FRAMING HEALTH MATTERS Social Media and Suicide: A Public Health Perspective David D. Luxton, PhD, Jennifer D. June, BA, and Jonathan M. Fairall, BS a systematic Web search of 12 suicide-associ- There is increasing evidence that the Internet and social media can influence suicide-related behavior. Important questions are whether this influence poses ated terms (e.g., suicide, suicide methods, how to a significant risk to the public and how public health approaches might be used kill yourself, and best suicide methods) to simu- to address the issue. To address these questions, we provide an overview of late the results of a typical search conducted ways that social media can influence suicidal behavior, both negatively and by a person seeking information on suicide positively, and we evaluate the evidence of the risk. We also discuss the legal methods. They analyzed the first 10 sites listed complexities of this important topic and propose future directions for research for each search, for a total of 240 different and prevention programs based on a public health perspective. (Am J Public sites. Approximately half were prosuicide Health. 2012;102:S195–S200. doi:10.2105/AJPH.2011.300608) Web sites and sites that provided factual in- formation about suicide. Prosuicide sites and Social media is a relatively new phenomenon recent increase in highly publicized cases of chat rooms that discussed general issues asso- ciated with suicide most often occurred within that has swept the world during the past suicide that involve social media has drawn the first few hits of a search. We should note decade. Social media fuses technology with national attention to this topic.7---9 Researchers that this study primarily focused on prosuicide social interaction via Internet-based applica- are also interested in whether the Internet in search terms and thus likely excluded many tions that allow the creation and exchange of general primarily helps or hinders suicide pre- suicide prevention and support resource sites. user-generated content.1 Social media plat- vention. Attempts to assess the extent of the Recupero et al.11 also conducted a study that forms, such as chat rooms, blogging Web sites Internet’sinfluence on suicide behavior are examined suicide-related sites that can be (e.g., Blogspot), video sites (e.g., YouTube), difficult because of the indirect and complex found using Internet search engines. Of 373 social networking sites (e.g., Facebook, My- association between Internet use and suicide. Web site hits, 31% were suicide neutral, 29% Space, Twitter, Google+), and electronic bulle- The myriad legal complexities involved, as well were antisuicide, and 11% were prosuicide. tin boards or forums, as well as e-mail, text as the important issues of freedom of speech The remaining sites either did not load or messaging, and video chat, have transformed and civil liberties, have also triggered debate. included “suicide” in the title but were not traditional methods of communication by Whether some of social media’sinfluence suicide sites (e.g., sites for movies and novels allowing the instantaneous and interactive on suicide behavior should be considered with “suicide” in their title or music bands sharing of information created and controlled a public health problem and how public health whose names included “suicide”). Together, by individuals, groups, organizations, and gov- approaches might be used to address this these studies have shown that obtaining pro- fl ernments. At the end of 2004, the popular in uence are relevant issues. In this article, we suicide information on the Internet, including social networking site Facebook had close to discuss the role of social media in suicide-related detailed information on suicide methods, is 2 1 million users; by June 2011, that number behavior and frame the issue from a public very easy. had risen to nearly 700 million users world- health perspective. We begin with discussion of A fundamental question is whether an as- 3 2 wide. Facebook has reported that an average the primary ways social media can have a nega- sociation exists between rates of Internet use, of 30 billion pieces of content (e.g., Web links, tive influence on suicide-related behavior and including social media, and population suicide news reports, photo albums, blog postings) we evaluate the evidence of this influence. We rates. Although limited, several preliminary are shared every month via the social media then provide examples of how social media can studies have begun to address this question. site. Social media has become fundamental in be used in the prevention of suicide. We also For example, Shah12 conducted a cross-na- the way many people and organizations com- discuss the legal complexities of this important tional study that examined the association municate and share opinions, ideas, and in- topic and propose future directions for research between general population suicide rates and formation. and prevention programs that are based on the