Lead: A comprehensive global analysis published 3 February 2026 finds that roughly 38% of new cancer diagnoses in 2022—about 7.1 million of 18.7 million cases—were linked to risk factors that can be changed or managed. The study, covering 36 cancer types across 185 countries and drawing on exposure data from 2012, identifies tobacco smoking, infections and alcohol use as the largest contributors. Lung, stomach and cervical cancers together account for nearly half of the cases the authors classify as avoidable. The research appears in Nature Medicine and is led by investigators including Hanna Fink of IARC.
Key takeaways
- 18.7 million new cancer cases were recorded globally in 2022; an estimated 7.1 million (≈38%) are attributable to 30 modifiable risk factors.
- Tobacco smoking was the single largest contributor among assessed risks, cited by the authors as accounting for about 15% of preventable cases.
- Infections (including viral and bacterial agents) were estimated to contribute roughly 10% of preventable cases, while alcohol contributed about 3%.
- Lung, stomach and cervical cancers together make up nearly 50% of all cases linked to modifiable risks in the analysis.
- The study synthesised exposure data from 2012 with 2022 case counts, covering 36 cancer types in 185 countries to estimate attributable fractions.
- Findings emphasise prevention measures — tobacco control, vaccination, infection control and alcohol policy — as primary routes to lower future cancer burden.
Background
The global cancer burden has been climbing for decades and is projected to increase further absent stronger prevention and control measures. Prior work often estimated the fraction of cancer deaths linked to avoidable causes rather than incident cases, and many analyses focused on single risk factors rather than an integrated set. The new study fills that gap by combining current incidence data with historical exposure estimates for multiple established carcinogens, aiming to quantify how many new cases could theoretically have been prevented.
Researchers selected 30 modifiable exposures that have well-established links to cancer risk—examples include tobacco use, alcohol consumption, several infectious agents, and selected occupational and environmental exposures. They analysed incidence for 36 cancer types across 185 countries, using exposure prevalence from 2012 so that latency between exposure and diagnosis could be partially accounted for. The investigators present attributable fractions: the share of cases estimated to be linked to each risk factor, and then aggregate these to produce regional and global totals.
Main event
The analysis found 18.7 million new cancer diagnoses in 2022 and attributed about 7.1 million (38%) to the assessed modifiable risks. Tobacco smoking emerged as the top contributor within those avoidable cases, followed by infections and alcohol. The authors report that lung, stomach and cervical cancers together represent almost half of the preventable-case total, reflecting both the high incidence of those cancers and their well-established links to modifiable causes.
Methodologically, the team combined cancer registry incidence with population-level exposure data from 2012, applying established relative-risk estimates from meta-analyses and cohort studies to compute population attributable fractions. The analysis was global in scope and stratified by cancer type and region, allowing identification of geographic patterns in preventable burden. The authors note that data quality and exposure measurement vary between countries, which they address through sensitivity analyses and uncertainty intervals.
The study was published as a peer-reviewed article in Nature Medicine (Fink et al., 2026) and accompanied by a Nature news summary. Co-author Hanna Fink, a cancer epidemiologist affiliated with the International Agency for Research on Cancer in Lyon, emphasized prevention policies as central to reducing future case counts. The paper complements earlier assessments—such as the Global Burden of Disease work—that focused on mortality or narrower risk sets.
Analysis & implications
Estimating that nearly 40% of incident cancers are linked to modifiable risks reframes how policymakers and health systems might prioritise resources. Where feasible, primary prevention (tobacco control, vaccination against oncogenic infections like HPV and hepatitis B, alcohol regulation, and reduction of occupational exposures) can avert cases before they occur, easing downstream demands on screening and treatment services. The finding that relatively few exposures explain a large share of preventable cases argues for targeted interventions with broad impact.
Geographic variation in attributable burden matters: countries with high prevalence of smoking, low vaccination coverage, or high infection prevalence will see a larger fraction of cases labeled preventable. That suggests policy levers differ by region—tobacco taxation and cessation programs may deliver the biggest gains in some settings, while vaccination and infection control may be the priority where infection-driven cancers predominate.
Economically, preventing millions of cases has implications for health spending and workforce productivity, though the study does not provide a direct cost-benefit analysis. Investment in proven prevention strategies can be cost-effective compared with lifetime treatment costs for cancer, but implementation requires political will, infrastructure and sustained funding. The authors also highlight that prevention reduces suffering and mortality in addition to costs.
Comparison & data
| Metric | Value |
|---|---|
| Total new cancer cases (2022) | 18.7 million |
| Estimated preventable cases | 7.1 million (≈38%) |
| Share attributed to tobacco | ≈15% of preventable cases |
| Share attributed to infections | ≈10% of preventable cases |
| Share attributed to alcohol | ≈3% of preventable cases |
The table condenses the study’s headline figures. Readers should note that the percentages for tobacco, infections and alcohol are presented relative to the set of cases designated as attributable to the listed modifiable risks; the study reports uncertainty ranges around each estimate. Differences in exposure measurement, latency assumptions and underlying relative-risk estimates contribute to those ranges and to regional variability.
Reactions & quotes
Study co-author Hanna Fink placed the findings in a public-health context, stressing prevention as an evidence-based way to lower future cancer burden. The statement below captures that emphasis and the policy implications the authors highlight.
“Reducing exposure to well-established risks—particularly tobacco, infection and harmful alcohol use—remains one of the most effective strategies to curb cancer incidence worldwide,”
Hanna Fink, International Agency for Research on Cancer (study co-author)
International health agencies noted the study reinforces existing prevention priorities. WHO advisers and independent public-health groups said the results align with long-standing recommendations on tobacco control, vaccination and alcohol regulation as central cancer-prevention tools.
“This analysis underscores the returns from proven prevention measures and the need to scale them equitably across regions,”
World Health Organization expert commentary (paraphrased)
Unconfirmed
- Exact allocation of the tobacco/infection/alcohol percentages between ‘of preventable cases’ and ‘of all cases’ varies by table and wording within the paper; readers should consult the full paper for precise denominators.
- Detailed country-level effectiveness of specific policy packages (for example, simultaneous tobacco taxation plus HPV vaccination) was not modelled in this analysis and remains to be quantified.
Bottom line
The study provides a comprehensive, peer-reviewed estimate that nearly two in five new cancers in 2022 were linked to changeable risks, underscoring prevention as a cornerstone of cancer control. Tobacco control, infection prevention and alcohol policy emerge as high-impact areas, but the balance of priorities will differ across regions depending on local exposure patterns and health-system capacity.
For policymakers, the implication is clear: scaling and tailoring proven prevention interventions could substantially reduce future incidence and the related social and economic costs. For researchers and funders, the study highlights where more granular, country-specific modelling and implementation research would help turn attributable fractions into measurable reductions in cases.
Sources
- Fink H. et al., Nature Medicine (2026) — peer-reviewed original study (journal).
- Nature news: More than one-third of cancer cases are preventable — news coverage summarising the study (media).
- International Agency for Research on Cancer (IARC) — research institute/WHO agency background and author affiliation (official research agency).