Does flavor-nutrient learning promote or protect against diet-induced obesity? Individual differences in conditionability predict resistance to weight gain in rats.

Flavor-nutrient learning (FNL) refers to learning associations between a foods flavor and the rewarding appetition signals that arise from post-oral nutrient sensing during or after a meal. In rodent models FNL reliably produces strong flavor preferences and increased intake of nutrient-paired flavors, implicating FNL as a presumptive obesogenic influence in the modern environment. However, evidence that FNL plays a causal role in diet-induced obesity is ambiguous. We have previously shown that degree of weight gain on a high-fat/sugar diet is associated with stronger FNL responses, but direction of causation was unclear. This paper reports three experiments investigating whether individual differences in FNL conditionability are linked to obesity proneness prior to obesity onset. Two experiments comparing selectively-bred obesity-prone vs resistant strains found no strain differences in FNL. A third study in lean, outbred rats evaluated whether baseline individual differences in FNL prospectively predict weight gain on a cafeteria diet. Unexpectedly, rats who showed the strongest learned increase in intake of a nutrient-paired flavor subsequently gained the least weight when switched to cafeteria diet, suggesting FNL protects against weight gain. In fact, individual differences in FNL explained a portion of variance in cafeteria weight gain over and above measured kcal intake, implying a function for FNL in adaptively modulating metabolic responses to energy intake. Collectively, several studies have now shown individual differences in obesity proneness to be either positively correlated, uncorrelated, or negatively correlated with FNL, calling for a more nuanced view of how appetition influences intake and energy balance.