SELECTION OF AN ASTRINGENCY REFERENCE STANDARD FOR THE SENSORY EVALUATION OF BLACK TEA

Astringent and bitter sensations are characteristic sensory qualities of black tea. Three different classes of potential astringent reference standards (two concentrations each of alum and tannic acid and three fruit juices) were evaluated in this study. The perceived astringency, bitterness and sourness of each were profiled using computerized time-intensity and compared with the astringent intensity of a standardized brew of black tea. The differences in temporal profiles of potential reference standards across taste attributes were evident and intensity ratings were found to be dependent upon the stimulus and its concentration. Both concentrations of tannic acid were evaluated as the highest in perceived bitterness. For the juices, a strong sour taste was perceived in addition to astringency. It was concluded that the best reference standard for the astringency of black tea is a solution of 0.7 g/L alum as it is low in perceived bitterness and sourness.     

Interactions of astringent substances

Two-component mixtures of astringent materials were rated forperceived intensity of astringent and taste attributes overtime. Components included alum (a complex salt), gallic acid(the monomeric component of hydrolyzable tannins), catechin(the monomeric component of condensed tannins) and citric acid.Mixtures of alum and gallic acid showed mixture suppression,in that the 50/50 mixture was less intense than either componentin astringency, drying, roughing and puckery/drawing sensations.Suppression was seen at concentration levels producing moderateto strong astringency but was absent or less pronounced at lowerconcentration levels. A similar pattern held for citric acid,although the suppressive effects were less pronounced. Catechinand gallic acid mixtures were additive. Sensory interactionsbetween astringent materials appears to depend on the substancesinvolved and their concentrations (or intensity levels).     

Astringency of Organic Acids is Related to pH

Astringency and sourness of lactic, acetic and citric acids,each adjusted to pH 3, 5 and 7, were evaluated in two experiments,one starting at equal concentrations in wt/vol before neutralizationand the second starting at equal molarity. Astringency and sournessdecreased with increasing pH. However, acids were differentiallysour at equal pH, consistent with previous findings. In contrast,the tactile attributes associated with astringency (drying,roughing of oral tissues and puckery/tightening sensations)were similar across acids; pH was the major influence on astringency.Strong dependence on pH suggests that astringency of these acidsis a direct result of their acidic properties, and not solelydue to the hydrogen bonding mechanisms previously suggestedas an explanation of astringency in tannin interactions withsalivary proteins. Chem. Senses 21: 397–403, 1996.     

EVALUATION OF BITTERNESS AND ASTRINGENCY OF (+)-CATECHIN AND (-)-EPICATECHIN IN RED WINE AND IN MODEL SOLUTION

Two stereoisomeric phenolic compounds, (+)-catechin and (-)-epicatechin, were rated for perceived intensity of oral astringency and bitterness by trained judges using the scalar method. Mouth drying and mouth roughening were also assessed, since they are often associated with astringency. Amounts of 375, 750 and 1500 mg/L of each compound were tasted in red wine, and in a model system, similar in composition to a dry table wine. Preliminary tests showed that these concentrations were above the threshold level but within the range found in wine. A control sample (model solution or wine without the addition of the above phenolic compounds) was also evaluated. The results showed that the two compounds were both bitter and astringent. The high (-)-epicatechin concentration was significantly more bitter and astringent than the equal concentration of (+)-catechin in the model solution. Mouth drying and roughening ratings showed a similar increasing pattern with the ratings of astringency particularly at the higher concentrations. However, these attributes were rated differently from astringency suggesting that although they contribute to astringency, they are not subsumed by it.     

Astringent Subqualities in Acids

Astringency, astringent subqualities (drying, roughing and puckering)and sourness were compared among six acids: hydrochloric, lactic,citric, acetic, fumaric and malic acids. The attribute profilesof organic acids were similar to each other but different fromhydrochloric acid, the only inorganic acid, which was the mostastringent and the least sour. In a second experiment, two inorganicacids (hydrochloric and phosphoric) and two organic acids (citricand malic) were tested at three concentration levels. At approximatelyequal levels of overall sensory impact, the inorganic acidswere alike in astringency and sourness, receiving higher ratingsfor roughing and drying, and lower ratings for sourness thanthe organic acids. Interactions with concentration (differencesin psychophysical functions) for the subquality of drying werenoted, in addition to the differences in the astringent subqualitiesof roughing and drying seen across acids in both experiments.The higher level of astringency for inorganic acids suggeststhat the current model for tannin binding to salivary proteinsas an explanation of astringency needs to be extended to includea direct pH-dependent effect. Chem. Senses 20: 593–600,1995     

