https://orcid.org/0000-0002-6319-6894
Abstract
The underlying mechanisms of taste interactions in humans are not well understood, and three mechanisms have been proposed, namely a chemical interaction, a peripheral physiological, and a central mechanism. In the present study, it was investigated which of these mechanisms causes the suppression of sweetness by citric acid. This was investigated using a split-tongue gustometer that can stimulate the two sides of the tongue with different stimuli simultaneously, enabling a comparison of sucrose and citric acid presented either separately on each side of the tongue simultaneously or in a mixture on one side. Two studies were conducted using low (Study 1; n = 50) and high (Study 2: n = 59) concentrations of sucrose (2.5% (w/w) and 10% (w/w), respectively), and citric acid (0.14% (w/w) and 0.18% (w/w), respectively). In neither of the studies was there a significant difference in sweetness intensity ratings between the two conditions where sucrose and citric acid were presented either separately or in a mixture form. However, both showed significantly lower sweetness ratings than without citric acid indicating suppression of the sweetness of sucrose from citric acid. This provides strong evidence for a central mechanism for the suppression of the sweetness of sucrose by citric acid. This mechanism seems to be equal in high and low concentrations of both sucrose and citric acid.
1. Introduction
Globally, there is a growing health concern regarding excessive sugar consumption. This has led to major initiatives by the food and drink industry to reduce added sugars and reformulate foods and beverages containing high amounts of sugar (Hutchings et al. 2019; Wang et al. 2019; Mora and Dando 2021). Understanding the underlying mechanisms of sweetness interactions with other compounds in the food matrix will help guide the reformulation of sweet foods and beverages. Foods and beverages are complex matrices and seldom contain only one taste component. Thus, tastes interact with each other in almost all foods and drinks we consume. The basic tastes have been shown to interact by enhancing or suppressing the intensities of other tastants (Breslin 1996; Keast and Breslin 2002a; Breslin and Huang 2006; Wilkie and Capaldi Phillips 2014).
The underlying mechanisms of taste interactions in humans are not well understood (Savant and McDaniel 2004; Breza and Contreras 2012; Maier and Katz 2013). It is generally accepted in the chemosensory literature that binary taste interactions can occur through one of three mechanisms: namely a chemical interaction mechanism, a peripheral physiological mechanism, or a central mechanism (Lawless 1979; McBride and Finlay 1990; Keast and Breslin 2002a; Bennett et al. 2012; Reyes et al. 2019).
Chemical interactions are assumed to alter a tastant’s affinity or availability to the taste receptor, either by structural changes of the tastant or interactions such as encapsulation (Keast and Breslin 2002a; Challa et al. 2005; Bennett et al. 2012). Peripheral physiological interactions occur in the peripheral taste transduction system. It could be an interaction between tastant and taste receptors (Breslin and Beauchamp 1995; Keast and Breslin 2002a; Keast 2003; Greene et al. 2011; Bennett et al. 2012) or signaling between taste receptor cells (Keast and Breslin 2002a; Chaudhari and Roper 2010; Roper 2021). Central interactions occur in the brain and have been demonstrated to result in mixture suppression (Kroeze and Bartoshuk 1985; Keast and Breslin 2002a). It has been suggested to relate to overlapping neural substrates for different tastes (Maier and Katz 2013). Thus, if one taste drives a neuron more strongly than another, the response of that neuron to a mixture of these two tastants is very similar to the response to the most effective stimulus alone, inhibiting the normal response to the other mixture component.
Studies have been conducted to investigate the mechanisms of taste interactions in humans (Kroeze 1979a; Lawless 1979; Kroeze and Bartoshuk 1985; Keast et al. 2001; Keast and Breslin 2002b; Savant and McDaniel 2004; Bennett et al. 2012). Lawless (1979) investigated the mechanisms of bitterness and sweetness interactions with an ingenious method using a so-called tongue box (McBurney 1966; McBurney et al. 1973; Kroeze 1979b). In this method, the tongue box separates the tongue laterally, creating two separate stimulation areas. Consequently, the two sides of the tongue can be stimulated simultaneously with different stimuli. Based on these studies, Lawless (1979) suggested a predominantly central basis of the mechanism of bitterness suppression by sweetness and vice versa.
