OIMYAKON, Russia - A few years ago, a retired math teacher named Tamara I. Vasileva began poring through diaries and records of births and deaths in this Siberian village of 950 people.
As the community's unofficial archivist, she studied the records of about 20 of the town's oldest families, dating back to the 1920s, and she noted something odd. Until roughly the 1960s, the records documented the lives of people who lived well into their 70s and 80s - one or two even into their hundreds.
But in the 1960s, as the Soviet Union's economy began to stagnate, the number of people in their 70s, 80s or 90s began to shrink. Among the small sample Vasileva studied, people were dying about five years earlier than they had a few decades before.
The names and figures Vasileva wrote on graph paper don't constitute a scientific study. But she had detected a tiny, local tremor of a far larger social and economic phenomenon, one that has brought premature death and grief to millions of Russians, and threatens this nation's future.
Since the collapse of the Soviet Union in 1991, life expectancy here has plummeted. No one really knows why. It is a phenomenon unprecedented outside of war, plague or famine. It also is one of the greatest medical mysteries of the early 21st century.
And it has broad implications for global health.
Russian men now live shorter lives than those in any other industrial country. Male life expectancy today is 59 years and 11 months. That is about 15 years less than men in Europe and the United States. It is also six years less than the average Russian lived in 1965.
Russian women on average live to be 73, about seven years less than in the West - and a year less than their grandmothers could expect to live in the 1960s.
A Russian man's risk of dying of stroke is seven times higher than for men in the United States. He is 12 times more likely than an American to die of alcohol poisoning. He is seven times more likely to suffer a fatal stomach cancer, and 78 times more likely to contract tuberculosis. Only the citizens of Colombia and El Salvador are more likely to die a violent death.
As well as experiencing an epidemic of premature deaths, Russia is suffering from a declining number of births. Russia's birth rate ranked 188th out of 193 countries surveyed in a 2002 United Nations report. The number of Russian women in their peak years of fertility - ages 20 to 29 - will fall from about 13 million to fewer than 8 million by 2012. After that, experts say, Russia's population will plummet.
According to the State Statistics Committee, Russia's population has slipped since 1989 from 148 million to 143.5 million, despite an influx of immigrants from Central Asia and parts of Eastern Europe. In January, Russia's National Security Council warned that by the year 2050 the nation's population could slip below 100 million.
The consequences of population loss are already visible across Russia, as forests and steppe reclaim scores of deserted villages. In years to come, Russia's factories and research institutes will find it difficult to recruit workers. School enrollments will shrink.
In a world in which national wealth depends on people with specialized skills, imagine a country with declining numbers of engineers, doctors and scientists.
Imagine the mindset of military officers with few soldiers but many nuclear weapons.
Imagine a nation that covers one-seventh of the earth's land surface but with a population smaller than that of Uganda or Vietnam.
At the end of the 1960s, mankind stood at the threshold of a new era in public health. Thanks to new vaccines, antibiotics and other medical advances, life expectancy was advancing around the world. The only significant exceptions were the Soviet Union and its satellites in Eastern Europe.
After two decades of growth, life expectancy here stalled and began a gradual retreat.
At the time, only a few bureaucrats and experts recognized the problem. Soviet scientists could publish work on the issue only in classified journals with strictly controlled circulation.
But Western researchers, relying on the few public studies, began to notice the trend. In a celebrated paper in the 1980s, an American scholar compared figures for births and subsequent first-grade school enrollments to estimate infant mortality. She discovered it was much higher in the Soviet Union than in the West.
Then came the catastrophe. As the Soviet Union broke apart in the late 1980s and early 1990s, life expectancy in Russia plummeted by six years.
"There was an explosion of mortality from 1992 to 1994," said Dr. Vladimir M. Shkolnikov, a former Soviet epidemiologist who now leads the Max Planck Institute for Demographic Research in Rostok, Germany. "We reached the absolute maximum."
The fall in life expectancy occurred throughout the former Soviet Union but was steepest by far in Russia and Ukraine. It wasn't just the shock of moving from a communist to a capitalist system that seemed to matter, but the rockiness of the transition.
In Poland, the Czech Republic and Slovenia, the political changes were generally welcomed. In those countries, there was a steady increase in life expectancy, even as incomes declined and unemployment rose.
At first, physicians and researchers distrusted the data about Russia.
"For a while, everybody believed this must be some kind of statistical error," said Martin Bobak, a senior lecturer at the International Center for Health and Society at University College, London.