Perceptual Properties of Benzoic Acid Derivatives

Sensory properties of equimolar concentrations of benzoic acidderivatives were examined using time-intensity procedures. Significantdifferences in maximum intensities (P < 0.005) were foundfor astringency, bitterness, prickling, sourness and sweetness.Although these compounds differed only in the number and positionof the hydroxy groups, they exhibited quite different profiles.Gentisic acid had the highest sourness and bitterness maximumintensity, salicylic and gentisic acids were highest in astringency,and m-hydroxybenzoic acid was the sweetest sample. Benzoic acidhad the highest intensity of prickling feeling which lasted20 s longer than salicylic acid and 40 s longer than the othersamples which elicited lowest intensity of prickling sensation.Chem. Senses 20: 393–400, 1995.     

THE TEMPORAL PERCEPTION OF MENTHOL1

Although menthol is a common ingredient in pharmaceutical and food products, its sensory properties have not been studied extensively. The objective of this study was to describe and compare the temporal properties of l- and d-menthol. The cooling, burning, and bitterness of two menthol isomers (l-, d-) each at 0.01, 0.02, 0.04, 0.08% (w/v) in aqueous solution were evaluated by 11 trained panelists using time-intensity methodology. The intensity of all three attributes were evaluated continuously from introduction of the sample into the mouth, through expectoration at 10 s, until the termination of the sensation. The l-menthol samples had a greater maximum intensity and longer total duration of cooling and burning sensations than the d-menthol samples. In addition, maximum intensity and total duration of cooling and burn increased with concentration. In contrast, the total duration of the burning sensation was only dependent upon concentration of the l-isomer. Increasing menthol concentration significantly increased maximum intensity and total duration of bitterness for both isomers.     

UNDERSTANDING MOUTHFEEL ATTRIBUTES: A MULTIDIMENSIONAL SCALING APPROACH

Multidimensional scaling (MDS) was used to study qualitative relationships among mouthfeel attributes encountered in oral healthcare products. Similarity estimates were obtained from a rapid sorting task and from pairwise similarity ratings. Configurations were interpreted as suggesting four groupings of oral sensations: numbing, astringency, pain and taste. The pain-associated sensations were further differentiated into thermally related sensations and chemically related sensations in some configurations. Two-dimensional solutions from the sorting task and from group-averaged similarity ratings were similar. Individual differences scaling solutions, however, showed unacceptably high stress in two dimensions, suggesting additional nuances in meaning to individual panelists that were not captured by group-averaged data or by sorting data.     

Psychophysical evidence that oral astringency is a tactile sensation

Aluminum potassium sulphate [AIK(SO₄)₂ 12H₂O—‘alum’]was tested for its ability to elicit oral astringency independentlyof its ability to elicit taste. First, the astringency producedby alum (10 g/l; 21.1 mM) on the non-gustatory surfaces betweenthe gum and the upper lip was measured. Subjects reported thatalum elicited sensations of astringency when the upper lip wasmoved laterally against the gum. Second, subjects dipped thetongue into two solutions—alum or a mixture solution thatapproximated the taste of alum—and attempted to determinewhich solution was more astringent. A slight tendency was foundfor subjects to identify the mixture as more astringent thanthe alum. However, the same subjects reported the alum to bemore astringent than the mixture when the same solutions wereapplied under the upper lip. These two experiments support thehypothesis that tactile stimulation is important for producingastringency, whereas taste stimulation of the anterior tongueis not. Third, after the application of alum, lubricating rinses(water, sucrose, Salivart®, corn oil, and the subjects'own saliva) were compared for their ability to decrease astringency.The lubricants reliably decreased original astringency, butto varying degrees depending upon their lubricating properties.All three experiments suggest that tactile sensations causedby increased friction (decreases,in salivary lubrication) betweenoral membranes are the primary basis of astringent sensations.     

