Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Mar 15.
Published in final edited form as: Physiol Behav. 2024 Feb 1;276:114480. doi: 10.1016/j.physbeh.2024.114480

The effect of multisensory context and experience on flavor preference decisions in rats

Alex Hua 1, Timothy V Dong 1, Joost X Maier 1
PMCID: PMC10922607  NIHMSID: NIHMS1964311  PMID: 38307360

Abstract

Flavor is perceived through multiple senses, including gustation and olfaction. Previous studies have shown that different sensory qualities that make up flavor are integrated to inform perceptual judgements. Psychophysical work in humans further suggests a prominent role for congruency (i.e., the learnt correspondence between taste and odor components of flavor through eating experience) in shaping multisensory interactions underlying perceptual judgments of flavor. However, eating experience cannot be controlled in humans, and depending on the type of judgement, these studies yielded mixed findings. Here, we used rats to test how experimentally-controlled experience with specific flavor mixtures (OdorA+TasteA and OdorB +TasteB) from weaning to adulthood affects subsequent flavor preference judgements in a series of two-bottle preference tests. In unisensory conditions, animals made odor or taste preference decisions (i.e., OdorA versus OdorB and TasteA versus TasteB, respectively). In multisensory conditions, animals made identical decisions, but the addition of the other modality rendered one solution congruent; the other one incongruent (e.g., OdorA+TasteA versus OdorB+TasteA). The results show that animals effectively learned congruency associations between the taste and smell components of experienced flavor mixtures. Comparing unisensory and multisensory conditions revealed no systematic effect of congruency on the magnitude of flavor preference, but preferences were less variable in multisensory compared to unisensory conditions. Results from a second group of naïve animals further demonstrate that increased reliability of preference judgements in multisensory conditions was independent of experience.

Keywords: taste, odor, retronasal, variance, reliability, congruency

Introduction

Most foods that constitute a typical diet consist of both taste and smell characteristics. During consumption, taste compounds stimulate the lingual epithelium and give rise to the perception of gustatory qualities (sweet, bitter, sour, salty and umami). At the same time, volatile compounds released in the oral cavity travel up the nasopharynx to stimulate the olfactory epithelium retronasally and give rise to the perception of a near-endless variety of qualities (Bushdid et al., 2014) that allow us to uniquely identify foods. Both gustatory and retronasal olfactory modalities contribute substantially to our sense of flavor (often referred to as “taste” in everyday language) (Murphy et al., 1977; Murphy and Cain, 1980; Rozin, 1982; Spence, 2016).

Previous work in humans demonstrated that both gustatory and retronasal olfactory signals contribute to flavor perception (Small and Green, 2012). Moreover, central integration of signals from both modalities is thought to enhance perceptual judgements. For example, detection of taste-smell mixtures is faster than can be predicted on the basis of independent processing of the two modalities (Veldhuizen et al., 2010a; Shepard et al., 2015). Some of these studies further suggest a role for experience in the ability to integrate gustatory and retronasal olfactory inputs (White and Prescott, 2007; Lim and Johnson, 2011, 2012). In particular, enhanced detection (Dalton et al., 2000; Shepard et al., 2015) and pleasantness (Schifferstein and Verlegh, 1996; Seo et al., 2013; Amsellem and Ohla, 2016; Fondberg et al., 2018) of multisensory mixtures has been shown to critically depend on “congruency” of taste and smell components of a mixture. Congruency is typically defined as learned associations between taste and odor components of multisensory flavor mixtures through consumption (Zellner et al., 1983; Stevenson et al., 1998; Yeomans et al., 2006; White and Prescott, 2007). However, work in humans precludes control over (or knowledge of) flavor experience, and congruency can at best be determined post-hoc, and is typically assumed based on arbitrary criteria (Murphy and Cain, 1980; Schifferstein and Verlegh, 1996; Dalton et al., 2000; Small et al., 2004; Nasri et al., 2011; Seo et al., 2013; Shepard et al., 2015). Indeed, findings regarding the effect of congruency on flavor judgements have been inconsistent. Intensity judgements generally do not appear to be affected by congruency (Murphy and Cain, 1980; Garcia-Medina, 1981; Hornung and Enns, 1986; Small et al., 2004; Lim et al., 2014), and whereas some studies have shown that the addition of a congruent odor enhances the pleasantness/unpleasantness of palatable/unpalatable tastes (Schifferstein and Verlegh, 1996; Seo et al., 2013; Amsellem and Ohla, 2016; Fondberg et al., 2018), others found no effect of congruency on pleasantness ratings (Small et al., 2004).

