日本語版

Research note 053 — 2026-09-07

Rereading Kitchen Prep Through Measured Bitterness

Faced with a bitter ingredient, the kitchen offers a lineup of steps: 「塩もみする」 (in translation: "salt-rubbing"), 「湯通しする」 (in translation: "blanching"), 「ワタを取る」 (in translation: "removing the pith and seeds"), 「砂糖や出汁を足す」 (in translation: "adding sugar or dashi stock"). Each of these steps is often described as an operation that draws out or weakens bitterness. But when these steps are checked one by one against research that directly measured bitterness itself, how well each one works turns out to differ considerably from step to step. Steps whose bitterness score moved, steps whose score did not move, and steps that are widely spoken of as prep but for which no material measuring bitterness itself yet exists — talk about these three without separating them, and it all gets pressed into a single story: 「下ごしらえをすれば苦味は抜ける」 (in translation: "do the prep, and the bitterness comes out"). Here, sensory evaluations and chemical analyses covering bitter gourd, cucumber, leafy greens, olives, and coffee are laid out, sorted into what worked, what did not work, and what has not yet been measured.

What this page establishes

• One study found that soaking bitter gourd in 3.5% brine as a prep step barely changed its bitterness score.
• One study reported that adding a small amount of sugar or salt to spinach and kale purees lowered bitterness intensity.
• Katsuobushi significantly lowered bitterness when the flakes themselves were added. In an experiment adding it to bitter gourd extract, the stock (extract liquid) also lowered bitterness, but not as much as the flakes, the report states.
• How much aku removal for wild greens actually reduces bitterness itself has not yet been measured.

Salt doesn't work as evenly as you'd think

Soaking bitter gourd in brine, as a prep step, is widely practiced as an operation to draw out bitterness. But in an experiment where fruit pieces were soaked in a 3.5% (w/v) NaCl solution at 25°C for 60 minutes before pressing the juice, the bitterness score after treatment was almost unchanged from untreated fruit. The evaluation was by a trained panel of 10.

How chloride ions work also differs depending on the metal ion they are paired with. In an experiment on brined olives where part of the NaCl was replaced with potassium chloride (5-15 g/L), the more that was replaced, the more the bitterness score fell in a straight line. This experiment was evaluated by a consumer panel of 100 people, each with 3-10 years' experience of regularly eating green olives, who rated 14 different formulations.

In the same series of experiments, however, adding calcium chloride and magnesium chloride (each 0-10 g/L) worked the other way, as factors pushing bitterness up. In the equation predicting bitterness, the coefficient for calcium chloride is +0.426 and the coefficient for magnesium chloride is +0.380 — both acting in the positive direction.

In a separate experiment on Brussels sprouts and cauliflower, no average change in the bitterness score from adding salt was seen. However, there was an individual difference: participants who perceived bitterness more strongly without salt showed a larger drop in bitterness after it was added, and the strength of that relationship is recorded as R2=0.211. Inside an experiment that, looking only at the average, 「効かなかった」 (in translation: "did not work"), there turns out to have been a layer of people for whom it did work.

Sugar lowered the bitterness of leafy greens

Separately from salt, there is also a prep step of adding sugar. In an experiment adding 0, 0.6, 1.2, or 1.8% sugar, or 0 or 0.2% salt, to spinach and kale purees, bitterness intensity fell for both vegetables, and other sensory attributes such as texture and aroma did not change. The evaluation was by a trained panel of 9 adults.

It was the solid, not the extract, that worked

In bitter gourd cooking, adding katsuobushi is a known prep step. Adding katsuobushi is reported to have lowered bitterness in both raw and cooked bitter gourd.

The same study also tried adding katsuobushi stock (extract liquid) to bitter gourd extract. Bitterness dropped significantly in this condition too (p<0.05), but not as much as when the flakes themselves were added. The authors think bitter compounds are adsorbed onto the solid.

Fermentation and inclusion agents lowered bitterness itself

In lactic-acid fermentation of bitter gourd juice, a significant difference in bitterness arose compared with the unfermented sample, and many panelists perceived a drop in bitterness in the fermented sample.

Among additives, adding 5% gum arabic to bitter gourd juice lowered the bitterness score from a control of 10.7 +/- 1.7 to 8.6 +/- 0.6, and at 15% it fell further, to 7.3 +/- 1.4. The evaluation was by the same trained panel of 10.

For beta-cyclodextrin, which encloses bitter molecules in a ring-shaped structure, adding 0.25% to bitter gourd extract reduced bitterness by about 33%, and at 0.75% or more it fell by more than 90%. This is a different way of working from salt or sugar, which weaken bitterness on the taste-receptor side — instead, it encloses the bitter molecule itself, making it harder for it to touch the tongue.

