{
  "id": "bitterness-in-coffee",
  "url": "https://coffearium.org/entries/bitterness-in-coffee/",
  "license": "CC-BY-4.0",
  "title": "Bitterness in coffee",
  "summary": "Why every coffee tastes bitter – the compounds behind it, from caffeine to roast-formed lactones and phenylindanes, how roast, brewing, species and sugar change it, and why people perceive it differently.",
  "domain": "science",
  "kind": "Concept",
  "stage": "cup",
  "topics": [
    "science"
  ],
  "topicSections": {},
  "lang": "en",
  "status": "verified",
  "reviewer": "Coffearium verification pass",
  "lastVerified": "2026-11-12T00:00:00.000Z",
  "literatureSearched": "2026-11-12T00:00:00.000Z",
  "updated": "2026-11-12T00:00:00.000Z",
  "added": "2026-11-12T00:00:00.000Z",
  "brief": [
    "All coffees have some bitterness. On the Specialty Coffee Association (SCA) descriptive form, cuppers tick “bitter” when it stands out.<sup><a href=\"#src-s1\">1</a></sup>",
    "Key bitter compounds form in roasting, when chlorogenic acids give bitter lactones and, in darker roasts, harsher 4-vinylcatechol oligomers.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s6\">6</a></sup>",
    "Bitter recognition thresholds were 9.8–180 µmol/L for lactones (in water) and 23–178 µmol/L for 4-vinylcatechol oligomers; caffeine’s detection threshold was 1.2 mM in another study.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s10\">10</a></sup>",
    "In a 2020 analysis of eight sensory studies, darker roasts went with more bitterness.<sup><a href=\"#src-s11\">11</a></sup>",
    "In a 1990 study, sucrose suppressed caffeine bitterness and coffee flavour; at matched intensity, coffee did not suppress sweetness.<sup><a href=\"#src-s13\">13</a></sup>"
  ],
  "quickFacts": [
    {
      "label": "Basic taste",
      "value": "Bitter, one of the five basic tastes"
    },
    {
      "label": "Main compound groups",
      "value": "Chlorogenic acid lactones, 4-vinylcatechol oligomers (mostly phenylindanes), caffeine, diketopiperazines, mozambioside and its roast products"
    },
    {
      "label": "Lactone thresholds (water)",
      "value": "9.8–180 µmol/L (recognition, one study)"
    },
    {
      "label": "4-Vinylcatechol oligomer thresholds",
      "value": "23–178 µmol/L (recognition, one study)"
    },
    {
      "label": "Caffeine detection threshold",
      "value": "1.2 mM (33 people, one study)"
    },
    {
      "label": "Roast degree",
      "value": "Darker roasts, more bitterness (2020 analysis of eight studies)"
    },
    {
      "label": "Arabica-specific bitter compound",
      "value": "Mozambioside; traces only in robusta"
    },
    {
      "label": "SCA bitter reference",
      "value": "Dark roasted coffee, about Agtron"
    },
    {
      "label": "In the CVA",
      "value": "A main-taste box on the descriptive form; no scored section of its own"
    }
  ],
  "aka": [
    "Coffee bitterness",
    "Bitter taste of coffee"
  ],
  "linkTerms": [
    "bitterness",
    "bitter taste",
    "bitter compounds",
    "bitter compound",
    "bitter taste receptors",
    "bitter taste receptor"
  ],
  "related": [
    "how-we-taste-coffee",
    "sweetness-in-coffee",
    "cup-taints-and-faults",
    "coffee-consumers-and-preference",
    "chlorogenic-acids",
    "caffeine",
    "caffeine-and-roast-degree",
    "acids-in-coffee",
    "roast-degree-and-colour",
    "what-extraction-is",
    "extraction-yield-and-tds",
    "brew-ratio-and-strength",
    "serving-temperature",
    "milk-and-milk-drinks",
    "cupping",
    "flavour-language",
    "coffea-canephora"
  ],
  "contested": [
    "Mozambioside threshold: a 2015 study measured a bitter recognition threshold of 60 ± 10 µmol/L<sup><a href=\"#src-s8\">8</a></sup>; a 2025 study from a related group reported 132 µM.<sup><a href=\"#src-s9\">9</a></sup> The 2025 abstract does not name the threshold type or say whether the method differed.",
    "Bitter genes and coffee intake: a 2018 Mendelian randomisation study found that people genetically predisposed to perceive caffeine as more bitter drank more coffee<sup><a href=\"#src-s15\">15</a></sup>; a 2021 study in UK and US cohorts found that genetically inferred caffeine sensitivity was more strongly associated with coffee taste preferences than genetically inferred bitter taste perception.<sup><a href=\"#src-s16\">16</a></sup> The two studies used different genetic measures, so their results are not directly comparable (our note)."
  ],
  "corrections": [
    {
      "date": "2026-11-12T00:00:00.000Z",
      "change": "Threshold statements narrowed: only the 2006 lactone study states water as the medium; the 2025 mozambioside value is no longer called a recognition threshold; the 2006 lactone thresholds are named as recognition thresholds.",
      "reason": "The abstracts of Lang et al. 2015, Frank et al. 2007, Bichlmaier et al. 2025 and Keast and Roper 2007 do not name the medium, and the 2025 abstract does not name the threshold type."
    },
    {
      "date": "2026-11-12T00:00:00.000Z",
      "change": "Brief, Explained and Key studies wording brought back to the sources: sugar suppressed caffeine bitterness and coffee flavour (not “the bitterness of coffee”); “much of coffee’s bitterness” replaced; the colour–bitterness link attributed to the analysis; mozambioside products described by their lower thresholds, not as “more potent”; 4-vinylcatechol oligomers described as strongly retained by the grounds.",
      "reason": "Statements were stronger than the abstracts of Calviño et al. 1990, Bichlmaier et al. 2025 and Blumberg et al. 2010 and the 2020 roasting analysis."