We previously used consumption tasks in rats to overcome some of the limitations of working with human subjects. Consumption is a behavioral read-out of hedonic flavor perception, and working with rats allows for experimentally manipulating flavor experience. Elliott & Maier (2020) showed that lifelong experience with specific taste-smell mixtures had no apparent effect on consumption of taste-smell mixtures during subsequent testing sessions. That is, the amount consumed from a unisensory taste or odor solution was not differentially affected by adding a congruent versus an incongruent component of the other modality. However, since this study used a one-bottle design, a direct comparison between congruent and incongruent flavors was lacking. Moreover, a second study (Maier and Elliott, 2020) found that gustatory and olfactory components of flavor stimuli were integrated to inform preference judgements in a manner that did not appear to depend on experience. Here, we exposed rats to specific taste-smell mixtures throughout development, thereby creating (in)congruent flavor combinations. Following training, we used a two-bottle design that allows us to directly contrast unisensory and multisensory preference decisions using congruent, incongruent and novel taste-smell mixtures. We assessed the effect of multisensory context on both the magnitude and the variability of flavor preference decisions. Based on work in humans, we hypothesized that multisensory congruency increases flavor liking. In the context of the present experiment, we thus predicted that adding a congruent compound of a different modality to a unisensory (taste or odor) flavor should increase the preference for that flavor. Based on our previous work in rats, we also considered the possibility that multisensory context affects flavor preferences in a manner that is independent of experience. Previous work suggests that access to gustatory and retronasal flavor qualities not only affects preference magnitude, but also increases the reliability (i.e., lowers the variability) of flavor preference decisions relative to unisensory stimuli (Maier and Elliott, 2020). In the context of the present experiment, we thus predicted that preferences in multisensory conditions were less variable than preferences in unisensory conditions.

Methods

Subjects

A total of 45 Long-Evans rats (20 males) were used for this study. Pregnant dams were obtained from www.criver.com and litters were kept together until weaning on postnatal day (PND) 21. All animals were kept on a 12 hour light cycle (6AM-6PM) and had ad lib access to food and fluids at all times during the experiment. All procedures took place in animals’ housing facilities. Animals were treated in accordance with the Guide for the Care and Use of Laboratory Animals, and all procedures were approved by the Institutional Animal Care and Use Committee of Wake Forest University School of Medicine.

Stimuli

Stimuli consisted of aqueous solutions of taste and/or smell compounds (obtained from www.fischersci.com and www.sigmaaldrich.com; >98% purity; dissolved in distilled water). Tastants: 10 mM saccharin (sweet), 58 mM sucrose (sweet), 50 mM sodium chloride (salty) and 100 mM monopotassium glutamate+0.17mM inosine monophosphate (umami). Odorants (both at 0.025% weight/volume): amyl acetate (CAS#: 628–63–7) and 2-hexanone (CAS#: 591–78–6). Since our lifelong exposure paradigm relies on animals’ willingness to consume the presented flavors, only palatable tastes were used.

Procedures

Two groups of animals were used in this study. One group underwent training followed by testing sessions; the other group underwent only testing sessions. Table 1 summarizes the groups and associated experimental phases. For animals in group 1, training occurred over a period of four weeks, from postnatal day (P) 21–23 to P47–49; testing started two days after the last training day, from P49–51 to P60–62. Animals in group 2 remained naïve until testing, from P56–67.

Table 1.

Overview of experimental timeline and procedures for group 1 and 2.

Training Testing
Block 1 Block 2
Group 1 (n=33)
# Days 24 6 6
Age ~22–48 ~50–55 ~56–61
Group 2 (n=12)
# Days - 6 6
Age - 56–61 62–67
Open in a new tab

Training

Each week for four weeks, animals were exposed to taste-smell mixtures for 6 consecutive training days. Each animal was exposed to two unique bimodal mixtures consisting of one tastant and one odorant (TA+OA, TB+OB). On each training day, animals received 20 ml of a single mixture for 16 hours (from 5–6PM to 9–10AM). Mixture identity alternated each training day (A-B-A-B-A-B), resulting in 3 exposures to each mixture per week (12 total over the course of training). During mixture exposure, animals were single-housed; in between exposures (from 9–10 AM to 5–6 PM), animals were pair-housed with a littermate and received ad lib water. Each block of six training days was followed by one day of ad lib access to plain water. For all animals, the taste component of one of the training mixtures was saccharin (per convention arbitrarily referred to as mixture A); depending on the animal, the other mixture (per convention referred to as mixture B) contained sodium chloride, sucrose or umami. Odor components used were identical for each animal (amyl acetate and 2-hexanone); the identity of the odor component paired with each taste was stable within each animal across training, but counterbalanced between animals. The training conditions are summarized in Table 2. To ensure equal exposure to mixtures A and B, the number of exposures to each mixture was identical for each animal, and we monitored consumption throughout the training period. Consumption did not differ between mixtures A and B (independent samples t-test comparing daily consumption of mixtures A and B: t789=0.96, p=0.34). Training resulted in the formation of two congruent flavors (i.e., TA+OA and TB+OB) and two incongruent ones (i.e., TA+OB, TB+OA).

Table 2.

