Presidential Column

Learning to Like Foods

In previous columns, we have distinguished between the hard-wired affect associated with taste (especially our love of sweet and salty tastes and dislike of bitter tastes) and the learned affect associated with flavor (i.e., retronasal olfaction). APS Fellow Anthony Sclafani has played a major role in our understanding of the mechanisms that result in learned preferences for food flavors.  In this interview, he and I discuss conditioned food preferences and the interaction of these preferences with hard-wired affect for taste. In addition, we examine the status of fat. Fat obviously has textural properties (e.g, greasy, oily, creamy, viscous, thick). But, does it have a taste as well? Are we born loving fat, do we have to learn to like it, or is our love of fat some combination of the two?

The name “Sclafani” is derived from the Greek “Aesculapii fanum” and means “sacred to the Greek god of medicine” (Wikipedia). Thus it is fitting that Anthony Sclafani has devoted his career to one of the most important modern medical concerns: obesity.

Sclafani was an undergraduate at Brooklyn College, getting his BS degree in 1966 in Psychology. He got his PhD in Biopsychology from the University of Chicago working in the laboratory of S.P. Grossman. He returned to Brooklyn College as an Assistant Professor in 1970, where he rose through the ranks to become a Full Professor in 1980 and a Distinguished Professor in 1994.

He received his first NIH grant in 1971 and has been funded by NIH ever since. His project “Carbohydrate Appetite, Fat Appetite, and Obesity” has been funded since 1984, culminating in a 10-year merit award in 2001.

Sclafani’s work has focused on our love of energy dense foods (high in sugar and fat).  His laboratory has used animal models (rats, mice) to reveal the neural circuitry that controls appetite.


Bartoshuk: One of the many phenomena for which you are famous is the “supermarket” or “cafeteria” diet. You showed that one of the easiest ways to get your lab rats to overeat and thus gain weight was to provide them with a diet containing a variety of commercially prepared, energy dense foods [Sclafani & Springer, 1976]. Most of us can attest that these foods will promote overeating and weight gain in us, too. When did you start studying “supermarket” foods?

Sclafani: Actually, as a graduate student I accidentally discovered that obese rats with hypothalamic lesions would readily eat Fruit Loops (a sweetened cereal) in a novel environment (initially the lab bench, later an open field test cage). This was newsworthy because these animals were thought to overeat because of a satiety deficit, whereas my finding suggested that they had an increased appetite for tasty foods. Several years later, we were interested in studying the eating behavior of obese rats that did not have brain lesions. In order to get otherwise normal rats to overeat, we fed them an assortment of supermarket foods including Fruit Loops, chocolate-chip cookies, and sweetened condensed milk. in addition to their regular lab chow diet. The results were dramatic: The rats virtually ignored the lab chow and overate the supermarket foods to extreme obesity. This finding stimulated renewed interest in the role of diet in promoting overeating.

Bartoshuk: What have you learned about why these foods have so much power over us?

Sclafani: We initially assumed that it was the palatable taste and variety of different foods that promoted overeating in our study. However, the metabolic effects of the foods, which were rich in sugar and fat, also presumably contributed to their obesity promoting effect. To distinguish between the importance of taste and nutrient factors, in the early 1980s we conducted what we thought would be a simple experiment. We fed one group of rats chow and a sweet sugar solution (32 percent sucrose) and another group a “bland” tasting modified starch solution (32 percent Polycose). To our surprise, both groups gained much more weight than did the control group fed only chow and water. The findings seemed to indicate that palatable sweet taste is not essential for the weight promoting effect of sugary drinks.

We were perplexed why our rats consumed so much of the Polycose drink. We considered two possibilities: that Polycose is inherently very tasty to rats or that it has a powerful post-ingestive effect that conditions a strong taste preference. After a series of experiments, we were convinced that rats and other rodents have a taste receptor for Polycose (i.e., starch-derived polysaccharides) that is distinct from their sweet taste receptor and that they experience both types of carbohydrates as very palatable. We summarized our findings in a series of 15 papers that were published as a special issue of Neuroscience and Biobehavioral Reviews in 1987 [Sclafani, 1987]. More recently, we confirmed the idea that different taste receptors mediate sugar and Polycose taste in a study showing that genetically modified mice missing the sweet taste receptor are attracted to the taste of Polycose but not sugar [Zukerman et al., 2009]. Alan Spector’s laboratory at Florida State obtained similar results and we published our papers together in the American Journal of Physiology last year [Treesukosol et al., 2009]. Because people report Polycose to be rather bland tasting, we assumed that our species lacks a Polycose taste receptor. However, recent findings in the sports nutrition field indicate that athletes who sip-and-spit a Polycose-like drink show improved performance  [Carter et al., 2004; Chambers et al., 2009]. The performance results along with fMRI data suggest that humans may indeed have a Polycose taste receptor. However, the Polycose in humans  receptor may activate brain metabolism circuits rather than produce a produce a pleasant taste experience.