"We thought maybe the Soviets did not report their deaths properly. So a French demographer got access to original files - handwritten papers from different regions - and reconstructed the data back to 1960. He found that the data were fine."
Russia was caught in an epidemic of early deaths from which it has not yet escaped.
Over the years, researchers have advanced several explanations for Russia's health catastrophe:
Russia remains addicted to cigarettes. About 60 percent of Russian men smoke, compared with 25 percent of American men, and Russian cigarettes are not subject to limits on tar or nicotine. But lung cancer rates here are only slightly higher than those in the United States. And lung cancer typically takes decades to develop. No increase in smoking, no matter how dramatic, could explain the sudden spike in deaths after 1990.
Some Russian researchers blamed rising death rates on radioactive and toxic wastes carelessly dumped during the Soviet era. But the increase in deaths occurred when factories were shutting down and toxic discharges were decreasing.
"Even if you have very high levels of pollution, it would increase mortality only a few percent," said Bobak, at the International Center for Health and Society. "It cannot account for millions of people dying."
As state support dwindled in the 1990s, health care deteriorated. Hospitals lacked access to modern pharmaceuticals and advanced equipment. Poorly paid doctors and nurses demanded bribes for all but the most basic treatment.
Demographers point out that the health care system, although strained, did not collapse. Consider, for example, two key measures of a country's health: infant mortality rates and maternal death rates.
Infant mortality in Russia - about 17 for every 1,000 live births in 1999, versus seven in the United States - rose briefly from 1990 to 1993. But the rate then began a decade of decline. Maternal deaths during childbirth, meanwhile, fell from 49 for every 100,000 births in 1989 to 39.7 in 2000, according to Russian Health Ministry statistics.
"Other countries that have gone through an economic crisis, you typically see an increase in mortality among the elderly and infants" because of declines in public health care, said Elizabeth Brainerd, an economist teaching at Williams College in Amherst, Mass. "It's nothing like you see in Russia, where it's prolonged and primarily among middle-aged men."
Vodka - alcohol distilled from fermented grain or potatoes ---- flows through Russian history and culture. Vodka has been used as currency when the ruble has failed, as medicine when hospital shelves are empty, as succor during the long, dark winters.
In May 1985, Soviet leader Mikhail S. Gorbachev campaigned to end the economic toll of drunk, sick or injured workers by limiting alcohol production and sales. Russians responded by drinking brake fluid, kerosene and tractor oil. After two years, the temperance effort collapsed in a flood of bootleg liquor.
But while it lasted, the campaign had a dramatic effect. In just the first 12 months of Gorbachev's initiative, life expectancy in the Soviet Union rose by three full years. "It was a natural experiment," said Shkolnikov of the Max Planck Institute. "By separating people from alcohol, it showed how important a factor alcohol was."
After the breakup of the Soviet Union, controls on alcohol vanished. Vodka became cheaper than bottled water. And everyone, it seemed, wanted a drink.
Between 1989 and 2000, consumption in Russia increased 45 percent, according to the World Health Organization. Over about the same period, the death rate from alcohol-related causes among men rose 245 percent. Among women, it rose more than 300 percent.
But alcohol, it seemed, couldn't explain everything.
The sudden decline in life expectancy here is chiefly due to a sharp increase in cardiovascular illness. And alcohol consumption - though linked to many illnesses - is not thought to contribute significantly to heart disease. Moderate drinking has been credited in studies with preventing heart attacks.
When challenged, Shkolnikov re-examined the studies that seemed to highlight alcohol as the explanation for early deaths. Perhaps, he said, it was the way Russians drank, and not just the amount, that mattered.
Adults here, he pointed out, don't just consume about 60 percent more alcohol per capita than western Europeans. By tradition, vodka here is drunk by the glassful in a single swallow. And few here drink just one glass at a sitting.
A late 1980s study among 25- to 64-year-old men in the Siberian city of Novosibirsk showed they drank on average about once a week.
But each time they drank, 70 percent of them consumed the equivalent of a full liter bottle of vodka each. (A liter is slightly more than a quart.)
Studies show that at least one-fifth of Russians are binge drinkers. And binge drinking is known, in some cases, to cause heart attacks and life-threatening arrhythmias.
Still, the alcohol skeptics weren't satisfied.
Bobak's group in London studied Shkolnikov's findings.
Yes, the researchers agreed, bingeing occasionally leads to death. But there are too few binge drinkers, Bobak said, and binge drinking itself doesn't kill people frequently enough to have had much of an impact.
Another cause was needed to explain the scale of the disaster.
