A new study concludes that men who experience persistently moderate or high levels of stressful life events over a number of years have a 50 percent higher mortality rate.

In general, the researchers found only a few protective factors against these higher levels of stress - people who self-reported that they had good health tended to live longer and married men also fared better. Moderate drinkers also lived longer than non-drinkers.

"Being a teetotaler and a smoker were risk factors for mortality," said Carolyn Aldwin, lead author of the study and a professor of human development and family sciences at Oregon State University. "So perhaps trying to keep your major stress events to a minimum, being married and having a glass of wine every night is the secret to a long life."

This is the first study to show a direct link between stress trajectories and mortality in an aging population. Unlike previous studies that were conducted in a relatively short term with smaller sample sizes, this study was modified to document major stressors - such as death of a spouse or a putting a parent into a retirement home - that specifically affect middle-aged and older people.

"Most studies look at typical stress events that are geared at younger people, such as graduation, losing a job, having your first child," Aldwin said. "I modified the stress measure to reflect the kinds of stress that we know impacts us more as we age, and even we were surprised at how strong the correlation between stress trajectories and mortality was."

Aldwin said that previous studies examined stress only at one time point, while this study documented patterns of stress over a number of years.

The study, out now in the Journal of Aging Research, used longitudinal data surveying almost 1,000 middle-class and working-class men for an 18-year period, from 1985 to 2003. All the men in the study were picked because they had good health when they first signed up to be part of the Boston VA Normative Aging Study in the 1960s.

Those in the low-stress group experienced an average of two or fewer major life events in a year, compared with an average of three for the moderate group and up to six for the high stress group. One of the study's most surprising findings was that the mortality risk was similar for the moderate versus high stress group.

"It seems there is a threshold and perhaps with anything more than two major life events a year and people just max out," Aldwin said. "We were surprised the effect was not linear and that the moderate group had a similar risk of death to the high-risk group."

While this study looked specifically at major life events and stress trajectories, Aldwin said the research group will next explore chronic daily stress as well as coping strategies.

"People are hardy, and they can deal with a few major stress events each year," Aldwin said. "But our research suggests that long-term, even moderate stress can have lethal effects."