Overview of training conditions and number of animals assigned in group 1.

n TasteA OdorA TasteB OdorB
6 Saccharin AA NaCl 2H
7 Saccharin 2H NaCl AA
5 Saccharin AA Sucrose 2H
5 Saccharin 2H Sucrose AA
5 Saccharin AA Umami 2H
5 Saccharin 2H Umami AA
Open in a new tab

Testing

Testing occurred over a period of 12 consecutive days. Each day, rats drank freely from two bottles containing taste, odor, or taste+odor mixture solutions for 16 hours (from 5–6PM to 9–10AM), effectively asking animals to make preference judgments in the following conditions: unisensory (OA vs. OB, TA vs. TB) or multisensory (OA+TA vs. OB+TA, OA+TB vs. OB+TB, TA+OA vs. TB+OA, TA+OB vs. TB+OB). Each bottle contained 200 ml of solution, ensuring animals had ad lib access to both solutions. Table 2 summarizes the testing conditions. In between tests, animals had ad lib access to water. Two blocks of testing were conducted. Within each block, each condition was presented once, in random order. To demonstrate that the testing procedure reliably measures preferences, we compared relative preferences in each condition between the two testing blocks. Preferences were significantly correlated between blocks (r=0.50, p<0.001), and preferences did not consistently differ between testing blocks (paired samples t-test comparing preferences in each animal and condition between the two blocks: t240=0.88; p=0.38), indicating stable preference throughout the testing phase.

Data analysis

Consumption was measured by weighing bottles before and after training/testing sessions (Consumption=WeightBefore–WeightAfter). Relative preference was calculated as ConsumptionBottleA/(ConsumptionBottleA+ConsumptionBottleB), where a score >0.5 indicates a preference for bottle A; a score <0.5 indicates a preference for bottle B. For a total of 33 bottles (3.1% of all bottles), consumption could not be determined due to spillage of displacement in the cage. Preferences in these instances were excluded from the analyses. Comparison of preference magnitude was based on the average preference over the two repetitions for each condition (i.e., blocks 1 and 2). Variance in preference was calculated across the two repetitions. Because preference and variance in preference were approximately normally distributed, we compared these two dependent variables between conditions defined by various independent variables using t-test and analysis of variance (ANOVA). Preferences were primarily compared between multisensory conditions and their corresponding unisensory conditions (factor Modality). We further broke down the data into cases where animals made odor decisions versus taste decisions (factor Decision), and into cases where different taste or odor compounds made up the congruent flavor (factors Taste and Odor Identity respectively). Finally, results were compared across groups that did or did not receive prior flavor experience (Factor Experience).

Results

We investigated the effect of multisensory context on the expression of flavor preferences in a two-bottle task. One group of rats (n=33; 15 males) was first exposed to specific taste-smell mixtures in a controlled manner from weaning to adulthood during a training phase (see Methods), and then subjected to a two-bottle preference testing phase. A second group of animals (n=12; 5 males) remained naïve until adulthood and were subjected to the same two-bottle preference tests without prior exposure. This design eliminates any possible confounding contribution of generalization between exposed and novel odor and taste compounds. The timeline of the experiments is illustrated in Table 1; Table 2 lists the two-bottle testing conditions.

Efficacy of congruency training on unisensory preferences

Extensive experience with a specific mixture of taste and odor components is thought to result in that mixture to be perceived as congruent. Thus, during training, each animal was exposed to two binary taste-smell mixtures. Although the use of animal models precludes direct assessment of congruency perception, our paradigm does allow us to test for the formation of associations between taste and odor components of mixtures. Based on previous work in rodents and humans showing that odors can acquire taste qualities when experienced in mixture (Holman, 1975; Stevenson et al., 1998; Stevenson et al., 2000a; Stevenson et al., 2000b; Yeomans et al., 2006; Blankenship et al., 2019; Christensen et al., 2022), we predicted that if animals formed associations between the taste and odor components of the mixtures they were exposed to during training, relative preferences for tastes and their associated odors (conditions #1 and #4, respectively) should be correlated. For example, if an animal was exposed to saccharin+amyl acetate (A) and sodium chloride+2-hexanone (B), the preference for saccharin over sodium chloride should be similar to the preference for amyl acetate over 2-hexanone. Indeed, preferences in the two conditions were significantly correlated (r=0.45, p<0.01; Figure 1), demonstrating that animals learned associations between multisensory components of the training mixtures. We did not observe significant differences within unisensory preferences. That is, taste preferences relative to saccharin did not differ significantly between sucrose, NaCl and umami (one-way ANOVA on preference with factor Taste: F2,32=2.94, p=0.07). As a consequence, odor preferences relative to the saccharin-associated odor did not differ between odors associated with sucrose, NaCl and umami (F2,32=2.14, p=0.14). We speculate this is due to the fact that we used moderate concentrations of tastes that are equally palatable. We also note that whereas taste and odor preferences are correlated, taste and odor stimuli are detected and processed by (at least partly) separate neural systems. Indeed, previous work has shown that naïve animals have different preferences for the monomolecular odorants used here (Blankenship et al., 2019; Maier et al., 2023). This offers a potential explanation for individual variability of differences in preference for tastes and their associated odors seen in Figure 1. In the following analyses, conditions in which animals made taste-based versus odor-based decisions are treated as different (factor Decision, see below).

