Main learnings

Was the pandemic effect on life-lost serious?
Yes, it was exceptional. Most European countries saw in the pandemic years the worst life expectancy losses since the end of the Second World War.

What made the difference in saving lives during the COVID-19 pandemic?
In the first pandemic year 2020, the countries with measured, timely and well-enforced lockdowns had the lowest excess death. In 2021, vaccines were available and countries that applied vaccines fast and wide saw lower losses.

Who was mostly affected?
In general, the excess death was highest for older persons, for males, inhabitants of large urban agglomer-ations, and in particular persons from already disadvan-taged segments of society.

And what about Denmark?
Denmark with its timely and well-enforced lockdowns, wide vaccination, and overall high level of population compliance experienced the lowest levels of excess death during the two pandemic years. Also, especially in 2020, Denmark successfully protected institutionalized populations in nursing homes.

Measuring the burden of the COVID-19 pandemic on mortality proved to be challenging. COVID-19 reported deaths revealed only a partial picture. This was due to the insufficiency of testing capacity, especially at the beginning of the pandemic, and the misclassification and inconsistencies in reporting COVID-19 deaths across time and countries. Researchers from CPop at SDU envisioned to assess the impact of the pandemic more comprehensively by means of two metrics: Excess death and life expectancy.

Researchers from CPop analyzed the excess death during the first COVID-19 wave in England and Wales (Aburto et al., 2019 JECH) and Denmark and Sweden (Rizzi and Vaupel, 2019 PNAS). Sweden reported the highest excess death risk with a 20% increase in deaths compared to the expected levels; in England and Wales, elevated mortality rates resulted in 15% increase in deaths compared with the expected level. In comparison, negligible excess death risk was reported for Denmark. The urge to provide systematic evidence of the impacts of the pandemic on population health in a cross-national perspective, led CPop researchers to analyzing a high-quality data set of harmonized all-cause mortality records for 29 countries, representing most of Europe, Chile, and the USA (Aburto et al., 2022 IJE).

Denmark and Norway were the only two countries out the 29 under study, that recorded increase in life expectancy for both males and females in 2020 (Figure 2). This does not mean that lives were not lost due to COVID-19 there. But the losses were relatively low and the increase in mortality due to COVID-19 was offset by mortality reductions among other causes (e.g., because of lower infection rates from other diseases). Danes were gaining a little more than two months of life expectancy yearly before the pandemic. In the first pandemic year 2020 the gain stayed positive although at somewhat lower level, 1.1 months.

A follow-up study covering the impact of the pandemic on life expectancy in 2021 saw more diverse impacts across countries (Schöley, et.al. 2022 Nat Hum Behav). Denmark again fared well in this comparison (Figure 3). In 2021, vaccines were available and countries that applied a vaccine fast and wide saw lower life-losses.Some countries experienced life expectancy bounce-backs from previous losses (among them most significantly Switzerland, Sweden, Spain, and Belgium). In more countries however, the life expectancy storm continued and even intensified. This was the case of Central and Eastern European countries. Also, USA with the worst life expectancy decline in the first year of pandemic did not rebound in 2021.

CPop researchers did not only analyze excess death and life expectancy at the national level. They zoomed in and stratified the population by regions and demographic characteristics to understand which subpopulations had been mostly affected, focusing on regional level, urban vs rural areas, age, sex, race, living arrangements, frailty, and socioeconomic status.

Estimation of excess death and life expectancy at the regional level showed that geographical spatial disparities in mortality occurred during the pandemic. An unequal impact of the first wave was found across French and Spanish regions (Ainhoa and Rizzi, 2020 Popul) with high density areas suffering the most. Comunidad de Madrid in Spain and Île-de-France in France, regions including the capital cities, recorded the highest excess death risk above age 70 (around 50% and 20% respectively). In a later paper, Kashnitsky et al. (Kashnitsky et al. 2023 study report) explored the differences in the mortality dynamics between urban and rural areas of 20 European countries during the two pandemic years 2020 and 2021. Authors found pronounced and significant differences, with urban areas being harder hit by COVID-19 mortality in most countries. Evidence pointed to the necessity of protecting densely populated areas during the spread of highly communicable diseases.

Regarding demographic characteristics, while the highest excess mortality was generally found at older ages across countries, gender differences in excess death risk existed, but with a nuanced and non-consistent picture. In an analysis of the first COVID-19 wave in Italy, for ex-ample, authors found excess death risk to be 25.0% both for men and women, mostly over age 70. Men generally had more excess deaths compared to women, but the only statistically significant difference between the two sexes occurred for the age categories 60–69 and 70–79 (Rizzi et al, Genus).

