200
The Human Bitterness Detection Threshold of Iso-α-acids and
Tetrahydro-iso-α-acids in Lager Beer
Kathryn M. Kolpin and Thomas H. Shellhammer,
1
Department of Food Science, Oregon State University, Corvallis, OR
ABSTRACT
J. Am. Soc. Brew. Chem. 67(4):200-205, 2009
The human taste thresholds of iso-α-acids (iso) and tetrahydro-iso-α-acids
(tetra) were measured in an unhopped lager beer matrix using the ASTM
method 1432 protocol. For each compound, 14 volunteers participated in
3 training sessions and 6–13 testing sessions, during which each panelist
was presented with a series of six 3-alternative forced-choice tests. Each
series was presented in increasing concentration by a factor of 1.5 or 1.3
over the previous dosed sample for iso and tetra, respectively. The individ-
ual panelist threshold concentrations were determined as the point on a sig-
moidal curve fitted to their percent correct response data where the panelist
correctly chose the dosed sample 66% of the time. The group-wise detec-
tion threshold was determined as the point on a rank probability plot where
50% of the panelists correctly identified the dosed sample. The group-wise
detection thresholds and standard deviations of thresholds for iso and tetra
were 7.1 mg/L (4.5–11.2 mg/L) and 2.7 mg/L (0.7–10.0 mg/L), respectively.
Threshold data obtained in this experiment were also analyzed using ASTM
method 679. Group and individual threshold results were compared across
both methods and were not statistically different on all but one testing day.
Keywords: ASTM method 679, ASTM method 1432, Bitterness, Hops,
Thresholds
RESUMEN
El sabor humano umbrales de iso-α-ácidos (iso) y tetrahidro-iso-α-ácidos
(tetra) se midieron en una matriz de cerveza lager sin lúpulo utilizando el
método ASTM 1432 protocolo. Para cada uno compuesto, 14 voluntarios
participaron en 3 sesiones de formación y 6–13 sesiones de pruebas, du-
rante el cual cada miembro del grupo de expertos se presentó con una serie
de seis alternativas de 3-forzado-pruebas de selección. Cada serie se pre-
sentó en el aumento de la concentración por un factor de 1.5 o 1.3 en los
últimos dosificado muestra del iso y tetra, respectivamente. El individuo ex-
perto concentraciones umbral se determinará como el punto en una curva
sigmoidia provistos por ciento a sus datos de la respuesta correcta cuando
el panelista eligió correctamente dosificado la muestra el 66% del tiempo.
El grupo de sabios umbral de detección se determinará como el punto en
un rango de probabilidad de la parcela donde el 50% de los panelistas iden-
tificado correctamente dosificado la muestra. El grupo de sabios y los um-
brales de detección de desviaciones estándar de los umbrales de iso y tetra
fueron 7.1 mg/L (4.5–11.2 mg/L) y 2.7 mg/L (0.7–10.0 mg/L), respecti-
vamente. Umbral de los datos obtenidos en este experimento también se
analizaron mediante el método ASTM 679. Umbral de grupo e individua-
les a través de los resultados se compararon ambos métodos y no fueron
estadísticamente diferentes en todas las pruebas, pero un día.
Palabras claves: Amargura, Lúpulo, Método ASTM 679, Método ASTM
1432, Umbrales
Hops are used as the primary bittering agents in nearly all types
of beer. Learning human taste detection thresholds—the lowest con-
centration detectable by taste—helps scientists and brewers define
and understand the complex mixture of chemicals that combine to
create the flavor we recognize as beer. The current increase in the
use of hop extracts in beer and the trend toward reducing the over-
all bitterness of American light lagers has made it relevant to find
the lowest amounts of iso-α-acids (iso) and tetrahydro-iso-α-acids
(tetra) detectable by taste in a similar beer matrix. The importance
of this knowledge was recognized in the mid-1970s and 1980s by
M. C. Meilgaard (6,7), who set out to create a library of taste and
aroma thresholds for a myriad of compounds that can be detected
in beer. He determined the threshold of iso in beer with 0–30 mg
of isohumulone per L to be between 7 and 15 mg/L using the as-
cending method of limits (ASTM method 679), which was the ac-
cepted method for determining group thresholds at that time (2,
7,8). Around the time that Meilgaard was publishing results for
thresholds of various compounds, G. R. Morrison (9,10) published
two articles questioning the statistical validity of the ascending meth-
od of limits used in Meilgaard’s threshold experiments (7). His
main criticism was that the method of ascending limits did not re-
quire individual panelist repetition (2). Because individual thresholds
vary moment to moment, concerns were raised that group thresh-
olds were based on estimates of individual thresholds, and thus,
the group threshold was statistically only an estimate itself. Once
these statistical shortcomings were noted, a modified version of
the method of ascending limits (ASTM method 1432) was introduced
(3). This method required each panelist to participate in several
repetitions of an ascending series of triangle tests to obtain statis-
tically defined individual thresholds, resulting in group thresholds
with stronger statistical significance (3). Although the method used
to gather the original data has been relegated to a “rapid method”
known as ASTM 679, researchers and industry scientists still use
the original resultant value of 7–15 mg/L when discussing the thresh-
old value of iso.
In 2002 the best-estimate threshold of tetra in unhopped beer
was found by Weiss et al (11) to be 8.50 mg/L. The term “best-
estimate threshold” suggests the implementation of ASTM method
679, but no sensory method was cited or described in the article.
To date, human threshold measurements of iso and tetra in beer
using ASTM method 1432 have not been published. The primary
objective of our research was to provide updated individual thresh-
old and group values for both iso and tetra in a well-defined beer
matrix using current standard methods. The secondary objective was
to compare methods 1432 and 679 using a common set of data.
EXPERIMENTAL
Sample Preparation
An unhopped lager beer was the medium for the sensory experi-
ments. It was brewed in the Oregon State University pilot beer brew-
ery with 90% of extract from pale malt (Great Western Malting Co.)
and 10% of extract from rice syrup solids (California Natural Prod-
ucts). The original gravity of the beer was 11.0°P, and the final
gravity was 2.2°P following fermentation at 14.5°C for 2 weeks
with lager yeast (Wyeast strain 2007). The final ethanol concentra-
tion was determined to be 3.47% by weight (1), and the real extract
was calculated as 4.26°P (1).
The iso solution used in this experiment (John I. Haas Hops Com-
pany) was found to be 31.1% (wt/wt) iso using spectrophotometry
(5) and 29.7% (wt/wt) iso using HPLC under conditions outlined
in the ASBC standard method for analysis of isomerized hop ex-
tracts by HPLC (1). Furthermore, the solution was composed of
33% (wt/wt) isocohumulone. The solution was dosed into the un-
1
Corresponding author. E-mail: tom.shellhammer@oregonstate.edu
doi:10.1094 /ASBCJ-2009-0706-01
© 2009 American Society of Brewing Chemists, Inc.