Research note 006 — 2026-08-30 (expanded 2026-09-08)
IBU Does Not Measure Bitterness
IBU (International Bitterness Units), familiar from craft beer labels. The bigger the number, the more bitter — or so it is thought, but what this unit actually measures is the absorbance at a specific wavelength, not the bitterness your tongue actually feels. And that gap has grown to a size that cannot be ignored in beer as it is brewed today. That does not make IBU meaningless. For comparisons made under matched conditions, and above all for traditionally kettle-hopped beers, it remains a useful index; the gap widens when modern techniques such as dry hopping, or comparisons across differing alcohol levels and hop compositions, come into play. Hop-forward styles that put bitterness front and center are reported, by the industry association's own data, to occupy about 69% of single-can sales in US craft beer, so the gap between the number and the sensation is not something to take lightly.
The definition of IBU, and an overview of what it does and does not measure, is gathered in the topic hub Beer bitterness and IBU. This page digs into one part of it: the gap between the number and the sensation.
beer — reported consumption share: About 69% of single-can sales of US craft beer were hop-forward (strongly bitter) styles, broken down as imperial IPA 26%, IPA 22%, hazy IPA 16%, and imperial hazy IPA 5%. In 12-can packs, IPA was the leading style by volume at 21%.contextual
Scope: The US craft beer market, per the Brewers Association's own analysis (based on NIQ data), as of 2025
Note: A self-published analysis article (2026-03-05) by a staff economist of the industry association Brewers Association. Counter-evidence to the generalization that 'mass production strips out bitterness' — bitterness-forward IPA-family styles still sit at the center of the US craft beer market.
About 69% of single-can sales were hop-forward styles (imperial IPA 26%, IPA 22%, hazy IPA 16%, imperial hazy IPA 5%). In 12-can packs, IPA led by volume at 21%.
Limitations: The source is the Brewers Association's own analysis article and is not connected to the raw NIQ data itself. The definition of the IPA-family group (including imperial IPA and hazy IPA) follows the article's own classification.
Show dataingredient-beer —has_reported_consumption_share_of→ (the claim sentence above)
claim: clm-h10-brewers-assn-ipa-dominant-share / type: cultural / updated: 2026-09-02
What this page establishes
In dry-hopped beers, multiple surveys of commercial beer have reported that the IBU value rises while the concentration of iso-α-acids actually falls.
A study covering more than 120 beer brands has shown that the relationship between IBU and sensory-rated bitterness intensity is nonlinear.
The human group detection threshold for iso-α-acids was 7.1 mg/L in one study that added them to an unhopped lager (it is not a threshold for telling differences apart).
Which combination of compounds best explains the gap between IBU and perceived bitterness has not yet been settled.
What the number called IBU points to
IBU is positioned as a measure that quantifies the level of iso-α-acids into which hop-derived α-acids convert during brewing, resting on the understanding that this level is tied to the bitterness of beer. Even among iso-α-acids, however, the cis-configured isomer is more bitter than the trans-configured isomer, and a single absorbance number does not distinguish this difference in stereochemistry.
bitterness units (BU / IBU) is defined as International Bitterness Units (IBU) is positioned as a measure that quantifies the level of iso-α-acids into which hop-derived α-acids convert during brewing, and the source states that the level of iso-α-acids correlates closely with the bitterness of beer. Among iso-α-acids, the source states that the cis form is more bitter than the trans form.preliminary
Scope: Layer: literature. A statement given as background on beer bitterness in a 2025 review (a comprehensive review on mouthfeel).
Note: The bitterness difference between the cis/trans isomers and the correlation between IBU and iso-α-acids are statements the review makes citing Bossaert et al. 2022, Yan et al. 2024, and Luo et al. 2020 (none of the original papers has been read directly; this is a secondhand citation).
"The cis form of iso-alpha acids is more bitter than the trans form... it is worth mentioning the level of iso-alpha acids closely correlates with the bitterness of beer and is quantified using the International Bitterness Units (IBUs)."
Limitations: The original cited papers (Bossaert 2022, Yan 2024, Luo 2020) were not checked directly by this team. This relies on the review's summary description.
