Individual differences in sensory threshold for aroma chemicals added to beer

Abstract

The flavor impression made by a given product on different people can be judged from the variation in flavor thresholds for its principal flavor constituents. Perception threshols were determined for 25 compounds. The compounds were purified to constant flavor and threshold. Twenty-two laboratories and a total of 400 tasters each examined from 1 to 18 of the compounds. These were added to bland beers, and Best-Estimate Thresholds (BET) were determined by ASTM Method E679, the Ascending Method of Limits. Several training sessions were given before any main test. Panel thresholds varied from a high of 17 g/liter for ethanol, to a low of 0.1 μg/liter for geosmin and 2-nonenal. The difference between panels as a rule was less than threefold and rarely exceeded fivefold. Those oftenquoted differences of 100-fold or more were rarely seen, and it is tempting to conclude that many of them are artifacts due to misflavoring by impurities or confusion on the part of untrained tasters. Individual differences in sensory threshold for a given compound can be described by the log standard deviation from the geometric mean of the panel. The log s.d. averaged 0.53 ± 0.15 (3.4-fold) over 60 panels. The log s.d. varied from the most uniform at 0.26 (1.8-fold) for SO 2 to the least uniform at 0.95 (8.9-fold) for geraniol. The variability was least for simple flavors such as sweet, salty, butyric and alcoholic, and highest for complex yet distinct flavors such as hop oil, geraniol, vanillin and 2-nonenal. Results pointed towards the existence of a bimodal distribution of thresholds for geraniol and vanillin. No taster was highly sensitive to more than a few compounds nor was any taster insensitive to more than half of the compounds.