Lancet Healthy Longev. 2026 Jul 12:100867. doi: 10.1016/j.lanhl.2026.100867. Online ahead of print.
ABSTRACT
BACKGROUND: The design of evidence-based interventions to reduce the burden of dementia requires knowledge of the prevalence and patterns of modifiable risk factors. However, most existing evidence comes from high-income countries (HICs). Hence, we aimed to quantify differences in the prevalence and patterns of dementia risk factors across diverse contexts.
METHODS: We conducted a comparative cross-sectional study using harmonised data from 14 countries and regions (Ireland, the USA, England, Northern Ireland, Eastern Europe, Western Europe, Northern Europe, Southern Europe, South Korea, Mexico, China, Malaysia, Brazil, and India), including HICs and low-income and middle-income countries (LMICs). We included individuals aged 50 years or older from 11 nationally representative ageing studies, using data from the most recent available study waves with refresher samples collected between 2009 and 2023. Respondents were excluded from specific analyses if data on the risk factors of interest were missing. We estimated the prevalence of 12 established binary dementia risk factors (low education, hearing loss, high LDL cholesterol, depression, physical inactivity, diabetes, smoking, hypertension, obesity, excessive alcohol consumption, social isolation, and vision loss) using descriptive statistics and examined patterns by age group (70 years and older and 50-69 years), gender, and education using Poisson models with robust variance estimation. We also compared the rank order of the prevalence of risk factors across countries and assessed risk factor co-occurrence and clustering.
FINDINGS: Data on 214 251 respondents were included in the study. We observed some variation in the prevalence and patterns (by age, gender, and education) of risk factors between HICs and LMICs. For example, low education had higher prevalence in many LMICs (85·6% [95% CI 84·8-86·5] in China vs 12·0% [95% CI 11·3-12·7] in the US), whereas obesity was more prevalent in HICs than in LMICs (44·9% [95% CI 43·3-46·5] in the US vs 13·3% [95% CI 12·9-13·7] in India). Risk factor distributions differed by age group, gender, and education, although patterns were not consistent across all settings. Risk factors commonly co-occurred across settings, with more than 50% of individuals having at least two risk factors across all countries and regions. Moreover, broadly similar clusters of risk factors-related to cardiovascular diseases, risky behaviours, and social or sensory factors-were observed across settings.
INTERPRETATION: Differences in the prevalence and patterns of dementia risk factors highlight the need to tailor prevention strategies to specific contexts. However, findings also reveal consistent patterns in risk factor co-occurrence and clustering, which could guide the design of multidomain interventions and policy approaches to reduce dementia risk across settings. Overall, these findings support the use of both context-specific and shared approaches to reduce the burden of dementia.
FUNDING: National Institutes of Health.
PMID:42437564 | DOI:10.1016/j.lanhl.2026.100867