On suicide clusters: More than contagion

To the Editor

Suicide clusters and ‘contagion’ are sometimes used as equivalent terms (Boyce, 2011), but this is not exact. Sensu stricto, the concept of suicide cluster does not consider the reasons behind the accumulation of cases, but rather just the accumulation of cases (Joiner, 1999). In contrast, suicide ‘contagion’ is just one way of explaining suicide clusters following the ‘infectious disease model’ (Johansson et al., 2006). Different mechanisms may explain why suicide can spread. For instance, direct contact with a suicide victim and word-of-mouth knowledge may explain the spread of suicide in geographically isolated, small areas, thus explaining point clusters (clusters of suicides localized both in space and in time). Indirect transmission via the dissemination of news concerning famous suicides in the mass media may better explain mass clusters (clusters of suicides localized in time but not space). While available literature supports the existence of point clusters, mass clusters might be explained by an ‘ecological fallacy’ (Davidson et al., 1989; Joiner, 1999). Whenever the information about a suicide reaches the recipient, its harmful effect operates by means of social learning, identification and/or imitation (Austin et al., 2011).

There certainly are other explanations apart from contagion that can explain suicide clusters. For instance, Joiner suggested that point-clustered suicides were the result of the coalescence of four factors: first, personal risk factors of suicide (e.g. personality disorders); second, the existence of precipitating life events; third, the lack of social support; and fourth, the tendency of these people to form assortative relationships (Joiner, 1999). Moreover, in a recent study of spatial suicide clusters in Scotland, the authors found three mass suicide clusters within 10,058 small areas for three periods (Exeter and Boyle, 2007). While they could not totally discard contagion, all three clusters were satisfactorily explained by the concentration of socio-economic deprivation.

More frightening is the potentially harmful effect of the mass media in the Internet era. Currently, there are more than 100,000 suicide-related sites on the Internet, and some of them give explicit instructions on how to commit suicide (Gallagher et al., 2003). But it is not just that we have access to more information and more quickly than ever before in human history. Electronic messaging and Internet-based social sites are the way an increasing number of people develop social networks, and this makes a huge difference. Recently, the mother of the ‘Facebook suicide woman’ asked why none of her daughter’s Facebook ‘friends’ did anything when she posted online that she would overdose (Anonymous, 2011). Apart from the ethical and legal consequences that these ‘friends’ may face, they failed to fulfil the universal duty of rescue; thanks to electronic messaging and Internet-based social sites, direct contact at any time is easier than ever (Austin et al., 2011). Given that the role of direct transmission in generating suicide point clusters does not seem to cast any doubt, electronic messaging and Internet-based social sites might lead to an increase in youth suicide (Austin et al., 2011). Thus, we can easily imagine a wave of copycat suicides stemming from a social networking website.