Bring evidence together
Connect health, environment and social context through a shared country-and-year data model.
Global health · Data · Evidence
Health is never
without reason.
Data helps us understand why.
GHEAS is an independent, data-driven platform for exploring how environment, development and health systems shape outcomes across countries and over time.
CONTRIBUTING FACTORS CONNECTED VIEW
Air quality
Income
Nutrition
Healthcare
Lifestyle
Climate
HEALTHOUTCOMESObserved over time
UNIT OF ANALYSISCountry × Year
METHODMulti-factor
PRINCIPLEEvidence first
A wider view of health01 Air & environment02 Disease burden03 Social determinants04 Health systems
01 — PURPOSE
健康，从来就不是没有缘由的。
数据帮助我们揭示其中的原因。
A health outcome rarely belongs to one cause. It emerges from exposure, behaviour, biology, resources, place and time.
GHEAS brings these fragments into view—carefully, transparently and without pretending that correlation alone is an answer.
The first complete release demonstrates that workflow end to end: source the data, record the rules, compare models, preserve uncertainty, and publish a reviewable evidence report.
02 — APPROACHFrom scattered data
From scattered data
to clearer evidence.
A completed first-study workflow designed to be repeated as the questions, datasets and methods become more sophisticated.
Make relationships visible
Use transparent statistics and clear visualizations to surface patterns, differences and uncertainty.
Explain with restraint
Distinguish association from causation, preserve source context and make every conclusion traceable.
03 — PLATFORM ARCHITECTURE
A growing structure for evidence, not just a single page.
The first GHEAS release now connects all five planned modules: source records, a repeatable pipeline, cross-sectional and panel analysis, interactive visual exploration, and a formal evidence report.
Data Library
Record datasets, indicators, versions, sources and licensing in a traceable way.
Data Pipeline
Standardize countries and years, then prepare repeatable cleaning and matching workflows.
Analysis Lab
Run descriptive comparisons, subgroup analysis, regression checks and sensitivity tests.
Visual Explorer
Turn results into maps, trends and charts that remain honest about uncertainty.
Evidence Report
Compile methods, limits and conclusion levels into a report that can be reviewed and reused.
05 — RELEASE MILESTONE
Research Paper 01 is complete and reproducible.
01Integrated a country-year data workflow.
02Presented descriptive evidence, model adjustments and interpretation limits.
03Completed two-way fixed effects, sensitivity checks and independent numerical verification.
06 — RESEARCH REGISTRYOne completed study.
One completed study.
One deeper test underway.
GHEAS now separates completed findings from active protocols. Every project shows its evidence stage before it shows a conclusion.
01 · COMPLETE
PM2.5 and life expectancy
A frozen country-year study with dual-track sensitivity analysis, a formal paper and reproducible outputs.
02 · ACTIVEPM2.5 and mortality evidence
A systematic review and Meta-analysis that tests how design, measurement and transparency shape published estimates.
OPEN REGISTRYFollow the evidence trail
See protocols, validated searches, screening progress and exactly which claims are—and are not—ready.
08 — WHAT IS READY
A complete first research release, not just a prototype.
A reviewer can inspect the framing, sources, model sequence, numerical outputs, evidence boundaries and paper without needing additional context.
01 · Narrative framing
The homepage explains the project purpose, the evidence philosophy and the first study in a coherent, presentation-ready sequence.
02 · Report-style analysis
The analysis page unifies provenance, descriptive evidence, adjusted models, interpretation and limits into one reviewable experience.
03 · Evidence boundary
The work explicitly states that the current result is descriptive and model-based, not a final causal claim.
04 · Reproducible release
The data, analysis programs, machine-readable results, formal paper and public download are linked as one release.
04 — FIRST STUDY Module 01 · Demonstration study
GLOBAL COUNTRY COMPARISON · MODULE 01 · VERSION 1.0
What can air quality tell us about life expectancy?
ENVIRONMENTPM2.5 exposure
HEALTH OUTCOMELife expectancy
COVERAGE200 countries · 1990–2023
The first dataset now connects 6,800 country-year observations. The study now includes descriptive comparisons, pre-specified multivariable adjustment, a 4,406-observation country panel, lag and sample sensitivity checks, and independent numerical verification.
Starting globally. Learning carefully. Expanding responsibly.Open the completed analysis