A UK study published in Neurology finds that limiting exposure to added sugar in the womb and during the first two years of life is associated with a lower risk of dementia decades later. The researchers examined adults born around the end of wartime sugar rationing in Britain and linked early‑life sugar restriction with delayed dementia onset, healthier brain structure and better cognitive test performance in their 60s.
Early-life sugar restriction and dementia risk
The study explored whether nutritional conditions in the earliest stages of development could influence how the brain ages, beyond what is already known about middle‑ and later‑life risk factors for cognitive decline. Dementia, including Alzheimer’s disease and vascular dementia, results from progressive brain changes that affect memory and daily function.
The researchers used a historical natural experiment created by the end of wartime sugar rationing in September 1953. The period of rationing had kept added sugar and sweets tightly controlled, while total calories and basic nutrients were maintained. When rationing ended, national sugar consumption rose sharply, offering a unique way to study long‑term health effects tied to early sugar exposure.
The work builds on a line of investigations leveraging this policy shift to examine how early dietary conditions relate to later health outcomes, including metabolic and respiratory health, and mood-related measures. In this latest study, the focus was dementia risk, brain structure and cognitive performance in older age.
The research team was led by Jiazhen Zheng, an assistant professor at the School of Life and Health Sciences at Shenzhen University of Advanced Technology in China, who also directed related work on the study of lung health. Co‑corresponding author S. W. Ricky Lee and colleagues at The Hong Kong University of Science and Technology (Guangzhou) contributed to the project.
“While growing evidence suggests that early-life nutrition can shape long-term health, remarkably little is known about whether sugar exposure during this period is associated with dementia many decades later,” Zheng said. “We were particularly interested in addressing this knowledge gap because most dementia prevention research focuses on risk factors in middle and older age. The end of sugar rationing in the United Kingdom in September 1953 provided a unique natural experiment to investigate this question.”
The UK Biobank provided the data, comprising adults born between 1951 and 1956. After applying several exclusions, the final sample included 64,737 participants who were on average 54.6 years old at recruitment and just over half were women.
Participants were grouped by how long they experienced sugar rationing in the first 1,000 days—from conception through the first two years of life—based on their birth dates relative to the 1953 policy change. About 40,963 people had rationing exposure: 10,682 in the womb only, 15,025 in the womb plus up to one year after birth, and 15,256 in the womb plus one to two years after birth. A further 23,774 participants were born after rationing ended and grew up in an environment of unrestricted sugar.
Researchers tracked dementia diagnoses through hospital records and death registries up to mid‑2023, adjusting for age, sex, race, birthplace, birth month, inflation‑adjusted food prices, maternal smoking, breastfeeding, and parental dementia history. They also accounted for adult socioeconomic factors and health behaviours such as smoking, alcohol use, physical activity and chronic diseases.
In a subset of over 7,500 participants who had MRI brain scans in their early 60s and approximately 6,400 to 20,600 who completed cognitive tests, the team examined brain volume, grey matter, white matter, the hippocampus and white matter hyperintensities, which reflect small vessel damage and tissue wear.
The investigators also tested whether adult type 2 diabetes and high blood pressure explained the link to dementia and checked two conditions unlikely to be affected by early sugar exposure—type 1 diabetes and melanoma—as negative controls. They then looked for patterns that would support a consistent story across clinical and brain‑imaging data.
Over the follow‑up, 504 of 23,774 unexposed participants were diagnosed with all‑cause dementia, including 227 with Alzheimer’s disease and 178 with vascular dementia. Raw figures showed higher dementia counts in the rationed groups because those individuals were older on average; after adjusting for age and other factors, the association pointed to a lower risk of dementia for those exposed to sugar rationing during the womb and early life.
Compared with those never exposed to rationing, people exposed during the womb plus their first year of life had a 21 per cent lower hazard of all‑cause dementia and a 23 per cent lower hazard of Alzheimer’s disease. Those exposed from conception through ages one to two showed a 23 per cent lower hazard of all‑cause dementia and a 28 per cent lower hazard of Alzheimer’s disease.
