We saw a very large reduction in caloric expenditure in people who were already very fit runners. So body weight changes over the course of the experiment are unlikely to explain it.