Suicide is the second leading cause of death in young people aged 10–24 years, globally . However, it is challenging to detect and intervene early, as many of those who go on to die by suicide will not have interacted with mental health services . Novel mechanisms that underpin suicidal behaviours are required.
The internet is now ubiquitous globally and is used for educational, recreational and social purposes. Social network sites (SNSs) and social media are web-based services entitling users to construct a personal profile, support user-generated content, connect with other users and support ways for members to collaborate [3,4]. However, as technology has progressed, the boundaries between internet, social networking sites (SNSs), social media, online gaming and digital technology have become increasingly blurred.
The links between self-harm, suicidal ideation and later suicide attempts are well established [5,6], but nomenclature in this field varies. In the United States, suicide attempts, nonsuicidal self-injury (NSSI) and self-harm (with undetermined intent) are described separately, in contrast to Europe where ‘self-harm’ is used more broadly [7–9]. We will focus on suicide attempts, one of the strongest known predictors of completed suicide, and will specifically focus on the evidence pertaining to adolescents aged 18 years and under, to be relevant to child and adolescent mental health service provision in most countries.
Seven systematic reviews published to date have found an association between increased screen time and worse mental health in young people [10▪▪], and the association between cyberbullying and suicidal behaviour is described in a recent meta-analysis [11▪▪]. However, the existing evidence suggests that the relationship between internet use and self-harm and suicidal behaviour is mixed [12▪▪] with potential for harm, but also scope to foster a sense of community, offering isolated young people supportive contacts . Guidance is required, but there remains a paucity of evidence to inform this. There are suggestions for strategies, such as: a ‘Family Media Use Plan’ , ‘Digital Literacy’ being taught in schools [15,16], increased support for parents  and the need for funding of new research, to ensure future guidance is evidence-based [17▪].
No previous review has specifically investigated the association between social media/internet use and completed or attempted suicide in adolescents (under 19 years). This review aims to fill that knowledge gap, as well as outline some of the recent developments in this field of enquiry.
This review is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and the protocol is registered on PROSPERO (ID CRD42018115259). English language publications, published from database inception up to 25 January 2019 were included. Studies with a majority of participants under 19 years old at the point of enrolment in the study were included.
We were interested in the following exposures: patterns and nature of social media and internet use, and content viewed or shared online. Due to the high prevalence of social media/internet use in young people, studies comparing different types, or levels, of internet or social media use were included. Only peer reviewed, observational studies, where the full text was available were included. Qualitative studies, case reports or comments/editorials were excluded.
The outcome of interest was suicide or suicide attempts in children and adolescents. Due to the practical challenges of determining intent from the wider ‘self-harm’ literature, only studies specifically stating ‘suicide attempts’ as their outcome were included. NSSI and self-harm, without suicidal intent or where intent was not specified, were not included.
The following databases were searched with a predefined search strategy: Medline, PsycINFO, EMBASE, HMIC and CINAHL. The search strategies were developed and adapted to include both subject headings (i.e. Child, Adolescent, Social Media, Internet, Suicide, Suicide Attempt) and keywords, that is, ‘Facebook’, ‘Instagram’, ‘hashtag*’, ‘suicid*’ relevant to each respective database. The full search strategies for each database are available in the online supplement. In addition, backward and forward citation searching was conducted and the reference lists of existing systematic reviews on similar topics were reviewed to identify any further relevant articles.
Data extraction and quality assessment
Abstract screening, data extraction and quality assessment of articles was completed by two independent reviewers (R.S. and S.E.). Full text articles were obtained and screened by the two reviewers where suitability could not be determined based on the title and abstract. Data was extracted using a predesigned data extraction form. Risk of bias was assessed using an adapted Newcastle-Ottawa scale (see supplementary material). Scores of 0–4 were considered as high quality (low risk of bias), 5–7 as moderate quality and 8–10 as low quality (high risk of bias).
