The research team analyzed data from 38,414 older adults recorded in the National Alzheimer’s Coordinating Center database, a large U.S. research resource that brings together clinical, cognitive and imaging information from Alzheimer’s Disease Research Centers. Participants contributed information from the NACC Uniform Data Set, while a subset also had magnetic resonance imaging data from the NACC MRI Data Set. The investigators classified each participant according to ten binary risk factors: hypertension, diabetes, hypercholesterolemia, alcohol misuse, smoking, depression, obesity, hearing loss, vision loss and low education. “Binary” means that each factor was counted as present or absent, rather than being assigned a detailed severity score. The researchers then added the factors to create a cumulative index. A person with none of the listed risks received a score of zero, while someone with all ten received a score of ten. This approach does not imply that every risk has identical biological effects; instead, it tests whether the overall accumulation of risks is related to brain health more consistently than isolated factors are.

The biological logic behind the index is straightforward. Blood vessels in the brain are extraordinarily small and tightly regulated, and they must deliver oxygen and nutrients while preserving the blood–brain barrier. Long-term high blood pressure can damage vessel walls and disrupt their ability to control blood flow. Diabetes can expose vessels and neurons to abnormal glucose levels and inflammation, while high cholesterol is associated with broader vascular disease. Smoking and harmful alcohol use can add toxic and cardiovascular stress. Obesity and depression may influence the brain through metabolic, inflammatory, hormonal and behavioral pathways. Hearing and vision loss can reduce sensory input, increase social isolation and make everyday cognitive tasks more demanding. Education is different from the other factors: it is not a disease or lifestyle exposure, and a low level of formal education cannot be treated as an individual failing. In dementia research, education is often used as a marker related to cognitive reserve, the brain’s capacity to tolerate pathology while maintaining function. By combining these heterogeneous factors, the study captures a broad profile of vulnerability rather than a single disease mechanism.

The outcomes covered both cognition and brain structure. Participants were assessed for global cognitive performance, clinical cognitive status and disease severity, as well as delayed recall and semantic fluency. Delayed recall tests the ability to retain and retrieve information after a time interval, a function closely linked to medial temporal-lobe structures that include the hippocampus. Semantic fluency asks people to generate words belonging to a category, such as animals, within a limited period; the task draws on language, memory retrieval and executive control. The researchers also examined three MRI markers: white matter hyperintensities, cerebral infarcts and hippocampal atrophy. White matter hyperintensities appear as bright regions on particular MRI sequences and are commonly associated with small-vessel disease, though their underlying tissue damage can vary. Infarcts are areas of brain injury caused by interrupted blood supply, including lesions that may have occurred without a dramatic recognized stroke. Hippocampal atrophy refers to reduced volume in a structure central to memory formation and retrieval and often implicated in neurodegenerative disease.

Across the analysis, a higher cumulative risk score tracked with poorer cognitive outcomes. Every additional risk factor was associated with worse global cognitive status, greater clinical severity, lower delayed-recall performance and lower semantic fluency. The same stepwise pattern appeared in the imaging data: accumulating risks were associated with higher odds of white matter hyperintensities, cerebral infarcts and hippocampal atrophy. This matters because vascular and neurodegenerative changes are often discussed as though they belong to separate categories. In reality, they can coexist and potentially amplify one another. Damaged small vessels may reduce the brain’s energy supply, impair waste clearance or weaken networks that connect distant regions. Vascular injury can also reduce the brain’s resilience when protein pathology or age-related neuronal loss is present. Meanwhile, degeneration of the hippocampus may undermine memory even when vascular lesions are modest. The study’s central message is therefore not that vascular disease explains every case of cognitive decline, but that the cumulative burden of vascular, sensory, psychiatric and educational factors is reflected in multiple dimensions of late-life brain health.

