It's well known that we secrete insulin in response to dietary carbohydrates. Specifically, we convert most non-fiber carbohydrates into glucose, whose presence in the blood is sensed by cells in the pancreas, which secrete insulin in response. But there's been a lot of work done in recent decades showing that the insulin response doesn't always match what we'd expect from this simple explanation.
For example, certain artificial non-nutritive sweeteners have been shown to elicit an insulin response. Certain fatty foods have a similar effect, including foods with little or no carbohydrate content, like butter. Why is this? What produces this response, if not the glucose-sensing islet β-cells in the pancreas? This study and others have posed the question, or at least acknowledged the gap in understanding: "the mechanism of the cephalic phase insulin response to meal ingestion has not been established in humans." This study, which looks at the subjects' metabolic responses to the taste of fat, rather than the ingestion of it, may give us a clue: it's your brain.
I have a testable hypothesis that fits what we know so far, and explains the positive correlation between reflexive insulin release and obesity. If you're on a typical American high-carbohydrate diet, your body is likely conditioned to expect a spike in blood sugar any time you eat a meal. If this conditioning is responsible for the "cephalic phase" insulin response, the stimuli I mentioned above should be expected to produce this result.
Insulin secretion can be conditioned in rats in response to arbitrary environmental cues, like site and sound, very much like Pavlov's salivating dog. There's a considerable body of work on the cephalic phase insulin response in humans, but I haven't seen any work that investigates whether it's a conditioned response. What I'd like to see is a similar test comparing the response in subjects on a typical diet versus subjects on a long term low carbohydrate diet. My prediction would be that the experimental group (the low-carbers), being deconditioned, would display little if any cephalic phase response.
How does a curious guy with an interest in this subject get a study like this done without spending his own money on it?