
Understand the 45% dementia estimate, why risk factors overlap, and how to use population research without mistaking it for a personal forecast.

The widely cited figure is a population attributable fraction, often shortened to PAF. Summaries of the 2024 Lancet Commission report describe a model of how dementia case numbers might change under a hypothetical removal of the relevant risk factors.
The unit being counted is dementia cases across a population. It is not a percentage of one person's brain health or a measure of how well that person has looked after themselves. The figure describes potential change across many lives, not a result assigned to each life.
The estimate also concerns both prevention and delay. Reading it only as a count of cases that could disappear leaves out part of the finding. A careful reading keeps both outcomes in view.
A headline that drops the words population, potentially or delay changes the meaning. The first tells us whose outcomes are being counted. The second signals an opportunity rather than a certainty. The third reminds us that shifting when dementia occurs is part of the estimate.
Population estimates combine how common a factor is with its association with dementia. An individual's odds reflect their own circumstances. The headline figure does not perform that personal calculation.
It would therefore be wrong to describe 45% as the reduction someone should expect after changing their habits. It is also not a starting probability from which someone can subtract points for each healthy choice. Neither use matches what the model estimates.
A factor can make a larger contribution to population totals when many people are exposed to it. That does not automatically make it the most relevant concern for a particular reader. Population reach and personal relevance answer different questions.
The Commission's factors can occur together, and some share pathways or circumstances. Its combined estimate adjusts for that shared contribution. Adding separate percentages without this adjustment could count the same potential contribution more than once.
The factors also differ in how common they are and in the strength of their links with dementia. That means the overall estimate is not a set of 14 equal shares. Removing a factor from a personal checklist cannot reveal how much of the combined figure remains.
The headline concerns dementia cases worldwide. This is a Commission population model, rather than a result from a single treatment trial. The research summaries do not give one sample size, a follow-up duration or a demographic breakdown for that model.
A worldwide estimate and a personal health history are different kinds of information. The year 2024 identifies the report update, not a stated length of follow-up. Worldwide scope should not be mistaken for a detailed description of who contributed data.
What the report summary does describe is a life-course framework. Its factors cover different stages rather than only choices made after retirement:
This framing helps explain why a present-day habit list is not the same as the Commission's model. For someone over 60, the practical question is which current needs deserve attention, not how to rewrite an earlier stage of life.
It is more useful to bring a present concern to a health appointment than to treat an education history or past exposure as a score. That is a practical reading of the model, not an individual calculation.
The model assumes that the relevant risk-factor associations are causal. An association means that a factor and dementia occur together in the data; causation means that changing the factor changes the outcome. The estimate itself does not prove that all the modeled relationships work that way.
It therefore does not establish that all dementia is preventable or that changing a factor guarantees an individual will avoid it. Nor does it explain why a particular person developed dementia. A diagnosis is not evidence that someone failed to make the right choices.
For everyday decisions, the useful response is to match broad guidance to real needs. The World Health Organization's dementia guidance identifies a range of health and lifestyle measures relevant to risk reduction. It provides a basis for discussion with a qualified healthcare professional, rather than a personal prevention formula.
An unresolved concern can be a better starting point than the largest-looking headline. Hearing or vision difficulties, blood pressure, cholesterol and blood sugar are topics raised in the research. Mood concerns or reduced social contact also belong in that conversation.
Personal decisions about assessment or care belong with a qualified professional. The role of this framework is to help identify a topic worth discussing, not to select a treatment.
A hearing concern and a wish to reduce future dementia risk can sit in the same appointment without being the same question. One asks about a current need; the other asks about a future outcome. Keeping those questions separate makes it easier to discuss help now without attaching an unsupported prevention promise to it.
WHO includes physical activity, not smoking and avoiding harmful alcohol use. It also lists a balanced diet and healthy weight. Social and cognitive activity are part of its guidance too.
Managing blood pressure, cholesterol and blood sugar also appears in WHO's guidance. Hearing aids when needed and reducing air-pollution exposure are further areas it identifies. There is no need to turn this range of options into an all-or-nothing personal test.
Instead, the practical aim is to identify a realistic area of support. Readers weighing different claims can use our guide to sorting stronger evidence from early findings to keep the type of evidence in view.
Choosing a focus need not mean dismissing the rest. It means using relevance and feasibility to make the discussion concrete. The framework offers topics for discussion rather than a prescribed order everyone should follow.
If movement, social contact or managing a health condition is difficult, a clinician or another appropriate professional can help discuss realistic options. That conversation can start with what is getting in the way rather than a demand to address every listed factor.
Some topics call for different kinds of support. A concern about cholesterol belongs in a health conversation; reduced social contact may call for a discussion about accessible opportunities. Asking which barrier matters most is more useful than ranking every factor by a population percentage.
For a broader perspective on responsibility, our discussion of risk reduction without personal blame keeps the focus on useful support. The goal is a workable next conversation, not a tally of past choices.
Future clinical research should test whether changing these factors reduces dementia cases or delays their onset, rather than only improving an intermediate health measure. Trials also need to show how combined approaches work when participants have overlapping needs. For adults over 60, useful results would describe who took part, how long they were followed and which outcomes changed. The next advance should be a clearer account of which kinds of support help which people under realistic conditions, so that population estimates can inform care without being mistaken for personal forecasts.
Using a population estimate when weighing health priorities requires context; FitBrainLab explains the research behind those everyday health decisions. For adults 60+ unsettled by alarmist dementia coverage, the publication offers calm accounts of cognitive aging that separate established knowledge from early or uncertain findings.
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