Figure 1. Efficacy of congruency training.

Figure 1.

Open in a new tab

Each data point represents preference for taste stimuli (i.e., condition #4) plotted against preference for taste-associated odor stimuli (i.e., condition #1) for each animal (n=33). Preferences in the two conditions were significantly correlated (r=0.45, p<0.01).

The effect of congruency on preference magnitude

Testing conditions were designed to directly test the effect of congruency on flavor preference decisions. In condition #1, animals made a unisensory preference decision between two odor stimuli they were previously exposed to. In the multisensory conditions #2 and 3, animals were confronted with the same odor choice, but in the presence of a taste stimulus that rendered one solution congruent; the other incongruent. Thus, in addition to a choice between two odors, animals made—at the same time—a choice between a congruent and incongruent flavor. Note that because the same taste stimulus is added to both bottles in the multisensory conditions, taste information could not by itself affect preference. Therefore, any difference between unisensory (#1) and multisensory (#2–3) conditions would reflect an effect of taste congruency on odor preference. More specifically, based on our hypothesis that congruency increases liking, we predict that preferences for the congruent flavor in multisensory conditions should be greater than preferences for the odor component of the congruent flavor in the corresponding unisensory condition. Conditions #4–6 follow a similar design, but constitute taste-based choices (instead of odor-based choices). For all analyses and graphs, preferences in multisensory conditions were expressed as preference for the congruent flavor relative to the incongruent flavor; preferences in the corresponding unisensory conditions were expressed as preference for the odor/taste component of the congruent flavor, depending on whether animals made odor/taste discriminations, respectively. Figure 2A shows that preferences in multisensory and corresponding unisensory conditions were highly correlated (r=0.73, p<0.001), demonstrating that odor-based and taste-based choices were consistent in unisensory and multisensory conditions. Figure 2B shows the difference in preference between multisensory and corresponding unisensory conditions for all animals, broken down by factors Modality (i.e., odor-based versus taste-based choice) and Taste (i.e., identity of the congruent taste). Unisensory and multisensory preferences did not differ when combining both taste and odor-based decisions involving all stimulus identities, although there was a (non-significant) trend toward lower preferences in multisensory versus unisensory contexts (t127=1.9, p=0.06), indicating that incongruent flavors were slightly preferred over congruent ones. Post-hoc comparisons revealed that this was particularly apparent in cases where animals made odor-based choices with NaCl as the congruent taste (t12=4.04, p<0.01). This was likely due to a ceiling effect, since animals in this condition showed the highest preferences in the corresponding unisensory condition, and differences in preference between unisensory and multisensory conditions were overall negatively correlated with unisensory preference magnitude (r=−0.28, p<0.01). That is, when unisensory preferences were higher, corresponding multisensory preferences tended to be lower. Two-way ANOVA on the difference in preference between unisensory and multisensory conditions did not yield an effect of Decision (F1,127=0.63, p=0.43) or Taste Identity (F3,127=2.74, p=0.05; trending toward significance due to ceiling effect), nor an interaction effect (F3,127=1.84, p=0.14). We also directly compared preferences in cases where the congruent odor was amyl acetate versus 2-hexanone. No significant effects of Odor Identity (i.e., amyl acetate versus 2-hexanone in odor-based decisions or amyl acetate-associated tastes versus 2-hexanone-associated tastes in taste-based decisions) or Decision (i.e., odor-based versus taste-based decisions) were observed (main effect of Odor Identity: F1,127=2.87, p=0.09; main effect of Decision: F1,127=1.51, p=0.22; interaction: F1,127=0.61, p=0.43). Thus, flavor congruency did not affect the magnitude of odor or taste preferences in our two-bottle test.

Figure 2. The effect of congruency on magnitude of flavor preferences.

Figure 2.

Open in a new tab

A. Each datapoint represents preference in a multisensory condition (expressed as preference for the congruent relative to the incongruent flavor), plotted against preference in the corresponding unisensory condition (n=132). No significant difference between the two conditions was observed. B. Bars represent differences in preference between multisensory and corresponding unisensory conditions, averaged (±SEM) across all animals and conditions. Preferences are shown separately for odor-based decisions (dark gray) and taste-based decisions (light gray), and for cases where saccharin (n=33), sucrose (n=10), sodium-chloride (n=13) and umami (n=10) was the congruent taste. Two-way ANOVA revealed no significant effects of Modality or Taste.