Loss of status
When Brainerd, the Williams College economist, went to Moscow in the early 1990s as a graduate student, she found a nation in turmoil.
During the Communist era, Soviet citizens lived in a society that guaranteed them an income, no matter how small; a place to live, no matter how crowded; a job, no matter how poorly paid. Now all that was gone.
Everyone suffered, of course. But she noticed that middle-aged men had the most trouble adjusting.
"You saw a lot of despair among this group - despair about not having a future, about not being able to provide for their families," Brainerd said.
And Russia's family men weren't just depressed. They were dying at a terrifying rate.
From the perspective of public health, it just didn't make sense. Normally, men in their 40s and 50s are among the healthiest in any society. "Just looking at the numbers, you could tell that something unprecedented was going on," she said.
Brainerd, among others, wondered whether psychological factors might play a role. Researchers thought the crisis might partly be explained by the work of Dr. Michael Marmot, a British physician and public health researcher.
In 1967, Marmot began what became known as the Whitehall studies, named for the central London neighborhood synonymous with the British civil service. The first study analyzed the health data of 18,000 civil servants, in an effort to measure the impact of job stress on heart disease.
As a mental state, stress is difficult to define. Physiologically, though, it is well understood. Under psychological stress, a person's levels of adrenaline, cortisol and other steroids rise. Blood pressure increases. Saliva production drops. The immune and reproductive systems are depressed. Together, these physiological changes constitute the classic "fight-or-flight" response.
Over short periods, this state of heightened physical and mental readiness improves a person's chances of survival. Prolonged, stress can weaken the heart, blood vessels, liver and other organs. That can leave people more vulnerable to infections and other diseases.
Marmot studied civil servants because their status was easy to determine: they are strictly ranked by pay grade and job responsibility. He assumed that the civil servants with more responsibility would suffer more stress, and more heart disease and strokes.
But when he published his findings in 1978, Marmot reported that men at the top lived far longer than their subordinates. Nor was the effect limited to cardiovascular illness. For low-ranking civil servants, the risk of death was four times greater from all major causes of death: gastrointestinal disease, renal illness, cancers, infections, homicides, suicides and accidents.
It wasn't just that higher-ranking officials lived longer than their lowliest subordinates - an effect that might be explained by differences in diet, education and other factors. Researchers found that a bureaucrat's risk of dying rose in small but measurable increments with each step down the pay grade ladder.
People with nearly identical backgrounds had significant differences in health.
The research, Marmot said, did not suggest that lifestyle was independent of health. Civil servants who smoked were more likely to get lung cancer than those who didn't. But lower-ranking bureaucrats who smoked were far more likely to get lung cancer than their bosses who smoked. An overweight boss, the study showed, had a much lower risk of heart disease than his overweight subordinates.
Status, in other words, predicted health. The lower a person's status, Marmot decided, the more lack of control he feels over his own life. The more frustration, the more stress. The more stress, the earlier he is likely to die.
The Whitehall findings were later replicated among civil servants around the world. Similar studies were done among primates, and among a variety of social groups.
In every case, status seemed to play a role. Donald A. Redelmeier and Sheldon M. Singh, two Toronto researchers, studied the medical histories of 1,649 Hollywood actors and published their findings in the Annals of Internal Medicine in 2001.
Oscar winners, they found, lived four years longer than their co-stars and actors who were nominated but did not win.
And what if an entire social system collapsed - Marmot and his followers asked - leading to a catastrophic loss of status? Could Marmot's findings be applied to a whole society?
Perhaps Russia was, in part, dying of despair.
A couple of years ago, Brainerd and her colleague Bruce Cutler, a Harvard economist, set out to settle the question of what was killing Russia.
They reviewed more than 50 studies. They also pored over data from the World Health Organization and the Russian Longitudinal Monitoring Survey, or RLMS, an effort launched in 1994 that has gathered health information from 4,000 families.
Like earlier researchers, Brainerd and Cutler found that the data ruled out most of the traditional explanations: poor health care, inadequate diet, pollution and poverty.
Their analysis, published by the National Bureau of Economic Research in November, partly vindicated the work of Shkolnikov and his team: By Brainerd and Cutler's calculations, about a quarter of the increase in mortality in Russia in the early 1990s was due to alcohol consumption.
The work also partly supported Bobak and Marmot's hypothesis about status loss. Brainerd and her associates concluded that what they termed "psycho-social stress" could explain about a quarter of Russia's excessive deaths in the early 1990s.
But the core of the mystery remains. Half of the increase in death rates remains unexplained. Russia's population is still shrinking, one premature death at a time.