Taste interactions depend on tastant and concentration (Keast and Breslin 2002a; Wilkie and Capaldi Phillips 2014). Sweetness intensity has consistently been shown to be suppressed by sourness in aqueous solutions in many different concentrations of different tastants (Pangborn 1960; Kamen et al. 1961; Frank and Archambo 1986; McBride and Finlay 1990; Schifferstein and Frijters 1991; Junge et al. 2020). Only a few studies have investigated the underlying mechanisms of the suppression of sweetness by sourness in humans. In a sequence of studies, Schifferstein and colleagues investigated the overall taste intensity of sweet-sour taste interactions (Schifferstein and Frijters 1990, 1991; Schifferstein 1992). However, not addressing the issue of peripheral versus central directly, their results indicated a primarily central mechanism of interactions on sweetness by sourness. However, studies on animals and animal tissue have indicated both peripheral physiological (Liu et al. 2005; Roper 2014; Qin et al. 2022) and central mechanisms (Maier and Katz 2013) for sweetness suppression from citric acid.
In the present study, we deploy a split-tongue approach inspired by previous split-tongue studies (Kroeze 1979a, 1979b; Lawless 1979; Kroeze and Bartoshuk 1985). The technical gustometer setup was further inspired by Andersen et al. (2019). The split-tongue method is a study design where the participants’ tongues are laterally divided into two different stimulation areas. This enables us to exploit the phenomenon that the two sides of the tongue are functionally identical and thus that they have similar taste sensitivities (Kroeze 1979b; McMahon 2001). This makes it possible to use the tongue as a control for itself in different taste stimulation combinations. Thus, two aqueous solutions will be presented to the tongue simultaneously. The core principle is that tastants can be presented either in mixture (i.e. a mixture of sucrose and citric acid) or simultaneously to each side of the tongue but physically separated (sucrose on one side of the tongue and citric acid on the other side), removing the possibility for the tastants to interact with each other. The difference between these two scenarios represents the peripheral physiological interaction and/or the chemical interaction, as the tastants only have the possibility to interact in the mixture scenario. Therefore, if no differences are found between these two scenarios, this would indicate a central interaction mechanism.
This study aimed to investigate how the perception of the sweetness of sucrose is affected by citric acid, and more specifically, whether the taste interaction mechanism is related to peripheral physiological interactions or central interactions. We hypothesize that (i) citric acid will suppress the intensity of the sweetness of sucrose, (ii) this suppression occurs at both high and low concentrations, and (iii) the suppression of the sweetness of sucrose from citric acid will not differ between the in mixture and the on each side of the tongue scenarios indicating a central suppression mechanism.
2. Materials and Methods
This paper presents two studies conducted using the same apparatus and overall setup. The setup is illustrated in Fig. 1. In short, the gustometer pumped two taste solutions simultaneously onto the participants’ tongues where a mouthpiece divided the tongue into two stimulation areas keeping the two taste solutions separated throughout the stimulation.
Illustration of setup and apparatus in the studies. Panel a) illustrates the mouthpiece that sprays the taste stimuli onto the two sides of the tongue. Panel b) shows the pump system that generates the flow of the stimuli. Panel c) illustrates the 5 different conditions as they are presented to the participant’s tongue. The orange line separates the two different stimulation areas. Stimuli are 1 = WA|WA (Water|Water), 2 = WA|CA (Water|Citric Acid), 3 = WA|SU (Water|Sucrose), 4 = WA|SU+CA (Water|Sucrose + Citric Acid), and 5 = CA|SU (Citric acid|Sucrose), where | indicates the separation on the tongue. All stimulations are counterbalanced by mirrored repetitions. For further illustrations of the lateralized setup see Andersen et al. (2019).
2.1 Gustometer
The gustometer is a custom-made 5-pump gustometer using CETONI neMESYS mid-pressure pumps (Cetoni GmbH, Korbussen, Germany), with 50 mL glass syringes (Cetoni GmbH, Korbussen, Germany) built following the template for the gustometer described by Andersen et al. (2019). A mouthpiece that simultaneously separated the tongue into two stimulation areas and kept the mouth fixed was constructed. The tongue separator attached underneath the mouthpiece was approximately 6 mm wide and 90 mm long. It was applied to participants’ tongues and had a small extension that helped participants fixate the separator to the tongue by holding it down to the tongue with their teeth. Participants were instructed to make sure that the tongue separator always touched their tongue throughout the study section and only remove it to change mouthpiece after the trials. A picture of the setup can be found in Supplementary Fig. S1.