One thousand six hundred and ninety references were identified through database searching and 1179 remained after removing duplicates. One thousand one hundred and twenty were excluded on title and abstract screen. Fifty-nine full texts were reviewed of which nine were eligible for inclusion in the review. Figure 1 (PRISMA diagram) shows further details, including reason for exclusion at the full text screening stage.
Of the nine identified studies, all were cross-sectional, published between 2012 and 2018. The age range of participants studied was 11–18 years. In five of the studies, the female participant rate was over 50%, most studies had an approximately equal proportion of male and female participants. Studies were conducted in a range of countries with sample sizes from n = 111 [18▪] to n = 221 265 . There were a total of 346 416 participants across the nine studies. Six articles were judged to be of high quality, two of medium quality and one low quality. The two medium-quality studies lacked information on nonrespondents and appropriate statistical tests. Ascertainment of the exposures and outcomes within each article showed low risk of bias, but there was significant heterogeneity between studies.
One study investigated psychiatric outpatients with Major Depressive Disorders , another a psychiatric inpatient setting [18▪], the rest were community-based or school-based samples. Two studies examined internet use as the outcome [21,22], the remainder analysed suicide attempts as the dependent outcome. Control groups in those with social media or internet use as the exposure were either lower level internet users , nonusers , occasional users  or those not meeting the defined threshold for pathological [18▪], problematic [18▪,22,25▪▪] or addicted  internet behaviours.
There was significant heterogeneity in both the method of assessing social media/internet use and the timescale of reported suicide attempts. Only one study investigated SNSs. Problematic Internet Use was the exposure of interest, classified as a score on the Young Internet Addiction Test (YIAT) of over 50, in two studies [20,25▪▪]. In another, Pathological Internet Use, defined as at least 5 on the Young's Diagnostic Questionnaire (YDQ) was the outcome . Lin et al. used the addiction cut-off score of 64 on the Chen Internet Addiction scale to define their exposure. The remaining studies were interested in hours of noneducational internet use  and cut-offs on the Korean Internet Addiction Self-Assessment Tool .
There was an association between increased social media/internet use and suicide attempts in seven studies when controlling for at least age and sex [adjusted odd ratios (ORs) ranged from 1.03 to 5.10] [23,25▪▪]. Alpaslan et al. found no relationship between suicide attempts and YIAT score while controlling for age and sex in Major Depressive Disorder cases and Fuchs et al.[18▪] did not report any multivariable results. The association between Social Networking Sites (SNS) and suicide attempts found by Sampassa-Kanyinga (OR 5.10, 95% CI 1.45–17.88) was found to be indirect and explained by cyberbullying victimization. Guo et al.[25▪▪] found that on path analysis, sleep disturbance was found to be a mediator in the link between Problematic Internet Use and suicide attempts. Cyberbullying victimization and sleep disturbance were not controlled for in other studies.
Kim  found both male and female ‘heavy internet users’ showed an increased attempted suicide rate compared with ‘no internet users’, but ‘occasional internet users’ had the lowest rate of attempted suicide. Similarly, Lee et al. found that the Korean Internet Addiction Self-Assessment Tool (KS scale) determined a ‘high risk’ of internet addiction user group (OR 1.91, 95% CI 1.71--2.14), and the ‘nonuser’ group (OR 1.33, 95% CI 1.25--1.42) were more likely to attempt suicide than the ‘potential-risk’ group (OR 1.20, 95% CI 1.04--1.38). These results indicate that some degree of internet use may be beneficial.
Unfortunately, because of the heterogeneity in exposures and outcomes, results could not be combined in meta-analysis to provide a meaningful result; however, all nine studies are summarized in Table 1.
The potential for harm related to the internet and social media is widely discussed in both the lay and research literature, but it remains difficult to determine, which aspects may be harmful and conversely, which may be supportive. Gaming Disorder in the 11th edition of the International Classification of Diseases and Internet Gaming Disorder in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM), are potential diagnoses requiring further research , and the importance of the ‘internet’ aspect is not clear. Internet Addiction, also referred to as problematic or pathological internet use, dominated the findings of this systematic review, but the addiction model of internet use does not account for more nuanced behavioural mechanisms, which we shall now discuss.
