Smoking and cancer

Tobacco smoking is the single largest preventable cause of cancer worldwide and is causally linked to at least 17 cancer types. Lung cancer is the most familiar association, but smoking also drives cancers of the mouth, throat, larynx, oesophagus, stomach, pancreas, liver, kidney, bladder, cervix, colon and rectum, as well as acute myeloid leukaemia. Risk rises with the number of cigarettes smoked per day and with the duration of smoking, and falls (slowly) after quitting.

Tobacco smoke is a complex aerosol containing more than 70 established carcinogens, including polycyclic aromatic hydrocarbons (PAHs) such as benzo[a]pyrene, tobacco-specific nitrosamines (NNK and NNN), aromatic amines, formaldehyde, and heavy metals like cadmium and arsenic. These compounds are metabolically activated to electrophiles that bind DNA and form bulky adducts, leaving a characteristic mutational fingerprint dominated by C to A transversions in lung tissue. Persistent DNA damage overwhelms repair, accumulates driver mutations in TP53, KRAS, and other genes, and eventually produces clonal expansion. Smoke also generates oxidative stress, suppresses ciliary clearance, induces chronic inflammation, alters DNA methylation patterns (notably promoter hypermethylation of tumour suppressor genes), and impairs anti-tumour immunity. Field cancerization, where carcinogen-exposed epithelium harbours pre-malignant changes far beyond a visible tumour, is a hallmark of tobacco-related cancers.

Recent work continues to map both the size and the shape of smoking-related cancer risk. The LUCIA cohort study notes that lung cancer remains the leading cause of cancer mortality globally, largely because diagnosis happens at advanced stages, and that current screening criteria built around age and smoking history miss never-smokers and people with non-smoking risks. A multicenter retrospective study reported that smoking and opium use were associated with worse tumour characteristics and survival in colorectal cancer. A modelling analysis projected that full implementation of the World Health Organization MPOWER tobacco control package across the Eastern Mediterranean Region from 2025 to 2050 would meaningfully reduce tobacco-attributable cancer incidence. Mutational signature analyses of head and neck cancers showed that oral cavity tumours in non-smoker, non-drinker, HPV-negative patients carry distinct molecular patterns, suggesting an emerging clinical entity not driven by traditional tobacco mutagenesis. In head and neck cancer treated with chemoradiotherapy, current smokers fared worse than former smokers, consistent with smoking-induced tissue hypoxia and radioresistance. Methylation studies in lung squamous cell carcinoma found promoter hypermethylation patterns linked with smoking-related interstitial fibrosis and emphysema, supporting field cancerization.

Lifetime risk scales steeply with intensity and duration: heavy long-term smokers can carry lung cancer risks 15 to 25 times that of never-smokers. Starting young matters because immature airway epithelium is more susceptible and the cumulative exposure is longer. Genetic variants in carcinogen-metabolising enzymes (for example, CYP1A1, GSTM1) and DNA-repair genes modulate individual susceptibility. Women appear somewhat more susceptible than men to lung cancer per pack-year on some analyses, though the question is unsettled. Secondhand smoke exposure raises lung cancer risk in non-smokers by roughly 20 to 30 percent, and a 2018 China survey found awareness of secondhand smoke harms was uneven, particularly among older and less-educated adults. Co-exposures (alcohol, asbestos, radon, occupational carcinogens) amplify risk multiplicatively rather than additively, especially for upper aerodigestive cancers.

This is among the most settled causal relationships in modern epidemiology, supported by decades of cohort studies, dose-response gradients, mechanistic toxicology, mutational signatures, and the consistent decline in lung cancer rates after population-level smoking reductions. Active research questions are now narrower: how to extend screening fairly to never-smokers and lighter smokers, how electronic cigarettes and heated tobacco products affect long-term risk (data are immature), how stigma affects screening uptake among current smokers, and how to model the cancer prevention impact of national tobacco control policies. Confounding is modest because the exposure is so strongly carcinogenic, but residual issues include misclassification of smoking intensity and changing product landscapes.

The evidence indicates that smoking causes a large share of cancer cases and deaths, and that quitting at any age lowers risk, with larger gains the earlier one stops. Risk does not vanish immediately after quitting, which is why former heavy smokers remain candidates for screening. Population-level policies (taxation, smoke-free laws, plain packaging, cessation support) have a strong track record of reducing future cancer burden.