The pumping system and the mouthpiece were connected by PTFE tubing (Mikrolab Aarhus A/S, Hoejbjerg, Denmark). The gustometer’s pumps generated the pressure initiating the flow of water or taste stimulus from the mouthpiece to the tongue. A 3-way valve controlled the inlet of solutions to the syringes and outlet to one of two mouthpieces. The purpose of this was to be able to mirror the stimulation in all sessions so that each condition could be replicated on the opposite side of the tongue, thus balancing the study design. Tongue separators were 3D printed (Original Prusa Mini+, Prusa Research a.s., Czech Republic) in food contact material grade filament (white colorFabb_XT, colorFabb, Netherlands).
The gustometer sprayed one of the 5 conditions in Fig. 1 onto the participant’s tongue. All were combinations of pure water (bottled, Salling and, Salling Group, Denmark); sucrose (Merck KGaA, Darmstadt, Germany) in water solution; citric acid (Merck KGaA) in water solution; or a mixture of sucrose and citric acid in water solution. The five conditions were (the “|” indicate separation on the tongue) (i) Water | Water (WA|WA), (ii) Water | Citric acid (WA|CA), (iii) Water | Sucrose (WA|SU), (iv) Water | Sucrose + Citric acid (WA|SU+CA), and (v) Sucrose | Citric acid (SU|CA). The concentrations of tastants were different for the two studies (Tables 1 and 3). All rinsing and stimulations were conducted with a flow of 1 mL/s.
Stimulus concentrations for Study 1.
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 1 mL) |
|---|---|---|---|---|
| Citric acid | – | 1.40 | 998.6 | 1,000 |
| Sucrose | 25.0 | – | 975.0 | 1,000 |
| Citric acid + Sucrose | 25.0 | 1.40 | 973.6 | 1,000 |
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 1 mL) |
|---|---|---|---|---|
| Citric acid | – | 1.40 | 998.6 | 1,000 |
| Sucrose | 25.0 | – | 975.0 | 1,000 |
| Citric acid + Sucrose | 25.0 | 1.40 | 973.6 | 1,000 |
Stimulus concentrations for Study 1.
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 1 mL) |
|---|---|---|---|---|
| Citric acid | – | 1.40 | 998.6 | 1,000 |
| Sucrose | 25.0 | – | 975.0 | 1,000 |
| Citric acid + Sucrose | 25.0 | 1.40 | 973.6 | 1,000 |
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 1 mL) |
|---|---|---|---|---|
| Citric acid | – | 1.40 | 998.6 | 1,000 |
| Sucrose | 25.0 | – | 975.0 | 1,000 |
| Citric acid + Sucrose | 25.0 | 1.40 | 973.6 | 1,000 |
2.2 Dependent variable
Participants rated sweetness and sourness using a horizontal generalized labeled magnitude scale (gLMS) with ticks and semantic labels at the following points: no sensation (0) barely detectable (1.4), weak (6), moderate (17), strong (35), very strong (51), and strongest sensation of any kind (100), as proposed by Hayes et al. (2013). Participants completed a training session with training in the gLMS with remembered sensations and a gustometer training session to get accustomed to the gustometer.
In the interest of clarity and brevity, only sweetness results will be reported in this paper.
2.3 Study 1—low concentration study
2.3.1 Preparation of taste stimuli
Stimulus solutions were prepared in volumes of 1000 mL per condition. The concentrations can be seen in Table 1. After sucrose and/or citric acid were weighed and added into a 1 L blue-capped bottle, water was added up to 1,000 g in total. Solutions were kept refrigerated at 5 °C until the day of use where they were tempered at room temperature before being administered to subjects. The solutions were used within 3 d after preparation.
2.3.2 Participants
Participants were university students in Aarhus, Denmark, and were recruited through advertisements on social media. Prior to participation, participants filled out a questionnaire with basic demographic questions serving both as a screening tool to exclude non-eligible and to obtain demographic information about participants. Questions in the questionnaire included age, gender, diabetes status, and known taste and smell impairment from COVID-19.
Only subjects between 18 and 35 years of age with no known taste and smell impairment from COVID-19 and not suffering from diabetes were invited to participate in the study. Characteristics of participants can be seen in Table 2.
