Estimating Biological Age in the Singapore Longitudinal Aging Study

Area of Science:

• Gerontology
• Biomarker Discovery
• Aging Research

Background:

• Biological age (BA) offers a more precise assessment of aging rate than chronological age (CA).
• Limited consensus exists on standardized BA measures for predicting lifespan and healthspan outcomes.
• This study addresses the need for robust BA assessment tools in aging populations.

Purpose of the Study:

• To investigate measurement sets of physiological biomarkers for biological age estimation.
• To compare the predictive accuracy of different algorithms (PCA, MLR, KD, machine learning) for BA.
• To evaluate the association of optimal BA estimates with health risks and outcomes.

Main Methods:

• Utilized data from 2,844 Chinese Singaporeans (aged 55-94) in the Singapore Longitudinal Aging Study (SLAS-2).
• Applied Principal Component Analysis (PCA), Multiple Linear Regression (MLR), and Klemera-Doubal (KD) methods for BA estimation.
• Explored machine learning approaches for predicting mortality and frailty over an 8-year follow-up.

Main Results:

• Selected 8 biomarkers for males and 10 for females, including estimated glomerular filtration rate and forced expiratory volume in 1 second.
• Biological age estimates robustly predicted frailty and mortality, outperforming chronological age.
• Biologically older individuals exhibited higher frailty, mortality, and poorer health characteristics.

Conclusions:

• Biological age is a more effective measure than chronological age for assessing lifespan and healthspan.
• The identified physiological markers effectively distinguish between biologically old and young individuals of the same chronological age.
• This validated set of biomarkers provides a reliable tool for aging research and clinical applications.