The effect of congruency on variance in preference decisions

The results above suggest that congruency may not affect the magnitude of flavor preferences. However, it is possible that preference decisions are otherwise affected by multisensory context. Previous work suggests that multisensory flavor preferences are more reliable than preferences for their unisensory components. We tested the effect of multisensory context on the reliability of taste-based and odor-based choices in our paradigm by comparing preferences from the same animal and condition across the two testing blocks (see Table 1). Variance in multisensory judgements within the same condition has previously been used to assess judgment reliability (Ernst and Banks, 2002; Maier and Elliott, 2020). Variance in preference across repetitions was significantly lower in the multisensory conditions as compared to the corresponding unisensory conditions (t102=2.39, p<0.05; Figure 3, left). Two-way ANOVA on the difference in variance between multisensory and unisensory conditions did not yield significant effects (Decision: F1,102=0.03, p=0.87; Taste Identity: F3,102=0.28, p=0.84; interaction: F3,102=0.45, p=0.72). Thus, multisensory context increased the reliability of preferences for flavor solutions.

Figure 3. Preference decisions are less variable in a multisensory context and independent of experience.

Figure 3.

Open in a new tab

Bars represent within-condition variance, averaged (±SEM) across conditions and animals (n=132, 132, 48 and 48, respectively). Two-way ANOVA on variance with factors Condition (unisensory, multisensory) and Experience (experience, no experience) yielded a significant effect of Condition (F1,321=4.30, p<0.05), but no effect of experience (F1,321=2.82, p=0.09).

The effect of multisensory context on preference variance in naïve animals

Preference is a relative measure, and it is therefore unclear whether the decrease in variance reported above resulted from more reliable judgements of congruent solutions, incongruent solutions or both. To test for the contribution of congruency to the reliability of flavor preferences, we performed an identical series of preference tests in a second group of naïve animals that were not experimentally exposed to taste-smell mixtures (see Table 2). Each animal in this group was tested on the same odor-based (amyl acetate versus 2-hexanone) and taste-based (saccharin versus sodium chloride) choices. Since we did not observe an effect of Taste in the data from group 1 presented above, only two taste compounds were used in group 2. Since these animals were not exposed to flavors, they had no opportunity to form taste-odor associations. As was the case for the first group of animals, multisensory preferences were correlated to their corresponding unisensory preferences (r=0.50, p<0.001). Figure 3 (right) shows variances in unisensory and multisensory preference across testing block 1 and 2 for this group of animals. Two-way ANOVA on variance with factors Modality (unisensory, multisensory) and Experience (experience, no experience) yielded a significant effect of Modality (F1,321=4.30, p<0.05), but no effect of experience (F1,321=2.82, p=0.09) or an interaction effect (F1,321=0.13, p=0.71). Thus, multisensory context increases the reliability of flavor preference decisions independent of experience with specific taste-smell mixtures.

Discussion

In the present study, we experimentally manipulated life-long experience with specific taste-smell mixtures to create congruent and incongruent multisensory flavor associations. Subsequent taste and smell preference testing in unisensory and multisensory context showed that congruency did not affect the magnitude of preferences. However, multisensory context did increase the reliability of preferences in a way that was independent of experience with specific taste-smell mixtures.

Our results show that taste-smell congruency does not increase (or decrease) taste or smell preferences as measured in a two-bottle choice test. Previous work in humans has provided evidence that perceptual judgements of multisensory flavor stimuli are affected by congruency (Dalton et al., 2000; Shepard et al., 2015; Amsellem and Ohla, 2016; Fondberg et al., 2018). For example, congruency enhances the detectability of taste-smell mixtures relative to incongruent mixtures or their unisensory components (Shepard et al., 2015). However, preferences observed in the present study are unlikely to have been affected by detectability since all stimuli used were well-above threshold. Future work using near-threshold concentrations of taste and odor stimuli during testing will investigate how congruency training affects flavor detection. Congruent tastes (as compared to incongruent ones) have also been shown to increase pleasantness ratings of odor stimuli (Seo et al., 2013), and congruent odors increase the saltiness/sweetness ratings of salty/sweet taste stimuli (Schifferstein and Verlegh, 1996; Nasri et al., 2011). In our design, preferences for congruent flavors were measured relative to incongruent flavors, and congruency-enhanced pleasantness (or incongruency-suppressed pleasantness) would thus be expected to have increased preferences. These findings seem at odds with the findings from the present study, however, note that there is a significant portion of the literature on human flavor perception that does not support a role for congruency in perceptual flavor judgements (Murphy and Cain, 1980; Garcia-Medina, 1981; Hornung and Enns, 1986; Small et al., 2004; Labbe et al., 2007; Lim et al., 2014). Below, we will discuss several factors that could contribute to discrepancies within the human psychophysics literature, and between perceptual judgments in humans and rats. First, humans made pleasantness judgements following explicit experimental instruction. Such judgments are subject to interpretation, and only indirectly relate to preference and choice behavior. Second, congruency is poorly defined in the extant literature, and few studies explicitly measured perceived congruency of stimuli (Schifferstein and Verlegh, 1996; Stevenson et al., 1998; Amsellem and Ohla, 2016). Therefore, choice of congruent stimulus pairings is relatively arbitrary. For example, sucrose-lemon (Schifferstein and Verlegh, 1996; Shepard et al., 2015), citric acid-grapefruit (White and Prescott, 2007) and saccharin-almond (Dalton et al., 2000)—pairings that may not necessarily be thought of as congruent—have all been used as congruent mixtures. Perhaps the most intuitive objective definition of flavor congruency includes taste-smell mixtures that have been experienced through eating (Stevenson et al., 1995; Yeomans et al., 2006; White and Prescott, 2007). However, life-long, subjective experience with an enormous variety of flavor stimuli precludes precise knowledge of which taste-smell combinations are congruent for a given individual. Our rat model allowed us to greatly simplify eating experience, and experimentally manipulate congruency of arbitrary taste and smell components in a highly controlled manner. Even though animal models preclude assessment of subjectively perceived congruency, our behavioral data demonstrate that animals effectively learned associations between taste and smell components experienced during training, consistent with the definition of congruency resulting from eating experience. Finally, it is possible that the effects of congruency observed in human subjects are species-specific. For example, it is possible that humans have evolved hard-wired connections between certain taste and smell components. It may also be the case that the subjective experience of congruency is unique to humans and affects pleasantness judgements.

