Pinpointing exact aspects of diets leading to weight gain as been difficult, part of the difficulty conducting human studies is due to controlling what is being consumed long enough to work out what are the most important factors, animal studies which are similar to humans can help point out the right direction; mice have many similarities to humans in physiology and metabolism.
30 different diets varying in protein, fat, and carbohydrate content were fed to model mice for 3 months, which is the human equivalent to 9 human years. 100,000+ measurements were made of body weight changes and micro MRI machines were used to measure body fat. Limitation of this study was due to it being based on mice rather than humans, as for reasons previously stated this study serves as a good clue to what effects different diets are likely to be in humans.
Unequivocal results of the study were the only thing that made animals get fat was consuming more fats in their diets. Carbohydrates up to 30% coming from sugars had no effect; combining fat with sugar had no more impact than that of fat alone. No evidence was found of low protein stimulated greater intake suggesting that there is no protein target. Effects of dietary fat appeared to be uniquely due to fat in the diet stimulated by reward centers in the brain stimulating greater intake.