Turbidity as a measure of salivary protein reactions with astringent substances

Binding of tannins to proline-rich proteins has been proposed as an initial step in the development of astringent sensations. In beer and fruit juices, formation of tannin-protein complexes leads to the well-known effect of haze development or turbidity. Two experiments examined the development of turbidity in human saliva when mixed with tannins as a potential in vitro correlate of astringent sensations. In the first study, haze was measured in filtered human saliva mixed with a range of tannic acid concentrations known to produce supra-threshold psychophysical responses. The second study examined relationships among individual differences in haze development and the magnitude of astringency ratings. Mostly negative correlations were found, consistent with the notion that high levels of salivary proteins protect oral tissues from the drying effects of tannic acid.     

THE INFLUENCE OF BITE SIZE AND MULTIPLE BITES ON ORAL TEXTURE SENSATIONS

The influence of bite size on sensory mouthfeel and afterfeel sensations was explored in two studies in which single bites of vanilla custard desserts were varied from itollmL (study 1) and in which series of five bites of two different custard desserts were presented consecutively (study 2). In single bites, sensations of perceived creaminess were enhanced with size whereas sensations of astringency and temperature were suppressed. Bite size also affected perceived thickness, but the direction of the effects varied with custard desserts. With multiple bites, creaminess sensations continued to increase, whereas other sensations were unaffected. Switching to a second series of bites of another vanilla custard dessert showed larger sensory effects, especially on astringency sensations but also to a lesser extent on sensations of thickness and fatty afterfeel. Single and multiple bite results are discussed in terms of possible peripheral mechanisms. Switching results are discussed in terms of possible central sensory contrast mechanisms.     

TASTE COMPONENTS OF CHEDDAR CHEESE: FRACTIONATION AND OPTIMIZATION OF CHEDDAR CHEESE TASTE IN WATER

To determine the taste components of Cheddar cheese, we fractionated one mild Cheddar cheese and one aged Cheddar cheese by water extraction, freeze‐drying and gel filtration. Salty, sour and umami were the three predominant tastes present in the fractions. Neither trigeminal sensations nor astringency was perceived. We used response surface methodology to reconstruct a mild Cheddar cheese taste and an aged Cheddar cheese taste in water. Less salt and less acid were needed to simulate the taste of mild Cheddar cheese, compared with aged Cheddar cheese. Our optimized water solutions (containing sodium chloride, lactic acid and monosodium glutamate) were as similar to the real cheese samples as were the water extracts of the standard cheeses. However, neither our optimized solutions nor the water extracts matched the taste of the actual cheeses.     

INTERACTION OF ASTRINGENCY AND TASTE CHARACTERISTICS1

The perception of astringency and basic taste in mixtures and their interaction effects were investigated by two procedures. In Experiment 1, focused and nonfocused testing procedures were compared using mixtures of low and high concentrations of alum and basic taste solutions. Both procedures yielded taste and astringency intensities that were modality‐dependent. Nonfocused testing was used in Experiment 2 to investigate the interactions of astringent phenolic (tannic acid) and nonphenolic (alum) compounds with each basic taste. Sweetness of sucrose increased with increased concentration with or without alum or tannin present. Changes in salty, bitter, and sour taste intensities were modality‐dependent. Astringency either remained unchanged or decreased with the addition of sucrose, sodium chloride, citric acid, or caffeine depending upon the taste concentration. Bitterness of tannin and alum at high concentrations was suppressed by the addition of sucrose, sodium chloride, or citric acid; sourness also decreased in the presence of sucrose or sodium chloride as well as a high level of caffeine.     

EFFECTS OF DISCONFIRMED CONSUMER EXPECTATIONS ON FOOD ACCEPTABILITY

Three studies were conducted to assess the effects of disconfirmed consumer expectations on food acceptability. In the first, disconfirmed expectations for the sensory attributes of an edible film had a negative effect on acceptability of the film. Greater disconfirmation resulted in lower acceptance and purchase intent. In the second study, written product information was used to establish three levels of expected acceptability and expected bitterness for a novel fruit beverage. Comparison of preexposure (expected) and postexposure (perceived) ratings of acceptability and bitterness supported an assimilation model of disconfirmation effects for conditions in which expectations of acceptability were high and expectations of bitterness were low. A contrast effect was observed for bitterness judgments when expectations of bitterness were high. Associative effects resulting from the expectation manipulation were observed on other sensory attributes. In the third study, expectations were manipulated to influence both direction (positive versus negative) and degree of disconfirmation for the acceptance of cola beverages. Results provided further support for an assimilation model of these effects.     

