Concordant decline in both subjective memory and objective cognition was associated with the greatest risk of incident functional dependence (HR 1.470, 95% CI 1.381–1.566), with smaller associations for subjective-only or objective-only decline, suggesting joint cognitive change may improve identification of older adults at risk.
Key Findings
Results
Concordant decline in both subjective memory and objective cognition showed the strongest and most consistent association with incident functional dependence across all four cohorts.
Pooled Cox analysis HR for concordant decline: 1.470 (95% CI 1.381–1.566)
Analysis drew on 62,329 participants included in fully adjusted outcome analyses
Results were consistent across CHARLS, HRS, ELSA, and SHARE cohorts
Pooled estimates were derived using random-effects meta-analysis of cohort-specific Cox models
Results
Subjective-only decline (worsening subjective memory without objective cognitive decline) was associated with a smaller but statistically significant increase in risk of functional dependence.
Pooled HR for subjective-only decline: 1.222 (95% CI 1.055–1.415)
The confidence interval does not include 1.0, indicating statistical significance
This association was smaller than that observed for concordant decline (HR 1.470)
Results
Objective-only decline (worsening objective cognition without subjective memory decline) was associated with the smallest risk increase among the three declining trajectory groups.
Pooled HR for objective-only decline: 1.116 (95% CI 1.038–1.200)
This was the weakest association among the three non-stable trajectory groups
The association remained statistically significant with the confidence interval excluding 1.0
Results
Concordant decline remained the strongest predictor of functional dependence after accounting for death as a competing event.
Competing-risk analyses were conducted across CHARLS, HRS, and SHARE (cohorts with adequate mortality information)
Pooled subdistribution HR for concordant decline: 1.386 (95% CI 1.273–1.509)
This finding held after treating death as a competing event, suggesting the association is not an artifact of differential mortality
Methods
The study included 86,562 participants for trajectory derivation across four large international aging cohorts spanning China, the United States, England, and Europe.
Cohorts included: CHARLS (China Health and Retirement Longitudinal Study), HRS (US Health and Retirement Study), ELSA (English Longitudinal Study of Ageing), and SHARE (Survey of Health, Ageing and Retirement in Europe)
86,562 participants were included in trajectory derivation; 62,329 were included in fully adjusted outcome analyses
Subjective memory and objective cognition were assessed at baseline (T0) and first follow-up (T1), with incident functional dependence examined after T1
Participants were classified into four trajectory groups based on joint changes in subjective and objective cognition
Results
Multiple sensitivity analyses supported the robustness of the primary findings linking concordant cognitive decline to functional dependence.
Sensitivity analyses addressed alternative decline thresholds, continuous change-score models, fixed follow-up windows, and baseline-level adjustment
Additional sensitivity checks included multiple imputation for missing data and competing-risk models
Results were consistent across these varied analytical approaches
What This Means
This research suggests that tracking both how people feel about their own memory (subjective memory) and how they actually perform on memory and thinking tests (objective cognition) over time can help predict who is most likely to lose independence in daily activities later in life. The study followed over 86,000 older adults from four countries—China, the United States, England, and across Europe—and found that people whose subjective memory complaints and objective test performance both worsened together ('concordant decline') faced about a 47% higher risk of becoming functionally dependent compared to those who remained stable. People who had only one type of decline—either complaints without test changes, or test changes without complaints—also had elevated risk, but not as high.
The findings were consistent across all four large international studies and held up under numerous additional checks, including analyses that accounted for the possibility that some participants died before becoming dependent. This suggests the results are robust and not explained by differences in how death rates varied across groups. The study used a harmonized approach, meaning the researchers carefully aligned the methods and measures across the different national studies to make the comparisons as valid as possible.
This research suggests that clinicians and researchers should consider monitoring both self-reported memory concerns and objective cognitive test performance together, rather than relying on just one type of measure, when trying to identify older adults who may be at higher risk of losing independence. People who report worsening memory and also show declining test scores may benefit from closer follow-up or earlier intervention, even before functional problems become apparent.
Yin X, Li Z, Fu R, Hong T, You Y, Zhao Y, et al.. (2026). Joint changes in subjective memory and objective cognition and risk of incident functional dependence: A harmonized analysis of four aging cohorts.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71752