Satiety, the suppression of appetite after food consumption, involves a complex integration of cognitive, sensory and post‐ingestive signals generated by the consumed product. A better understanding of how these different signals interact could allow the development of novel products optimised to produce satiety, thereby helping to counteract the effects of obesogenic lifestyles. This report highlights recent studies, conducted largely as part of the BBSRC DRINC initiative, which examined how beliefs about satiety before food ingestion and the sensory experience during ingestion together influence how the consumer responds to food. These studies highlight the integrative nature of satiety and pave the way for updated models of satiety and novel products that should aid consumers' ability to regulate their appetite.

Novel food products that have been optimised to aid appetite control in the face of an obesogenic environment offer a useful and novel approach to tackling the obesity crisis. Factors that promote short‐term overeating, such as large portion size, variety and higher energy density (e.g. Malik et al. 2013), are widely recognised, whereas those that promote the inhibition of appetite – satiety – are less well established. However, a series of recent studies, conducted as part of the portfolio of research funded through the Biotechnology and Biological Sciences Research Council (BBSRC) Diet and Health Research Industry Club (DRINC) initiative at the University of Sussex, is starting to shed light on the complexity of satiety, and more importantly, how consumer expectations about how filling products will be contribute to subsequent responses to ingested nutrients. Satiety, the suppression of appetite following nutrient (energy) ingestion, has long been recognised as a complex phenomenon involving cognitive, sensory, post‐ingestive and post‐absorptive signals generated by the experience of food when ingested and digested (see Halford & Harrold 2012), ideas neatly captured by the satiety cascade model (Blundell et al. 1987). However, while we have a wealth of knowledge on the impact of each of the key satiety signals independently, how these different factors integrate is less clear. Improving our understanding of how these signals interact might allow manufacturers new insight into ways to optimise the satiating effects of their products and so help consumers regulate appetite. The research highlighted here describes recent studies that examined how integration of consumer beliefs, the sensory experience during ingestion and nutrient‐generated post‐ingestive signals modify satiety, paving the way for design rules for future product development.

