His supervisors told him the difference was inconsequential and urged him to move on to more important topics. But over his 40-year career as a brain researcher and university administrator, he never let go of the question.

Decades of research — often conducted in his spare time and with piecemeal funding — led him to a surprising hypothesis: that progesterone, a natural female hormone that protects fetuses in the womb, may actually protect and heal injured brains. His work slowly helped overturn medical orthodoxy that states that brain tissue, once injured, stays that way. Now he and colleagues plan a large-scale human trial over the next several years. While the outcome is far from assured, the effort could produce a new treatment for the estimated 10 million people world-wide who suffer traumatic brain injuries each year.

Dr. Stein’s journey shows just how difficult it is to challenge the medical establishment, which often begrudges ideas outside the mainstream. It also underscores how difficult it is for a lone researcher to persevere without drug-company or other major financial support: For many years, Dr. Stein held administrative jobs and had to moonlight to continue his research. Drug companies tend to focus more on blockbuster drugs they design than on naturally occurring ones with minimal profit potential.

“This is probably the most promising breakthrough in improving outcomes for traumatic brain injury,” says Gregory O’Shanick, national medical director of the Brain Injury Association of America, which advocates for families of people disabled from brain injury. He first heard Dr. Stein present his findings at an international medical meeting in 1992. “It’s absolutely astonishing that it’s taken this long,” he says.

Dr. Stein, an energetic and wise-cracking 68-year-old researcher at Emory University, was the only member of his family to take up a career in science. “If you grow up in a tenement in the Bronx,” Dr. Stein says, “the last thing you want is to work with rats.”

As an undergraduate at Michigan State University in the early 1960s, he helped pay some of his school costs by working in a state mental hospital and the psychiatric ward of a veterans’ hospital.

He grew appalled by the ineffective treatments of the day, such as blasting hopelessly psychotic patients with water from a high-pressure hose to shock them out of their condition and wrapping them in wet bedsheets. Some of these patients had been given lobotomies, an operation that mimics a form of brain injury. The therapies “seemed medieval to me and I was convinced there had to be a better way,” he says.

A few years later, as a doctoral candidate at the University of Oregon in 1964, he stumbled onto what would turn out to be his life’s work. His job was to surgically injure rats’ brains to help determine which parts control memory and consciousness. But he was struck by an anomaly: Why was it that about 30% of the injured rats didn’t act impaired at all?

Medicine’s prevailing view held that injured brain tissue would never heal, so his professors found his data of little interest. “I was told, ‘Don’t waste your time on that. Stick to your topic.’” They explained the differences as “natural variation,” he said. But he was skeptical.

Dr. Stein took the question with him later that year to the Massachusetts Institute of Technology, where he embarked on a prestigious postdoctoral fellowship. His supervisor, Steven Chorover, and others at MIT urged him to stick with the memory work. “The work he wanted to pursue was not something we were working on,” recalls Dr. Chorover.

Dr. Stein still wanted to figure out why those brain-injured rats seemed to recover. But he says he concluded that he wouldn’t win tenure if he pursued the question.

In 1966, with a wife and young child to support, he left MIT to take a job as a psychology professor and director of the brain-research lab at Clark University in Worcester, Mass.

His growing interest in the possibility of recovery from brain injury put him in a tiny minority. Most neurologists at the time still agreed with Nobel laureate Santiago Ramn y Cajal, who wrote in 1913, “In the adult brain, nervous pathways are fixed and immutable. Everything may die, nothing may be regenerated.”

Starting in the late 1960s, Dr. Stein began publishing research that suggested the Nobel winner was wrong.

His lab began methodically studying precisely why some rats stayed smart despite injury. The researchers would place rats in a large vat of water. The rats had to swim to reach a safe platform in a test called a “water maze.” Then the scientists surgically damaged the animals’ brains to study what happened after injury: Would they still be able to maneuver through the maze? The rats that recovered quickly were all female, although not all of the females recovered.

Dr. Stein considered whether the explanation might be something complex, like molecular or genetic differences between males and females. But investigating that would take much more time and money than seeing if a female hormone might yield some clues. First his team evaluated estrogen, but didn’t find a major correlation. Then they tried progesterone — a female hormone that helps protect fetuses from injury during pregnancy.

In these early experiments, Dr. Stein tested female rats to see if they would recover better or worse at different times during their hormonal cycles that resemble human menstruation. Progesterone levels rise and fall during these cycles, and these early studies did indeed show that female rats that were high in progesterone recovered faster.