Iso-α-acids are not the only hop-derived bitter components either. A 1989 commentary article in the Journal of the Brewing Society of Japan cites a report that there are more than 90 hop-derived bitterness components, with isohumulone (the representative iso-α-acid) as the principal one. What IBU's absorbance mainly reflects is just one representative group within that much larger breakdown.
bitterness is documented in 『ビールの味を特徴づけているのは，ホップ由来の爽快な，ほろ苦い苦味であり，小若と古久保は，ホップ由来の苦味成分は90以上あり，主成分はイソフムロンであることを報告している。』 (in translation: What characterizes the taste of beer is the refreshing, mellow bitterness derived from hops, and Kowaka and Kokubo report that there are more than 90 hop-derived bitterness components, the principal one being isohumulone.)contextual
Scope: Ishibashi Tokuo, "On the Sensory Evaluation of Beer," Journal of the Brewing Society of Japan, vol. 84, no. 6 (1989), p. 362. A commissioned commentary article in the Journal of the Brewing Society of Japan (1989).
Note: A statement citing a report by Owaka and Kokubo (the original paper is only indicated by citation number 24 within this article, and that original paper itself has not been checked directly).
Verbatim: 『ビールの味を特徴づけているのは，ホップ由来の爽快な，ほろ苦い苦味であり，小若と古久保は，ホップ由来の苦味成分は90以上あり，主成分はイソフムロンであることを報告している。』 (in translation: What characterizes the taste of beer is the refreshing, mellow bitterness derived from hops, and Kowaka and Kokubo report that there are more than 90 hop-derived bitterness components, the principal one being isohumulone.)
Limitations: The original paper by Owaka and Kokubo (citation number 24 in the text) itself has not been checked directly; this is a secondhand citation via this article.
Source: ビールの官能評価について (石橋徳雄, 1989) · full text checked
Show datasensation-bitter —is_documented_in→ (the claim sentence above)
The dry-hopping reversal: IBU rises while the iso-α-acids fall
In dry hopping — steeping hops in beer after fermentation, a technique used for IPA aroma — measurements of commercial beer have observed a reversal in which IBU rises while the iso-α-acid concentration that matters actually falls. What was pushing IBU up was a different class of compounds dissolving out of oxidized hops: humulinones. The unit moves in the direction opposite to the compound it is supposed to measure — a textbook breakdown of a proxy measure.
Scope: Pilot-scale dry hopping of a hop-free ale (0-16 g/L, 0-72 h), and a survey of 15 Pacific Northwest commercial beer brands sampled before and after dry hopping
Note: The 2 halves move in opposite directions: the conventional IBU proxy rises while the compound class it is supposed to stand in for falls. Central counter-evidence against treating IBU as a reading of iso-α-acids.
A trained sensory panel quantified the increase in bitterness when a hop-free ale was dry-hopped at 0-16 g/L over 0-72 h, alongside IBU and hop acid/polyphenol measurements. The increase in beer bitterness (IBU) from dry hopping was attributed to the extraction of humulinones and, in some cases, polyphenols. A parallel survey of 15 commercial beer brands sampled before and after dry hopping found a decrease in total iso-α-acid concentration in the majority of samples tested, showing that the dry-hopping process can remove iso-α-acids from beer.
Limitations: Abstract-level extraction. The number of panelists, magnitude of IBU, and iso-α-acid concentration are not given in the abstract, nor are sensory results corresponding to the commercial beer survey reported. The abstract as a whole still regards iso-α-acids as the main contributor to beer bitterness; this is a limitation on the IBU proxy, not a replacement of iso-α-acids.
A review by an independent group (the Kraków University of Agriculture) synthesizes reports from multiple labs, stating, "A decrease in iso-α-acids, and a concomitant increase in α-acids, humulinones, and polyphenols, which are extracted from hops during the process, are well documented in the literature." It cites adsorption onto the hop material (citing demonstrations by Oladokun et al.) as the mechanism for the iso-α-acid decrease, names humulinones as "the main factors responsible for the increase in perceived bitterness in beers after the dry-hopping process," and gives their relative bitterness as 66% (±13%) of iso-α-acids. This shows that the two-directional movement reported by Parkin 2017 — IBU rising while iso-α-acids fall with dry hopping — is reproduced across the literature as a whole.
Limitations: A review, not new primary data. Many of the cited primary sources are from the Shellhammer/OSU lineage (the same lab as Parkin), so this does not substitute for a fully independent primary replication, but it does show convergence of reports across multiple countries and labs.
