Stump the Scientist Scenes from AAAS Science After Hours


February 17, 2001

Complex Hearts
by Pearl Tesler

Here's a test. Forget about genetics and your HMO and imagine for a second that you really are in control of your health. You can choose one of these two resting heart rates to be yours. Which do you want, the nice smooth one or the wacky bumpy one?

Surprise: that nice smooth pattern belongs to someone with severe congestive heart failure. The bumpy pattern is a healthy heart.

Dr. Ary Goldberger, Professor of Medicine at Harvard Medical School, explains that the heart is a complex system, one that displays the characteristics common to complexity: nonlinearity, emergent phenomena, and fractal structure.

The bumps in a healthy heart rate pattern are examples of a fractal pattern. By definition, a fractal is a pattern that displays self-similarity and scale invariance. What this means is, the closer you look, the more structure you see, and these smaller structures are little copies of the larger structures. A frequently-cited example is a tree. Each branch mimics the structure of the entire tree, and each branch branches off into smaller branches that continue the pattern.

Showing graphs of a healthy heartbeat pattern, Goldberger notes that these graphs show fractal structure. A graph that shows the meanderings of heart rate over 300 minutes looks just like the one that spans only 30 minutes. Zooming in again, the graph that covers only 3 minutes looks roughly identical.

What is Complexity?
Complexity, also called "chaos" or "nonlinear dynamics," refers to systems that have lots of interdependent factors. These systems tend to display the following traits:

nonlinearity (output isn't proportional to input)
• bifurcations, or abrupt changes
• fractal structure (scale invariance and self-similarity)
• period doubling
• nonlinear waves (solitons, spirals, scrolls)
• chaos

emergent phenomena (strange attractors)

When this healthy randomness breaks down, watch out. When discernible patterns begin to show up, it often signals disease. In graphs of patients experiencing altitude sickness and sleep apnea, patterns like sinusoidal waves appear.

Goldberger says this trait isn't exclusive to hearts. "The output of many systems actually becomes more regular with pathologies. Patients with a wide range of disorders often display strikingly ordered dynamics." As examples, he cites Parkinson's, nystagmus (periodic motions of the eyes), and obsessive compulsive disorder.

Goldberger hopes that research into the complex dynamics of biological systems can translate into useful diagnostics. A better understanding of "healthy disorder" and its ominous disappearance could lead to "really smart" heart monitors or, in neuroscience, a device to forecast seizures.

<< Back to Dispatches