On ordered subpopulations and population mortality at advanced ages

Area of Science:

• Biostatistics
• Survival Analysis
• Stochastic Modeling

Background:

• Heterogeneous populations exhibit complex mortality patterns.
• Fixed frailty models are used to analyze survival data with unobserved heterogeneity.
• Understanding hazard rate shapes is crucial for accurate survival predictions.

Purpose of the Study:

• To investigate hazard (mortality) rate shapes in ordered subpopulations.
• To analyze deceleration and leveling off (plateaus) in hazard rates.
• To explore lifetime distributions resulting in asymptotically flat hazard rates.

Main Methods:

• Utilizing fixed frailty models with multiple frailty parameters.
• Applying vitality models where initial resources are depleted via stochastic processes.
• Considering stochastic processes like the Wiener process with drift and the gamma process.

Main Results:

• Demonstrated how ordered subpopulations influence hazard rate shapes.
• Identified conditions leading to decelerating and asymptotically flat hazard rates.
• Provided examples of lifetime distributions exhibiting mortality plateaus.

Conclusions:

• Heterogeneity and specific vitality models can generate mortality plateaus.
• The study offers insights into long-term survival patterns in complex populations.
• Findings are relevant for actuarial science, demography, and reliability engineering.