Does tea smell different after the first sip? In a 2018 study, participants drank jasmine-flavored oolong tea in ten successive sips. The first sip produced a lower peak aroma intensity than later sips, and it took longer for the aroma to reach its peak.
However, participants had not practiced evaluating the actual tea beforehand. Besides becoming familiar with drinking it, they may have been getting used to attending to the aroma after swallowing and reporting it with a device. The study explores a caveat when evaluating food from a single sip or mouthful.
Tracking aroma for 60 seconds after swallowing
The researchers wanted to examine how aroma traveling from the mouth to the nose changes over the brief experience of drinking a cup in several sips. They considered both the possibility that repeated drinking would weaken the sensation through adaptation and the possibility that familiarity with the aroma or evaluation task would make it easier to perceive.
The participants were 25 adults aged 20–54: 11 women and 14 men. They took ten 10 mL sips of a commercial jasmine-flavored oolong tea containing no sugar or milk, at a room temperature of approximately 24°C. Each time, they measured out their own serving from the same bottle with its branding visible.
Immediately after swallowing, they continuously rated aroma intensity for 60 seconds. Pulling a ring worn on a finger moved an on-screen indicator from 0, meaning no sensation, to 5, meaning very strong. A cross-sectional diagram of the head and an explanation directed their attention to aroma traveling from the mouth to the nose.
They checked how to operate the device beforehand, but there were no practice trials involving drinking and rating the tea. They cleaned their mouths only before the first trial and did not rinse between sips. After each 60-second evaluation, approximately 30 seconds elapsed before evaluation of the next sip began.
A lower peak and a slower rise on the first sip
The researchers calculated six measures for each trial: maximum intensity, time to maximum intensity, area under the curve, duration near the peak, rate of increase, and rate of decrease. Five participants were excluded because the necessary points on their curves could not be identified, leaving 20 participants for these comparisons.
Mean maximum intensity was 3.71 for the first sip and 4.13–4.32 for sips two through ten. The time to peak was approximately 14.9 seconds for the first sip and approximately 3.8–8.4 seconds for later sips. Compared with every subsequent trial, the first sip had a significantly lower maximum intensity and a significantly longer time to peak.
The first sip also had a lower area under the curve, which sums intensity over time, and a slower rate of increase. There were no significant trial differences in the duration near the peak or the rate of decrease. The differences were especially apparent as aroma intensity rose after swallowing.
Differences lingered for the first few seconds of the second sip
A separate analysis included all 25 participants and divided the 60-second curves into two-second intervals. The first sip was lower than later trials mainly from the beginning to around 15 seconds. The second sip also had some early intervals with lower ratings than subsequent trials.
Later trials did not show differences as large as those involving the first sip. The evaluation did not reveal a simple adaptation pattern in which aroma steadily weakened toward the tenth sip.
One passage in the original paper’s discussion describes the first trial as higher, contrary to the results. This article follows the direction consistently reported in the numerical table, results section, and conclusion: the first sip was lower and slower to reach its peak.
Modeling the decline in aroma mathematically
The researchers also modeled the decline in aroma using an exponential decay curve. Guided by the overall mean curve, they used a common fitting interval beginning around 11 seconds into each trial. The average correspondence with the measured curves was approximately 0.66 for the first sip and approximately 0.84–0.91 for later trials, with a significantly poorer fit for the first sip.
However, trial differences in the curve coefficient and the constant related to the rate of decline were not significant. The paper discusses a possible slightly faster decline for sips two through four, but this was a numerical tendency. It does not establish that the rate of aroma decay reliably changed as participants continued drinking.
The first sip may also have served as practice
The authors suggest that the first sip acted as a warm-up. Participants may initially have been unfamiliar with attending to aroma after swallowing rather than taste intensity, and with translating the perceived intensity into movements of the ring. Because early differences persisted into the second sip, the authors infer that a single sip may not have provided enough preparation for some participants.
Other possibilities include aroma compounds remaining in the mouth or nose, repeated exposure making the next aroma easier to anticipate, and changes in olfactory processing. However, neither residual aroma compounds nor brain activity were measured, and there were no control conditions that could separate these explanations.
Because participants rated aroma intensity alone, other sensations such as bitterness or astringency may also have entered their ratings. This was a study of 25 people and one tea with visible branding. It did not show that the tea became more enjoyable or better liked from the second sip onward.
Summary
When participants repeatedly drank the same jasmine-flavored oolong tea, the first sip produced a lower maximum and overall aroma intensity and took longer to reach its peak than later sips. Early differences remained on the second sip, while the decline in aroma showed no significant trial differences.
The authors emphasize familiarity with both drinking and evaluation. On the first sip, encountering the tea’s aroma and practicing the rating task may occur together. The study suggests that research on beverage aroma should not treat the first sip and later sips as equivalent conditions.
The study was funded mainly by Suntory Global Innovation Center, with additional support from JSPS KAKENHI grants 26245073 and 16K04418. At the time, two coauthors were affiliated with Suntory and two with NTT Data Institute of Management Consulting. Company-affiliated authors also contributed to the conception and design. The authors declared no conflicts of interest, and the acknowledgments state that the main funder did not control interpretation, writing, or publication.
References
Gotow N, Omata T, Uchida M, et al. Multi-Sip Time–Intensity Evaluation of Retronasal Aroma after Swallowing Oolong Tea Beverage. Foods. 2018;7(11):177. doi:10.3390/foods7110177