Characteristics of subjects in Study 1.
| Subjects | |
|---|---|
| N | 47 |
| Number of women (%) | 34 (72.3) |
| Mean of age ± SD (range) | 25.4 ± 3.3 (18–34 years) |
| Subjects | |
|---|---|
| N | 47 |
| Number of women (%) | 34 (72.3) |
| Mean of age ± SD (range) | 25.4 ± 3.3 (18–34 years) |
SD, standard deviation, N, number of subjects.
Characteristics of subjects in Study 1.
| Subjects | |
|---|---|
| N | 47 |
| Number of women (%) | 34 (72.3) |
| Mean of age ± SD (range) | 25.4 ± 3.3 (18–34 years) |
| Subjects | |
|---|---|
| N | 47 |
| Number of women (%) | 34 (72.3) |
| Mean of age ± SD (range) | 25.4 ± 3.3 (18–34 years) |
SD, standard deviation, N, number of subjects.
2.3.3 Study procedure
Prior to participation, participants received an oral introduction to the study, the gustometer, and the scale. Participants completed a 3-item training where they evaluated imagined sensations using the gLMS. After the gLMS training, participants completed a gustometer trial stimulation to familiarize themselves with the gustometer followed by the actual experiment.
The experimental procedure was as follows: 2 s water rinse, 5 s break, 1.5 s stimulus with either one of the 5 taste conditions shown in Fig. 1 and with the tastant concentrations shown in Table 1. After stimulation, participants rated sweetness and sourness using a computer in front of them without removing their tongue from the tongue separator on the mouthpiece. The 5-s break after the water rinse ensured that the rinse water had time to drip off before the stimulus to avoid mixing between the water rinse and stimulus. Pilot testing ensured that 2 s were sufficient to rinse the tongue showing no significant decrease in perceived sweetness over 10 trials containing sucrose in different presentations (left side, right side, or both). This procedure was repeated 5 times to present all 5 taste conditions. The 5 conditions were presented to the participants’ tongues in randomized order. After a short break, the procedure was repeated but mirroring the side of the different stimuli (if sucrose was at the left side of the tongue for a specific condition during the first trial, sucrose would be at the right side during the second trial of that condition). Thereby each condition was presented in two trials. Further, the order of the second trial for each condition was also randomized. After the study, participants received a small monetary reward. The study duration for each participant was between 20 and 30 min.
2.4 Study 2—high concentration study
2.4.1 Preparation of taste stimuli
Stimulus solutions were preparations of sucrose and/or citric acid, solubilized in water prepared in 1,000 g per stimulus in blue-capped bottles. Stimulus concentrations can be seen in Table 3. Solutions were kept refrigerated at 5 °C until the day of use where it was tempered at room temperature before being administered to subjects. Solutions were used within 3 dafter preparation.
Stimulus concentrations for Study 2.
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 0.1 g) |
|---|---|---|---|---|
| Citric acid | – | 1.80 | 998.2 | 1,000 |
| Sucrose | 100.0 | – | 900.0 | 1,000 |
| Citric acid + Sucrose | 100.0 | 1.80 | 898.2 | 1,000 |
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 0.1 g) |
|---|---|---|---|---|
| Citric acid | – | 1.80 | 998.2 | 1,000 |
| Sucrose | 100.0 | – | 900.0 | 1,000 |
| Citric acid + Sucrose | 100.0 | 1.80 | 898.2 | 1,000 |
Stimulus concentrations for Study 2.
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 0.1 g) |
|---|---|---|---|---|
| Citric acid | – | 1.80 | 998.2 | 1,000 |
| Sucrose | 100.0 | – | 900.0 | 1,000 |
| Citric acid + Sucrose | 100.0 | 1.80 | 898.2 | 1,000 |
| Stimulus | g sucrose (± 0.1 g) | g citric acid (± 0.01 g) | Water (± 0.1 g) | Total (± 0.1 g) |
|---|---|---|---|---|
| Citric acid | – | 1.80 | 998.2 | 1,000 |
| Sucrose | 100.0 | – | 900.0 | 1,000 |
| Citric acid + Sucrose | 100.0 | 1.80 | 898.2 | 1,000 |
A pilot study was conducted to determine the concentration of citric acid with a sourness perceived as the same intensity as the sweetness of 10% (w/w) sucrose. The study was conducted with a trained sensory panel (n = 11) and the intensity of both sweetness of sucrose and the sourness of citric acid was evaluated using gLMS.
