American Health & Medical Research Association
Evidence review

Four in 10 cancers trace to changeable risks, and the math has fine print

Smoking, blood pressure, blood sugar, weight and a handful of other risks drive much of cancer, dementia and heart disease. This article covers how "percent preventable" numbers are built and what they can't tell you.

By the Health Research USA editorial team

Published Last updated October 2026

Written from the sources listed below; links checked October 2026. It touches on medicines, doses or health conditions and has not yet been reviewed by a licensed clinician, so use it to prepare for a conversation with yours. How we research and review

Sources15, listed below Reading time15 minutes

Where the evidence stands

  • Strong evidence Smoking is the largest changeable cause of U.S. cancerA handful of risks, led by blood pressure, drive most U.S. coronary deathsAvoiding five heart risks by 50 tracks with a decade or more free of heart disease
  • Promising but early About 45% of dementia could be preventedPrograms aimed at individuals prevent dementia
  • Evidence says no, or unsafe Separate "percent preventable" shares can be added togetherA population share tells you your personal risk

Each use is rated separately, from the scorecard in this guide. Safety cautions are spelled out in the guide itself. How we grade

On this page
  1. Cancer: smoking still dwarfs every other factor
  2. Dementia: 14 factors, and the top two surprise people
  3. Heart disease: blood pressure leads, and progress has stalled
  4. Worldwide: the No. 1 risk changed between two editions
  5. How a "percent preventable" number gets built
  6. What the numbers can't tell you about your own body
  7. Modifiable for whom? Some levers belong to communities
  8. The cross-disease risk factor map
  9. What the evidence says
  10. Questions readers ask
  11. The bottom line

Key takeaways

  • An American Cancer Society study linked about 4 in 10 U.S. cancer cases and about half of cancer deaths in 2019 to changeable risks, with smoking alone tied to about 3 in 10 deaths.
  • The 2024 Lancet Commission tied about 45% of dementia to 14 changeable factors, led by hearing loss and high LDL cholesterol, but critics call that figure a ceiling, not a forecast.
  • In the Global Burden of Disease 2023 study, high blood pressure was the top risk worldwide, and the 88 risks studied accounted for 1.27 billion of the 2.8 billion healthy years lost.
  • “Percent preventable” shares overlap and can add up to more than 100%, so they can’t be stacked, and they describe populations, not your personal odds.
  • In a 2025 study of 2 million people, 50-year-olds free of five major heart risks lived a decade or more longer without heart disease, yet some still got it.

In July 2024, American Cancer Society (ACS) researchers published a tally that reads like good news for anyone who believes in prevention. They studied about 1.78 million cancers diagnosed in 2019 among U.S. adults 30 and older3. They estimated that 713,340 of them, about 4 in 10, were tied to risks people or communities can change13.

The team linked about half of cancer deaths, 262,120 in all, to those same risks1. Cigarette smoking led by a wide margin1.

On July 31, 2024, an international commission writing in The Lancet tied about 45% of dementia worldwide to 14 changeable factors5. And in July 2026, a U.S. analysis linked nearly 9 in 10 deaths from coronary heart disease, the kind caused by narrowed heart arteries, to 12 such risks12.

These numbers back up what natural-health readers have long said: how and where we live shapes how we get sick. But each comes from the same kind of calculation, with assumptions built in, and it describes populations, not any one person.

Cancer: smoking still dwarfs every other factor

The 2024 ACS study, led by Farhad Islami, covered 30 kinds of cancer. Besides habits, it counted ultraviolet (UV) radiation and seven infections, including HPV and hepatitis B and C2.

Ranked by share of cases, the top five were12:

  • Cigarette smoking: 19.3% of cases and about 30% of cancer deaths
  • Excess body weight: 7.6%
  • Alcohol: 5.4%
  • UV radiation: 4.6%
  • Physical inactivity: 3.1%

Shares swung widely by type: every cervical cancer was tied to a changeable risk, as were about 92% of melanomas and 88% of lung cancers, but under 5% of ovarian cancers2.

