Lengthen your life, do some exercise.

It is wise to know that elderly people who like to eat out, play cars, go to movies, and take part in other social activities live an average two and a half years longer than more reclusive people, a new study finds.

Health experts typically recommend that the elderly stay active, but the study suggests that simply mixing with other people may offer as great a benefit as regular exercise.

Experts say the work by Harvard University researchers, published in a British Medical Journal, goes further than previous studies because it compares different types of activities and concludes that the benefits of social and productive pursuits are equivalents to and independent of the merits of exercise.

"There is a positive nugget for people who can't exercise, or won't exercise," said Richard Suzman, associate director behavioral and social research at the US National Institute of Aging, who was not connected with the study.

"It shows there are other alternatives that look like they might be just beneficial," Suzman said.

It is widely assumed that staying active is good for people because of the physical component, said Thomas Glass, an assistant professor of health and social behavior at Harvard's School of Public Health, who led the study.

"That social activities involving almost no physical exertion played a measurable role at all in length of life is really quite something," he said. "This is perhaps the strongest circumstantial evidence we've had to date that having a meaningful purpose at the end of life lengthens life. It may not be an old wives' tale after all."

Glass and his colleagues followed 2,761 residents of New Haven, Connecticut (USA), aged 65 an older for 13 years to see how exercise, social and productive activity were related to their chances of dying during the study's duration.

Factors known to contribute to longer life, such as superior health or education, did not influence the result significantly, the researchers said.

Subjects were asked how frequently they were involved in 14 common activities categorized as mainly social, productive or physical.

Physical activities examined were walking fitness exercises and active sports or swimming.

Social activities included church attendance, going to restaurants and sporting events, day trips, playing cards or bingo and participation in social groups.

Pursuits categorized as productive included gardening, preparing meals, shopping, employment and paid or unpaid community work.

The three types of activities to overlapped to some extent and many people were in more than one category. People in each category were compared only to those in other categories.

Among these who participated in social activities, the most activities were 19 percent less likely to die during the study's duration than those least engaged in social activities. Glass said, those most socially active lived about two and a half years longer, he said.

When it came to physical fitness, those who exercised most frequently were 15 percent less likely to die before the study's end than their more sedentary counterparts. They also survived about two and a half years longer.

Those who were most engaged in productive pursuits were 23 percent less likely to die than those least involved in such pursuits. The difference in longevity between those two groups was about four years.

Because the categories were not compared with each other, however, the findings cannot be constructed to mean that being constructive is better than exercising when it comes to extending life, Glass said.

The results remained consistent even when gardening and shopping, which involve some physical activity were excluded from the calculations, the researchers noted.

And among those who exercised the least, the people who were most heavily involved in social and productive pursuits lived longer than those who rarely engaged in such activities.

When each activity was examined individually, doing a lot, as opposed to not much, extended life in almost every case regardless of the activity, Glass said.

"This is pretty impressive because of what they've done to sort out that it wasn't physical activity of physical health that was responsible," said Dr. Harold G. Koenig, as associate professor of psychiatry and medicine at Duke University Medical Center who was not connected with the study.

From course.