One of the most interesting aspects of the paper is that the investigators measured a variety of food properties (energy density, fat, starch, sugar, fiber, water content, palatability), and then determined which of them explained the SI values most completely.
As I've explained on this blog before, the brain "shuts down" feelings of fullness when you consume highly palatable foods, and this can be mimicked by administering drugs (such as marijuana or opiates) that stimulate these brain pathways directly (2, 3). That's why there's always room for dessert. The mechanism of this effect has been partially worked out, and involves communication between brain regions that regulate reward and food intake (4, 5, 6).
Lucky for us, the investigators measured the palatability of each food and compared it to that food's SI. If food palatability does indeed shut down the mechanisms in the brain that normally constrain food intake, we would expect that foods rated as more palatable would have a lower SI. Here's what they found:
Palatability was one of the two strongest predictors of SI. It was tied at first place with energy density (calories per gram of food)*. Lower energy density, and higher water, fiber and protein content were associated with higher SI, while palatability and fat content were associated with lower SI. Starch and sugar content were not associated with SI. SI predicted subsequent food intake, such that foods with a higher SI led to less calorie intake two hours later.
This study, along with many others, suggests that focusing on simple foods that have a lower energy density leads to greater fullness and less subsequent food intake, and conversely that highly palatable energy-dense foods promote excessive food intake. Potatoes, sweet potatoes, meats, fish, vegetables, fruits, rice and beans are foods with a moderate level of palatability and energy density, and are consequently helpful for weight loss and maintenance. Conversely, baked goods, candy, ice cream and fried foods have the lowest SI, reflecting their extreme palatability and energy density. These are exactly the same foods people eat to relieve stress, which reinforces the fact that they are hyper-palatable and hyper-rewarding (7). In my opinion, these are among the most fattening foods, and the obesity literature as a whole supports this.
The paper finished with this statement, some of which may sound familiar to regular WHS followers:
...simple, 'whole' foods such as the fruits, potatoes, steak and fish were the most satiating of all foods tested. Interestingly, many plant foods such as beans, lentils and potatoes contain antinutrients which can delay or inhibit the absorption of nutrients or affects gastrointestinal hormone release. These factors could contribute to their greater satiating powers.
* Also, water content, but this is basically a proxy for energy density.