The research, based largely on data from World War I draft cards, suggests that keeping off excess weight in youth,
farming and fathering a large number of offspring all help men live past a century.

One finding in particular was unexpected, the researchers said.

"We were surprised that having more than three children is beneficial to longevity -- based on previous studies by other authors,
and common sense, quite the opposite could be expected," said study co-author Leonid Gavrilov, who conducted the study with his wife,
Natalia Gavrilova, both of the University of Chicago's Center on Aging.

Gavrilov, a leader in longevity research, was to present the findings Monday at the Gerontological Society of America annual meeting, in San Francisco.

The husband-and-wife team have long mined vital statistics and other data, looking for clues to why some people live into extreme old age.

Just last year, they reported one new finding: Babies born to women under 25 years of age were twice as likely to live to 100 years of age compared to infants born to moms aged 25 or older.

The new research in men was spurred by the fact that a treasure trove of information about 20th-century American males has now been put online: World War I Draft Registration Cards.

From 1917 to 1918, almost all adult males aged 46 or under were required by law to fill out these cards, which asked them to detail a number of physical and social attributes.

In their study, Gavrilov and Gavrilova first used Social Security data to locate 240 men born in 1887 who lived to be at least 100.

In 171 of those cases, the men's physical and social attributes at age 30 were recorded on their WW I draft cards -- giving the researchers a snapshot of their lives at the time.

The Chicago team then compared that data against draft card information for a randomly selected group of American men who were also born in 1887 but who did not reach 100.

Some surprising findings emerged. First of all, a man's chances of reaching 100 rose along with the number of children he had produced by age 30.

Compared to childless men of the same age, a 30-year-old man in 1917 who had one to three children had a 61 percent increased chance of living past a century, the data showed. However, a man's chances for extreme longevity almost tripled if he had fathered four or more children by age 30, the study found.

That's at odds with a prevailing theory in longevity research that holds that "there is a trade-off between the number of children and [parental] longevity," noted Arnold Mitnitski, a longevity researcher and associate professor of medicine, mathematics and statistics at Dalhousie University in Halifax, Canada.

He described the study as "very well-done, very clean."

Theoretically, a household full of young kids should deplete a family's resources and undermine the longevity of parents, Mitnitski said. And yet, young dads with many children lived much longer than other men in this sample.

"This may be due to the support by the children when the person becomes older," Gavrilov speculated. Alternatively, siring many children "could be an indicator of good general health and attractiveness on the marriage market, leading to earlier marriage and hence to more kids by age 30," he said.

In other words, the same robust health that boosted a man's marriage prospects and fertility might also promote long life, Gavrilov reasoned.

Another finding, replicated in prior studies, was that being a farmer (as listed on the Draft Card) more than doubled a man's odds of living into the triple digits.

"The most popular hypothesis [there] is that people in the past had poor sanitation in towns, and hence a high infection load early in life," Gavrilov said. Farms were more isolated, and so farmers were less likely to contract life-limiting illnesses, he reasoned.

Not unexpectedly, overweight -- what the researchers described as a "stout" physique -- reduced a man's likelihood for very long life. In fact, slender or medium-built men were twice as likely to reach the century mark compared to stout types.

But, "surprisingly, there is not much difference between the slender and the medium body build, in terms of survival chance to 100," Gavrilov said. That runs counter to the results of animal studies that have suggested that low-calorie diets, and resultant skinniness, boosts longevity.

Other characteristics -- including marital and immigration status at 30 -- had little or no impact on longevity.

Finally, taller men were only marginally more likely to live to 100 than their shorter peers, the team found. Experts have long linked shorter adult height to the types of childhood infections that might shorten lifespan, so this finding is also a bit of a puzzle, Gavrilov said.

"We need to make a larger study and take a closer look at the links between adult height, childhood infection and longevity," he said.

He stressed that findings for women would no doubt be different, for a variety of reasons. "We need to find [similar] data for women to get the answer," Gavrilov said