I’m not sure that’s right. People have different metabolisms and also it would be nigh impossible to control for activity levels. One person’s 2500 calorie diet may lead to weight loss, another person’s might not, grapefruit or no grapefruit. Wouldn’t it make more sense to have the subjects continue on with their normal eating habits? The only problem would be, as Jackmanii points out, if subconsciously the grapefruit-eating crowd adjusted their dietary habits (although one aspect of this was controlled for in the placebo + apple juice vs grapefruit capsule + apple juice group.)