Association between frailty index and self-reported stomach pain among middle-aged and older adults: Evidence from the China Health and Retirement Longitudinal Study and machine-learning analysis
Guobin Cheng, Yan Geng, Linya Lyu, Fang Dai, Jingji Huang, Wei FuBackground
Frailty reflects accumulated health deficits and reduced physiological reserve across multiple systems. Although frailty has been associated with various adverse health outcomes, its relationship with self-reported stomach pain remains unclear. This study aimed to investigate the association between frailty index and self-reported stomach pain and evaluate the predictive performance of frailty index–based models using machine-learning approaches.
Methods
In total, 24,280 participants aged ≥45 years enrolled in the baseline survey of the 2011–2013 China Health and Retirement Longitudinal Study were included. A frailty index was constructed using multiple health-related deficits. Multivariable logistic regression and restricted cubic spline analyses were performed to assess the association and nonlinear relationship between frailty index and stomach pain. Subgroup analyses were conducted to explore potential effect modification by major comorbidities. Machine-learning models incorporating frailty index and clinical characteristics were developed, and SHapley Additive exPlanations analysis was performed to interpret feature contributions.
Results
A higher frailty index was significantly associated with increased likelihood of self-reported stomach pain in the fully adjusted model (odds ratio = 2.17, 95% confidence interval: 1.95–2.42,
Conclusion
A higher frailty index was associated with greater likelihood of self-reported stomach pain among middle-aged and older adults. These findings suggest that frailty represents an important marker of gastrointestinal symptom burden. Machine-learning analyses further demonstrated the value of frailty index in identifying individuals with stomach pain. However, longitudinal studies are needed to clarify causal relationships.