Sensitivity to Viscosity Changes and Subsequent Estimates of Satiety across Different Senses

Authors: Pellegrino, Robert1,2; Jones, Jourdan1; Shupe, Grace E.1; Luckett, Curtis R.1

1 Department of Food Science, Institute of Agriculture, University of Tennessee, Knoxville, Tennessee

2 Smell & Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Dresden, Germany

Corresponding author: Curtis R. Luckett; Department of Food Science, University of Tennessee, 2510 River Drive, Knoxville, TN 37996, U.S., [email protected]

Keywords: Viscosity, Satiety, Tactile, Auditory, Vision

1 Abstract While it is widely accepted that texture is a multisensory property, little research has been published regarding how we use senses other than touch to assess texture. In beverages, humans use texture (i.e. viscosity) information to estimate calories and expected satiety. This study was designed to compare and contrast the sensitivity of humans to changes in viscosity and estimated satiety through different sensory modalities. Milk samples of varying viscosities were constructed, and 49 participants were asked to perform a series of 2-alternative forced choice tests and identify which sample was thicker. Sensitivity to viscosity changes across different sensory modalities was determined by having each participant consume the samples, listen to the sample pouring, and observe clear vials of the samples. Using vision, participants were notably less sensitive to changes in viscosity when compared to hearing or oral tactile sensations. Interestingly, oral tactile sensations and hearing were almost identical in their viscosity difference threshold (0.346 cP and 0.361 cP, respectively). Similar patterns were observed when the participants were asked to estimate how full they would be after consuming beverage stimuli varying in viscosity. Expected caloric values and satiation were found to change with thickness level when participants were assessing the stimuli through the auditory or tactile modalities. However, these measures of expected caloric value and satiation did not change as a function of viscosity for visual assessment, suggesting the assessment of caloric density and satiation are linked to specific sensory modalities’ ability to detect viscosity. This study highlights the relative importance of vision, audition, and touch to forming our sensory judgements regarding viscosity and subsequent satiety estimations. Introduction

Viscosity represents a primary mechanical characteristic in food and is determined by the molecular components of the food such as water, protein, carbohydrates and fat (Vliet & Walstra, 1980).

Consequently, these components and others help cue individuals to changes in viscosity to determine perceptual outcomes. For instance, wine judges use viscosity to determine alcohol content by visually inspecting the speed with which the “legs” descend the inside wall of a glass (Nurgel & Pickering, 2005) and shoppers use the thickness of yogurt to determine fat content (Bruzzone, Ares, & Giménez, 2013).

Indeed, humans use viscosity as a strong indicator of nutritional properties of a beverage (McCrickerd et al. 2012). These properties include, caloric density, satiety (fullness), and satiation (fullness over time) .

For instance, thick beverages elicit a higher expected satiety than thin beverages (McCrickerd, Chambers,

Brunstrom, & Yeomans, 2012), and this perception relates to physiological changes in the body such as increased insulin and pancreatic polypeptide (PP) (Yeomans, Re, Wickham, Lundholm, & Chambers,

2016). The aforementioned findings bolster the concept of a satiety cascade, which states early cognitive and sensory information integrate with post-ingestive signals and behaviors such as a suppressed appetite after eating (Blundell, Rogers, & Hill, 1987).

Typically, surface-contacting physical instruments such as rheometers measuring resistance to stress can be used to objectively characterize the viscosity of liquids (i.e. beverages). However, measuring human sensitivity to viscosity is different than many other food related perceptions, in that an absolute threshold cannot be calculated. More clearly, a liquid cannot have a complete absence of viscosity, so to measure sensitivity, difference thresholds can be used. Difference threshold, or just noticeable difference (JND), is the minimum change in stimulus intensity before a change in perception is noted (Lawless & Heymann, 2010). For the purposes of viscosity, this can be thought of as the change in viscosity needed for a population to notice a sample is thicker than another. In determining difference thresholds, a series of 2-alternative force choice tests (2-AFC) are commonly used (Ulrich & Miller,

2004; Ulrich & Vorberg, 2009).