    }
  ],
  "sources": [
    {
      "id": "s1",
      "title": "SCA Standard 103-2024 Coffee Value Assessment: Descriptive Assessment",
      "authors": "Specialty Coffee Association",
      "year": 2024,
      "publisher": "Specialty Coffee Association",
      "url": "https://sca.coffee/s/AW_SCA-103_Descriptive-Assessment_Sept2024_Secured.pdf",
      "tier": "A/B",
      "tierNote": "A for the taste lexicon, form and procedure the standard specifies; B otherwise.",
      "quote": "Though all coffees have bitterness, if bitterness stands out, this box should be selected.",
      "quotes": [
        {
          "text": "Dark roasted coffee (about Agtron #35)",
          "for": "bitter reference in the taste lexicon"
        },
        {
          "text": "up to two main tastes that best represent the coffee",
          "for": "up to two main tastes"
        }
      ],
      "scope": {
        "claim": "guidance",
        "context": "SCA descriptive assessment form and procedure (taste lexicon, CATA boxes), September 2024 version.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s2",
      "title": "SCA Standard 104-2024 Coffee Value Assessment: Affective Assessment",
      "authors": "Specialty Coffee Association",
      "year": 2024,
      "publisher": "Specialty Coffee Association",
      "url": "https://sca.coffee/s/AW_SCA-104_Affective-Assessment_Sept2024_Secured.pdf",
      "tier": "A/B",
      "tierNote": "A for what the standard specifies; B for its explanations",
      "quote": "The score is the result of a linear equation that takes into account the affective results from all cupping sections, and the numbers of non-uniform and defective cups, to produce a total that equals 100 when all sections score at 9 and equals 79 when all sections score at 5 (assuming all cups are uniform and non-defective).",
      "quotes": [],
      "scope": {
        "claim": "definition",
        "context": "SCA affective assessment and cupping score, September 2024 version.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s3",
      "title": "World Coffee Research Sensory Lexicon: Unabridged Definition and References, Version 2.0",
      "authors": "World Coffee Research",
      "year": 2017,
      "publisher": "World Coffee Research",
      "url": "https://worldcoffeeresearch.org/download/5806d8c7-cd7c-4882-a81a-2850bd6fbd36",
      "tier": "B",
      "quote": "The fundamental taste factor associated with a caffeine solution.",
      "quotes": [
        {
          "text": "Because coffee has significant bitterness compared with other food products, panelists usually only evaluate four to six samples in a session",
          "for": "few samples per session because of bitterness"
        }
      ],
      "scope": {
        "claim": "definition",
        "context": "WCR descriptive-panel lexicon: attribute definitions and intensity references.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s4",
      "title": "Bioresponse-guided decomposition of roast coffee beverage and identification of key bitter taste compounds",
      "authors": "Frank, O.; Zehentbauer, G.; Hofmann, T.",
      "year": 2006,
      "publisher": "European Food Research and Technology 222(5–6), 492–508",
      "url": "https://portal.fis.tum.de/en/publications/bioresponse-guided-decomposition-of-roast-coffee-beverage-and-ide/",
      "doi": "10.1007/s00217-005-0143-6",
      "tier": "A",
      "quote": "Sensory-guided fractionation of a roasted coffee brew by means of sequential solvent extraction, ultrafiltration, and RP",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Lab: sensory-guided fractionation of one roasted coffee brew; recognition thresholds in water.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s5",
      "title": "Structure determination and sensory analysis of bitter-tasting 4-vinylcatechol oligomers and their identification in roasted coffee by means of LC-MS/MS",
      "authors": "Frank, O.; Blumberg, S.; Kunert, C.; Zehentbauer, G.; Hofmann, T.",
      "year": 2007,
      "publisher": "Journal of Agricultural and Food Chemistry 55(5), 1945–1954",
      "url": "https://pubmed.ncbi.nlm.nih.gov/17269788/",
      "doi": "10.1021/jf0632280",
      "tier": "A",
      "quote": "The structures of these bitter compounds show strong evidence that they are generated by oligomerization of 4-vinylcatechol released from caffeic acid moieties upon roasting.",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Lab: model roasting of caffeic acid; compounds then confirmed in a coffee brew by LC-MS/MS; medium of thresholds not stated in abstract.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s6",
      "title": "Quantitative studies on the influence of the bean roasting parameters and hot water percolation on the concentrations of bitter compounds in coffee brew",
      "authors": "Blumberg, S.; Frank, O.; Hofmann, T.",
      "year": 2010,
      "publisher": "Journal of Agricultural and Food Chemistry 58(6), 3720–3728",
      "url": "https://portal.fis.tum.de/en/publications/quantitative-studies-on-the-influence-of-the-bean-roasting-parame/",
      "doi": "10.1021/jf9044606",
      "tier": "A",
      "quote": "Quinides 4-11, exhibiting a coffee-typical bitter taste profile, were found to be preferentially formed under slight to",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Lab roasts (260 °C for 60–600 s; 240 s at 190–280 °C) and hot-water percolation of the ground coffee.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s7",
      "title": "LC-HRMS-based identification and quantification of novel 2,5-diketopiperazines in roasted coffee",
      "authors": "Kaur, S.; Schmidt, D.; Hölscher, J.; Kuhnert, N.",
      "year": 2026,