2.4.2 Participants
Participants were university students in Aarhus, Denmark, and were recruited through a participant pool from the Cognition and Behavior Lab, Aarhus University. Only subjects between 18 and 35 years of age with no known taste and smell impairment from COVID-19, and not suffering from diabetes, were invited to participate in the study.
Participants filled in a questionnaire with basic demographic questions after the study. Questions included were age, gender, and nationality. Characteristics of participants can be seen in Table 4.
Characteristics of subjects in Study 2.
| Subjects | |
|---|---|
| N | 57 |
| Number of women (%) | 32 (56.1) |
| Mean of age ± SD (range) | 23.9 ± 3.0 (18-29 years) |
| Subjects | |
|---|---|
| N | 57 |
| Number of women (%) | 32 (56.1) |
| Mean of age ± SD (range) | 23.9 ± 3.0 (18-29 years) |
SD, standard deviation, N, number of subjects
Characteristics of subjects in Study 2.
| Subjects | |
|---|---|
| N | 57 |
| Number of women (%) | 32 (56.1) |
| Mean of age ± SD (range) | 23.9 ± 3.0 (18-29 years) |
| Subjects | |
|---|---|
| N | 57 |
| Number of women (%) | 32 (56.1) |
| Mean of age ± SD (range) | 23.9 ± 3.0 (18-29 years) |
SD, standard deviation, N, number of subjects
2.4.3 Study procedure
Prior to participating, participants received an oral introduction to the study, the gustometer, and the scale. Participants completed a 15-item training where they evaluated imagined or remembered sensations using the gLMS. After the gLMS training, participants completed a gustometer trial stimulation to familiarize themselves with the gustometer. After this, the actual experiment was initiated.
The study procedure was as follows: 1.5 s water rinse, 4 s break, 1.5 s stimulus with either of the 5 taste conditions shown in Fig. 1 with the tastant concentrations shown in Table 3. After stimulation, participants rated sweetness and sourness using a computer in front of them without removing their tongue from the tongue separator on the mouthpiece. The water rinse was reduced from 2 s to 1.5 s to ensure sufficient rinse water in the syringe for the full session. It was ensured that even 1.5 s was sufficient to rinse the tongue by evaluating pilot data for the effect of presentation order, where no significant effect was found. Further, the lower amount of rinse water allowed a shorter break to let rinse water disappear from the tongue thus lowering the break to 4 s. The 5 taste conditions were presented to the participants’ tongues in a randomized order in duplicates. After a short break, the procedure was repeated with the different stimuli being mirrored on the tongue sides. Thus, each condition was presented in quadruplicates. After the study, participants received a small monetary reward. The study duration for each participant was between 30 and 40 min.
2.5 Software
The questionnaire for demographic screening for Study 1 was conducted using Compusense Cloud (Compusense Inc., Guelph, Ontario, Canada). The data collection during the studies was conducted using PsychoPy (V2021.1.3) and the plugin for the CETONI pumping system (Andersen et al. 2019).
2.6 Ethics approval
Both studies were submitted to the Institutional Review Board at Aarhus University for ethical evaluation and obtained their approval (study 1: approval number 2020-63; study 2: approval number 2021-120). All participants gave their written consent prior to participation.
2.7 Data analysis
All data analysis was conducted in R (V4.1.2, R Core Team 2020) using RStudio (V2021.09.0). Data were handled using the Tidyverse package (Wickham et al. 2019) and plots were constructed using the packages ggplot2 (Wickham 2016) and ggstatsplot (Patil 2021). In the plots are included both F test for analysis of variance, Student’s t-test for pairwise comparison, as well as Bayesian statistics. All these tests were evaluating the model Sweetness ~ Condition, where Sweetness is the average of the 2 or 4 gLMS rating ratings from each participant, and Condition is the 5 conditions Water | Water (WA|WA), Water | Citric acid (WA|CA), Water | Sucrose (WA|SU), Water | Sucrose + Citric acid (WA|SU+CA), and Sucrose | Citric acid (SU|CA). Responses for each participant are averaged before plotting, thus each participant only contributes with one average value for each condition. Evaluations were done in duplicates (Study 1) and quadruplicates (Study 2), respectively.
Pairwise comparisons were conducted using Student’s t-test with Holm adjustments for multiple comparisons. Significant differences are indicated with asterisks. All significance tests were conducted using an alpha of 0.05. For Bayesian statistics, effect sizes were estimated using partial ω2, and the prior width for Bayes factor calculation was 0.707.