Most studies measuring viscosity perception in humans have concentrated on only oral tactile sensations; however, texture evaluation is multisensory in nature. As defined by the International

Standards Organization (ISO), texture is “all the mechanical, geometrical and surface attributes of a product perceptible by means of mechanical, tactile, and, where appropriate, visual and auditory receptors”. Texture sensations, such as crispness, have been shown to be dependent on multiple sensory modalities. Crispness is heavily dependent on the acoustics during chewing (Vickers & Bourne, 1976) while vision aids with expectations and surface texture (Chen, 2007). Currently, the role of various sensory modalities in constructing viscosity judgements and expected satiety is unknown.

Two experiments were conducted to understand how we use sensory information from different modalities to assess beverage thickness, satiety, satiation, and caloric density. The first experiment was designed to determine the sensitivity of auditory, oral tactile, and visual sensory input to changes in viscosity. The second experiment was designed to assess how the three sensory modalities from experiment 1 (audition, oral tactile, and vision) are used to estimate the nutritional properties of a beverage.

Experiment 1

Materials and Methods

Participants

Forty-nine volunteers (32 women) participated in this study with ages ranging from 21 to 64 (32.2 ±

11.2). Based off a prescreener, only individuals that reported a liking for dairy products with moderate, to high consumption were chosen to participate. Individuals who reported visual, auditory, or gustatory/olfactory impairments (chronic or acute) were not selected to participate. Additional exclusion from participation included pregnancy, abnormal oral health, and dietary restrictions (e.g. dairy allergy).

All participants were asked to not smoke or eat 1 hour prior to the start of the study. All participants signed an informed consent and were compensated for their time spent participating. This experiment

(and the follow-up) protocol was conducted according to the Declaration of Helsinki for studies on human subjects and approved by the University of Tennessee IRB review for research involving human subjects

(IRB #16-03417-XP).

Preparation of stimuli

Iota-carrageenan (IC) was used to thicken milk at several concentrations, ranging from 0.00% to 0.111%

(w/v). To account for possible effects of reference viscosity, two samples were used as the reference milk

1.714 cP and 1.768 cP. Both references were the same for each modality condition in the study, visual, auditory, and oral tactile. The milk stimuli were made my mixing IC with 1% milk (Great Value, Wal-

Mart, Bentonville, AR) and heated under constant agitation to 63 °C. The milk was then cooled to 25 °C in an ice bath, with constant agitation. Next, the thickened milk was homogenized for 3 minutes using a high-speed homogenizer (T25 Ultra Turrax, IKA Works, Staufen, Germany), and cooled to 4 °C. Lastly, the mixture was homogenized again for 3 minutes. The samples were stored at 4°C prior to serving.

The viscosity of each sample was measured with a rheometer (AR 2000, TA Instruments, New

Castle, Delaware, USA) using parallel plate geometry (40mm diameter). Measurements were at 25°C, using a shear rate of 143 s-1. This procedure was also used to measure the viscosity of different milk types, reduced-fat milk (1% and 2%), half-half, and heavy cream (Great Value, Wal-Mart, Bentonville,

AR).

Stimuli presentation For the visual condition, 30 mL scintillation vials were filled with 20 mL of milk at a specific IC concentration. These vials had a white top with a clear glass body through which the milk was visible and easily moved by titling the vial back and forth. Vials were presented at 4 °C.

To produce the audio, 150 mL of each milk stimuli at 4 °C was poured from a height of 20 cm into a 400 ml beaker. The sound produced, was recorded using a microphone (Microsoft PnP, Redmond,

WA) connected to a PC running Audacity 2.1.0. The samples were poured at a rate in which the pour was completed in 5.0 ± 0.5 sec. Each sample was recorded multiple times and only files that exhibited traits of a continuous pour were used. The raw audio files were treated with a noise reduction (12 dB, 6 sensitivity, 3 bands).

For the oral tactile condition, 20 mL of a milk was placed in a 60 mL plastic black container with a fitted top. Milk samples were served at 4 °C.

Procedure

The tests were spread over a three-day period, one for each of the three modalities – visual, auditory, and oral tactile. A series of two-alternative force choice (2-AFC) tests were used to determine the sensitivity for each modality. In this method, two milk samples were presented at the same time, either in vials, as auditory clips, or in opaque containers for visual, auditory and oral tactile modality conditions respectively. For each condition, a reference and a thicker non-reference were presented in a randomized presentation order (left or right, or first or second for sounds clips). The participants were asked to select which sample was thicker using the sense under investigation for the testing day. For auditory, two recordings were presented on a screen allowing the individual to play the sound of each (at 70dB through headphones [Lightweight On-Ear Headphones, AmazonBasics, Amazon, Seattle, WA, USA] and then choose the thicker sounding clip. The order of pairing was randomized across participants. After the evaluations of all pairs, demographic information such as age and gender were recorded. All evaluations took place in a controlled sensory booth and took approximately 15 minutes to complete. Individuals were compensated for their participation at the conclusion of the study.