Context: Brands inferred to be dry-hopped (inferred from humulinone detection: O, Q, AA, T, V) showed high polyphenol and humulinone concentrations, e.g., TPC>250 mg/L and up to 3 mg/L, and were described sensorially as harsh and progressive in bitterness. It states that a recent Hopsteiner report suggested humulinones are approximately 65% as bitter as iso-α-acids (citing Steiner, 2015).
Limitations: Whether dry hopping was used is an inference not confirmed with the manufacturer. The primary source for the 65% relative bitterness of humulinones is an industry report (Hopsteiner, Steiner 2015), which is close to the "66% (±13%)" cited in the Klimczak 2023 review used elsewhere in this evidence, but both may trace back to the same primary industry report, so this cannot be called a fully independent replication (needs checking).
As an example of a primary study (Maye and Smith) summarized in the review, it is recorded that when Cascade hops were dry hopped at 1 lb./bbl. (equivalent to 386.54 g/hL) for 3 days at 16°C, iso-α-acid concentration fell from 51 mg/L to 32 mg/L, while alpha acid and humulinone concentrations both increased from below the detection limit to 13 mg/L.
Limitations: These are figures from a primary study (Maye and Smith) cited by the review; this entry has not checked the primary paper itself. This is a single example under one dosage, contact time, and temperature condition, and cannot be generalized.
Scope: Dry-hopped beer in a pilot-scale study, a survey of 15 commercial brands, and a hop-freshness trial across 3 varieties
While iso-α-acids were the main contributor to beer bitterness overall, humulinones (oxidized α-acids) and polyphenols were also potentially significant contributors to bitterness, especially in strongly dry-hopped beers, and the IBU increase from dry hopping was attributed to the extraction of humulinones.
Limitations: Abstract-level extraction. The phrase 'potentially significant' is the authors' own hedge and is preserved here as well. The abstract reports no omission/addition test with isolated compounds.
The introduction states, 'For a long time, dry hopping was not considered to contribute to bitterness, since no isomerisation of alpha acids occurs,' and states that humulinones are oxygenation products of α-acids, that their concentration increases during hop storage, and that they carry approximately 66% of the bitterness of iso-α-acids. The conclusion states, 'In dry hopped beers, the main source of bitterness is humulinones, which are increasing and, at a certain HSI, started to drop; this is due to the further oxidation of humulinones to oxidation products.'
Limitations: The figure '66% of the bitterness of iso-α-acids' and its historical framing are cited from prior literature within this paper's introduction, not measured in this study, so it is recorded as context and not promoted to a claim. The description 'main source of bitterness' is the authors' conclusion about dry-hop-only beers and is not generalized to kettle-hopped beers.
The review writes 「静置型(static)のドライホッピングでは、24時間で平均75%のフムリノンが13.3-15.0°Cの工程温度で抽出されるのに十分」 (in translation: "in static dry hopping, 24 hours is sufficient for an average of 75% of humulinones to be extracted at a process temperature of 13.3-15.0°C") (citing Lafontaine and Shellhammer).
Limitations: These are figures from a primary study (Lafontaine and Shellhammer) cited by the review; this entry has not checked the primary paper itself. They cannot be generalized to dynamic (agitated) dry hopping (the authors themselves specify static).
bitterness units (BU / IBU) does not reliably track Perceived bitterness intensity or iso-α-acid concentration in dry-hopped beersupported
Scope: Dry-hopped and highly hopped beers. Pilot-scale experiments and commercial-product surveys across 3 independent studies
In the majority of the commercial beers surveyed, dry hopping raised IBU while total iso-α-acid concentration fell. In other words, the measured change in IBU did not track the compound class that IBU is conventionally said to measure.
Limitations: Abstract-level extraction; the abstract does not give the magnitude of the IBU increase or iso-α-acid decrease, or how many of the 15 brands showed a decrease.
Setting the BU in Table 5 beside the bitterness intensity in Table 6 (10 assessors) in order of HSI 0.3-0.7: Celeia had BU 24, 28, 23, 22, 21 against intensity 2.9, 3.1, 2.2, 1.2, 1.2; Aurora had BU 22, 29, 39, 35, 34 against 2.4, 2.8, 4.0, 4.2, 4.3; Styrian Wolf had BU 21, 25, 45, 42, 35 against 2.2, 2.4, 3.3, 3.7, 4.2. They mostly moved in the same direction, but from HSI 0.5 to 0.7 there are cases where BU fell while intensity rose (the text states that this intensity difference for Aurora was not significant). The paper writes that BU correlated with the level of humulinones.