To ensure the reliability of the results and as a test of the robustness of estimation, besides the analyses described above, the results were analyzed using generalized linear mixed model and pairwise comparisons using Tukey’s Honest Significant Difference test. Model specifications and results from these can be found in the Supplementary Information. However, the results from these analyses showed similar results and were not in contrast to the performed analyses.
3. Results
3.1 Study 1—low concentration
The concentrations of both sucrose and citric acid in Study 1 were chosen to be low but perceivable and of similar intensities (sucrose was perceived as having the same intensity of sweetness as citric acid was perceived to be sour). This was found to be the case for these concentrations when tasted conventionally in a consumer study (Junge et al. 2020).
Results from Study 1 are shown in Fig. 2. Both the F test (P value = 6.62 × 10−17, F value = 35.72, DF (unadjusted) = 230, effect size (ω2) = 0.24), and the Bayesian Factor (BF = 44.1) show strong evidence for the alternative hypothesis that the 5 conditions are not the same. Investigating differences in means of the sweetness intensity ratings, the 5 conditions fall into 3 groups. The conditions without sucrose are the lowest and the WA|CA-condition was only slightly higher than the WA|WA-condition (P value = 0.03). The highest sweetness intensity rating was WA|SU, and this condition was significantly higher than all other conditions in this study. The conditions WA|SU + CA and SU|CA were not significantly different from each other in sweetness intensity ratings (P value = 0.84) but different from all other conditions. These two conditions were in composition and concentration the same, where the only difference was whether the tastants were physically mixed and presented on the same side of the tongue (WA|SU + CA) or physically separated as presented simultaneously but on each side of the participants’ tongues (SU|CA).
Sweetness intensity ratings from Study 1. Mean for each condition is shown as a red dot, and P values for pairwise differences are shown above the graphs. F test showed significant effects of condition on sweetness ratings (P value=6.62 × 10−17, F value= 35.72, DF (unadjusted) = 230, effect size (ω2)= 0.24, n = 47 participants); BF = 44.1. For pairwise comparisons, only significant pairwise differences are shown. WA|WA = Water | Water, WA|CA = Water | Citric acid, WA|SU = Water | Sucrose, SU|CA = Sucrose | Citric acid, WA|SU + CA = Water | Sucrose + Citric acid. Means are WA|WA = 4.94, WA|CA = 6.64, WA|SU = 18.88, SU|CA = 13.70, WA|SU+CA = 13.45. * indicate P < 0.001); WA|WA versus WA|SU (P< 0.001); WA|WA versus SU|CA (P < 0.001); WA|CA versus WA|SU + CA (P < 0.001); WA|CA versus WA|SU (P < 0.001); WA|SU versus WA|SU + CA (P = 0.002); WA|SU versus SU|CA (P = 0.006).
The lower sweetness intensity ratings of the two conditions containing both sucrose and citric acid (WA|SU + CA and SU|CA) compared to sucrose only condition (WA|SU) show a sweetness suppression from citric acid (P values 0.002 and 0.006, respectively). Citric acid suppressed the sweetness of sucrose both when presented in the mixture (WA|SU + CA) and when they were presented simultaneously but on separate sides of the tongue (SU|CA). This indicates that under these conditions, the sweetness from sucrose was suppressed by citric acid and that this suppression was most probably due to a central interaction mechanism. If the suppression was due to a peripheral interaction, the suppression effect would only occur in the WA|SU + CA-condition, where the tastants are physically mixed. However, here it occurs in both conditions, and it is, therefore, most probably due to central interactions.
3.2 Study 2—high concentration
The concentrations of both sucrose and citric acid were chosen to be comparable to those of commercial beverage products and are thus relatively high concentrations. The concentration of sucrose was established by a screening of the sugar content of commercially available sucrose-sweetened beverages and was determined to be 10% (w/w).
Besides the higher tastant concentration in Study 2 compared to Study 1, the number of trials for each condition presented to the participants was doubled as well as the gLMS training was increased.