Statistical Analysis

To assess sensitivity to changes in viscosity, regression models predicting the ability of correctly identifying the thicker sample were made for each modality. Several models were fit, examining the relationship between the percent correct answers and the difference in viscosity. Using minimum Akaike information criterion (AICc) estimation, a 3-parameter exponential model was chosen.

For experiment 1 and 2, all analysis was done in JMP (version 13.0; SAS Institute, Cary, NC,

USA), with a statistically significant difference defined as p < 0.05.

Results

In general, sensitivity to minute changes in viscosity was observed across all three senses. Figure 1 demonstrates the model-fitted regression lines for each modality.

After reaching the percent correct often associated with difference threshold (76 %), auditory and tactile continued to show higher sensitivity to changes in viscosity (ΔV). When ΔV was high (greater than

2 cP), auditory and tactile showed extremely high frequency of correct answers, while vision failed to reach those high levels of certainty until ΔV was greater than 3.5 cP. While it is easy to focus on the absolute difference thresholds, higher ΔV’s provide insight into more realistic changes in fat-induced viscosity changes which relay a caloric-density shift. For instance, typical changes in viscosity from milk with 1% fat to 2% fat contents is approximately 0.11 cP.

Experiment 2

Materials and Methods

To examine how different senses estimate satiety, participants were presented thin and thick milk samples in the same manner as Experiment 1 and asked several questions about ingestive outcomes. It was hypothesized that participants are less aware of their sensitivities to viscosity through senses other than tactile. Therefore, we asked participants to report their confidence level in these judgements.

Participants

Fourty-eight volunteers (31 women) participated in this study with ages ranging from 21 to 64 (32.6 ±

11.2). Similar to Experiment 1, only individuals that reported a liking for dairy products with moderate to high consumption were chosen to participate. Individuals who reported visual, auditory, or gustatory/olfactory impairments (chronic or acute) were not selected to participate. Additional exclusion from participation included pregnancy, abnormal oral health, and dietary restrictions (e.g. dairy allergy).

All participants were asked to not smoke or eat 2 hours prior to the start of the study.

Stimuli

Milk stimuli were made at 2 high and 2 low viscosity levels using varying amounts of iota-carrageenan.

The change in the viscosity (ΔV) within the thin and thick samples was small (0.62 cP) while the ΔV between thinnest and thickest sample was large (4 cP). Each concentration was made using the same procedure as Experiment 1 using 1% milk (Great Value, WalMart, Bentonville, AR) and each thickness level was verified with a rheometer using the same procedure as Experiment 1. All samples were prepared in the same manner as Experiment 1 for visual (in vial), auditory (sound clip) and oral tactile (in black container) conditions.

Procedure

Testing took place in a well-lit, off-white booth over three consecutive days between 8 – 12 AM. One sensory modality was addressed on a single day. Within the booth, was a computer monitor with keyboard to record answers as well as an empty, label-free 8 oz plastic bottle. At the beginning of the test, participants were informed that many questions would relate to an 8 oz. serving size, and to use this bottle as a reference. A reminder to use this reference appeared for each question about satiety or satiation.

Before any samples were presented, participants were asked to rate “How hungry are you right now?” on a 100-point line scale with anchors from “Not hungry” to “Extremely hungry”. Next, they were given a sample or sound clip and asked to drink, visualize, or listen, respectively. During sampling, panelists were asked a series of questions: 1) “This sample has a desirable texture.” (7-point Likert scale, “Strongly disagree” to “Strongly agree”), 2) “How full do you think you would get drinking an 8 oz serving of this sample?” (100-point line scale, “Not at all” to “Extremely”). 3) “For how long do you think you will feel full from drinking an 8 oz serving of this sample?” (100-point line scale, “Hungry again at once” to “Full for 5 or more hours”), and 4) “How many calories would you expect to be in an 8 oz serving of this sample?” (text box to enter number). These questions were repeated for all samples which were served in a randomized order. After all samples were evaluated, participants were asked “In general, how sure were you in your answers regarding the caloric values of the samples presented today?” on a 100-point line scale from “Extremely unsure” to Extremely sure”.

