A new parametric model to assess delay and compression of mortality

Area of Science:

• Demography
• Actuarial Science
• Public Health

Background:

• Decreasing mortality shifts age patterns (mortality delay) or changes age-at-death distribution (mortality compression/expansion).
• Existing models struggle to differentiate mortality delay from compression across the full age spectrum.
• Previous models often focused only on adult ages and did not distinguish compression mechanisms above/below the modal age at death.

Purpose of the Study:

• Introduce a novel parametric model, the Compression-Delay (CoDe) model, to describe the full age pattern of mortality.
• Assess mortality compression at different life stages and mortality delay.
• Quantify the distinct contributions of mortality delay and compression to life expectancy changes.

Main Methods:

• Developed the CoDe model with 10 parameters, five time-varying, to capture mortality dynamics across life stages.
• Incorporated specific functions for infant/background mortality and a mixed logistic model for adult ages.
• Included modal age at death as a parameter to account for mortality delay.

Main Results:

• The CoDe model demonstrated an excellent fit for the full age pattern of mortality in Japanese, French, American, and Danish populations (1950-2010).
• Mortality delay accounted for approximately two-thirds of the increase in life expectancy at birth.
• Young-age mortality compression contributed one-third; late adulthood showed no strong compression, and old-age compression had a slight negative impact on life expectancy.

Conclusions:

• The CoDe model is a valid tool for describing the complete age pattern of mortality.
• The model effectively disentangles the effects of mortality delay and compression across different life stages.
• This approach provides crucial insights into the drivers of increasing life expectancy.