Limitations: The paper itself does not report a correlation between BU and intensity; this is a reading of the 2 tables side by side. 10 assessors. The paper states that all experiments were done in duplicate. The iso-α-acid concentrations in these beers are very low in absolute terms, consistent with bittering by dry hopping only, so they should not be compared directly with kettle-hopped beers.
Context: Intensely dry-hopped styles contained the highest hop phenol concentrations, up to 159 mg/L, matching or exceeding the humulinones and iso-α-acids in the same beer. The authors state that while humulinones are a known contributor to the soft bitterness of hazy styles, the role of hop phenols remains poorly understood, and conclude that hop phenols may be playing a larger role than previously thought in the bitterness and mouthfeel of modern dry-hopped beers.
Limitations: A composition study with no sensory testing (stated in the text); 42 commercial beers. The 159 mg/L is the sum of quantified compounds. Treated as compositional context, not direct evidence of a mismatch between perceived bitterness and IBU.
Parallel measurement of the BU method and the HPLC method (direct quantification of iso-α-acids) in 34 commercial lager brands found a tendency for BU to read higher than HPLC in 60% of samples (BU 8-36 mg/L vs. HPLC total 8-41 mg/L). Furthermore, CATA sensory evaluation of 10 brands with differing chemical profiles found that many of the 13 terms differed significantly among brands (p<0.05), with perceived "quality" differences splitting systematically by type of hop product (conventional hopping / dry hopping / tetrahydro products). The same dataset reports both the mismatch between BU and HPLC-measured compound levels and the differences in sensory descriptors among 10 brands of differing composition. It does not directly test how well BU alone explains the sensory results.
Limitations: The "overestimation" by BU was observed in only about 60% of samples; the rest were comparable or BU was in fact lower than HPLC (e.g., brand AA was BU 34 mg/L but HPLC 41 mg/L). The sensory side is qualitative CATA description (frequency data), not absolute intensity, and no correlation coefficient on an intensity scale is reported.
To begin with, the unit and perception were never linked by a straight line
The gap existed even before dry hopping. A study of strongly hopped commercial beers reports that the relationship between the bitterness unit and rated bitterness intensity was nonlinear (the abstract we could check does not show the shape of the curve). Separately, a study that added iso-α-acids to an unhopped lager and tested 14 people found a group detection threshold of 7.1 mg/L. That is a threshold for telling whether the compound is present at all; it cannot tell us how many IBU of difference people can distinguish between hopped beers.
beer shows a nonlinear relationship between: Bitterness Units (BU) and rated bitterness intensitysupported
Scope: More than 120 highly hopped commercial beer brands
Note: Model performance outside highly hopped beers has yet to be verified.
In strongly hopped beers, Bitterness Units (BU) predicted rated bitterness with a nonlinear response.
Limitations: Scope of application is strongly hopped commercial beers.
Limits: The relationship between Bitterness Units (BU) and rated intensity modeled in highly hopped beers is further constrained by the finding that dry hopping raises IBU through the extraction of humulinones and polyphenols while removing iso-α-acids in the majority of the commercial beers surveyed. Therefore, even at the same nominal IBU, the composition behind the IBU input to this kind of model differs between dry-hopped and kettle-hopped beers.
Limitations: This is a cross-study inference; the source itself neither tests nor comments on the Hahn 2018 model. Abstract-level extraction; the magnitude of IBU or concentrations was not obtained.
A review by an independent group states in writing, "It is important to note that bitterness is not a direct equivalent to IBU. In some cases, the value of IBU can increase disproportionately to the perceived bitterness level," making explicit the non-equivalence and non-proportionality between IBU and perceived bitterness. In the dry-hopping context, humulinones and polyphenols contaminate the IBU measurement, so a single linear correspondence from BU to perceived intensity does not hold — an overall judgment pointing the same direction as Hahn 2018's nonlinear finding.
Limitations: A review; it does not give a quantitative function (the shape of the nonlinearity) for the discrepancy. The description is mainly in the dry-hopped beer context.
Context: In 16 conventionally hopped commercial lager brands, the relationship between the rank-rating bitterness intensity from 14 trained assessors and the HPLC-quantified concentration of 6 bitter compounds (iso-α-acids and their derivatives) was well modeled by PLS regression, with a correlation of 0.92. This shows a strong correspondence between compound concentration and perceived bitterness in the concentration range typical of conventional lagers, and functions as a scope limitation showing that the Hahn 2018 nonlinearity emerges in the highly hopped range. It also reports that time-intensity profiles showed large qualitative and quantitative individual differences among assessors.