For vascular dementia, the estimates trended in the same direction but did not reach statistical significance. In analyses of when dementia onset occurred, exposure through ages one to two was associated with delaying all‑cause dementia by about 2.55 years, Alzheimer’s by about 2.87 years and vascular dementia by about 2.49 years.
“The magnitude of the associations was particularly noteworthy,” Zheng said. “These are population‑level estimates rather than predictions for any individual, but even modest shifts in dementia occurrence could have meaningful public health implications given the growing global burden of the disease.”
An additional analysis accounting for competing risks of death attenuated the results somewhat: the sugar‑ration exposure was still linked to roughly a 15 per cent lower likelihood of all‑cause dementia, but associations with Alzheimer’s and vascular dementia were no longer statistically significant.
The link appeared stronger in women than in men, a difference the authors suggested might relate to biological differences in how female neural circuits respond to early nutritional environments. The findings held regardless of whether participants carried APOE gene variants associated with higher Alzheimer’s risk. Mediation analyses indicated that lower adult rates of type 2 diabetes and high blood pressure explained about 18.0 per cent and 11.6 per cent of the association, respectively (25.5 per cent when combined), while birth weight accounted for only 2.7 per cent. The results were specific to dementia risk and did not apply to type 1 diabetes or melanoma, as expected. A validation analysis in an independent British aging study, with an American cohort for comparison, supported the primary patterns.
The brain imaging and cognitive data aligned with the clinical findings. Those who experienced early‑life sugar rationing had slightly bigger total grey matter volumes and marginally smaller white matter hyperintensities. Specifically, grey matter averaged 618.9 cubic centimetres in the rationed group versus 618.5 in the unexposed group, an adjusted difference of 3.27 cubic centimetres. White matter hyperintensities averaged 3.72 versus 3.84 cubic centimetres, an adjusted reduction of 0.64 cubic centimetres.
On cognitive testing, the rationed group performed better on measures of complex processing speed, verbal and numerical reasoning, nonverbal pattern reasoning and a timed attention task. Differences in total brain volume and several other cognitive measures were not statistically significant. The authors described the differences as modest, noting that the grey‑matter increase represented about half a per cent of total grey matter volume and cognitive gains were similarly small.
“What struck us most was how strongly early‑life nutritional exposure appeared to be associated with brain health more than half a century later,” Zheng explained. “Beyond dementia diagnoses, we also observed favourable differences in brain MRI measurements, including greater grey matter volume and fewer white matter hyperintensities, alongside better performance on certain cognitive tests.”
The study’s authors emphasised that the findings show associations, not causation, and cautioned against dietary overreach. They noted that the observational design meant direct measurements of maternal or infant sugar intake were unavailable, and that the rationing exposure was defined by birth date rather than individual consumption. They also highlighted that the UK Biobank cohort tends to be healthier, wealthier and predominantly white, which may limit broader applicability. The researchers suggested randomized trials to explore whether practical nutritional guidance to limit added sugars during pregnancy and infancy could bolster brain development and long‑term cognitive resilience.
“The key message is not that people should eliminate sugar entirely, but rather that limiting excessive added sugar during these sensitive developmental periods may be beneficial,” Zheng added. “Importantly, our findings demonstrate associations, not definitive causation.”
Lead author Zheng concluded that dementia prevention may require a life‑course perspective. “We often think of dementia as a disease of old age, but some of the factors associated with vulnerability or resilience may originate decades before symptoms appear,” he said. “Understanding these early influences could open new opportunities for preventive research and public health strategies.”
The study, “Association of Sugar Restriction in Utero Through Age 2 Years on Dementia Risk Later in Life,” lists Jiazhen Zheng, Gregory Y.H. Lip, Zhuoni Zhang and S. W. Ricky Lee as authors.
