Nicotine and Cotinine Metabolism

When nicotine enters the body it is broken down, mostly into cotinine, and cotinine is then cleared. Cotinine lasts far longer than nicotine, which is why it is the standard measure of tobacco exposure: it is what a laboratory tests to check whether someone smokes, whether they have quit, and how much second-hand smoke they are getting.

That makes it a measuring instrument. And a measuring instrument that varies between people for reasons unrelated to what it is measuring is worth understanding.

The question the study was actually asking

This research did not begin as a curiosity about metabolism. It began with an unexplained fact: smokers' lung cancer risk differs by race and ethnicity even after adjusting for how much they smoke.

One candidate explanation is dose. Two people who smoke the same number of cigarettes do not necessarily absorb, metabolise and clear the same amount of nicotine — and if the amount that stays in the body differs, so does the exposure that matters.

The design, which is the good part

The study genotyped 2,239 smokers from the Multiethnic Cohort and measured nicotine and its metabolites in urine. The composition of that sample is why it can answer the question at all:

• 437 European American
• 364 African American
• 453 Latino
• 674 Japanese American
• 311 Native Hawaiian

Almost every study on this site is overwhelmingly European-ancestry, and we say so on almost every page. This one was built the other way round, because the difference between groups was the question. It is worth pointing at as a design rather than passing over.

What it found

Of 11,892,802 variants analysed, 1,241 were strongly associated with cotinine glucuronidation, and 490 of those were also associated with nicotine glucuronidation. The gene doing most of the work is UGT2B10 — a UDP-glucuronosyltransferase, one of the enzymes that attaches a sugar to a compound to make it water-soluble enough to be excreted.

Two of the positions on this page, rs141360540 and rs115219551, are in UGT2B10 itself.

Why a test that depends on genotype is a problem worth naming

Cotinine is used to decide things. Insurers use it to price policies. Transplant programmes and some surgical services use it to verify abstinence. Studies use it to classify people as smokers or non-smokers.

All of that assumes the number means the same thing in everybody. If clearance is partly genetic, then two people with identical smoking behaviour can produce different numbers, and the direction of that difference is not random with respect to ancestry.

This does not make the test useless — it remains far better than asking people. It does mean a cotinine level is a measurement with a person-specific scale factor in it, and that is not usually how it is treated.

CYP2A6 — the enzyme that does most of the work, and a lung cancer connection

The same research group followed up a year later with a second paper, using the same 2,239-smoker Multiethnic Cohort, this time on CYP2A6 — the enzyme responsible for nicotine's first breakdown step. Patel et al. 2016 found 248 variants in and around CYP2A6 associated with metabolic activity — a tightly linked cluster of signals in one gene region, not 248 independent discoveries.

The more consequential finding came from testing those same activity-linked variants against lung cancer risk in a much larger set — 13,479 cases and 43,218 controls. Faster CYP2A6 activity was directionally linked to higher lung cancer risk for 81% of the variants tested, reaching statistical significance for six of them. The proposed mechanism: someone who clears nicotine faster tends to smoke more, or smoke more intensely, to maintain the same nicotine level — a gene that changes clearance speed changes exposure, and exposure is what drives the cancer risk.

Ten of this page's variants, mostly in or near CYP2A6 — including rs113029345 and rs7247903 directly in the gene — come from this study.

Where genotype has actually been tested against a treatment choice

This is a rare case on this site where a genotype-linked measurement has been tested in a real randomized trial against a treatment decision, rather than only observed to correlate with something. Lerman et al. 2015 randomized 1,246 smokers, split by their nicotine metabolite ratio into "normal" and "slow" metabolizers, to the nicotine patch, varenicline, or placebo. Normal metabolizers quit more successfully on varenicline than on the patch; slow metabolizers did equally well on the patch, with fewer side effects than varenicline. No consensus clinical guideline yet recommends testing the metabolite ratio before choosing a smoking-cessation medication, but this is real trial evidence that the ratio predicts which treatment works better for whom — a stronger result than most of what this site can point to.

What eighteen positions tell you

Nothing you should act on for most of these. There is no consensus clinical guideline that uses nicotine or cotinine clearance speed to change a prescription — the trial above is a real step in that direction for smoking-cessation drug choice specifically, but it is one trial, not an adopted guideline. Nothing here changes how to stop smoking on your own.

What does change outcomes is well known and has nothing to do with genotype: stopping, with help. See smoking cessation medicines for where a genotype does briefly enter that conversation, and smoking behaviour and nicotine dependence for the genetics of the habit itself rather than of the metabolite.