Adequate life-expectancy reconstruction for adult human mortality data

Area of Science:

• Demography
• Biostatistics
• Actuarial Science

Background:

• Life tables are essential for calculating life expectancy but struggle with right-censoring in open-ended age groups.
• Conventional methods can distort life expectancy, particularly when the last age group represents a large population segment.
• The gamma-Gompertz-Makeham model offers a robust framework for adult mortality, addressing censoring effectively.

Purpose of the Study:

• To quantify the discrepancies between gamma-Gompertz-Makeham life expectancy and existing life-table databases.
• To highlight populations where life expectancy values may need revision.
• To advocate for the gamma-Gompertz-Makeham model in life expectancy calculations.

Main Methods:

• Utilizing the gamma-Gompertz-Makeham model to calculate life expectancy.
• Comparing model-based results with data from major human life-table databases.
• Applying the model to datasets with data quality issues, scarcity, and censoring, including hunter-gatherer populations.

Main Results:

• Gamma-Gompertz-Makeham life expectancy values are comparable to those from complex, high-quality databases.
• The model effectively handles severe censoring due to data aggregation in the last age group.
• Demonstrated applicability to data-scarce populations, such as hunter-gatherers.

Conclusions:

• The gamma-Gompertz-Makeham model provides reliable life expectancy estimates, especially when data are incomplete or censored.
• It offers a unified approach for diverse populations, including historical ones for mortality forecasting.
• Revising life expectancy trends using this model enhances demographic and mortality forecast accuracy.