Abstract
Background:
Little is known about the factors that determine vulnerability to subsequent suicide in the community following a celebrity suicide. Our objective was to investigate the link between an alleged celebrity suicide and further suicidal behaviour in the community in India.
Methods:
Relevant news articles that reported suicidal behaviour in the population were retrieved from online news portals of regional and English language newspapers in the immediate month following the actor’s death. A deductive analysis of the retrieved suicide news articles was carried out using a pre-designed data extraction form.
Results:
A total of 1160 relevant news articles were identified from the local language (n = 985) and English (n = 175) newspapers. For a sizeable percentage of these reports (n = 65, 5.6%), the media reported links with celebrity suicide. Odds of subsequent suicide among young (Odds Ratios [OR] – 9.24), female (OR – 1.94), unemployed (OR – 7.26), those without precipitating life events (OR – 2.94) or mental illness (OR – 1.69) were higher among those with link to celebrity suicide; likewise, odds of death by hanging (OR – 49.84) and leaving a suicide note (OR – 2.03) were higher among those linked to celebrity suicide. English newspapers (OR – 4.23) were more likely to report events linked to celebrity suicide than local language newspapers
Conclusion:
Persons who died by suicide by hanging after a celebrity suicide are more likely to be young, female, unemployed, have a mental disorder or precipitating life events. Suicide prevention efforts must focus on this group and prevent the same method of suicide like that of the celebrity.
Introduction
Growing evidence in the past decade indicates a spike in suicide ideation (Fu & Yip, 2007), attempts (Cheng et al. 2007a; Jeong et al., 2012), and rates (Cheng et al., 2007b) in the general population following a celebrity suicide. This phenomenon has been attributed to a modelling effect, otherwise called the Werther effect, wherein detailed and imbalanced media portrayals of suicides may trigger suicide contagion (Phillips, 1974). Indeed, research has suggested that this association may even be causal (Pirkis & Blood, 2001). This association may be particularly pronounced if the deceased is a celebrity; as an example, studies that measured the impact of well-known celebrity suicides found more significant imitative effects (14.3 times) than non-celebrity suicides (Stack, 2003).
Several gaps remain in our understanding of what may trigger this spike in suicide rates. Apart from imbalanced media portrayals, pre-existing psychiatric morbidity (Hawton et al., 2000), greater identification with the deceased (Stack, 1987), social learning (Mesoudi, 2009), and nature of the celebrity (politician vs. entertainment celebrity) (Kim et al., 2013) have all been proposed as plausible drivers of the observed association.
Less is known about why some people are more vulnerable to copycat effects than others. Prior research shows this phenomenon to be more salient among specific subgroups; suicide rates were noted to be elevated among those belonging to the same gender and age group as the celebrity (Yip et al., 2006), and there was a rise in the use of the same suicide method as the celebrity (Fu & Chan, 2013). However, discordant findings have also been noted (Park et al., 2016). These differences may stem from cultural differences in attitudes and values towards suicide, the nature of the celebrity, and the quality of media reporting in the respective countries.
The above issue has not been investigated so far in India, the second-most populous nation in the world belonging to suicide dense WHO South-East Asian region that contributes 39% of global suicides (Vijayakumar, 2017). The suicide rates in India are among the highest in the world. The estimated annual suicide rate is 21 per 100,000 population (vs. global suicide rates of 11.4/100,000), equating to 250,000 deaths yearly with devastating social, economic, and emotional consequences (World Health Organization [WHO], 2014).
Against this background, we undertook the present research with the primary objective of investigating correlates of subsequent suicidal behaviour in the general population following alleged celebrity suicide in India. This study attempted to analyse and interpret the suicide news reports published in online newspapers from various parts of the country. In a lower-middle-income country with limited resources and a wide mental health gap, such efforts would assist in directing resources towards those who may need it the most. The reference event was the alleged suicide of Sushant Singh Rajput (SSR), a young and famous entertainment celebrity who has starred in several movies that garnered fame and national attention, and who died by suicide on June 14, 2020.
Methods
Study setting and data collection
This was a cross-sectional study conducted between June 14, 2020 (the day the actor allegedly died by suicide) to July 15, 2020 (a month after the event). A team of seven bilingual investigators with expertise in India’s various regional languages (Hindi, Tamil, Kannada, Odia, Marathi, and Bangla) searched online news portals of selected regional and English newspapers for news articles on deaths due to suicides. The team represented cities located in all the four zones of the country; Lucknow (North Zone), Bhubaneswar (East Zone), Puducherry (South Zone), and Mumbai (West Zone), thus enhancing the representativeness of data.