Wine tannins,saliva proteins and membrane lipids

Polyphenols have been part of human culture for about 6000 years. However, their mode of action in relation to wine tasting while eating is only beginning to be understood. This review, using analytical techniques and physicochemical concepts, attempts to summarize current knowledge and present an integrated view of the complex relationship between tannins, salivary proteins, lipids in food and in oral membranes. The action of tannins on taste sensations and astringency depends on their colloidal state. Although taste sensations are most likely due to interactions with taste receptors, astringency results from strong binding to proline-rich salivary proteins that otherwise lubricate the palate. Tannins disorder non-keratinized mucosa in mouth, possibly perturbing taste receptor function. The 10–15% ethanol present in wines potentiates this action. Cholesterol present in large quantities in keratinized mucosa prevents any disordering action on these oral membranes. Polyphenols bind strongly to the lipid droplets of fatty foods, a situation that reduces the astringency perceived when drinking a tannic wine, the so-called “camembert effect”. Based on binding constants mainly measured by NMR, a comprehensive thermodynamic model of the interrelation between polyphenols, salivary proteins, lipids and taste receptors is presented.     

Effect of compound sequence on bitterness enhancement

The nature and occurrence of carry-over effects, i.e. the response to a stimulus is influenced by previous samples, were examined for selected bitter compounds. A time-intensity procedure was used to rate the bitterness of six compounds (caffeine, denatonium benzoate, limonin, naringin, quinine and sucrose octa-acetate). For each subject concentrations of these compounds were determined that were approximately equal in intensity to 1.18 x 10(-5) M limonin. To test carry-over effects of each compound the 36 paired sequences (pairs) were evaluated. Within a session three pairs were tested, between which two-stage rinses were used to remove any effects of the previous pairs. Within a pair only water rinses were used between stimuli. For all compounds carry-over or sensitization effects were observed in which values for maximum intensity, rate of onset and total area under the time-intensity curve were higher for a compound when tested in the second position than in the first. In addition, the degree of sensitization and susceptibility to sensitization were compound-specific. Caffeine increased the bitterness by the largest amount for all other compounds, while it was least affected. Regardless of the compound in the first position, the bitterness of quinine and denatonium were most enhanced.     

The influence of acid on astringency of alum and phenolic compounds

Astringency of aqueous solutions of phenolic compounds (grape seed tannins, tannic acid, catechin and gallic acid) increased upon addition of citric acid, whereas the astringency of alum was reduced. Astringency of alum was decreased equivalently by addition of equi-sour levels of lactic acid, citric acid or hydrochloric acid. The difference between alum and the phenolic compounds is speculated to result from chemical modifications affecting binding of the astringents with oral proteins rather than cognitive differences. Chelation of the aluminum ion in alum by acids reduces its availability for interacting with salivary proteins or epithelial proteins. In contrast, the increased astringency produced upon acidification of phenolic compounds is speculated to result from the pH driven increase in the affinity of the phenols for binding with proteins. These results suggest that alum cannot be used interchangeably with phenolic astringents in psychophysical studies.      

REGIONAL VARIATION IN SWEET SUPPRESSION

Differences in the sweet‐blocking efficacy of 2‐(4‐methoxyphenoxy) propanoic acid (PMP) for different sweeteners (sucrose and aspartame) and for various exposure areas of the mouth were found. Twenty participants rated sweetener solutions with and without PMP for sweetness, sourness, saltiness, bitterness and umami for stimulation of anterior tongue, posterior tongue and whole‐mouth areas. For sweetness ratings, suppression was significant for all stimulation areas. In the presence of PMP, stimulation of the posterior tongue yielded significantly higher sweetness ratings than stimulation of the anterior tongue for aspartame but not for sucrose. Sourness and bitterness ratings were significantly higher for anterior tongue than posterior tongue stimulations for aspartame but not for sucrose. The increases in sourness ratings in the presence of PMP were likely because of the sour taste PMP has at the concentration used. Results imply a difference between the front and the back of the tongue in the mechanisms involved in the perception of sweetness.