Mechanisms of harm
There is growing evidence of specific mechanisms by which social media/internet use can be harmful, especially in relation to young people. In a recent meta-analysis, cybervictimization was associated with suicide attempts (OR 2.57, 95% CI 1.69–3.90) [11▪▪]. Furthermore, a link was reported between cyberbullying, suicidal ideation and self-harm, highlighting it as an extremely important area of focus when considering social media/internet risks.
Exposure to social media/internet has the potential to both suggest and reinforce negative thoughts and behaviours. There is an association between comments on Instagram with increasing severity of self-injury, suggesting social media may act to reinforce harmful behaviours . Themes such as self-loathing, loneliness and feeling unloved were found in content analysis of 3360 randomly selected Tumblr posts from 17 depression-related accounts; 82% of posts were related to depression, suicide or self-harm . There are differences in how social media platforms are used, for example, different trends in image posting between Twitter and Tumblr or Instagram [31▪▪]. Understanding the functions across online platforms that are supportive or detrimental for different age groups, or populations, will be important to guide clinicians’ line of enquiry, risk assessment as well as recommendations about social media/internet use and future interventions.
The internet provides an unrivalled opportunity to access information but sometimes the ideas gleaned can be detrimental for vulnerable young people. Capacity (dispositional, acquired and practical) to make a suicide attempt is an important factor in the Three-Step Theory of suicide  and it is not yet known if, or how, social media/internet could impact this. The Darknet is an under-researched entity, with potentially significant implications for risk assessment of suicidal youth because of the anonymity conferred by not being indexed by conventional search engines. Morch et al. found fewer websites devoted to suicide on the Darknet via the Tor browser (4%), compared with the Surface Web (23.1%)  but this appears to be the only study of its kind. Violent methods are associated with increased risk of suicide [35▪], and the internet, especially the Darknet, has the potential to increase availability to harmful means. Determining, which young people access different levels of web content and whether access to the Darknet is associated with more violent or effective means of suicide could be important areas of future research.
Internet search trend data has the potential for wide reaching possibilities in terms of surveillance and detection of those at risk of suicide. Chandler found a positive correlation between search intensity of suicide-related terms and the number of suicides across America between 2006 and 2014, particularly for youth. This methodology has also been replicated in the UK , but in both cases, not specifically looking at adolescent populations. This may suggest an association between suicide internet search activity and suicide. However, the usefulness of search trends on a population level to understand risk, especially for specific groups, such as adolescents, is limited at present.
With the rise of social media/internet, there is concern about the potential for electronic communication to facilitate clusters of youth suicides . Following a potential suicide cluster in the United States, online social networking was identified as having broad relevance, both positive and negative, on how young people hear about a suicide, the impact on them, their perceptions of the environment afterwards and recovery . In response to this potential risk, some research has generated evidence-based guidelines to help youth discuss suicide-related themes safely on social media [40▪], though more research is required to understand this complex phenomenon.
Support and a force for good
Frost et al. found one-third of young people with a history of self-injurious behaviour had used the internet to seek help in relation to self-harm. Over half of these online help-seekers perceived that they had more support available to them online than offline . In their analysis of depression-related accounts on Tumblr, Cavazos-Rehg et al. found that 9% of posts involved direct interaction with others and of these, 47% provided emotional support or reassuring messages.
Intensive community treatment can reduce need for hospital admission among adolescents  and there is increasing evidence that both ‘self-driven’ and ‘socially-driven’ processes can decrease suicide attempts . Youth-Nominated Support Team Interventions (which include youth-nominated caring adults) have potential to reduce mortality in suicidal adolescents . No intervention studies have been done using social media for suicide prevention, but there are examples of social media sites designed for suicide prevention, including sites with potential to reach those at risk of suicide , building on the increasing self-harm and suicide intervention evidence base.