The figure barely moved from an earlier ACS study of 2014 data, which found 42% of cases and 45% of deaths4. Its authors called those estimates likely conservative, because some risks lacked good national data4.

Prevention here isn’t only about habits. Vaccines protect against HPV and hepatitis B, though senior author Ahmedin Jemal called U.S. HPV vaccine uptake “suboptimal”1.

Dementia: 14 factors, and the top two surprise people

The Lancet Commission on dementia, led by Gill Livingston, kept 12 factors from its 2020 report and added two: high LDL cholesterol in midlife and untreated vision loss in later life56.

Removing all 14 could, in theory, prevent about 45% of dementia, up from 40% for 12 factors in 202057. Hearing loss and high LDL cholesterol had the largest shares, at 7% each. Less schooling and social isolation came next, at 5% each. Most other factors landed between 1% and 3%7.

The method matters more than the list. Because many of these risks cluster in the same people, the commission trimmed each share for overlap, using a measure called “communality”7. The 14 trimmed shares add up to the 45% total7.

The inputs also shift between reports. For hearing loss, the share of people counted as having it jumped from about 32% to 59%, while its extra risk fell from 1.9 times to 1.4 times. The hearing share ended up near where it started, 8% in 2020 and 7% in 20247.

The commission’s message is hopeful. Much of the risk sits in midlife, which it defines as ages 18 to 65, and acting at any age can lower risk6.

In a July 2026 Science Norway article, Norwegian researchers Anders Fjell and Kristine Walhovd pushed back, arguing the benefits are oversold8. In their view, 45% is an upper limit that assumes every risk could be erased, and it rests mainly on observational studies, which can’t prove cause. Some “risks,” such as hearing loss or depression, may be early signs of brain disease rather than causes8.

A 2025 review by Etuini Ma’u, Carol Brayne and colleagues rated the evidence that programs aimed at individuals prevent dementia as “modest at best”9.

Heart disease: blood pressure leads, and progress has stalled

Our article on prevention versus treatment covered the INTERHEART study. Newer U.S. data agree.

The July 2026 analysis, published in JAMA Cardiology by Global Burden of Disease researchers, examined about 473,000 U.S. deaths from coronary heart disease in 2023. According to a summary by the American College of Cardiology, it linked 89% of them to 12 changeable metabolic, behavioral and environmental risks12.

High systolic blood pressure, the top number, was tied to 47% of those deaths, unhealthy diet to 39% and high LDL cholesterol to 29%12.

The trend is mixed. Since 1990, the contribution of smoking fell by a third, while those of excess weight and high blood sugar rose 55% and 39%12. Coronary deaths fell 59% from 1990 to 2023, but only 21% after 201012.

In 2022, the American Heart Association (AHA) turned these risks into a checklist, Life’s Essential 8. It scores four habits (diet, activity, nicotine and, newly added, sleep) and four measurements (weight, cholesterol, blood sugar and blood pressure)13.

A 2025 study in the New England Journal of Medicine shows the stakes. The Global Cardiovascular Risk Consortium pooled data on 2,078,948 people in 39 countries14. It compared 50-year-olds with none of five risks (high blood pressure, high cholesterol, diabetes, smoking and an unhealthy weight) with those who had all five14.

Women with none of the five lived about 13 more years free of cardiovascular disease, such as heart attacks and strokes. Men gained nearly 1114. In absolute terms, about 24 in 100 women with all five risks developed cardiovascular disease by 90, compared with 13 in 100 with none. For men, it was 38 in 100 versus 2114.

Change in midlife still counted. Among people 55 to 60, controlling high blood pressure was linked to the most added years free of heart disease, and quitting smoking to the most added years of life. The authors note these are associations, not proof14.

Worldwide: the No. 1 risk changed between two editions

The Global Burden of Disease (GBD) study, run by the Institute for Health Metrics and Evaluation, estimates the ill health each risk causes. It counts disability-adjusted life years, or DALYs: years lost to early death plus years lived with illness.

