alone

RESEARCH ARTICLE
Social Relationships and Mortality Risk: A Meta-analytic Review
Julianne Holt-Lunstad, Timothy B. Smith, J. Bradley Layton
Editors' Summary
Background
Humans arenaturally social. Yet, themodernway of life in industrialized countries isgreatly reducing the quantity and quality of social relationships. Many people in these countries no longerlive inextended families oreven neareach other. Instead, they oftenlive on the otherside of the country orevenacrossthe worldfrom their relatives. Manyalsodelay gettingmarried andhaving children. Likwise,more andmore people of all ages in developed countries are living alone, and loneliness isbecoming increasinglycommon. In the UK,according to arecent survey by the Mental Health Foundation, 10% of people often feel lonely, a thirdhave aclosefriend or relativewho they think isvery lonely, andhalf think that people are getting lonelier ingeneral. Similarly,across the Atlantic, over thepasttwodecades therehas been a three-fold increase in thenumber of Americanswho say theyhave noclose confidants. There isreasonto believe that people arebecomingmore socially isolated.

Why Was This Study Done?
Some experts think that social isolation is bad forhuman health. Theypoint to a 1988review offiveprospective studies (investigations inwhich thecharacteristics of apopulation are determined and then thepopulation is followed to seewhether any of thesecharacteristics are associated withspecific outcomes) that showed that people with fewer social relationships dieearlier on average than those withmore social relationships. But,even though manyprospective studies of mortality (death)have included measures of social relationships since thatfirstreview, theidea that a lack of social relationships is a riskfactor fordeath is still not widelyrecognized by healthorganizations and thepublic. In this study, therefore, the researchers undertake a systematicreview and meta-analysis of the relevantliterature to determinethe extent towhich social relationships influence mortality risk andwhich aspects of social relationships aremostpredictive of mortality. A systematicreview usespredefinedcriteria to identify all the research on a giventopic; a meta-analysis uses statistical methods to combine the results of several studies.

What Did the Researchers Do and Find?
The researchers identified 148prospective studies thatprovideddata onindividuals' mortality as afunction of social relationships and extracted an “effect size”fromeach study. Aneffect size quantifies the size of adifferencebetweentwogroups―here, thedifference inthe likelihood ofdeathbetweengroups thatdiffer in terms of their social relationships. The researchers then used a statistical method called “randomeffectsmodeling” tocalculate the averageeffect size of the studiesexpressed as anodds ratio (OR)―the ratio of the chances of anevent happening in onegroup to the chances of the sameevent happening in thesecondgroup. They report that the average OR was 1.5. That is, people withstronger social relationships had a 50% increased likelihood of survival than those with weaker social relationships. Put another way, an OR of 1.5 means that bythe timehalf of ahypothetical sample of 100 peoplehas died, there will befivemore peoplealive withstronger social relationships than people with weaker social relationships. Importantly, the researchersalso report that social relationships weremorepredictive of the risk ofdeath in studies that consideredcomplex measurements of social integration than in studies that consideredsimpleevaluations such asmarital status.

What Do These Findings Mean?
Thesefindings indicate that the influence of social relationships on the risk ofdeath arecomparable withwell-established riskfactors for mortality such assmoking andalcohol consumption and exceed the influence of other riskfactors