The scientists from the University of Colorado at Boulder, found that genetically identical worms responded to stress in greatly different ways.

They say that the worms with more active stress reactions lived much longer than worms with less active stress proteins.

Apparently more active stress responses suggest the animal is coping with the stress.

They suggest their findings will almost certainly apply to humans in some respect.

Researcher Shane Rea tested more than 100,000 nematodes known as Caenorhabditis elegans, a worm favored by scientists because it is easy to work with.

Caenorhabditis elegans despite its tiny size, is genetically complex and has much in common with "higher" animals such as humans.

The team genetically engineered the little transparent worms to carry a jellyfish gene called green fluorescent protein, which glows green under certain light.

This gene was then tagged to a gene called hsp-16.2, a stress protein found in most organisms, that is associated with the health of cells.

The more active the hsp-16.2 gene was, the brighter the worms glowed green and, presumably, the better they coped with stress.

In a typical experiment, the research showed that, under identical conditions, the worms that glowed the brightest green lived about 16 days, compared to three days for those that glowed the most weakly.

Professor Thomas Johnson, who helped lead the study, says they have shown it is possible to predict the life span in an organism, on the first day of adult life, based on how it responds to stress, and this has not been done before.

Johnson says the findings have implications for human longevity and health.

The researchers say they doubt the stress gene itself decides how long an animal lives, but instead reflects some as-yet unknown trait.

According to most scientists, the life span of living creatures is affected by a combination of genetic, environmental and chance factors, and studies done in twins suggests that genes are only about 15 percent to 30 percent responsible for how long an otherwise healthy person will live.

Rea says this study begins to address the question of why genetically identical organisms raised in identical environments still age at different rates.

In future, he says, it might be possible to test people for stress compounds such as hsp-16.2 and predict how long they will live, and even be able to adjust each stress-response system and set them for maximum longevity, which is believed to be about 120 years.

The report is published in this week's issue of the journal Nature Genetics.