The present results are consistent with previous work in rats showing that taste-smell interactions in the context of preference judgements are independent of prior experience with specific taste-smell mixtures. Using an identical training procedure to the one used here followed by a series of one-bottle consumption tests, Elliott and Maier (2020) showed that the amount animals consumed from taste-smell mixture solutions constituted a weighted sum of the amount consumed from the unisensory component solutions. This multisensory operation did not depend on congruency. Consistent with this finding, Maier and Elliott (2020) measured preferences for taste, smell and mixture solutions in a series of two-bottle tests without experimentally manipulating prior exposure to specific taste-smell mixtures, and showed that mixture preferences could be explained by a weighted sum of unisensory component preferences. This pattern of results is consistent with the framework of maximum-likelihood integration (Ernst and Bülthoff, 2004). According to this framework, perceptual judgements of multisensory stimuli are the result of “weighting” the unisensory component judgements, as has been observed for judgements of multisensory stimuli in a variety of modalities and task contexts, including visual-haptic (Ernst and Banks, 2002), auditory-haptic (Bresciani and Ernst, 2007) and visual-vestibular (Butler et al., 2010). This operation is highly adaptive in that the weight placed on a given unisensory component increases with its reliability, resulting in multisensory judgements that are more reliable (i.e., less variable) than either unisensory component judgement in isolation. Our finding that flavor preference judgements are more reliable in a multisensory as compared to a unisensory context is consistent with the idea that taste and smell components of mixtures are integrated to inform preference judgements that are more reliable than the unisensory component judgements. Our findings further suggest that multisensory integration of flavor components is independent of experience with specific taste-smell mixtures. This independence would allow for the ability to evaluate any multisensory flavor stimulus, regardless of whether it has been experienced before. Future work that independently quantifies variance in taste, smell and mixture conditions will establish whether multisensory preferences can be accurately predicted using the maximum likelihood integration framework.

In summary, the present work provides insight into the multisensory computations underlying flavor preference judgements: the ability to integrate taste and smell components of flavor solutions results in more reliable preference judgements, and this ability does not depend on prior experience with these taste and smell components (not together in mixture nor in isolation). Our rat model allowed us to experimentally manipulate flavor experience in a controlled manner, overcoming some major challenges presented by work on the topic in human subjects. Rats also present an experimentally-tractable model system in which to study the circuit-level mechanisms underlying multisensory integration of flavor inputs. Previous work has established multisensory convergence of taste and smell inputs to several brain regions, including classical “association” areas such as the orbitofrontal and anterior cingulate cortex (Rolls et al., 1996; de Araujo et al., 2003; Small et al., 2004), as well as primary olfactory (piriform) (Maier et al., 2012; Maier et al., 2015; Avery et al., 2020; Maier et al., 2023) and gustatory cortices (Veldhuizen et al., 2010b; Vincis and Fontanini, 2016; Samuelsen and Fontanini, 2017). Recent work has made important progress in enhancing our understanding of how circuits in these regions integrate their multisensory inputs (Escanilla et al., 2015; Idris et al., 2023), and future work combining behavioral and circuit-level approaches will investigate how the neural computations performed in these circuits produce the adaptive advantages associated with multisensory integration of flavor.

Table 3.

Overview of testing conditions

# Modality Bottle 1 Odor Bottle 2 Odor Bottle 1&2 Taste
1 Unisensory OdorA OdorB No Taste
2 Multisensory OdorA OdorB TasteA
3 Multisensory OdorA OdorB TasteB
Bottle 1 Taste Bottle 2 Taste Bottle 1&2 Odor
4 Unisensory TasteA TasteB No Odor
5 Multisensory TasteA TasteB OdorA
6 Multisensory TasteA TasteB OdorB
Open in a new tab

Highlights.