Writing style could help detect suicidal youth via online platforms through identification of: internal attribution, excessive self-focus and higher psychological pain and cognitive constriction . Social media content on self-harm is not always used to actively encourage others to self-harm, but predominantly to express difficult emotions and inspire recovery [31▪▪]. For some young people, the anonymous potential of social media/internet may make it an easier place to express themselves and find support, beyond what can be offered via conventional means.
Strengths and limitations of this review
We did not include studies with participants 19 years or over, or which referred to ‘suicidal behaviour’ where suicide or suicide attempts were not specified. These narrow criteria, as well as exclusion of grey literature and non-English language publications may have excluded relevant studies. Further, the heterogeneity of exposure and outcome measures made synthesizing evidence challenging and prevented the combination of studies in a meta-analysis. However, the specific nature of our question has highlighted the existing literature on suicide, suicide attempts and social media/internet use and will be relevant to child and adolescent clinicians, as well as highlighting areas of interest for future study.
Current evidence suggests that excessive or ‘problematic’ use of social media/internet does impact suicide risk, specifically increasing the risk of suicide attempts. Longitudinal studies are vital to establish the direction of the potential association, the impacts of potential confounders, such as sleep disturbance and cyberbullying and recommendations on safe amounts of use. As internet and social media platforms develop, more understanding of the specific risks and mechanisms associated with different types of digital activity, by different population groups will be essential to understand risk and pave the way for specific interventions.
Financial support and sponsorship
R.S. and S.E. have been employed as National Institute of Health Research (NIHR) Academic Clinical Fellows. S.E. also received salary support from an MQ Data Science Award and from the Psychiatry Research Trust. R.D. is supported by a Clinician Scientist Fellowship from the Health Foundation in partnership with the Academy of Medical Sciences. D.O. is supported by research grants with the Medical Research Council (grant mr/r004927/1) and the National Institute for Health Research.
The authors acknowledge infrastructure support from the National Institute for Health Research (NIHR). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
- ▪ of special interest
- ▪▪ of outstanding interest
1. Patton GC, Coffey C, Sawyer SM, et al. Global patterns of mortality in young people: a systematic analysis of population health data. Lancet 2009; 374:881–892.
2. Kidger J, Heron J, Lewis G, et al. Adolescent self-harm and suicidal thoughts in the ALSPAC cohort: a self-report survey in England. BMC Psychiatry 2012; 12:69.
3. Boyd DM, Ellison NB. Social network sites: definition, history, and scholarship. J Comput Mediat Commun 2007; 13:210–230.
4. The SAGE Handbook of Social Media Research Methods 2017.
5. Mars B, Heron J, Klonsky ED, et al. Predictors of future suicide attempt
among adolescents with suicidal thoughts or nonsuicidal self-harm: a population-based birth cohort study. Lancet Psychiatry 2019; 6:327–337.
6. Victor SE, Klonsky ED. Correlates of suicide attempts among self-injurers: a meta-analysis. Clin Psychol Rev 2014; 34:282–297.
7. National Collaborating Centre for Mental Health (UK). Self-Harm: The Short-Term Physical and Psychological Management and Secondary Prevention of Self-Harm in Primary and Secondary Care. Leicester (UK): British Psychological Society; 2004. (NICE Clinical Guidelines, No. 16.) Available from: https://www.ncbi.nlm.nih.gov/books/NBK56385/
. [Accessed 9 April 2019].
8. Nock MK. Self-injury. Annu Rev Clin Psychol 2010; 6:339–363.
9. Hawton K, James A. Suicide and deliberate self harm in young people. BMJ 2005; 330:891–894.
10▪▪. Stiglic N, Viner RM. Effects of screentime on the health and well being of children and adolescents: a systematic review of reviews. BMJ Open 2019; 9:e023191.
This systematic review, reviews investigated associations between a range of health and wellbeing outcomes and screentime in young people. There was some evidence of associations between screentime and depression, but weak or insufficient evidence of associations between screentime and factors, such as suicidal ideation.
11▪▪. John A, Glendenning AC, Marchant A, et al. Self-harm, suicidal behaviours, and cyberbullying in children and young people: systematic review. Journal of Medical Internet Research 2018; 20:e129.