prevalence of Internet users, using data Suicide is a considerable public health a public health perspective. from the World Health Organization’s and the problem; more than 30 000 suicide deaths United Nations Development Program’sWeb in the United States and nearly 1 million SOCIAL MEDIA AND SUICIDE RISK sites. Shah showed that the prevalence of suicide deaths worldwide occur every year.4,5 Internet users was positively correlated with The role that the Internet, particularly social An immense quantity of information on the general population suicide rates. Multiple re- media, might have in suicide-related behavior topic of suicide is available on the Internet and gression analysis indicated that the prevalence is a topic of growing interest and debate.6 The via social media. Biddle et al.10 conducted of Internet use was independently associated Supplement 2, 2012, Vol 102, No. S2 | American Journal of Public Health Luxton et al. | Peer Reviewed | Framing Health Matters | S195 FRAMING HEALTH MATTERS with general population suicide rates in men developed in some way through the use of resulted in the deaths of 208 people.32 This (P = .001) and approached statistical signifi- the Internet has been referred to as a cybersui- suicide outbreak was blamed on the introduc- cance for women (P = .074). Hagihara et al.13 cide pact.23 Traditional suicide pacts have tion of the gas-related method on message conducted a time-series analysis with data typically developed among individuals who boards via the Internet. Family members, from 1987 to 2005 and reported a statistically know each other, such as a couple or friends.23 paramedics, and caregivers were reported to reliable positive correlation between general A primary characteristic that differentiates have been injured or even killed in attempts to population male suicide rates in Japan and the cybersuicide pacts from traditional suicide save suicide victims because of the toxic gas prevalence of households using the Internet pacts is that these pacts are usually formed methods used.32 (P < .05). We should note that the results of among complete strangers.21 The use of Another concern is the media contagion these studies cannot be considered conclusive online chat rooms and virtual bulletin boards effect.33---36 The media’sinfluence on suicidal because of cross-sectional designs and the and forums can provide an unmediated avenue behavior, especially suicide methods used, has possibility of ecological fallacy. to share one’s feelings with other like-minded been well documented,37---43 and social media There are several specific ways that social individuals, which can be easier than talking may possibly increase the risk of the media media can increase risk for prosuicide behav- about such thoughts and feelings in person.24---26 contagion effect, especially among young peo- ior. Cyberbullying and cyber harassment, for The first documented use of the Internet to ple. A recent study by Dunlop et al.24 specifi- example, are serious and prevalent prob- form a suicide pact was reported in Japan in cally examined possible contagion effects on lems.14---19 Cyberbullying typically refers to 2000. It has now become a more common suicidal behavior via the Internet and social when a child or adolescent is intentionally and form of suicide in Japan,27 where the suicide media. Of 719 individuals aged 14 to 24 repeatedly targeted by another child or teen in rate increased from 34 suicides in 2003 to years, 79% reported being exposed to suicide- the form of threats or harassments or humili- 91 suicides in 2005.28 South Korea now related content through family, friends, and ated or embarrassed by means of cellular has one of the world’s highest suicide rates traditional news media such as newspapers, phones or Internet technologies such as e-mail, (24.7/100 000 in 2005), and evidence exists and 59% found such content through Internet texting, social networking sites, or instant that cybersuicide pacts may account for almost sources. Additional analysis revealed no link messaging.17 Cyber harassment and cyber one third of suicides in that country.29 Cur- between social networking sites (e.g., Face- stalking typically refer to these same actions rently, a dearth of published information is book) and suicidal ideation, but it did find when they involve adults. A review of data available regarding the number of cybersuicide a connection between suicidal ideation and collected between 2004 and 2010 via survey pacts in the United States. The problem of suicide-related content found on online forums. studies indicated that lifetime cyberbullying cybersuicide pacts has gained international Video-sharing Web sites have also gained victimization rates ranged from 20.8% to attention, however, and more research is in presence and popularity on the Internet, 40.6% and offending rates ranged from 11.5% needed to understand social media’s impact on especially since the creation of YouTube in to 20.1%.18 the formation of Internet-based suicide pacts.