The effects of suicide prevention measures reported through a psychoeducational video: A practice in Japan

Abstract

Background:

As the suicide rate in Japan has remained high since 1998, various suicide prevention measures have been implemented in Japanese local communities.

Aims:

To report our findings on the effect of a psychoeducational video as a suicide prevention measure in a Japanese rural town.

Methods:

Questionnaires were randomly mailed to 2,000 residents aged between 30 and 79 years. Within 4 weeks, volunteers in the town visited the residents individually and collected the questionnaires. The variables reported in this study are demographics, awareness of suicide prevention measures available in the town, whether the residents watched the video, help-seeking from advisers regarding suicidal ideation and financial problems and attitudes towards suicide.

Results:

We analysed data collected from 1,118 people who reported their demographics (i.e. sex, age, and job) and whether they had watched the video. By conducting a series of logistic regression and multiple regression analyses and controlling for demographic variables, we found that watching the video had substantial psychoeducational effects.

Conclusion:

Despite conducting a cross-sectional study, our new suicide prevention measures were considered effective for psychoeducation. However, further studies using a longitudinal design are needed.

Keywords

Attitude community survey educational activities preventive psychiatry suicide

Introduction

As the suicide rate in Japan has remained high since 1998, various suicide prevention measures have been implemented in Japanese local communities (for a review, see Yamashita et al., 2005). Because depression is one of the major causes of suicide (Conwell et al., 1996), most suicide prevention measures are based on a medical model; that is, they focus on early detection of and intervention for depression. While such prevention measures are considered effective, there are still some unresolved issues. First, suicide prevention measures focusing on early detection of and intervention for depression cannot stop the occurrence of depression itself. Second, although suicide prevention measures that are based on a medical model may strengthen the association between medical professionals and community residents, it is difficult to foster a social support network for these measures among residents, which is considered important for the prevention of depression and suicide (Antonucci, Fuhrer, & Dartigues, 1997). Third, since an attitude of acceptance towards suicide may exist as a result of social background (Dumesnil & Verger, 2009; Sakamoto, Tanaka, Neichi, & Ono, 2004), it is important to alter such attitudes. In order to resolve these limitations, a sociopsychological model is needed. Using a combined sociopsychological and medical model, we began implementing suicide prevention measures in 2000 in the town of Nagawa, a rural town in Aomori prefecture, located in northern Honshu (Japan’s main island) (Ono, 2004; Sakamoto, Tanaka, Neichi, & Ono, 2004). After implementing the preventive measures, the risk of suicide for elderly females in Minami district was found to be 74% less than the historical trend (Oyama et al., 2006).

In this article, we report our new suicide prevention measures and their psychoeducational effectiveness. We used a video that we made and circulated in residential areas. The town of Nagawa was combined with the adjacent town (the former Nanbu) and village (the former Fukuchi village) in 2006 to form the present town of Nanbu. Thus, the scope of the suicide prevention measures was extended from the former Nagawa town to the present Nanbu town. The aim of the new measures was to inform residents about the town’s suicide prevention measures. Moreover, in relation to the sociopsychological model, the measures had other purposes, as shown in Table 1.

Table 1.

The purposes, measures and expected effects of making and circulating the video.