In the 2021 edition, published in May 2024, particle air pollution ranked first, at 8.0% of all DALYs worldwide, just ahead of high blood pressure at 7.8%10. In the 2023 edition, published in October 2025, high blood pressure moved to first, at 8.4%, followed by pollution, high blood sugar and smoking11.

The swap looks dramatic, but the two were nearly tied. Their 2021 uncertainty ranges overlapped almost entirely: 6.7% to 9.4% for pollution and 6.4% to 9.2% for blood pressure10. Treat these rankings as a rough order, not a race result.

The 2023 study tied 1.27 billion of the world’s 2.8 billion DALYs to the 88 risks it measured11. That’s about 45% by our arithmetic; the authors call it nearly half. Since 2010, age-adjusted rates fell for most leading risks but rose for high body mass index11.

How a “percent preventable” number gets built

All of these figures rely on a measure called the population attributable fraction, or PAF. In plain words, it asks: if this risk vanished and nothing else changed, what share of cases would disappear?

This is a made-up example. Picture a town with 1,000 new cases of a disease a year, and 800 if no one smoked. Smoking’s PAF would be 200 out of 1,000, or 20%.

Researchers get there by combining how common a risk is with how much it raises the odds, then applying that share to real counts47. Change either input and the answer moves, as the hearing-loss example shows.

A 2007 article by researcher Beverly Levine in Preventing Chronic Disease, a CDC journal, lists the assumptions underneath15:

  • The risk estimates are unbiased.
  • The risk causes the disease, rather than just traveling with it.
  • Removing the risk doesn’t change other risks.
  • To call cases “preventable,” a real way to remove the risk must exist.

Levine also explains why separate PAFs can’t be added. One heart attack might have been avoided by removing either high blood pressure or high cholesterol, so it counts under both. PAFs for several factors therefore usually sum to more than 100%15.

The U.S. heart data show it. By our arithmetic, 47% plus 39% plus 29% is 115%, yet all 12 risks together accounted for 89% of deaths12.

Timing and definitions add wrinkles. The ACS team assumed a 10-year gap between exposure and cancer, which may not fit every cancer3. And broad definitions inflate a PAF: in Levine’s extreme example, simply being older than 15 accounts for 100% of breast cancer15. Our article on telling a link from a cause covers other traps.

What the numbers can’t tell you about your own body

A PAF describes a crowd. It can’t say why one person got sick, or how much your own risk would fall if you changed one habit.

In the lifetime study, even among 50-year-olds with none of the five risks, about 13 in 100 women and 21 in 100 men still developed cardiovascular disease by 9014. Healthy habits lowered the odds. They did not erase them.

So if you or someone you love got sick after years of careful living, these numbers don’t mean anyone failed. About 6 in 10 U.S. cancer cases in 2019 weren’t tied to any of the risks the ACS measured1. Geir Selbæk, a contributor to the Lancet report, warned that prevention messages shouldn’t breed stress or shame8.

Your own odds depend on your age, family history and test results. Our article on finding your health priorities explains free risk calculators that give a personal estimate.

Modifiable for whom? Some levers belong to communities

“Modifiable” sounds personal. But several of the biggest risks aren’t fully in your hands.

Particle pollution ranked first or second worldwide in both GBD editions and adds 3% to the dementia total71011. You can stay indoors on a smoky day, but you can’t clean a region’s air alone.

Income, schooling and access to care matter too. The AHA weighed them, along with stress and structural racism, but left them out of its score until they can be measured well13. Donald Lloyd-Jones, who led the writing group, still called them foundational to heart health13.

Shared action has a track record: since 1990, the contribution of smoking to U.S. coronary deaths fell by a third, and that of particle pollution by three-quarters12. Because a PAF tracks how common a risk is, our reading is that those drops largely reflect fewer smokers and cleaner air15. The 2025 dementia review cites falling dementia rates in wealthy countries as similar evidence9.