Bartoshuk: What about your studies on the post-ingestive effects of carbohydrates?

Sclafani: Yes. As we were investigating Polycose taste, we also studied the post-ingestive effects of carbohydrates. Our experiments revealed that Polycose has potent “reward” effects in the gut that condition strong flavor preferences. Our experiments involved fitting animals with a stomach feeding tube, which we called the “electronic esophagus,” and training them to drink different flavored saccharin solutions  [Elizalde & Sclafani, 1990]. As they drank one flavor (e.g., cherry), a computer turned on a pump that infused Polycose into the stomach as long as the rat drank. In other training sessions, the rats drank a different flavored solution (e.g., grape) and were infused automatically with water. After two or more training sessions, we gave the rats a two-bottle choice test and invariably they selected the flavor that had been paired with Polycose during training. The preferences were usually quite strong (about 90 percent) and persisted over many test sessions even when the animals were no longer being infused. In subsequent experiments, we demonstrated that stomach infusions of sucrose and glucose also conditioned strong flavor preferences, but most interestingly, fructose had weak or no conditioning effect.

We conclude from these studies that although rats are inherently attracted to the distinctive tastes of sucrose and Polycose, their liking of these tastes (and any associated flavors) is further enhanced by the post-ingestive conditioning actions of the carbohydrates. Several studies demonstrate that a similar conditioning process strengthens the flavor preferences of humans. For example, Leanne Birch and coworkers  [Birch et al., 1990] offered young children distinctively flavored drinks at snack time each day. Both flavored drinks were sweetened with aspartame, but one also contained a “tasteless” maltodextrin similar to Polycose. After several exposures (conditioning trials), the children expressed a preference for the flavor that had been paired with maltodextrin during training.

Bartoshuk: Your experiments indicate that animals can taste carbohydrates in their stomach. How is that possible?

Sclafani: Actually, although we infused the carbohydrates into the stomach in our early studies, subsequent experiments revealed that their rewarding action is being mediated by intestinal receptors. The identity of these receptors is the $64,000 question right now. Recently, scientists from several different laboratories found that sweet, bitter, and other taste receptors found in the mouth also exist in the gut — so the gut literally does taste food after it is consumed. This is an exciting discovery, and the function of these gut taste receptors is under intense study. Conceivably, the sweet taste receptors in the gut may be responsible for the flavor conditioning actions of sucrose but our old and new data refute this idea. In particular, fructose stimulates sweet taste receptors but has at best only a weak conditioning effect in the gut. More recently, we found that mice missing the sweet taste receptor show a normal flavor conditioning response to sucrose in the gut. We also know that for Polycose and sucrose to condition flavor preferences they must first be digested to glucose, which indicates that the conditioning process is mediated by a glucose sensor rather than a sweet receptor. There are several types of glucose sensors in the body, but we don’t know yet which one is involved in flavor conditioning.

Bartoshuk: In addition to sugar, dietary fat contributes to food palatability. What do we know about the sensory appeal of fat?

Sclafani: As you know, the palatability of fat-rich foods has long been attributed to the textural properties of solid and liquid fats. Consistent with this idea, years ago we and others reported that rats would readily consume oily drinks prepared with non-nutritive (mineral oil) as well as nutritive oils (i.e., mineral oil and corn oil)  [Ackroff et al., 1990; Mindell et al., 1990]. More recent findings from several laboratories suggest that rodents also have a taste for fatty acids that enhances their preference for nutritive oils and fats. One putative fatty taste receptor is CD36, and we confirmed a prior report that mice missing CD36, unlike normal mice, do not prefer solutions containing small amounts of fatty acids or soybean oil  [Sclafani et al., 2007a]. We found similar deficits in mice missing other taste cell signaling components  [Sclafani et al., 2007b]. The genetically modified mice, however, readily learned to prefer concentrated fat solutions, which we attribute to the post-ingestive reward effects of dietary fat. We first reported 20 years ago that stomach infusions of fat condition flavor preferences in rats and more recently obtained similar effects in mice including knockout mice missing the putative CD36 taste receptor  [Lucas & Sclafani, 1989; Sclafani et al., 2007a]. Thus, as in the case of sugars, the palatability of fat has taste and post-ingestive components, but it also includes mouth feel sensations such as creamy texture. The multiple sensory and post-ingestive aspects of fat appetite are the subject a new edited book  [Montmayeur & le Coutre, 2009].

Bartoshuk: Your work has emphasized the important role of learning in food preference and appetite. What do we know about the brain centers involved in such learning?