      "publisher": "Food Research International 223, 117871",
      "url": "https://doi.org/10.1016/j.foodres.2025.117871",
      "doi": "10.1016/j.foodres.2025.117871",
      "tier": "A",
      "quote": "DKP concentrations determined in aqueous infusions of coffee were in the range of reported bitterness threshold concentrations.",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "LC-HRMS of roasted coffees and their aqueous infusions; thresholds taken from earlier literature.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s8",
      "title": "Mozambioside Is an Arabica-Specific Bitter-Tasting Furokaurane Glucoside in Coffee Beans",
      "authors": "Lang, R.; Klade, S.; Beusch, A.; Dunkel, A.; Hofmann, T.",
      "year": 2015,
      "publisher": "Journal of Agricultural and Food Chemistry 63(48), 10492–10499",
      "url": "https://agris.fao.org/search/en/records/65de73650f3e94b9e5cf59b1",
      "doi": "10.1021/acs.jafc.5b04847",
      "tier": "A",
      "quote": "Sensory evaluation revealed a bitter taste recognition threshold of 60 (±10) μmol/L.",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Green and roasted arabica and robusta samples; one sensory threshold panel; medium not stated in abstract.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s9",
      "title": "Contribution of mozambioside roasting products to coffee's bitter taste",
      "authors": "Bichlmaier, C.; Fröhlich, S. M.; Brychcy, V.; Graßl, A.; Behrens, M.; Lang, R.",
      "year": 2025,
      "publisher": "Food Chemistry 469, 142547",
      "url": "https://portal.fis.leibniz-lsb.tum.de/en/publications/contribution-of-mozambioside-roasting-products-to-coffees-bitter-/",
      "doi": "10.1016/j.foodchem.2024.142547",
      "tier": "A",
      "quote": "The major roasting products exhibited lower thresholds (27–80 μM) than 1 (132 μM).",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Roasting experiments, UHPLC-MS/MS, human panel with TAS2R43 genotyping; threshold type and medium not stated in abstract.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s10",
      "title": "A complex relationship among chemical concentration, detection threshold, and suprathreshold intensity of bitter compounds",
      "authors": "Keast, R. S. J.; Roper, J.",
      "year": 2007,
      "publisher": "Chemical Senses 32(3), 245–253",
      "url": "https://deakin.figshare.com/articles/journal_contribution/A_complex_relationship_among_chemical_concentration_detection_threshold_and_suprathreshold_intensity_of_bitter_compounds/20553975",
      "doi": "10.1093/chemse/bjl052",
      "tier": "A",
      "quote": "The results suggest a complex relationship between chemical concentration, detection threshold",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "33 adults; detection thresholds and intensity of caffeine, quinine-HCl and PROP; medium not stated in abstract.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s11",
      "title": "Roasting Conditions and Coffee Flavor: A Multi-Study Empirical Investigation",
      "authors": "Münchow, M.; Alstrup, J.; Steen, I.; Giacalone, D.",
      "year": 2020,
      "publisher": "Beverages 6(2), 29",
      "url": "https://www.mdpi.com/2306-5710/6/2/29",
      "doi": "10.3390/beverages6020029",
      "tier": "A",
      "quote": "With respect to the direction of the correlations, colour (i.e., a darker roast) was positively correlated with bitterness and negatively correlated with sweetness, acidity, and fruitiness (Table 4).",
      "quotes": [],
      "scope": {
        "claim": "finding",
        "context": "Eight descriptive-analysis studies; colour and timing varied together only in Studies 1 and 2.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s12",
      "title": "Coffee Flavor: A Review",
      "authors": "Seninde, D. R.; Chambers, E., IV",
      "year": 2020,
      "publisher": "Beverages 6(3), 44",
      "url": "https://www.mdpi.com/2306-5710/6/3/44",
      "doi": "10.3390/beverages6030044",
      "tier": "A",
      "quote": "This accounts for a large proportion of the higher bitterness usually found in coffee brews made from Robusta beans.",
      "quotes": [],
      "scope": {
        "claim": "cited",
        "context": "Review of published coffee flavour studies; brewing and roast findings cited from individual studies.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s13",
      "title": "Interactions in caffeine–sucrose and coffee–sucrose mixtures: evidence of taste and flavor suppression",
      "authors": "Calviño, A. M.; García-Medina, M. R.; Cometto-Muñiz, J. E.",
      "year": 1990,
      "publisher": "Chemical Senses 15(5), 505–519",
      "url": "https://academic.oup.com/chemse/article/15/5/505/261627",
      "doi": "10.1093/chemse/15.5.505",
      "tier": "A",
      "quote": "Thus, the suppression of bitterness and coffee flavor qualities increase when sucrose levels increase, and similarly, the suppression of sweetness increases when caffeine or coffee levels rise.",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Taste panel; caffeine–sucrose and coffee–sucrose mixtures, compared at matched perceived intensity.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s14",
      "title": "The impact of individual variations in taste sensitivity on coffee perceptions and preferences",
      "authors": "Masi, C.; Dinnella, C.; Monteleone, E.; Prescott, J.",
      "year": 2015,
      "publisher": "Physiology & Behavior 138, 219–226",
      "url": "https://pesquisa.bvsalud.org/ses/resource/pt/mdl-25446205",
      "doi": "10.1016/j.physbeh.2014.10.031",
      "tier": "A",
