摘要：Background: The longevity gains recorded in high-income countries since the 1960s are mainly due to a reduction in mortality from chronic degenerative diseases, which particularly affect older individuals. In recent years the adult modal age at death (M) gained increasing recognition as a lifespan indicator for monitoring improvements in old-age survival. However, studies of M by cause of death are lacking. Objective: This work investigates trends in M by leading causes of death in Canada over the 1974–2011 period and identifies the causes of death that have been more responsive to improvements in lifestyle behaviors and medical progress. Methods: We extend a recent method for estimating the all-cause M using a flexible P-spline approach to the context of cause-of-death analysis. Using data from the Canadian Vital Statistics Database for the 1974‒2011 period, we derive cause-specific modal age-at-death estimates and compare them in terms of levels and time-trends. Results: Although modal age-at-death estimates for heart diseases, cerebrovascular diseases, and the three types of cancers studied (breast/prostate, colorectal, and trachea, bronchus, and lung) differ greatly in terms of levels, they have all followed a steady upward trend since the mid-1970s in Canada. Moreover, the increase in cause-specific modal age estimates occurred at a strikingly similar pace for most causes, except for breast cancer (females) and heart diseases (males), whose modal ages rose at a substantially faster pace. Contribution: Our study introduces an innovative method for estimating cause-specific modal ages at death and provides the first available estimates of time-trends in M by leading causes of death.
关键词：age-at-death distribution;Canada;cancer;cardiovascular diseases;cause of death;longevity;modal age at death;old-age mortality;Poisson P-splines;smoothing