Limitations: Abstract checked only (via a single retrieval route: a publisher-deposited abstract on OpenAlex). The 0.92 correlation is the PLS model's goodness of fit, not a test of linearity itself. Individual compound concentrations, not BU (IBU), are used as the explanatory variable.
Limits: From an analysis of 34 commercial lager brands, 10 with differing compositions were selected for sensory evaluation. Beers high in polyphenols and hop acids were perceived as harsh and progressive in bitterness; conventionally hopped beers as sharp and instant; and those using tetrahydro-iso-α-acids as diminishing, rounded, and acidic. In the abstract, the authors conclude that the hopping strategy brewers adopt affects the nature, temporal profile, and intensity of bitterness perception, but the sensory test in the accepted manuscript recorded CATA selection frequencies and did not score intensity.
Limitations: Abstract checked only. A cross-study inference, not a direct test of the Hahn 2018 model. Individual intensity values and compound concentrations are not in the abstract.
The correlation coefficients between lager beer sensory bitterness intensity (rank rating) and the concentration of each acid were: cohumulone 0.54, humulone 0.74, adhumulone 0.67, tetrahydrocohumulone -0.23, tetrahydrohumulone -0.01, and tetrahydroadhumulone 0.07 — limited for any single acid, while the multivariate (PLS) model reached 0.92.
Limitations: Full text checked. See the paper for sample size and panel composition.
In regressions of Bitterness Units (BU) against the concentration of each bitter compound class, R² was 0.5911 for iso-α-acids alone, 0.7569 for α-acids, 0.8739 for α-acids and their conversion products, and 0.8904 for the total.
Limitations: The main text is behind a paywall and was not checked; only the Supporting Information was verified. This shows that BU is not a linear indicator of any single compound class, but this data does not address the relationship to sensory bitterness.
iso-α-acids (isohumulones) has a human detection threshold of Group-level detection threshold of 7.1 mg/L in a hop-free lager beer matrixsupported
Scope: ASTM method 1432 three-alternative forced-choice test in a hop-free lager beer matrix, 14 volunteers
In a hop-free lager beer matrix, the group-wise detection threshold for iso-α-acids was 7.1 mg/L, with 4.5-11.2 mg/L given alongside in parentheses. The group value was taken as the point on a rank-probability plot where 50% of panelists correctly identified the spiked sample, and the individual threshold was read from a fitted sigmoid curve at 66% correct.
Limitations: Abstract-level extraction. In the abstract's wording ('the group-wise detection thresholds and standard deviations of thresholds... were 7.1 mg/L (4.5-11.2 mg/L)'), whether the parenthetical is a standard-deviation-derived interval or the observed range across panelists is ambiguous, and it is recorded here verbatim without interpretation. The detection threshold is not a value for supra-threshold intensity, and the hop-free matrix is not a finished hopped beer. It is not a threshold for telling differences apart (a difference threshold), and it is not used for converting differences in IBU.
Behind this nonlinearity are circumstances that the amount of iso-α-acids alone cannot explain. A study examining more than 120 strongly hopped commercial beer brands proposed a separate way to predict sensory-rated bitterness intensity based on isohumulone (an iso-α-acid), humulinone, and ethanol concentrations. How it compared with prediction from BU is not stated in the abstract we could check.
Scope: The principal hop-derived bitter substance in beer. Its human detection threshold has been measured in a hop-free lager base
Note: An isomerization-product class within the hop bitter acid family (compound-hop-bitter-acids). / 2026-08-31: calibration-downgraded from established to supported because the supporting evidence has no full-text-checked source. May be re-upgraded once full text is checked.
A study that measured the human bitterness detection threshold of iso-α-acids itself, directly demonstrating its identity as a bitter stimulus (group mean 7.1 mg/L).
In sensory plus instrumental analysis of more than 120 commercial beer brands, sensory bitterness intensity showed a nonlinear relationship with BU, and a bitterness prediction based on isohumulone (iso-α-acid), humulinone, and ethanol concentrations was proposed. The contribution of each variable and the difference in performance from prediction by BU are not shown in the abstract we checked.
