In a nationally representative Chilean sample, machine learning models identified sex differences in frailty predictors, with number of health conditions being predominant for both sexes, physical activity more salient for men, and leisure/recreational activities more pronounced for women.
Key Findings
Results
The predominant predictor of physical frailty was the number of health conditions present in both male and female community-dwelling older adults.
The sample consisted of 9,123 individuals aged 60 and older from the National Disability and Dependency Survey in Chile
The sample included 5,363 women and 3,760 men
Supervised machine learning regression models were applied to identify the most relevant predictors
The finding was consistent across both sexes
Results
Physical activity emerged as the second most salient predictor of frailty for men.
This sex-specific finding was identified using supervised machine learning regression models
Physical activity was more salient for men than for women as a frailty predictor
The sample of men included 3,760 individuals aged 60 and older
Results
Engagement in leisure and recreational activities demonstrated a more pronounced correlation with frailty for women compared to men.
Leisure and recreational activity engagement was identified as a key sex-specific predictor for women
This contrasted with physical activity being the second most salient factor for men
The sample of women included 5,363 individuals aged 60 and older
These findings suggest gender-responsive strategies may be needed for frailty prevention
Results
Age was a significant predictor of physical frailty in both male and female older adults.
Age was identified as significant across both sexes using machine learning regression models
Participants were aged 60 and older
The finding was consistent regardless of sex
Conclusions
Sex differences exist in the predictors of physical frailty among community-dwelling older adults in Chile.
The study used a nationally representative sample from the National Disability and Dependency Survey
Supervised machine learning regression models were applied to a sample of 9,123 individuals
While some predictors (number of health conditions, age) were shared, physical activity and leisure/recreational activities showed sex-specific patterns
The findings were interpreted as offering 'valuable insights for developing gender-responsive strategies to prevent or delay the progression of physical frailty'
Araya A, Iriarte E, Fernández-Lorca M, Cornejo C. (2026). Sex Differences in Predictors of Frailty among Community-Dwelling Older Adults: Machine Learning Approach.. Clinical nursing research. https://doi.org/10.1177/10547738261417607