Statistical Analysis

For each modality, a repeated-measures ANOVA was created for the ratings of satiety, satiation, and caloric density estimation with the viscosity of the sample as the predictor. Three contrasts with

Bonferonni corrections were used to test differences within the thin and thick samples and the average difference between thin and thick samples.

To determine differences in hunger and confidence of answers across modalities, a repeated- measure ANOVA was used. Pairwise comparisons were done with Tukey-Kramer HSD method.

Additionally, indirect effects of confidence to estimate expected ingestive ratings between the thick and thin samples were analyzed with mixed repeated-measures ANOVA.

Results and Discussion The participants’ pre-experiment hunger levels did not differ across the 3 experiemntal sessions (p =

0.4139). Table 1 shows the results of Experiment 2. Overall, all senses estimated higher satiety, satiation, and caloric density as the milk samples became thicker. Using all senses, participants estimated differences in expected satiety between the thick and thin samples, but not within them. However, vision did not show satiation nor caloric density differences as samples moved from thin to thick. Through audition, participants estimated more and larger differences among the samples than the other two senses, having higher mean separation for caloric estimators between thick and thin samples, and significantly estimating a higher satiety and satiation for the more viscous sample within the thick sample set.

Contrary to our hypothesis, the participants did show a similar confidence in auditory judgements

(M=45.15, SD=16.34) compared to oral tactile (M=48.40, SD=16.61;p = 0.23; Figure 2). Additionally, the participants rated their confidence lower in vision (M=41.02, SD=16.48) compared to oral tactile (p <

0.001) and trending toward auditory (p = 0.084). There were no indirect effects of confidence on estimating ingestive outcomes between the thin and thick samples (p > 0.05).

Overall Discussion

In the current study, we demonstrate the high sensitivity of senses to detect changes in thickness and how these changes align with estimations of caloric density and post-meal satiation. The thicker the beverage, the more calories individuals expected the beverage to contain and the longer they expected to be full.

This perception may rise from using thickness as a proxy for increased fat content. Indeed, several studies point to thickness as a main attribute to creaminess along with smoothness (de Wijk, Prinz, &

Janssen, 2006; Kokini, 1987; Kokini & Cussler, 1983). In relation to dairy fat, thickness and smoothness from evenly spaced globules of fat found in homogenized milk produce the sensation of creaminess

(Richardson, Booth, & Stanly, 1993) and an increased perception in creaminess is related to an increase in fat content (Clegg, Kilcast, & Arazi, 2003; Prindiville, Marshall, & Heymann, 1999). A 2013 study showed that artificially thickened yogurts increases the perception of creaminess (Bruzzone et al., 2013). Our research points to thickness alone can estimate caloric density similar to previous research

(Hogenkamp, Stafleu, Mars, Brunstrom, & de Graaf, 2011; McCrickerd et al., 2012; Yeomans et al.,

2016) and this estimation happens long before it enters the mouth.

Oral tactile and auditory proved to be the most sensitive senses to changes in viscosity and reflectively, were also more discriminating in caloric estimations when compared to vision. In other words, auditory stimuli may have more of a role than currently known in evaluating the nutritional content of beverages. There is very little research on how we use our auditory system to assess food characteristics. However, several experiments from Velasco et al. have shown the accuracy of detecting auditory changes of water poured into a variety of vessels at different temperatures (Velasco, Jones, King,

& Spence, 2013a, 2013b). In these studies, individuals are presented with clips of water being poured into different vessels (glass, porcelain, paper, and plastic) and asked to correctly identify if the water was hot or cold. People were not only able to detect differences above chance, but in follow-up studies were aware of the differences in pitch and tempo (Wang & Spence, 2017). Interestingly, in another study, prior to presenting water samples, less than half of the sixty individuals thought they would be able to decipher differences (Velasco et al., 2013a). To the latter point, we were surprised to find participants so confident in their estimation of calories which was at par with oral sensation and above visual. To our knowledge, this is the first study to demonstrate the ability of estimating expected calories with auditory cues; however, the expectation of auditory cues on textures has been widely explored (Spence & Wang, 2015).

