Temporary Shock or Lasting Scar? Life Expectancy Trajectories Since COVID-19

Five years after the onset of the pandemic, most high-income countries have yet to fully recover their pre-pandemic life expectancy trajectories, revealing that COVID-19 caused lasting, multi-year disruptions with distinct recovery patterns rather than a uniform, short-lived mortality shock.

The Gist

Imagine the world's population health as a long, steady marathon. For decades, runners (people in high-income countries) were getting faster and faster, shaving seconds off their time every year. This was the "pre-pandemic trend": a smooth, upward climb in how long people were expected to live.

Then, in 2020, a massive storm hit the track. This was the COVID-19 pandemic.

This new paper, written by a team of demographic scientists, asks a simple but crucial question: Five years after the storm, have the runners recovered their pace, or has the track itself been permanently damaged?

Here is the breakdown of their findings using everyday analogies:

1. The "Deficit" (The Missing Miles)

The researchers didn't just look at how many people died; they looked at the gap between where we should have been and where we actually are.

Think of it like a bank account. Before the pandemic, your life expectancy was a savings account that grew by a few dollars every year. The pandemic was a massive withdrawal.

• The Counterfactual: This is the "what if" scenario. It's the balance your account would have had if the storm never happened.
• The Deficit: This is the difference between your actual balance and that "what if" balance.
• The Finding: In 2024, five years later, 31 out of 34 countries still have a negative balance. They haven't just lost a little money; they are still operating below their pre-storm trajectory.

2. Four Different Stories (The Four Types of Runners)

Not all countries reacted the same way. The authors found four distinct "personas" in how these countries handled the storm:

• The "First Wave Peak" (The Early Screamers):
  • Countries: Italy, Spain, UK, Belgium.
  • The Story: These countries got hit hard right at the start (2020). It was a massive shock. But, like a runner who trips early but gets up and sprints hard, they recovered relatively quickly. By 2024, they were close to their old pace, though they still carry a "cumulative debt" of lost years.
• The "Second Wave Peak" (The Late Bloomers of Doom):
  • Countries: USA, Bulgaria, Poland, Chile.
  • The Story: These countries had a rough start, but the real disaster happened in 2021. It was like a runner who trips, gets up, and then immediately trips again, harder. This group suffered the biggest total loss. The USA, for example, has the largest "debt" of lost life years of any country studied. They haven't bounced back yet.
• The "Late Peak" (The Delayed Impact):
  • Countries: Japan, Norway, Denmark.
  • The Story: These countries were like runners who stayed in the shelter while the storm raged outside. They had almost no impact in 2020 or 2021. But when the storm finally broke through in 2022, it hit them. They are still recovering, and their "debt" is still fresh and unresolved.
• The "Prolonged Depression" (The Slow Bleed):
  • Countries: Netherlands, Portugal, France.
  • The Story: These countries didn't have one giant crash. Instead, they had a slow, persistent leak. Every year, they lost a little bit of life expectancy, never having a "peak" year of disaster, but never really recovering either. It's like a runner who is constantly slowing down, year after year.

3. The "Scarring" vs. The "Bounce Back"

There are two theories about what happens after a disaster:

1. The Bounce Back (Temporary Shock): The storm kills the weakest runners, and the remaining runners are so healthy and motivated that they run faster than before to make up for lost time. The track is fine; we just need to catch up.
2. The Scar (Lasting Damage): The storm damaged the track itself. Maybe the runners are injured (Long COVID), the medical clinics are broken, or the stress of the event made everyone sicker. The track is now permanently slower.

What the paper found:
It's mostly The Scar.
While some countries (mostly in Eastern Europe) are showing signs of a "bounce back," most countries are still running slower than they were before. The data suggests that for many places, the pandemic didn't just pause the race; it changed the terrain.

4. Who Got Hurt the Most?

• The USA: Had the worst cumulative loss. It's like a runner who started the race with a heavy backpack (pre-existing health issues) and then got hit by the storm. They are still far behind where they should be.
• Middle-Aged Runners: In countries like the US and Eastern Europe, the biggest losses weren't just among the elderly (who were expected to be most vulnerable). There was a surprising amount of "missing time" for people aged 40–60. This suggests the storm didn't just kill the frail; it hurt the working-age population through heart issues, drug overdoses, and other indirect effects.

The Bottom Line

Five years later, the pandemic is not just a "bad year" that we have moved past. For most of the wealthy world, it has left a lasting scar on our life expectancy.

We haven't just lost a few years of progress; we have lost the momentum we had. The question now isn't just "when will we get back to normal?" but "is the track broken, and do we need to rebuild the whole road?"

The authors warn that if we don't fix the underlying issues (healthcare access, social stress, chronic disease), we might be stuck running a slower race for decades to come.

Technical Summary

1. Problem Statement

The COVID-19 pandemic caused the most significant decline in life expectancy (LE) in many high-income countries since World War II, reversing decades of steady improvement. A critical unanswered question remains: Is this a temporary mortality shock with a rapid "bounce-back" to pre-pandemic trajectories, or does it represent a lasting structural scar that has permanently altered the level or slope of mortality improvement?

Historically, LE reversals (e.g., post-1950) were often short-lived, followed by rapid recovery. However, recent trends suggest mortality improvements have already slowed in many high-income nations. The authors investigate whether the pandemic has induced a transient deviation (Scenario A), a permanent level shift (Scenario B), or a structural break in the rate of improvement (Scenario C).

2. Methodology

The study employs a counterfactual forecasting approach to quantify "life expectancy deficits" (the difference between observed and expected LE) across 34 high-income countries from 2020 to 2024.