Results from Study 2 are shown in Fig. 3. Similar to Study 1, both the F test (P value = 2.22 × 10−25, F value= 35.72, DF (unadjusted) = 280, effect size (ω2) = 0.32) and the Bayesian Factor (BF = 78.6) show strong evidence for the alternative hypothesis that the 5 conditions are not the same. Investigating differences in means of the sweetness intensity ratings, the conditions without sucrose were rated lowest in sweetness intensity, as in Study 1. However, WA|CA-condition was rated significantly higher in sweetness intensity than the WA|WA-condition (P value < 0.001). As in Study 1, WA|SU was rated highest in sweetness intensity. Study 2 also showed the sweetness intensity of sucrose to be suppressed by citric acid both presented in the mixture and presented on each side of the tongues, as both WA|SU + CA and SU|CA were significantly lower than WA|SU in sweetness intensity ratings (P value < 0.001 for both). As in Study 1, it was found that the mixed condition (WA|SU + CA) and the separate condition (SU|CA) were not significantly different from each other (P value = 0.63) but differed from all other conditions in sweetness intensity ratings. Study 2 likewise study 1 showed that citric acid suppressed the sweetness intensity of sucrose both when presented in mixture (WA|SU + CA), and when the tastants were presented simultaneously but at separate sides of the tongue (SU|CA).
Sweetness intensity rating from Study 2. Mean for each condition is shown as a red dot, and P values for pairwise differences are shown above the graphs. F test showed significant effects of condition on sweetness ratings (Pvalue = 2.22 × 10−25, F value= 35.72, DF (unadjusted) = 280, effect size (ω2) = 0.32, n = 57 participants); BF = 78.6. For pairwise comparisons, only significant pairwise differences are shown. WA|WA = Water | Water, WA|CA = Water | Citric acid, WA|SU = Water | Sucrose, SU|CA = Sucrose | Citric acid, WA|SU + CA = Water | Sucrose + Citric acid. Means are WA|WA = 2.64, WA|CA = 6.28, WA|SU = 19.93, SU|CA = 15.21, WA|SU + CA = 14.74. * indicate P < 0.001; WA|WA versus WA|SU + CA (P < 0.001); WA|WA versus WA|SU (P < 0.001); WA|WA versus SU|CA (P < 0.001); WA|CA versus WA|SU + CA (P < 0.001); WA|CA versus WA|SU (P < 0.001); WA|SU versus WA|SU + CA (P < 0.001); WA|SU versus SU|CA (P < 0.001).
Accordingly, these results indicate that similar to the low concentrations in study 1, the sweetness from sucrose was suppressed by citric acid at high concentrations and that this suppression is most probably due to a central interaction mechanism.
Another result from this study was the apparent sweetness of citric acid alone. WA|CA was rated significantly higher than WA|WA in sweetness intensity, indicating that participants perceived sweetness from this condition.
4. Discussion
Study 1 and Study 2 had slightly different protocols. Study 2 intended not only to replicate Study 1 with different tastant concentrations but also add further rigor in the form of providing the participants with more training in the use of gLMS and increasing the number of repetitions of each condition. Citric acid was rated as sweeter than water in both studies. This seems counter-intuitive; however, previous studies have sporadically shown increased sweetness of citric acid in the presence of no or low amounts of sweeteners (Fabian and Blum 1943; Prescott et al. 2001; Wilkie and Capaldi Phillips 2014). Another possible explanation could be a carry-over effect from sucrose even though this effect has been tried to be overcome by randomization (Lawless and Heymann 2010).
In both studies, a suppression of sweetness from citric acid was found. This is in accordance with most previous studies that have shown citric acid to suppress the sweetness of sucrose at different concentrations of both tastants (Pangborn 1960; Frank and Archambo 1986; McBride and Finlay 1989; Green et al. 2010; Junge et al. 2020). In both studies, the sweetness was suppressed by about a fourth, which is not uncommon in aqueous solutions (Junge et al. 2022).
The condition where sucrose and citric acid were mixed (WA|SU+CA) and the condition where the two tastants were presented at each side of the tongue (SU|CA) were not significantly different in either of the two studies. This indicates a primarily central mechanism for the sweetness suppression from citric acid. This is the case in both low (Study 1) and high (Study 2) concentrations of both tastants (sucrose and citric acid). It is to the authors’ best knowledge the first time it has been shown that sweetness suppression from citric acid, primarily occurs through a central mechanism in humans. Schifferstein and Frijters (1990) showed that sweetness suppression from citric acid depended on citric acid concentration and not on sucrose concentration. This is consistent with a central suppression mechanism, as a peripheral mechanism most likely would imply a dose-dependent response from both tastants, as seen in Qin et al. (2022).