Sound is not only an integral part of textures, but can be manipulated to influence expectations of them such as the aforementioned temperature, carbonation, and crispness (Velasco et al., 2013b; Zampini &

Spence, 2004, 2005).

Although counter-intuitive, many studies have shown a visual disconnect in what the eyes see and what actually goes into the body. In other words, things that are visually matched by weight or size may have high energy density with lower expected satiety such as a chocolate bar (Brunstrom, Shakeshaft, & Scott-Samuel, 2008). Similarly, one study has shown that visual appearance (as perceived volume) only explained a small proportion of variance in expected satiation (Brunstrom, Collingwood, & Rogers,

2010). However, even though at a smaller extent than the other senses, vision can be used to differentiate expected satiety levels between the thin and thick samples. Similarly, other studies have shown that physical characteristics of food, such as density, have an effect on satiety and consumption (Arboleya,

García-Quiroga, Lasa, Oliva, & Luis-Aduriz, 2014; Osterholt, Roe, & Rolls, 2007).

While this study gives clear evidence of viscosity sensitivity to beverages across sensory modalities, it was not designed in a way to answer some important questions. For example only 2 references, which were extremely close in viscosity (ΔV = 0.054 cP), were used in this study. To better understand the psychophysics of viscosity assessment across the sensory modalitites, a wider range of starting viscosities would be needed. Additionally, while this study focused solely on viscosity/thickness, it is possible other textural attributes of the stimuli were modified by the addition of IC. Lastly, the sensitivity to viscosity was analyzed on the study population as a whole. For a more detailed assessment of how sensitivity to viscosity across different modlalities influence caloric judgements, attempts should be made to look at individual participants.

Future directions should seek to build off the findings of this study and previous research showing that modifying sensory cues related to texture could be a way to combat modern issues related to excessive caloric intake. Previous research has shown that expected ingestive outcomes influences eating behaviors such as bite size and oral processing time. For instance, higher expectation of caloric value may cause someone to take smaller bites and increase oral processing time leading to less consumption

(Forde, Almiron-Roig, & Brunstrom, 2015; McCrickerd, Chambers, & Yeomans, 2014). Taking these findings into account, the possibility of modifying sensory properties of food to decrease portion selection and overconsumption becomes apparent.

Conclusion Our study provides further evidence that subtle changes to sensory-cues related to texture can affect expected satiety and satiation. In addition, it is evident that the assessment of caloric density, satiety, and satiation are linked to specific sensory modalities’ ability to detect viscosity. Of the three sensory inputs assessed in this study, vision was identified as less sensitive to viscosity and used to a lesser extent to judge estimated caloric density, satiety, and satiation. On a higher level, this study provides conclusive evidence that viscosity is multisensory in nature. References