Sclafani: Good question. The brain collects the sensory information from the mouth, nose, and gut, integrates it with current metabolic signals of energy need and supplies and food-related memories. The brain quickly decides what, when, and where to eat and when to stop eating. Our understanding of this complex process, while still rudimentary, is rapidly expanding based on innovative research in laboratories around the world. My lab has focused on the brain circuits involved in flavor preference learning, and we have identified some critical sites in the brainstem (parabrachial nucleus) and forebrain (amygdala)  [Touzani & Sclafani, 2009]. In collaboration with Rich Bodnar’s laboratory at Queens College, we have also explored the neurochemical basis of learned food preferences. Dopamine circuits in the nucleus accumbens, amygdala, and elsewhere are essential for the development of new food preferences, although they are less involved in the expression of previously acquired preferences. This is consistent with the extensive research literature on dopamine reward functions. To our surprise, brain opioid receptors, which have a well-established role in food palatability, do not appear to be critical for flavor preference learning although this is still under investigation. The importance of other neurotransmitters implicated in reward processing to food preference learning awaits detailed study.

Finally, let me add that the research in my laboratory would not be possible without the dedication of my long-time collaborators Karen Ackroff and Khalid Touzani and the contribution of graduate and undergraduate students and research technicians. ♦


Ackroff, K., Vigorito, M., & Sclafani, A. (1990). Fat appetite in rats: The response of infant and adult rats to nutritive and nonnutritive oil emulsions. Appetite, 15, 171-188.

Birch, L.L, McPhee, L., Steinberg, L., & Sullivan, S. (1990). Conditioned flavor preferences in young children. Physiology & Behavior, 47, 501-505.

Carter, J.M., Jeukendrup, A.E., & Jones, D.A. (2004). The effect of carbohydrate mouth rinse on 1-h cycle time trial performance. Medicine and Science in Sports and Exercise, 36, 2107-2111.

Chambers, E.S., Bridge, M.W., & Jones, D.A. (2009). Carbohydrate sensing in the human mouth: Effects on exercise performance and brain activity. Journal of Physiology, 587, 1779-1794.

Elizalde, G., & Sclafani, A. (1990). Flavor preferences conditioned by intragastric Polycose infusions: A detailed analysis using an electronic esophagus preparation. Physiology & Behavior, 47, 63-77.

Lucas, F., & Sclafani, A. (1989). Flavor preferences conditioned by intragastric fat infusions in rats. Physiology & Behavior, 46, 403-412.

Mindell, S., Smith, G.P., & Greenberg, D. (1990). Corn oil and mineral oil stimulate sham-feeding in rats. Physiology, 48, 283-287.

Montmayeur, J-P., & le Coutre, J. (2009). Fat detection: Taste, texture, and post ingestive effects. Boca Raton, FL: Taylor & Francis.

Sclafani, A. (1987). Carbohydrate taste, appetite, and obesity: An overview. Neuroscience and Biobehavioral Reviews, 11, 131-153.

Sclafani, A., Ackroff, K., & Abumrad, N. (2007a). CD36 gene deletion reduces fat preference and intake but not post-oral fat conditioning in mice. American Journal of Physiology: Regulatory, Integrative, and Comparative Physiology, 293, R1823-R1832.

Sclafani, A., & Springer, D. (1976). Dietary obesity in adult rats: Similarities to hypothalamic and human obesity syndromes. Physiology & Behavior, 17, 461-471.

Sclafani, A., Zukerman, S., Glendinning, J.I., & Margolskee, R.F. (2007b). Fat and carbohydrate preferences in mice: The contribution of a-gustducin and Trpm5 taste signaling proteins. American Journal of Physiology: Regulatory, Integrative, and Comparative Physiology, 293, R1504-R1513.

Touzani, K., &  Sclafani, A. (2009). Learned flavor aversions and preferences. In R.S. Larry (Ed.), Encyclopedia of neuroscience (pp. 395-399). Oxford, United Kingdom: Academic Press.

Treesukosol, Y., Blonde, G.D., & Spector, A.C. (2009). T1R2 and T1R3 subunits are individually unnecessary for normal affective licking responses to polycose: Implications for saccharide taste receptors in mice. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 296, R855-R865.

Zukerman, S., Glendinning, J.I., Margolskee, R.F., & Sclafani, A. (2009). T1R3 taste receptor is critical for sucrose but not polycose taste. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 296, R866-R876.

APS regularly opens certain online articles for discussion on our website. Effective February 2021, you must be a logged-in APS member to post comments. By posting a comment, you agree to our Community Guidelines and the display of your profile information, including your name and affiliation. Any opinions, findings, conclusions, or recommendations present in article comments are those of the writers and do not necessarily reflect the views of APS or the article’s author. For more information, please see our Community Guidelines.

Please login with your APS account to comment.