Structural breaks in Italian age-specific mortality trends: assessing the cohort-effect hypothesis

Using Human Mortality Database cohort life-tables for Italy, we reconstructed calendar-year log-mortality rates for five-year age groups between 50–54 and 85–89 from 1980 to 2019. An unconstrained segmented-regression search located one or two structural breaks per series; almost all fell within 2008-2010. Treating 2008 as a common breakpoint, basic interrupted time-series models then estimated separate slopes before and after that year. The post-2008 slope is significantly flatter for essentially every age group, implying that annual improvements in age-specific mortality slowed by roughly 0.6 % to 2.2%. Men experience the sharper early-old-age slowdown, but the effect tapers after their mid-sixties; among women it grows with age and peaks near 80. The near-synchronous timing of the breaks across cohorts, their consistency across sexes, and their alignment with the onset of Italian health-care austerity are difficult to reconcile with a cohort-driven explanation. Instead, they point to a period effect that dampened mortality gains almost overnight.