Natural-health values fit here in two ways. Self-reliance covers the habits that are yours: food, movement, sleep, no tobacco. Community covers clean air, safe streets and smoke-free spaces, which neighbors win together. Our next article, on health promotion, shows how communities have done it.

The cross-disease risk factor map

This map is ours, built from the studies above. The shares come from different studies, places and years, so compare them loosely and never add them.

Risk Diseases it feeds Share estimates A first step
Smoking and nicotine Cancer, heart disease, dementia 19.3% of U.S. cancer cases and about 30% of cancer deaths12; 2% of dementia7; No. 4 worldwide in 202311 Quit support at any age, as the Lancet Commission urges6
High blood pressure Heart disease, stroke, dementia 47% of U.S. coronary deaths12; 2% of dementia7; No. 1 worldwide in 202311 A blood pressure reading; it’s one of the AHA’s four core measurements13
High LDL cholesterol Heart disease, dementia 29% of U.S. coronary deaths12; 7% of dementia7 A cholesterol test; no fasting needed for the AHA measure13
High blood sugar Heart disease, dementia No. 3 worldwide in 202311; diabetes 2% of dementia7 Ask about a glucose or A1c test13
Unhealthy diet Heart disease, cancer 39% of U.S. coronary deaths12; several diet factors in the cancer estimate2 A later article covers eating patterns
Excess weight, inactivity Cancer, heart disease, dementia 7.6% and 3.1% of U.S. cancer cases12; 1% and 2% of dementia7 A later article covers activity
Alcohol Cancer, dementia 5.4% of U.S. cancer cases1; 1% of dementia7 A later article covers alcohol
UV radiation Skin cancer 4.6% of U.S. cancer cases; about 92% of melanomas12 Sun protection; see related reading
Hearing loss Dementia 7%7 A hearing check; the commission urges prompt hearing-aid referral6
Air pollution Heart disease, dementia 8.0% of healthy years lost worldwide in 202110; 3% of dementia7 Mostly a community lever
Social isolation, less schooling Dementia 5% each7 A later article covers social connection
Cancer-causing infections Cancer 100% of cervical cancers in the 2024 estimate2 Ask about HPV and hepatitis B vaccines1
Sleep outside 7 to 9 hours Heart health Part of the AHA score, with 7 to 9 hours as the adult target13; no share estimated here A later article covers sleep

Smoking and high blood pressure feed every major disease here. If you’re choosing where to start, discuss those two with your clinician.

What the evidence says

Claim Evidence level What it rests on
Smoking is the largest changeable cause of U.S. cancer Strong evidence Top factor in two ACS estimates, using 2014 and 2019 data14
A handful of risks, led by blood pressure, drive most U.S. coronary deaths Strong evidence About 89% of 2023 deaths tied to 12 risks in a large modeled analysis12
Avoiding five heart risks by 50 tracks with a decade or more free of heart disease Strong evidence 2 million people in 39 countries; observational14
About 45% of dementia could be prevented Promising but early Consistent links, but cause unproven for several factors; 45% is a ceiling78
Programs aimed at individuals prevent dementia Promising but early Evidence “modest at best,” per a 2025 review9
Separate “percent preventable” shares can be added together Evidence says no Overlapping causes push sums past 100%1215
A population share tells you your personal risk Evidence says no PAFs describe groups; risk stays above zero even without major risks1415

Questions readers ask

If 4 in 10 cancers are tied to changeable risks, did my habits cause my cancer?

Not necessarily. A PAF describes a population, not one person’s illness, and people with no major risks still get sick14.

Why do the percentages change from one report to the next?

They depend on how common each risk is, how strongly it’s linked to disease and how it’s defined715. The Lancet Commission’s smoking share, for example, fell from 5% to 2% between its 2020 and 2024 reports7.

Is it too late to lower my risk in my 50s or 60s?

The research is encouraging. In the 2025 lifetime study, controlling high blood pressure or quitting smoking between 55 and 60 was linked to more healthy years14. The Lancet Commission’s advice is that action helps at any age6.

