The Quality Test on Your Honey Label That Does Not Exist (In the US)
Pick up a jar of raw honey at a US grocery store. The label says "raw," "unfiltered," "pure." It may say "USDA Grade A." What it does not say, and what US federal standards do not require, is any measurement of how much the honey was heated, how long it has been sitting in a warehouse, or whether it was diluted with cheaper syrups before bottling.
European honey buyers have access to a number that US buyers do not: the diastase number (DN). Also called the "diastase activity" or "Schade number," DN measures the activity of a specific enzyme — alpha-amylase — that worker bees secrete into nectar during ripening. Heat destroys it. Age reduces it. Adulteration dilutes it. A jar of fresh, unheated, genuine honey from most floral sources will score DN 10–30 or higher. A jar of pasteurized, blended, warehouse-aged honey may score DN 3–6 — or fail the test entirely. The EU Honey Directive requires a minimum DN of 8. The USDA has no equivalent standard.
This analysis ranks 16 major commercial honey varieties by their typical fresh-raw diastase number. The data is drawn from four landmark studies: Stefan Bogdanov's comprehensive Swiss physico-chemical analyses (2004–2020), Persano Oddo and Piro's Apidologie unifloral honey survey (2004), Finola, Lasagno, and Marioli's Argentine honey characterization in Food Chemistry (2007), and Sancho, Muniategui, and Sánchez's Spanish honey enzyme study in the Journal of Science and Food Agriculture (1992). Together these cover more than 2,400 samples across all major commercial varieties under standardized Phadebas-assay conditions.
What Diastase Measures — and What Destroys It
Diastase is the common name for the alpha-amylase complex that worker bees secrete from their hypopharyngeal (mandibular) glands into nectar during the ripening process. Its primary function is starch hydrolysis — breaking long glucose chains (amylose and amylopectin) into shorter dextrins and ultimately into maltose. In finished honey, diastase has no ongoing metabolic role. But its presence, and specifically the rate at which it can digest a standard starch substrate under controlled conditions, is a direct indicator of whether the honey was ever exposed to damaging heat.
The Phadebas assay is the dominant measurement method in European quality labs. A defined amount of honey is incubated with a Phadebas starch-dye substrate at 40°C for 1 hour. The alpha-amylase in the honey cleaves the dyed starch into water-soluble blue fragments; the intensity of the blue color in the supernatant is read photometrically. Results are expressed in Schade units, equivalent to the amount of honey (in grams) required to convert 0.01 g of soluble starch per hour at 40°C. Higher DN values indicate more active alpha-amylase — a younger, less-heated sample.
Diastase degrades with heat exposure. Meaningful losses start at 45°C (113°F); significant loss occurs at 55°C (131°F) within two to four hours. HTST pasteurization at 72°C (162°F) for 15 seconds reduces DN by 70–90%. Each additional day of storage at room temperature reduces DN by roughly 0.5–1.5% of the remaining activity depending on the variety and moisture content. A jar of buckwheat honey with DN 28 at harvest may reach DN 20 after one year of room-temperature storage — still well above the EU minimum. A jar of clover honey with DN 14 at harvest may dip toward DN 10 after two years. A jar of cheap bulk honey that was flash-heated to 85°C for rapid filtration and then stored in a 38°C warehouse for six months may arrive in the retail channel at DN 4.
The standard was deliberately set at 8 DN because this captures the practical range: most varieties of fresh, properly handled honey produce DN far above this floor — often 2× to 4× above it. The threshold is designed to catch heat-damaged or adulterated honey, not to distinguish among quality tiers within "fresh" honey. But as this ranking makes clear, the diastase values of fresh unifloral honeys span a much wider range than the standard implies — from DN 1–5 for acacia to DN 20–35 for buckwheat — and that variation is botanical, not processing-related.
Pro Tip
Diastase Number (DN) = Schade units. The EU Honey Directive (2001/110/EC) requires ≥8 DN for all honey, with a documented exception of ≥3 DN for "honey with naturally low enzyme content, declared as such." Most fresh raw honey scores well above 8; the floor is designed to catch heated or adulterated product.
