A ration’s end, and a cancer signal decades later
A PNAS study of 64,761 UK Biobank adults used the end of British sugar rationing in September 1953 as a natural experiment on the first 1,000 days of life. The question is not whether sweet food is vaguely unwise. It is whether the amount of sugar an infant and toddler actually consumed—set, in this case, by a national ration rather than by a household’s taste—left a mark that could still be read in adult cancer rates, in leukocyte telomeres, and even in decades-later diets.
The contrast the researchers exploited is blunt. People whose early childhood fell under rationing (~40g sugar/day) later showed sharply lower adult cancer rates than cohorts born after intake roughly doubled to ~80g. That doubling was not a slow cultural drift. It was the lifting of a wartime and postwar control. When the ration ended in September 1953, the sugar available in a British child’s day jumped from about 40 grams to about 80 grams. Half a century on, that jump is still visible in the bodies of UK Biobank participants.
The paper sits in PNAS—the Proceedings of the National Academy of Sciences—and it treats history as an instrument. Randomizing tens of thousands of infants to high or low sugar would be unethical. The British state did something close by accident: it constrained sugar, then it did not. Birth timing relative to September 1953 decided whether a child’s first 1,000 days unfolded under scarcity or under the doubled allowance.
Why the first 1,000 days
The first 1,000 days of life—roughly the stretch from early gestation through the second birthday—are the window developmental science already treats as unusually plastic. Organs are finishing. Metabolic set points are being written. Taste preferences are forming. The new work takes that window as the exposure period and asks what happens when the exposure is sugar, measured not by a food-frequency questionnaire filled out by a parent decades later, but by a national ration that applied across the country.
Rationing held intake near ~40g sugar/day. After the ration ended, intake roughly doubled to ~80g. The study does not claim that 40 grams is an ideal pediatric target, nor that 80 grams is a unique modern excess. It claims that the difference between those two early-life regimes, imposed by policy rather than by household whim, sorted a generation into two sugar histories—and that those histories later sorted adult cancer rates.
Because the cutoff was a calendar date, the design is a natural experiment. Children whose early childhood fell under rationing spent that formative stretch at the lower gram count. Cohorts born after the cap came off spent the same stretch in a world where intake had roughly doubled. Near the cutoff, exposure was partial. That gradient is what lets the authors talk about people who were fully exposed to the low-sugar early childhood versus people who were not.
A natural experiment is not a perfect trial. Families still differed in wealth, in what else was on the plate, and in later smoking, work, and medical care. What the design does hold roughly constant is the sudden, nationwide change in sugar itself. September 1953 is not a metaphor. It is the month the ration ended, and therefore the month the early-life sugar environment changed for an entire birth cohort.
Sixty-four thousand adults, one biobank
The analysis drew on 64,761 UK Biobank adults. The UK Biobank is a long-running resource of volunteers who have given blood, answered questionnaires, and allowed their health records to be followed into middle and later life. It is not a pediatric clinic. It is a warehouse of adult outcomes. That is the point. The sugar exposure happened in infancy and toddlerhood. The cancers, the telomeres, the Granzyme B, and the later diets were measured in adulthood.
A resource of this size is what makes rare and site-specific cancers countable. Liver and intrahepatic bile duct cancers are not everyday diagnoses. Neither is a clean comparison of rectal, lung, prostate, and breast cancers in the same cohort, against an early-life exposure that no living participant can accurately recall gram by gram. Linking 64,761 adult records back to whether those adults’ early childhood fell under rationing is how a policy change from 1953 becomes a paper in PNAS.
The participants are adults. The exposure is not, in the first instance, their current dessert habit, though that habit turns out to differ too. The exposure is whether the first 1,000 days of life happened while Britain still capped sugar near ~40g a day or after the cap came off and intake rose to ~80g. The biobank supplies the later chapters. The ration supplies the first page.
Sharply lower adult cancer rates
People whose early childhood fell under rationing later showed sharply lower adult cancer rates than cohorts born after intake roughly doubled. Sharply lower is the study’s own summary of the overall cancer picture, not a decorative adjective grafted onto a tiny relative risk. The site-specific numbers underneath that summary are more specific still.
