Treatment reaches the right people only if they're identified before symptoms arrive
In the long arc of medicine's struggle against silent disease, a team of researchers led by the University of Tartu has developed two mathematical models capable of identifying heart failure risk years before symptoms emerge — drawing on the health records of nearly 700,000 Estonians to build tools that any clinician, anywhere, could use with data already in hand. The work, published in the European Heart Journal, reflects a broader ambition: to shift cardiovascular care from reaction to anticipation, and to close the stubborn health gap between Eastern and Western Europe that history and habit have quietly widened.
- In Estonia, cardiovascular disease claims nearly half of all lives, yet it often goes undetected until the damage is already well underway.
- Two new models — SCORE2-HF for the general population and SMART2-HF for those with prior cardiovascular events — can project heart failure risk decades in advance using only routine clinical data.
- Eastern and Central Europe carry a disproportionately high heart failure burden, shaped by elevated smoking rates, alcohol patterns, and what some researchers describe as an epidemiological inheritance from the Soviet era.
- The models require no specialized equipment or exotic testing, making them viable for immediate integration into everyday clinical practice across diverse health systems.
- SCORE2-HF is expected to be recommended in future European Society of Cardiology guidelines, potentially transforming routine check-ups into early-warning systems for millions of patients.
Heart failure often goes undetected until a patient arrives at a clinic with an unrelated complaint — by which point the disease is already advancing. Laura Lõo, a junior research fellow at the University of Tartu, has spent recent years working with international colleagues to change that reality. Their answer is two predictive models, published in the European Heart Journal, that can identify who will develop heart failure long before the first symptom appears.
The first model, SCORE2-HF, is built for the general population. It uses routine measurements — blood pressure, BMI, smoking status, diabetes history, medication use — to estimate a person's risk of developing heart failure over the next thirty years. The second, SMART2-HF, is designed for patients who have already experienced a cardiovascular event, projecting their ten-year risk of progression. Neither model demands anything beyond what a doctor already collects.
Building them required scale. Lõo's team drew on BIG-HEART, a nationwide Estonian database covering nearly 700,000 people. A small country became a vast research laboratory, and the models were validated against other health systems to ensure broader applicability. What the data also revealed was a troubling geographic pattern: heart failure risk is significantly higher in Eastern and Central Europe than in the West, a gap shaped by smoking rates, alcohol consumption, and health behaviors that some researchers trace to the social conditions of the Soviet era.
For Estonia, the stakes are especially high. Lõo stressed that effective treatment depends on identifying at-risk patients before symptoms arrive — and these models make that possible during ordinary clinical visits. The immediate horizon is adoption: SCORE2-HF is expected to be recommended in future European Society of Cardiology guidelines, potentially turning routine check-ups into early-warning systems. The deeper question is how quickly research can become the daily practice of medicine.
Heart failure kills. In Estonia, cardiovascular disease accounts for nearly half of all deaths, yet the condition often goes undetected until a patient shows up at a doctor's office with an unrelated complaint. By then, the damage is already advancing. Laura Lõo, a junior research fellow in public health at the University of Tartu, has spent recent years working with international colleagues to change that calculus. Their solution: two new mathematical models that can predict who will develop heart failure years before the first symptom appears. The work, published in the European Heart Journal, represents a shift from reactive medicine to something closer to prevention.
The first model, called SCORE2-HF, is designed for the general population. It takes routine health measurements—blood pressure, body mass index, smoking status, diabetes diagnosis, blood pressure medication use—and projects a person's risk of developing heart failure over the next three decades. The second, SMART2-HF, targets people who have already had a cardiovascular event and estimates their ten-year risk of progressing to heart failure. Neither requires exotic testing or specialized equipment. Both work with data that any doctor already collects.
Building these models required scale. Lõo and her team drew on BIG-HEART, a nationwide Estonian database containing health and social information on everyone over 36 living in the country in 2012. That meant analyzing health indicators from nearly 700,000 Estonians. A small nation became a laboratory. "Although Estonia is a small country, it is extremely gratifying to see that we can successfully contribute to major international research projects with our data," Lõo said. The international collaboration validated the models against other health systems, ensuring they would work beyond Estonia's borders.
What the research revealed, however, was troubling geography. Heart failure risk is not evenly distributed across Europe. Eastern and Central Europe, including Estonia, show significantly higher risk levels than Western Europe. The gap appears rooted in lifestyle and historical patterns. Smoking rates remain elevated. Alcohol consumption patterns trace back decades. Some researchers have connected present-day health behaviors in former Soviet countries to the social and economic conditions of that era—a kind of epidemiological inheritance.
For the Estonian health system, the implication is clear: early detection matters more here than in wealthier Western nations. Prevention cannot be passive. Lõo emphasized that "treatment reaches the right people at the right time" only if those people are identified before symptoms arrive. The new models make that identification possible during routine clinical encounters.
The practical next step is adoption. SCORE2-HF is likely to be recommended in future clinical practice guidelines from the European Society of Cardiology. Doctors would use it as a screening tool, flagging patients at high risk and initiating preventive treatment before heart failure develops. For a health system already burdened by cardiovascular disease, catching the disease in its mathematical infancy—years before it becomes clinical—could reshape outcomes. The question now is whether the models will move from research journals into the daily work of medicine.
Notable Quotes
Treatment reaches the right people at the right time only if those people are identified before symptoms arrive— Laura Lõo, University of Tartu
Heart failure is often detected only when a person visits a doctor for another reason— Laura Lõo, University of Tartu
The Hearth Conversation Another angle on the story
Why does heart failure so often go undetected until it's advanced?
Because it's silent. People don't feel it coming. They show up at the doctor for something else—shortness of breath they attribute to age, fatigue they blame on work—and only then does someone listen closely enough to hear the problem.
And these models change that by predicting it before symptoms?
Exactly. They use data doctors already have. Blood pressure, weight, whether you smoke. The math finds the pattern in 700,000 people's lives and says: this person is on a path toward heart failure.
But why is the risk so much higher in Estonia than in Germany or France?
That's the harder question. It's not genetics. It's how people live. Smoking is more common. Drinking patterns are different. Some of it traces back to Soviet times—the way those decades shaped behavior that persists today.
So the model is good, but it's also a mirror of inequality?
Yes. It shows us where prevention is most urgent. In Estonia, you can't wait for symptoms. You have to find people early.
Will doctors actually use these models, or will they sit in journals?
That's the real test. If the European Society of Cardiology recommends SCORE2-HF in their guidelines, it becomes standard. Then it's not optional. It becomes how you practice.