      "quote": "While coffee liking was unrelated to PROP status, PROP non-tasters (NTs) added more sugar to the coffee samples",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "Participants grouped by PROP taster status and fungiform papilla density; coffee samples and standard solutions.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s15",
      "title": "Understanding the role of bitter taste perception in coffee, tea and alcohol consumption through Mendelian randomization",
      "authors": "Ong, J.-S.; Hwang, D. L.-D.; Zhong, V. W.; An, J.; Gharahkhani, P.; Breslin, P. A. S.; Wright, M. J.; Lawlor, D. A.; Whitfield, J.; MacGregor, S.; Martin, N. G.; Cornelis, M. C.",
      "year": 2018,
      "publisher": "Scientific Reports 8, 16414",
      "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6237869",
      "doi": "10.1038/s41598-018-34713-z",
      "tier": "A",
      "quote": "Higher caffeine perception was also associated with increased risk of being a heavy (>4 cups/day) coffee drinker (OR 1.207 [1.126, 1.294]).",
      "quotes": [
        {
          "text": "A standard deviation (SD) higher in genetically predicted bitterness of caffeine was associated with increased coffee intake (0.146 [95%CI: 0.103, 0.189] cups/day)",
          "for": "0.146 more cups a day per SD (causal estimate, abstract)"
        }
      ],
      "scope": {
        "claim": "finding",
        "context": "UK Biobank, up to 438,870 participants of white-British ancestry; single-SNP Mendelian randomisation.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s16",
      "title": "Genetic determinants of liking and intake of coffee and other bitter foods and beverages",
      "authors": "Cornelis, M. C.; van Dam, R. M.",
      "year": 2021,
      "publisher": "Scientific Reports 11, 23845",
      "url": "https://doaj.org/article/4862499fe88e4467b3f11999b3158bb0",
      "doi": "10.1038/s41598-021-03153-7",
      "tier": "A",
      "quote": "Contrary to our hypotheses, genetically inferred caffeine sensitivity was more strongly associated with coffee taste preferences than with genetically inferred bitter taste perception.",
      "quotes": [],
      "abstractOnly": true,
      "scope": {
        "claim": "finding",
        "context": "UK and US cohorts; genetic association with coffee and other bitter foods.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    },
    {
      "id": "s17",
      "title": "Influences of Product Temperature on Emotional Responses to, and Sensory Attributes of, Coffee and Green Tea Beverages",
      "authors": "Pramudya, R. C.; Seo, H.-S.",
      "year": 2018,
      "publisher": "Frontiers in Psychology 8, 2264",
      "url": "https://www.frontiersin.org/articles/10.3389/fpsyg.2017.02264/full",
      "doi": "10.3389/fpsyg.2017.02264",
      "tier": "A",
      "quote": "all participants self-reported that they habitually drink one or more cups of coffee with no added condiments",
      "quotes": [],
      "scope": {
        "claim": "finding",
        "context": "79 black-coffee drinkers; CATA for brewed coffee at 5, 25 and 65 °C.",
        "checked": "2026-11-12T00:00:00.000Z"
      },
      "accessed": "2026-11-12T00:00:00.000Z"
    }
  ],
  "markdown": "## What bitterness is\n\nBitter is one of the basic tastes, and the Specialty Coffee Association (SCA) lists it in its taste lexicon for cupping.<sup><a href=\"#src-s1\">1</a></sup> How taste receptors work is explained under how we taste coffee. Every coffee has some bitterness. The SCA’s cupping standard says so directly.<sup><a href=\"#src-s1\">1</a></sup>\n\nTo show tasters what bitter means, SCA Standard 103-2024 uses a dark roasted coffee of about Agtron #35.<sup><a href=\"#src-s1\">1</a></sup> The World Coffee Research (WCR) Sensory Lexicon, written for trained panels, uses a caffeine solution instead.<sup><a href=\"#src-s3\">3</a></sup>\n\nCoffee’s bitterness does not come from one substance. Laboratory studies have identified several groups of bitter compounds in roasted coffee.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s7\">7</a>,<a href=\"#src-s8\">8</a></sup> Most of them form during roasting.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s7\">7</a></sup>\n\n## The bitter compounds\n\n- **Chlorogenic acid lactones.** Also called quinides. Roasting forms them from the chlorogenic acids of the green coffee.<sup><a href=\"#src-s4\">4</a></sup> A 2006 study identified them as key bitter compounds of a roasted coffee brew.<sup><a href=\"#src-s4\">4</a></sup> Their bitterness was described as typical of coffee.<sup><a href=\"#src-s6\">6</a></sup>\n- **4-Vinylcatechol oligomers.** Small chains built from 4-vinylcatechol, which roasting releases from caffeic acid; many are phenylindanes.<sup><a href=\"#src-s5\">5</a></sup> They form in darker roasts as the lactones break down, and taste harsher.<sup><a href=\"#src-s6\">6</a></sup>\n- **Caffeine.** Present in the green coffee and stable in heat.<sup><a href=\"#src-s12\">12</a></sup> It is bitter, but more is needed to taste it than for the roast compounds above.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s10\">10</a></sup>\n- **Diketopiperazines.** Small ring-shaped molecules formed by heat from amino acids.<sup><a href=\"#src-s7\">7</a></sup> In a 2026 study, their levels in coffee infusions were in the range of reported bitterness thresholds.<sup><a href=\"#src-s7\">7</a></sup>\n- **Mozambioside and its roast products.** A bitter compound found in arabica seeds, with only traces in robusta.<sup><a href=\"#src-s8\">8</a></sup> Roasting turns part of it into products that are bitter at lower concentrations.<sup><a href=\"#src-s8\">8</a>,<a href=\"#src-s9\">9</a></sup>\n\nA threshold is the lowest concentration at which tasters notice a substance or name its taste. It shows how little is needed. In a 2007 study, it did not predict how bitter stronger caffeine solutions tasted.