Global dementia study: about 10 new cases a year per 1,000 followed, estimate higher in poorer countries

Estimates of how many people have dementia are easy to find. Estimates of how fast new cases arise, drawn from the same kind of data in rich and poor countries alike, are not. An international team has now pooled individual records from 22 long-running studies in 18 countries and arrived at a single figure: about 10 new cases of dementia a year for every 1,000 people followed.

The study, published on October 12 in JAMA Neurology, also reports higher rates in women than in men and in low- and middle-income countries than in high-income ones. Both gaps come with ranges that overlap, so neither is settled. And 22 studies, however carefully combined, cannot stand in for every country on six continents.

Why a global figure for dementia incidence was missing

Dementia means a decline in memory and thinking severe enough to interfere with daily life. Two different numbers describe how common it is. Prevalence is the count of everyone living with the condition at one time. Incidence is the rate at which new cases appear. Incidence is harder to measure, because it takes following the same people for years.

Most of that following has been done in wealthy countries. The new paper opens with the problem: global dementia incidence estimates are limited by underrepresentation of low- and middle-income countries and heterogeneous analytic methods. Heterogeneous means varied. Studies that define dementia differently, or analyze their data differently, cannot simply be added up.

The gap matters because of where dementia is concentrated. The authors of a 2019 analysis wrote that around two thirds of people with dementia worldwide live in low-income and middle-income countries. Such countries are often grouped under the label LMICs. Prevention planning leans on estimates of how much dementia is tied to risk factors that can be changed. The same paper noted that most data to calculate these risk factors come from high-income countries only.

Reviews of the field have run into the same wall. One systematic review of trends over time searched the literature to the end of 2017. In total, 10,992 articles were identified and 43 retained. Only one of the 43 reported on a low- or middle-income country, using data from Nigeria.

A pooled study of European and American cohorts in 2020 set out its observations and ended by calling for work on determining their validity in geographically and ethnically diverse populations.

How the global dementia study pooled its data

The new analysis comes from a consortium known as COSMIC, short for Cohort Studies of Memory in an International Consortium. Ashleigh Vella is the first author, followed by a long list of investigators from the member studies. In an earlier description of the consortium, its organizers said the aim was to study dementia risk factors more equitably across high-income countries (HICs) and low- and middle-income countries (LMICs).

A cohort is a group of people followed over time. The cohorts used here began enrolling people at various dates: baseline recruitment commenced between January 1990 and December 2011 and follow-up data were collected through the end of 2021. To be included, a study had to have diagnosed dementia by clinical assessment. In the paper’s wording, the analysis covered regions where data on clinically based dementia diagnoses were available.

That left individual-level records from 22 population-based longitudinal cohorts, with 38,418 participants. Longitudinal means that the same people were assessed repeatedly. The studies were spread widely, spanning 18 countries or territories across 6 continents.

Working from individual records, instead of from each study’s published totals, let the team apply one method to everyone. They counted person-years at risk. One person followed for ten years contributes ten person-years, as do ten people followed for one year each. They corrected statistically for participants who dropped out. Then they calculated rates for each age and sex and combined them.

Because the cohorts differ in their mix of ages, raw rates would not be comparable. So the rates were standardized to the 2023 global age-sex population structure. Standardized means recalculated as if every group had the same age and sex make-up, here that of the world population in 2023.

What the study found on dementia incidence

Across 287,632 person-years at risk, 3001 participants developed dementia.

The standardized rate was 9.7 new cases per 1,000 person-years, with a range from 7.89 to 11.97. The range is the span within which the true rate probably lies, given the uncertainty in the data. It does not describe the spread between studies or countries.

In plain terms, about 10 of every 1,000 participants would be expected to develop dementia in a year of follow-up. That is a rate among people enrolled in studies of aging, with the balance of ages and sexes adjusted to match the world’s. It does not describe a population of every age, children and young adults included.

By sex, the rate was 10.6 per 1,000 person-years in women, with a range from 8.61 to 13.11. In men it was 8.6, with a range from 6.94 to 10.72.

By national income, the rate was 13.6 in low- and middle-income countries, with a range from 9.67 to 19.24. In high-income countries it was 8.3, with a range from 6.57 to 10.54.

In both comparisons the ranges overlap. The women’s range and the men’s share the stretch from 8.61 to 10.72. The two income groups share a narrower stretch, from 9.67 to 10.54. The paper reports each rate as higher. Overlap of that kind means the true gap could be smaller than the headline numbers suggest, and the range for poorer countries is especially wide.

These averages hide the largest influence of all, which is age. The 2020 study of seven cohorts in Europe and the United States found the incidence rate ranging from about 4 per 1,000 person-years in individuals aged 65-69 years to 65 per 1,000 person-years for those aged 85-89 years. A single global figure of about 10 sits between those extremes because it blends the ages the cohorts covered.

