Early Population-Level Impact of Helicobacter pylori Eradication on Gastric Cancer Deaths in Japan: A Counterfactual Analysis of Short-Term Divergence

This study demonstrates that the nationwide expansion of *Helicobacter pylori* eradication therapy in Japan since 2013 has already contributed to a 10.4% reduction in gastric cancer deaths by 2021, preventing an estimated 1,427 deaths through a counterfactual analysis framework that accounts for the disease's long natural history.

The Gist

The Big Picture: A "Time-Travel" Experiment

Imagine Japan is a giant garden where a specific type of weed (Stomach Cancer) has been growing for a long time. For decades, the main cause of this weed was a tiny, invisible pest called H. pylori bacteria.

In 2013, the Japanese government decided to launch a massive campaign to spray a "pesticide" (eradication therapy) on anyone with chronic stomach inflammation. They paid for it through national insurance, so almost everyone could get treated.

The Problem: Weeds don't disappear overnight. Even if you spray them today, it takes years for the garden to look different. So, by 2021 (only 8 years later), critics might have said, "The garden still looks pretty full of weeds. Did the spraying actually work?"

The Solution: This study acts like a time-travel simulation. The researcher asked: "What would the garden look like today if we had never started spraying the pesticide in 2013?"

How They Did It: The "Two-Layer" Detective Work

The researcher didn't just count the dead weeds; she built a complex model to understand why they were dying. She used a two-step detective approach:

1. Layer 1: The "Biological Blueprint"
   She built a model that understands how the weed grows. She didn't just look at the current year; she looked at:

   • Age: Older people are more likely to have the weed.
   • History: How long has the person been infected?
   • Cohorts: Different generations of Japanese people have different infection rates (older generations had more bacteria).
   • Analogy: Imagine she built a weather forecast that accounts not just for today's rain, but for the droughts and storms of the last 50 years.

2. Layer 2: The "What-If" Scenario (Counterfactual)
   She ran two simulations side-by-side:

   • Scenario A (Reality): The garden where they sprayed the pesticide starting in 2013.
   • Scenario B (The "No-Spray" World): A parallel universe where they didn't spray anything after 2013, and the treatment rates stayed low like they were before.

The Results: The Gap Widens

When she compared the two worlds, she found a clear difference:

• In the Real World: Stomach cancer deaths dropped from about 48,600 in 2013 to 41,600 in 2021.
• In the "No-Spray" World: Deaths would have only dropped slightly (to about 45,600) because of natural trends, but not nearly as much.

The "Divergence":
By 2021, there was a gap of 4,030 fewer deaths in the real world compared to the "No-Spray" world.

• Analogy: Imagine two runners. One is running normally (the "No-Spray" world). The other runner (Real World) started using a new pair of super-shoes in 2013. By 2021, the runner with the shoes is 4,000 meters ahead. That distance is the "extra" lives saved.

The "Aha!" Moment

The study found that the "pesticide" (eradication therapy) was responsible for preventing 1,427 deaths between 2013 and 2021.

• In 2015, the effect was tiny (only 17 lives saved).
• By 2021, the effect was huge (417 lives saved that year alone).

This proves that even though cancer takes a long time to develop, primary prevention (stopping the cause before the disease starts) can show results much faster than people thought—within just a decade.

Who Benefited Most?

The "super-shoes" worked best for people aged 50 to 79.

• Why? This group had the highest rate of infection, but they were also the ones most likely to get treated once the insurance coverage started. It's like spraying the part of the garden where the weeds were thickest.

The Takeaway

This paper is a victory lap for public health policy. It proves that:

1. You don't have to wait 30 years to see if a prevention strategy works.
2. Japan's decision to pay for H. pylori treatment for everyone with stomach inflammation was a smart move that is already saving thousands of lives.
3. The "Gap" is growing: Every year, the difference between "what happened" and "what would have happened" gets bigger, meaning the strategy is becoming even more effective over time.

In short: Japan sprayed the weeds, and even though it's only been a few years, the garden is already noticeably healthier than it would have been otherwise.

Technical Summary

1. Problem Statement

• Context: Gastric cancer (GC) is the third leading cause of cancer-related death in Japan, with Helicobacter pylori (H. pylori) infection accounting for 98% of cases. In 2013, Japan became the first country to implement nationwide insurance coverage for H. pylori eradication therapy for patients with chronic gastritis.
• The Challenge: While the long-term impact of eradication is theoretically understood, quantifying its short-term impact on GC mortality is methodologically difficult.
  • GC has a long natural history (latency period), meaning deaths today reflect infections and exposures from decades ago.
  • Observed mortality trends are confounded by overlapping structural factors: population aging, cohort replacement (younger generations with lower infection rates), and evolving clinical practices.
  • Conventional "before-and-after" or simple trend analyses cannot isolate the causal contribution of the eradication policy from these background demographic shifts.
• Objective: To quantify the early, population-level reduction in GC deaths attributable specifically to the 2013–2021 expansion of H. pylori eradication, distinguishing it from natural mortality trends.

