Nearly three decades of Framingham data link faster biological aging in midlife to cardiovascular disease and mortality, with women showing an additional signal for dementia risk.

Paper: Association of Accelerated Biological Age in Midlife with Cardiovascular Events, Dementia, and All-Cause Mortality in the Framingham Offspring Study. Image Credit: Faizal Ramli / Shutterstock

A recent study published in the journal Scientific Reports with nearly three decades of follow-up suggests that accelerated biological age (ABA) in midlife is associated with a higher risk of adverse health outcomes, particularly dementia in women.

The findings suggest that ABA could complement conventional risk factors in identifying people with poorer multisystem health, although its added predictive value was limited and clinical applications require prospective validation.

As age advances, physiological functions eventually decline. As a result, age-related changes can contribute to increased disease and death risks. The chronological age (CA) of an individual is calculated as the number of years elapsed since birth. While the CA would be the same for two individuals born in the same year, their biological age (BA) estimates may vary, based on genetic, lifestyle, and environmental factors. CA reflects elapsed time since birth, whereas BA estimates physiological aging and can vary among people of the same chronological age.

BA estimation, by integrating several clinical biomarkers, may therefore better reflect the dynamic changes in biological processes associated with aging compared to CA. Recent studies have linked ABA to worse health outcomes. However, sex-stratified differences in the long-term influence of ABA on health outcomes remain unclear.

Many previous studies employed cross-sectional designs or relatively short- to moderate-term follow-up periods and often focused on individual outcomes.

About the study

In the present observational study, researchers explored longitudinal associations between BA, cardiovascular disease (CVD), dementia, and death by sex and its utility compared with CA.

The team estimated BA based on anthropometric parameters and biochemical markers for up to 3,046 Framingham Offspring Study (FOS) participants across three cycles between 1991 and 2001, with sample sizes varying by outcome.

Using biomarkers such as left ventricular mass indexed by height (LVMI), Mini-Mental State Examination (MMSE), pulmonary function measures, grip strength, and gait speed, the researchers constructed principal components (PCs) from selected biomarkers and applied the Klemera and Doubal method (KDM) to estimate BA separately for men and women.

Markers such as forced vital capacity (FVC) and forced expiratory volume in one second (FEV1) helped assess lung function. Kidney function indicators included creatinine, cystatin C, and estimated glomerular filtration rate (eGFR).

In addition to vital signs, the researchers also assessed total cholesterol (TC), free thyroxine (fT4), C-reactive protein (CRP), and carotid-femoral pulse wave velocity (cfPWV). Glucose regulation markers, such as blood glucose (BG), hemoglobin A1c (HbA1c), and insulin measures, were also considered for BA estimation, with the biomarkers available and selected varying by examination, sex, and outcome.

Based on standardized residuals from the regression of BA on CA, participants were classified as having accelerated, concordant, or decelerated biological age (DBA). Time-dependent Cox regression models were used for CVD and mortality, while competing-risk models accounting for death were used for dementia.

Additional analyses, including restricted cubic splines, assessed the shape and robustness of these associations, while incidence rates were calculated per 1,000 person-years.

The team also considered variables such as CA, body mass index (BMI), smoking status, diabetes, and medication use, with prevalent CVD included in mortality models and apolipoprotein E ε4 (APOE4) status and education included in dementia models.

For sensitivity analyses, the researchers incorporated the eighth examination cycle conducted between 2005 and 2008, adding 74 participants to the analytic cohort. Concordant BA was used as the reference instead of DBA in an additional analysis. The researchers also incorporated coronary artery calcium (CAC) and bone mineral density (BMD) as candidate biomarkers for BA estimation at examination eight.

Results and discussion

In the community-based study sample, the mean age of participants was 55 years, and most (55%) were female. Participants with DBA had an estimated BA that was, on average, 3 years lower than their CA. Male participants with ABA had BA estimates 4 years above their CA, whereas female participants had BA estimates 2.5 years above their CA.

Over median follow-up periods of 23 years for CVD and 26 years for dementia and mortality, 713 CVD cases and 265 incident dementia cases were reported. In addition, 1,105 participants died. Compared to men with DBA, those with ABA were at an increased risk of CVD (hazard ratio [HR], 1.6) and mortality (HR, 2.3).

Women who biologically aged faster showed similar patterns, with an additionally heightened dementia risk. In particular, for female participants classified as having ABA, the HR values for CVD, dementia, and death were 1.8, 2.1, and 1.8, respectively.

The dementia association reached statistical significance when women with ABA were compared with those with DBA, although the authors characterized the finding as borderline, and it was not significant when concordant BA was used as the reference.

Hormonal factors such as declining estrogen levels have been proposed in previous research as one possible contributor to greater susceptibility to cognitive decline after menopause, although the mechanisms remain uncertain.

Sensitivity analyses generally supported the CVD and mortality findings, although the strength and statistical significance of the dementia associations varied across analyses.

Conclusion and Future Directions

The findings suggest that people with accelerated biological age may be at an increased risk of experiencing CVD, dementia, and death. In addition, ABA may provide complementary information about cumulative physiological changes across multiple organ systems, although adding it to standard clinical risk factors produced only minimal improvements in predictive performance.

Based on the findings, ABA determination could potentially support more targeted monitoring or preventive strategies, but these applications remain exploratory.

However, the findings should be confirmed in prospective studies involving more diverse populations across different ages and racial and ethnic backgrounds to improve generalizability. The observational design also precludes causal conclusions, and residual confounding cannot be ruled out.

Further investigations could explore the biological mechanisms underlying differential aging between males and females and determine whether biological-age measures provide clinically useful information beyond established risk factors.

Journal reference:

• Pan, S., Beiser, A. S., Mitchell, G. F. et al.(2026). Association of Accelerated Biological Age in Midlife with Cardiovascular Events, Dementia, and All-Cause Mortality in the Framingham Offspring Study. Scientific Reports.DOI: 10.1038/s41598-026-64591-9, https://www.nature.com/articles/s41598-026-64591-9