Purpose	Sub-purpose	Measures	Effects expected in this study (hypotheses)
Informing residents of suicide prevention measures available in the town	Informing residents of suicide prevention measures available in the town	Making and circulating the video around the town. The mayor’s participation in the video as a commissioned welfare volunteer.	The proportion of people who know that the town has suicide prevention measures will be higher among people who watched the video than among those who did not. (Hypothesis 1).
	Informing residents of mental health consultations by public health nurses	Making a video in which family members consult a public health nurse about their father’s mental anguish.	The proportion of people who know that mental health consultations by public health nurses are available to the residents will be higher among people who watched the video than those who did not (Hypothesis 2).
Facilitating help-seeking behaviour (consultations)	Facilitating help-seeking for suicidal ideation	Making a video in which a father who had concealed his annoyance disclosed it to his family and neighbours, and then consulted a public health nurse, and finally recovered from his depression.	The proportion of people who will list family and/or relatives, friends, public health nurses, medical doctors and psychiatrists as advisers in the case of having suicidal ideation will be higher among people who watched the video than among those who did not (Hypothesis 3). People who watched the video will list more types of advisers in the case of having suicidal ideation than will those who did not watch the video (Hypothesis 4).
	Informing residents of various advisers in case of troubles (especially financial problems)	Distributing a leaflet in which names and telephone numbers of various advisers are listed (e.g. crisis intervention by telephone, law centres, child consultation centres and the local municipal office).	The proportion of people who list family and/or relatives, friends, the municipal office, commissioned welfare volunteers, the social welfare council, lawyers and bank officers as advisers in the case of having financial problems will be higher among people who watched the video than among those who did not (Hypothesis 5).People who watched the video will list more types of advisers in the case of having financial problems than will those who did not watch the video (Hypothesis 6).
Facilitating the well-being and concern for family and neighbours	Educating residents about the importance of mutual support in the neighbourhood and family	By video circulation (i.e. passing on the video to neighbours), we had the residents rethink their neighbourhood bonds. Making a video where neighbours and family members talk about how to solve a problem of a distressed family member.	We did not report the results in this article.
Reducing the attitude of acceptance towards suicide	Sending a clear message that ‘Suicide can be prevented’	In the video, a famous psychiatrist speaks to the audience about the importance of suicide prevention. The video shows that suicide can be prevented if family, neighbours and the municipal office workers work together and help the distressed person.	People who watched the video will believe in the feasibility of suicide prevention more than will those who did not watch the video (Hypothesis 7).
	Sending a clear message that ‘Tour life is important and precious’	In the video, we had junior high school students shout ‘Your life is important and precious!’ The video shows that life is worth living, and if one chooses to continue to live, they will certainly have many more wonderful experiences in their life.	People who watched the video will be less accepting of suicide than will those who did not watch the video (Hypothesis 8).

In order to achieve these goals, we decided to make a psychoeducational video using a volunteer staff living in the town. The video was completed in November 2007 and presented to the residents for the first time at the town’s mental health forum conducted in March 2008. Next, the video was circulated around the town using the Japanese ‘kairanban’ (circular notice) system. This is a system for circulating official community notices to residents in Japan, in which residents receive a notice from their neighbours and are asked to read it and then pass it on to other neighbours. After residents received the video, they were asked to watch it and then pass it on to their neighbours within 3 days. Because the area and number of households of the present Nanbu town are very large, the video circulation was conducted first in the former Nagawa town, then in the former Nanbu town and finally in the former Fukuchi village. By the end of March 2009, the circulation of the video in the whole town was completed. After the video circulation, meetings were held in various districts in the town for residents to view and discuss the video. In order to assess the effect of the video circulation, we conducted a questionnaire survey.

Methods

Participants and procedure

The authors randomly selected 2,000 residents aged between 30 and 79 years on the basis of resident registration. The questionnaires were mailed to these 2,000 people, among which 28 questionnaires were returned to the authors due to long-term absences. Within 4 weeks, health-promotion volunteers in the town visited the residents individually and collected the questionnaires. As a result, the authors received responses from 1,331 people (494 men, 751 women, 86 unknown; response rate = 66.6%).

Ethical considerations

The authors obtained permission for this study from the mayor of the town and received cooperation from public health nurses in the town. In administering the study, all participants were informed through a cover letter that their responses to the questionnaire were anonymous and that participation was voluntary. The participants were also asked to put the questionnaires into a sealed envelope when they handed them in.