Dr. Stein thought he had a big part of the answer to the question that had been vexing him for years. The medical establishment, however, largely shrugged off the results.

A naturally occurring hormone like progesterone, some forms of which have been available generically for infertility, is of little interest to drug makers. That’s because the substance probably can’t gain secure patent protection. That shut off a major avenue of potential funding for his research. “Big pharma likes more of an airtight protection,” says Todd Scherer, director of the Office of Technology Transfer at Emory, Dr. Stein’s current academic home.

Caroline Loew, senior vice president for science and regulatory affairs at Pharmaceutical Research and Manufacturers of America, the main drug-industry trade group, says drug companies need patent protection for their research investments. But she suggested the situation could benefit from a government action along the lines of the federal Orphan Drug Act, which increased financial incentives for companies researching rare diseases. A similar idea “may apply to this case,” she says.

Dr. Stein was turned down for half a dozen or more grants from the National Institutes of Health during the 1970s and 1980s. Zaven S. Khachaturian, a leading scientist and former NIH official, says his ideas were “really creative but the NIH system never gave them the good scores they deserved.” At one point, he says, “I just told Don Stein that sometimes it doesn’t pay to keep hitting your head against the wall,” Dr. Khachaturian says.

Dr. Stein recalls feeling “shaken” by the denials, while at the same time growing more determined to prove his case. He kept up a steady drumbeat of research, published in a wide range of journals such as Science, Brain Research, Experimental Neurology and others.

During his time at Clark, Dr. Stein was given to jeans, long hair and shooting his mouth off in faculty meetings or challenging guest speakers, even eminent ones. “Eyebrows would go up whenever Don’s hand would go up,” recalls Julio Ramirez, a former student, now a professor of neuroscience at Davidson College.

But Dr. Stein also was developing into a powerful speaker himself. One speech in Copenhagen in the late 1980s ended up offering a lifeline for his research.

Diane Bistany, then a senior officer at reinsurance company General Re Corp., heard the talk and was impressed. Dr. Stein spoke for 40 minutes without notes, pacing the stage as he made an impassioned argument for the notion that progesterone could be crucial to recovery for thousands of people with traumatic brain injury.

There are five million patients in the U.S. alone disabled from brain injury. “Not only was it his knowledge, but he had a passion for his research and conveyed that this research was going to get somewhere and mean something,” Ms. Bistany says. Gen Re soon contributed grants of $50,000 a year for eight years, keeping his research going.

Around the same time, Dr. Stein grew frustrated at Clark, which was emphasizing clinical psychology and less so the research he preferred. He applied for various positions at other universities, but his ideas, while intriguing enough to win him speaking invitations, still seemed too far out of the mainstream to win him a pure research job.

Finally, in 1988, he landed an administrative position, as dean of the graduate school and associate provost for research at Rutgers University in Newark, N.J. He would work as dean until about 4 p.m., then return to the lab for several hours of work in the evening along with his research team.

Their work progressed step by step. First, did progesterone lower swelling in the brain? If so, did it matter if the animals were females? What did estrogen do, if anything? When did the progesterone have to be given? At what dose? Did progesterone affect animals’ memories?

He found that female rats with much higher, pregnancy-level amounts of progesterone did far better than other rats in following mazes. Even male rats also recovered far better from injury when given the hormone, performing just as well as the highest-performing females. The stuff worked when given up to 24 hours after injury.

Then, in one crucial experiment in 1991, his team tested the amount of cerebral edema, or brain swelling, in brain-injured rats with high progesterone levels versus others with none. Edema is important because of its role in causing brain damage to proliferate.

Robin L. Roof, then a postdoctoral fellow in Dr. Stein’s lab, was waiting for the results while vacationing in a Michigan cottage. When a colleague left a message that there was “no difference,” she thought the experiment had failed. But it turned out the colleague meant that progesterone-protected, injured rats did just as well as rats with no injury at all. Immediately, says Dr. Roof, now a researcher at Pfizer Inc., she realized “this was career-changing research.”

Still, why would a hormone that rises and falls with menstrual cycles and enables fertility, protect the brain? Unlike estrogen, progesterone doesn’t produce visible female sex characteristics. It is present in men at low levels. It rises sharply during pregnancy and helps protect the fetus. Through multiple studies, Dr. Stein and colleagues concluded that it protects the brain in similar fashion.

In 1995, Dr. Stein joined Emory as dean of the graduate school and vice provost. University administrators neglected to provide him lab space, thinking his brain-injury research was little more than a hobby. So for six years his team worked in a moldy double-wide trailer between a parking lot and an industrial-size trash bin. To complete the picture, they installed plastic flamingos in front.