In contrast, Liu et al. (2005) showed that the TRPM5, a receptor that is part of the taste transduction cascade for sweet, bitter, and umami tastes (Liman 2007), was blocked by extracellular acidity, providing a possible peripheral mechanism for the suppression of sweetness from citric acid. However, as indicated by Roper (2014), the existence of a peripheral mechanism does not rule out a central mechanism, as it could be speculated that the central mechanism overrides the peripheral mechanism. The results from the present study indicated that if it is the case, that both a central and a peripheral mechanism contribute to sweetness suppression from citric acid, the peripheral mechanism most likely does not provide suppression to an extent that is perceivable in the concentrations investigated here. Another objection that the effects are in fact evidence for a central mechanism is the lateralization of nerve fibers on the tongue. Although the chorda tympani nerve fibers generally convey ipsilateral taste sensation, (McManus et al. 2011), studies indicate some degree of peripheral overlap of the nerve fibers on the tip of the tongue (Berteretche et al. 2008; Rusu et al. 2008). This crossing might imply that some of the effects interpreted as central in this study could, in fact, be part of a peripheral mechanism through the nerve fiber crossings.
Previous studies using split-tongue studies have found evidence of different mechanisms for the suppression of different tastant pairs. Kroeze and Bartoshuk (1985) investigated the suppression of bitterness from quinine hydrochloride with sucrose and sodium chloride, respectively. They found that the bitterness suppression by the sweetness of sucrose seemed was from a central mechanism, whereas they found indications that the suppression of bitterness from sodium chloride originated from a peripheral mechanism. The results from Kroeze and Bartoshuk (1985) show that both peripheral and central suppression mechanisms are possible. However, in his study of bitterness and sweetness interactions, Lawless (1979) found evidence for a central mechanism for both the suppression of the bitterness of quinine sulfate by sweetness from sucrose and the suppression of sweetness of sucrose by the bitterness of quinine sulfate. The results from our study that sweetness seemingly is suppressed by citric acid in a central mechanism is in line with these findings, indicating that a central mechanism is not uncommon for sweetness suppression.
It is puzzling that comparing the sucrose-containing conditions in Study 1 with the same in Study 2, the 4-fold increase in tastant intensity between the two studies did not result in profound increases in intensity ratings. One contributor to this, given the concentration increase and the increased number of repetitions, could be a lowered taste intensity perception due to adaptation (Kroeze 1979a, Kroeze 1979b; Ganzevles and Kroeze 1987). Another possible reason for this could be that the concentration differences are not perceived as profound when stimulated only on the tongue as compared to the whole mouth. Meiselman (1971) showed that when stimulating the dorsal tongue area with a continuous flow, changes in concentration resulted in lower perceived differences than when evaluating the same concentrations using a whole mouth stimulation. The dorsal tongue stimulation in the study by Meiselman (1971) was similar to the stimulation in the present study. Thus, maybe the concentration differences are just not perceived as strongly as one would expect from a whole-mouth stimulation. Further, there are differences in the study procedures between the two studies that could contribute to the absence of profound differences. Among these factors are an increased amount of training and an increased number of trials per condition that might play a role in this. Especially the amount of training could influence this. The purpose of the gLMS is that it can encompass various types of sensations. While this enables comparisons between sensations of different natures, it also requires training to be used consistently (Green et al. 1996). Therefore, the participants in Study 2 who received more training might have been more familiarized with the scale and, therefore, used a smaller area as they were more aware of the span of the sensations the scale should encompass.
Given the difficulties of using the gLMS, and the general high within-subject variability in taste intensity ratings, the relatively low number of trials per subject (2 and 4, in each study, respectively) could have contributed with a high variability and increased risk of error.
5. Conclusion
The present split-tongue studies showed that citric acid suppressed sucrose and there were strong indications of a central mechanism for the suppression of the sweetness of sucrose by citric acid in humans independently of high or low concentrations of both sucrose and citric acid. However, a separate peripheral mechanism cannot be ruled out based on these studies. The findings strongly suggest a primarily central mechanism for the suppression of the sweetness intensity of sucrose by citric acid.
Supplementary material
Supplementary material can be found at http://www.chemse.oxfordjournals.org/
Funding
This project was funded by CiFOOD Aarhus University Center for Innovative Food Research, Sino-Danish Center for Education and Research, and the Graduate School of Technical Sciences, Aarhus University.
Conflict of interest statement
None declared.
Data availability
The datasets generated for this study are available upon request to the corresponding author.
References
Author notes
Present address