Questionnaire

The authors investigated the following variables: (1) demographics (i.e. sex, age, job and cohabitants), (2) prior knowledge about suicide prevention measures available in the town, (3) whether the video was watched, (4) advisers on suicidal ideation, (5) advisers on financial problems, (6) attitudes towards suicide, (7) real and ideal level of familiarity with relatives and neighbours, (8) social support and (9) depressive symptoms. Because the purpose of this study was to examine Hypotheses 1–8 shown in Table 1, we analysed variables (1)–(6).

Awareness of suicide prevention measures in the town

One of the major purposes of the measures was to inform residents of the suicide prevention measures available in the town, especially informing them of the availability of mental health consultations by public health nurses. Thus, people were required to answer these two questions: ‘Do you know that the town has suicide prevention measures?’ and ‘Do you know that consultations by public health nurses about your mental health problems are available?’ Participants answered dichotomously (yes or no).

The influence of watching the video

In order to assess the influence of watching the video, people were asked, ‘Have you ever watched the video for mental health titled “Inochi wo tsunagu WA no machi”?’ (Our peaceful town: Connecting together for a healthy life). Participants were asked to choose from the following items: (1) Watched the video at a public-viewing platform, (2) Watched the video via video circulation, (3) Watched the video at other local gatherings and (4) Have not seen the video. We combined those who selected the former three items into a Video-watching group and those who selected the fourth item were categorised into a Non-watching group.

Advisers on suicidal ideation and financial problems

Another major purpose of our measures was to facilitate help-seeking behaviour. We examined the selection of advisers on suicidal ideation by using a case vignette. Specifically, people were presented with vignettes depicting suicidal ideation and asked to select from among the advisers shown in Table 3. We also examined the selection of advisers on financial problems by using a case vignette and asking people to select from among the advisers shown in Table 3. In order to investigate the variety of advisers from whom people seek help, the total number of advisers selected was calculated for suicidal ideation and financial problems separately.

Attitudes towards suicide

One of the main goals of our measures was to reduce the attitude of acceptance towards suicide. Thus, we assessed attitudes towards suicide by evaluating responses to the following three items: (1) ‘Suicide may be forgiven if there are enough reasons’, (2) ‘It is better that I die than survive and become a burden to my family’ and (3) ‘Suicide is a personal matter’. We assessed aspects of different attitudes towards suicide (i.e. the feasibility of suicide prevention) using the following four items: (1) ‘Suicide can be prevented if a family is concerned for the suicidal family member’, (2) ‘Suicide can be prevented if neighbours are concerned for each other’, (3) ‘Suicide can be prevented if government and other concerned authorities take counter measures’ and (4) ‘Suicide cannot be prevented’ (reverse item). Participants answered these items dichotomously (agree or disagree; answers were scored 1 and 0, respectively). Since these three and four items had sufficient internal consistency (Cronbach’s alpha = .649 and .744, respectively), we summed the three- and four-item scores and named them attitude of acceptance score and feasibility of suicide prevention score, respectively.

Data analyses

We analysed the data collected from 1,118 (450 men, 668 women) people who reported their demographics (i.e. sex, age and job) and whether they had watched the video. Because the number of people who did not answer the question on cohabitants exceeded 10%, these data were not deleted from the final data pool. Although this study aimed to examine the psychoeducational effects of the video on outcome variables (i.e. awareness of suicide prevention measures in the town, availability of consultations by public health nurses, advisers on suicidal ideation and financial problems and attitudes towards suicide), some demographics might correlate with both video-watching and outcome variables, and thus the effects might be explained by these demographic variables, not by video-watching. In order to investigate this possibility, we first examined the relationships between video-watching and demographic variables by conducting a series of chi-square tests. If any demographic variable was significantly correlated with video-watching, it was necessary to consider the demographic variable as a potential confounder. Second, we conducted a series of chi-square tests and t tests in order to examine the relationships between video-watching and outcome variables. Third, we examined the relationships between demographic variables and outcome variables by conducting a series of chi-square tests and analyses of variance. Finally, we conducted a series of logistic regression or multiple regression analyses in order to partial out the influence of demographic variables that were correlated with both outcome variables and video-watching. Incomplete data were dealt with by pairwise deletion. Data analyses were carried out by using SPSS version 20.0.

