"The first thing to understand," says Peter Cavanagh, "is that there's no such thing as standing still."

Cavanagh, a lean, nimble man with a neat reddish beard, pops up from his chair to demonstrate.

"Even when we think we're standing still, we're engaged in continuous sway," says the 46-year-old distinguished professor of locomotion studies, biobehavioral health, medicine, and orthopedics. "The body oscillates in an apparently random way -- Call it micro-sway." He stands with his hands at his sides.

"We have looked at patterns of people who have tried to stand still."

He goes on swaying imperceptibly for a minute.

"To make it worse," he says at last, "all you have to do is close your eyes." He does so, and his movement grows noticeable.

"Next," says Cavanagh, "tip the head back."

Now we're swaying.

"The balance organs of the inner ear work best when the head is erect," Cavanagh says, nose pointed to the ceiling, in his faded British accent. "Tipping the head back puts them at a disadvantage."

If he wanted to take things further, Cavanagh would now place a thickness of foam, or something equally unstable, under his feet.

"On the soles of our feet," he explains, "we have hundreds of receptors that respond continually to pressure. We also have joint-angle sensors in our ankles that note how the angle between foot and leg changes during sway. Standing on foam decreases the efficiency of this measuring system."

No foam handy, Cavanagh retakes his seat.

As director of Penn State's Center for Locomotion Studies (CELOS), Cavanagh has put numerous experimental subjects -- safely tethered to an overhead track -- through similarly disorienting paces, as part of a continuing effort to understand the biomechanics of balance. Keeping ourselves upright is, after all, the first step in locomotion, in getting ourselves, bipedally, from one place to the next: from the car to the mall, say, or from the couch to the refrigerator. Anybody who thinks it's a snap to get this system running in synch needs to have a look at a 10-month-old baby taking it for an early test spin.

Oh, sure, soon enough we get pretty smooth. We come to take it for granted, just as later we do the ability to drive a car. If all goes well, we cruise along, what once seemed impossible having become second nature.

But should a "pathology" develop in one of the control systems, our "postural stability" is going to drop -- and chances are that we are going to drop, too, hitting the deck with a painful thud, or worse.

Such pathology develops naturally, unfortunately, with age. The body's control systems slowly degrade. When the effects of aging are combined with those of a degenerative disease like diabetes, the change can be dramatic.

People with diabetes suffer many side-effects. One of the more common of these is neuropathy -- loss of feeling -- in the feet. The anesthetic effect of neuropathy can be gruesomely complete: Cavanagh tells of a carpenter who, having trouble removing his boot, summoned his wife to help. What she noticed -- and he couldn't feel -- was a nail driven clear through the boot's sole, its point protruding from the leather upper.

Even when not so advanced, neuropathy is a serious problem; and one of the things it affects is balance.

"People with diabetic neuropathy sway as much with their eyes open and their heads forward as non-diabetics do with eyes closed and head back," Cavanagh reports. Their lack of sensitivity makes these diabetics, many of them elderly and fragile, more susceptible to falls, and thus to serious injury.

As Cavanagh's colleague, assistant professor Ge Wu, phrases it, "They dont know where their feet are."

# "The foot," Cavanagh says, "is the key element in locomotion."

A skeletal foot sits on the table, reinforcing the point. The model is nicely flexible; its numerous small bones are carefully labeled, the joints connected with tiny springs.

"The feet -- and the legs. We put these organs under the microscope. We want to understand them from the engineering perspective," Cavanagh says, of the work being done in CELOS.

He started into locomotion as a doctoral student, at the Royal Free Medical School in London. There he scrutinized walking by measuring the electrical activity in muscle fibers. "I found that with sensitive equipment you can detect unequivocally and precisely when the muscle is working." Typically, this result got him to thinking practically: "Suppose you had somebody with one limb paralyzed and the other intact," he thought. "If you had a robotic device to move the paralyzed leg, conceivably the signal from the good leg could tell the other one when to move."

About that time, Cavanagh says, "I got into running. I ran marathons -- Boston, Avenue of the Giants, Boston again. And I got interested in energetics, the mechanics of locomotion. It's like two cars that look the same but one uses twice as much gas to go the same distance. Why is it that two people can have the same body weight, the same structure, yet one can run on 20 percent less oxygen?"

He also became interested in running injuries, which is what first swung his attention to feet. "With the running boom -- this was the late '70s -- had come this boom in injuries. Overuse injuries. And I realized that footwear was the key to understanding these injuries, and to preventing them: controlling excessive motion and cushioning."