Fully exposed participants had about 69% lower risk of liver and intrahepatic bile duct cancers. That is the headline site. Liver cancer and intrahepatic bile duct cancer—cancers of the liver itself and of the bile ducts that run inside it—were the diagnoses with the largest reported reduction among people whose entire early window sat under the ration. A 69% lower risk is not a rounding error. It is the kind of difference that, if it holds, would force pediatric nutrition and cancer epidemiology to talk to each other.
Reductions also appeared for rectal, lung, prostate, and breast cancers. The account used here does not attach a separate percentage to each of those sites. What it does attach is the fact of reductions—plural—across a spread of organs that do not share a single obvious mechanism. Rectal cancer is a disease of the large bowel’s last segment. Lung cancer is a disease of the airway. Prostate cancer is a male reproductive-tract disease. Breast cancer is a female reproductive-tract disease. Finding reductions at all four, on top of the about 69% drop in liver and intrahepatic bile duct cancers, is what makes the early-sugar story look like a systemic early-life imprint rather than a one-organ curiosity.
The comparison is between ration-era early childhood and post-ration cohorts. It is not a claim that sugar is the only cause of these cancers, and it is not a claim that every sweetened spoon in a modern nursery maps onto a 69% change. It is a claim that being fully exposed to the low-sugar first 1,000 days—the ~40g sugar/day world—went with sharply lower adult cancer rates than being born into the ~80g world.
Telomeres, Granzyme B, and a slower biological clock
Cancer incidence was not the only adult mark. The same fully exposed people also showed longer leukocyte telomeres (~2.2 fewer biological aging years). Leukocyte telomeres are the protective caps on the chromosomes of white blood cells. They shorten as cells divide and as organisms age. Longer caps, in this kind of analysis, are read as a younger biological age. The study translates the difference into time: about 2.2 fewer biological aging years among people whose early life sat under rationing.
That is a modest number of years and a large claim about when those years were set. The sugar difference was in the first 1,000 days. The telomeres were measured in UK Biobank adults. The implication is that an early ~40g versus ~80g contrast can still be detected in white-blood-cell chromosome caps decades later, as a gap of roughly 2.2 biological years.
They also showed lower Granzyme B. Granzyme B is an enzyme tied to immune cytotoxic activity—the machinery certain immune cells use to kill damaged or infected targets. The study reports it as lower in the ration-exposed group. It does not, in the facts at hand, spell out a complete mechanistic story from infant sugar to this enzyme to a tumor. What it does is place lower Granzyme B beside longer leukocyte telomeres and sharply lower adult cancer rates, as part of the same adult phenotype that traces back to early sugar.
Taken together, the biology is not only “fewer tumors.” It is a cluster: longer leukocyte telomeres, ~2.2 fewer biological aging years, lower Granzyme B, and a suite of site-specific cancer reductions led by liver and intrahepatic bile duct disease at about 69%, with further reductions for rectal, lung, prostate, and breast cancers.
The taste that lasted
The early-life groups did not only differ in their cells. They also showed decades-later diets with less sugar. People whose early childhood fell under rationing were still eating less sugar as adults, long after the ration was a history-book fact. That finding matters because it splits—and then reunites—the interpretation.
One reading is purely biological: the first 1,000 days imprinted metabolism, immune tone, and telomere dynamics, and those imprints produced later cancer differences regardless of what people ate afterward. Another reading is behavioral: a low-sugar start trained a preference, that preference persisted, and decades-later diets with less sugar are themselves protective. The study, as argued by its Cambridge co-author, does not force a choice. It holds both.
A natural experiment from September 1953 cannot perfectly separate a baby’s developing physiology from that baby’s later grocery list. What it can show is that the same people who spent those 1,000 days at ~40g sugar/day arrived in the UK Biobank with less sugar in their adult diets and with the cancer, telomere, and Granzyme B differences already described. The diet finding is not a footnote. It is one of the facts the authors want infant nutrition policy to notice.
