and especially problematic because it is associated with body mass index – potentially giving rise to spurious associations with obesity and food intake.

We propose a new method for processing self-report food intake data that largely eliminates the association with systematic misreporting bias of total food intake. Our measure of food intake is determined by relative intakes of protein, carbohydrate, fat, and sugar intake, which we subject to genome wide association analyses. The current preliminary sample size is N ≈ 90,000, which is projected to more than double in size in the near future. Here, we present our preliminary findings. These findings give rise to new insights into the biological basis of food preferences, and quasi-causal associations between food intake and health outcomes.