Mortality deceleration and mortality selection: three unexpected implications of a simple model.

Area of Science:

• Demography
• Biostatistics
• Gerontology

Background:

• Unobserved heterogeneity in mortality risk is a significant factor in population studies.
• Mortality deceleration, the slowing of mortality's increase with age, is often used to infer this heterogeneity.

Purpose of the Study:

• To challenge the assumption that mortality deceleration patterns directly reveal underlying heterogeneity in mortality risk.
• To explore the relationship between mortality selection and deceleration patterns using computational modeling.

Main Methods:

• Development of a simple computational model with two subpopulations exhibiting Gompertz mortality.
• Simulation of cohort mortality patterns to analyze aggregate deceleration and selection effects.

Main Results:

• Aggregate mortality can decelerate even when most of the cohort is frail.
• The model demonstrates the possibility of multiple decelerations and acceleration due to mortality selection.
• Simulated patterns are consistent with real-world cohort data.

Conclusions:

• Mortality deceleration patterns may offer limited insight into the heterogeneity that produces them.
• Results challenge conventional heuristics linking selection to deceleration and inferences about social inequality.
• Standard parametric models may misestimate deceleration timing due to oversimplified assumptions about plateauing.