  • Taste and smell aspects of flavor become congruent through experience

  • The effect of congruency on perceptual judgements is poorly understood

  • We used rats to control experience and directly measure preference judgements

  • Congruency did not affect the magnitude of flavor preferences

  • Multisensory context increased the reliability of preference judgments

Acknowledgements

Many thanks to Victoria Elliott and Chad Collins for help with data collection. This work was supported by NIH R01 DC016063 and DC020212 (to JXM).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Amsellem S, Ohla K (2016) Perceived Odor-Taste Congruence Influences Intensity and Pleasantness Differently. Chem Senses 41:677–684. [DOI] [PubMed] [Google Scholar]
  2. Avery JA, Liu AG, Ingeholm JE, Riddell CD, Gotts SJ, Martin A (2020) Taste quality representation in the human brain. J Neurosci 40:1042–1052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Blankenship ML, Grigorova M, Katz DB, Maier JX (2019) Retronasal odor perception requires taste cortex, but orthonasal does not. Curr Biol 29:62–69.e63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bresciani JP, Ernst MO (2007) Signal reliability modulates auditory-tactile integration for event counting. Neuroreport 18:1157–1161. [DOI] [PubMed] [Google Scholar]
  5. Bushdid C, Magnasco MO, Vosshall LB, Keller A (2014) Humans can discriminate more than 1 trillion olfactory stimuli. Science 343:1370–1372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Butler JS, Smith ST, Campos JL, Bülthoff HH (2010) Bayesian integration of visual and vestibular signals for heading. J Vis 10:23. [DOI] [PubMed] [Google Scholar]
  7. Christensen BA, Triplett CS, Maier JX (2022) Odor mixture perception during flavor consumption in rats. Behav Neurosci 136:300–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dalton P, Doolittle N, Nagata H, Breslin PA (2000) The merging of the senses: integration of subthreshold taste and smell. Nat Neurosci 3:431–432. [DOI] [PubMed] [Google Scholar]
  9. de Araujo IE, Rolls ET, Kringelbach ML, McGlone F, Phillips N (2003) Taste-olfactory convergence, and the representation of the pleasantness of flavour, in the human brain. Eur J Neurosci 18:2059–2068. [DOI] [PubMed] [Google Scholar]
  10. Elliott VE, Maier JX (2020) Multisensory interactions underlying flavor consumption in rats: the role of experience and unisensory component liking. Chem Senses 45:27–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ernst MO, Banks MS (2002) Humans integrate visual and haptic information in a statistically optimal fashion. Nature 415:429–433. [DOI] [PubMed] [Google Scholar]
  12. Ernst MO, Bülthoff HH (2004) Merging the senses into a robust percept. Trends Cogn Sci 8:162–169. [DOI] [PubMed] [Google Scholar]
  13. Escanilla OD, Victor JD, Di Lorenzo PM (2015) Odor-taste convergence in the nucleus of the solitary tract of the awake freely licking rat. J Neurosci 35:6284–6297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fondberg R, Lundström JN, Blöchl M, Olsson MJ, Seubert J (2018) Multisensory flavor perception: The relationship between congruency, pleasantness, and odor referral to the mouth. Appetite 125:244–252. [DOI] [PubMed] [Google Scholar]
  15. Garcia-Medina MR (1981) Flavor-odor taste interactions in solutions of acetic acid and coffee. Chemical Senses 6:13–22. [Google Scholar]
  16. Holman E (1975) Immediate and delayed reinforcers for flavor preferences in rats. Learn Motiv 6:91–100. [Google Scholar]
  17. Hornung DE, Enns MP (1986) The contributions of smell and taste to overall intensity: a model. Percept Psychophys 39:385–391. [DOI] [PubMed] [Google Scholar]
  18. Idris A, Christensen BA, Walker EM, Maier JX (2023) Multisensory integration of orally-sourced gustatory and olfactory inputs to the posterior piriform cortex in awake rats. J Physiol 601:151–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Labbe D, Rytz A, Morgenegg C, Ali S, Martin N (2007) Subthreshold olfactory stimulation can enhance sweetness. Chem Senses 32:205–214. [DOI] [PubMed] [Google Scholar]
  20. Lim J, Johnson MB (2011) Potential mechanisms of retronasal odor referral to the mouth. Chem Senses 36:283–289. [DOI] [PubMed] [Google Scholar]
  21. Lim J, Johnson MB (2012) The role of congruency in retronasal odor referral to the mouth. Chem Senses 37:515–522. [DOI] [PubMed] [Google Scholar]
  22. Lim J, Fujimaru T, Linscott TD (2014) The role of congruency in taste-odor interactions. Food Quality and Preference 34:5–13. [Google Scholar]
  23. Maier JX, Elliott VE (2020) Adaptive weighting of taste and odor cues during flavor choice. J Neurophysiol [DOI] [PMC free article] [PubMed]