Results

Demographic variables and their associations with video-watching (Table 2)

We found that sex, age, job and area were significantly associated with video-watching. Then, we examined the adjusted residuals (non-parametric equivalent of z-scores) for the cell percentage of each subgroup. An adjusted residual score greater than 1.96 for a given subgroup percentage indicated that the subgroup differed significantly (p < .05) from the overall group percentage. We found that the following characteristics were significantly associated with watching the video: women (adjusted residual = 4.5), people aged between 60 and 69 years (adjusted residual = 2.9) and housewives (adjusted residual = 4.1). Characteristics such as men (adjusted residual = −4.5), people aged between 30 and 39 years (adjusted residual = −2.1) and office workers (adjusted residual = −2.8) were associated with the non-watching group. Therefore, we restructured age groups as follows: those between 30 and 39 years, those between 60 and 69 years and others. Similarly, we restructured job groups as follows: office workers, housewives and others.

Table 2.

The association between video-watching and demographic variables.

	Video-watching (n = 214)		Non-watching (n = 904)		χ²	P
	n	%	n	%	χ²	P
Sex
Male	57	27	393	44	20.40	<.001
Female	157	73	511	57
Age (years)
30–39	20	9	134	15	13.26	.010
40–49	43	20	160	18
50–59	53	25	239	26
60–69	68	32	201	22
70–79	30	14	170	19
Cohabitants
Alone	11	5	53	6	4.48	.482
Spouse	47	22	170	19
Parent(s) or child(ren)	82	38	325	36
Grandparents) and parents), or parent(s) and child(ren)	38	18	151	17
Other	14	7	69	8
NA	22	10	136	15
Job
Farmer	39	18	147	16	23.23	.001
Self-employed	15	7	79	9
Office worker	40	19	255	28
Housewife	39	18	79	9
Part-time worker	25	12	123	14
Unemployed	54	25	206	23
Other	2	1	15	2
Area
Fukuchi	88	41	295	33	7.17	.028
Nagawa	67	31	363	40
Nanbu	59	28	246	27

NA: no answer.

Differences between video-watching and non-watching groups in outcome variables (Table 3)

Table 3.

The association of video-watching with knowledge about suicide prevention and advisers on suicidal ideation and financial problems.

	Video-watching		Non-watching		χ²	p
	n	%	n	%	χ²	p
Knowledge about suicide prevention measures in the town
Awareness of suicide prevention measures in the town	165	78	380	42	85.64	<.001
Availability of mental health consultations by public health nurses	161	76	460	51	42.21	<.001
Advisers on suicidal ideation
No one	27	13	155	17	2.28	.131
Family and/or relatives	151	71	545	60	7.34	.007
Friends	84	39	278	31	5.33	.021
Public health nurses	34	16	74	8	10.90	.001
Medical doctors	63	29	197	22	5.25	.022
Psychiatrists	62	29	210	23	2.80	.095
Advisers on financial problems
No one	12	6	71	8	0.97	.326
Family and/or relatives	183	86	668	74	12.22	<.001
Friends	37	17	174	19	0.32	.575
Municipal office	23	11	70	8	1.67	.196
Social welfare council	21	10	58	6	2.55	.111
Commissioned welfare volunteers	14	7	53	6	0.05	.829
Lawyers	54	25	215	24	0.13	.721
Bank officers	39	18	152	17	0.15	.695

Incomplete data were dealt with by pairwise deletion.

Awareness of suicide prevention measures in the town

Compared with the non-watching group, the video-watching group was more aware of the town’s suicide prevention measures and the availability of mental health consultations by public health nurses.

