Of 18 indicators, at EU level
– improved
Every mortality and survival indicator improved. Among lifestyle and diet indicators, only smoking moved meaningfully.
Data analysis · 27 EU member states · 2014–2024
Better medicine, same habits: a decade of health in Europe
Over the last ten years Europeans became much less likely to die of heart disease, stroke or cancer. Over the same ten years they did not get any thinner, more active or better fed. This report tracks 18 indicators across all 27 EU countries to separate what improved from what stood still, and shows where things got worse.
Heart & circulatory deaths
–
age-standardised rate, 2014 →
Adult obesity
–
was in 2014
Diabetes prevalence
–
was in 2014
Price of vegetables since 2015
–
vs for all consumer prices
- Treatment won the decade. Deaths from heart and circulatory disease fell and cancer deaths . Preventable and treatable deaths are both down, and life expectancy rose to years.
- Prevention stood still. Adult obesity, child obesity, diabetes, physical inactivity, alcohol and sugar are all within a rounding error of where they were in 2014. Meanwhile healthy food became the fastest-rising item on the shopping list.
- The gaps widened. Less-educated Europeans became less active while graduates became more active. In parts of the East, lifestyle risks rose sharply: ’s adult obesity rose by percentage points.
01 · The scorecard
Seven indicators improved, eleven stood still, none got worse
Each indicator compares 2014 (2015 for smoking) with the latest year available. A change counts as real only if it exceeds a stated threshold: 0.5 years of life expectancy, 1 percentage point for most prevalence rates, or 5% for death rates and food supply. The bar shows progress in units of that threshold: right is better, left is worse.
Sources: Eurostat, WHO GHO, FAOSTAT · EU-27 values for WHO and FAO series are population-weighted means of member states
02 · What medicine won
Death rates from the big killers fell steadily, except during the pandemic
Age-standardised death rates remove the effect of an ageing population, so these declines are real gains in survival. The steepest fall was in heart and circulatory disease, which is still Europe’s leading cause of death. Diabetes deaths are the exception: they never fell.
Death rates by cause, EU-27, indexed to 2014 = 100
Age-standardised death rate per 100,000. Below 100 = fewer deaths than in 2014.
Source: Eurostat hlth_cd_asdr2 · 2020–2021 include the COVID-19 pandemic
View data
03 · What habits didn’t change
Ten years of lifestyle data, and the lines are almost flat
These are the main drivers of chronic disease. At EU level only smoking declined meaningfully. Each chart starts at zero so that small movements are not exaggerated. The thin grey band shows the spread between member states (lowest to highest).
Source: WHO Global Health Observatory, modelled age-standardised estimates (NCD-RisC for obesity and diabetes).
04 · Where it got worse
The EU average hides a divergence between East and West
The map shows the change over the decade. Red means the country moved in the unhealthy direction, blue means it improved. The colour scale is centred on zero for every indicator.
Ranking
Change over the decade, worst first.
Boundaries © EuroGeographics / GISCO · Lambert equal-area projection
View data
05 · What’s on the plate
Europe eats about the same as ten years ago, but healthy food costs more
FAO food balance sheets measure the food available per person, before household waste. That is an upper bound on what people eat, but it is consistent over time. Supply barely moved. Prices did: vegetables and fruit rose faster than sweets, ready meals and inflation as a whole.
Food available per person, EU-27
Change 2013–15 → 2021–23 (3-year averages), kg per person per year.
Source: FAOSTAT Food Balance Sheets
View data
Food prices, EU-27, 2015 = 100
Harmonised consumer price index, annual average.
Source: Eurostat prc_hicp_ainr (ECOICOP v2)
View data
06 · Who is left behind
Education predicts health habits, and the gap is widening
The European Health Interview Survey (2014 and 2019 waves) breaks behaviours down by education. In every indicator, people with only basic education fare worse. Between the two waves, graduates became more active (), while people with basic education became less active ().
Health behaviours by education level, EU-27, 2019
% of adults. Hover a row for the 2014 values.
Source: Eurostat EHIS (hlth_ehis_bm1e, fv3e, fv7e, sk1e, pe9e) · education by ISCED 2011 level
View data
07 · 27 countries, 18 indicators
made the most progress; the least
Each cell is a country’s change on one indicator in threshold units: blue = improved, red = worsened, grey = no real change. Rows are sorted by average progress (capped at ±3 per indicator so that a single extreme swing cannot dominate).
Progress by country and indicator, 2014 → latest
Hover a cell for the values. Empty cells are FAO series excluded because of a suspected statistical break.
Own calculation from Eurostat, WHO and FAO data
View data
Values: 2014 (or first year) → latest, with the EU-27 latest value for reference
Did obesity rise more where people moved less?
Change in insufficient physical activity (2014→2022) vs change in adult obesity (2014→2024), by country.
Pearson r = (95% bootstrap CI , n = 27). The CI includes zero, so the link is suggestive, not conclusive.
View data
What does hold up: countries where obesity rose more also saw diabetes rise more (r = ,
95% CI ). There is no cross-country link between falling smoking and falling cancer deaths
(r = ). That is expected, because cancer reflects smoking from decades earlier.
08 · Methodology
How this analysis was built
Three official sources are harmonised into one fact table. Two commands rebuild everything from scratch.
ExtractEurostat API · WHO OData · FAOSTAT bulk
→
HarmoniseDuckDB SQL · one fact table
→
Scorethresholds · verdicts · progress
→
Validatebreak detection · bootstrap CIs
→
PublishSVG/JS report · GISCO map
Sources
- Eurostat: life expectancy (
hlth_hlye), preventable & treatable mortality (hlth_cd_apr), causes of death (hlth_cd_asdr2), self-perceived health (hlth_silc_01), EHIS behaviours, HICP food prices - WHO GHO: obesity (adult, 5–19), diabetes, smoking, alcohol, insufficient activity (modelled, age-standardised where available)
- FAOSTAT: Food Balance Sheets, food supply in kg per person per year
Data quality handled
- Break in 2013: Eurostat death rates use a new European standard population from 2013, so the analysis starts in 2014.
- Coding artefacts: Poland’s ischaemic heart disease deaths rise 44% due to a cause-of-death coding change. The scorecard uses all circulatory diseases, which is robust to that change (Poland −24%).
- Noisy food balances: FAO series use 3-year moving averages. country series with a single-year jump above 40% are flagged as statistical breaks and excluded from the verdicts.
- Healthy life years are not used for comparisons. They rest on a self-reported limitation question whose wording changed in several countries, producing implausible jumps (e.g. Italy +11 years).
Scoring rules
- Change = latest − baseline (2014; 2015 for smoking, the first WHO estimate year after 2014).
- Verdict: “Improved” / “Worsened” only if the change exceeds the indicator’s threshold in the healthy / unhealthy direction; otherwise “No real change”.
- Progress = signed change ÷ threshold (positive = better). Country ranking = mean progress, capped at ±3.
- EU-27 values for WHO and FAO series are population-weighted means of member states.
Limitations
- Latest years differ: death rates to 2023, life expectancy and WHO obesity to 2024, diabetes and inactivity to 2022, food supply to 2023.
- WHO figures are model estimates with uncertainty intervals; FAO measures food available, not food eaten.
- Threshold choices are a judgement. They are stated openly so that results can be checked.
- Cross-country correlations (n = 27) describe patterns, not causes.