  24. Maier JX, Wachowiak M, Katz DB (2012) Chemosensory convergence on primary olfactory cortex. Journal of Neuroscience 32:17037–17047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Maier JX, Idris A, Christensen BA (2023) Taste-Odor Association Learning Alters the Dynamics of Intraoral Odor Responses in the Posterior Piriform Cortex of Awake Rats. eNeuro 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Maier JX, Blankenship ML, Li JX, Katz DB (2015) A multisensory network for olfactory processing. Curr Biol 25:2642–2650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Murphy C, Cain WS (1980) Taste and olfaction: independence vs interaction. Physiol Behav 24:601–605. [DOI] [PubMed] [Google Scholar]
  28. Murphy C, Cain WS, Bartoshuk LM (1977) Mutual action of taste and olfaction. Sens Processes 1:204–211. [PubMed] [Google Scholar]
  29. Nasri N, Beno N, Septier C, Salles C, Thomas-Danguin T (2011) Cross-modal interactions between taste and smell: Odour-induced saltiness enhancement depends on salt level. Food Quality and Preference 22:678–682. [Google Scholar]
  30. Rolls ET, Critchley HD, Treves A (1996) Representation of olfactory information in the primate orbitofrontal cortex. J Neurophysiol 75:1982–1996. [DOI] [PubMed] [Google Scholar]
  31. Rozin P (1982) “Taste-smell confusions” and the duality of the olfactory sense. Percept Psychophys 31:397–401. [DOI] [PubMed] [Google Scholar]
  32. Samuelsen CL, Fontanini A (2017) Processing of intraoral olfactory and gustatory signals in the gustatory cortex of awake rats. J Neurosci 37:244–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Schifferstein HN, Verlegh PW (1996) The role of congruency and pleasantness in odor-induced taste enhancement. Acta Psychol (Amst) 94:87–105. [DOI] [PubMed] [Google Scholar]
  34. Seo HS, Iannilli E, Hummel C, Okazaki Y, Buschhuter D, Gerber J, Krammer GE, van Lengerich B, Hummel T (2013) A salty-congruent odor enhances saltiness: functional magnetic resonance imaging study. Hum Brain Mapp 34:62–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Shepard TG, Veldhuizen MG, Marks LE (2015) Response times to gustatory-olfactory flavor mixtures: role of congruence. Chem Senses 40:565–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Small DM, Green BG (2012) A proposed model of a flavor modality. In: The Neural Bases of Multisensory Processes (Murray MM, Wallace MT, eds). Boca Raton (FL). [Google Scholar]
  37. Small DM, Voss J, Mak YE, Simmons KB, Parrish T, Gitelman D (2004) Experience-dependent neural integration of taste and smell in the human brain. J Neurophysiol 92:1892–1903. [DOI] [PubMed] [Google Scholar]
  38. Spence C (2016) Oral referral: On the mislocalization of odours to the mouth. Food Quality and Preference 50:117–128. [Google Scholar]
  39. Stevenson R, Prescott J, Boakes R (1995) The acquisition of taste properties by odors. Learning and Motivation 26:433–455. [Google Scholar]
  40. Stevenson RJ, Boakes RA, Prescott J (1998) Changes in odor sweetness resulting from implicit learning of a simultaneous odor-sweetness association: an example of learned synesthesia. Learning and Motivation 29:113–132. [Google Scholar]
  41. Stevenson RJ, Boakes RA, Wilson JP (2000a) Resistance to extinction of conditioned odor perceptions: evaluative conditioning is not unique. J Exp Psychol Learn Mem Cogn 26:423–440. [DOI] [PubMed] [Google Scholar]
  42. Stevenson RJ, Boakes RA, Wilson JP (2000b) Counter-conditioning following human odor–taste and color–taste learning. Learning and Motivation 31:114–127. [Google Scholar]
  43. Veldhuizen MG, Shepard TG, Wang MF, Marks LE (2010a) Coactivation of gustatory and olfactory signals in flavor perception. Chem Senses 35:121–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Veldhuizen MG, Nachtigal D, Teulings L, Gitelman DR, Small DM (2010b) The insular taste cortex contributes to odor quality coding. Front Hum Neurosci 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vincis R, Fontanini A (2016) Associative learning changes cross-modal representations in the gustatory cortex. Elife 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. White TL, Prescott J (2007) Chemosensory cross-modal stroop effects: congruent odors facilitate taste identification. Chem Senses 32:337–341. [DOI] [PubMed] [Google Scholar]
  47. Yeomans MR, Mobini S, Elliman TD, Walker HC, Stevenson RJ (2006) Hedonic and sensory characteristics of odors conditioned by pairing with tastants in humans. J Exp Psychol Anim Behav Process 32:215–228. [DOI] [PubMed] [Google Scholar]
  48. Zellner DA, Rozin P, Aron M, Kulish C (1983) Conditioned enhancement of human’s liking for flavor by pairing with sweetness. Learning and Motivation 14:338–350. [Google Scholar]

RESOURCES