Phenotypic Age Acceleration is Associated with Frailty but Not NT-proBNP-Defined Cardiac Dysfunction: A Longitudinal Cohort Study with Community-Dwelling Older Chinese Adults

Technical Summary: Phenotypic Age Acceleration and Geriatric Phenotypes in Community-Dwelling Older Chinese Adults

Problem Statement
Frailty and cardiac dysfunction are highly prevalent geriatric conditions that frequently co-occur in older adults, yet the biological mechanisms linking them remain unclear. A central unresolved question is whether these syndromes arise from a shared "common soil" of biological ageing or from distinct, pathway-specific processes. While Phenotypic Age Acceleration (PhenoAgeAccel)—a composite biomarker derived from nine routine clinical biomarkers and chronological age—has been validated as a predictor of mortality and morbidity in Western cohorts, its differential association with frailty versus cardiac dysfunction has not been systematically tested. Specifically, it is unknown whether PhenoAgeAccel uniformly predicts both conditions or exhibits selective associations that would support distinct intervention strategies.

Methodology
This study employed a prospective, community-based cohort design involving 3,631 community-dwelling older adults (aged ≥60) from Shanghai, China.

• Exposure: The primary exposure was PhenoAgeAccel, calculated at baseline (2023) using the Levine et al. algorithm. This involved nine routine biomarkers (albumin, creatinine, glucose, CRP, lymphocyte percentage, MCV, RDW, ALP, WBC) and chronological age. PhenoAgeAccel was defined as the residual from regressing PhenoAge on chronological age.
• Outcomes: At the 2025 follow-up, participants were classified into four mutually exclusive groups based on NT-proBNP levels (≥300 pg/mL) and/or EHR-recorded heart failure diagnoses, and the Fried frailty phenotype (score ≥3):
  1. Robust (Reference)
  2. Cardiac Dysfunction Only
  3. Frailty Only
  4. Comorbid (Both conditions)
• Statistical Analysis: The primary analysis utilized multinomial logistic regression with hierarchical adjustment (Model 1: age/sex; Model 2: socioeconomic/behavioral factors; Model 3: cardiometabolic comorbidities). Results were reported as Relative Risk Ratios (RRRs).
• Sensitivity Analyses: Robustness was tested using stricter cardiac thresholds (NT-proBNP ≥450 pg/mL), an incident cohort (excluding baseline NT-proBNP 300–450 pg/mL), and age stratification (<75 vs. ≥75 years). Longitudinal trajectories of natriuretic peptides were assessed using linear mixed models.

Key Results

• Differential Association: In the fully adjusted model, higher PhenoAgeAccel was significantly associated with an increased risk of Frailty Only (RRR = 1.028 per 1-year increase; 95% CI 1.011–1.046; P=0.002). This corresponds to an estimated 15% increased risk over 5 years and 32% over 10 years.
• Lack of Cardiac Association: Conversely, PhenoAgeAccel showed no statistically significant association with Cardiac Dysfunction Only (RRR = 1.005; P=0.675) or the Comorbid group (RRR = 1.014; P=0.454).
• Robustness: These findings remained consistent across all sensitivity analyses, including those using stricter NT-proBNP thresholds (≥450 pg/mL) and the incident cohort.
• Age Stratification: The association between PhenoAgeAccel and frailty appeared stronger in adults younger than 75 years (RRR = 1.032) compared to those aged 75 and older (RRR = 1.020, non-significant).
• Longitudinal Trajectories: PhenoAgeAccel was not associated with baseline natriuretic peptide levels or their rate of change over the 2-year follow-up period.

Key Contributions

1. Pathway Specificity: The study provides evidence that PhenoAgeAccel is selectively associated with frailty but not with NT-proBNP-defined cardiac dysfunction in a community-dwelling Chinese population. This challenges the "common soil" hypothesis for these specific phenotypes, suggesting that systemic multi-organ vulnerability (captured by PhenoAge) and cardiac-specific hemodynamic stress may operate via distinct biological ageing pathways.
2. Methodological Rigor: By utilizing mutually exclusive outcome groups (separating single-phenotype states from comorbid states) and rigorous sensitivity analyses, the study isolates the specific predictive value of PhenoAgeAccel for frailty, avoiding the conflation of systemic vulnerability with organ-specific pathology.
3. Clinical Relevance: The findings suggest that PhenoAgeAccel may serve as a specific tool for stratifying frailty risk in primary care, potentially complementing functional assessments, while indicating that it may not be a suitable standalone biomarker for predicting incident cardiac dysfunction in this demographic.

Significance and Claims
The authors conclude that PhenoAge, as a specific composite ageing biomarker, does not uniformly capture risk for both frailty and cardiac dysfunction. Instead, it appears to primarily reflect systemic multi-organ vulnerability rather than cardiac-specific hemodynamic stress. The study posits that these findings support the existence of at least partially distinct pathways linking biological ageing to these geriatric syndromes. Consequently, the authors suggest that PhenoAgeAccel could complement functional frailty risk stratification in community primary-care settings, with implications for tailoring integrated care planning and quality-of-life interventions. The paper explicitly notes that while the dissociation suggests distinct pathways, the limited power for the comorbid group precludes definitive exclusion of a meaningful effect in that specific subgroup.