Coffee brew is a beverage made by grinding roasted beans and extracting the components in water or hot water. Bitterness arises not only from caffeine but is rearranged by roasting byproducts and extraction conditions.
Reading the bitterness
The photograph shows roasted beans, but the subject of this entry is a single cup of coffee made by grinding them and extracting the components in water or hot water. As the roasting method and the conditions for passing the hot water through the ground coffee vary, so too does the assembly of bitterness that reaches the cup. This bitterness cannot be explained by the word caffeine alone; multiple component groups are involved, including quinides, oligomers of 4-vinylcatechol, and mozambioside roasting byproducts.
For this reason, it is not possible to determine the bitterness of the extracted liquid on the basis of the bean name alone. Roasting conditions change the generation of bitter compounds, and the manner in which hot water passes through changes the amount of those compounds that reach the cup. Which candidate appears strongest is also influenced by the analytical method, and instrument taste-sensor responses cannot be equated with the breakdown of bitterness perceived by humans.
Nor is taste sensation on the drinking side fixed. In a trial with 155 subjects, a change occurred immediately after drinking coffee in which sweetness perception increased and bitterness perception decreased; this change was repeated even in decaffeinated coffee. In large-scale genetic statistics, bitterness sensitivity to quinine and PROP was negatively related to coffee intake, while bitterness sensitivity to caffeine was positively related. Simple logic does not hold — that a person sensitive to bitterness will necessarily avoid coffee.
Structured claims — open a section to read each claim's confidence, evidence and limits
Composition and content (5)
coffee brew has candidate contributors to its bitterness: Nicotinic acid, L-lactic acid, and nicotinamide in the research fractionpreliminary
Scope: 4 stages of roast level evaluated by instrumental chemical analysis and electronic taste sensor
Note: Candidate ranking was by PLS-R and taste sensor, not by human sensory panel.
By PLS-R, nicotinic acid, L-lactic acid, and nicotinamide were identified as candidates; addition of these substances increased the bitter response of the taste sensor.
Limitations: The abstract contains no mention of human sensory panel.
Limits: Research combining human sensory and receptor systems has identified quinides and mozambioside byproducts as bitter contributors, whereas the list of candidates from electronic taste sensor (nicotinic acid, lactic acid, nicotinamide) does not overlap. Candidates depend on the measurement system.
Limitations: Not a direct replication, but rather a note on the divergence of candidate lists due to methodological differences.
Show data
ingredient-coffee-brew —has_candidate_bitterness_contributors→ (the claim sentence above)claim: clm-coffee-sensor-candidates / type: composition / updated: 2026-08-30
coffee brew has bitter contributors other than caffeinesupported Scope: Multiple lines of evidence: human sensory testing, receptor assays, and quantitative chemistry
Note: A group of counterexamples to the common belief that 「コーヒーの苦味＝カフェイン」 (in translation: "coffee's bitterness equals caffeine"). The quantitative contribution ratio is not recorded, since the full text was not confirmed.
Roasting-generated compounds other than caffeine (quinides, 4VC oligomers) define the bitterness quality, and the bitterness signature can be controlled through roasting/extraction conditions.
Limitations: Only the abstract was checked.
Suggests that the group of roasting-degradation products of mozambioside contributes to coffee's bitterness profile.
Limitations: Only the abstract was checked.
Context: Given that caffeine itself is not a suitable representative of bitterness, coffee's bitterness cannot be reduced to caffeine alone.
Limitations: This is not direct contribution-ratio data.
Show data
ingredient-coffee-brew —has_bitter_contributors_other_than→ compound-caffeineclaim: clm-coffee-bitter-beyond-caffeine / type: composition / updated: 2026-08-30
Scope: Quantification by UV-Vis spectrophotometry (330 and 272 nm, 70% ethanol extraction), compared against a standard calibration curve
Note: หนองน้ำดำ (Nong Nam Dam) village is not necessarily the same growing area as Doi Tung/Doi Chang, but it connects to this cluster in the context of arabica grown under a Royal Project in northern Thailand. Only the abstract was checked (the full-text PDF was rasterized and could not be extracted).