Habit is not a soft variable here. If early sweetness teaches the palate, then the ~80g world after rationing did not only deliver more sugar in the crib. It delivered a template for how sweet food is supposed to taste. The ~40g world delivered a different template. Decades later, those templates were still showing up on the plate. That is a developmental story as much as a metabolic one.
What adult cutbacks can and cannot match
Cambridge co-author Weilong Zhang cautioned that adult cutbacks still help but may not match early-life magnitude. That is the sentence that keeps the paper from being read as fatalism. Cutting sugar in adulthood is still worth doing. Zhang’s caution is that those later cuts may not match the size of the effect tied to the first 1,000 days. The early window, on this evidence, is larger.
He is arguing that both biological imprint and lasting food preferences matter for infant nutrition policy. Both. The biological imprint is the cellular and cancer story: telomeres, Granzyme B, liver and intrahepatic bile duct cancers, rectal, lung, prostate, and breast. The lasting food preferences are the decades-later diets with less sugar. Policy that only lectures adults about dessert, in this framing, arrives after the more powerful window has closed. Policy that only talks about infant feeding without noticing that early tastes persist would miss the other half.
Zhang’s affiliation is Cambridge. His role is co-author. His caution is not a dismissal of adult behavior and not a claim that infant sugar is destiny written in a single gene. It is a magnitude claim: the early-life contrast between ~40g and ~80g looks bigger than what adult restraint typically delivers, and the reasons look like a mix of biology and habit.
The practical residue is uncomfortable in a culture that treats sugar as a willpower problem for grown-ups. If adult cutbacks still help but may not match early-life magnitude, then the highest-yield lever is not another round of adult advice. It is the sugar that reaches a child in those 1,000 days. Zhang’s argument puts that lever in the hands of infant nutrition policy, not only in the hands of individual parents improvising in a supermarket aisle.
What a ration can teach a clinic
British sugar rationing ended in September 1953. That date is the hinge. Before it, a child’s early diet was capped near 40 grams of sugar a day. After it, intake roughly doubled to ~80g. A PNAS analysis of 64,761 UK Biobank adults treats that hinge as a natural experiment on the first 1,000 days of life.
The adult picture on the low-sugar side of the hinge is consistent across several domains. Adult cancer rates were sharply lower. Fully exposed participants had about 69% lower risk of liver and intrahepatic bile duct cancers, with reductions also for rectal, lung, prostate, and breast cancers. They had longer leukocyte telomeres, amounting to about 2.2 fewer biological aging years. They had lower Granzyme B. They still ate less sugar decades later.
None of those facts is a randomized trial of baby formula. None is a license to invent a precise pediatric gram target beyond the historical ~40g and ~80g the ration and its end produced. They are, together, a case that early sugar is not only a dental problem and not only a childhood weight problem. It is a candidate contributor to later cancer, later biological age, and later appetite.
Infant nutrition policy is the audience Zhang is pointing at. If biological imprint and lasting food preferences both matter, then the sugar that reaches an infant in those 1,000 days is a policy object, not only a parenting style. The British ration was never designed as a cancer-prevention program. In this PNAS reading of the UK Biobank, it behaved like one.
The study does not name a safe adult dose that erases an early ~80g start. Zhang’s caution runs the other way: adult cutbacks still help, and they may not match early-life magnitude. For clinicians, feeding-guidance authors, and anyone writing rules for what goes into an infant’s mouth, that is the practical residue of a September 1953 policy change, measured in 64,761 adult lives: the sugar that hits in the first 1,000 days is the sugar that is hardest to outrun.
The fully exposed group is the clearest illustration. Their early childhood sat entirely under rationing. Their later risk of liver and intrahepatic bile duct cancers was about 69% lower. Their other cancer reductions, their longer leukocyte telomeres, their lower Granzyme B, and their decades-later diets with less sugar all point in the same direction. Cohorts born after the doubling to ~80g sit on the other side of that line. Between those two groups lies the argument this PNAS paper is making: baby sugar is not a closed chapter of childhood. It is a chapter that adult cancer rates still appear to be reading.