Limitations: Abstract checked only. This is a contribution estimate from a regression model, not a sensory test of isolated iso-α-acids. The brand set is skewed toward highly hopped beers.
The compositional ratio of hop acids to polyphenols — that is, differences in hopping strategy, in what blend the hops were added — was also linked to differences in how often descriptors of bitterness quality and temporal character were chosen, in an evaluation where 6 trained panelists picked from 13 terms for 10 brands (CATA). It was not an evaluation that scored intensity.
beer has its bitterness character influenced by Hop bitter acid and polyphenol profile, and hopping strategysupported
Scope: Commercial lager beers in the cited study
Note: Terms such as sharp, progressive, and rounded are descriptors from that study's own panel, not fixed categories applicable to every beer.
A trained panel of 6 evaluated 10 brands with CATA using 13 terms (3 replicates each), and differences in hop acid and polyphenol profiles were associated with differences in how often descriptors of bitterness quality and temporal character were chosen. Intensity was not scored (the end of the published abstract also mentions intensity, but this is not supported by what the accepted manuscript measured).
Limitations: The commercial product set and the panel's vocabulary may not generalize to every beer style.
Hop-derived components are not the only factor either. In a study where 10 panelists evaluated 9 commercial products differing in alcohol content (3 brands × 3 categories) by choosing the most dominant sensation over time (TDS), the time windows in which bitterness became dominant differed between product groups. However, this was neither a test that scored bitterness intensity nor one that kept the hop composition fixed and changed only the alcohol, so it cannot be said that alcohol content alone changes how bitterness plays out.
beer — perceived bitterness is affected by: Among commercial beer groups differing in alcohol category, the time windows in which bitterness became temporally dominant differed (the effect of alcohol alone was not shown)supported
Scope: 9 commercial products differing in alcohol content (3 brands × 3 categories); TDS (choosing the most dominant sensation at each moment) by a trained panel of 10
Note: This is not a mechanistic claim that alcohol directly enhances bitterness perception. It includes concurrent differences among the beer groups.
In TDS, bitterness reached significant dominance (above the chance selection rate) at 23-45 seconds for regular beer and 26-41 seconds for low-alcohol beer. Astringency did not reach significant dominance in regular beer, and reached it at 50-63 seconds in low-alcohol beer. The significance line is a comparison with the chance selection rate, not a test between alcohol groups.
Limitations: A comparison of separate commercial products that measured neither hop compounds nor IBU. It is not a test that scored bitterness intensity, and it includes product differences other than alcohol content. It does not show a mechanism.
β-acid conversion products blend into the same number too
IBU's absorbance lumps together various hop-derived compounds into one reflection. Beyond iso-α-acids, the group of compounds formed when hop β-acids convert under wort-boiling conditions also contributes to bitterness, with recognition thresholds ranging, depending on the compound, from 7.9 to 90.3 µmol/L. The lowest is cohulupone at 7.9 µmol/L, and the next lowest is hydroxytricyclocolupone at 14.7 µmol/L. The former is described (within the scope of the abstract) as a short-lasting bitterness resembling iso-α-acids, and the latter as a coarse bitterness that lingers long at the back of the tongue and in the throat.
Scope: Beta-acid transformation products from heat treatment of colupulone under wort-boiling conditions. Subsequently quantified in commercial beer
Note: Shows that hop beta acids, not only alpha-acid derivatives, contribute measurable bitterness, and that congeners with similar threshold ranges differ in the time course of their bitterness. The number of panelists and the threshold method are not stated in the abstract. External audit N2-2 (2026-10-20): the numerical recognition thresholds lack the number of panelists, the threshold method and the tasting liquid in the abstract, which falls under the holding rules of research protocol §5 and §8, so the status was lowered from supported to preliminary.
Heat treatment of the hop beta-acid colupulone under wort-boiling conditions gave cohulupone, hulupinic acid, nortricyclocolupone, and multiple tricyclocolupone epimer pairs as the main bitter products. Depending on structure, the recognition threshold ranged from 7.9 to 90.3 µmol/L. The lowest were cohulupone at 7.9 µmol/L (described as a short-lasting, iso-alpha-acid-like bitterness) and hydroxytricyclocolupone at 14.7 µmol/L (described as a long-lingering, prickly bitterness perceived at the back of the tongue and throat). Concentrations differed markedly between beer types.