Selection of advisers regarding suicidal ideation

Compared with the non-watching group, the participants in the video-watching group were more likely to seek advice onsuicidal ideation from family and/or relatives, friends, public health nurses and medical doctors. When we examined the total number of advisers selected by the residents who had answered the questionnaire, we found that when the participants had suicidal ideation, the video-watching group listed more types of advisers (M = 1.84, standard deviation (SD) = 1.24) than did the non-watching group (M = 1.44, SD = 1.14; t(1116) = 4.53, p < .001; effect size d = 0.34).

Selection of advisers regarding financial problems

A significant difference between the video-watching and non-watching groups was found only when people sought the help of family and/or relatives. When we examined the total number of advisers chosen, we found that when the participants had financial problems, the video-watching group listed more types of advisers (M = 1.73, SD = 1.12) than the non-watching group did (M = 1.52, SD = 1.04; t(1116) = 2.65, p = .008; effect size d = 0.20).

Attitudes towards suicide

As expected, the video-watching group had a lower score regarding an attitude of acceptance towards suicide; the mean acceptance-attitude scores for the video-watching and non-watching groups were 0.63 (SD = 0.94) and 0.87 (SD = 1.03), respectively (t(1030) = 3.02, p = .003, effect size d = 0.24). The video-watching group also believed in the feasibility of suicide prevention more than the non-watching group did; the mean feasibility of suicide prevention scores of the video-watching and non-watching groups was 2.25 (SD = 1.44) and 2.02 (SD = 1.45), respectively (t(1000) = 1.97, p = .049, effect size d = 0.16). However, because the effect size d did not reach the criteria for a small effect (i.e. 0.20), the ‘significant’ difference observed in the t test does not show a clear difference between the groups.

Differences among demographics (sex, age, job and area) in outcome variables

We examined the relationships between outcome variables and demographics that were correlated significantly with video-watching. Sex (men/women) was significantly related with the following outcome variables: awareness of suicide prevention measures in the town (χ²(1) = 9.76, p = .002), availability of mental health consultations by public health nurses (χ²(1) = 31.96, p < .001), advisers (friends and public health nurses) on suicidal ideation (χ²(1) = 6.94, p = .008; χ²(1) = 7.17, p = .007, respectively), the total number of advisers on financial problems (t(1116) = 2.86, p = .004) and feasibility of suicide prevention (t(1030) = 2.60, p = .009). Age (30–39 years/60–69 years/others) was significantly correlated with the following outcome variables: awareness of suicide prevention measures in the town (χ²(2) = 21.17, p < .001), availability of mental health consultations by public health nurses (χ²(2) = 37.76, p < .001), advisers (friends, public health nurses and medical doctors) on suicidal ideation (χ²(2) = 24.59, p < .001; χ²(2) = 22.02, p < .001; χ²(2) = 14.38, p = .001, respectively), the total number of advisers on financial problems (F(2, 1117) = 3.16, p = .043) and feasibility of suicide prevention (F(2, 1001) = 10.72, p < .001). Job (office workers/housewives/others) was significantly correlated with the following outcome variables: availability of mental health consultations by public health nurses (χ²(2) = 21.77, p < .001), advisers (friends and public health nurses) on suicidal ideation (χ²(2) = 23.96, p < .001; χ²(2) = 12.72, p = .002, respectively), advisers (family and/or relatives) on financial problems (χ²(2) = 6.28, p = .043) and the total number of advisers on financial problems (F(2, 1117) = 4.23, p = .015), feasibility of suicide prevention (F(2, 1001) = 5.75, p = .003). Area (Fukuchi/Nagawa/Nanbu) was significantly correlated with the following outcome variables: awareness of suicide prevention measures in the town (χ²(2) = 15.83, p < .001), availability of mental health consultations by public health nurses (χ²(2) = 10.16, p = .006), advisers (friends) on suicidal ideation (χ²(2) = 6.62, p = .037) and attitude of acceptance towards suicide (F(2, 1031) = 4.20, p = .015).