Among the MRI measures, hippocampal atrophy showed the largest indirect pathway linking cumulative risk burden to cognition. In statistical mediation analysis, an indirect pathway is a relationship in which one variable is associated with an outcome partly through an intermediate measure. Here, the researchers tested whether structural brain markers could account for part of the association between the risk score and cognitive performance. The result suggests that hippocampal shrinkage may be especially important in connecting accumulated risk exposure with poorer cognitive function. White matter hyperintensities and infarcts also showed indirect pathways, but they were smaller in the evaluated models. These findings should not be interpreted as proof that risk factors physically shrink the hippocampus in a simple linear chain. Mediation in cross-sectional data cannot establish timing, and the same unmeasured factors may influence risk exposure, brain structure and cognition simultaneously. Still, the pattern is biologically plausible: vascular dysfunction can affect the hippocampus, which has a demanding metabolic profile and a vulnerable blood supply, while white matter injury can disrupt communication between memory-related regions and wider cognitive networks.

The study also highlights why “brain health” cannot be reduced to a scan or a memory test. A bright spot on an MRI does not automatically equal dementia, and hippocampal atrophy can reflect several processes, including normal aging, vascular injury and neurodegeneration. Conversely, people with substantial brain pathology may continue to function well for years, in part because of cognitive reserve, social support and compensatory brain networks. Education may contribute to that reserve, although the researchers treated low education as one of the measured risk factors rather than as a protective mechanism that can be summarized by a single number. Hearing and vision are similarly complex. Treating sensory loss may improve communication and daily functioning even if it does not reverse established brain lesions. Depression can affect concentration and test performance directly, while also sharing biological and social pathways with dementia risk. These overlapping mechanisms help explain why the association between a larger risk burden and poorer cognition emerged across several tests rather than in one narrow domain.

The authors’ conclusions align with a growing shift in dementia prevention research toward multidomain intervention. Earlier studies, including the FINGER randomized trial, have tested combinations of exercise, dietary guidance, cognitive training and monitoring of vascular risk, rather than relying on a single treatment. The new analysis does not test an intervention and cannot show that lowering the risk score will prevent dementia. It does, however, offer a large-scale snapshot of how clustered risks correspond to both vascular lesions and neurodegenerative-appearing changes. That distinction is crucial. People cannot alter their educational history, genetic background or every aspect of aging, and not every risk factor is equally controllable. Yet blood pressure, diabetes, cholesterol, smoking, harmful alcohol use, obesity, depression and untreated sensory impairment are all potential targets for clinical care or public-health action. Addressing them may benefit the heart, blood vessels, mobility, mood and independence as well as the brain.

The analysis has limitations that temper its headline-grabbing implications. Because the data were cross-sectional, the researchers could not determine which risk factors came first, how long participants had been exposed to them or whether changes in risk preceded changes in brain structure. The NACC database is an invaluable research resource, but people enrolled through Alzheimer’s Disease Research Centers may not perfectly represent the wider older population. The simple yes-or-no scoring system also treats risks as equal units, even though duration, severity and timing probably matter. A decade of poorly controlled hypertension is not biologically identical to a recently diagnosed case, and hearing loss may have different consequences depending on access to effective treatment. MRI availability may further narrow the imaging sample. Finally, associations can be shaped by confounding factors that were not fully captured. Even with these caveats, the size of the dataset and the consistency across cognitive and imaging outcomes make the results difficult to dismiss. The message is both sobering and hopeful: accumulated risk is visible in the aging brain, but several components of that burden are potentially addressable long before severe cognitive impairment appears.

The investigative team, led by Prof. Iris Shai from Ben-Gurion University of the Negev and Reichman University, alongside collaborators from Harvard University, Leipzig University, and Tulane University, meticulously tracked 533 middle-aged men and women who participated in controlled dietary intervention trials spanning more than a decade. Repeated MRI assessments quantified visceral fat accumulation alongside detailed neuroanatomical measurements and comprehensive cognitive evaluations using the Montreal Cognitive Assessment (MoCA).

Unlike generalized obesity metrics such as body mass index (BMI), which showed no predictive power for brain structural changes, the study zeros in on visceral adiposity as a specific and biologically relevant marker influencing cerebral integrity. Participants exhibiting lower cumulative visceral fat consistently demonstrated preservation of crucial brain structures—including total brain volume, cortical gray matter, and the hippocampus—regions integral to memory formation and cognitive resilience. Correspondingly, there was a deceleration in ventricular enlargement, an established hallmark of neurodegeneration.