<sup><a href=\"#src-s10\">10</a></sup> No study consulted for this entry measured what share of a cup’s bitterness each group carries.\n\n## How roast degree changes bitterness\n\nDarker roasts are more bitter. In a 2020 analysis of eight sensory studies, darker roasts went with more bitterness and less acidity.<sup><a href=\"#src-s11\">11</a></sup> In the two studies that varied colour, darker colour also went with less sweetness and fruitiness.<sup><a href=\"#src-s11\">11</a></sup> Longer roasting phases also went with more bitterness, but the authors found colour the stronger of the two predictors.<sup><a href=\"#src-s11\">11</a></sup>\n\nThe kind of bitterness changes as well. Lactones form mainly at light to medium roasts and give a bitterness typical of coffee.<sup><a href=\"#src-s6\">6</a></sup> Under more severe roasting they break down into 4-vinylcatechol oligomers with a harsher bitterness.<sup><a href=\"#src-s6\">6</a></sup> Heated caffeic acid gave a bitterness reminiscent of a strongly roasted espresso.<sup><a href=\"#src-s5\">5</a></sup>\n\nNot every bitter compound follows this path. Caffeine is heat-stable, so roasting changes it little.<sup><a href=\"#src-s12\">12</a></sup> The entry on caffeine and roast degree gives the numbers. Mozambioside, by contrast, is partly broken down during roasting.<sup><a href=\"#src-s8\">8</a></sup>\n\n<figure class=\"iso-fig\">\n<div data-figure=\"bitterness-in-coffee-roast-sequence\"></div>\n<ol class=\"fig-legend\">\n<li data-n=\"1\"><span><b>Green coffee.</b> Holds the precursors: chlorogenic acids, caffeine and, in arabica, mozambioside.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s8\">8</a>,<a href=\"#src-s12\">12</a></sup></span></li>\n<li data-n=\"2\"><span><b>Light to medium roast.</b> Chlorogenic acid lactones form mainly here; their bitterness is typical of coffee.<sup><a href=\"#src-s6\">6</a></sup></span></li>\n<li data-n=\"3\"><span><b>Roast products of mozambioside.</b> Part of the mozambioside breaks down into products with lower bitter thresholds.<sup><a href=\"#src-s8\">8</a>,<a href=\"#src-s9\">9</a></sup></span></li>\n<li data-n=\"4\"><span><b>Diketopiperazines.</b> Formed by heat from amino acids.<sup><a href=\"#src-s7\">7</a></sup></span></li>\n<li data-n=\"5\"><span><b>Dark roast.</b> The lactones break down into 4-vinylcatechol oligomers, including phenylindanes, with a harsher bitterness.<sup><a href=\"#src-s5\">5</a>,<a href=\"#src-s6\">6</a></sup></span></li>\n<li data-n=\"6\"><span><b>Caffeine.</b> Heat-stable; carried through every stage.<sup><a href=\"#src-s12\">12</a></sup></span></li>\n</ol>\n<figcaption>Which bitter compound groups are present at each stage of roasting. Sequence only: the drawing shows no amounts and no curves. Our own drawing. Schematic, not to scale.</figcaption>\n</figure>\n\n## Brewing, species and green coffee\n\nBitter compounds do not leave the grounds at the same speed. In a 2010 study, lactones with one caffeic acid unit dissolved quickly in hot water.<sup><a href=\"#src-s6\">6</a></sup> Lactones with two units came out slowly, and the 4-vinylcatechol oligomers were strongly held back by the grounds.<sup><a href=\"#src-s6\">6</a></sup> Mozambioside passed almost completely into the brew.<sup><a href=\"#src-s8\">8</a></sup>\n\nThe recipe matters. In studies cited by a 2020 review, cold brews and espressos made with more coffee per volume of water were more bitter.<sup><a href=\"#src-s12\">12</a></sup> In one study it cites, slow-drip cold brew was more bitter than steeped cold brew.<sup><a href=\"#src-s12\">12</a></sup> Whether a stronger brew is also a more bitter one at the same extraction is disputed; the entry on brew ratio and strength compares the views.\n\nThe species matters too. Roasted robusta holds about twice as much caffeine as arabica.<sup><a href=\"#src-s12\">12</a></sup> According to the review, this explains much of the higher bitterness usually found in robusta brews.<sup><a href=\"#src-s12\">12</a></sup> Arabica has a bitter compound of its own: mozambioside.<sup><a href=\"#src-s8\">8</a></sup> No study consulted for this entry compared the bitterness of washed and natural coffees directly.\n\n## Sugar, milk and temperature\n\nSugar can lower bitterness. In a 1990 study, sucrose suppressed both the bitterness of caffeine and the flavour of coffee; more sugar suppressed more.<sup><a href=\"#src-s13\">13</a></sup> At similar perceived strength, coffee did not suppress the sweetness of the sugar.<sup><a href=\"#src-s13\">13</a></sup> Sweetness in coffee covers the other side of this balance.\n\nDrinkers use sugar differently. In a 2015 study, people who barely taste the bitter compound PROP (propylthiouracil) added more sugar to coffee than highly sensitive tasters.<sup><a href=\"#src-s14\">14</a></sup> No study consulted for this entry measured how milk changes coffee bitterness; milk itself is covered under milk and milk drinks.\n\nTemperature shifts the picture. In a 2018 study with 79 black-coffee drinkers, 84% chose “bitter” for coffee at 5 or 25 °C, but 67% at 65 °C.<sup><a href=\"#src-s17\">17</a></sup> Other studies found different patterns; the entry on serving temperature compares them.\n\n## Why people taste it differently\n\nPeople differ in how bitter the same coffee tastes. In the 2015 study, people highly sensitive to PROP rated the bitterness of coffee as stronger.<sup><a href=\"#src-s14\">14</a></sup> Yet how much they liked coffee did not depend on their PROP sensitivity.<sup><a href=\"#src-s14\">14</a></sup>\n\nGenes play a part. In a 2018 UK study, people with gene variants that make caffeine taste more bitter drank more coffee.