Arboleya, J.-C., García-Quiroga, M., Lasa, D., Oliva, O., & Luis-Aduriz, A. (2014). Effect of highly aerated food on expected satiety. International Journal of Gastronomy and Food Science, 2(1), 14– 21. https://doi.org/10.1016/J.IJGFS.2013.12.002 Blundell, J. E., Rogers, P. J., & Hill, A. J. (1987). Evaluating the satiating power of foods: implications for acceptance and consumption. Food Acceptance and Nutrition / Edited by J. Solms ... [et Al.]. Retrieved from http://agris.fao.org/agris-search/search.do?recordID=US201302689288 Brunstrom, J. M., Collingwood, J., & Rogers, P. J. (2010). Perceived volume, expected satiation, and the energy content of self-selected meals. Appetite, 55(1), 25–29. https://doi.org/10.1016/J.APPET.2010.03.005 Brunstrom, J. M., Shakeshaft, N. G., & Scott-Samuel, N. E. (2008). Measuring “expected satiety” in a range of common foods using a method of constant stimuli. Appetite, 51(3), 604–614. https://doi.org/10.1016/j.appet.2008.04.017 Bruzzone, F., Ares, G., & Giménez, A. (2013). Temporal aspects of yoghurt texture perception. International Dairy Journal, 29(2), 124–134. https://doi.org/10.1016/J.IDAIRYJ.2012.10.012 Chen, J. (2007). Surface Texture of Foods: Perception and Characterization. Critical Reviews in Food Science and Nutrition, 47(6), 583–598. https://doi.org/10.1080/10408390600919031 Clegg, S., Kilcast, D., & Arazi, S. (2003). The Structural and compositional basis of creaminess in food emulsion gels. In Proceedings of the third international symposium on food rheology and structure. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.602.5670 de Wijk, R. A., Prinz, J. F., & Janssen, A. M. (2006). Explaining perceived oral texture of starch-based custard desserts from standard and novel instrumental tests. Food Hydrocolloids, 20(1), 24–34. https://doi.org/10.1016/J.FOODHYD.2005.02.008 Forde, C. G., Almiron-Roig, E., & Brunstrom, J. M. (2015). Expected Satiety: Application to Weight Management and Understanding Energy Selection in Humans. Current Obesity Reports, 4(1), 131– 140. https://doi.org/10.1007/s13679-015-0144-0 Hogenkamp, P. S., Stafleu, A., Mars, M., Brunstrom, J. M., & de Graaf, C. (2011). Texture, not flavor, determines expected satiation of dairy products. Appetite, 57(3), 635–641. https://doi.org/10.1016/j.appet.2011.08.008 Kokini, J. L. (1987). The physical basis of liquid food texture and texture-taste interactions. Journal of Food Engineering, 6(1), 51–81. https://doi.org/10.1016/0260-8774(87)90021-5 Kokini, J. L., & Cussler, E. L. (1983). Predicting the Texture of Liquid and Melting Semi-Solid Foods. Journal of Food Science, 48(4), 1221–1225. https://doi.org/10.1111/j.1365-2621.1983.tb09196.x Lawless, H. T., & Heymann, H. (2010). Measurement of Sensory Thresholds (pp. 125–147). https://doi.org/10.1007/978-1-4419-6488-5_6 McCrickerd, K., Chambers, L., Brunstrom, J. M., & Yeomans, M. R. (2012). Subtle changes in the flavour and texture of a drink enhance expectations of satiety. Flavour, 1(1), 20. https://doi.org/10.1186/2044-7248-1-20 McCrickerd, K., Chambers, L., & Yeomans, M. R. (2014). Does modifying the thick texture and creamy flavour of a drink change portion size selection and intake? Appetite, 73, 114–120. https://doi.org/10.1016/J.APPET.2013.10.020 Nurgel, C., & Pickering, G. (2005). Contribution of glycerol, ethanol and sugar to the perception of viscosity and density elicited by model white wines. Journal of Texture Studies, 36(3), 303–323. https://doi.org/10.1111/j.1745-4603.2005.00018.x Osterholt, K. M., Roe, L. S., & Rolls, B. J. (2007). Incorporation of air into a snack food reduces energy intake. Appetite, 48(3), 351–358. https://doi.org/10.1016/J.APPET.2006.10.007 Prindiville, E. A., Marshall, R. T., & Heymann, H. (1999). Effect of Milk Fat on the Sensory Properties of Chocolate Ice Cream. Journal of Dairy Science, 82(7), 1425–1432. https://doi.org/10.3168/jds.S0022-0302(99)75369-5 Richardson, N. J., Booth, D. A., & Stanly, N. L. (1993). Effect of homogenization and fat content on oral perception of low and high viscosity model creams. Journal of Sensory Studies, 8(2), 133–143. https://doi.org/10.1111/j.1745-459X.1993.tb00208.x Spence, C., & Wang, Q. (2015). Sensory expectations elicited by the sounds of opening the packaging and pouring a beverage. Flavour, 4(1), 35. https://doi.org/10.1186/s13411-015-0044-y Ulrich, R., & Miller, J. (2004). Threshold estimation in two-alternative forced-choice (2AFC) tasks: The Spearman-Kärber method. Perception & Psychophysics, 66(3), 517–533. https://doi.org/10.3758/BF03194898 Ulrich, R., & Vorberg, D. (2009). Estimating the difference limen in 2AFC tasks: Pitfalls and improved estimators. Attention, Perception, & Psychophysics, 71(6), 1219–1227. https://doi.org/10.3758/APP.71.6.1219 Velasco, C., Jones, R., King, S., & Spence, C. (2013a). “Hot or Cold?” On the Informative Value of Auditory Cues in the Perception of the Temperature of a Beverage. In ((( ABA ))) Audio Branding Academy Yearbook 2012/2013 (pp. 175–187). Nomos Verlagsgesellschaft mbH & Co. KG. https://doi.org/10.5771/9783845243559-175 Velasco, C., Jones, R., King, S., & Spence, C. (2013b). The Sound of Temperature: What Information do Pouring Sounds Convey Concerning the Temperature of a Beverage. Journal of Sensory Studies, 28(5), 335–345. https://doi.org/10.1111/joss.12052 Vickers, Z., & Bourne, M. C. (1976). A psychoacoustical theory of crispness. Journal of Food Science, 41(5), 1158–1164. https://doi.org/10.1111/j.1365-2621.1976.tb14407.x Vliet, T., & Walstra, P. (1980). Relationship between viscosity and fat content of milk and cream. Journal of Texture Studies, 11(1), 65–68. https://doi.org/10.1111/j.1745-4603.1980.tb00308.x Wang, Q. (Janice), & Spence, C. (2017). The Role of Pitch and Tempo in Sound-Temperature Crossmodal Correspondences. Multisensory Research, 30(3–5), 307–320. https://doi.org/10.1163/22134808-00002564 Yeomans, M. R., Re, R., Wickham, M., Lundholm, H., & Chambers, L. (2016). Beyond expectations: the physiological basis of sensory enhancement of satiety. International Journal of Obesity, 40(11), 1693–1698. https://doi.org/10.1038/ijo.2016.112 Zampini, M., & Spence, C. (2004). The role of auditory cues in modulating the perceived crispness and staleness of potato chips. Journal of Sensory Studies, 19(5), 347–363. https://doi.org/10.1111/j.1745-459x.2004.080403.x Zampini, M., & Spence, C. (2005). Modifying the multisensory perception of a carbonated beverage using auditory cues. Food Quality and Preference, 16(7), 632–641. https://doi.org/10.1016/J.FOODQUAL.2004.11.004 Table 1. Estimation of expected ingestive outcomes within and between thin and thick milk samples.

