Scientists say it may be possible to predict longevity based on the levels of certain proteins and hormones in the body. Their research involved an assessment of body systems that are not usually suspected of causing premature death.

A routine part of a doctor's examination usually involves a blood pressure check and blood test for cholesterol. If either of those numbers is too high, chances are the doctor will prescribe medication.

But, scientists are finding out there are a host of other body systems that do not get measured that are equally important to good health. These include proteins involved in glucose metabolism and stress hormones.

In a study published in the journal Proceedings of the National Academy of Sciences, researchers studied 13 so-called biomarkers, which included not only cholesterol and blood pressure, but also immune system markers and stress hormones.

They wanted to find out which combinations of biomarkers increased the risk of death.

In a study that began in 1988, researchers measured levels of these markers in a group of 1,200 healthy men and women in their 70s and followed them for 12 years.

Lead investigator Tara Gruenewald of the Division of Geriatrics at the University of California in Los Angeles says that in men, the presence of two biomarkers most strongly predicted early death: elevations of a stress hormone called cortisol and immune system proteins that measure inflammation.

She says inflammation is a response by the body to infection and injury.

"And we also know that levels of inflammation increase with other types of experiences such as exposure to stress and poor sleep patterns, as well as markers of hormones, hormone stress markers," said Tara Gruenewald.

In women, the study found that elevated blood pressure in combination with any abnormal biomarker significantly increased the risk of death.

Gruenewald says it is hard to say whether measuring the same 13 biomarker combinations in people in their 30's or 40's could predict whether they are likely to die prematurely.

But she says investigators selected the markers because they are indicators of the major regulatory systems in the body.

"And so I would suspect that if we were seeing biomarkers at the same levels that we saw in these old adults when folks were younger decades earlier that it would be my hypothesis that the same biomarkers at high levels would most likely predict the greater likelihood of death and disease in later adulthood," she said.

But Gruenewald says biomarkers are not fixed and can be changed early enough in life, for example through exercise, to improve the chances of a long life.