<sup><a href=\"#src-s15\">15</a></sup> A 2021 study found genetic sensitivity to caffeine’s effects more closely linked to coffee taste preferences than bitter perception.<sup><a href=\"#src-s16\">16</a></sup> In a 2025 study, sensitivity to mozambioside products went with intact versions of one bitter receptor gene, *TAS2R43*.<sup><a href=\"#src-s9\">9</a></sup>\n\n## Bitterness in cupping\n\nOn the SCA descriptive form, bitterness has no intensity scale of its own. Fragrance, aroma, flavour, aftertaste, acidity, sweetness and mouthfeel are rated instead.<sup><a href=\"#src-s1\">1</a></sup> Bitter appears in a short list of main tastes, where the taster ticks up to two that best represent the coffee.<sup><a href=\"#src-s1\">1</a></sup> The standard asks tasters to tick bitter when it stands out, especially in the context of the other coffees tasted.<sup><a href=\"#src-s1\">1</a></sup>\n\nThe affective form, SCA Standard 104-2024, scores eight sections from fragrance to overall.<sup><a href=\"#src-s2\">2</a></sup> None of them is bitterness, and the standard does not mention it.<sup><a href=\"#src-s2\">2</a></sup> Trained research panels work differently: they rate bitter intensity against caffeine references.<sup><a href=\"#src-s3\">3</a></sup>\n\n<!-- deep-dive -->\n\n## Reference values\n\n| Parameter | Value | Basis |\n|---|---|---|\n| Chlorogenic acid lactones (quinides), recognition thresholds in water | 9.8–180 µmol/L, depending on structure | Frank et al. 2006<sup><a href=\"#src-s4\">4</a></sup> |\n| 4-Vinylcatechol oligomers, recognition thresholds | 23–178 µmol/L (10 compounds) | Frank et al. 2007<sup><a href=\"#src-s5\">5</a></sup> |\n| Mozambioside, bitter threshold | 60 ± 10 µmol/L, recognition (2015); 132 µM, type not stated (2025) | Lang et al. 2015<sup><a href=\"#src-s8\">8</a></sup>; Bichlmaier et al. 2025<sup><a href=\"#src-s9\">9</a></sup> |\n| Mozambioside roast products, bitter thresholds | 27–80 µM (type not stated) | Bichlmaier et al. 2025<sup><a href=\"#src-s9\">9</a></sup> |\n| Caffeine, detection threshold | 1.2 ± 0.12 mM (mean, 33 people) | Keast and Roper 2007<sup><a href=\"#src-s10\">10</a></sup> |\n| Mozambioside in green coffee | Arabica 396–1188 nmol/g; robusta < 5 nmol/g | Lang et al. 2015<sup><a href=\"#src-s8\">8</a></sup> |\n| Mozambioside in roasted arabica | 232 ± 37 nmol/g | Lang et al. 2015<sup><a href=\"#src-s8\">8</a></sup> |\n| Mozambioside extracted into the brew | 86–98% | Lang et al. 2015<sup><a href=\"#src-s8\">8</a></sup> |\n| Mozambioside and main roast products in coffee powders | 21.0–170.4 nmol/g; mixture at coffee levels recognised as bitter by 80% of panellists | Bichlmaier et al. 2025<sup><a href=\"#src-s9\">9</a></sup> |\n| Diketopiperazines identified | 33, of which 23 newly reported | Kaur et al. 2026<sup><a href=\"#src-s7\">7</a></sup> |\n| Sucrose with caffeine or coffee | Overall taste or flavour intensity suppressed by about 30–40% | Calviño et al. 1990<sup><a href=\"#src-s13\">13</a></sup> |\n| “Bitter” chosen, black coffee, 79 drinkers | 84% at 5 and 25 °C; 67% at 65 °C | Pramudya and Seo 2018<sup><a href=\"#src-s17\">17</a></sup> |\n| SCA bitter example | Dark roasted coffee, about Agtron #35 | SCA 103-2024<sup><a href=\"#src-s1\">1</a></sup> |\n\nThe thresholds above come from different panels and methods. The 2006, 2007 and 2015 studies report recognition thresholds, at which the taste is named; the caffeine study reports a detection threshold, at which something is first noticed; the 2025 abstract does not name the type.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s8\">8</a>,<a href=\"#src-s9\">9</a>,<a href=\"#src-s10\">10</a></sup> Of the abstracts consulted, only the 2006 study names the medium (water); none reports a threshold measured in coffee.<sup><a href=\"#src-s4\">4</a></sup> The values are therefore a guide to relative potency, not a ranking of each compound’s role in a cup.\n\n<figure>\n<div data-figure=\"bitterness-in-coffee-thresholds\"></div>\n<figcaption>Reported bitter thresholds for compound groups formed in or found in coffee, in µmol/L; the lactone values were measured in water, and the other abstracts do not name the medium. Spans cover several compounds of a group. Caffeine, for comparison, had a detection threshold of 1.2 mM (1,200 µmol/L) in a separate panel, far off this scale.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s8\">8</a>,<a href=\"#src-s9\">9</a>,<a href=\"#src-s10\">10</a></sup></figcaption>\n</figure>\n\n## Bitter compound groups in detail\n\n**Lactones.** A 2006 study split a roasted coffee brew into fractions, guided at each step by a taste panel.<sup><a href=\"#src-s4\">4</a></sup> It identified seven caffeoyl and feruloyl quinides as intense bitter compounds, among them 3-O-caffeoyl-γ-quinide and 4-O-caffeoyl-γ-quinide.<sup><a href=\"#src-s4\">4</a></sup> A further, highly complex and intensely bitter fraction held lactone isomers esterified with p-coumaric, caffeic, ferulic, 3,4-dimethoxycinnamic and quinic acid; three dicaffeoyl quinides from it were isolated.<sup><a href=\"#src-s4\">4</a></sup> Thresholds, measured for the first time, depended strongly on structure.<sup><a href=\"#src-s4\">4</a></sup>\n\n**4-Vinylcatechol oligomers.** When caffeic acid was heated alone, it gave an intense bitterness reminiscent of a strongly roasted espresso-type coffee.<sup><a href=\"#src-s5\">5</a></sup> Mass spectrometry and NMR identified 10 bitter compounds in the reaction products, and each was confirmed in a coffee brew.<sup><a href=\"#src-s5\">5</a></sup> Their structures indicate that they form by oligomerisation of 4-vinylcatechol released from caffeic acid units during roasting.<sup><a href=\"#src-s5\">5</a></sup> The chlorogenic acids entry explains where caffeic acid comes from.