Long-term mortality rate changes and inverse patterns for pandemic and post-pandemic excess mortality

The COVID-19 pandemic and pandemic response may have long-term consequences. The cumulative impact may be better appraised when post-pandemic years are also considered. For 38 populations with reliable death registration data, we estimated excess deaths for 2020-2025 with granular age stratification. One model (Fa) compared deaths against the mean of 2017-2019. Three other trend models considered changes in mortality rates after 2003 (or after a country reached $20,000 per capita income) factoring trend-of-trends (TTa), including shrinkage (STTa), and factoring also the 2024-2025 data for trend-of-trends calculation (STTa+). Slopes (weighted mean -0.62%/year in 2019) and slopes-of-slopes (weighted mean +0.104%/year^2) for age-stratified mortality rates were highly heterogeneous across populations. On model average, 4 populations (Luxembourg, Ireland, Sweden, New Zealand) had cumulative death deficits during 2020-2025, while another 6 (Chile, Bulgaria, Japan, Greece, Slovakia, Italy) had >4% excess deaths. Differences were more prominent during 2020-2023, while 33/38 countries had death deficits in 2024-2025. Total 2020-2025 excess deaths were 0.92-2.49 million (2020-2023: 1.99-2.88 million; 2024-2025: a deficit of 0.39-1.08 million). Lack of age stratification and use of unchanged linear trends for the baseline (as in Lee-Carter modeling) grossly biased excess death estimates upwards. Socioeconomically more vulnerable populations had higher pandemic deaths, but a more pronounced post-pandemic death deficit. Excess death estimates require careful consideration of changing population age structure and long-term mortality trajectories. Post-pandemic deficits, especially in more vulnerable populations, may reflect pandemic deaths of people with modest life expectancy, paid off in 2024-2025.