The influence of video-watching after controlling for demographic variables

In order to examine the psychoeducational effect of the video, a series of logistic regression and multiple regression analyses were conducted in which we entered video-watching and demographics that correlated significantly with outcome variables.

Awareness of suicide prevention measures in the town

Video-watching was significantly related to the awareness of suicide prevention measures available in the town (odds ratio (OR) = 4.84, 95% confidence interval (CI): 3.38, 6.94, p < .001) as well as the availability of mental health consultations by public health nurses (OR = 2.72, 95% CI: 1.91, 3.89, p < .001).

Advisers on suicidal ideation and financial problems

Video-watching was significantly related to help-seeking from family/relatives (OR = 1.58, 95% CI: 1.14, 2.18, p = .006), friends (OR = 1.54, 95% CI: 1.12, 2.13, p = .009), public health nurses (OR = 1.80, 95% CI: 1.15, 2.82, p = .011) and medical doctors (OR = 1.41, 95% CI: 1.01, 1.98, p = .044) in the case of suicidal ideation, and it was also significantly related to help-seeking from family/relatives in the case of financial problems (OR = 2.15, 95% CI: 1.43, 3.25, p < .001). Multiple regression analysis suggests that video-watching facilitated help-seeking; watching the video made the residents more aware of the different types of advisers available in the case of suicidal ideation (β = 0.13, p < .001) and financial problems (β = 0.08, p = .009).

Attitudes towards suicide

The results of multiple regression analyses suggested that video-watching reduced the attitude of acceptance towards suicide among the residents (β = −0.08, p = .011), while video-watching did not increase the feasibility of suicide prevention (β = 0.05, p = .109).

Discussion

The purpose of this study was to examine the effects of psychoeducational measures related to suicide prevention in a Japanese local community. Of the eight hypotheses shown in Table 1, five were supported, two were partially supported and one was not supported. Thus, we believe that our measures generally succeeded in achieving the goals of our study.

The first purpose of circulating the video was to inform residents of suicide prevention measures available in the town. Thus, we expected that knowledge about suicide prevention measures in the town (Hypothesis 1) and availability of mental health consultations by public health nurses (Hypothesis 2) would be more prevalent in the group who watched the video than in the group that did not. Judging from the results of logistic regression analyses, these two hypotheses were supported. Because mental health consultations by public health nurses play an important role in community mental health in Japan (Ono, 2004), informing the residents of the available consultations was expected to facilitate help-seeking behaviours towards public health nurses; this may contribute to a reduction in depression and prevention of suicide.

The second purpose of the psychoeducational measures was to facilitate residents’ help-seeking behaviours, especially when they experience suicidal ideation or financial problems. Thus, we expected that in the case of suicidal ideation or financial problems, people who watched the video were more likely to seek advice from various people shown in Table 3 (Hypotheses 3 and 5) and to list more types of advisers (Hypotheses 4 and 6) than those who did not. The results from logistic regression analyses and multiple regression analyses showed that Hypotheses 3, 4 and 6 are mostly supported, while Hypothesis 5 was weakly supported. The video may facilitate seeking advice from non-professionals (e.g. family and/or relatives, friends) and from some professionals (e.g. public health nurses, medical doctors), which is expected to reduce suicide. However, the effects of watching the video on seeking advice from psychiatrists were not significant. This might be due to the difficulty in consulting with psychiatrists in the town; because there is no psychiatric hospital in the town, medical doctors sometimes deal with patients’ mental problems. The psychoeducational effects were also weak with regard to help-seeking for financial problems. This may show a limitation of the video’s psychoeducational effects.

The third purpose of our psychoeducational measures was to reduce the attitude of acceptance towards suicide. Thus, we expected that people who watched the video would believe in the feasibility of suicide prevention (Hypothesis 7) and would be less accepting of suicide (Hypothesis 8) than were those who did not watch the video. According to the results from multiple regression analyses, Hypothesis 7 was not supported, but Hypothesis 8 was supported. Because an attitude of acceptance towards suicide is thought to support suicidal behaviours (Dumesnil & Verger, 2009; Sakamoto et al., 2004), it is important that the psychoeducational measure succeed in reducing this attitude. However, the psychoeducational effect of enhancing the feasibility of suicide prevention was not significant; residents might consider that various factors (e.g. illness, ageing) relate to suicide and those factors are in most cases beyond their control.