Limitations: Abstract-level extraction. Panel size, threshold method, and tasting matrix are not stated in the abstract, and no DoT (dose-over-threshold) calculation is reported, so the abstract does not establish that these compounds actually exceed threshold in the beers analyzed.
The breadth of these conversion products is also borne out by recent analysis. A study examining hop-forward commercial craft and lager beers by tandem mass spectrometry newly identified twelve compounds, including tricyclocohumol, as hop bitter acid derivatives. Compared with wheat beer, which is low in hop aroma, the mass signals for these compounds came out clearly higher in the hop-forward beers. This study itself did not run a sensory evaluation of bitterness and does not show which compound is bitter to what degree, but it does show that this many diverse compounds exist at once behind the single absorbance number called IBU.
tri- and tetracyclic iso-α-acid transformation products is documented as having Analysis of hop-rich commercial craft and lager beers by UHPLC-MS and tandem MS identified 12 compounds, including tricyclocohumol, as “hops bitter acid derivatives.” Compared with wheat beer, which is low in hop-derived aroma, these mass traces were markedly higher in the hop-rich beers.supported
Scope: Layer: food. Commercial craft and lager beers (hop-rich) and wheat beer (a control low in hop aroma).
"By tandem mass spectrometry we were able to identify twelve hops bitter acid derivatives like cohumulinic acid [C14H20O4], hulupinic acid [C15H20O4], cohulupone [C19H26O4], (ad)humulone[C21H30O5], tricyclocohumol [C20H30O6], or tetracyclohumol [C20H30O6] on level two by comparison of fragmentation patterns and intensities with literature data." "...the corresponding mass traces are decisively higher in hops richer craft and lager beers verifying their discriminating character."
Limitations: This is a “level two” identification (assignment by comparison with literature data), not confirmation with authentic reference standards. Quantitative values and thresholds for individual compounds are not given in the text.
Bonus: bitterness moves not only in 「量」 (in translation: "quantity") but also in 「質」 (in translation: "quality")
When beer is stored, iso-α-acids break down and turn into a group of conversion products with a harsh, lingering bitterness. An observation has also been reported that dry hopping with aged hops lowers bitterness quality. Even with the same IBU number, the bitterness inside is not the same.
Scope: trans-specific conversion products formed during beer aging. Quantified in stored commercial beer
Note: Explains the change in the quality (not only intensity) of beer bitterness with storage. The threshold range given in the abstract is for the compound set, not per-compound values.
Tricyclocohumol, tricyclocohumene, isotricyclocohumene, tetracyclocohumol, and epitetracyclocohumol were identified as previously unknown trans-specific iso-α-acid conversion products. The abstract states that these compounds show a harsh, lingering bitterness with threshold concentrations ranging from 5 to 70 µmol/L, and their formation during beer aging was confirmed by HPLC-MS/MS.
Limitations: Abstract-level extraction. The 5-70 µmol/L range is given for the compound set as a whole, with no per-compound assignment. The solvent or matrix, number of panelists, and threshold method for the threshold test are not stated in the abstract.
Quantitative LC-MS/MS profiling of 56 hop-derived sensometabolites found that proton-catalyzed cyclization of trans-iso-α-acids is the quantitatively dominant reaction, generating tricyclic and tetracyclic compounds (tricyclocohumol, tricyclocohumene, isotricyclocohumene, tetracyclocohumol, epitetracyclocohumol) with a persistent, harsh bitterness, and that these accumulate in beer as storage time and temperature increase. The authors propose controlling the beer's initial pH and keeping storage temperature low to delay the onset of the harsh bitter aftertaste.
Limitations: Abstract-level extraction. The number of beers, storage temperature, duration, and pH values are not stated in the abstract. The recommendations on pH and temperature are the authors' inference from the chemistry, not a reported sensory storage trial.
Context: An independent group quantified the degradation of trans-iso-α-acids during storage of commercial lager beer by HPLC, derived kinetic parameters (activation energy, rate constant), and concluded that the iso-α-acid loss 'can be used in the beer industry to predict the alteration of the bitterness of beer during storage.' This independently confirms the precursor chemistry (storage-driven trans-iso-α-acid loss) behind the harsh-bitterness conversion products.