\n\n**Diketopiperazines.** These cyclic dipeptides form by heat-driven condensation of amino acids and are known for their bitterness in roasted foods such as coffee.<sup><a href=\"#src-s7\">7</a></sup> A 2026 study identified 33 in coffee; cyclo(Pro-Leu), cyclo(Pro-Val) and cyclo(Pro-Tyr) were the most abundant.<sup><a href=\"#src-s7\">7</a></sup> Their levels in coffee infusions lay in the range of reported bitterness thresholds.<sup><a href=\"#src-s7\">7</a></sup>\n\n**Mozambioside.** A furokaurane glucoside, isolated in 2015 from a highly polar, bitter fraction of a roasted coffee brew.<sup><a href=\"#src-s8\">8</a></sup> Its near absence from robusta makes it a proposed marker for arabica.<sup><a href=\"#src-s8\">8</a></sup> A 2025 study characterised seven of its roast products; the main ones had lower bitter thresholds than mozambioside itself.<sup><a href=\"#src-s9\">9</a></sup> The receptors that respond to it, TAS2R43 and TAS2R46, are described under how we taste coffee.\n\n**Caffeine.** Caffeine is heat-stable and bitter.<sup><a href=\"#src-s12\">12</a></sup> Its detection threshold lies far above the recognition thresholds of the roast compounds.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s10\">10</a></sup> In a 2007 study, caffeine’s threshold did not predict how bitter it tasted at higher concentrations.<sup><a href=\"#src-s10\">10</a></sup>\n\n## Formation during roasting\n\nThe 2010 study roasted beans at 260 °C for 60–600 s, and for 240 s at 190–280 °C.<sup><a href=\"#src-s6\">6</a></sup> Lactones formed preferentially at slight to medium roast degrees.<sup><a href=\"#src-s6\">6</a></sup> With more severe roasting they degraded into 4-vinylcatechol oligomers with a harsher bitterness, in line with the sensory results.<sup><a href=\"#src-s6\">6</a></sup>\n\nThe 2020 multi-study analysis covered eight sensory studies; six used a trained panel and two semi-trained coffee professionals.<sup><a href=\"#src-s11\">11</a></sup> In the two studies analysed by correlation, darker colour went with more bitterness.<sup><a href=\"#src-s11\">11</a></sup> Time to first crack and development time were both associated with a consistent rise in bitterness; the authors found colour the stronger predictor.<sup><a href=\"#src-s11\">11</a></sup> The authors link the rise to the progress of the Maillard reaction, including melanoidins and bitter chlorogenic acid fragments.<sup><a href=\"#src-s11\">11</a></sup>\n\nIn a cold brew study cited by the 2020 review, dark roasts gave more bitter taste and bitter aftertaste than the same coffees at medium roast.<sup><a href=\"#src-s12\">12</a></sup> Green arabica held 396–1188 nmol/g of mozambioside, roasted arabica 232 ± 37 nmol/g.<sup><a href=\"#src-s8\">8</a></sup> The authors read this as partial breakdown during roasting.<sup><a href=\"#src-s8\">8</a></sup>\n\n## Extraction\n\nThe 2010 study followed the bitter compounds during percolation with hot water.<sup><a href=\"#src-s6\">6</a></sup> Caffeoylquinic acids and monocaffeoyl quinides were released fast, dicaffeoyl quinides slowly, and the 4-vinylcatechol oligomers were strongly retained by the grounds.<sup><a href=\"#src-s6\">6</a></sup> Mozambioside, by contrast, was 86–98% extracted.<sup><a href=\"#src-s8\">8</a></sup> The chlorogenic acids entry also gives a brewing textbook’s account of how phenylindanes are extracted over a brew.\n\nSensory studies point the same way for strength. The 2020 review cites a cold brew study in which a higher ratio of coffee to water gave a stronger bitter taste and aftertaste.<sup><a href=\"#src-s12\">12</a></sup> It also cites an espresso study in which a high ratio gave more bitter brews than a low one.<sup><a href=\"#src-s12\">12</a></sup> These comparisons changed strength, not extraction alone. Whether extraction yield and strength separately raise bitterness is set out under extraction yield and TDS.\n\n## Taste–taste interactions\n\nThe 1990 study compared caffeine–sucrose and coffee–sucrose mixtures. Both showed hypoadditivity: the mixture tasted about 30–40% less intense overall than the sum of its parts.<sup><a href=\"#src-s13\">13</a></sup> The amount of suppression depended on the strength of the suppressing component and on its taste quality.<sup><a href=\"#src-s13\">13</a></sup> Compared at equal perceived intensity, sucrose suppressed caffeine bitterness and coffee flavour, while caffeine and coffee left sweetness unaffected.<sup><a href=\"#src-s13\">13</a></sup>\n\nAcidity and bitterness move in opposite directions with roast degree.<sup><a href=\"#src-s11\">11</a></sup> That is a change in composition, not proof that one taste masks the other. No study consulted for this entry tested whether sourness suppresses bitterness in coffee.\n\n## Genetics of bitter perception\n\nThe 2018 UK Biobank study used gene variants as instruments in Mendelian randomisation.<sup><a href=\"#src-s15\">15</a></sup> The variant for PROP lies in *TAS2R38* on chromosome 7; those for caffeine and quinine lie in the bitter receptor gene cluster on chromosome 12.<sup><a href=\"#src-s15\">15</a></sup> One standard deviation higher genetically predicted caffeine bitterness went with 0.146 more cups of coffee a day.<sup><a href=\"#src-s15\">15</a></sup> The same step for PROP and quinine went with 0.021 and 0.081 fewer cups.<sup><a href=\"#src-s15\">15</a></sup> Higher caffeine perception also went with higher odds of drinking more than 4 cups a day (odds ratio 1.207).<sup><a href=\"#src-s15\">15</a></sup>\n\nThe 2021 study started from a gap: genome-wide studies of coffee drinking had found variants in caffeine pathways, not in taste perception.