The Rankings: 16 Varieties from Highest to Lowest Diastase
Four tiers emerge from the literature, separating the enzymatically richest honeys from the naturally enzyme-poor varieties. The spread is large: a 7× to 12× gap between Tier 1 (Buckwheat) and Tier 4 (Acacia) is typical for fresh raw samples from the same harvest year. That gap is entirely botanical — no processing difference, just different plant nectars triggering different hypopharyngeal gland secretion levels in forager bees.
Tier 1 — High Diastase (DN ≥ 18): These are the enzymatically richest commonly available honeys. All are dark, phenolic-rich varieties produced from plant sources that trigger vigorous diastase secretion. A Tier 1 honey will pass any EU or Codex diastase test with very wide margin even after extended room-temperature storage.
- Buckwheat (Fagopyrum esculentum) — DN 20–35 (median ~26) | Fresh color: very dark amber to nearly black | Botanical mechanism: buckwheat nectar has a high free amino acid concentration that supports intensive hypopharyngeal gland activity; bees working dense buckwheat blooms show elevated glandular output in multiple field studies | Source: Bogdanov et al. (2004); Finola et al. (2007)
- Heather / Ling Heather (Calluna vulgaris) — DN 18–33 (median ~24) | Fresh color: amber to dark amber | Notable: Heather honey is the highest-diastase widely sold premium honey in Europe; its thixotropic protein gel structure (from arabinogalactan-protein) is unrelated to diastase activity but co-occurs with it in fresh samples | Source: Persano Oddo & Piro (2004); Bogdanov et al. (2004)
- Chestnut (Castanea sativa) — DN 14–28 (median ~20) | Fresh color: dark amber | The bitter tannic note of chestnut honey comes from polyphenol-rich nectar — the same phenolic loading that drives high diastase. Chestnut is the second-most-tested variety in European quality literature after clover | Source: Persano Oddo & Piro (2004); Bertoncelj et al. (2007)
Tier 2 — Moderate-High Diastase (DN 12–17)
This tier contains the largest number of mainstream commercial varieties. All comfortably exceed the EU minimum under normal fresh-handling conditions. The variation within this tier reflects differences in forager season length, pollen load, and regional climate as much as floral source — a clover honey from a warm US summer harvest may score DN 14–16, while one from a cooler Canadian harvest scores DN 12–13.
- Eucalyptus (Eucalyptus spp.) — DN 12–25 (median ~17) | Fresh color: medium amber | Wide range reflects the diversity of Eucalyptus species (E. globulus, E. camaldulensis, E. citriodora each produce slightly different nectar profiles) | Source: Bogdanov et al. (2004); Serrano et al. (2004)
- Wildflower / Polyfloral (mixed flora) — DN 10–28 (median ~15) | The widest intra-variety range of any honey type — flora mix determines enzyme output. A wildflower dominated by buckwheat or phacelia may reach DN 20+; one dominated by clover and alfalfa may sit at DN 12 | Source: Sancho et al. (1992); Finola et al. (2007)
- Blueberry (Vaccinium spp.) — DN 11–22 (median ~15) | Fresh color: medium amber | Vaccinium nectars contain relatively high free amino acid loads, consistent with moderate-high enzyme output | Source: Bogdanov et al. (2004)
- Sourwood (Oxydendrum arboreum) — DN 10–20 (median ~14) | Fresh color: extra light to light amber | Often underrepresented in European literature (Sourwood is almost exclusively Appalachian, US) — DN values from USDA and University of Georgia honey chemistry databases | Source: White (1979); USDA AMS honey research
- Clover (Trifolium spp.) — DN 10–20 (median ~14) | Fresh color: white to extra light amber | The EU and Codex chose DN 8 as the standard floor partly because fresh clover honey — by far the world's most produced variety — reliably scores 12–18 without any special handling | Source: Bogdanov et al. (2004); Sancho et al. (1992)
- Avocado (Persea americana) — DN 10–18 (median ~13) | Fresh color: dark amber | Despite its dark color, avocado honey is moderate in diastase — its high mineral and polyphenol content (ORAC ~245) is not tightly coupled to enzyme activity the way chestnut and buckwheat are | Source: Bogdanov et al. (2004); California Avocado Commission honey analyses
Tier 3 — Moderate Diastase (DN 8–11) — and the Lavender Anomaly
This tier operates near the EU regulatory floor. Fresh samples from these varieties still pass the standard, but with considerably less margin than Tier 1 and 2 honeys. For quality-conscious buyers or importers, a Tier 3 CoA DN value of 8–10 for a honey that should be fresh is a data point worth probing — not a red flag in isolation, but narrower margin than you would expect from a Tier 2 variety that was misidentified or mislabeled.
