top of page

What evidence do we have that social networks affect health? There are many such examples. When we consider married couples, mortality is lower compared to unmarried spouses, with death of one partner resulting in a higher mortality in the survivor (Christakis 2008a). When one spouse is in hospital, the other has an increased risk of mortality, caring for an ill spouse can lead to higher risk of physical health problems in the remaining spouse, a finding which extends to mental illness in the ill spouse. Depression can also result in the caregiver (Christakis 2008a). Interestingly wife caregivers seem to suffer more than husband caregivers.


What about other non-spousal relationships between two people? parental ill health can lead to ill health in children e.g., maternal depression can lead to the child having a higher risk of behaviour problems, depression, addiction, etc. (Christakis 2008). Relationships between friends can also have health related effects e.g., peer subculture is risk factor in the development of addiction, with drug cessation programmes involving the addicts peer group tending to be more successful (Christakis 2008a).


There also exists evidence that influences can occur via wider networks e.g., sexual behaviour. One study found that 52% of college students involved in relationships were connected within one network with the network being vulnerable to the breaking of social rules. The former finding was replicated in an investigation into HIV in Africa but not the latter effects (Christakis 2008a). Another study looking at the Framingham data, investigated the spread of happiness in a large network, they found people were more likely to be happy in contact with a happy friend, this effect was also found if the friend’s friend was happy and if the friends friends friend was happy, with the effect size becoming smaller with greater distance from the person and the other who was happy. Being in the centre of the network meant people were more likely to be happier, after controlling for age, education and total numbers of connected family and friends and this effect appears not to work the other way around i.e., happiness does not lead to becoming a central member of the network (Fowler 2008). Another study utilizing Facebook found that those who kept their profiles private tended to have friends who kept their profiles private, were female, preferred popular music and was active on Facebook (Lewis 2008).


Finally, it must be borne in mind that social networks can lead to behaviours that can affect health e.g., lack of exercise as determined by the prevailing norm in the particular social network. The effects that social networks can be gender specific e.g., women tend to have more close friendships, mobilise, and provide more social support than men, which can have the interesting effect of exposing women to more social stressors. Paradoxically social support may lead to feeling obliged to the network and in women it can cause mental distress. Socioeconomic circumstances can influence social network effects, in that the poorer tend to be the least likely to receive help (Kawachi 2001).

bottom of page