The number of people newly developing dementia each year could more than double to 24.7 million by 2050, according to a large international analysis that found the burden was already unevenly distributed between countries and between women and men.

The study, published online in JAMA Neurology, estimated that 11.4 million people developed dementia worldwide in 2023. Researchers projected the annual total forward using age- and sex-specific incidence rates and the expected 2050 population, assuming those rates remain constant over time.

The analysis combined individual-level information from 38,418 participants in 22 population-based studies across 18 countries and territories on six continents. The cohorts contributed 287,632 person-years of follow-up, during which 3,001 participants developed dementia based on clinical diagnoses.

Researchers calculated a global age- and sex-standardized incidence of 9.7 new cases for every 1,000 person-years. After adjusting for age, incidence was 10.6 per 1,000 person-years among women and 8.6 among men.

The divide by national income was larger. Low- and middle-income countries recorded a standardized incidence of 13.6 cases per 1,000 person-years, compared with 8.3 in high-income countries. The authors said the findings create a more consistent benchmark for planning prevention, diagnosis and memory-care services.

The work was conducted through the Cohort Studies of Memory in an International Consortium, led by researchers at the University of New South Wales Sydney's Centre for Healthy Brain Ageing. It applied common statistical methods across studies, addressing a long-standing problem in dementia research: countries have often used different diagnostic and analytical approaches, while lower-income populations have been underrepresented.

The projection concerns incidence, meaning new cases arising during a period, rather than prevalence, which counts everyone living with the condition. More than 57 million people currently live with dementia, and earlier research has projected that total could approach 150 million by 2050, UNSW said.

The authors cautioned that their study does not explain why incidence was higher among women or in lower-income countries. The differences may reflect a complicated mix of health, education, economic conditions, environment and access to care. More diverse local research is needed before assigning causes.

The 2050 figure is also a projection, not a fixed prediction. It assumes today's age- and sex-specific incidence rates continue. Better prevention could lower the total, while worsening health conditions or incomplete diagnosis could change the path in either direction.

The crisis will arrive first as a family burden

A number as large as 24.7 million can feel abstract. In practice, every new diagnosis reorganizes a household. Relatives become caregivers, work hours are reduced, savings are redirected and health systems must provide years of support rather than a single course of treatment.

Governments often discuss ageing as a distant fiscal problem. This study shows that the pressure is measurable now and that waiting carries a heavy cost. Clinics, trained specialists, community support, long-term care and protections for unpaid caregivers cannot be created quickly after demand has already overwhelmed them.

The higher measured incidence in low- and middle-income countries exposes a particularly harsh imbalance. Many of the societies facing rapid population ageing have the fewest neurologists, diagnostic services and social-care systems. Families may carry most of the responsibility in private, leaving the scale of the crisis hidden from official budgets until it becomes unavoidable.

Wealthy democracies have their own responsibility. They should invest in research and domestic care capacity, but also help expand reliable data, professional training and affordable diagnosis internationally. A global health strategy that discovers treatments unavailable to most patients would be scientific progress without adequate human delivery.

Prevention cannot eliminate every case, and public messaging must not blame patients. Dementia has many causes, including factors people cannot control. But reducing modifiable risks, identifying cognitive decline earlier and treating cardiovascular or metabolic disease can still improve outcomes at a population level. The honest position is neither fatalism nor a promise that lifestyle choices guarantee protection.

The study's greatest value may be the clarity of its warning. Population ageing is predictable, and the likely demand for care can be estimated decades in advance. Leaders therefore cannot claim surprise when hospitals, families and public finances come under pressure.

A humane response treats memory care as essential infrastructure. It protects dignity, supports caregivers and directs resources toward communities that research has historically overlooked. The projected doubling is not merely a medical statistic; it is a test of whether societies plan for vulnerability before millions of families are forced to face it alone.