On the morning of July 2, 2006, Sgt. Nick Paupore was driving the lead Humvee in a convoy near Kirkuk, in northern Iraq, when a roadside bomb blew off his right leg above the knee.
Within 48 hours, he was at Walter Reed Army Medical Center in Washington, where he has spent the past 18 months recovering. Soon after arriving, Paupore began to feel excruciating pain - in his missing leg.
"It felt like someone was shocking me, like someone was putting an electrode on the back of my ankle," says Paupore, 32.
He tried several painkillers, including methadone, but the pain didn't let up. Then a Navy neurologist, Dr. Jack W. Tsao, asked him to try a new approach that requires patients to move the intact limb while watching the action in a mirror.
Not surprisingly, Paupore was skeptical, and said no thanks.
He's not skeptical now.
Tsao eventually persuaded Paupore to try the therapy. After several weeks the shocks had almost disappeared.
"As soon as I started the treatment, I noticed a remarkable change," says Paupore, who has stopped taking painkillers. "I could see really big improvement, really fast."
No one quite understands how the therapy works, other than a suspicion that it reduces painful nerve impulses, probably in the brain.
Whatever it does, the treatment could revolutionize how doctors deal with phantom limb pain, as the problem is called. There are 2 million amputees in the United States, a number that has increased markedly in recent years with the rise in diabetes. More than 700 U.S. soldiers have lost limbs after being wounded in Iraq and Afghanistan.
Although there are no hard numbers, experts believe phantom limb pain afflicts from 10 percent to 50 percent of amputees. The ailment is often impervious to treatment. Many patients end up on drugs such as Oxycontin or Percocet; as powerful as the medicines are, they seldom work.
A few weeks ago, Tsao, 41, published his results in The New England Journal of Medicine. The study examined 18 veterans, all leg amputees suffering from phantom limb pain.
The soldiers were divided into three groups of six: One group received mirror treatment; another underwent treatment using a covered mirror, while the third didn't use a mirror, but visualized moving the amputated limb.
Those who used an uncovered mirror had significant pain relief. Few in the other groups got relief, and some actually got worse. When the covered-mirror and visualization patients tried the mirror, almost all improved.
"This is a beautifully designed study that shows, without a doubt, that patients are helped by the mirror," says Vilayanur Ramachandran, a neuroscientist at the University of California at San Diego, who invented mirror treatment 11 years ago.
Tsao's study is the first to test the mirror approach in such a rigorous way - a key step for establishing scientific proof.
"It amazes me that nobody ever did follow-up studies" on Ramachandran's work, says Tsao, an associate professor of neurology at the Uniformed Services University of the Health Sciences, the military's medical school in Bethesda. A key reason was the scarcity of amputees in any one place. But with a steady stream of injured veterans from the wars in Iraq and Afghanistan, Walter Reed had enough for a study, Tsao said.
Ramachandran says Tsao's research could increase the popularity of the treatment, which is not widely used.
"People are suffering needlessly because they're not getting this treatment," says Ramachandran. "Given how easy and inexpensive this is, it's worth trying."
Since the study appeared, Tsao has heard from doctors and therapists around the country interested in trying mirrors with their patients.
The technique is remarkably simple. The patient sits on a flat surface with his legs straight out. He puts a rectangular, 6-foot-long mirror lengthwise between his legs, with the reflective side facing the intact limb.
He then moves his good leg and watches the movement in the mirror. The reflection creates the illusion of two whole limbs moving in unison. As this is happening, the patient imagines moving the amputated leg in the same way as the uninjured one.
Patients go through this process for 15 minutes a day for a month, using a range of leg movements.
Sitting on an exercise bench in Walter Reed's rehabilitation center, Paupore says mirror therapy saved him.
"The pain wasn't something I could learn to live with," he says.
A gentle man with a crew cut, Paupore joined the Army in 2004 out of a sense of patriotism. Before enlisting, he worked in a wine and beer warehouse in Traverse City, Mich.; he hopes to find work as a civil servant for the federal government.
The mechanisms behind phantom limb pain remain mysterious. The problem could stem from a kind of cortical confusion. The brain has specific regions devoted to receiving sensory input from the body. In an uninjured person, this information - temperature, position in space, hardness and so on - flows steadily from nerve endings to cortex.
In amputees, the brain no longer receives information from the lost limb. But the receiving station still works, and without incoming data, it can go haywire.
"We think the brain cells may be firing off randomly," says Tsao. The result: The amputee feels burning, throbbing, spasms or any number of other unpleasant sensations in the absent limb. One of Tsao's patients describes feeling as if a spike were being driven continuously through his foot.
Many researchers, including Tsao, suspect that mirror therapy eases these sensations by creating the illusion of two whole limbs.
Although patients know that the mirror does not reflect reality, other parts of the brain are not so savvy. These sensory input regions accept the mirror's evidence: The amputated limb reappears, provides the brain with sensory data and stops the random firing.
Tsao's research provides more evidence for an increasingly popular idea, that pain, particularly the chronic sort, is more a problem of brain than body.
"The key is not what happens in the body part; it's what the brain perceives is going on," says Oxford University neuroscientist Lorimer Moseley, who studies how the brain senses pain. "Any pain that anyone experiences is because the brain has constructed it."
Moseley expresses some skepticism about Tsao's results and says that larger studies are needed. He noted that mirror therapy sometimes loses effectiveness when used over longer periods.
Scientists believe that mirror treatment might help with other pain conditions. Dr. Eric Altschuler, a researcher at the University of Medicine and Dentistry of New Jersey, has used mirror therapy to help stroke patients. It not only reduces pain, but can restore use of paralyzed limbs.
"Mirror therapy is potentially a tremendous paradigm shift in therapy for many different problems," says Altschuler.
Tsao is starting two more mirror studies. One will examine whether mirrors can help arm amputees, while the other will use MRI scans to find out what happens in the brains of phantom limb pain patients before, during and after mirror treatment.
But even this work will likely leave many questions unanswered.
"Anybody who says they understand the brain," says Altschuler, "is a liar, a fool or both."