But meanwhile NIH began to back his progesterone research, and the Centers for Disease Control and Prevention also began giving him grants. His studies continued to show positive findings with progesterone, and scientists elsewhere began to confirm them. Often, such confirming studies get done because other scientists start out skeptical.

Dr. Robert Vink, chairman of neurosurgical research at the University of Adelaide in Australia, was among the skeptics, but he grew intrigued by Dr. Stein’s work. Dr. Vink has confirmed that progesterone is beneficial in brain-injured animals in numerous ways, such as lowering brain swelling and cell death, and improving animals’ cognitive abilities. He says Dr. Stein is “persevering. He’s got data, 10 or 12 years now, showing that progesterone in animals is neuroprotective. There’s no doubt about it.”

Still, others remain on the fence. David A. Hovda, a prominent University of California, Los Angeles, neurosurgery professor, says he is still unconvinced that progesterone will prove to be an effective human treatment. But he says he admires Dr. Stein’s work. “Don Stein has a history of stirring the pot,” he says.

In the past decade, Dr. Stein says he and his team have repeatedly run into walls when trying to discern the cellular and molecular mechanism through which progesterone works. But at other times, they did have success on this “cellular pathway” research, and now are finding that there are at least three or four such mechanisms.

By 2000, the findings of Dr. Stein and other brain scientists were swaying the textbooks. One leading neurology tome, “Principles of Neuroscience,” said in its 2000 edition that functions such as thought, language and memory “are all made possible by the serial and parallel interlinkages of several brain regions, each with specific functions. As a result, damage to a single area need not result in the loss of an entire faculty as many earlier neurologists predicted.”

Still, as a practical matter, this was just nifty rat science. Dr. Stein’s hypotheses wouldn’t really matter until they were borne out in humans.

After hearing Dr. Stein lecture a decade ago, Arthur Kellermann, then Emory’s chief of emergency medicine, resolved to get human studies going. He introduced Dr. Stein to David Wright, a young Emory emergency doctor with research ambitions. Drs. Wright and Kellermann wrote an NIH grant application in 1999 for the first phase of human study. Two years later, the federal agency approved a grant of $2.2 million for the first stage of human research.

Over the next three years, the study focused on 100 head-injured patients who had been brought into the emergency room at Grady Memorial Hospital in downtown Atlanta. Some patients received standard treatment to control bleeding and fevers along with state-of-the-art head-injury treatment. Others were also given intravenous progesterone, at triple the highest natural levels at the end of pregnancy.

One Saturday morning in 2005, Dr. Stein was driving north of Atlanta on a shopping trip with his wife when a stern-sounding Dr. Kellermann called him. Dr. Kellermann said he had just learned the study’s findings, adding, “Pull over to the side of the road.”

Dr. Stein froze, fearing that decades of research with animals would prove useless, that progesterone might have turned out to raise the death rate in humans for some unforeseen reason.

His heart was thumping as Dr. Kellermann told him the results: Patients on progesterone had a death rate of just 13% from their head injuries, less than half the 30% death rate of those on standard treatment. And progesterone showed no negative side effects. The 100-subject study was too small to prove that progesterone caused the lowered death rate, but the findings were consistent with animal research. Don Stein was so elated that he had to ask his wife to take over the driving.

In the respected journal “Annals of Emergency Medicine” this past April, Dr. Stein and his researchers summarized the study: “Moderate traumatic brain injury survivors who received progesterone were more likely to have a moderate to good outcome than those randomized to placebo.”

The story is still far from over. Before progesterone can be approved as a treatment, Dr. Stein’s findings must be proved in a larger study of humans. But as he and his team have persisted in research, he has himself become the mainstream of neuroscience. His animal research now has been replicated in dozens of studies at numerous research institutions.

Dr. Stein and his Emory team have applied for NIH funds to do a 1,000-patient study, which will give the definitive word. The NIH already has given an initial $229,000 grant to plan the study, but Emory hasn’t yet officially applied for the full grant. Such a trial could take five years or more. Meanwhile, Emory’s technology-transfer office is “optimistic” about developing and marketing progesterone as a treatment for brain injury, says Dr. Scherer, director of that office.

For Dr. Stein, the results of the clinical study in emergency head-trauma patients were further reason for enthusiasm — “a tremendous culmination,” as he puts it.

“Most bench scientists work for years to discover a truth about nature,” he says. “Very few of us ever get to have a major impact on people’s lives. How can you not be excited?”