The assignment of direction and causality within networks of observational associations is problematic outside randomized control trials, and the presence of a causal relationship between body mass ...