Discussion

In general, the most common/frequent methods for suicide are hanging, pesticide self-poisoning and firearms; however, suicide methods exhibit remarkable various characteristics depending on the social/cultural backgrounds of each country.19 27 28 Indeed, the most common/frequent method for suicide in Japan was hanging, accounting for more than 60%.2 Japanese frequency of hanging suicide is higher than those in European countries.18 29 30 Total counts during 2009–2018, charcoal-burning suicide was remarkably the common suicide method of Japanese men with ranking second, whereas female charcoal-burning suicide was relatively low with ranking third (almost equal to the counts of poisoning at fourth rank). Furthermore, in Europe, it has been considered that jumping is not common suicide method,20 whereas in East countries, jumping was one of the common/frequent methods for suicide.23 31 32 Indeed, total counts during 2009–2018, jumping was also one of the common methods for suicide at third rank of man and at second rank of woman. Therefore, Japanese features of suicide methods are different from those in Europe, but additionally, the suicide methods between man and woman in Japan are also different. A correspondence analysis indicated that the correspondence maps between genders were more similar rather than among countries.27 In this correspondence analysis study, a notable exception that was that European male firearms for suicide replaced female poisoning.27 Therefore, analysing the impact of EFECBSC, which is one of the important programmes in comprehensive suicide prevention, on the gender-specific characteristics of suicide methods provide important implications for the improvement of future suicide prevention programmes.

It has been established that ‘means restriction’, which suppresses the access to common/frequent suicide methods, provides the reduction of suicide mortality.19 A number of studies reported the effectiveness of means restriction for suicide prevention, via suppression of access to drugs, pesticides, charcoal, gun and barriers on bridges, railway platforms and high-rise buildings.11–20 In Japan, the limitation of use of hypnotics decreased suicide mortality by poisoning in 1956, whereas, following then, there have been no intervention or postevaluation analyses regarding the effects of means restriction on suicide mortality.15 23 Prevention programmes in Japan have enhanced community welfare and social safety nets, rather than ‘means restriction’.5–8 The strategy of Japanese suicide prevention programmes is influenced by the features of Japanese suicide methods, most common suicide method was hanging, that is hardly regulated by ‘means restriction’.2 18 The majority place of hanging was the person’s home, custody, prisons and psychiatric wards.18 Most common tools for hanging were rope, belts, electric flex, rafters/beams, bannisters, hooks, doorknobs and trees.18 Furthermore, a study of lethal suicide attempters reported that 24% took less than 5 min between decision and suicide action and 70% within 1 hour.33 It has been established the theory of means restriction that suppression of access to lethal suicide methods converts to other methods with lower risk of lethality or allows time for reduction of suicide impulsivity during suicide impulsivity.14 19 Therefore, the characteristics of hanging suicide, the place of execution (place of daily life) and the tools (fundamental items used in daily life) make it difficult to set the intervention targets for an effective means restriction against hanging suicide.18

In spite of these hardships regarding prevention of hanging suicide, the decrease in hanging contributed significantly to the decrease in the number of suicides in Japan.2 It is easy to expound that combination between ELM activities within relatively wide-range live spheres (prefecture) and enhancement of qualitative activities of community-based (municipality) gatekeepers contributes to reduction of total and male hanging suicides; however, unexpectedly, municipal PCS related to the decrease in female hanging suicide mortality. Previous EFECBSC cost-effectiveness analysis studies could not detect the significant effects other than the negative effects of municipal PCS on suicide mortality of elders.5 6 Although this study cannot specify the detailed mechanism of this valuable finding, the positive effects of municipal PCS on prevention of female hanging suicide, which is the most frequent suicide method of Japanese woman, probably provide some suggestions for improving the ECFCBSC, which failed to detect clear contribution on a decrease in female suicide mortality due to additional qualitative researches.