We included all newspaper reports related to suicidal behaviour published during the study period. We excluded news items where the focus was on reporting the celebrity suicide, reactions to celebrity suicide from prominent personalities, expert opinions, and columns related to the celebrity suicide.
All investigators underwent an online training session to clarify questionnaire responses and data coding. During the online training sessions, all the investigators were introduced to the aims and objectives of the study, the news articles to be searched for, the study questionnaire (meanings and relevance of various study variables included in the questionnaire), interpreting and coding the responses. The session ended with clarification (if any) and responding to the queries of the investigators. This session was conducted by a psychiatrist (first author) with prior experience in conducting media suicide reporting studies (Menon et al., 2020b). The investigators also got the opportunity to clarify their doubts by putting their queries in a closed group (WhatsApp) chat box, which was meant for this study.
Data extraction
For analysis and interpretation of the news reports, relevant data were extracted in a structured format using a pre-designed google form questionnaire developed by the investigators after referring to the existing research in this area. The questionnaire initially explored the report’s demographic details such as the level of a news report (national/regional), state/city of publication, date of the news publication, name of the publication, and language of the news report.
Next, the following demographic details of the individual and event were extracted: name, age, sex, marital status, educational attainment, and occupation, as well as the number of days since the passage of the reference event (celebrity suicide)
This was followed by items that captured suicide-related variables: risk factors for suicidal behaviour (issues related to interpersonal relations, finances, legal troubles, health, academic difficulties, workplace-related issues, or any other issue mentioned explicitly in the report), mode of suicide, associated life events (present/absent), description of a life event (if present), recovery of suicide note (present/absent), reference to mental health issues (presence of illness, nature of illness if present, any treatment information), substance use (yes/no), type of suicide (attempted, completed or extended), suicide pact (present/absent), reference to abetment of suicide in the report (yes/no), number of persons who died by suicide in the particular report, mentioning links with the celebrity suicide (yes/no) and any discussion about Mental Health Care act 2017 (Duffy & Kelly, 2019) in the report (yes/no). Mention of the link with celebrity suicide in the news report, was an important suicide variable. All news reports, that discuss the suicide of the index person (described in the particular news report) giving reference to the suicide of the celebrity (e.g. describing the person as the fan of the celebrity, discussing the resemblance of the index suicide with the celebrity suicide or describing the impact of celebrity suicide on the psychological well-being of the index person), were coded as “mentioning the link with celebrity suicide”.
Statistical analysis
Data were analysed and described in terms of percentages, proportions, and ratios. A comparison between different variables was made by applying the chi-square test. For significance, the p-value was taken as less than 0.05. A logistic regression analysis was used to estimate the odds ratio (OR) of factors associated with the news linked to SSR death.
Ethical approval
As the study involved information available in the public domain, no formal ethics committee approval was sought for carrying out the present work.
Result
Sample description
A total of 1160 reports of several languages was analysed. Reports in the Hindi language constituted the single largest category (n = 272, 23.4%); the distribution of other language news reports is shown in Table 1. Majority of the respondents were male (n = 757, 65.3%); a large proportion were married (n = 478, 41.2%), and employed (n = 392, 33.8%). However, a good proportion of variables were missing in the index report (Table 1). Most of the reports (n = 1009, 87%) described a single suicide; hanging was the most favoured method of suicide (n = 748, 64.5%).
Characteristics of news reports.
A total of 65 reports (5.6%) explicitly mentioned links with celebrity suicide. Following the death of SSR, there were, on average, 28 suicide news reported per day. The trend (number) of suicide reporting declined only 1 month after the initial case. Figure 1 shows the trend or number of suicide news per day following the SSR suicide on June 14.
Trend of suicide with respect to the celebrity suicide.
Comparison of suicide-related parameters between suicides linked and not linked to celebrity suicide
News reports linked to the celebrity suicide described individuals or attempts with the following characteristics more commonly: age ⩽24 years (OR: 9.24), females (OR: 1.94), unemployed (OR: 7.26), without inciting life events (OR: 2.94), died by hanging (OR: 49.84), and had suicide note (OR: 2.03). English news reports (OR: 4.23) were significantly more likely to publish events with a link to celebrity suicide (Table 2).
Comparison of suicide-related parameters between suicides linked and not linked to celebrity suicide.
Reference variable, while calculating Odds Ratio.
It also included Students, housewives and retired persons (getting pension). It refers to persons, who don’t have any money earning occupation.
