Clinical data-driven classification of pre-frailty reveals sex-specific patterns - Data from the Berlin Aging Study...

Area of Science:

• Gerontology
• Biomarkers
• Machine Learning

Background:

• Frailty significantly impacts older adults' quality of life.
• Lack of standardized frailty assessment tools leads to heterogeneity.
• Understanding frailty's early stages is crucial for effective treatment.

Purpose of the Study:

• Investigate determinants of physical frailty using machine learning.
• Identify sex-specific biomarkers for pre-frailty.
• Enhance current diagnostic scores with monitorable characteristics.

Main Methods:

• Analysis of cross-sectional biochemical and survey data from the Berlin Aging Study II (BASE-II).
• Inclusion of 1512 participants aged 60 and above.
• Application of machine-learning techniques to assess Fried's 5-item frailty phenotype.

Main Results:

• Identified sex-specific predictors of pre-frailty.
• Low appendicular lean mass and high fat composition in men.
• Vitamin D deficiency and high white blood cell counts in women.

Conclusions:

• Findings highlight potential clinical applications for sex-specific pre-frailty biomarkers.
• Results emphasize the need to consider sex differences in frailty research and clinical practice.
• Further research is needed due to the small sample size of fully frail individuals.