Cold War nuclear tests nearly doubled the carbon-14 in the atmosphere, leaving an isotopic timestamp inside cells born during those decades — and researchers later used that accidental fallout to answer a question anatomy had failed to settle: whether the adult human heart can make new cells

Anyone alive in 1963 ate fallout. Not enough to hurt them, but enough to leave a mark, and that mark is still sitting inside their cells.

Between 1955 and 1963, above-ground nuclear tests hurled neutrons into the sky, and those neutrons struck nitrogen atoms and converted them into carbon-14, a slightly heavier and mildly radioactive version of ordinary carbon. It oxidised into carbon dioxide, spread around the planet, and moved straight into the food chain. By 1963, atmospheric carbon-14 had roughly doubled, as documented by the Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory.

Then the testing stopped, and levels began to slide as the excess dissolved into the oceans and drifted into plants. A sharp spike, then a long smooth decline. Plotted against time, that shape is a calendar, one accurate to about two years according to a 2018 review of nuclear physics applications co-written by physicist Marco Durante.

DNA is what makes it useful. Most molecules in a cell churn constantly, but once a cell divides for the final time, the carbon in its DNA is locked. Whatever ratio the atmosphere happened to hold that year stays there permanently. Measure it, match it against the curve, and a cell will tell you its birthday.

The question anatomy could not settle

Skin, gut lining, blood: nobody doubted those are replaced. Heart muscle was assumed to be different, a fixed allocation issued at birth and spent down for the rest of a person’s life. That assumption explains a lot of cardiology. It is why a heart attack leaves scar tissue where working muscle used to be.

Testing it was the hard part. The usual way to catch a dividing cell is to feed someone a chemical tag such as bromodeoxyuridine and see which cells absorb it, but as reported by NOVA, the compound is toxic and possibly carcinogenic. No ethics committee was going to approve dosing healthy volunteers to settle a textbook dispute.

A hunch, and a great many horse heads

Jonas Frisén, at the Karolinska Institute in Stockholm, suspected the bomb curve could do the job without touching a living patient. He has described it as a high-risk project, and only one postdoc in his lab was willing to take it on. Kirsty Spalding spent part of her early work outside a slaughterhouse, where horses went in one side and workers emerged from the other holding severed heads for her to take away. Hours of chipping through each one followed, all for the neurons inside. She has said the whole experience was exactly as disgusting as it sounds.

Forensics had already shown the principle worked on people. In the 1990s, Viennese police asked physicist Walter Kutschera to determine which of two elderly sisters, found dead in the same flat, had died first, with a substantial estate at stake. Rapidly renewing tissue gave him the answer.

Their first major biological result appeared in Cell in 2005. Non-neuronal cells in the adult cortex turned over steadily. Neurons in the occipital cortex were as old as the person carrying them.

What the heart samples showed

By 2009, the same trick had a new target: the heart. Olaf Bergmann and colleagues sorted heart muscle nuclei using antibodies against cardiac troponin, extracted the genomic DNA, and measured its carbon-14 by accelerator mass spectrometry, a technique sensitive enough to count individual atoms. They published the result in Science: cardiomyocytes, the heart’s own muscle cells, did renew, at roughly 1 per cent a year at age 25, tapering to 0.45 per cent by 75. Fewer than half of a person’s heart muscle cells get replaced across a normal lifespan.

Decades of medical dogma, settled by a single decimal point.

The follow-up made it messier

In 2015 the same group examined tissue from 29 deceased people of various ages and reported in Cell that the full complement of heart muscle cells is present within a month of birth. Childhood growth comes from those cells swelling, not multiplying. Over a long life, per Karolinska Institutet, only around 40 per cent are exchanged.

Other cells in the same organ move far faster. Endothelial cells lining the blood vessels turn over at more than 15 per cent a year, while mesenchymal cells, the heart’s connective tissue, turn over at under 4 per cent in adulthood.

At least three different clocks are running inside that single organ.

The numbers remain contested

Two papers from one laboratory using one technique do not close a field. Jeremy Elser and Kenneth Margulies at the University of Pennsylvania, writing in PLOS ONE, built a hybrid mathematical model and relaxed some of the underlying assumptions. Much higher turnover rates, climbing with age instead of falling, also fit the radiocarbon data. It’s now widely accepted that adult hearts do make some new muscle. How much, and how usefully it might be stimulated after a heart attack, is still open.

The ruler is disappearing

Burning coal and oil releases carbon containing no carbon-14 whatsoever, which dilutes the signal every year. In 2021, per Eos, the atmospheric concentration slipped below pre-bomb levels for the first time since the 1950s. Bruce Buchholz, the Livermore physicist behind many of these measurements, expects the window to close for anyone born after the mid-2020s.

Nobody set out to build a clock accurate to two years for dating human cells. Three governments trying to frighten each other left it lying around. It worked for roughly seventy years, and now it is switching itself off. Whoever tackles the next question of this kind will have to design the instrument on purpose, which is a much harder way to get lucky.