Sensory‐enhanced satiety A series of studies using modified drinks (Yeomans & Chambers 2011; Chambers et al. 2013; McCrickerd et al. 2014; Yeomans et al. 2014) clearly demonstrate how enhancing consumers' expectations of satiety modifies their responses to nutrient ingestion. In a typical study, participants consumed different versions of the same basic product (typically a drink combining fruit juice and dairy products) and then satiety was assessed by measuring subsequent experience of appetite and intake at a later test meal. All studies combined two key manipulations of the test drink. The sensory characteristics of these drinks were varied using flavourings (e.g. cream flavour) and thickening agents (e.g. tara gum) so that the drinks were either relatively thin and less creamy (low sensory: LS) or noticeably thicker and more creamy (high sensory: HS). This was combined with manipulation of the overall energy content of these drinks to give lower (LE: 70–90 kcal) and higher energy (HE: 270–300 kcal) versions in a way that was not noticeable to consumers. The logic in this design was that the HS versions would generate stronger expectations of satiety than would the LS versions, which was confirmed through sensory testing (McCrickerd et al. 2012). It was then predicted that these expectations would modify the extent to which consumers experienced actual satiety generated by the real difference in nutrient content between LE and HE versions. This hypothesis was strongly supported in all four studies to date: the extent to which the hidden added energy in the HE versions generated satiety depended on the sensory characteristics of the drink, with weak satiety when energy was less expected (LS drinks) and greater satiety when energy was predicted by the thick, creamy sensory characteristics (HS drink). One way of thinking about satiety in this context is to ask, if you were to consume an extra 200 kcal at one meal, would you eat 200 kcal less at the next? If you did so, you would be changing your behaviour to compensate perfectly for the extra energy you had consumed earlier. So expressing changes in intake at the test ad libitum meal as a percentage of the additional energy in the HE relative to the LE versions offers a crude satiety index. When the data from our recent studies were analysed in this way, adding energy to a LS drink, which generated weak satiety expectations, leads to poor compensatory reductions in intake at the subsequent test meal (accounting for between 6% and 48% of added energy). But when the energy was added to the HS drinks, which were thicker, creamier and more predictive of satiety, compensation was much greater (70–100%). These data show a profound influence of consumer expectations on the experience of satiety, suggesting that consumption of energy when it is not expected may be a significant risk factor for short‐term overconsumption. These subtle effects of sensory cues might also explain the apparent difference in satiating difference between some macronutrients. In many studies, the addition of protein generates greater satiety than does the same energy consumed as carbohydrate (for recent review, see Westerterp‐Plantenga et al. 2012). However, in a study when protein was added with minimal changes to the sensory characteristics of a food product, the satiety advantage of protein was lost, whereas when added carbohydrate was matched sensorially to protein, these manipulations had the same effects on satiety (Bertenshaw et al. 2013). Sensory‐enhanced satiety is not limited to textural cues like viscosity and flavour characteristics like creaminess, examined in the BBSRC DRINC‐funded studies. In a separate but related series of studies, similar effects were seen when ingestion of protein was cued by umami taste (Masic & Yeomans 2013, 2014a). These studies again used a pre‐load design, testing a soup with flavour varied by adding monosodium glutamate (MSG) and energy content varied by adding protein or carbohydrate. The rationale behind this approach was that the umami taste system may have evolved as a way of identifying the likely presence of protein in foods (e.g. Naim et al. 1991): consequently, the addition of MSG was predicted to trigger an expectation of protein that might enhance the satiating effects of actual protein ingestion. As predicted, consumers experienced greater satiety in response to added protein, compared with carbohydrate, but only when MSG had been added. The added MSG increased the level of compensation from 32% to 64% of the added energy in the protein‐rich soup but had no effect when the added energy was carbohydrate. A subsequent study (Masic & Yeomans 2014b) using a stronger umami signal, based on a combination of MSG and inosine 5′‐monophosphate (IMP), found some evidence that the experience of umami alone enhanced satiety, but again the umami cue greatly increased the ability to compensate for covert nutrient manipulation by reducing intake at the next meal (from 44% without umami to 70% with). These studies imply that expectations about satiety may be macronutrient specific, and may promote better utilisation of the cued‐macronutrient. Taken together, these series of closely related studies all suggest that expectations that a product may be nutrient‐rich (i.e. satiating), cued by textural, taste and flavour cues, can enhance the consumer's ability to respond to actual ingested nutrients by reducing their energy intake at a later meal. These data suggest that it is the careful matching of sensory characteristics to nutrient content that is the key to maximising post‐ingestive satiety. This is in contrast to the approach taken in the formulation of traditional diet products, where the aim has been to keep the sensory characteristics constant but decrease the actual nutrient content.

Cognitively‐enhanced satiety If expectations about how filling a product will be, generated by the sensory characteristics of a food, can alter consumers' satiety responses to ingested nutrients, then can the same effect be achieved by changing expectations through cognitive approaches, such as product labelling and contextual information? There has been a marked increase in the number of products available to consumers that make explicit reference to their ability to counter post‐ingestive hunger, with these claims based on the products' ingredients (usually higher levels of protein and fibre). However, considering the findings of sensory‐enhanced satiety, it is plausible that these products exert some effects on satiety by generating strong expectations that they will be filling. A further study conducted as part of the BBSRC DRINC programme lends support to that argument (McCrickerd et al. 2014). Again, based on a pre‐load design with covert energy manipulation, participants consumed the same LE and HE dairy‐juice drinks used in previous studies, but rather than enhancing satiety expectations by manipulating sensory characteristics, the key manipulation was persuading the consumers that the product was either specially designed to suppress appetite or to assuage thirst. Whereas compensation for covert energy in the absence of satiety expectations was weak (6%), this increased to 35% when the products were consumed with the belief that they suppressed appetite. Although this cognitive effect was not as strong as a HS manipulation (70% compensation), it clearly demonstrates that beliefs alone can enhance nutrient‐derived satiety.