Data Sources:

• Death Counts: Sourced from the Short-Term Mortality Fluctuations Database (HMD-STMF), Human Mortality Database (HMD), UN World Population Prospects (WPP-2024), and national statistical offices (e.g., ONS, CDC).
• Population: Mid-year population estimates converted to person-years of exposure.
• Scope: 34 countries (including USA, UK nations, EU members, Japan, Taiwan, Chile, etc.) covering the period 2000–2024.

Analytical Pipeline:

1. Data Harmonization:
   • Ungrouping: Age-specific death counts from weekly data (STMF) were harmonized to single-year age groups using the Penalized Composite Link Method (PCLM) to handle varying age bin widths.
   • Bias Correction: A Poisson model was used to estimate correction factors, aligning STMF weekly data with official annual HMD totals to correct for systematic reporting biases.
2. Counterfactual Forecasting:
   • A Poisson-Lee-Carter model was fitted to pre-pandemic data (2000–2019) using the StMoMo R package.
   • Future mortality rates (2020–2024) were forecasted assuming the continuation of pre-pandemic trends (random walk with drift for the time index $\kappa_t$).
3. Life Table Construction & Deficit Calculation:
   • Period life tables were constructed using a standard piecewise-exponential model.
   • Life Expectancy Deficit ($e_0^{deficit}$): Calculated as $e_0^{actual} - e_0^{expected}$. Negative values indicate a deficit (excess mortality).
   • Decomposition: The Arriaga method was used to decompose deficits by age group.
4. Uncertainty Quantification:
   • Simulation: A parametric bootstrap (250 Poisson replicates) was applied to death counts, combined with random walk simulations of the Lee-Carter forecast.
   • Inference: P-values were derived to determine if observed deficits were statistically significant given the null hypothesis of continued pre-pandemic trends.

3. Key Contributions

• Extended Timeline: Provides the first comprehensive analysis of LE trajectories through 2024, five years after the pandemic's onset, moving beyond initial 2020–2022 snapshots.
• Typology of Recovery: Identifies four distinct mortality trajectory typologies rather than a uniform global experience:
  1. First Wave Peak: Large 2020 deficits with gradual recovery (e.g., Italy, Spain, Sweden).
  2. Second Wave Peak: Largest deficits in 2021 with sharper rebounds (e.g., USA, Bulgaria, Eastern Europe).
  3. Late Peak: Minimal early impact, deficits emerging in 2022 (e.g., Japan, Norway, Taiwan).
  4. Prolonged Depression: Persistent, smaller deficits without a sharp peak (e.g., Netherlands, Portugal, Israel).
• Counterfactual Rigor: Moves beyond simple excess death counts to trend-adjusted LE deficits, accounting for the fact that some countries were already experiencing stagnating mortality improvements prior to 2020.

4. Key Results

• Incomplete Recovery: As of 2024, 31 out of 34 countries still had lower life expectancy than predicted by pre-pandemic trends. Only Iceland, Luxembourg, and (marginally) Scotland showed non-significant deficits in 2024.
• Cumulative Impact: Across the 2020–2024 period, cumulative deficits were statistically significant in nearly all countries.
  • Highest Cumulative Deficits: USA (-1.73 years), Bulgaria (-1.49 years), Latvia (-1.47 years).
  • Lowest Cumulative Deficits: Denmark (-0.43 years), Sweden (-0.41 years).
• Age-Specific Patterns:
  • "Second Wave Peak" countries (e.g., USA, Eastern Europe) showed significant excess mortality in younger age groups (40–59), suggesting factors beyond direct viral mortality (e.g., drug/alcohol use, cardiovascular stress).
  • "First Wave Peak" countries saw deficits concentrated in older ages (>60).
  • Mortality Displacement: Evidence of "mirroring" (faster recovery than expected at older ages) was found in a subset of Eastern European countries (Bulgaria, Croatia, Hungary, Lithuania, Slovakia), suggesting the removal of frail individuals. However, this was not a dominant global pattern.
• Sex Differences: Males generally experienced larger deficits than females, with the USA and Latvia showing the largest sex gaps.
• Trajectory Shifts: The persistence of deficits in 2024 suggests that for many countries, the pandemic has not just caused a temporary dip but has shifted the baseline level of life expectancy or slowed the rate of improvement (Scenarios B and C).

5. Significance and Implications

• Demographic Projections: The findings challenge the assumptions of recent UN World Population Prospects (2024), which assumed a return to pre-pandemic trajectories. The authors argue this assumption is premature for most high-income nations.
• Public Health Policy: The persistence of deficits, particularly in mid-life (40–59) in countries like the USA, indicates that the pandemic's impact extends beyond direct viral mortality. It highlights the role of indirect effects (health system disruption, behavioral changes, chronic disease management).
• Long-term Scarring: The study suggests that for many countries, the pandemic may have induced a "structural break" in mortality trends. If the rate of mortality improvement has slowed permanently, the long-term implications for aging populations, pension systems, and healthcare burdens are profound.
• Inequality: The analysis underscores that recovery is not uniform; pre-existing vulnerabilities (e.g., obesity, diabetes, healthcare access) and policy responses (vaccination rates, lockdown stringency) significantly shaped the magnitude and duration of the deficits.

In conclusion, the paper argues that the COVID-19 pandemic was not a uniform, short-lived shock. Instead, it created multi-year disruptions with variable recovery patterns, leaving a "lasting scar" on the life expectancy trajectories of most high-income countries five years post-onset.