'Blood age' vs. bowel cancer: Epigenetic clock predicts risk in postmenopausal women

The researchers showed that epigenetic age - age calculated from DNA methylation in white blood cells - is associated with future risk of colorectal cancer (CRC). In postmenopausal women, "older" blood and accelerated aging according to the epigenetic clock predicted a higher probability of developing CRC after years of observation. At the same time, a healthy diet smoothed out the increased risk, and tolerability of early removal of both ovaries (premenopausal oophorectomy) against the background of accelerated epigenetic aging, on the contrary, was accompanied by a noticeable increase in risk. The authors believe that prediagnostic (before cancer is detected) markers of biological aging can help more accurately select women for screening.

Background of the study

Colorectal cancer is typically an age-associated tumor: about 90% of new cases occur after age 50. But “passport” years are a poor explanation for why peers develop the disease so differently: real, biological age reflects the accumulated influences of the environment and habits (obesity, activity, nutrition), which rewire the epigenome - primarily the DNA methylation pattern. Hence the interest in the epigenetic clock (DNAm-age): it is consistent with chronological age in many tissues and captures the influence of lifestyle on molecular functions over time.

At the same time, DNAm signals in tumor tissue behave paradoxically: due to the “rejuvenation” of the clone (expansion of the pool of stem/progenitor cells), the clock in the tumor itself often shows the “youngest” age, which makes them weak predictors of risk. It is more logical to look for a prognostic marker before diagnosis and in the blood: leukocytes are an accessible tissue where the epigenetic clock can record the systemic mechanisms of carcinogenesis through immune and metabolic circuits of the blood and cumulative effects of lifestyle.

To reduce the noise from racial differences in epigenome aging and sample heterogeneity, we focused on white postmenopausal women from the WHI prospective cohort and the EPIC-Italy validation sample, measuring prediagnostic DNAm age using three established clocks (Horvath, Hannum, Levine/PhenoAge). This design allows us to assess whether accelerated epigenetic aging predicts future CRC risk and whether this association varies with modifiable factors such as diet quality, activity, anthropometry, and reproductive interventions such as premenopausal bilateral oophorectomy.

The motivation is practical: if “blood age” really captures vulnerability to CRC and its association is mitigated by a healthy diet, such a marker can be built into personalized screening trajectories - who to call for colonoscopy earlier and monitor more often. At the same time, comparing blood with tissue kits (TCGA, GEO) clarifies why prediction by tumor tissue turns out to be “younger” and predicts risk worse, while pre-diagnostic blood is more informative for prevention.

How was it carried out?

The work is based on data from the WHI, a large prospective cohort of postmenopausal women in the United States. The database included 955 white participants without cancer at baseline; over an average of 17 years of follow-up, 29 of them developed primary CRC. Epigenetic age was estimated using three recognized "clocks" - Horvath, Hannum, Levine (PhenoAge) - in the DNA of peripheral blood leukocytes taken long before diagnosis. The results were confirmed in an independent EPIC-Italy cohort (79 CRC cases and 340 controls) and compared with tissue data from TCGA and GEO, where methylation was analyzed in the tumor and adjacent normal tissue. The models took into account familiar risk factors (BMI, circumference, nutrition, alcohol, smoking, activity), and the composition of leukocytes for the "clock".

The main thing is in numbers

The authors directly tested: if the epigenetic age is older than the passport age (acceleration), what happens to the risk of CRC in the following years?

• For every “+1 year” of epigenetic age, there was approximately a +10% increase in the risk of future CRC. When analyzed by “decades,” the effect was even more dramatic: +10 years of DNAm age corresponded to a ~4-fold increase in risk.
• When accelerated/decelerated aging was expressed as categories (ACC - accelerated vs. DCC - decelerated), women with accelerated aging had a shorter time to cancer and a risk approximately 5-10 times higher than women with decelerated aging.
• Modifiable factors are not just “background”: in women with a healthy diet, there was no increased risk among “accelerated aging” compared to “slow aging.” Early bilateral oophorectomy combined with accelerated aging was associated with a significantly higher risk of CRC.

Why is this important?

CRC is an age-associated tumor, but passport age is a poor explanation for why two peers have different risks. The epigenetic "clock" records biological traces of the environment and habits - from excess weight to activity - and therefore can become a pre-screening biomarker, especially if combined with a lifestyle questionnaire. It is also critical that the tissue clock in the tumor itself paradoxically gives a "younger" age due to the "rejuvenation" of cancer cells (expansion of the stem/progenitor pool), so it is pre-diagnostic blood that looks more suitable for prognosis.

Details worth a closer look

In the section on associations of the “clock” with habitual risk factors, the authors found the expected trends: the higher the BMI and waist-to-hip ratio, the “older” the clock, and physical activity was associated with less acceleration (about minus “one year” according to AgeAccelDiff). Some behavioral signals were ambiguous and depended on whether we looked at the entire sample or only at women who subsequently developed CRC (e.g., for alcohol and smoking). This emphasizes that the “clock” is the cumulative imprint of many influences, and not the line of a single habit.

What this might mean in practice

The idea is simple: to assess the “age of blood” years before a possible disease and to strengthen prevention precisely where the acceleration of aging is most pronounced.

• Who is especially relevant: women 50+ with risk factors for CRC (obesity, low activity, high WHR), as well as women with premenopausal bilateral oophorectomy.
• How to act wisely: increase screening and lifestyle interventions in the group with accelerated epigenetic aging, focusing on a quality diet - in the study it was this that neutralized the additional risk.
• Where are the limits of the method: today it is a scientific tool, not a routine analysis; standards, thresholds and an assessment of clinical benefit are needed over and above the usual predictors (family history, polyps, stool occult blood test, colonoscopy).

Strengths and Limitations

This is a prospective design with pre-diagnostic blood and validation in an independent cohort; the authors additionally compared blood with tissue data. However, the number of cases in the main sample is small (n=29), and the entire main cohort were post-menopausal white women, which limits transferability. Subgroup inferences by lifestyle combinations may be unstable due to small numbers; the validation sets had few covariates, and the methylation platforms differed (450K vs EPIC) - although the authors checked that this did not break the "clock" estimate. Finally, the epigenetic clock is not yet a causal factor, but a risk marker.

What's next?

The team proposes to create a “colonic” epigenetic clock that takes into account the microbiome and diet, and test whether lifestyle interventions reduce epigenetic acceleration and the actual risk of CRC. If algorithms and thresholds can be standardized, “blood age” could be incorporated into personalized screening trajectories: who should have a colonoscopy earlier, who should be monitored more often, and who needs basic tests.

A short reminder from the article

• What are AgeAccelDiff and IEAA?
  AgeAccelDiff is "how much the clock is ahead of the passport age"; IEAA is "intrinsic" aging acceleration (the remainder of the "clock" after taking into account the composition of blood cells). Both indicators were used as continuous and binary metrics.
• What "clock" was counted?
  Classic Horvath, Hannum and Levine (PhenoAge) - they are in good agreement with passport age in different tissues and are sensitive to the sum of genetic and behavioral effects.
• How much has the risk increased?
  ~+10% for each “additional” epigenetic year, ~×4 for each “+10 years”; ACC (acceleration) vs. DCC (slowdown) - approximately ×5-10. These are estimates within the studied samples; for practice, they need to be confirmed in larger studies.

Source: Jung S.Y., Pellegrini M., Tan X., Yu H. Epigenetic age and accelerated aging phenotypes: a tumor biomarker for predicting colorectal cancer. Aging (Albany NY), 17:1624–1666. https://doi.org/10.18632/aging.206276