How sweet it is…or is it?

As a perception that is universally positive, sweetness is unique. Moreover, sweetness is typically liked outside of its usual context of a food or drink. This is in contrast to all other sensory signals. Salt is great when attached to chips, red is terrific when painted onto a Ferrari, and Gb peerless when part of your favourite tune. But out of context, they loose much of their appeal. Not so, sweet taste. In an earlier discussion, I talked about the ability of sweet tastes to elicit smiles from human and animal newborns reflecting the adaptive priority given to our ability to sense carbohydrates.

Even if preference is always shown for sweetness, it, like every other sensory quality, are liked to greater and lesser extents, in part depending on the type of food or beverage. The role of learning here is obvious – even if we have taken sugar in our tea or coffee for decades, we can quickly learn to like less or even no sugar through repeated exposure. Hence, given different histories of food consumption, we would expect individuals to vary in how sweet foods or beverages should ideally be.

But what is less considered is whether there might be consistent patterns of individual differences in sweetness preferences that either can’t easily be explained by, or may even precede, food choices. Such patterns in bitterness perception and preference have become familiar in recent years through research on the compound 6-n-propylthiouracil or PROP (see for example [
1]). Moreover, the idea of subsets of consumers similarly showing reliable differences in their liking for sweet tastes isn’t new. In 1970 [2], Rose-Marie Pangborn described patterns of variations in both sweet and salty taste preferences. Many of us would identify with the idea of having a “sweet tooth” and research on ‘sweet liking’ has indicated that this is around two thirds of us. Meanwhile, the remaining 30% are known as ‘sweet dislikers’, but this is a misnomer, since it really reflects a preference for sweetness per se that peaks at a lower optimal than it does for the sweet likers – not an actual dislike.

One approach to quantifying the sweet liker/disliker dimension has been to simply obtain the preferred sucrose concentration using a forced-choice procedure. While this is useful and reliable as a measure, it will miss differences in patterns of liking that become evident across a range of sucrose concentrations. Distinct groups of sweet likers and dislikers can also be segregated by patterns of increasing or decreasing preferences for the sweetness of sucrose concentrations. Typically, sweet likers show a monotonic increase as sweetness increases, reaching an asymptote at or close to the highest concentration. Dislikers, on the other hand, often show a steep decline that begins at relatively low concentrations. But it is also clear that some individuals fail to be especially impressed by sweetness at any concentration, instead showing a moderate liking that is insensitive to variations in concentration. These patterns, and the proportions of consumers contained in each, are yet to be definitive, and what’s missing in all the research is estimates of how people vary obtained from large population samples.

The important issue arising from categorizing consumers as sweet likers, dislikers or ‘indifferents’ is whether or not classification into groups predicts anything useful. After all, within a culture, there is more or less general agreement about optimal sweetness levels: fruit should be sweet, but beer shouldn’t. Such ‘agreement’ can be seen as the consequence of cumulative experience of the typical products that are on the market. Nevertheless, given the prevailing sweetness levels of foods and drinks, do some consumers reject some products because they are just too sweet? In other words, does classifying consumers into sweet liking groups predict which of the available foods will be chosen? And if it does, what are the implications for public health?

As we stand, none of these questions can be answered with any certainty. But increasingly, there are research findings that point to a tentative conclusion that sweet liking might a key determinant of food preferences. Julie Mennella and her team recently evaluated their method of assessing sweet liking in a large sample of adults, adolescents and children [
3]. Amongst other findings, they noted a significant positive association between preferred concentrations sucrose in solution and the sugar content of their favourite breakfast cereal.

Since our foods likes (and dislikes) are learned, the importance of the sweet liking dimension is also illustrated by findings that it can influence preference development. In this study [
4], a novel odour was repeatedly paired in solution with the sweet taste of saccharin. This ought to produce a liking for the odour, since the “positive feelings” for the sweetness are transferred to the odour, a process known as evaluative conditioning. And so it did – but only for the sweet likers. Consistent with their classification as individuals who do not find sweetness especially rewarding, the sweet dislikers failed to develop a liking for the novel odour. Since evaluative conditioning is thought to be a very common mechanism for food preference development, the implications of such a finding are critical in helping us understand how food preferences – and hence food choices - can vary within populations.

Just one of the obvious next questions, given the energy provided by carbohydrate sweeteners, is whether or not sweet liking is a risk factor for obesity. Again, we need much larger samples of consumers from which ask this question, but Mennella’s study, plus another one from her lab, failed to find a relationship between body weight (BMI) and preferred sucrose concentration.

Unlike the case with PROP sensitivity, for which variations in specific genes are implicated, there is little strong evidence one way or another that sweet liking is determined by early food experiences, or genetics, or a combination of both. However, Mennella also provides interesting data that bear on the question of the origin of the sweet liker dimension. Her sample was analysed in terms of demographic factors including race, education, income and sex. Her black participants preferred a significantly higher sucrose concentration, as did those with a lower income. Boys preferred a higher concentration than girls, and children higher sweetness than adults. None of these findings rule out a role for taste genetics, but it is important to note that variations in sweet liking reflects hedonics - and not sensitivity to - sweetness. In other words, sweetness of equal intensity is liked differently. Mennella herself notes that black mothers often feed sugar water during infancy and suggests that this is one plausible origin of the greater liking for higher sweetness levels in this group.

If it seems premature to conclude that sweet liking will be an important predictor of food choices, the research to date nevertheless supports a view that this variable is worth further consideration. As with variations in PROP sensitivity and the recently reported thermal tasting, we don’t yet know what ultimate benefits measurement of sweet liking will bring, but understanding the impact of person-to-person differences in consumers’ patterns of sensory and hedonic responses is unlikely to be wasted effort. Sweet liking
could be the factor that mediates the tortuous path between responses to products in consumer evaluations and what products consumers actually choose to eat.

  1. Tepper, B.J., Nutritional Implications of Genetic Taste Variation: The Role of PROP Sensitivity and Other Taste Phenotypes. Annu. Rev. Nutr., 2008. 28: p. 367-388.
  2. Pangborn, R.M., Individual variation in affective responses to taste stimuli. Psychon. Sci., 1970. 21(2): p. 125-126.
  3. Mennella, J.A., et al., Evaluation of the Monell Forced-Choice, Paired-Comparison Tracking Procedure for Determining Sweet Taste Preferences across the Lifespan. Chem Senses, 2011. 36: p. 345-355.
  4. Yeomans, M.R., J. Prescott, and N.G. Gould, Individual differences in responses to tastes determine hedonic and perceptual changes to odours following odour/taste conditioning. Q. J. Exp. Psychol., 2009. 62(8): p. 1648-1664.