
Northumbria's five-factor dementia-risk score targets research in lower-income settings. Read about its evidence, intended use and limits.

On November 2, 2026, Northumbria University announced that an international team it led had developed and validated a five-factor dementia-risk score for low- and middle-income countries. The university said the study was published in the American Journal of Epidemiology.
Northumbria says the score is primarily intended to support research and recruitment into future studies and trials. It is not designed for routine screening of the general population.
The work addresses a substantial global need. The World Health Organization reports that 57 million people were living with dementia worldwide in 2021. More than 60% lived in low- and middle-income countries.
Those figures give the study's setting a clear purpose. Developing a tool for research in lower-income settings addresses where much of the global dementia burden lies, rather than treating those settings as an afterthought.
Professor Dame Louise Robinson of Newcastle University said dementia-risk tools have largely come from wealthier countries. Her view, quoted in Northumbria's announcement, helps explain the study's focus on building a model for different settings.
In the same announcement, Professor Blossom Stephan of Curtin University called the study a “valuable foundation” for more inclusive dementia-prevention research. She said further validation across diverse regions is needed.
A 2026 systematic review in GeroScience provides useful context for that next step. It found that other dementia-risk scores varied in predictive performance by score, age group and follow-up period. It also advised paying attention to how closely the intended users resemble the population in which a score was validated. The review did not independently establish the new Northumbria model's performance.
According to Northumbria, the team developed and tested the model using data from more than 11,000 older adults across seven sites on three continents. The sites were in China, Cuba and the Dominican Republic. They also included Puerto Rico, Venezuela, Mexico and Peru.
The university identifies five inputs:
The social-engagement information includes whether someone lives with others, sees friends or family, or takes part in community or religious groups. Physical activity includes everyday movement alongside leisure activity, rather than referring only to planned exercise.
Northumbria says the information can be gathered through conversation with an older adult. A family member or carer can contribute where relevant. No blood tests, brain scans or specialist equipment are needed for the information collection described.
The score ranges from 0 to 9. The university reports an estimated three-year dementia risk of about 1% for a score of 0 and nearly 19% for a score of 9.
Northumbria credits collaborators across the 10/66 Dementia Research Group network. It says funding came from the National Institute for Health and Care Research using UK international development funding, with additional support from Dementia Australia.
The immediate development is a research option built from data gathered in the settings described. Dr Eduwin Pakpahan of Northumbria said the model could help identify people for future studies and trials. He described it as a step toward addressing the lack of risk-prediction tools built for lower-income settings.
The practical feature is not simply that the score is short. It uses information that Northumbria says can be collected without specialist testing equipment. For teams planning research, that makes the type of information required a central part of the model's appeal.
For a research team, the questions raised by this design are concrete: Does the model match the people a study hopes to recruit? Is its three-year focus suited to the research question being asked? Those are questions about study design, not tasks for an older adult to work through alone.
There is also a useful lesson for anyone reading health news: ask which decision a tool is meant to support. In this case, the stated decision concerns identifying potential research participants. That is the practical purpose against which the announcement should be judged.
For adults following brain-health research, this is a chance to look beyond the equipment mentioned in a headline. The central issue is whether a method fits its intended task and setting. FitBrainLab's discussion of brain-health choices beyond blood testing offers a related perspective on keeping tests in context.
The reported risks are estimates from a model, not diagnoses or personalized forecasts for every individual. That distinction matters when reading risk factors without treating them as fate. The score is not presented as a self-screening method or a replacement for clinical assessment. Personal concerns about memory or daily functioning belong in a discussion with a qualified health professional.
Northumbria also reports a model accuracy score of 0.75 and describes 0.70 as a threshold regarded as clinically useful. The release does not provide enough detail to independently assess what “accuracy” means here or how that threshold applies to this model. The figure should therefore not be converted into a percentage of people correctly diagnosed.
The five factors are selected predictors, not a complete list of influences on dementia risk. A predictor is information used to estimate an outcome. A prevention recommendation asks a different question: whether changing something changes that outcome.
Northumbria's report does not show that changing any one factor guarantees prevention or explain how the score should guide personal treatment decisions. The list combines health information with social circumstances and difficulty with a daily task. It should not become a blame-based checklist for older adults or their families.
The lasting value of this development may lie less in the simplicity of the score than in the question behind it: whose lives are represented in dementia research? The WHO figures make lower-income settings central to that question, rather than peripheral to it.
Northumbria's work offers a concrete example of building a risk model with those settings in mind. Its attention to information available through conversation gives the study a clear practical focus. That is a meaningful direction for cognitive-health research because useful knowledge should be relevant to the people it aims to serve.
The key takeaway is the value of more inclusive research design. Progress in cognitive health is not only about more advanced technology. It is also about asking useful questions in the communities that need better evidence.
A short dementia-risk score is easy to describe but needs careful context before its meaning is clear. FitBrainLab addresses the gap between a study headline and its evidence through research-led explanations for readers aged 60 and older, helping them judge claims about memory without alarm or false certainty.
Follow FitBrainLab for research-led insights on memory, focus, brain aging, nutrition and mental fitness after 60. Stay connected for new articles, practical guidance and ideas for a sharper, more engaged life.



Build habits that support memory, focus and a curious, connected life.
Read the Blog