Health message from the heart

By DUNCAN GRAHAM-ROWE

How do you tell a healthy heart from one that could stop without warning?

By measuring variations in the length of the heartbeat, according to a team of researchers in Greece.

The finding could provide a way to screen for people at risk of sudden cardiac death.

Such people's heartbeat often looks perfectly healthy by conventional criteria.

Yet a quarter of a million people die each year in the US alone when their heart suddenly stops and - like the soccer player Marc-Vivien Foe who collapsed and died last year while playing for Cameroon - many of them have had no history of heart problems.

Even a person's ECG, or electrocardiogram, can look normal for much of the time. In patients with Brugada syndrome, for example, abnormal electrical signals sporadically stop their hearts from pumping properly.

Long QT syndrome is a similar condition, which can strike young, fit adults.

Standard approaches to analysing ECGs tend to focus on the peaks and troughs of the trace.

Instead, Panayiotis Varotsos of the University of Athens has been studying the variation in the length of time it takes for the heart to complete one beat.

The amount of variation in the rate of heartbeats is already used to measure aerobic fitness, with more variation meaning a fitter heart. However, for Varotsos the crucial test is the variation in the length of each beat, and whether this variation is random.

He adapted equations he had previously used to describe physical systems such as earthquakes to predict that, in a healthy heart, these variations will have some degree of order.

But if there is something wrong with the heart, it should disrupt that order, making the variation more random.

To test the theory, Varotsos and his colleagues analysed 95 sample ECGs taken from public databases of people with various heart conditions and 10 from healthy patients.

He found that the beats of the diseased hearts varied more randomly.

Varotsos says the method could be used as an initial screen to flag all types of heart problems.

A lot of research as gone into discovering ways to identify cardiac diseases from an ECG.

Some have used data mining techniques - screening blind for any effect that comes up, while other studies have looked for chaotic signatures that might distinguish unhealthy hearts.

But so far no method has stood up to scrutiny in clinical trials.

Varotsos believes his discovery has a better chance of turning out to be real because he used a physical model of how the heart works to predict a specific effect.

If it proves reliable, the method could be particularly useful for screening those who have a family history of sudden cardiac death.

This may not be enough to give rise to a nationwide screening programme.

Instead, Varotsos suggests that cardiologists could apply his method to Holter monitors - the portable ECG devices that are used to monitor patients thought to be at risk.

HOW IT WORKS


* Researchers say an uneven length of heartbeat is a pointer to potentially fatal heart disease.

* They suggest the method could be used to monitor patients known to be at risk.