This systematic review and meta-analysis found that victims of cyberbullying are at greater risk of self-harm and suicidal behaviour. Cyberbullying perpetrators were at increased risk of suicidal ideation and suicidal behaviours.
12▪▪. Marchant A, Hawton K, Stewart A, et al. A systematic review of the relationship between internet use, self-harm and suicidal behaviour in young people: the good, the bad and the unknown (1932-6203 (Electronic)).
Comprehensive systematic review investigating a wide range of internet mediums and associations with self-harm/suicidal behaviour in young people aged under 25 years. They outlined both positive and negative influences and found that though there was potential for harm, there were also scope for use of the internet to support young people.
13. Daine K, Hawton KF, Singaravelu V, et al. The power of the web: a systematic review of studies of the influence of the internet on self-harm and suicide in young people (1932-6203 (Electronic)).
14. Media use in school-aged children and adolescents. Pediatrics 2016; 138:e20162592.
15. Commissioner Cs. Growing up digital: a report of the growing up digital taskforce. In: Commissioner Cs, editor. London 2017.
16. Government H. Internet Safety Strategy - Green paper. 2017.
17▪. Committee HoCSaT. Impact of social media and screen-use on young people's health. In: Committee HoCSaT, editor. 2019.
Recent United Kingdom guidance from the House of Commons Science and Technology Committee inquiring into the impact of social media and screen-use on young people's health including a summary of the evidence and recommendations including resources for schools and parents and regulation of social media sites.
18▪. Fuchs M, Riedl D, Bock A, et al. Pathological internet use—an important comorbidity in child and adolescent psychiatry: prevalence and correlation patterns in a naturalistic sample of adolescent inpatients. BioMed Res Int 2018; 2018:1629147.
This article is included in the narrative review and summarized above and in Table 1.
19. Lee SY, Park EC, Han KT, et al. The association of level of internet use with suicidal ideation and suicide attempts in South Korean adolescents: a focus on family structure and household economic status. Can J Psychiatry 2016; 61:243–251.
20. Alpaslan AH, Soylu N, Kocak U, Guzel HI. Problematic Internet use was more common in Turkish adolescents with major depressive disorders than controls. Acta Paediatr Int J Paediatr 2016; 105:695–700.
21. Kaess M, Durkee T, Brunner R, et al. Pathological Internet use among European adolescents: psychopathology and self-destructive behaviours. Eur Child Adolesc Psychiatry 2014; 23:1093–1102.
22. Rikkers W, Lawrence D, Hafekost J, Zubrick SR. Internet use and electronic gaming by children and adolescents with emotional and behavioural problems in Australia - results from the second Child and Adolescent Survey of Mental Health and Wellbeing. BMC Public health 2016; 16:399.
23. Sampasa-Kanyinga H, Hamilton HA. Social networking sites and mental health problems in adolescents: the mediating role of cyberbullying victimization. Eur Psychiatry 2015; 30:1021–1027.
24. Kim JY. The nonlinear association between internet using time for noneducational purposes and adolescent health. J Prevent Med Public Health 2012; 45:37–46.
25▪▪. Guo L, Luo M, Wang WX, et al. Association between problematic Internet use, sleep disturbance, and suicidal behavior in Chinese adolescents. J Behav Addict 2018; 7:965–975.
This article is included in the narrative review and summarized above and in Table 1.
26. Lin IH, Ko CH, Chang YP, et al. The association between suicidality and Internet addiction and activities in Taiwanese adolescents. Compr Psychiatry 2014; 55:504–510.
27. International statistical classification of diseases and related health problems (11th Revision): World Health Organization; 2018. Available at: https://icd.who.int/browse11/l-m/en#/http://id.who.int/icd/entity/1448597234
. [Accessed 19 June 2019].
28. Arlington V. Diagnostic and statistical manual of mental disorders, fifth edition. USA: American Psychiatric Association; 2013.
