Modelling the Non-linear Associations between Age and Health: Implications for Care

Background
Ageing is a complex biological process that weakens the body's defences against various health problems and diseases. This decline is not uniform or linear; rather, it varies significantly among individuals. Understanding the non-linear progression of ageing is crucial for identifying critical periods where intervention might be most beneficial. This study aimed to explore nonlinearities in the association between age and health outcomes.

Methods
Using generalised additive models, we analysed data from the 2020 wave of the United States Health and Retirement Study. This modelling approach allows for the detailed examination of nonlinearities in health outcomes across different ages. Our sample consisted of adults aged over 50 (n = 1338; mean = 68; range = 50 – 95). We assessed seven health outcomes (heart conditions, diabetes, cholesterol, blood pressure, fatigue, back pain, and headaches) and seven protective behaviours (prostate exams, mammograms, cholesterol screenings, pap smears, flu vaccinations, dental check-ups, and smoking patterns) reflecting various aspects of health status and behaviour.

Results
The analysis indicated a significant and pronounced non-linear relationship between age and certain health outcomes, namely headaches, blood pressure, diabetes, and heart conditions. Additionally, ageing was non-linearly associated with specific health behaviours, including the uptake of prostate exams, mammograms, flu vaccinations, and smoking prevalence. Notably, distinct patterns emerged concerning health behaviours in relation to age-specific screening policies.

Conclusion
The findings of this study reveal that the probability of developing health problems or engaging in health protective behaviours do not increase/decrease linearly in direct proportion to age. Rather, the probability curve exhibits fluctuations, presenting periods of heightened odds interspersed with phases of relative stability. Such non-linearity may be partially attributed to policy decisions but could also reflect a complex interplay of factors such as weakened immune systems, access to healthcare, personal health beliefs, and the evolving perception of risk as one grows older.