Discussion
While the media linking the suicide to the celebrity suicide was considered as the group of interest in our study, it is possible that there may be other suicides in the community which may also be linked to the SSR case but were not captured or reported by the media. In a large country like India with inefficient civil registration systems, widespread under reporting of suicides and associated factors (because national data is based entirely on police registration of death with significant inter-state variations in reporting), the costs of doing a nationwide community surveillance study to identify all related suicides is challenging. Hence, we chose media reporting of suicide as an indicator to generate preliminary data in this regard for a question which has not been investigated in India thus far, but which is nevertheless relevant to ask.
Main findings of the study
This study investigated the correlates of further suicide in the population in the subsequent month following an alleged celebrity suicide. The study’s main findings were that a significant percentage (more than 5%) of subsequent suicides in the population were linked to celebrity suicide in the immediate month following the event. Such individuals were more likely to be young, female, unemployed, without precipitating life events, die by hanging, not have a mental illness, leave a suicide note, and have suicide pacts. English newspapers were more likely to report events linked to celebrity suicide than local language newspapers.
Age and gender-specific effects
Our findings are consistent with prior reports that have pointed to a rise in suicide rates among those who belonged to the same age group and used the same method of suicide as the celebrity (Cheng et al., 2007b; Fu & Chan, 2013; Fu & Yip, 2009; Hawton et al., 2000; Yip et al., 2006). However, in the present study, females were more likely to be impacted through the deceased celebrity was a man. There is a possibility that the female fans of a celebrity may project their feelings towards the celebrity as their dream partner and loss of the celebrity is likely to evoke a kind of grief reaction in them, which might be similar to getting departed or losing the real-life partner. Prior evidence in this regard has been inconsistent; a rise in suicides in the same gender (Cheng et al., 2007b; Fu & Yip, 2009; Schmidtke & Schaller, 2000; Yip et al., 2006), a rise among females (Myung et al., 2015) and no gender-specific impact in some celebrity suicides (Park et al., 2016) have all been reported.
We speculate that this could be because the influence of celebrity suicide on subsequent suicides is dependent on a complex interaction of psychosocial factors and processes rather than a linear stimulus-response relationship. Possible explanations include a greater vulnerability of women to the adverse effects of a celebrity suicide on the general population (Chen et al., 2012b; Hawton et al., 2000; Myung et al., 2015) as well as increased susceptibility to the negative media portrayal of celebrity suicides (Chen et al., 2012b; Cheng et al., 2007b). However, socio-cultural interpretation of media reports and misinformation form an important component in the ‘Werther effect’. Such effects of media, the portrayal of the ‘celebrity suicide’ and ‘fantasising’ it might be important risk factors. These factors merit further investigation.
Mode of suicide
In the present study, suicides linked to celebrity suicides more commonly used hanging as the mode of death; this was the same method that the celebrity also used. Literature suggests that the most common method of death in subsequent suicide closely associates with that used by the deceased celebrities (Chen et al., 2012b; Cheng et al., 2007b; Hegerl et al., 2013). Prior investigators have explained this phenomenon on the basis of greater identification with the deceased (Fu & Yip, 2009). Though hanging is the common mode of suicide in India and commonly young and less-educated people adopt this modality, after the celebrity suicide by hanging, the attributions of media might have shifted towards suicides done by hanging. As the suicide of the celebrity was a trending news across the country, it is possible that people, police as well as media attribute their assumptions more firmly towards the suicides by hanging.
Mediating factors
Contrary to our findings, previously published studies have shown that certain factors such as pre-existing mental illness may be a risk factor that determines vulnerability for subsequent suicides following celebrity suicide; this includes evidence from both retrospective (Hawton et al., 2000) and prospective studies (Fu & Yip, 2007). However, much of the evidence in this regard comes from developed nations where mental illness is found in nearly 90% of those who die by suicide (American Psychiatric Association [APA], 2003; Mann et al., 2005). In contrast, social determinants play a more significant role in suicide in Asia (Chen et al. 2012a; Hendin et al., 2008). This may explain our findings that those without mental illness were at a higher risk of suicide following celebrity suicide.
Prior studies have noted an increased proportion of those who experienced adverse life events among those who die by suicide following a celebrity suicide (Hawton et al., 2000); this was not observed by us. This may have to do partly with the limitations of data collection - media reporting of suicide is often limited by incompleteness of data and space constraints. However, this may also point to cultural differences in the way that celebrity suicide is perceived. It may also mean that celebrity suicide can trigger suicidal thoughts not only among those with pre-existing vulnerability but also practically among anyone in the community, as shown previously (Fu & Yip, 2007).
