The 67% cancer jump assumes risk at every age stays put

A new global report estimates 20.6 million new cancer cases in 2024 and projects 34.4 million a year by 2050. That rise comes from a bigger, older world population, with the cancer rate at each age held at today's level. The projection gives no range, and the 2024 figures are themselves estimates, mostly projected from pre-2020 data.

8 October 2026 · How we rate evidence

To keep the number of new cancer cases in 2050 no higher than in 2024, cancer rates around the world would have to fall by about 2% a year, every year, for 26 years. That figure, from a new global cancer report, says more than the number in the headlines.

The claim going around. The report, published on 8 July in CA: A Cancer Journal for Clinicians by researchers at the American Cancer Society and the International Agency for Research on Cancer (IARC), the World Health Organization's cancer agency, estimates 20.6 million new cancer cases and 9.8 million cancer deaths in 2024. It projects 34.4 million new cases a year by 2050. The society's press release said the burden was "Projected to Surge 67% by 2050," and ScienceDaily reposted it on 7 September as "Cancer cases could surge 67% worldwide by 2050." On the numbers, that coverage was fair: both say the rise comes solely from population growth and ageing.

What they did. The 2024 figures come from GLOBOCAN, IARC's database of estimates for 186 countries and 34 cancer types. Where a country has a population-based cancer registry, which records every new diagnosis in a defined area, plus reliable death records, the estimates start from those. Where it doesn't, they are filled in from partial registries, or worked back from deaths using modelled ratios of deaths to cases. The authors are candid about the gaps: only about one country in three reports high-quality data on new cases, and one in four on deaths. Most figures are also projected forward from data collected before 2020.

The 2050 figure is a what-if, not a forecast of how cancer will change. The team took each age group's 2024 cancer rate, held it fixed, and applied it to the population expected in 2050: 9.7 billion people, up from 8.2 billion, and older. The paper gives no uncertainty range for either year.

What they found. On that assumption, yearly new cases rise from 20.6 million to 34.4 million, 67% more. Population growth is the smaller part: 9.7 billion is 18% more people than 8.2 billion. Most of the rest comes from ageing (our arithmetic: 1.67 divided by 1.18 is about 1.41). Worldwide, cases per 100,000 people rise from about 250 to about 355, while the risk at any given age stays the same.

The relative rise is steepest where cancer care is thinnest. Countries are grouped by the Human Development Index (HDI), a UN score combining income, schooling and life expectancy:

How firm the group figures are is a close call: they rest on the thinnest cancer records and have no independent check of their own, but the relative rises depend mostly on how fast populations age, which is known far better.

Survival is where the gap bites. Women in Western Africa are diagnosed with breast cancer at about half the rate of women in Australia and New Zealand, yet die of it at about twice the rate, according to the paper's figures.

What it doesn't show. It doesn't predict how cancer rates will actually change, and the authors say so: holding rates constant "does not account for the dynamic nature of cancer causes and their control." Their own review shows rates moving both ways. Stomach cancer has been falling for a century. Lung cancer is falling among men in several rich countries but still rising among women in many. Colorectal cancer is rising in adults under 50 in the US and elsewhere.

The headline number also depends on how the sum is done. The four country groups in the table add up to about 31.5 million cases in 2050, a 53% rise, not 34.4 million. The paper's caption says the world total "does not equal the sum" and points to IARC's methods page, without giving the reason. One plausible reason, our reading: a single world calculation applies the world's average rates to a population that is shifting toward lower-HDI countries, where rates are lower. The gap is about 3 million cases a year, and on the smaller total the yearly fall needed to hold cases steady is about 1.6%, not 2%.

The starting point is uncertain too. A separate modelling study this year, funded by the American Cancer Society, estimated that about 31% of people worldwide who develop 17 cancers die without being diagnosed, and about two-thirds in Western Africa. As diagnosis improves there, recorded cases could climb faster than demography alone would predict, with no change in risk.

An independent check reached a similar answer. The Global Burden of Disease study, published in The Lancet in 2025, let risk factors and rates change and still forecast a 61% rise in cases from 2024 to 2050, with a range of 42% to 81%. The two projects lean on largely the same registries, so they share blind spots.

What it could mean for people. Does a 67% rise mean your own chance of getting cancer is going up? Not in this projection. By construction, a 60-year-old in 2050 has the same risk as a 60-year-old in 2024. What changes is how many people are 60, 70 or 80. For health systems, that means more diagnoses, treatment and end-of-life care to provide, much of it in countries that already struggle. The report stresses prevention, citing earlier work that nearly half of cancer deaths worldwide are potentially avoidable, through tackling tobacco, infections, alcohol and excess weight, or through earlier detection and treatment.

Why it's still interesting. The projection is partly a record of success: more cancer is largely what happens when more people live long enough to get it. The 2% figure turns that into a benchmark. Falls of that pace happen for single cancers: breast cancer death rates dropped 2% to 3% a year in the decade to about 2017 in several European countries, Australia and New Zealand, the paper notes, though for deaths, not new cases. Whether the world can manage it for new cases, while populations age fastest where cancer doctors are fewest, is the real question behind the 67%.

Open questions.

Will cancer rates themselves rise or fall by 2050? The projection holds them still. Tracking age-standardised rates in high-quality registries, and comparing them with forecasts that model smoking, weight, infections and screening, would show which way they are heading. It matters: a steady 2% yearly fall means about 20 million new cases a year in 2050; no change means 34 million.

How many cancers go uncounted today? Where registries are sparse, the 2024 figures are estimates, and many people with cancer are never diagnosed. New population-based registries, and surveys that establish causes of death in communities without them, would measure this directly. The answer decides whether a future rise in recorded cases in poorer countries means more cancer or better counting.

How close have earlier projections come? Earlier GLOBOCAN editions also projected future cases from population change. Comparing what they projected for the 2020s with what registries later recorded would show how much weight a 26-year projection can bear. That tells planners whether to budget for 34.4 million, or for a range.

A note on timing. The paper and the society's release came out on 8 July 2026. ScienceDaily's repost followed two months later, on 7 September; the study hadn't changed.