How could a high loss of follow-up of those in the LOW FAT DIET group potentially affect the estimate of the measure of association for high dietary fat intake and hypertension?