The risks of rebound hunger If the consumer's experience of satiety is stronger when the post‐ingestive effects of nutrient ingestion are predicted by the product's sensory qualities, what happens if expectations of satiety are high but actual nutrient delivery is less than expected? Put another way, can you fool consumers into feeling full through expectations alone (like a placebo effect)? One surprising finding of the studies outlined earlier is that when consumers expected to feel full but the actual product was LE, participants often felt more hungry just prior to the test meal, and crucially they then consumed more, than if they had consumed the same amount of energy without the expectation of satiety. One interpretation of this rebound hunger effect is that expectations generate preparatory physiological responses in anticipation of the ingested nutrients, possibly involving cephalic phase responses such as insulin release (Smeets et al. 2010). However, when the actual level of nutrients ingested is very low (typically 60–90 kcal), there is a mismatch between the physiological state of the body and sensed nutrient content of the gut following ingestion that leads to the experience of rebound hunger. This idea has important implications for the formulation of diet products. Could low‐energy products that consumers believe will suppress their urge to eat actually promote increased intake later because of rebound hunger?

Real‐world implications What new insights does the work reviewed here provide outside the academic understanding of the nature of satiety? The key insight could be called ‘honest signalling’: the idea that satiety responses are dependent on the match between what consumers expect to happen and their actual experience once they consume a product. To illustrate this idea, consider how filling we expect a drink to be. Most of the time beverages are consumed because we consider ourselves to be thirsty. If that beverage happens to contain appreciable amounts of nutrients, any effects of those nutrients on satiety may be reduced by the consumer's expectation that the drink quenches thirst rather than reduces hunger. Ingesting nutrients when they are unexpected could promote energy storage rather than utilisation, which would fit with concerns that energy‐rich drinks are a risk factor for weight gain. In contrast, soups are consumed with the expectation that they will suppress hunger, and energy consumed as soup is reliably more filling than energy consumed as drinks (e.g. Mattes 2005). Here both the general belief that soups will be filling and the sensory characteristics of soups reliably predict nutrients and so the cues signal the post‐ingestive effects: they act as ‘honest signals’. If we extrapolate from drinks and soups to products in general, these ideas raise real issues about how foods are labelled in relation to their actual energy content. If foods are labelled as low energy, they are likely to generate expectations of weaker satiety. This may, counter‐intuitively, be counter‐productive for consumers trying to control their appetite since the low satiety expectations may reduce the impact of the actual nutrient content on satiety. Conversely, products, such as diet foods, that have been carefully developed to ‘taste’ like their high energy equivalent, but which have reduced energy content, might lead to high satiety expectations that when not followed by nutrient delivery lead to rebound hunger. Thus well‐intentioned attempts to produce products of benefit to weight‐concerned consumers may unwittingly impact negatively on appetite control.

More questions than answers The study of cognitive and sensory‐enhanced satiety is still in its early days. The research outlined here suggests that it is important to consider how psychology and physiology integrate in the control of appetite. But to be able to fully utilise these findings to develop novel satiating food products, there are numerous questions that need to be answered. How do cognitive and sensory signals of satiety interact? Are cognitive manipulations effective only when they are confirmed by sensory experience? If expectations improve the satiating effects of nutrients, can they be used to offset small reductions in nutrient content or portion size? And how low can nutrient content be reduced before the risk of rebound hunger and consumer dissatisfaction offsets any benefits to consumers? These and other questions need to be addressed by future research. But the recent findings highlighted in this report demonstrate that to understand satiety, it is essential to look at a product in terms of its labelling and context, sensory profile and nutrient content. When all three of these aspects are optimised, then food and drink products will be both satisfying to ingest and less likely to lead to short‐term overconsumption.

Acknowledgements Studies on cognitive and sensory enhancement of satiety in drinks reported in this paper were funded by the BBSRC under the DRINC initiative (Grant No. BB/H004645/1): studies on umami and satiety were funded as a PhD studentship by Ajinomoto Co, USA.

Conflict of interest The author has no conflicts of interest.