Heterogeneity in the association between cardiovascular disease and cognitive decline: a pooled analysis of five prospective cohort studies from 24 countries.
Li J, Cui L, et al. • Alzheimer's & dementia : the journal of the Alzheimer's Association • 2026
Substantial heterogeneity (I²=97.5%) exists in the association between cardiovascular disease and cognitive decline across five multinational cohorts, with opposite directions of association observed between Chinese and Western cohorts, potentially explained by variations in heart disease composition and medication use.
Key Findings
Results
Substantial heterogeneity was observed across cohorts in the association between cardiovascular disease and cognitive decline.
I² = 97.5%, p < 0.001 across five multinational cohorts
The pooled sample included 83,159 adults aged 50 years or older
Data came from CHARLS, Health and Retirement Study, Survey of Health Ageing and Retirement in Europe, English Longitudinal Study of Ageing, and Mexican Health and Aging Study
Countries represented spanned 24 nations
Linear mixed models followed by random-effects meta-analysis were used to estimate associations
Results
The direction of the association between cardiovascular disease and cognitive decline was opposite between Chinese and Western cohorts.
CHARLS (the Chinese cohort) showed an opposite direction of association compared to the four Western cohorts
CHARLS differed from the other four cohorts across multiple characteristics including region, sex, body mass index, and education
The association reversed in CHARLS when analyses were restricted to participants with CVD not taking medication
Descriptive stratification analyses identified CHARLS as the primary source of heterogeneity
Results
Heart disease, rather than stroke, was the primary driver of accelerated cognitive decline in Western cohorts.
Heart disease correlated with accelerated cognitive decline across four Western cohorts
Heart disease-to-stroke ratios ranged between 1.8:1 and 5.2:1 across the four Western cohorts
Stroke alone showed no significant link with cognitive decline in any cohort
Mutually adjusted models for stroke and heart disease were fitted as sensitivity analyses
Separate time-varying models for stroke and heart disease were also examined
Results
Medication use among CVD patients may partially explain heterogeneity in the CVD-cognitive decline association.
The association between CVD and cognitive decline reversed in CHARLS when analyses were restricted to participants with CVD not taking medication
Variations in medication use across the five cohorts were identified as a potential source of heterogeneity
Variations in the composition of heart disease across cohorts were also identified as a potential explanatory factor
Subgroup and sensitivity analyses generally supported the main findings
Methods
Time-varying CVD models were used as supplementary analyses to examine the robustness of findings.
Time-varying CVD models were fitted in supplementary analyses
Mutually adjusted models for stroke and heart disease were included
Separate time-varying models for stroke and heart disease were also run as sensitivity analyses
Subgroup and sensitivity analyses generally supported the main findings of the primary analysis
What This Means
This research suggests that the relationship between cardiovascular disease (CVD) and mental decline in older adults is not consistent across different countries and populations. By pooling data from over 83,000 people aged 50 and older across 24 countries, the researchers found extreme variability (97.5% heterogeneity) in whether and how CVD relates to faster cognitive decline. Most strikingly, the Chinese cohort showed the opposite pattern compared to cohorts from Western countries, meaning CVD was associated with cognitive decline in different directions depending on where people lived.
The study also found that within CVD, it matters what type of cardiovascular condition people have. In Western populations, heart disease (such as heart attacks or heart failure) was strongly linked to faster cognitive decline, while stroke alone was not significantly associated with cognitive decline in any cohort. The ratio of heart disease to stroke cases varied considerably across cohorts (between 1.8:1 and 5.2:1), which may help explain why results differed so much between populations. Additionally, medication use appeared to play a role — when the analysis was limited to Chinese participants with CVD who were not taking medication, the direction of the association flipped.
This research suggests that findings about CVD and cognitive health from one country or population may not apply universally, and that factors like disease composition, medication access and use, and population characteristics (such as education levels and body weight) may all shape this relationship. The findings highlight the importance of studying diverse global populations separately rather than assuming one-size-fits-all conclusions about heart health and brain health.
Li J, Cui L, Xu Y, Ding Y, Huo J, Niu D, et al.. (2026). Heterogeneity in the association between cardiovascular disease and cognitive decline: a pooled analysis of five prospective cohort studies from 24 countries.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71843