Underlying these associations, the research implicates glycemic control and insulin sensitivity as central mechanistic mediators. Elevated fasting glucose and HbA1c levels emerged as significant predictors of accelerated brain atrophy, whereas traditional blood lipid profiles and inflammatory markers bore no such correlation. This metabolic signature supports a model wherein insulin resistance and chronic dysregulation of glucose metabolism compromise cerebral perfusion and blood-brain barrier integrity, precipitating selective gray matter loss and hippocampal vulnerability.

The study’s rigorous design, incorporating repeated MRI measurements alongside cognitive testing, allowed for the unprecedented characterization of cumulative visceral fat exposure over time. Notably, reductions in visceral fat achieved during an 18-month dietary intervention forecasted brain structural preservation even 5 to 10 years later, independently of overall weight loss. This critical finding delineates a clear path for targeted interventions focusing on visceral adiposity rather than general weight metrics.

Dr. Dafna Pachter, the study’s first author, emphasized the limitations of relying solely on weight as a biomarker for metabolic health, highlighting that modest reductions in visceral fat can yield substantial neuroprotective effects. This insight calls for a paradigm shift in clinical and public health approaches, advocating for more precise, MRI-informed assessments of abdominal fat and metabolic status.

The longitudinal nature of the study further solidifies the causal inference between visceral adiposity and neurodegeneration. Subgroup analyses involving three separate brain MRI scans over five years revealed that individuals with persistently elevated visceral fat experienced notably accelerated hippocampal volume loss and ventricular expansion. Importantly, this effect was not paralleled by subcutaneous fat compartments, underscoring the unique pathogenic role of visceral fat in brain aging.

The ramifications of these findings are far-reaching. Identifying visceral fat reduction—and by extension, optimized glucose metabolism—as modifiable targets in midlife opens novel preventative avenues for dementia and age-related cognitive decline. Unlike irreversible genetic predispositions, these metabolic factors offer actionable intervention points amenable to lifestyle modification and clinical therapies.

This seminal research, published in the high-impact journal Nature Communications, represents the most extensive and longitudinal MRI-based investigation to date connecting specific fat depots to brain aging trajectories. The integration of advanced imaging, rigorous dietary interventions, and comprehensive metabolic profiling delivers a multidimensional perspective essential for future neuroprotective strategies.

From a methodological standpoint, the use of repeated, high-resolution MRI scans provided unparalleled precision in quantifying both cerebral and abdominal adiposity changes over time. The combination of neuroimaging biomarkers—including total brain volume, gray matter, hippocampal occupancy score, and ventricular metrics—with cognitive testing furnishes a robust framework for monitoring disease progression and intervention efficacy.

Moreover, the study’s multinational collaboration harnessed expertise across epidemiology, clinical nutrition, neurology, and metabolic research. This interdisciplinary approach was critical in teasing apart the complex interplay between metabolic health and brain integrity, positioning the findings at the forefront of translational neuroscience.

Prof. Iris Shai envisions these findings driving a new clinical focus on glycemic control and visceral fat as measurable, modifiable determinants of brain aging, emphasizing the potential to slow degeneration and diminish the risk of cognitive decline. As the global population ages and the burden of neurodegenerative diseases escalates, this research offers hope for impactful prevention rooted in metabolic health optimization.

The funding for this groundbreaking endeavor came from prominent institutions, including the German Research Foundation’s LeiCeM center of excellence and the ERA-4-Health initiative, underscoring the initiative’s international significance and the scientific community’s commitment to confronting brain aging through innovative metabolic research.

In summation, this groundbreaking study substantiates the pivotal role of visceral adiposity in brain aging and provides compelling evidence that targeted reduction of abdominal fat can attenuate neurodegeneration and enhance cognitive function in midlife. These findings herald a transformative era in preventive neurology—one where metabolic health interventions become integral to preserving brain vitality across the lifespan.

Subject of Research: People

Article Title: Sustained visceral fat loss is associated with attenuated brain atrophy and improved cognitive function in late midlife

News Publication Date: 26-Mar-2026

Web References: https://doi.org/10.1038/s41467-026-71141-4

Image Credits: Credit: Nir Slakman

Keywords: Fat storage, Metabolism, Dietary counseling, Magnetic resonance imaging