Does quitting smoking lower colorectal cancer risk?

01

Where a ‘roughly half after ten years’ headline can go wrong

A 2013 study followed 134,204 women and men in two US health-professional cohorts. Smoking information was updated every two years. Its molecular analysis included 1,260 new colorectal cancers with suitable tumour specimens; it did not obtain tumour blocks for every cancer diagnosed in the cohorts.

For cessation of 10–19 years compared with current smoking, the adjusted hazard ratio for CIMP-high cancers was 0.53 (95% confidence interval 0.29–0.95). For all analysed colorectal cancers, it was 0.96 (0.75–1.23). Removing the words ‘CIMP-high’ changes the claim. A hazard ratio compares rates of occurrence during follow-up; it is not the reader's absolute probability or a guarantee of half their personal risk.

[2]

02

A tumour label is not a label you can give yourself

CIMP describes a pattern of chemical marks on DNA called methylation. In this study it was measured in tumour tissue and used to classify the cancers that occurred. CIMP-high and CIMP-low/negative were different outcomes. These labels do not mean that everyone who smokes has a known future tumour type; nor are they a hereditary diagnosis that can be read from a smoking record.

The study was observational and exploratory. Most participants were White health professionals; missing specimens and remaining confounding, including past smoking amount, limit interpretation. A weak or statistically uncertain trend in another subtype does not prove quitting has no value. Conversely, the stronger CIMP-high finding does not establish the same effect for every subtype or population.

[2]

03

Check what the study counted before using its number

An anatomical category such as colon or rectum is not the same as a molecular category. A study may combine anatomical sites while separating molecular features, as this one did. The 2020 review also included studies of new cancer and cancer deaths; those endpoints answer different questions. A result for new cases cannot simply be called a survival or recurrence benefit after diagnosis.

Colorectal adenomas are another distinct outcome. NCI discusses their association with smoking separately from invasive cancers and deaths. An adenoma result is not automatically a count of cancers prevented. Before trusting a headline, retain its outcome, comparison group and study population—not just its most striking percentage.

[1][2][3]

04

Quitting and appropriate care belong alongside each other

Age, family history, inherited conditions, previous polyps and inflammatory bowel disease can still matter after quitting. A quit date or this table cannot decide screening eligibility, replace an agreed follow-up or select a molecular test. Discuss the smoking change and your actual question with the responsible professional; do not cancel existing care on the strength of a research headline.

Early colorectal cancer may have no symptoms. Blood in the stool or persistent changes in bowel habits need a health-care conversation, not an explanation based solely on quitting. For separate help stopping smoking in England, NHS local services are an entrance; confirm access locally. This page does not need symptoms, smoking history or test results entered into it.

[3][4][5]

05

What to keep in mind

• Keep the subtype in the headline when it belongs to the result.
• New cancer, adenomas, recurrence and death are different endpoints.
• A smoking record does not assign a molecular subtype or settle screening.