- A new study of nearly 447,000 adults, published in JCO Oncology Advances and presented at AACR 2026, modeled individualized cancer risk across 38 cancer-specific models.
- Cancer risk varied widely between individuals of the same age and sex, with the highest-risk group carrying close to twice the lifetime risk of the lowest.
- Optimizing modifiable behaviors meaningfully reduced lifetime risk and roughly halved the spread between individuals.
- Ten-year risk overlapped across age groups: a higher-risk 50-year-old can match an average 65-year-old.
- Age and sex alone are not enough to guide modern cancer screening.
A new way of looking at cancer risk
Cancer screening in the United States is built primarily around age. Most major guidelines define when to start, how often to test, and when to stop based on age thresholds, with sex and a handful of high-impact factors layered on top.
That framework has saved lives. It is also incomplete.
In a study published in JCO Oncology Advances and presented at the AACR 2026 Annual Meeting, our team and collaborators projected individualized 10-year and lifetime cancer risk across the UK Biobank, a prospective cohort of 446,795 adults. The goal was to understand how much risk varies between individuals once a broad set of factors, including genetics, lifestyle, environmental exposures, and medical history, is taken into account, and how much of that risk could shift if modifiable behaviors were optimized.[1]
The analysis included 118 distinct variables, drawn together across 38 cancer-specific models. Each individual was assigned a personalized 10-year and lifetime risk for developing any cancer. The resulting distributions tell a more nuanced story than population averages do.
What the study found
Lifetime cancer risk varies widely
Across the cohort, the distribution of lifetime cancer risk was wide and right-skewed in both sexes. The most striking finding was the gap between individuals at the high and low ends.
Men at the 90th percentile of risk carried a lifetime cancer risk roughly 1.4 times the median and 1.8 times that of men at the 10th percentile. For women, the spread was wider still, with the 90th percentile carrying roughly 1.7 times the median and 2.3 times the 10th percentile.[1]
In practical terms, two adults of the same age and sex can carry meaningfully different lifetime cancer risk depending on their underlying profile.
Modifiable factors compress the curve
The analysis then asked a counterfactual question: what would the distribution of risk look like if every individual's modifiable risk factors were set to an ideal state?
The answer was striking. Median lifetime cancer risk fell substantially in both sexes, but the more interesting finding was what happened to the spread. The interquartile range — the gap between people at the 25th and 75th percentile of risk — compressed by roughly half.[1]
Modifiable behaviors do not flatten cancer risk to zero. Genetics, family history, and accumulated environmental exposures all continue to shape risk. But a substantial portion of the variability we see between individuals is attributable to factors people can act on.
Ten-year risk overlaps across age groups
The most policy-relevant finding may be how much short-horizon risk overlaps between age groups once individual factors are accounted for.
In men, those aged 50 to 59 at the 90th percentile of risk carried a higher 10-year cancer risk than men aged 60 to 70 at the 25th percentile. The same pattern held for women, where those aged 40 to 49 at the 90th percentile carried higher 10-year risk than women in their 50s near the median, and matched some women in their 60s.[1]
In other words, a higher-risk individual in their 40s or 50s can carry the same near-term cancer risk as an average individual a decade or two older. Age remains an important signal, but it is not a sufficient one.
Why this matters for screening
Most current screening recommendations are built around the premise that risk rises predictably with age. The new findings reinforce that this premise is directionally correct but operationally imprecise.
Two implications follow.
First, age-based thresholds may under-screen high-risk individuals at younger ages. A 45-year-old in the top decile of risk for their age and sex may benefit from earlier or more comprehensive screening than current guidelines specify.
Second, the same thresholds may over-screen low-risk individuals at older ages. For someone with a favorable risk profile, the marginal benefit of certain screening tests may be smaller than the population-level recommendation suggests.
This is not an argument against age-based screening. Age remains one of the strongest single predictors of cancer risk. It is an argument for layering individualized risk on top of age, particularly as new modalities, including multicancer early detection blood tests and advanced imaging, expand the menu of available tools.
What this means in practice
For most adults, three takeaways are worth holding onto.
Your risk is not fixed by your age. Two people the same age can carry meaningfully different cancer risk. Family history, lifestyle, environmental exposures, and underlying medical conditions all matter, and they accumulate differently across individuals.
Modifiable behaviors compress the curve. Smoking, alcohol intake, body composition, physical activity, and diet remain among the most important levers individuals can pull. Optimizing them does not eliminate risk, but it shifts the entire distribution downward and narrows the spread.
Screening decisions benefit from a personalized lens. When to start, how often, and which modalities to use are questions best answered against an individual's full risk profile, not age alone.
How Catch fits in
Catch is built on the idea that individuals deserve a clear, comprehensive view of their own cancer risk, integrating the same kinds of factors used in this analysis: family history, genetics where available, lifestyle, environmental exposures, and medical history. Each member receives a personalized lifetime and 10-year risk estimate across 21 cancers, along with a screening and prevention plan calibrated to their profile rather than to population averages.
The findings from this study are part of why we built Catch the way we did. Cancer risk is not one-size-fits-all, and the people we serve deserve guidance that reflects that.
Cancer risk varies widely between individuals, even at the same age. A study of 446,795 adults, presented at AACR 2026 and published in JCO Oncology Advances, demonstrates that 10-year and lifetime cancer risk are shaped by a wide constellation of factors, that modifiable behaviors meaningfully compress the risk curve, and that age and sex alone are not sufficient inputs to optimal screening strategy.
The implication is not that age-based guidelines should be discarded. It is that they should be augmented. As the tools available for cancer screening continue to expand, individualized risk assessment will increasingly determine who benefits most, and when.
.svg)