The bottom line

The big estimates agree on the direction: a short list of changeable risks sits behind much of cancer and heart disease and, less certainly, dementia. Smoking and high blood pressure recur most. That supports the natural-health instinct to put prevention first.

Read the exact percentages as estimates for a whole population, not a verdict on anyone’s choices. Use them to pick a starting point and get your numbers checked, and remember that clean air and safe streets come from communities, not just individuals.

About this article

This article was researched using reports from the American Cancer Society, the Lancet Commission on dementia, the Global Burden of Disease study, the American Heart Association and the CDC, plus peer-reviewed studies and expert commentary, all listed below. It was reviewed in October 2026. It is general education, not medical, legal or financial advice.

Sources

  1. American Cancer Society. “40% of cancer cases and almost half of all deaths in the U.S. linked to modifiable risk factors” (on Islami F, et al., CA: A Cancer Journal for Clinicians, 2024). News release via ScienceDaily, July 2024. sciencedaily.com
  2. The ASCO Post. “40 Percent of Cancer Cases and Almost Half of All Cancer Deaths in the United States Linked to Modifiable Risk Factors.” July 12, 2024. ascopost.com
  3. AJMC. “New Study Reveals 40% of US Cancer Cases, Nearly Half of Deaths Linked to Modifiable Risk Factors.” July 11, 2024. ajmc.com
  4. American Cancer Society. “More than 4 in 10 Cancers and Cancer Deaths Linked to Modifiable Risk Factors” (on Islami F, et al., CA: A Cancer Journal for Clinicians, 2017). November 21, 2017. cancer.org
  5. Alzheimer Europe. “2024 Lancet Commission underscores potential dementia risk reduction, identifying 14 modifiable risk factors.” 2024. alzheimer-europe.org
  6. Diabetes & Primary Care. “Distilled: Reducing dementia risk” (on Livingston G, Huntley J, Liu KY, et al., “Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission,” The Lancet, 2024;404:572–628). diabetesonthenet.com
  7. Canadian Audiologist. Analysis of hearing loss in the 2024 Lancet Commission report, with 2020 and 2024 attributable fractions. canadianaudiologist.ca
  8. Kjørstad E. “There is strong evidence that lifestyle and health measures can help reduce the risk of dementia.” Science Norway, July 14, 2026. sciencenorway.no
  9. Global Brain Health Institute. “Are modifiable risk factors for dementia really modifiable?” (on Ma’u E, Cullum S, Röhr S, Brayne C, Current Opinion in Psychiatry, 2025). May 27, 2025. gbhi.org
  10. GBD 2021 Risk Factors Collaborators. “Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990–2021.” The Lancet, May 16, 2024 (summary via IHME). healthdata.org
  11. GBD 2023 Disease and Injury and Risk Factor Collaborators. “Burden of 375 diseases and injuries, risk-attributable burden of 88 risk factors, and healthy life expectancy in 204 countries and territories, 1990–2023.” The Lancet, 2025;406:1873–1922 (record via Health Research Board National Drugs Library). drugsandalcohol.ie
  12. American College of Cardiology. “GBD 2023: High SBP, Diet and High LDL-C Top Drivers of US IHD Deaths” (Journal Scan of Benziger CP, et al., JAMA Cardiology, 2026). July 22, 2026. acc.org
  13. American Heart Association. “American Heart Association adds sleep to cardiovascular health checklist” (on Lloyd-Jones DM, et al., “Life’s Essential 8,” Circulation, 2022). News release via Newswise, June 29, 2022. newswise.com
  14. Global Cardiovascular Risk Consortium. “Global Effect of Cardiovascular Risk Factors on Lifetime Estimates.” New England Journal of Medicine, 2025 (copy hosted by Universidad Nacional de La Plata). med.unlp.edu.ar
  15. Levine B. “What Does the Population Attributable Fraction Mean?” Preventing Chronic Disease (CDC), 2007;4(1). cdc.gov