How much cancer is actually preventable, and what moves the needle most

"How much cancer can you actually prevent?" is one of the most searched and most muddled questions in health. The muddle comes partly from headlines that alternate between blaming everything and blaming nothing. The best estimate for a whole population is neither: a large analysis found that a substantial minority of cancers trace to risk factors people or societies can change — and that the factors are very unequal in how much they matter.

The headline number

Researchers at the American Cancer Society estimated the share of US cancers attributable to a defined list of modifiable exposures: cigarette smoking, secondhand smoke, excess body weight, alcohol, eating red and processed meat, low intake of fruit, vegetables, fibre and calcium, physical inactivity, ultraviolet radiation, and six cancer-associated infections [s1]. In adults aged 30 and older in 2014, they estimated that 42.0 percent of all incident cancers — 659,640 of 1,570,975 cases, excluding non-melanoma skin cancers — and 45.1 percent of cancer deaths were attributable to those factors [s1].

So a little under half of cancers, by this accounting, are linked to something modifiable. That is a large, real figure. It is also not 100 percent: the majority of the arithmetic in this study is attributable to identified causes, but a great deal of cancer risk comes from age, inherited genetics, and chance mutations that no lifestyle change addresses. The study's authors add that their figure may even understate the modifiable share, because not every likely risk factor is established firmly enough to quantify [s1]. Prevention has enormous potential; it is not omnipotence.

One factor dominates

The most important thing the numbers show is how lopsided the contributions are. Cigarette smoking alone accounted for the highest proportion of cases, 19.0 percent, and of deaths, 28.8 percent [s1]. Excess body weight was next, at 7.8 percent of cases and 6.5 percent of deaths, followed by alcohol at 5.6 percent of cases and 4.0 percent of deaths [s1]. In other words, smoking is responsible for roughly as much cancer as the next several factors combined — a scale that reframes what "cancer prevention" should focus on.

That lopsidedness carries through to specific cancers. Lung cancer had by far the most cases (184,970) and deaths (132,960) attributable to the evaluated factors, followed distantly by colorectal cancer (76,910 cases and 28,290 deaths) [s1] — a direct consequence of tobacco's dominance.

Prevention that has already worked

This is not a hypothetical. US cancer mortality has fallen for decades, and the American Cancer Society attributes the aversion of more than 4 million deaths since 1991 to a combination of reductions in smoking, earlier detection for some cancers, and improved treatment [s2]. The single biggest driver of that decline — falling smoking rates — is the population-level proof that changing a modifiable risk factor changes the cancer burden.

Reading the smaller factors honestly

The factors below smoking, weight and alcohol are real but individually smaller, and this is where public understanding often goes wrong. Diet-related factors, physical inactivity and UV exposure each contribute, and the six cancer-associated infections — which include HPV and hepatitis viruses, addressable through vaccination and treatment — are a meaningful and often overlooked category [s1]. But none approaches the magnitude of tobacco. A person optimising every dietary detail while ignoring smoking has the priorities inverted relative to what the evidence shows.

What this means for a reader

The evidence supports a clear, unglamorous hierarchy. Not smoking, and not being exposed to others' smoke, is by a wide margin the most powerful cancer-prevention step available [s1]. Maintaining a healthy weight, limiting alcohol, staying physically active, protecting skin from UV, and taking up vaccination against cancer-causing infections all add to that, in roughly that order of established magnitude. None of this guarantees an individual will or will not develop cancer — much of cancer risk remains outside anyone's control — but at a population level, a little under half the burden is linked to factors that can be changed, and the largest share of that is concentrated in one. This is general information, not personal medical advice; individual risk and prevention are worth discussing with a clinician.

Sources

• [s1] Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States, CA: A Cancer Journal for Clinicians, 2017-11-21
• [s2] Cancer statistics, 2024, CA: A Cancer Journal for Clinicians, 2024-01-17