「ปริมาณกรดคลอโรจีนิกในสารสกัดกาแฟทุกชนิดในเมล็ดดิบ 2.46-3.82%...สูงกว่าในเมล็ดคั่วที่มีปริมาณ 0.78-1.75%(confirmed by agreement between the full text's Table 1 and running text; the abstract's stated "0.78-1.78%" is an error)...ปริมาณคาเฟอีนในสารสกัดเมล็ดกาแฟคั่วทุกชนิดมีปริมาณ 1.76-4.15% สูงกว่าในเมล็ดดิบ(the green-bean side is 0.43-0.82%, newly confirmed in Table 2)」 (in translation: "The chlorogenic acid content in extracts of every coffee type in green beans was 2.46-3.82%...higher than the 0.78-1.75% in roasted beans...the caffeine content in extracts of every roasted coffee type was 1.76-4.15%, higher than in green beans") (both statistically significant at the 0.05 level). The full text confirms the extraction method (moka pot, 12 g beans/120 ml water, about 6 minutes, filtered hot and then cooled, diluted 1:100 and 1:200 with 70% ethanol), the measuring instrument (Metertech SP8001 UV-Vis spectrophotometer, CGA at 330 nm/caffeine at 272 nm, calibration curve R²=0.9998/0.9996), and the pH range (5.34-6.03, weakly acidic in both groups).
Limitations: The full text was OCR-extracted (because the PDF was rasterized, pdftotext could extract only 8 bytes, so it was read using Tesseract OCR). The roasted-bean CGA upper bound stated in the abstract, "0.78-1.78%," is an error; based on agreement between the full text's Table 1 and the running text, the correct figure is "0.78-1.75%" (FF=1.755% is the maximum). Descriptions of physiological-effect claims (blood glucose, inflammation, antioxidant activity, lipid reduction) were excluded from coverage.
Confirmed in the full text (OCR) the quantitative data underlying the abstract's summary values. 6 green-bean samples (CA-CF) and 6 roasted-bean samples (FA-FF), 1 of which was arabica from Nong Nam Dam village (the remaining 5 were commercially purchased arabica from areas in northern Chiang Rai province run by the Royal Project Foundation). Extraction used a moka pot (12 g beans/120 ml water/about 6 minutes, filtered hot and then left to cool at room temperature, diluted 1:100 and 1:200 with 70% ethanol); a UV-Vis spectrophotometer (Metertech SP8001) measured CGA at 330 nm and caffeine at 272 nm, with 3 replicates. The calibration curve followed the method of Belay and Gholap (2009), with R²=0.9998 (CGA)/0.9996 (caffeine). pH across all samples was in the weakly acidic range of 5.34-6.03. All roasted samples met the minimum caffeine content of 1% required for 『真正コーヒー』 (in translation: "genuine coffee") under Thai Industrial Product Standard มผช.268/2558 and Ministry of Public Health Notification 197/2543.
Limitations: Because the data were OCR-extracted, there is a risk that some digits were misread; the figures in this row are limited to those confirmed as mutually consistent across the Table itself and its multiple restatements in the running text. The authors themselves note the limitation that the lower CGA values compared with the prior study (Belay and Gholap 2009) may be due to a difference in the heating conditions during extraction (the moka pot uses boiling water). Descriptions of health effects (blood glucose, inflammation, antioxidant activity, lipid reduction) were excluded from coverage.
