Excess mortality at ages 20–64 from 2020 to 2024: a comparative analysis of 16 European countries, 2025, Alicandro et al.

[SNT Gatchaman]

Excess mortality at ages 20–64 from 2020 to 2024: a comparative analysis of 16 European countries

Spoiler: Authors

Gianfranco Alicandro; Alberto Giovanni Gerli; Carlo La Vecchia

INTRODUCTION
Emerging evidence suggests that young and middle-aged adults may experience postacute sequelae of COVID-19, increasing their risk of cardiovascular complications. However, it remains unclear whether these risks translated into higher all-cause mortality.

METHODS
We analysed mortality and population data from 16 European countries to estimate excess mortality among individuals aged 20–64 years during the period 2020–2024. Excess mortality was defined as the difference between observed and expected age-standardised mortality rates. Expected values were modelled using three approaches based on different assumptions: (1) a long-trend linear model based on 2010–2019 data; (2) a non-linear model using the same baseline and (3) a short-trend linear model based on 2015–2019 data.

RESULTS
In 2020 and 2021, all models consistently estimated substantial excess mortality in several countries, with the highest estimates in 2021 for Bulgaria (34.7%–37.3% in males, 45.1%–51.6% in females), Hungary (21.0%–28.9% in males, 30.7%–35.4% in females), Poland (19.1%–24.1% in males, 23.6–24.0 in females) and Slovakia (32.1%–36.6% in males, 34.2%–38.1% in females). In 2022, model estimates diverged, with significant excess mortality persisting only under the long-trend linear model. By 2023, residual excess mortality was found only among males in Austria (according to the long-trend linear and non-linear models) and Finland (according to the long-trend linear model), while in 2024 persisted only in Austria under the same models. Assuming a continuing linear mortality trend appeared to overestimate excess mortality in some age groups.

CONCLUSIONS
Our analysis indicates substantial excess mortality among adults aged 20–64 in 2020 and in 2021 in Europe but does not support a sustained postpandemic increase in all-cause mortality. The estimates were sensitive to the choice of the baseline period and modelling approach, underscoring the importance of presenting results under multiple analytical assumptions.

Web | DOI | PDF | BMJ Public Health | Open Access

 

[Kalliope]

Just adding the last paragraph in the article, which I thought was a bit clearer than the conclusion in the abstract:

In conclusion, excess mortality among the population aged 20–64 years was substantial and heterogeneous across Europe in 2020–2021. In the following years, point estimates and their associated uncertainty became highly sensitive to modelling choices, with linear trend assumption generally producing higher and statistically significant excess estimates. Therefore, our analysis highlights the importance of considering multiple scenarios when projecting expected mortality, and that a persisting declining assumption cannot be assumed for all age groups and countries.

 

[Utsikt]

Deleted a previous comment because it misunderstood the premise of the paper.

If I’ve understood it correctly this time, they are testing various methods of establishing the expected mortality based on pre-pandemic data.

They are comparing a 10-year linear model, a 10-year non-linear model an a 5-year linear model.

Using very narrow age bands (5 years except for 20-29), they claim that the 10-year linear model consistently produces a higher rate of excess mortality.

So my main question is: are those narrow age bands realistic? Could the narrow bands introduce more variance and essentially cloud the assessments?