However, some limitations should also be noted. First, considering that our sample consisted of healthy female undergraduate students, we cannot generalize these results to other groups of interest, such as males or people with overweight or obesity. Second, although our sadness induction was very effective, we cannot be sure if any additional emotions were elicited in our participants (e.g., anger, nostalgia, remorse, anxiety). This seems to be an insurmountable issue however, as it would be very hard, if not impossible, to elicit one single, pure emotion in humans. If additional emotions were elicited, this could have influenced the subjective, physiological or behavioural measures. A third limitation concerns the stimuli we used. When participants were presented with the CS−, this was followed by an empty spoon. We chose this empty spoon to resemble the CS+ trials (in which a spoon with chocolate mousse was presented) as closely as possible, while still making clear to participants that no eating would occur. However, it is possible that the spoon was not neutral enough, given its typical association with eating, which caused participants to still create a link between the CS− vase and food. Although this did not influence conditioned expectancies and desires to eat, it could have influenced the null findings with regard to actual food intake. Fourth, although widely used in eating research with student samples, the BIS-11 has not been validated in this population.