Show data
compound-chlorogenic-acid —has_measured_content_in→ ingredient-coffee-brewclaim: clm-northern-arabica-chlorogenic-caffeine / type: composition / updated: 2026-09-01
Scope: Connection of the GI-standard product classification (including green beans, roasted beans, and ground coffee) to the existing ingredient-coffee-brew entry, a beverage extracted from roasted beans, as a geographically restricted version of that entry
Context: The GI registry defines DoiTung Coffee as a product (Class: Processed Foodstuffs) that includes raw coffee beans, roasted coffee beans, and ground coffee, and it is positioned in the same downstream product lineage as the existing ingredient-coffee-brew entry (roasted-bean extract).
Show data
ingredient-doi-tung-coffee —is_associated_with→ ingredient-coffee-brewclaim: clm-doitung-coffee-brew-link / type: taxonomy / updated: 2026-09-01
Note: Connects to the existing entry ingredient-coffee-brew.
Context: Describes kopi as a type of coffee, calling it Hainanese-style kopi (coffee).
Show data
ingredient-kopi —is_a_type_of→ ingredient-coffee-brewclaim: clm-kopi-coffeebrew-link / type: taxonomy / updated: 2026-09-02
Sensory (9)
coffee brew shows the sensory response bitternesspreliminary Scope: Electronic taste-sensor response to fractionated coffee extract
Note: The cited experiment does not provide confirmation by human sensory evaluation.
Fractionated coffee extract generated a bitter response value in an electronic taste-sensing system.
Limitations: Instrument response is not human bitterness rating.
Show data
ingredient-coffee-brew —has_sensory_response→ sensation-bitterclaim: clm-coffee-bitter / type: sensory / updated: 2026-08-30
coffee brew has_consumption_reduced_by Genetically high bitterness perception (the classical bitterness aversion model)contested
Scope: Mendelian randomization in UK Biobank (n = 438,870, white British ancestry)
Note: This holds for quinine and PROP bitterness (-0.081 / -0.021 cups/day per SD), but reverses for caffeine bitterness, in which higher intake increases. Because the direction differs by substance, the rule 『苦味感受性が高いほど苦味飲料を避ける』 (in translation: "the higher the bitterness sensitivity, the more one avoids bitter beverages") cannot be treated as general.
Genetically predicted PROP and quinine bitterness 1 SD higher was associated with decreased coffee intake (respectively -0.021 [-0.031, -0.011] and -0.081 [-0.108, -0.054] cups/day). This is consistent with the classical bitterness aversion model.
Limitations: Single SNP per trait. Instrumental variables estimated from a small twin sample (n = 1757). Intake was self-reported.
Contradicts: The direction reverses for caffeine: genetically predicted caffeine bitterness 1 SD higher was associated with increased coffee intake (+0.146 [0.103, 0.189] cups/day) and higher odds of heavy consumption (>4 cups/day) (OR 1.207 [1.126, 1.294]); tea showed a mirror pattern (caffeine bitterness -0.045 [-0.061, -0.030]).
Limitations: Instrumental variable caveats are the same as in the supporting row; the reversal is compound-specific and is not a general contradiction of bitterness aversion.
Show data
ingredient-coffee-brew —has_consumption_reduced_by→ (the claim sentence above)claim: clm-bitter-perception-lowers-coffee-intake / type: sensory / updated: 2026-08-30
coffee brewImmediately after intake, sweetness sensitivity rose and bitterness sensitivity fell. The same was true for both regular coffee and decaf can remain changed at: supported
Scope: Taste and smell sensitivity testing before and after coffee intake in 155 people
Note: Smell sensitivity did not change, and the effect was reproduced with decaf, so it cannot be attributed to caffeine alone. A short-term effect.
After regular coffee intake, sweetness sensitivity increased and bitterness sensitivity decreased, and the same held when repeated with decaf. Smell sensitivity did not change.
Limitations: A short-term sensitivity effect; does not measure food preference or long-term adaptation.
