日本語版

Topic hub — 2026-09-19

African bitter leaves

Bitterleaf (Vernonia amygdalina) is one of the few foods with bitterness in its name. Its bitterness has been traced to a molecule: in a sensory test with a paper disc, the minimum amount of vernonioside A1 to produce a bitter taste was 7 µg, and the intensity was almost equal to that of quinine sulfate. That puts it level with the reference compound for bitterness.

Meanwhile, the process that makes the leaf edible — boiling, soaking through several changes of water, macerating by hand — has its endpoint written in the sources only as "to a desirable level." The molecule is identified; how far to take it out is not specified. This page lays that difference in precision out as it stands.

What this page establishes

• Vernonioside A1, the bitter principle of bitterleaf, produced bitterness from 7 µg in a paper-disc sensory test, at an intensity reported as almost equal to quinine sulfate.
• Vernonioside B1, which differs only in the presence of a hydroxyl at C(16), showed no bitterness even at 200 µg in the same test.
• The process for removing the bitterness is recorded, but its endpoint is written as "to a desirable level," with no quantitative value.
• Total phenolic content of the washed powder was 0.445 mg gallic acid equivalent per g against 24.08 mg/g for the aqueous extract. That washing removes compounds is shown.
• For African nightshade leaves the bitterness is attributed to anti-nutrients such as solanaceous glycoalkaloids, and blanching at 95 degC for 5 minutes and similar treatments are described as lowering it. No quantitative sensory figure is cited.

One case taken down to the molecule

From the dried leaves of bitterleaf, the main ingredient of Cameroonian ndolé, the steroidal glycoside vernonioside A1 has been isolated. In a sensory test using a paper disc of 1 cm² and 0.2 mm thickness, the minimum amount producing a bitter taste was determined as 7 µg, and its intensity was almost equal to that of quinine sulfate.

The same study goes as far as the structural condition for bitterness. Vernonioside B1 differs from A1 only in the presence of a hydroxyl at C(16), and showed no bitterness at 200 µg in the same test. From that difference the authors suggest hydroxylation at C(16) may be involved in producing bitterness. That is the authors' own tentative wording, not an experiment testing the whole B series.

Few entries in this collection narrow the presence or absence of bitterness down to the position of a single hydroxyl. That precision is what makes the gap with the next section so wide.

Removing it is part of the cooking

The leaf is not eaten as it comes. It is boiled, or soaked through several changes of water. Freshly harvested leaves are macerated with cold or hot water to reduce the bitterness to a desirable level, as the source puts it.

The step appears inside the food records as a matter of course. In Igbo food documentation, onugbu (Vernonia) is recorded as bitter and as usually washed before use. How it is washed, where it stops, and what bitterness remains are not written there.

What comes out has been measured once, on the compositional side. The washed powder, which has been through the customary repeated trituration with water, had a total phenolic content of 0.445 mg gallic acid equivalent per g against 24.08 mg/g for the aqueous extract. But that contrast is between two forms prepared differently, and the figure for the unwashed powder itself is not given in the text.

So it is shown that washing removes compounds, and not shown how far to remove them. Within this page, what sets the endpoint is the cook's own tongue.

The nightshade side

The other bundle here is African nightshade leaves. They are boiled and eaten as an accompaniment to the staple maize meal, and are sometimes fermented with milk.

The bitterness is attributed to anti-nutrients such as solanaceous glycoalkaloids. No correspondence between individual compounds and bitterness thresholds is given. The precision is not that of the bitterleaf case.

The process is recorded in the same shape: traditional preparation including blanching (for example 95 degC for 5 minutes), cooking and fermentation are described as lowering the bitterness and improving palatability. No quantitative sensory value for the reduction is cited.

The broad-leafed Solanum scabrum (mnavu) is named as a popular leafy vegetable in West, Central and East Africa.

For the berries rather than the leaves, a different yardstick applies. The berries carry structurally diverse glycoalkaloids with large reported variation in content and composition, and the source states that safe consumption depends on chemical screening. Drawing the line by analysis rather than by tongue is the same arrangement as in theseeds hub.