There are three main aspects of our prevention measures leading to psychoeducational effects: the medium (i.e. video), the video content and the method of delivering the psychoeducational information (i.e. video circulation). First, regarding the medium, the visual medium is more advantageous compared to print media (e.g. newspapers and magazines). For example, people can obtain both visual and auditory information from a video, which might enhance the psychoeducational effect (Klimes-Dougan, Yuan, Lee, & Houri, 2009). Second, in making the video, we paid close attention to certain points in the content. For example, to arouse residents’ interest, we selected a theme and settings that are closely relevant to their daily lives, and we decided to use their regional dialect. The appearance of the mayor of the town and a famous psychiatrist in the video might also have heightened its reliability. Third, we used the ‘kairanban’ (circular notice) system to distribute the video, which might have contributed to its effectiveness. In this system, residents passed the video on to their neighbours after viewing it. Because the video aimed to remind residents of the bonds within their neighbourhood, video circulation via the ‘kairanban’ system might have fulfilled this purpose.

There are some limitations to this study. First, in order to ascertain a causal relationship between video-watching and outcome measures, further studies using a longitudinal design are needed, although the authors verified the video’s psychoeducational effect in short-term experimental settings (Sakamoto et al., 2013). Second, the rate of watching the video was low (about 20%). Third, significant psychoeducational effects were not shown for some outcome variables (e.g. feasibility of suicide prevention). Despite these shortcomings, in general this study shows the effectiveness of suicide prevention measures that use psychoeducational video.

Footnotes

Funding

The part of the suicide prevention measures was supported by the Univers Foundation, and the research was partially supported by the Japan Society for the Promotion of Science, Grants-in-Aid for Scientific Research (No. 21530742).

References

Antonucci

T. C.

Fuhrer

Dartigues

J. F.

(1997). Social relations and depressive symptomatology in a sample of community-dwelling French older adults. Psychology and Aging, 12, 189–195.

Conwell

Duberstein

P. R.

Cox

Herrmann

J. H.

Forbes

N. T.

Caine

E. D.

(1996). Relationships of age and axis I diagnoses in victims of completed suicide: A psychological autopsy study. American Journal of Psychiatry, 153, 1001–1008.

Dumesnil

Verger

(2009). Public awareness campaigns about depression and suicide: A review. Psychiatric Services, 60, 1203–1213.

Klimes-Dougan

Yuan

Lee

Houri

A. K.

(2009). Suicide prevention with adolescents: Considering potential benefits and untoward effects of public service announcements. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 30, 128–135.

Ono

(2004). Suicide prevention program for the elderly: The experience in Japan. Keio Journal of Medicine, 53, 1–6.

Oyama

Ono

Watanabe

Tanaka

Kudoh

Sakashita

Yoshimura

(2006). Local community intervention through depression screening and group activity for elderly suicide prevention. Psychiatry and Clinical Neurosciences, 60, 110–114.

Sakamoto

Tanaka

Kameyama

Takizawa

Fujishima

Ono

(2013). Suicide prevention activities using a psychoeducational video in Nanbu, Aomori Prefecture. Japanese Journal of Mental Health, 28, 56–63. (in Japanese).

Sakamoto

Tanaka

Neichi

Ono

(2004). Where is help sought for depression or suicidal ideation in an elderly population living in a rural area of Japan? Psychiatry and Clinical Neurosciences, 58, 522–530.

Yamashita

Takizawa

Sakamoto

Taguchi

Takenoshita

Tanaka

… Watanabe

(2005). Suicide in Japan: Present condition and prevention measures. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 26, 12–19.