With a little help from your friends

Behind the Headlines

Thursday July 29 2010

The study could not explain why relationships may improve health

“Loneliness is a killer” and is “as bad for your health as alcoholism, smoking and overeating”, reported the Daily Mail.

The research behind this story looked at studies on the effects of social relationships on health and disease. It found that having stronger social relationships generally increased a person’s likelihood of survival over an eight-year period by 50% compared to people with fewer social relationships.

The researchers point out that the 148 studies they included in their analysis measured the strength of social relationships in very different ways, and that they did not all take into account other factors that could influence the risk of death.

Although this study found that people generally lived longer if they had larger networks of family and friends, it does not spell doom for those with smaller or less intimate social networks. The study could not show why there was a relationship between social networks and health. The researchers put forward two theories: that social networks may provide a cushion against stress or that they may encourage healthier lifestyles.


Where did the story come from?

The study was carried out by researchers from Brigham Young University, which also funded the research along with TP Industrial Inc. The study was published in the open access peer-reviewed journal PLoS Medicine.

The newspapers reflected the content of the research accurately.


What kind of research was this?

This was a systematic review and meta-analysis that investigated the extent to which social relationships influence the risk of death and which aspects of these relationships contributed to the risk.

The researchers said that there are two possible ways that social relationships may influence health. The first is that friends and family may help people deal with stress and prevent it from building up. The second reason is that they encourage people to have healthy behaviours, as people with these social networks are more likely to conform to social norms of health and self-care.

A systematic review is the best way to find all the information available in a research area. However, the researchers reported that there was a large variation in how social relationships were measured, making it more difficult to directly compare different studies. The researchers focussed on three major components of social relationships that were consistently evaluated to see how each of these components contributed to health:

  • the degree of integration in social networks
  • the social interactions that are intended to be supportive (received social support)
  • the beliefs and perceptions of support availability held by the individual (perceived support)


What did the research involve?

The researchers searched for studies published between 1990 and 2007 using medical and scientific journal databases. They used the search terms “mortality”, “death”, “decease(d)”, “died”, “dead” and “remain(ed) alive” in addition to search words linked to social isolation, including “social”, “interpersonal”, “support”, “network”, “integration”, “participation”, “cohesion”, “relationship”, “capital” and “isolation”. If they found a review article using this search strategy, they looked at the references included in its reference list to make sure they had not omitted any relevant studies.

The researchers included studies that provided quantitative data on the association between death and an individual’s social relationships. As they were interested in the effect of social relationships on disease, they excluded studies where people had died as a result of suicide or injury. They also excluded studies where the measurement of social support was a treatment (for example, a support group for bereavement counselling) and studies that looked at the effects of having a pet or having faith in a god. Studies where individuals gave support (such as caregivers) rather than received support were also excluded.

Each paper was assessed and rated by four researchers. Data within the studies were often reported as odds ratios or hazard ratios, which show the risk of an outcome in people with a particular risk factor relative to those without the risk factor. The researchers used various statistical techniques to combine results between studies to give overall odds ratios for the risk of disease or death.


What were the basic results?

The researchers identified 148 studies that met their inclusion criteria and had data that could be pooled. The studies included data from over 300,000 individuals from North America (51%), Europe (37%), Asia (11%) and Australia (1%). The average age of the participants was 64 years. The participants were followed for an average of seven-and-a-half-years, and an average of 29% died during follow-up.

The researchers found that when they combined data from studies that had taken any type of measure of social interaction (for example, the number of people in a social network, amount of received support and amount of perceived support), the odds of survival was increased by 50% in those with stronger social interactions compared to weaker social interactions (odds ratio [OR] 1.50, 95% confidence interval [CI] 1.42 to 1.59).

In the 63 studies that only looked at the structural measures of social relationships (for example, how many people are in a social network), the odds of survival increased by 57% for individuals with stronger social networks compared to weaker ones (OR 1.57, 95% CI 1.46 to 1.70).

Twenty-four studies looked at functional measures of social relationships only, such as received or perceived social support. They found that stronger functional social relationships were associated with a 46% increase in odds of survival (OR 1.46, 95% CI 1.28 to 1.66).


How did the researchers interpret the results?

The researchers say that individuals with adequate social relationships have a 50% greater likelihood of survival (over 7.5 years) than those with poor or insufficient social relationships. They report that the size of this effect is similar to or larger than that seen with smoking, obesity and physical activity.

They suggest that their meta-analysis should promote further research into treatments that explicitly account for social relationship factors across levels of healthcare such as prevention, evaluation, treatment, compliance and rehabilitation. They also suggest that hospitals and clinics could involve patient support networks in implementing and monitoring treatment regimes.



The study showed an association between the strength of social relationships and lower mortality rates. However, as the researchers highlighted, many different measures of the strength of social interactions are used, and this study does not classify what constitutes a healthy social life. Owing to the high variability of the studies included in this meta-analysis, the researchers looked at social relationships in very broad terms. There are some limitations to this research, some of which the authors highlight:

  • Behavioural risk factors for disease, such as smoking and drinking alcohol, were adjusted for in some but not all of the studies included in this meta-analysis. Some of the effect on mortality may, therefore, be due to these and other factors that were not adjusted for in the analyses, rather than social factors.
  • The participants included in the individual studies varied greatly. For example, some studies included people who had diseases at the start of the study and looked at how their social relationships affected the outcome of their disease, while other studies looked at people who were healthy at the start of their follow-up.
  • Most measures of social relationships did not take into account the quality of the relationship and assumed that all relationships were positive.
  • The researchers used various standard methods to try to assess whether publication bias could have contributed to their results. Publication bias means that scientific research that has negative results (finds no association) is less likely to be published than studies that have positive results. Although the researchers say that, based on their calculations, this is unlikely to have had an effect, publication bias is an important factor in this type of meta-analysis and may still have affected the results.
  • It is difficult to directly compare the effects of loneliness seen in this meta-analysis with results seen in other studies of smoking, physical inactivity or other unhealthy behaviours, as the people participating in the studies may not be comparable.

The link between social relationships and disease is likely to be complex, as people who are seriously ill may have reduced capacity for social interaction and this could affect their social networks. It may also represent something that happens naturally with aging: that as people get older, their social networks get smaller. This study highlights that social relationships may play a role in health and disease, but further work is needed to assess what type of social interactions are helpful and the biology underlying this.

Analysis by Bazian

Edited by NHS Choices

Links to the headlines

Can your friendships save your life? The Guardian, July 28 2010

Popular people live longer. The Daily Telegraph, July 28 2010

Having pals 'helps you live longer'. Daily Mirror, July 28 2010

Loneliness is a killer: It's as bad for your health as alcoholism, smoking and over-eating, say scientists. Daily Mail, July 28 2010

Links to the science

Holt-Lunstad J, Smith TB, Bradley Layton J. Social Relationships and Mortality Risk: A Meta-analytic Review. PLoS Medicine 7(7)


How helpful is this page?

Average rating

Based on 0 ratings

All ratings

Add your rating