An increase charcoal-burning suicide is a specific phenomenon in East/Southeast Asia in the early 21st century. After extensive media reporting about charcoal-burning suicide in Hong Kong, initially, charcoal-burning suicide was increased in Chinese-speaking countries, Hong Kong (1998), Taiwan (1999), Macau and Singapore (2001), followed by different language countries, Japan (2002) and South Korea (2008).13 In Hong Kong, the government, police and community had collaborated to implement means restrictions on charcoal burning resulting in prevention of charcoal-burning suicide.13 16 17 Initially, in also Japan, internet involvement was noted, as young Japanese charcoal-burning suicides were involved in several suicide agreements on charcoal-burning suicides.13 34 35 In 2005, a voluntary guideline was regulated by associations managing internet provider service in Japan to suppress suicidal attempts via internet35 without implementation of direct means restrictions against charcoal burning17; however, charcoal-burning suicide continued to increase until 2009. In the present study, municipal DLL and ELM decreased male and female suicide mortalities, respectively. Therefore, charcoal-burning suicide, which was once increased by the internet at the domestic level, could not be improved by suppressing the increasing factors alone, but prevention of charcoal-burning suicide required additional careful actions, including community-based municipal DLL and ELM. In other words, the risk and preventative factors for suicide are not necessarily identical, but combination between suppressing suicide risk and performing prevention factors of suicide probably contributes to effective suicide prevention. Based on the situation where many young people use social networking service (SNS) as a means of daily communication, recently MHLW has been promoting a ‘consultation service using SNS for the suicide prevention’,3 a few years later, the evaluation regarding this programme probably provide the important aspects about the impact of virtual support for suicide prevention via internet.

Although the reduction in jumping suicide was mild compared with those of suicides by hanging and charcoal burning, the impacts of EFECBSC subdivisions on jumping suicide were complicated. Total jumping suicide was decreased and increased by respective prefectural TCS and ELM, respectively. Male jumping suicide was also decreased and increased by respective prefectural TCS and municipal ELM, whereas female jumping suicide was increased by municipal PCS. Prefectural TCS was one of the least effective subdivisions of ECFCBSC, with only a weakly suppression on the suicide mortality of young populations.5 6 The possibility that prefectural TCS contributes to the reduction of jumping suicides, in a situation where the decrease in jumping suicide is stagnant, it is important information for encouragement of suicide prevention programmes in the future. Contrary, the interpretation of mechanisms of increasing male jumping suicide by prefectural and municipal ELM, which were the most effective subdivisions of EFECBSC,5 6 is incomprehensible and requires careful consideration.

It has been well known that means restriction leads to switching suicide methods, so-called ‘means substitution’.19 The complicated responsiveness of jumping suicide to EFECBSC subdivisions possibly suggests that prefectural and municipal ELM and municipal PCS generate ‘mean substitution’ from hanging and charcoal-burning suicides to jumping suicide. Indeed, prefectural ELM decreased and increased total hanging and jumping suicides, showing the exact opposite effects (figure 5). Municipal PCS also decreased and increased female hanging and jumping suicides, showing the exact opposite effects (figure 5). In the present study, stepwise multiple regression analysis could not detect the significant effect of execution amount of municipal ELM on make charcoal-burning suicide (figure 3), whereas additional linear regression analysis detected the inhibitory effect of municipal ELM on male charcoal-burning suicide (online supplemental appendix 3). Therefore, the possibility, that municipal ELM decreased male charcoal burning but increased jumping, cannot be completely denied. Based on these dynamics of suicide mortalities, prefectural ELM and municipal PCS probably replace a part of hanging suicides to jumping suicides, and municipal ELM also replace a part of charcoal-burning suicide to jumping suicide. Although drastic legal mean restrictions have not been implemented against charcoal burning or hanging suicides in Japan,23 a part of charcoal-burning and hanging (more frequent suicide methods rather than jumping in Japan), possibly, was unexpectedly substituted to jumping by municipal PCS and prefectural and municipal ELM. In other words, means substitution can be generated by not only with means restriction but also with welfare/social safety nets and regional social protection. Clarification of whether prefectural TCS selectively suppressed jumping suicide or mean substitution induced reaction can provide important aspects for improving future regional suicide prevention programmes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Effects of the execution amounts of EFECBSC subdivisions on total (men plus women), male and female EBSMR trends by suicide methods. Blue and red columns indicate significant factors for reduced and increased EBSMR trends by suicide methods, hanging, charcoal burning, jumping, poisoning and throwing detected by stepwise multiple regression analysis. DLL, development programme of leaders and listeners; EBSMR, empirical Bayes standardised mobile ratio; EFECBSC, Emergency Fund to Enhance Community-Based Suicide Countermeasure; ELM, enlightenment; IVM, intervention model; PCS, personal consultation support; TCS, telephone consultation support.