Removing the pith and seeds cut away where bitterness concentrated

In cucumber, juice pressed from the stem end of a bitter cultivar was found to contain cucurbitacin C at 2.0±3.5 mg/L (n=4), a higher concentration than in other parts of the fruit. In fruit of non-bitter cultivated varieties, this compound itself was not detected. Cutting off the end near the stem cap can be explained as an operation that removes the very part where the bitter component is concentrated. As for the practice of rubbing the two cut surfaces together, no material verifying it through direct measurement of bitterness was found.

The reason this compound is not detected in non-bitter cultivars is explained as a change that occurred on the gene side during domestication. A mutation in the gene that controls the bitterness biosynthesis pathway was selected for and fixed, lowering its expression in the fruit, while a separate gene controlling the same pathway keeps its expression in the leaves.

Changing the water temperature doesn't always move bitterness

The strength of the heat applied does not always move bitterness. In drip coffee, under conditions that held brew strength and extraction yield at target values, changing the brewing water temperature among 87°C, 90°C, and 93°C showed no statistically significant difference in the trained panel's bitterness score. Many attributes, bitterness included, moved not with the water temperature itself but with the concentration of dissolved compounds.

Aku removal for wild greens still has no direct measurement

Every prep step covered so far is based on material that measured either the bitterness score or the content of bitter substances. Aku removal for wild greens, on the other hand — soaking warabi or zenmai in water with ash or baking soda, or boiling bamboo shoots together with rice bran or rice-rinsing water — turned up no material that directly measured bitterness itself. In the prep ledger, too, these lines sit as unverified rows that cannot be classified as bitterness decreasing, staying the same, or increasing. The phrases "removing the aku" and "removing the egumi" are heard often, but whether what is moving there is bitterness has not, at this point, been measured.

Honest notes

• The bitter-gourd brine soak and the gum arabic addition are both results from the same 1 study and the same trained panel of 10, under conditions intended for juice processing. These conditions differ from the home operation of rubbing with salt and wringing out the liquid.
• The effects of potassium, calcium, and magnesium on olives are results from a mixture experimental design using a single cultivar and a single processing style, and the optimization metric has not reached a fully optimal solution.
• The katsuobushi effect was checked in full text (2026-09-26). The panels were small: 8 people for chanpuru and raw bitter gourd, 5 for the bitter gourd extract experiment. Chanpuru is a dish with several ingredients, so it is not a comparison isolating katsuobushi alone.
• The effects of sugar and salt on leafy greens are based on a search-engine summary, and figures specific to spinach or kale alone have not been separated out.

Frequently asked

Does salt-rubbing remove bitterness from bitter gourd?

The answer here is that 「塩もみをすれば苦味が抜ける」 (in translation: "salt-rubbing removes the bitterness") cannot be said as a single blanket statement. In an evaluation by a trained panel of 10, the bitterness score of fruit pieces soaked in 3.5% brine at 25 degrees for 60 minutes was almost no different from unsoaked fruit. Under the same treatment, sourness was significantly lower, suggesting that what soaking in brine moves is sourness rather than bitterness. The cases where the bitterness score itself is reported to have dropped are experiments on lactic-acid-fermented bitter gourd juice, and on juice with gum arabic or beta-cyclodextrin added — so even under the same 「下ごしらえ」 (in translation: "prep"), how well it works differs by method.

Is aku removal for wild greens meant to remove bitterness?

Using ash or baking soda, boiling with rice bran added. Material directly measuring how much these procedures, handed down for wild greens, actually move bitterness has not yet been reached. For other ingredients where bitterness has been directly measured, such as bitter gourd, olives, and leafy greens, prep is known to work quite differently from one ingredient to the next: some cases, like sugar or fermentation, lower bitterness, while others, like brine, barely move it at all. Whether what actually moves during aku removal is bitterness has not, at this point, been measured.

What is not known

• No direct measurement comparing the bitterness intensity of bitter gourd's flesh separately from its wata (the white part around the seeds; pith and seeds) is found in this material.
• The reason itself that calcium chloride and magnesium chloride push up olives' bitterness is shown only as a coefficient in the model equation; the mechanism at the compound level has not yet been measured.
• The effect that differences in coffee grind size or brew time have on bitterness falls outside the scope of the water-temperature comparison this material covers.
• No material measuring bitterness under conditions other than juice processing — for the common home cooking operation of rubbing with salt and wringing out the liquid itself — has been found.

Related: bitter melon /cucumber /katsuobushi (smoked-dried bonito) /table olives /coffee brew /the prep ledger /Research note 007: Did Salt-Rubbing Ever Remove Bitter Gourd's Bitterness? /Research note: A World Map of Debittering