2. Methodology

The study employs a novel two-layer analytic framework combining etiologic modeling with counterfactual analysis.

A. Layer 1: Structured, Time-Dependent Multilayer Model (Incidence Reconstruction)

Instead of relying on unobservable transition parameters, the authors reconstructed GC incidence using a multiplicative formulation integrating four biological and epidemiologic components:

1. Baseline Incidence: GC incidence at age 40 in 2021 (from national cancer registry).
2. Ageing Factor: Biological risk increase associated with aging, independent of cohort exposure.
3. Duration-Related Exposure: Cumulative historical exposure not explained by aging alone.
4. Prevalence Factor: Age- and cohort-specific H. pylori prevalence derived from a meta-regression of 170,752 individuals.

Formula: $GC\ Incidence_{a,t} = Baseline \times Ageing_a \times Duration_a \times Prevalence_{a,t}$

B. Layer 2: Counterfactual Analysis

The study compares two scenarios over the period 2013–2021:

1. Observed Scenario: Actual GC deaths reflecting real-world eradication uptake.
2. Counterfactual Scenario: A modeled trajectory of expected GC deaths if eradication uptake had remained at pre-2013 levels.

• Divergence Calculation: The difference between the counterfactual and observed deaths ($GC\ deaths_{counterfactual} - GC\ deaths_{observed}$) represents the total reduction in mortality.
• Attribution: A state-transition (Markov) model was used to estimate the specific number of deaths prevented by eradication (comparing standard screening vs. screening + eradication).
• Metric: The proportion of the total mortality reduction attributable to eradication was calculated as:
  $$\frac{GC\ deaths\ prevented\ by\ eradication}{GC\ deaths\ difference}$$

Data Sources: National cancer statistics, vital statistics, published literature on eradication uptake, and regional datasets. Modeling was performed using TreeAge Pro 2025.

3. Key Results

• Mortality Trends:
  • Observed Deaths: Declined from 48,632 (2013) to 41,624 (2021).
  • Counterfactual Deaths: Declined more modestly from 49,794 (2013) to 45,654 (2021).
• The Divergence: The gap between observed and counterfactual deaths widened steadily, increasing from 1,162 in 2013 to 4,030 in 2021.
• Deaths Prevented by Eradication:
  • The model estimated that eradication prevented a cumulative total of 1,427 GC deaths between 2013 and 2021.
  • Annual prevented deaths increased from 17 in 2015 (first measurable effect) to 417 in 2021.
  • The effect was most pronounced in adults aged 50–79, aligning with the age distribution of eradication uptake and high-risk infection cohorts.
• Attributable Proportion:
  • The proportion of the total observed–counterfactual mortality reduction explained by eradication rose from 0.6% in 2015 to 10.4% in 2021.

4. Key Contributions

• Methodological Innovation: The study introduces a two-layer framework that successfully disentangles the causal impact of a specific public health policy (eradication) from complex demographic confounders (aging, cohort replacement). This overcomes the limitations of conventional time-series or interrupted time-series analyses.
• Early Impact Evidence: It provides empirical evidence that primary prevention strategies (eradication) can yield detectable population-level mortality benefits within a single decade, despite the long latency of gastric carcinogenesis.
• Quantification of Short-Term Effects: It moves beyond theoretical lifetime models to quantify the actual short-term divergence in national mortality trends, offering a template for evaluating other cancer prevention policies.

5. Significance and Implications

• Policy Validation: The findings validate Japan's 2013 policy decision, demonstrating that nationwide H. pylori eradication is an effective primary prevention strategy that is already altering national cancer trajectories.
• Global Relevance: For high-prevalence countries where GC remains a major cause of death, this study suggests that systematic eradication can produce measurable results relatively quickly, especially when integrated with risk-based screening.
• Future Outlook: As younger cohorts with lower infection rates replace older high-risk cohorts, the relative contribution of eradication to mortality reduction is expected to increase further.
• Framework Utility: The proposed analytic framework serves as a robust tool for evidence-based evaluation of short-term reductions in cancer deaths attributable to specific interventions in high-prevalence settings.

Conclusion: The study concludes that H. pylori eradication has already contributed to a 10.4% reduction in the expected decline of GC deaths in Japan by 2021, proving that primary prevention can drive rapid, measurable changes in national cancer mortality trends.