← All posts

Why some people turn red drinking — flush response and intoxication

Some people flush after a single drink; others never change colour no matter how much they have. The difference isn't willpower or tolerance — it's an inborn constitution. In East Asia, including Japan, the flush-prone type is very common: roughly 40% of adults.[2][3] Here's why it happens, and how it relates to getting drunk — grounded in Japan's MHLW.

The flush is acetaldehyde

The alcohol (ethanol) you drink is broken down mostly in the liver, in two steps:

  1. Ethanol → acetaldehyde (via enzymes such as ADH)
  2. Acetaldehyde → acetate (via the enzyme ALDH2), then on to water and carbon dioxide.

The acetaldehyde in the middle is toxic — it's behind the facial flushing, racing heart, headache, and nausea. When ALDH2 works well, acetaldehyde is cleared quickly to harmless acetate. When ALDH2 is weak, acetaldehyde builds up, and you flush. That's the "flush response" (medically, the flushing reaction).[1]

Roughly three types

ALDH2 activity comes in three inborn types:

TypeALDH2 activityTendency
Fully activeWorks fullyRarely flushes
Low activitySharply reducedFlushes on a little
InactiveBarely worksStrong reaction on even a small amount

Among Japanese, roughly half are fully active, about 40% low-activity, and about 5% inactive.[2][3] The low-activity type is often assumed to have "half" the activity, but because ALDH2 works as a group of four molecules, one weak copy drags the whole enzyme down to about a sixteenth of full activity — which is why it comes out as "can drink a little, but flushes quickly."[2]

Fully active rarely flushes
Low activity flushes on a little
Inactive strong reaction on a little
The three rough ALDH2 types (approximate share among Japanese: fully active ~half, low-activity ~40%, inactive ~5%)

It's inborn — practice doesn't change it

This type is set from birth and doesn't change with practice. People sometimes say you can "train yourself to drink," but what changes is tolerance (getting used to it) — not ALDH2's activity, and not how readily acetaldehyde accumulates. Even if you flush less visibly than before, what's happening inside is the same.[2]

Same amount, different intoxication

This is the key point. Drink the same amount, reach the same blood alcohol concentration, and how strongly it hits still varies by constitution. Flush-prone people feel the effects of acetaldehyde earlier and more strongly, and report feeling more intoxicated at the same concentration.[4] "Same pace as everyone else" doesn't mean "same buzz."

One important note: flush response doesn't change how fast the alcohol itself clears. What's at work is on the acetaldehyde side; the blood-alcohol curve and the time to clear don't vary much with flush response. What differs is "how a given concentration feels." We cover clearing time itself in the article on how long alcohol takes to clear.

Flush and hangovers

Being flush-prone also relates to how readily you get a hangover. Research shows low-activity people tend to get a hangover on smaller amounts,[5] and Japan's MHLW notes that people with the inactive, flush-showing ALDH2 are prone to hangovers.[1] What sets hangover severity, and how it can be estimated, is in the hangover-forecast article.

How Nomicho reflects it

In Nomicho you set your flush response in your profile (doesn't flush / flushes a little / flushes easily). Choosing a flush-prone setting makes the estimate reach the stages of intoxication sooner, and feeds the hangover forecast too. What it changes is not the concentration number itself, but how that concentration is likely to feel (the stage estimate) — so it's not one-size-fits-all maths, but a guide weighted toward your own constitution. It's an estimate, though, not a measuring device — not something to decide whether you're okay to drive by. For how the in-the-moment "how drunk am I right now" gets estimated, see the real-time prediction article.

Sources

  1. [1] MHLW e-healthnet, "Flushing reaction" (Japanese; on the flush-prone constitution and hangovers)
  2. [2] MHLW e-healthnet, "Aldehyde dehydrogenase 2 (ALDH2)" (Japanese; the three types and enzyme activity)
  3. [3] Chen et al. ALDH2 and the population biology of the flush response. 2022. PMC9235878 (frequency of the ALDH2 variant in East Asia).
  4. [4] Wall, Thomasson, Schuckit & Ehlers. Subjective feelings of alcohol intoxication in Asians with genetic variations of ALDH2. Alcohol Clin Exp Res. 1992. PMID 1443441 (more intense subjective intoxication at the same BAC).
  5. [5] Yokoyama et al. Alcohol flushing, ALDH2 genotype and hangover. 2005. PMID 16046871 (low-activity types get a hangover on smaller amounts).

For exactly which equations and primary sources Nomicho uses, see the methodology.

This article is general information, not medical advice. The figure is a conceptual illustration and the shares are approximate.