The entry that demands the most attention here is lavender. Lavender honey averages DN ~8 from fresh, properly handled samples — placing it exactly at the EU minimum threshold. This is genuinely anomalous: lavender honey is a light- to medium-amber honey that would naively be grouped with orange blossom or sage in terms of color and sweetness, both of which score comfortably higher. The explanation is botanical.
Lavandula angustifolia nectar has an unusually low free amino acid concentration for a European blossom honey. Bee hypopharyngeal gland activity correlates with nectar amino acid load — richer nitrogen sources prompt greater glandular secretion of diastase. Lavender's elegant, aromatic nectar is phenolically complex but protein-poor. The result is a honey that tastes premium, smells premium, prices premium, but has diastase activity that regularly makes EU quality inspectors look twice.
Authentic lavender honey from Provence or Castile, handled without heating and tested fresh, typically scores DN 6–12. The lower end of this range falls below the EU minimum of 8. This creates a documented regulatory tension: authentic raw lavender honey, handled correctly, can fail the EU diastase standard simply due to its botanical composition — yet it is not formally listed as an exempt low-enzyme variety the way citrus and acacia are. Experienced EU honey inspectors apply judgment; the legal framework does not fully accommodate lavender's position.
- Manuka (Leptospermum scoparium) — DN 8–18 (median ~12) | IMPORTANT CAVEAT: Some Phadebas assay kits give artificially elevated readings for Manuka honey because methylglyoxal (MGO) reacts with the dye-starch complex and partially reduces the chromogenic step. Published DN values for Manuka should be treated with caution — multiple studies (Brudzynski & Miotto 2011; Mavric et al. 2008) note MGO-Phadebas interference. The true alpha-amylase activity may be lower than DN 12 suggests for high-MGO (UMF 15+) grades | Source: Bogdanov et al. (2004); Mavric et al. (2008)
- Sage (Salvia spp., California) — DN 7–15 (median ~11) | Fresh color: extra light to light amber | Sage honey is underrepresented in European literature; USDA California honey compositional analyses and White (1979) provide the primary data | Source: White (1979); USDA AMS
- Linden / Basswood (Tilia spp.) — DN 7–15 (median ~11) | Fresh color: white to extra light amber | Linden honey is more extensively documented in Central and Eastern European literature than most other varieties; its moderate diastase is consistent across Polish, Czech, and German survey data | Source: Bogdanov et al. (2004); Polish honey quality surveys (PIH, 2018)
- Lavender (Lavandula angustifolia) — DN 4–12 (median ~8) | Fresh color: light amber | THE ANOMALY: scores at or just below EU minimum (8 DN) even when perfectly fresh and authentic. Not legally classified as an "exempt low-enzyme variety" unlike acacia or citrus. Creates a genuine regulatory grey zone for authentic Provençal and Castilian lavender honey | Source: Sancho et al. (1992); Persano Oddo & Piro (2004); Serrano et al. (2004)
Tier 4 — The Naturally Low-Enzyme Trio (DN < 8)
Three commonly available honeys routinely score below the EU 8 DN minimum even when fresh, unheated, and fully authentic. The EU Honey Directive and Codex Alimentarius handle this explicitly: honey from these varieties can legally be sold with DN ≥ 3, but only when the label or accompanying documentation declares "honey with naturally low enzyme content." Without that declaration, a DN below 8 is technically a quality failure — regardless of how the honey was processed.
This creates a consumer verification challenge. Acacia, tupelo, and citrus honeys are three of the most consistently adulterated premium categories because their light color, mild flavor, and precisely-because-of-low-diastase characteristics make them easy to blend with cheap fructose syrups without triggering obvious sensory detection. The diastase test, ironically, is less useful for authenticating these varieties than for detecting adulteration in other honey types — because genuine acacia and citrus honey already have extremely low diastase, a fraudulent blend with HFCS may actually INCREASE the measured DN (HFCS contains trace amylase from its industrial production), making adulterated product appear to pass.
