Cherish your old friendships, or learn to make new ones - they could well increase your lifespan.
New research by the Centre for Ageing Studies at Flinders University suggests that a network of good friends is more helpful than close family ties in prolonging life among older people.
The report, published recently in the Journal of Epidemiology and Community Health, draws on data from the Australian Longitudinal Study of Ageing (ALSA), which began in Adelaide in 1992. It was written by PhD student Lynn Giles with Professor Gary Andrews and Professor Mary Luszcz of the Centre for Ageing Studies, and Mr Gary Glonek, senior lecturer in Statistics at Adelaide University.
The study used a series of interviews with a large cohort of people aged 70 and upwards to assess how economic, social, behavioural and environmental factors affected their health and wellbeing.
In total, almost 1500 people were asked how much personal and phone contact they had with their various social networks, including children, relatives, friends, and confidants. The group was then monitored annually for four years and then approximately every three to four years for a decade.
Ms Giles said that evidence suggesting that social networks contributed to survival was not new.
"What hasn't been done before is to break down which social networks might be most beneficial," she said.
"It looks as if friends are the most important in terms of survival."
The research team also considered the impact of factors likely to influence survival rates, such as socio-economic status, health, and lifestyle.
Close contact with children and relatives had little impact on survival rates over the 10 years, while a strong network of friends and confidants significantly improved the chances of survival over that period. Those with extensive networks outlived those with the least friends by 22 per cent.
Ms Giles said the study did not aim to belittle blood ties.
"We are not writing off families or saying that children are unimportant," Ms Giles said.
Indeed, a separate study showed that family ties and contacts were the most important factor in staving off disability among the elderly.
In terms of staying alive, however, having friends came up trumps.
The beneficial effects on survival persisted across the decade, irrespective of other profound changes in individuals' lives, including the death of a spouse or close family members, and the relocation of friends to other parts of the country.
The authors speculate that friends may influence health behaviours, such as smoking and drinking, or seeking medical help for troubling symptoms. Friends may also have important effects on mood, self-esteem, and coping mechanisms in times of difficulty.
The duration of the friendships was not revealed by the study, and Ms Giles suspects it may not be a critical factor.
"We don't know how long these people have had their networks of friends for - they may be long-standing friendships or they may be comparatively recent," she said.
"The central message is that maintaining a sense of social embeddedness through friends and family appears pretty important for survival, and it seems that non-kin relationships are particularly important."