29. Brown RC, Fischer T, Goldwich AD, et al. #cutting: Nonsuicidal self-injury (NSSI) on Instagram (1469-8978 (Electronic)).
30. Cavazos-Rehg PA, Krauss MJ, Sowles SJ, et al. An analysis of depression, self-harm, and suicidal ideation content on Tumblr. Crisis 2017; 38:44–52.
31▪▪. Shanahan N, Brennan C, House A. Self-harm and social media: thematic analysis of images posted on three social media sites. BMJ Open 2019; 9:e027006.
Thematic analysis of visual content of 602 images from Instagram, Tumblr and Twitter. Very few images tagged as self-harm were found to show explicit representation of self-harm and none were encouraging or overtly sensationalizing self-harm. Key themes were: ‘communicating distress’, ‘addiction and recovery’, ‘gender and the female body’ and ‘identity and belonging’.
32. Klonsky ED, May AM. The three-step theory (3ST): a new theory of suicide rooted in the ‘Ideation-to-Action’ framework. Int J Cogn Ther 2015; 8:114–129.
33. Morch CM, Cote LP, Corthesy-Blondin L, et al. The Darknet and suicide (1573-2517 (Electronic)).
34. Biddle L, Derges J, Mars B, et al. Suicide and the internet: changes in the accessibility of suicide-related information between 2007 and 2014. J Affect Disord 2016; 190:370–375.
35▪. Beckman K, Mittendorfer-Rutz E, Waern M, et al. Method of self-harm in adolescents and young adults and risk of subsequent suicide (1469-7610 (Electronic)).
This article is not directly relating to social media/internet use, but reports longitudinal findings to inform assessing and predicting suicide risk.
36. Chandler V. Google and suicides: what can we learn about the use of internet to prevent suicides? Public Health 2018; 154:144–150.
37. Arora VS, Stuckler D, McKee M. Tracking search engine queries for suicide in the United Kingdom. Public Health 2016; 137:147–153.
38. Robertson L, Skegg K, Poore M, et al. An adolescent suicide cluster and the possible role of electronic communication technology. Crisis 2012; 33:239–245.
39. Heffel CJ, Riggs SA, Ruiz JM, Ruggles M. The aftermath of a suicide cluster in the age of online social networking: a qualitative analysis of adolescent grief reactions. Contemp School Psychol 2015; 19:286–299.
40▪. Robinson JA-OhooX, Hill NTMA-Ohoo, Thorn P, et al. The #chatsafe project. Developing guidelines to help young people communicate safely about suicide on social media: a Delphi study (1932-6203 (Electronic)).
This reports findings from a Delphi study, which has informed the creation of evidence-based guidelines to support young people, parents and professionals to communicate about suicide in a safe and supportive way on social media.
41. Frost M, Casey L. Who seeks help online for self-injury? Arch Suicide Res 2016; 20:69–79.
42. Ougrin D, Corrigall R, Poole J, et al. Comparison of effectiveness and cost-effectiveness of an intensive community supported discharge service versus treatment as usual for adolescents with psychiatric emergencies: a randomised controlled trial. The Lancet Psychiatry 2018; 5:477–485.
43. Iyengar U, Snowden N, Asarnow JR, et al. A further look at therapeutic interventions for suicide attempts and self-harm in adolescents: an updated systematic review of randomized controlled trials (1664-0640 (Print)).
44. King CA, Arango A, Kramer A, et al. Association of the Youth-Nominated Support Team Intervention for Suicidal Adolescents With 11- to 14-Year Mortality Outcomes: Secondary Analysis of a Randomized Clinical Trial. LID - 10.1001/jamapsychiatry.2018.4358 [doi]. (2168-6238 (Electronic)).
45. Robinson J, Cox G, Bailey E, et al. Social media and suicide prevention: a systematic review. Early Interv Psychiatry 2016; 10:103–121.
46. Barak A, Miron O. Writing characteristics of suicidal people on the internet: A psychological investigation of emerging social environments. Suicide Life Threat Behav 2005; 35:507–524.