Thin ΔV Thick ΔV Average Thin vs Thick ΔV Auditory Low High T value Low High T value Thin Thick T value Satiety 44.08 (19.48) 42.42 (17.04) 0.47 54.48 (19.89) 61.98 (18.93) 9.54** 43.25 (18.22) 58.23 (19.68) 76.12*** Satiation 43.27 (17.5) 40.1 (15.28) 1.85 52.31 (19.09) 58.31 (18.68) 6.65* 41.69 (16.42) 55.31 (19.03) 68.58*** Calories 0.37 3.12 157.29 97.71*** 154.9 (74.52) 159.69 (90.73) 205.31 (119.51) 219.17 (117.98) (82.62) 212.24 (118.33)

Oral Tactile Satiety 33.79 (13.23) 37.21 (15.59) 2.19 44.98 (18.95) 42.77 (15.43) 0.92 35.5 (14.49) 43.88 (17.22) 26.32*** Satiation 35.63 (13.12) 35.44 (14.05) 0.01 46.08 (17.04) 41.98 (13.75) 3.96 35.53 (13.52) 44.03 (15.54) 33.94*** Calories 0.34 1.13 140.06 30.71*** 141.9 (79.86) 138.23 (78.34) 168.02 (83.7) 161.35 (80.04) (78.71) 164.69 (81.53)

Vision Satiety 44.71 (17.98) 41.92 (16.51) 1.03 47.15 (16.26) 49.08 (15.7) 0.50 43.31 (17.22) 48.11 (15.93) 6.12* Satiation 43.23 (18.04) 39.98 (14.34) 2.26 43.38 (14.08) 46.6 (15.97) 2.23 41.6 (16.29) 44.99 (15.06) 4.91 Calories 2.61 0.91 151.57 2.20 157.15 (86.45) 146 (79.57) 155.52 (77.41) 162.13 (73.54) (82.83) 158.82 (75.18)

At Bonferroni adjusted values * p < 0.05, ** p < 0.01, *** p < 0.001. Figure 1. Fitted response observed between the proportion of correct responses and the change in viscosity (ΔV) of milk compared against the reference in 2-AFC paired tests.

Figure 2. Confidence in Estimating Calories Across Different Sensory Modalities.