Where bitterness was not measured

This bundle also contains entries that hold no bitterness claim of their own. Rather than drop them, the absence is written down.

Utazi is the local name for Gongronema latifolium, used as a flavouring in fish, chicken and goat pepper soups and in nkwobi and abacha. Neither its bitterness intensity, nor how much is used, nor how often, is measured in these sources.

The kola nut is plainer still. Caffeine and theobromine are recorded as important constituents, yet there is not one claim about the bitterness of the kola nut itself. Caffeine and theobromine are treated as carriers of bitterness on the coffeeand cacao hubs instead.

What is recorded thickly is use and trade rather than taste. It is chewed in West Africa, offered at naming ceremonies and weddings as a mark of hospitality and prosperity, and Cola nitida was for centuries a principal commodity of the north-south trade linking the forest belt with the savanna and the southern Sahara.

On the Igbo side, onugbu soup is given as an example of the regional cuisine. The source is not a primary ethnography written for cultural description.

Questions people ask

How bitter is bitterleaf?

On the compound side there is a number: vernonioside A1 produced bitterness from 7 µg in a paper-disc test, at an intensity almost equal to quinine sulfate. But that concerns an isolated compound; nothing on this page measures the bitterness intensity of the leaf itself or of a dish.

How much washing is enough?

It is not settled. The sources say to reduce it to a desirable level, and record neither the method, the endpoint, nor the bitterness remaining. On the compositional side, the washed powder's total phenolic content of 0.445 mg gallic acid equivalent per g is lower than the aqueous extract's 24.08 mg/g. That something comes out is shown; how much should come out is not.

Are bitter leaves all handled the same way?

No. Bitterleaf is washed to take the bitterness out before use; African nightshade goes through blanching or fermentation; utazi is used in small amounts as a flavouring, with no removal step recorded at all. Among leaves called bitter, some have it taken out and some are used for it.

Is the kola nut bitter?

Within this page there is not one claim about the bitterness of the kola nut itself. What is recorded is that it contains caffeine and theobromine, that it is chewed, that it is offered at ceremonies, and that it was a principal trade commodity. Its bitterness is not absent; it is unmeasured.

What is not known

• The 7 µg figure for vernonioside A1 comes from a tentative sensory test on an extract from a single lot of dried leaves, with no panel size or statistics reported.
• The suggestion that hydroxylation at C(16) separates bitter from non-bitter is the authors' tentative wording, not an experiment testing the non-bitterness of the whole B series.
• No source states the endpoint of the washing that removes bitterness. What remains after how much washing is unknown.
• The contrast between washed powder (0.445 mg/g) and aqueous extract (24.08 mg/g) is between differently prepared forms, and the figure for unwashed powder itself is not given in the text.
• For African nightshade, no correspondence is given between the individual compounds held responsible and bitterness thresholds, and no quantitative sensory value for the reduction is cited.
• Utazi and the kola nut have no bitterness measurement. They are placed in this bundle while holding no bitterness claim of their own.

Read in the encyclopedia

• Vernonioside A1 — the steroidal glycoside bitter from 7 µg, level with quinine sulfate
• Bitterleaf debittering — boiling, soaking, macerating; the endpoint is unwritten
• African nightshade leaves — a leafy vegetable that goes through blanching and fermentation
• Solanum scabrum (mnavu) — leaves eaten, berries drawn on by analysis
• Utazi leaves — used as a flavouring; bitterness unmeasured
• Kola nut — contains caffeine and theobromine, holds no bitterness claim
• Igbo food culture — where onugbu soup is recorded as an example of the regional cuisine

Go deeper in the research notes

• A world map of debittering — techniques for taking bitterness out, by region and by process
• Eating bitter leaves — ways of eating leafy greens around the world
• Aku-nuki and unmeasured bitterness — what to do with a process whose endpoint is never written
• Bitter in the name — when a food's own name says bitter

Source ledger

The primary sources and limitations for the claims on this page can be checked from the source links inside each card. The full list is on the Japanesesource ledger.