Can Fruitiness Change Even When Aroma Compounds Stay the Same?

If the same aroma compounds are added in the same amounts, will a sip taste equally apple-like or pear-like? A liquid also contains alcohol, acids, and compounds contributing sweetness. Their combinations may change the impression of a mouthful.

In 2019, Andrew R. Poveromo and Helene Hopfer used white-wine-like model solutions with matching aroma mixtures to study how perceptions changed after a sample entered the mouth. They aimed to identify components involved in the time course of flavor, taste, and mouthfeel, and to capture differences that a single intensity rating might miss.

Twelve model wines with a shared aroma mixture

The team prepared 12 model solutions resembling white wine, not actual commercial wines. They varied ethanol at three levels, 10%, 12%, and 14%; glycerol at two levels, 5 and 15 g/L; and caffeic acid at two levels, 20 and 60 mg/L. This 3 × 2 × 2 design produced 12 samples.

Glycerol occurs in wine and contributes to sensations including sweetness. Caffeic acid was chosen as a representative wine phenolic compound. The aroma mixture, acidity, pH, and amount of a model protein were held constant. The aroma mixture included several esters and linalool.

The 12 assessors were aged 23–65, drank white wine at least two or three times a month, and had experience in sensory evaluation or wine research. Eight were women and four men. After training in the method and terminology, they evaluated each sample three times over six sessions.

Selecting everything perceived at each moment

The method was Temporal Check-All-That-Apply, or TCATA. The screen displayed eight attributes: apple, pear, and citrus flavors; sweet, sour, and bitter tastes; astringency; and warm/hot sensation, plus other. Participants selected everything they perceived at that moment and deselected attributes when they no longer perceived them. They were not limited to choosing the single most prominent sensation.

Participants took the sample into their mouths, started the timer, moved it around for 10 seconds, and spat it out. They continued selecting attributes, including lingering sensations, for up to three minutes. Analyses compared selected attributes in 30-second intervals.

These records did not directly score sensory intensity. They captured selection frequency and duration, so more frequent selection cannot simply be converted into a claim that a sensation was a certain number of times stronger.

Alcohol related to early fruitiness and longer-lasting bitterness

Of the components varied, ethanol affected temporal sensory differences most broadly. Higher concentrations tended to produce more frequent selection of bitterness and warm/hot sensations, whereas lower concentrations tended to produce more apple and citrus flavor and sweetness selections.

The timing of differences varied by attribute. Ethanol concentration significantly affected apple flavor and sweetness during the first 30 seconds. For citrus, effects occurred during the first 60 seconds and again at 120–149 seconds.

Ethanol’s effect on bitterness, meanwhile, continued through 119 seconds after the start. Warm/hot sensations were affected during most of the evaluation period. Even with a common aroma formulation, the underlying matrix affected early fruitiness differently from tastes and irritation lingering later.

Changing glycerol also changed pear flavor

Higher-glycerol samples elicited more selections of sweetness and pear flavor. For example, glycerol concentration significantly affected pear flavor during the 30–59-second interval. Sweetness was also affected by glycerol in that interval.

Higher glycerol tended to reduce selection of bitterness and astringency. Effects on sweetness appeared during the first 120 seconds, bitterness during the first 90 seconds, and astringency during the first 60 seconds. The ethanol–glycerol combination also affected bitterness later in the evaluation. Component effects cannot necessarily be understood by simply adding their individual effects.

Caffeic acid showed no significant individual effects on the evaluated attributes at the tested concentrations of 20 and 60 mg/L. The authors suggest that the small concentration difference may partly explain this. The finding does not establish that phenolic compounds have no sensory effects under all conditions.

Why fruitiness can change without adding more aroma

The authors emphasize interactions between sensations. For example, sweetness from glycerol may become associated with a pear-like fruity impression. Experience of eating sweet fruit may underlie the tendency to perceive fruit aroma and sweetness together.

The authors discuss mutual suppression between tastes as an explanation for less frequent bitterness selection when sweetness increases. They also consider taste–aroma interactions in the relationship between bitterness at higher ethanol levels and less frequent selection of apple flavor.

These interpretations are consistent with the results, but the study did not directly measure integration in the brain. Even when aroma-compound concentrations in the liquid are matched, changes in composition can affect their release from it. Sensory evaluation alone cannot fully separate physical and chemical changes in release from perceptual interactions.

Combining single ratings with time-based evaluation

Another implication concerns how evaluation methods are used. The authors compared these findings with an earlier study that rated the same model samples using single intensity scores. They note that some differences were detected by both approaches, while others only became visible when tracked over time.

For example, if differences in bitterness or hot sensations persist or emerge after an initial-impression rating has ended, the first score cannot fully capture the finish. Conversely, because TCATA records whether a sensation is perceived, it has limitations in expressing fine intensity differences. The authors propose combining the methods rather than replacing one with the other.

The study involved 12 trained assessors and model solutions with a limited set of components. Real wine is more complex, and the authors identify validation in real wines as a future task. These concentration comparisons cannot directly predict the flavor of particular brands or determine a preferred alcohol level.

Summary

Even with the aroma mixture held at the same concentration, varying ethanol and glycerol changed the time course of apple, pear, and other flavor selections. Fruitiness may emerge not only from the amount of aroma compounds but also from its relationship with accompanying tastes and mouthfeel.

Moreover, differences did not occur only at the start of a sip. Understanding how components relate to aroma, taste, and finish requires considering when sensations arise and how long they last, alongside their intensity at a given moment. This study brings that temporal perspective to wine evaluation.

References

Poveromo, A. R., & Hopfer, H. (2019). Temporal Check-All-That-Apply (TCATA) Reveals Matrix Interaction Effects on Flavor Perception in a Model Wine Matrix. Foods, 8(12), 641. DOI: 10.3390/foods8120641 · Full original article.

Original article: © 2019 the authors. Licensee MDPI, Basel, Switzerland. CC BY 4.0. The study received funding from the US Department of Agriculture. Hopfer disclosed receiving speaking or consulting fees from food companies; the authors state that these organizations and funders had no role in the study design, analysis, or publication decisions.