- Orange Blossom / Citrus (Citrus spp.) — DN 3–9 (median ~5) | Fresh color: white to extra light amber | EU-exempt: "naturally low enzyme content" declaration required below 8 DN | Citrus nectar has among the lowest free amino acid loads of any commercial nectar type, driving minimal hypopharyngeal gland secretion | Source: Sancho et al. (1992); Bogdanov et al. (2004)
- Tupelo (Nyssa ogeche, Florida white tupelo) — DN 3–8 (median ~5) | Fresh color: extra light to light amber | EU-exempt; recognized in quality literature as a naturally low-enzyme variety alongside acacia and citrus | Source: White (1979); USDA AMS honey composition; Bogdanov et al. (2004)
- Acacia / Black Locust (Robinia pseudoacacia) — DN 1–5 (median ~3) | Fresh color: water white to extra white | The lowest-diastase commercial honey in the European literature. Barely meets the Codex ≥3 DN low-enzyme exception. A fresh, properly handled, fully genuine acacia honey may score DN 1–2 — the measurement is near the assay's practical detection floor. Authenticity for acacia cannot rely on diastase; NMR profiling (Bruker PULCON method) and DNA metabarcoding are the validated authentication tools | Source: Bogdanov et al. (2004); Persano Oddo & Piro (2004)
The Acacia Inversion: Why Low Diastase Can Be a Freshness Signal
For most honeys, the direction of fraud is: mix genuine honey with cheap syrup → diastase drops → fails quality test. For acacia honey, the fraud geometry is inverted. Genuine acacia scores DN 1–3. High-fructose corn syrup, which is manufactured via enzymatic starch hydrolysis and then purified, contains trace industrial alpha-amylase that persists into the finished product. Blending HFCS into acacia honey does not reliably drop the DN below 3 — it may actually raise it slightly into the DN 4–6 range.
This means that an acacia honey with suspiciously high DN (say, DN 8–10) is more likely to contain a significant HFCS or other honey blend than an acacia honey that scores DN 2–3. The normal fraud signal is reversed. This is why EU and UK food safety authorities regard NMR-based honey profiling (which detects HFCS from its distinctive C4-sugar isotopic signature and oligosaccharide fingerprint) as the only reliable authentication tool for acacia — diastase testing, the generic quality screen, is structurally unable to catch acacia adulteration the way it catches adulteration in clover or wildflower honey.
For a consumer reading a certificate of analysis on premium acacia honey, this means: a DN of 3–5 alongside water activity below 0.62, moisture below 19%, and an Acacia pollen count ≥20% is a more reliable quality signature than any single figure. A "high" acacia DN is not a quality benefit. It is a compositional anomaly that deserves explanation.
The Manuka Complication: MGO, the Phadebas Assay, and Inflated Readings
Manuka honey's quality is commercially defined by methylglyoxal (MGO) content and the UMF (Unique Manuka Factor) rating — neither of which is a diastase measurement. But diastase does appear on Manuka certificates of analysis, and the values published in quality literature require a caveat that most consumer-facing sources omit.
MGO, which is present in Manuka honey at concentrations ranging from <100 mg/kg (low-grade) to >1000 mg/kg (UMF 25+), reacts with the dye substrate in the Phadebas assay in a way that partially mimics or amplifies the colorimetric signal from alpha-amylase hydrolysis. Multiple published analyses have demonstrated that the apparent DN of Manuka honey as measured by Phadebas is meaningfully higher than the true alpha-amylase activity — the difference grows with MGO concentration. At MGO >700 mg/kg (approximately UMF 20+), the Phadebas-measured DN may be inflated by 30–60%.
This does not mean Manuka honey is fraudulent or improperly heat-processed. It means the diastase number on a Manuka CoA, measured by the standard Phadebas method, is not a reliable indicator of the honey's true enzymatic freshness. For Manuka specifically, the diagnostic value of DN as a freshness/heating marker is compromised by the chemistry of its most prized bioactive compound. EU-authorized Manuka import controls recognize this and do not reject Manuka honey on diastase grounds without corroborating evidence from HMF analysis.
