How does education bias dementia diagnosis?
Education can skew dementia tests because most cognitive assessments rely on skills like reading, vocabulary, and test-taking experience, which are directly tied to schooling. A 2023 study comparing dementia screening in Indonesia and the UK found that common tests like the Hopkins Verbal Learning Test (HVLT) and the Mini-Mental State Exam (MMSE) had lower sensitivity and specificity for dementia when used across different education levels [2]. This means they missed some cases of dementia and falsely flagged others, especially in people with less education. The study concluded that using culturally and educationally sensitive tests can prevent overestimates of dementia prevalence [2].
A 2024 study that mapped a dementia classification algorithm (the 10/66 algorithm) to US data found that all five algorithms it tested overestimated the education gradient in dementia [1]. In plain terms, these algorithms made it look like having less education was a much stronger risk factor for dementia than it actually is, because they were biased by the cognitive tests they used. This suggests that some of the apparent link between low education and high dementia rates is an artifact of how we measure dementia, not a true difference in brain health.
Does education actually protect against dementia?
Yes, but the evidence is not as strong as you might think. A 2022 study that looked at descriptive data found that people who finished only compulsory schooling had a significantly higher dementia risk compared to those with more education [3]. However, when the same study used two different statistical methods to test whether this was a causal effect (i.e., that education directly reduces dementia risk), the results were negative and not statistically significant [3]. The author concluded that they could not verify a significant causal impact, though the consistent pattern of negative effects suggests that a real effect might exist but was missed due to limited statistical power [3].
A 2021 study on racial and educational disparities in the US found stark differences: at every education level, Black Americans could expect to live more years with dementia than White Americans, and the highest burden was among Black adults without a high school diploma [4]. This shows that education and race combine to shape dementia risk, but the relationship is not simple—education alone does not erase disparities. The study simulated that these differences in life expectancy with dementia translate into large differences in dementia prevalence across race-education groups [4].
So is the dementia epidemic overestimated or not?
It depends on the population and the diagnostic method. In low- and middle-income countries (LMICs), where education levels are lower and cognitive tests are often imported from Western settings, dementia prevalence may be overestimated because the tests are not culturally or educationally appropriate. The 2023 study found that some Indonesian cohorts using standard screening methods reported very high dementia prevalence (29–32% in people over 60) [2], which is far above typical rates in high-income countries. This suggests that part of the 'epidemic' in these settings could be a measurement artifact.
However, in high-income countries like the US, where education levels are higher and tests are better calibrated, the epidemic is not necessarily overestimated. The 2024 study found that the 10/66 algorithm performed well in classifying dementia in the US, with 87% sensitivity and 93% specificity [1], meaning it correctly identified most cases and rarely misclassified healthy people. So the risk of overestimation is not uniform—it is highest where education and culture mismatch the diagnostic tools. The bottom line: better education levels can both protect against dementia and bias our ability to measure it, and the true size of the epidemic depends on how we account for these factors.
About These Sources
This answer is built on 4 studies (3 peer-reviewed, 1 preprint) — published from 2021 to 2024, 1 from 2024 or later, 2 in Q1 journals — selected as the most relevant from 4 studies that passed quality screening, drawn from 51 papers retrieved from a database of over 500 million.
Sources used in this answer
Assessing the 10/66 dementia classification algorithm for international comparative analyses with the United States
In a US study, the 10/66 dementia algorithm had high sensitivity (87%) and specificity (93%) for classifying dementia, but all five algorithms tested overestimated the education gradient in dementia, meaning they made low education look like a stronger risk factor than it actually is.
Diagnostic Paradigms and Prevalence Estimates
Comparing Indonesian and UK cohorts, common cognitive tests like the HVLT and MMSE had lower sensitivity and specificity for dementia across education levels, and using culturally sensitive tests can prevent overestimates of dementia prevalence.
The Effects of Education on Dementia Risk
Descriptive data showed higher dementia risk for those with only compulsory schooling, but two causal analyses found negative and non-significant effects of education on dementia risk, suggesting a possible protective effect that could not be confirmed due to limited statistical power.
Racial and Educational Disparities in Dementia and Dementia-Free Life Expectancy.
In the US, Black Americans at every education level had more years of life with dementia and higher dementia prevalence than White Americans, with the largest disparities between Black adults without a high school diploma and White adults with some college or more.