Would seem more obtuse to me to assume that a gain of 14% for mice in a lab-situation would translate to a similar gain for humans in a real-life situation.