<sup><a href=\"#src-s16\">16</a></sup> Using UK and US cohorts, it tested whether taste variants were linked more strongly to black coffee than to coffee with milk or sweetener.<sup><a href=\"#src-s16\">16</a></sup> Instead, genetically inferred caffeine sensitivity was linked more strongly to coffee taste preferences, and the pattern extended to tea and dark chocolate.<sup><a href=\"#src-s16\">16</a></sup> The authors suggest conditioned preferences, in which taste and the body’s response to caffeine are hard to separate.<sup><a href=\"#src-s16\">16</a></sup>\n\nReceptor genes also shape sensitivity to single coffee compounds. In the 2025 mozambioside study, panellists’ sensitivity to its roast products correlated with intact *TAS2R43* gene loci.<sup><a href=\"#src-s9\">9</a></sup> In the 2015 study, people with more fungiform papillae rated bitter test solutions as stronger, but did not differ in coffee bitterness.<sup><a href=\"#src-s14\">14</a></sup> PROP super-tasters rated sourness, bitterness and astringency as stronger in coffee and in standard solutions.<sup><a href=\"#src-s14\">14</a></sup>\n\n## Measuring bitterness\n\nIn cupping, bitterness is recorded as presence, not intensity. The descriptive form’s main-tastes list covers both flavour and aftertaste, and the taster records up to two main tastes.<sup><a href=\"#src-s1\">1</a></sup> The intensity of the other sections is rated on 15-point scales.<sup><a href=\"#src-s1\">1</a></sup> The affective score is a linear function of the eight section scores and the counts of non-uniform and defective cups.<sup><a href=\"#src-s2\">2</a></sup> Bitterness can therefore lower a score only through those sections (our reading).\n\nDescriptive panels measure bitterness as an intensity. The WCR lexicon defines bitter as the basic taste associated with a caffeine solution and gives caffeine solutions as references.<sup><a href=\"#src-s3\">3</a></sup> It names coffee’s marked bitterness, compared with other foods, as one reason why panels assess only a few samples per session.<sup><a href=\"#src-s3\">3</a></sup> The analytical work cited here combined such sensory tests with liquid chromatography and mass spectrometry.<sup><a href=\"#src-s4\">4</a>,<a href=\"#src-s5\">5</a>,<a href=\"#src-s6\">6</a>,<a href=\"#src-s8\">8</a></sup>\n\n## Open questions\n\n- **Shares of total bitterness.** No study consulted for this entry measured what share of a brewed coffee’s bitterness each compound group carries.\n- **Thresholds in coffee.** None of the thresholds consulted was reported for a coffee matrix; only the 2006 lactone study names water as the medium.<sup><a href=\"#src-s4\">4</a></sup>\n- **Mozambioside threshold.** Reported values differ by a factor of about two between 2015 and 2025 (see Contested).<sup><a href=\"#src-s8\">8</a>,<a href=\"#src-s9\">9</a></sup>\n- **Processing and milk.** No study consulted for this entry isolated the effect of washed or natural processing, or of milk, on coffee bitterness.\n- **Taste or caffeine.** Whether genetic links between bitter perception and coffee intake run through taste or through caffeine’s effects is not settled.<sup><a href=\"#src-s15\">15</a>,<a href=\"#src-s16\">16</a></sup>\n\n## Key studies\n\n<div class=\"study\">\n<p><strong>Frank et al., 2006 – The bitter compounds of a coffee brew.</strong> A roasted coffee brew was split into fractions under the guidance of a taste panel. Caffeoyl, feruloyl and dicaffeoyl quinides formed from chlorogenic acids emerged as key bitter compounds, with thresholds in water measured for the first time.</p>\n<p class=\"meta\">European Food Research and Technology 222(5–6), 492–508 · <a href=\"https://doi.org/10.1007/s00217-005-0143-6\">doi:10.1007/s00217-005-0143-6</a></p>\n</div>\n\n<div class=\"study\">\n<p><strong>Blumberg et al., 2010 – Roast, percolation and bitter compounds.</strong> Roasting time and temperature were varied and the bitter compounds followed into the brew. Quinides peaked at light to medium roasts and gave way to harsher 4-vinylcatechol oligomers, which were strongly retained by the grounds during percolation.</p>\n<p class=\"meta\">Journal of Agricultural and Food Chemistry 58(6), 3720–3728 · <a href=\"https://doi.org/10.1021/jf9044606\">doi:10.1021/jf9044606</a></p>\n</div>\n\n<div class=\"study\">\n<p><strong>Lang et al., 2015 – An arabica-specific bitter compound.</strong> Mozambioside was isolated from a bitter fraction of roasted coffee brew and quantified in green and roasted beans. Arabica held large amounts, robusta only traces; roasting reduced it, and brewing extracted almost all of it.</p>\n<p class=\"meta\">Journal of Agricultural and Food Chemistry 63(48), 10492–10499 · <a href=\"https://doi.org/10.1021/acs.jafc.5b04847\">doi:10.1021/acs.jafc.5b04847</a></p>\n</div>\n\n<div class=\"study\">\n<p><strong>Calviño et al., 1990 – Sugar and coffee bitterness.</strong> Caffeine–sucrose and coffee–sucrose mixtures were rated by a taste panel. Sucrose suppressed bitterness and coffee flavour more as its level rose, while coffee did not suppress sweetness at matched intensity.</p>\n<p class=\"meta\">Chemical Senses 15(5), 505–519 · <a href=\"https://doi.org/10.1093/chemse/15.5.505\">doi:10.1093/chemse/15.5.505</a></p>\n</div>\n\n<div class=\"study\">\n<p><strong>Cornelis and van Dam, 2021 – Genes, black coffee and caffeine.</strong> UK and US cohort data tested whether bitter taste genes predict drinking coffee black. Genetically inferred caffeine sensitivity was the stronger link to taste preferences, which the authors suggest may represent conditioned taste preferences.</p>\n<p class=\"meta\">Scientific Reports 11, 23845 · <a href=\"https://doi.org/10.1038/s41598-021-03153-7\">doi:10.1038/s41598-021-03153-7</a></p>\n</div>"
}