Modeling biological age and its link with the aging process

Area of Science:

• Gerontology
• Biostatistics
• Epidemiology

Background:

• Individual health status heterogeneity in older adults arises from genetic and environmental factors.
• Chronological age (CA) is a limited proxy for physiological decline, necessitating better biological age (BA) indicators.
• Existing BA estimators often rely on restrictive assumptions about the CA-BA relationship.

Purpose of the Study:

• To propose and validate novel biological age (BA) estimators using structural equation modeling (SEM).
• To develop BA estimators that do not require arbitrary assumptions about the relationship between chronological age (CA) and BA.
• To provide tools for empirically testing assumptions in BA estimation.

Main Methods:

• Utilized the US National Health and Nutrition Examination Survey (1988-1994) dataset.
• Developed two BA estimators based on SEM, modeling the BA-CA relationship.
• Compared SEM-based BA estimates against principal components analysis (PCA), multiple linear regression (MLR), and Klemera-Doubal's method (KD).

Main Results:

• SEM-based BA estimates significantly differed from PCA and MLR.
• SEM-based BA estimates showed comparable results to KD but with superior predictive power.
• The proposed SEM approach offers flexibility in modeling the CA-BA relationship and testing assumptions.

Conclusions:

• The novel SEM-based BA estimators provide a more accurate and flexible approach to assessing biological aging.
• These estimators can serve as valuable indicators of accelerated aging, improving upon traditional CA metrics.
• The SEM framework allows for robust validation and interpretation of biological age estimates.