Context: The decrease in bitter sensitivity (F(1,99)=4.7975, p=0.031) was larger among participants who did not drink coffee daily (n=25) and smaller among daily drinkers. Verbatim: "the decrease in bitter sensitivity (F(1,99) = 4.7975, p = 0.031) was found to be larger in participants that did not consume coffee daily (n = 25; decrease in mean bitter sensitivity score of -1.0)... We found no effects of gender on the changes in sensitivity"
Limitations: A moderating factor not in the abstract. An exploratory subgroup analysis.
Show data
ingredient-coffee-brew —can_remain_changed_after→ (the claim sentence above)claim: clm-coffee-short-term-taste-sensitivity / type: sensory / updated: 2026-08-31
coffee brew — perceived bitterness is suppressed by: sweetnesspreliminary Scope: Layer: food. A human psychophysical test of mixtures of brewed coffee and sucrose (abstract only; concentration and number of participants not confirmed).
The source measured taste-taste and flavor-taste interactions (suppression) for caffeine-sucrose mixtures and coffee-sucrose mixtures. Both mixtures showed broad hypoadditivity in overall taste/flavor intensity, with a suppression rate of roughly 30-40%. Mutual suppression between components was also observed (within the scope of the abstract).
Limitations: The full-text original has not been obtained; this is based on the abstract and a secondary summary. The coffee cultivar and roast conditions used were not confirmed.
Show data
ingredient-coffee-brew —has_perceived_bitterness_suppressed_by→ sensation-sweetclaim: clm-w67a-coffee-sucrose-flavor-suppression / type: sensory / updated: 2026-09-06
coffee brew — perceived bitterness is affected by: Brew temperature (for dark roast, raising it 4°C → 22°C → 92°C significantly increased bitterness) and total dissolved solids. Sourness was also higher at 92°C; bitterness and sourness did not cancel each other out but moved in the same directionsupported
Scope: Layer: food. Full-immersion-brewed coffee (light roast and dark roast) was brewed at 4°C, 22°C, and 92°C, and a trained panel of 14 (10 women, 4 men) evaluated 27 attributes. Bitterness, sourness (sour), burnt, smoky, and rubber notes all increased together with rising total dissolved solids (TDS), and for dark roast these attributes increased significantly with rising brew temperature (Liang, Batali, Ristenpart & Guinard 2024).
Note: An example where bitterness and sourness do not suppress each other, but move in the same direction depending on brew strength. Placed here as an example of co-variation rather than mutual suppression between sourness and bitterness.
A trained panel of 14 (10 women, 4 men) performed a descriptive analysis of 27 attributes. For dark roast, “bitter, burnt, smoky, rubber, astringent, whiskey, and earthy notes significantly increased with brewing temperature” (4°C → 22°C → 92°C). Sourness is reported in a separate sentence: “sour and citrus were significantly higher in the 92 °C brews for both roast levels.” Bitterness and sourness showed a positive correlation with total dissolved solids (TDS).
Limitations: The pattern of temperature dependence differs by roast degree — for example, for light roast the attributes barely changed between 4°C and 22°C but rose sharply at 92°C.
Show data
ingredient-coffee-brew —has_perceived_bitterness_affected_by→ (the claim sentence above)claim: clm-w67b-coffee-bitter-sour-covary-tds / type: sensory / updated: 2026-09-06
coffee brew — perceived bitterness is unchanged by: Under conditions where brew strength (TDS) and percent extraction (PE) were fixed at target values, varying the brew water temperature across 87°C, 90°C, and 93°C showed no statistically significant difference in the trained panel's bitterness score (the effect of brew temperature was p=0.11 in multivariate analysis). Attributes including bitterness changed together with TDS, but did not correlate with brew temperature itself.supported
Scope: Layer: food. Drip-brewed coffee. This is a comparison under conditions where brew strength and percent extraction were experimentally controlled, so it may not generalize directly to home brewing, where raising the water temperature also tends to move the concentration. A descriptive sensory analysis of 31 attributes by a trained panel of 12 (8 women, 4 men, ages 19-30).
