Dementia Risk Varies Drastically by Country, Global Study of 214,000 Finds

Dementia prevention cannot rely on a single global playbook, according to a sweeping new analysis of risk factors across 14 countries and regions. The study, presented at the Alzheimer's Association International Conference and published in Lancet Healthy Longevity, analyzed data from 214,251 older adults enrolled in 11 long-running aging studies collected between 2009 and 2023. It represents the most detailed international comparison of dementia risk profiles to date.

The findings reveal stark regional differences. Low education was a dementia risk factor for 85.6% of older adults in China compared with just 12% in the United States. Obesity was prevalent as a risk factor in 44.9% of Americans but only 13.3% of people in India. These contrasts suggest that prevention programs must be calibrated to local conditions rather than applied uniformly across populations.

Despite these differences, certain risk clusters emerged consistently across countries. Cardiovascular risks — including high cholesterol, hypertension, and diabetes — tended to cluster together, as did behavioral risks like smoking and excessive alcohol use. Poor hearing, poor vision, and low education also frequently co-occurred. Over 50% of individuals carried at least two risk factors simultaneously, and in 11 of the 14 regions, more than 20% of participants had four or more risk factors.

Experts commenting on the study noted that the consistent clustering of cardiovascular, behavioral, and sensory-social risk factors supports multidomain intervention models. However, they cautioned that it remains unclear whether cluster-based prevention strategies would outperform established approaches like the POINTER and FINGER trials.

A key limitation is the study's cross-sectional design, which prevents conclusions about how risk profiles change over time. Most prior dementia research has focused on high-income countries, so this study helps fill a critical knowledge gap — but primary longitudinal data from underrepresented regions is still needed to confirm these patterns.