At Penn State, Cavanagh became a recognized authority on running-shoe design. He consulted for major shoe manufacturers, designing a number of shoes, including one that was fitted out with a miniature computer that recorded lap times, calories burned, and other crucial data. ("An idea," he says, "whose time hasn't come.") In 1980, he authored The Running Shoe Book.

In 1981 Cavanagh and then graduate student Ewald Henning invented an electronic device for measuring the pressure applied to the soles of the feet during standing and walking. By standing on this pressure platform, a person could obtain a color-coded computerized image of the bottoms of the feet which pinpointed areas of high pressure. The device would be useful for identifying potential problem areas before they resulted in injury.

Yet while developing this imaging device, Cavanagh had begun to tire of sports medicine. For one thing, he says, he grew leery of the use that was being made of his data by running-shoe manufacturers. As he grew older, he wanted to make more of a contribution to society. He envisioned another use for the pressure platform.

"I realized that it could be applied in a very different context," he says, "to help people who wanted not to run faster but to keep their feet."

# Neuropathy, the loss of feeling, can strike the arms as well as the legs, hands as well as feet. It usually occurs symmetrically -- if one limb is affected, so is the other -- and it is worse in distal areas, those places farthest from the body's center. But it is especially a problem in the feet.

"The nerves just die," says Jan Ulbrecht, a diabetologist and member of the CELOS faculty. "It's a biochemical process that we can't explain yet."

What is known is that this loss of the protective sensation of pain in a heavy-stress area like the bottom of the foot can lead to real trouble. Unnoticed rubbing or chafing from something as minor as a tight shoe can easily result in an ulcer, which goes equally unfelt. Left untreated for weeks, months, even years, the ulcer becomes infected, and the infection spreads, eventually to bone. Too often, the final result is catastrophic: Each year over 60,000 lower-limb amputations are performed in the United States as a result of diabetes-related complications. In addition to the human toll -- amputation's immense physical and emotional impact, and the poor prognosis for long-term survival -- this drastic procedure is an economic nightmare. Each amputation costs $50-60,000 to perform.

The thing is, Cavanagh says, many of these amputations are preventable.

Until recently, by the time a diabetic patient developed a serious foot ulcer, it was all over. Such ulcers were almost impossible to heal, and even if a clinician managed to heal one, it would soon return. Loss of blood circulation was blamed for most foot problems, and standard practice was to amputate sooner rather than later, in order to prevent even worse problems.

In 1991 Cavanagh and Ulbrecht opened up the Diabetic Foot Clinic as a joint program of CELOS and the Nittany Valley Rehabilitation Hospital. The idea was to learn more about the diabetic foot in order to improve treatment. One of their goals was a series of tests to predict who would get foot ulcers and where.

Today, having treated more than 750 patients at University Park and at a second clinic in the Hershey Medical Center, Cavanagh and Ulbrecht argue that neuropathy, not poor circulation, is the major culprit in the diabetic foot. They have successfully healed ulcers using a special weight-bearing cast and a tough antibiotic regimen. Testimonials abound from diabetics who faced amputation. In September 1994, Cavanagh, Ulbrecht, and Gregory Caputo of Hershey, along with colleagues at the Harvard Medical School, published a definitive guide to management of the diabetic foot in the New England Journal of Medicine, emphasizing careful screening and early detection.

The key, says Cavanagh, is vigilance. Especially after an initial ulcer has been healed, "The patient remains at lifetime risk," Cavanagh stresses. "You cannot relax your guard."

In addition to frequent, thorough examinations, the foundation for preventing recurrence is protective footwear.

Improving footwear design is one of the lab's current concerns. Even at the clinic, Cavanagh says, "prescribing footwear is not the science we would like it to be." It's largely a hit-or-miss process.

"What we do is, we give a patient a specially designed shoe and say, Tell us how it goes. But what if it doesn't go well? The stakes are too high. Every ulcer creates a very high risk for losing a limb."

A better way to prescribe, Cavanagh suggests, would be to do the guesswork on a computer model, to predict on a simulated foot where problems are likely to occur, instead of correcting for them after the fact. A pair of shoes could be designed in accordance with the prediction of where pressures are likely to be highest. "Then, once they're as good as they can be, we bring the patient into the lab and measure the interaction with the foot using this." He holds up a thin green insole, made of what looks like molded rubber. It's actually a portable version of the pressure platform: the insole contains 100 electrodes, which are wired to a computer on the patient's waist.

Such complex modeling, involving so many variables, requires considerable computer power. CELOS post-doctoral fellow David Lemon is working on refining the mathematics using the resources of the Pittsburgh Supercomputer Center.

Meanwhile, CELOS researchers continue to investigate other alternatives to below-the-knee amputation. Ph.D. student John Garbalosa recently presented results of a study that looked at the efficacy of a partial amputation technique.