Dementia incidence in low- and middle-income countries

A higher rate in poorer countries runs against much of the older literature. A 2012 study in the Lancet, a British medical journal, began by observing that most earlier studies from countries with low or middle incomes report rates that are substantially lower than those from countries with high incomes.

That study, by a network called the 10/66 Dementia Research Group, suspected the low rates were partly an artifact of how dementia was defined. It followed people aged 65 and older in Cuba, the Dominican Republic, Venezuela, Peru, Mexico and China. Of 12,887 interviewed at the start, 8,137 who were free of dementia were seen again three to five years later.

The answer depended heavily on the definition. Using criteria the group had designed to work across cultures, incidence ran from 18.2 to 30.4 per 1,000 person-years across the sites. Using a stricter standard set of criteria, it ran from 9.9 to 15.7. The first set of rates was 1.4 to 2.7 times the second.

The 10/66 authors also gave a reason to expect more dementia in poorer settings. Cognitive reserve is the idea that education and mentally demanding work build up a capacity that lets the brain cope with damage for longer. Where schooling has been scarce, they reasoned, a higher incidence of dementia might be expected in these settings than in high-income countries.

The 2019 analysis of risk factors pointed the same way. It estimated the share of dementia tied to nine modifiable risk factors, such as less childhood education, high blood pressure, hearing loss and smoking. Modifiable means open to change. Worldwide that share was put at 35%. It was 39.5% in China, 41.2% in India and 55.8% in a Latin American sample. The authors concluded that the dementia prevention potential in India, China, and this sample of Latin American countries is large, and greater than in high-income countries.

So the new estimate of 13.6 against 8.3 fits a line of reasoning that predates it. It does not prove that reasoning. Differences in how each cohort identified dementia could also contribute.

Dementia incidence in women and men

More women than men live with dementia. A global forecast published in The Lancet Public Health in 2022 estimated that there were more women with dementia than men with dementia globally in 2019, by a ratio of about 1.7 to 1. Much of that is survival: women live longer, and dementia is a disease of old age.

Whether women develop it at a higher rate at the same age is a separate question, and earlier pooled studies mostly said no. The European and American analysis described rates at each age as notably similar between men and women. A 2023 review that assembled 205 studies with close to a million participants in 43 countries concluded that globally, there are no sex differences in age-specific dementia incidence. The one exception it found was at age 90 and over, where women’s rates were higher.

That review’s authors raised a caution that applies here as well. As women live longer than men, this may lead to higher rates of dementia observed among women, they wrote. In other words, a difference in survival can pass for a difference in risk.

The new paper’s figures, 10.6 for women and 8.6 for men, are standardized by age. They lean toward a higher rate in women, with overlapping ranges. They are best read as one more data point in a question that remains open.

What the 2050 dementia projection assumes

The team carried its rates forward. Applying them to the world population expected in 2050 gave 24.7 million new cases in that year. The paper attaches a condition: the projection was made assuming constant incidence over time.

That figure counts new cases in a single year. It is a different quantity from the 2022 forecast of people living with dementia, which begins from the premise that the number of people with dementia is expected to increase as populations grow and age. That forecast went from 57.4 million people in 2019 to 152.8 million in 2050. It found the drivers varied by region, with population growth contributing most to the increases in sub-Saharan Africa.

The assumption of constant rates is a real limit, because rates have been moving. In the seven European and American cohorts, the incidence rate of dementia declined by 13% per calendar decade between 1988 and 2015. Whether anything similar is happening elsewhere is unknown, for lack of data. If rates fall, 24.7 million will prove too high. If the risk factors that are more common in poorer countries spread, it could prove too low.

What the global dementia estimates cannot show

Twenty-two studies for the world. The cohorts span 18 countries or territories. Most of the world’s countries are not among them, and a cohort drawn from one city or region may not represent its whole country.

Overlapping ranges. The gaps between women and men and between income groups are estimates with wide margins. The low- and middle-income figure, from 9.67 to 19.24, is the least certain.

Diagnosis across settings. Every study used a clinical diagnosis, but clinical practice and assessment tools differ between sites. The consortium itself has described developing methods to reliably assess dementia from limited data in under-resourced regions as work in progress.

Older cohorts. Recruitment ran from 1990 to 2011. People who entered a study three decades ago grew up with different schooling, diets and medical care from those reaching old age now.

Who joins and who stays. People who volunteer for long studies, and who remain in them, tend to be healthier than those who do not. The analysis corrected for drop-out statistically, which reduces that problem without removing it.

A projection, not a forecast. The 2050 number holds today’s rates fixed and changes only the population.

All-cause dementia. The figures combine every type of dementia. They say nothing about Alzheimer’s disease or any other cause on its own.

The authors present the work as a benchmark. In their words, the harmonized estimates provide a comparable benchmark for global dementia incidence and inform planning for prevention, diagnosis, and memory-care services.

Whether memory or thinking changes in a particular person point to dementia is assessed by their doctor.

The figure of about 10 per 1,000 is a starting line drawn from the places that have data, and most of the world’s countries are not yet among them.