Implications
From a prevention perspective, our study has several implications which can broadly be divided as follows:
➢ Universal prevention strategies – This includes raising awareness about the negative impact of a celebrity suicide on population mental health. Here, the media can play an essential role by including educational information, such as the likelihood of suicide contagion, when reporting a celebrity suicide.
➢ Selective strategies – Given the association between the mode of death of the celebrity and subsequent suicides, restricting access to means may be an effective suicide prevention method. This may be easier to implement in cases of insecticide or chemical poisoning rather than ropes and other accessories. In case of hanging, suggestions include lowering the height of possible anchor points, installing anti-suicide break-away closet rods and shower curtain rods, and focussing on early warning signs of suicide risk (Baker et al., 2013). Responsible and balanced media portrayal of a celebrity suicide, a practice not adhered to in Asia (Menon et al., 2020a, 2020c), would assist in preventing suicide contagion (Ueda et al., 2014).
➢ Indicated strategies – This includes early identification of risk and close monitoring of vulnerable individuals. This could be facilitated by the periodic screening of individuals deemed to be vulnerable by family and friends. The use of artificial intelligence techniques to monitor vulnerable individuals based on content analysis of social media posts (Roy et al., 2020) may represent a cost-effective prevention strategy.
➢ Postvention activities – This could be targeted at both bereaved family members and vulnerable individuals such as adolescents and young adults through appropriate postvention programs at schools and universities.
Study limitations
Our study had several limitations. First, only online versions of purposively selected printed newspapers were covered due to ease of access. Other forms of mass media, such as television, and print media, such as tabloids and evening dailies, were not studied. Second, we have not covered social networking sites such as Facebook, Twitter, and Instagram, all of which are popular among the young. Given the evidence of suicide contagion about posts about celebrity suicide in social media (Fahey et al., 2018; Ueda et al., 2017), this limitation may have special significance.
Third, links between further suicides and celebrity suicide were inferred based on news reports in online newspapers; the actual cause of suicide was not directly investigated, and therefore, it remains unclear whether and to what extent the subsequent suicides were influenced by the index event.
Fourth, single investigators were involved in the analysis and coding of suicide reports for all included newspapers, suggesting the possibility of observer bias. Several methods were used to minimise this bias. Online training of all investigators on data coding was carried out before beginning the study. This was done through group video chat by an experienced psychiatrist with considerable experience in conducting similar media-based studies. Next, individual questions in the data collection form were designed so that most items had yes/no answers in response to single words/themes; this minimised the need for complex coding or judgement calls. Finally, an online group was created where the investigators were encouraged to post any queries that emerged during interpretation of the news reports and coding; this assisted in resolving any ambiguities during the process. Furthermore, the data collection form mandated all investigators to enter the links to individual news items compulsorily; this facilitated a random check of reliability in coding (done by the first author) and helped remove duplicate entries.
Fifth, this study was limited to 1 month following the death of the celebrity; hence, the long-term impact of celebrity suicides on subsequent suicides could not be assessed. However, our study covers the effective duration of 2–4 weeks, shown to be the period where celebrity suicides (Hawton et al., 2000; Stack, 1987) and suggestion (Phillips, 1974) exert the maximum triggering impact.
The present study was based only on a single celebrity suicide and the incident happened during the COVID-19 pandemic. It is difficult to differentiate the impact of COVID-19 and other circumstantial factors on the suicidal behaviour. To improve generalizability, we suggest that future studies investigate the impact of more numbers and varied celebrity suicide types. Simultaneously controlling for the effect of seasonal variations and adjusting for confounders such as unemployment may provide a clearer picture.
Conclusion
This study discussed the impact of celebrity suicide in the community during an ongoing global crisis (when the mental health expected to be compromised). Individuals who indulge in suicidal behaviour following a celebrity suicide by hanging are more likely to be young, female, unemployed, have a mental illness or precipitate life events, leave a suicide note and die by hanging themselves. These results suggest that suicide prevention efforts following a celebrity suicide should be directed at young women, unemployed people including students and housewives and focus on restricting access to means for the same suicide method as the celebrity. Similarly, cautious media reporting may be also helpful in the prevention of subsequent suicides. Mental health service provisions for vulnerable individuals may reduce the negative impact of a celebrity suicide on population mental health. More research is needed to uncover the underlying complex processes and pathways that may mediate subsequent suicide following a celebrity suicide.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