Note: A counterexample to the naive expectation that “changing the water temperature changes the bitterness.” Note that this is conditional on TDS being held constant.
With brew strength (TDS) and percent extraction (PE) fixed at target values, 3 brew-temperature levels — 87°C, 90°C, and 93°C — were compared. Many attributes, including bitterness, changed together with TDS, while the main effect of brew temperature was not statistically significant in multivariate analysis (p=0.11).
Limitations: This is a result under drip brewing at a specific brew strength/percent extraction setting; whether it generalizes to home brewing, where TDS tends to vary, needs separate confirmation.
Show data
ingredient-coffee-brew —has_perceived_bitterness_unchanged_by→ (the claim sentence above)claim: clm-w68c-coffee-brew-temperature-unchanged / type: sensory / updated: 2026-09-06
coffee brew — perceived bitterness is unchanged by: caffeinesupported Scope: Layer: food. Sensory evaluation on a 5-point scale, 0 (not detectable) to 5 (strongly detected), by 18 panelists (8 women, 10 men, ages 23-31). Caffeine was added to a decaffeinated roasted coffee beverage (a commercial capsule, equivalent to 54 g/L).
"Surprisingly, there was no significant difference in the bitter taste between the decaffeinated sample (3.8) and the sample...where caffeine was added with a value...of 4.0"
Limitations: Limited to a single commercial decaffeinated product (an arabica blend, espresso strength 6/12).
Show data
ingredient-coffee-brew —has_perceived_bitterness_unchanged_by→ compound-caffeineclaim: clm-w70a-decaf-roast-caffeine-added-no-bitterness-increase / type: sensory / updated: 2026-09-07
coffee brew — perceived bitterness is affected by: caffeinepreliminary Scope: Layer: Food. QDA by a trained panel of 18 and consumer preference survey of 120. Traditional and decaffeinated espresso sweetened with sucrose, sucralose, or stevia (multiple rebaudioside A concentrations) were compared.
Regular espressos scored higher than decaffeinated ones for bitterness and other attributes. Decaffeinated samples had lower pH (5.24–5.28 vs 5.40–5.43). In Table 3, sensory acidity was higher on average for regular espresso in all 5 pairs (for example 4.93 vs 3.78), which conflicts with the abstract and conclusions stating that decaffeinated espresso was more acidic.
Limitations: Acidity runs in opposite directions in the table and in the abstract and conclusions, so it cannot be settled.
Show data
ingredient-coffee-brew —has_perceived_bitterness_affected_by→ compound-caffeineclaim: clm-w70a-decaf-espresso-lower-bitterness / type: sensory / updated: 2026-09-07
coffee brew — perceived bitterness is affected by: In a sensory evaluation adding quaker beans (from immature beans, in 3 color grades) to natural-processed specialty coffee at 7 levels of contamination, quaker beans increased bitterness and astringency and brought negative sensory descriptors to aroma and flavor. For quaker beans with an Agtron color of 82.8 or above, a significant sensory difference appeared only once contamination reached 7 beans (10.8%) or more in 1 cup (equivalent to 65 beans).preliminary
Scope: Layer: food. An experiment contaminating natural-processed specialty coffee with quaker beans (from immature beans), with sensory evaluation and volatile compound analysis. The number and training status of the sensory panel are not stated in the abstract.
Beans with a color equal to or above Agtron 82.8 negatively affected the sensory characteristics of natural specialty coffee only once 7 Quaker beans were present in 1 cup (65 beans). Quaker beans made the beverage more astringent and bitter and added negative aroma and flavor sensory notes.
Limitations: The number and training status of the sensory panel, and details of the statistical method, were not confirmed at the abstract stage.
Show data
ingredient-coffee-brew —has_perceived_bitterness_affected_by→ (the claim sentence above)claim: clm-w70a-quaker-beans-increase-bitterness-astringency / type: sensory / updated: 2026-09-07