"Many surgeons," Cavanagh explains, "if they see an ulcer on the big toe, will take the whole leg below the knee, figuring that's eventually what's going to happen anyway. The recurrence of infection is high.

"But we think that's because of inadequate management. The partial foot has extra-special needs for protection. Given that protection, partial amputation can be quite successful."

# Out on the lab floor, Ge Wu is conducting a test. A young woman in jeans and a red sweater stands on a small raised platform. The woman is fitted snugly with a black harness that loops around her shoulders, crosses her chest, and encircles her thighs like a parachute rig; the harness is connected to a strap hooked to the ceiling. At a silent signal, the platform suddenly jerks back a few inches. The woman lurches forward, her hands fly up, and she recovers -- all in an instant.

Wu is trying to understand how people maintain their balance, given a sudden "sensory challenge": Or, put negatively, what makes them fall.

It's not as much fun as it looks. "Falls," says Cavanagh, "are the leading accidental cause of injury and death in the elderly."

Balance, as Cavanagh has already demonstrated, is the function of a complicated system. Three systems, actually: the visual, the vestibular (inner ear), and the proprioceptive (that's the sensory apparatus in feet and ankles). These control systems interact in a dense network of signal and feedback.

All in that instant after the bus jerks to a stop or the escalator pulls your feet out from under you, the brain gathers and integrates information from eyes, ears, and feet in order to make the proper response, to initiate commands to the appropriate muscles, to scream out the warning: Mayday! Mayday! We're about to fall!

"The thing that makes it tricky," says Wu, an engineer, "is that the control mechanism is over-redundant."

The human body, it seems, is designed with a certain amount of built-in overlap. To some degree, the systems cover for one another. While this is great for our survival, Wu acknowledges, it makes the system that much harder to comprehend.

She combines two approaches. The first is experimental, and involves manipulating environmental conditions to separate out the role of each physiologic system. In this context, Cavanagh notes, patients with diabetic neuropathy are valuable test subjects: their lack of feeling can reveal the proprioceptive contribution.

The second approach involves modeling the particulars of balance and posture on the computer.

"This is especially challenging, because mechanically the human body has so many degrees of freedom," Wu says. Her model human has six simplified joints, at the hips, knees, and ankles. Unlike many such models, it simulates movement in three dimensions.

Already, Wu has been able to simulate the amount of torque or force that acts at each joint in the effort to maintain balance, given a perturbation like the platform jerk-back. The amount of torque increases with distance from the body's center: the ankle is the most important joint for balance control. "This is consistent with experimental findings," Wu says, "which confirms that our model is working." Her eventual hope is to use the model to predict falls, showing exactly what types of movements put a body at risk.

For now, though, understanding falls means strapping subjects into the test harness and letting fly. And to that approach, Wu, whose training is in the design of precision instruments, has made a substantial contribution.

As a Ph.D. student at Boston University, she developed a device called the integrated kinematic sensor, or IKS, which combines three different sensors to provide direct readings of three important variables of body movement: orientation, speed, and acceleration. (Other sensors, she explains, rely on measuring one variable and calculating the rest, a procedure which increases the amount of noise and error.)

Its real-time operation makes the IKS especially valuable. Wu has used visual feedback to show neuropathic diabetics when, in response to perturbation, their center of gravity was outside the area of their support base -- the block formed by the outsides of their feet. ("When this happens," says Wu, "you're going to fall.")

"We're trying to see if this kind of training is helpful in improving postural stability," she says.

# Don Streit comes at falling from another angle. Streit, like Wu, is a mechanical engineer. But instead of preventing people from falling, he has focused on trying to lessen their injuries when they do -- and on facilitating their recovery from injuries that can't be avoided.

One of Streit's ideas is "soft" flooring.

Streit's floor consists of two thin polyurethane sheets sandwiched around a layer of tiny columns, made of the same material. Strong enough to withstand ordinary foot traffic, the floor is designed to "give" against the greater force of someone falling: the columns buckle slightly and then rebound. Computer simulations have shown that the new floor can lessen the force of an impact by 45 percent over conventional flooring. This forgiving quality should significantly reduce the number of debilitating injuries like hip fractures, a common problem among the elderly.

"If you landed on your hip," says Streit, "the flooring would actually tend to wrap around the hip. It conforms to the part of the body that impacts it."

In the lab, Streit and Cavanagh have tested the new flooring using a simulated hip joint made of foam-covered bone. This winter, test sections were installed at a central Pennsylvania nursing home.