Limitations: Does not identify tricyclic or tetracyclic conversion products or conduct a sensory test. The sensory attribution of the harsh, lingering bitterness to those compounds still relies solely on the Munich (Intelmann/Hofmann) group.
beer — bitterness quality is reduced by: Dry hopping with hops of a storage index (HSI) above 0.4. Bitterness quality scored 4.9-5.0 on a 1-5 scale at HSI 0.3, and declined at higher HSIpreliminary
Scope: A finding specific to dry hopping (when hops with HSI > 0.4 are used). In kettle hopping, a sister study found no bitterness-quality problem at high HSI when boiled for at least 45 minutes
Note: Quality and intensity behave differently: quality declined monotonically, while measured BU peaked at an intermediate HSI in 2 of 3 cultivars. 10 assessors, one brewing trial per cell.
Bitterness quality scored highest with the freshest hops at HSI 0.3 (on a 1-5 scale: Celeia 4.9 +/- 0.21, Aurora 4.9 +/- 0.21, Styrian Wolf 5.0 +/- 0.16), and declined at higher HSI. The bitterness-intensity response was cultivar-dependent. The abstract states that beer brewed with hops above HSI 0.4 showed deviations in aroma and bitterness compared with beer made from fresh hops, and the source concludes that, since bitterness quality decreases as HSI rises, humulinone oxidation products could have a negative impact on the perceived bitterness of hops.
Limitations: 10 assessors and a single 1-5 quality scale. No statistical test for the decline in quality is cited here. The explanation attributing the decline to humulinone oxidation products is a mechanism proposed by the authors ('could have a negative impact'), not a measured causal verification.
Limits: In a sister kettle-hopping study from the same group (the same 3 cultivars, HSI 0.3-0.7, the same panel design of 10 assessors), the effect of HSI on bitterness depended on boiling time: verbatim 'The quality of bitterness after 5 min of boiling decreased with increasing HSI' (Celeia at 5 min: 4.8 +/- 0.4 at HSI 0.3 versus 1.5 +/- 0.3 at HSI 0.6). However, verbatim 'If boiling is conducted for a long enough time (at least 45 min), then aged hops are not problematic' (regarding bitterness), and the abstract states 'The old hops were not problematic in terms of bitterness or early hopping time.' The kettle limit values were HSI 0.5 (Aurora, Celeia) and 0.6 (Styrian Wolf), against verbatim 'the limit for dry hopping at 0.4 for Aurora and Celeia and at 0.5 for Styrian Wolf.'
Limitations: Because this is the same laboratory as the dry-hopping study, it is not an independent replication but rather limits the scope of the claim (hop addition mode and boiling time). One brewing trial per cell.
This note's primary sources are a mix of ones checked in full text and ones that remain abstract-only (see each card's review-status indicator). The description of IBU's definition and the cis/trans bitterness difference is based not on the original paper but on the description in a review that cites it, and its confidence level is treated conservatively.
This is not to say IBU is meaningless — it remains a useful measure for traditionally kettle-hopped beer, and the gap widens mainly in combination with modern techniques such as dry hopping, or with comparisons that cross differences in alcohol content and hop composition.
Frequently asked
Does a higher IBU number mean a beer tastes more bitter?
A study of more than 120 strongly hopped commercial beer brands reports that the relationship between IBU and rated bitterness intensity was nonlinear rather than a straight line. The same IBU difference cannot be read as the same difference in bitterness. What the current material supports is that a higher IBU does not necessarily mean a beer is felt to be more bitter. The conditions and results of each study are laid out in Do beers with higher IBU taste more bitter?.
Why can a dry-hopped beer taste restrained in bitterness even as its IBU rises?
Multiple surveys examining commercial beer before and after dry hopping report that the rise in IBU comes from the extraction of humulinones and polyphenols, while the iso-α-acid concentration that actually matters falls instead. Iso-α-acids are compounds positioned as the main contributor to beer bitterness, so when their concentration falls while only the IBU measure rises, the correspondence between the number and the sensation is prone to break down in dry-hopped beer. However, this survey remains at the level of an abstract description, and the size of the rise or fall for each individual brand has not been confirmed.
What is not known
The description of IBU's definition and the bitterness difference between cis/trans isomers remains a secondhand citation via a review that cites it, not the original paper, and the original itself has not been checked.
Whether β-acid conversion products actually exceed threshold in real beer and contribute to perceived bitterness cannot be confirmed from the threshold-test abstract alone.
The relationship between the compositional ratio of hop acids and polyphenols and the temporal bitterness descriptors remains at the level of an abstract description, and the effect size for each individual component has not been measured.