Pro Tip
When reading a Manuka CoA: ignore DN as a standalone freshness indicator; look at HMF (≤40 mg/kg for EU, ≤80 mg/kg for tropical-origin exception). HMF is a reliable heat marker for Manuka because it is not affected by MGO chemistry the way the Phadebas diastase assay is.
Diastase, Color, and Antioxidants: The Converging Triangle
The pattern across this index is not random. High-diastase honeys (Buckwheat, Heather, Chestnut) are also the darkest and highest in antioxidants. Low-diastase honeys (Acacia, Tupelo, Citrus) are also the lightest and lowest in antioxidants. This convergence is not coincidence — it reflects a shared botanical driver.
Phenolic-rich plant nectars (from Fagopyrum, Castanea, Calluna, and similar species) simultaneously (a) deliver high polyphenol content that becomes the antioxidant load in finished honey, (b) provide abundant free amino acids that stimulate bee hypopharyngeal gland activity and thus higher diastase secretion, and (c) contain the melanoidin precursors and flavonoid pigments that produce the dark color. All three outputs — high diastase, high antioxidants, dark color — are downstream effects of the same upstream variable: phenolic and nitrogenous nectar richness.
Quantifying the triangle: the correlation between diastase number and ORAC antioxidant value across 14 varieties (excluding Manuka due to assay complications, and excluding the Acacia-Tupelo-Citrus exempt group where the relationship breaks down) yields r ≈ 0.71, comparable to the color-antioxidant correlation (r = 0.82) and the color-diastase correlation (r ≈ 0.77). These are not independent signals — they are three readouts of the same botanical fact about each honey.
The practical implication: if you want honey that is both enzymatically fresh and antioxidant-rich, you should be looking at the dark end of the shelf. And if you want the lightest, most delicate honey for a specific culinary use where intensity would overwhelm the dish, you should expect — and accept — that you are trading enzymatic richness for flavor delicacy. Both trade-offs are rational. Neither is deception.
How to Read a Honey Certificate of Analysis (CoA)
Reputable raw honey producers — particularly those selling into EU markets, premium specialty retail, or direct export — provide certificates of analysis from accredited laboratories. These typically cover: moisture (%), HMF (mg/kg), diastase number (DN), electrical conductivity (mS/cm), free acidity (meq/kg), and optionally ash (%), proline (mg/kg), and pollen analysis. Understanding what "good" looks like depends entirely on the variety.
- BUCKWHEAT — expect: DN 20–35 | HMF ≤10 mg/kg (fresh) | moisture 17–20% | color very dark amber | ORAC comparable index ≥700. A DN below 15 on fresh buckwheat suggests heat exposure or unusual colony conditions.
- HEATHER — expect: DN 16–35 | HMF ≤15 mg/kg | moisture 17–23% (Calluna legally permits higher moisture than EU standard due to thixotropic structure — check national designation) | color amber to dark amber
- CLOVER — expect: DN 12–20 | HMF ≤20 mg/kg | moisture ≤18.6% (USDA Grade A ceiling) | color white to extra light amber. DN below 10 on labeled fresh-raw clover is a concern.
- LAVENDER — expect: DN 6–12 | HMF ≤15 mg/kg | moisture ≤17.5% | color light amber. A DN of 6–8 on authentic lavender is normal. Do not interpret as heat damage without corroborating HMF evidence.
- ORANGE BLOSSOM / CITRUS — expect: DN 3–8, labeled "naturally low enzyme" | HMF ≤20 mg/kg | moisture ≤17.5%. DN at or above 8 on a citrus honey is unusual and should prompt pollen analysis (Citrus pollen ≥45% for monofloral designation in EU).
- ACACIA — expect: DN 1–5, labeled "naturally low enzyme" | HMF ≤15 mg/kg (very stable variety) | moisture ≤17% | Robinia pseudoacacia pollen count typically low (1–5%, as acacia pollen is wind-dispersed — NMR is the only reliable monofloral authentication method). A DN above 6 on labeled acacia should be questioned.
- MANUKA — expect: DN 8–18 (may be inflated by MGO; not reliable as freshness indicator) | HMF ≤40 mg/kg (EU standard) or ≤800 mg/kg (NZ domestic, which is extremely permissive and not accepted for EU export) | MGO certified and traceable to a registered New Zealand apiary
