What a child eats in the first two years may echo for decades, according to a large study exploiting an accident of history — the end of Britain’s postwar sugar rationing.
Researchers at the University of Cambridge followed more than 64,000 people born in Britain between 1951 and 1956 using the UK Biobank, publishing in PNAS.
Why this is a natural experiment
The methodological problem in nutrition research is that people who eat differently differ in many other ways. Families limiting sugar tend to be wealthier, better educated and healthier in numerous respects, so any association between diet and outcome is hopelessly entangled with everything else.
The end of sugar rationing in September 1953 broke that entanglement. Children born a few months apart, into otherwise comparable families, experienced very different sugar exposure during the first 1,000 days — conception to age two — for reasons entirely unrelated to their parents’ characteristics.
That is close to random assignment, delivered by policy rather than by researchers. It is why this study can say more than a conventional observational comparison, even though nobody was randomised.
What they found
People with more rationing — less sugar — early in life had lower rates of several cancers, including a 36% lower breast-cancer rate and a 69% lower liver-cancer rate in the analysis.
They also had longer telomeres, a marker of biological ageing, equivalent to roughly 2.2 years slower ageing.
And at age 50 they tended to eat less sugar and have healthier diets — a finding that turns out to be central to interpreting the rest.
Two mechanisms, pulling in different directions
The researchers propose two explanations, and their relative contributions are unclear.
The biological account holds that early sugar exposure affects metabolic and immune development during a period when systems are still being configured, leaving durable physiological differences — the same logic underlying developmental origins of adult disease more broadly.
The behavioural account is simpler and arguably more likely to explain much of the effect. Early exposure shapes taste preferences, those preferences persist, and people who grew up eating less sugar continue to eat less of it. The observation that the low-sugar group still ate better diets at age 50 supports this directly.
The distinction matters enormously for what anyone should do about it. If the mechanism is biological imprinting during a critical window, the opportunity closes at age two. If it is preference formation, the effect operates through seventy years of subsequent eating — which is both more modifiable later and less remarkable as a finding.
How to read the cancer numbers
A 69% lower liver-cancer rate sounds enormous, and relative reductions in rare outcomes routinely mislead. Liver cancer is uncommon, so a large relative difference may rest on a modest number of cases and represent a small absolute difference in individual risk.
Breast cancer is more common, making the 36% figure more substantial in absolute terms — though it is also an outcome influenced by reproductive history, hormones, alcohol and body weight over a lifetime, several of which correlate with diet.
What a natural experiment still cannot do
This is an association study, not a randomised trial, and the researchers say so.
The rationing cutoff was clean, but it was not the only thing that changed in postwar Britain. Rationing of other foods ended around the same period, and the broader economy, food supply and living conditions were changing rapidly. Anyone born after the cutoff experienced a different food environment in more respects than sugar alone.
The UK Biobank also has a well-documented limitation: participants are healthier, wealthier and less diverse than the general population, and everyone in this analysis survived to enrol in their sixties — excluding anyone who died earlier, which could distort the picture in either direction.
Why it still matters
The finding aligns with existing paediatric guidance to avoid added sugar in the first two years, which was already recommended for reasons including dental health, taste development and displacement of more nutritious foods.
What this adds is a long-horizon argument. Advice about infant nutrition is usually justified by near-term outcomes, and demonstrating measurable differences seven decades later — even with all the caveats — makes a stronger case for a period during which parental decisions are highly modifiable.
The honest summary
Strong associations from an unusually well-designed natural experiment, with a plausible mechanism that may be as much behavioural as biological, in a population with known selection issues.
Why the first 1,000 days keeps recurring in research
The window this study examines is not an arbitrary choice, and it has become a standing focus of developmental research for reasons independent of sugar.
Conception to age two is when most organ systems complete their structural development, when the immune system is calibrated by early exposures, when the gut microbiome is established, and when the brain undergoes its most rapid growth. Systems being built are more sensitive to their inputs than systems already running, which is the general principle behind developmental origins of adult disease.
Evidence for that principle comes largely from historical accidents like this one — famines, rationing schemes, policy changes that divided otherwise similar populations. Those natural experiments are the only ethical way to study the question, which is why researchers keep returning to them, and why their limitations are shared across the whole literature rather than specific to this paper.
That is genuinely useful evidence, and it is not proof that limiting sugar before age two prevents cancer. This summarises observational research and is not medical advice.