"Twenty five nursing homes have already called me, wondering how they can get it," Cavanagh says. Because the special flooring costs up to four times as much as standard vinyl, he adds, the idea, if the flooring does indeed prove effective at reducing injury, would be to install it in special "fall wings" frequented by people whose risk of falling is high.

Another of Streit's devices looms just across the lab. It's a large revolving arm, like a crane, whose "bucket" is a seat -like harness. The variable gravity rehabilitation system, he calls it, or, for short, the rehab device. It was designed with hip-fracture repairs and hip-replacement surgeries in mind, as a way to speed the rehabilitation process.

"With these procedures," Cavanagh explains, "rehab is very traditional -- there is none until the patient can walk with the physical therapist's help. Even then, one false move, a slip, and you can do major damage."

With Streit's device, things will be much safer -- and easier. Once a patient is secured in the harness, the revolving arm assumes most of the weight. Computer controls allow an attendant to dial in what percentage of body weight he or she wants the patient to bear, increasing the load as the patient gets stronger.

"With this device, a therapist will be able to take a patient the day after surgery -- but weighing only 25 percent of their body weight," says Cavanagh. "This is rehab under very controlled conditions."

# Hanging from the ceiling nearby, inoperative for now, is yet another mechanical attempt to defy gravity. This one looks both less substantial and more involved than Streit's crane -- a bona fide contraption. It goes by the name of the Penn State Zero Gravity Locomotion Simulator.

Astronauts in zero-gravity, Cavanagh explains, lose bone mass very quickly -- a kind of osteoporosis that begins within 24 hours of being in space. "If there is no countermeasure, astronauts arriving on Mars after a two-year flight would face a very real possibility of spontaneous broken bones."

The problem, it seems, is lack of activity: specifically, the absence of enough stress on the musculoskeletal system to keep the bones producing the necessary calcium to replace what is lost.

To help counteract this effect, the space shuttle is equipped with an exercise treadmill. But, says Cavanagh, treadmill running in zero gravity may not be giving astronaut bones the jolt they need. "If you look at NASA film of crew members exercising, you can see that they're just running on the balls of their feet without making heel contact."

Nobody really knows just how much stress is needed to stimulate bone production. In order to study the problem, NASA needs an environment of zero-gravity -- or its equivalent. Unfortunately, that's no easy thing to achieve.

Underwater simulation is one approach that has been tried.

"This is good for some things," Cavanagh says, "but not for exercise. The viscous drag is too much."

Another option is the KC-135, a modified 727 jet that flies Keplerian trajectories -- parabolic arcs -- over the Gulf of Mexico, producing, for its passengers, weightless interludes of 20 to 30 seconds.

Airplane time, however, is expensive, and the ride can be pretty harsh. Cavanagh, himself a certified pilot, calls his one KC-135 flight "my single worst experience." (The plane was scheduled to make 42 parabolas. "After nine, I lost all interest in science. I lost all interest in life.")

There had to be a better way. Sparked by an illustration in an old Russian cosmonaut-training manual, Cavanagh and graduate student Brian Davis built the zero-gravity simulator.

The device works by suspending a person horizontally three feet off the floor, using a combination of elastic cords attached by cuffs to the arms legs, torso, chest, and head -- "like a marionette," Cavanagh says. As the subject moves, the cords use pulleys to counteract the force of gravity, simulating weightlessness. The subject runs on a treadmill mounted on the wall "below" his or her feet.

Cavanagh and his team hope to use the simulator to measure the effects of weightless exertion, and eventually to design exercises that will provide the necessary stress.

In addition, he says, lessons learned in mock zero-gravity may have down-to-earth importance. Exercise could well prove to be a useful adjunct to the standard endocrine supplementation for treatment of 'ordinary' osteoporosis.

The idea brings a smile to the clinically-oriented Cavanagh.

"Once again," he says, "this is the thing I love. Research reduced into practice that influences people's lives."

Peter R. Cavanagh, Ph.D., is distinguished professor of locomotion studies, biobehavioral health, medicine, and orthopedics, and director of the Center for Locomotion Studies, Room 10 Intramural Building, University Park, PA 16802. Jan S. Ulbrecht, M.D., is associate professor of biobehavioral health and medicine, and medical director of the diabetes foot clinic at the Nittany Valley Rehabilitation Hospital. Ge Wu, Ph.D., is assistant professor of exercise and sports science and mechanical engineering. Donald Streit, Ph.D., is associate professor of mechanical engineering.

CELOS projects reported above have received funding from the National Institutes of Health, the American Diabetes Association, the Diabetes Research and Education Foundation, the Veterans Administration, and the Whitaker Foundation.

In 1994, Peter Cavanagh received the Borelli award of the American Society for Biomechanics.