Given the demographic analysis of the impact of the COVID-19 pandemic on mortality, researchers contributed further to the unresolved debate on the effectiveness of non-pharmaceutical interventions (NPIs) in fighting the pandemic. They analyzed two interesting cases to estimate the lockdown policy effectiveness.

First, Ege et al. (2023 Sci Rep) used the rare case of Sweden to estimate the causal impact of not implementing a lockdown on excess death outcomes from the end of February 2020 throughout the end of September 2020. Using state-of-the-art synthetic cohort methods, researchers constructed a counterfactual version of Sweden, a synthetic control, to estimate what would have happened to actual Sweden, the treated unit, had it imposed a mandatory lockdown (Figure 5). Authors found that by the end of September 2020 approximately 10% of all deaths in Sweden could have been avoided, had Sweden implemented a mandatory lockdown.

Second, Ege et al. (2023 study report) studied the short, yet stringent, lockdown implemented in 7 Danish municipalities in late 2020, when a mutated SARS-CoV-2 virus variant showed spillover infections between the mink population and humans. This case represents a unique opportunity to estimate the causal effect of a timely lock-down policy on infection rates. Infections, rather than excess deaths, were used here as outcome of interest given the small population size (hence, too small number of deaths). High testing rates consistent across all Danish municipalities lend credibility to the outcome measure. While the rest of Denmark remained with few restrictions in late 2020, 7 municipalities in Northern Jutland were targeted, not because of high infection numbers, but as a part of the overall response to the widespread circulation of mink-derived variants of SARS-CoV-2 in the human population in Northern Jutland. Applying a state-of-the-art synthetic cohort approach, it became evident that timely, short-lived, stringent measures were able to abruptly halt additional infections, indicating a substantial and significant reduction in daily infections when compared to the respective synthetic counterfactual municipality (Figure 6 reports the case of Hjørring).

Finding showed consistent reductions of daily infection numbers ranging between 19% to 43% in the target municipalities. Researchers could conclude that a strict and short lockdown policy controlled the spread of the virus.

1. Quantifying life expectancy losses globally

1. Aburto, JM, et al. 2022. “Significant Impacts of the COVID-19 Pandemic on Race/Ethnic Differences in USA Mortality.” Proceedings of the National Academy of Sciences (PNAS). Access here.

2. Aburto, José Manuel. 2021a. “The Need for All-Cause Mortality Data to Aid Our Understanding of the COVID-19 Pandemic in Latin America.” American Journal of Public Health. Access here.

3. Aburto, JM, et al. 2021. “Estimating the Burden of the COVID-19 Pandemic on Mortality, Life Expectancy and Lifespan Inequality
in England and Wales: A Population-Level Analysis.” Journal of Epidemiology and Community Health. Access here.

4. Aburto, JM, et al. 2022. “Life Expectancy Declines in Russia during the COVID-19 Pandemic in 2020.” International Journal of Epidemiology. Access here.

5. Aburto, JM., et al. 2021. “Quantifying Impacts of the COVID-19 Pandemic through Life-Expectancy Losses: A Population-Level Study of 29 Countries.” International Journal of Epidemiology. Access here.

6. Levin, A. et al. 2022. “Assessing the Burden of COVID-19 in Developing Countries: Systematic Review, Meta-Analysis, and Public Policy Implications.” BMJ Global Health. Access here.

7. Mena, G. and Aburto, JM. 2022. “Unequal Impact of the COVID-19 Pandemic in 2020 on Life Expectancy across Urban Areas in Chile: A Cross-Sectional Demographic Study.” BMJ Open. Access here.

8. Schöley, et.al. 2022. “Life Expectancy Changes since COVID-19.” Nature Human Behaviour. Access here.

2. Estimation of seasonal mortality

9. Leger, A. and S. Rizzi. 2022. “Estimating Excess Mortality in French and Spanish Regions during the First COVID-19 Wave with the Later/Earlier Method” Population. Access here.

10. Rizzi, S. et al. 2022. “High Excess Deaths in Sweden during the First Wave of COVID-19: Policy Deficiencies or ‘Dry Tinder’?” Scandinavian Journal of Public Health. Access here.

11. Rizzi, S, and J W. Vaupel. 2021. “Short-Term Forecasts of Expected Deaths.” Proceedings of the National Academy of Sciences. Access here.

12. Schöley, J. 2021. “MOCY Database on ‘Timeseries of Weekly Death Counts and Co-Variates by Country, Sex and Age. Database available at GitHub here.

13. Léger, A, and Rizzi S. 2023. “Month-to-Month All-Cause Mortality Forecasting: A Method to Rapidly Detect Changes in Seasonal Patterns.” Accepted American Journal of Epidemiology.

14. Missov, T. I. n.d. “Estimates of Life Years Lost and Harvesting.” – DRAFT