Epigenetic Age Estimates in Blood Shift by Decades with Cell Population: A Within-Donor Variance Partition Across Two Platforms and Two Study Designs
This study demonstrates that epigenetic age estimates in blood vary by decades depending on the specific cell population assayed, a variation large enough to dominate biological signals and confound cross-study comparisons, thereby establishing cell composition as a critical methodological variable that requires standardized reporting.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
The Biological Timekeeper's Great Mix-Up
Imagine your body is a bustling city, and inside every neighborhood (your tissues), there are tiny workers called cells. For years, scientists have been trying to build a "biological clock" to tell us how fast our city is aging. They do this by looking at a chemical tag called DNA methylation, which acts like a sticky note on our genetic instructions. As we get older, these sticky notes change in predictable ways. By counting them, researchers can estimate your "epigenetic age"—a number that often predicts how long you might live better than your actual birthday does. This has become a super-popular tool for studying everything from heart disease to how pollution affects us.
But here's the catch: almost all of this research is done using blood samples because they are easy to get. The problem is that "blood" isn't just one thing. It's a chaotic soup of different cell types—some are the immune system's soldiers (like T-cells), others are the cleanup crew (like monocytes), and some are the heavy-duty fighters (like granulocytes). Scientists have been treating a bowl of mixed fruit salad as if it were a single apple. They've been comparing studies that used whole blood with studies that used only specific, sorted cells, assuming the results would be the same. This paper asks a simple but terrifying question: Does it actually matter which specific cells you pick out of that blood soup to measure your age?
The Great Cell Switcheroo
This study, led by Frederic Scheer, decided to test that idea with a clever experiment. Think of it like a "blind taste test" for aging, but instead of ice cream flavors, the scientists were tasting different types of blood cells from the exact same people.
They started with a group of six healthy volunteers. For each person, they didn't just take one blood sample; they took ten different versions. They pulled out the whole blood, then they carefully separated out just the T-cells, just the B-cells, just the monocytes, and so on. It was like taking one person, slicing them into ten different "cell-only" versions, and asking the biological clock, "How old is this person now?"
The results were mind-boggling. When they switched the cell type, the estimated age of the same person jumped around wildly.
- Using the Horvath clock (a popular, general-purpose clock), the estimated age shifted by a median of 15.6 years just by changing the cells.
- Using the Hannum and PhenoAge clocks, the shift was even crazier: a median of 36.7 years.
- In one extreme case with the PhenoAge clock, the same person's estimated age swung by 54.0 years depending on which cells were tested.
To put that in perspective: the difference in age between two random strangers is usually around 36 years. This study found that just by changing the type of blood cell you test, you can create a fake age difference as big as the difference between two different people.
The "Young" and the "Old" Cells
The study also discovered that different clocks see different cells as "old" or "young," and they don't always agree.
- CD8+ T-cells (a type of immune cell) were a major source of confusion. The Hannum and PhenoAge clocks thought these cells were incredibly young—sometimes estimating them to be 30.9 years younger than the person's actual age! In fact, for some people, the PhenoAge clock gave a negative number (like -3.8 years) for these cells, which is biologically impossible and suggests the clock was completely confused by the data.
- Monocytes (another immune cell), on the other hand, were consistently read as "older" than the person actually was.
The researchers also looked at what happens when scientists in the real world swap one blood sample for another. In many studies, researchers might use "whole blood" (the whole soup) in one group and "PBMC" (just the white blood cells) in another. The study found that making this switch shifts the Hannum clock estimate by a median of 5.6 years and the PhenoAge clock by 4.5 years. If you compare a group of patients using whole blood to a group using sorted cells, you might think the patients are nearly a decade apart in age, when in reality, they might be identical.
The "Multi-Tissue" Hero and the "Blood-Only" Losers
You might think, "Well, maybe the clocks trained specifically on blood would be better at handling blood cells." The study tested this idea and found it to be false.
- The clocks trained only on whole blood (Hannum and PhenoAge) were the most sensitive to the cell mix-up, swinging wildly.
- The clock trained on many different tissues (skin, blood, mouth, etc.), known as the Horvath multi-tissue clock, was the most stable. It still shifted by 15.6 years, but that was much better than the 36+ year swings of the others.
- Interestingly, a newer clock trained on skin and blood was also quite stable, suggesting that training a clock on a variety of tissues makes it tougher against cell mix-ups, even if it's still not perfect.
The Bottom Line
The paper concludes that epigenetic age in blood isn't just a number about you; it's a number about you plus the specific cells you happened to test. The effect of the cell type is so huge that it can completely drown out the real biological differences scientists are trying to find.
The authors argue that we can no longer ignore how blood samples are prepared. If a study uses whole blood and another uses sorted cells, comparing them is like comparing apples to oranges—or in this case, comparing a 20-year-old's age to a 50-year-old's age just because of the lab technique. The study doesn't say these clocks are useless, but it demands that scientists report exactly which cells they used, because without that info, the age estimate is just a guess. As the paper puts it, blood preparation should be treated as a "primary methodological variable," meaning it's as important as the experiment itself.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.