Sorry, when you say it isn't experimental, are you proposing a study where we abuse people enough to give them an unhealthy body image and then see if they gain weight? That seems... well, unlikely to pass an IRB, for one.