Journal of Affective Disorders 167 (2014) 322–325
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Brief report
Suicide from carbon monoxide poisoning in South Korea: 2006–2012 Young-Rim Choi a, Eun Shil Cha a, Shu-Sen Chang b,c, Young-Ho Khang d, Won Jin Lee a,n a
Department of Preventive Medicine, Korea University College of Medicine, 73, Inchon-ro, Seongbuk-gu, Seoul 136-705, South Korea Hong Kong Jockey Club Centre for Suicide Research and Prevention, The University of Hong Kong, Pokfulam, Hong Kong SAR, China Department of Social Work and Social Administration, The University of Hong Kong, Pokfulam, Hong Kong SAR, China d Institute of Health Policy and Management, Seoul National University College of Medicine, Seoul, South Korea b c
art ic l e i nf o
a b s t r a c t
Article history: Received 10 April 2014 Received in revised form 13 June 2014 Accepted 13 June 2014 Available online 24 June 2014
Backgrounds: Suicide from carbon monoxide poisoning by burning coal briquette or barbecue charcoal increased rapidly in some East Asian countries in the recent decade. The purpose of this study was to examine trends in suicides from carbon monoxide poisoning in South Korea and their epidemiologic characteristics. Methods: We presented age-standardized mortality rates of carbon monoxide suicide and compared them with those of suicide by other methods using registered death data from Statistics Korea (South Korea) from 2006 to 2012. Logistic regression analysis was conducted to estimate odds ratios of carbon monoxide suicide by socio-demographic characteristics before and after the marked increase in carbon monoxide suicide in September 2008. Results: The number of carbon monoxide suicides in South Korea was only 34 in 2006 but rapidly increased to 267 in 2008 and was 1125 in 2012, with the age-standardized rates of 0.06 (2006), 0.48 (2008), and 1.97 (2012) per 100,000 population respectively (a striking 3,183% increase in 2006–2012). Suicide by carbon monoxide poisoning showed greater odds ratios among men, younger age groups, single or the divorced, and those with high education and non-manual jobs compared with suicides by other methods. Limitations: This study only used data for fatal self-poisoning by carbon monoxide (non-fatal cases not included) and had no information on the sources of carbon monoxide. Conclusions: Carbon monoxide suicides substantially increased in South Korea over the relatively short study period and showed some distinct socio-demographic characteristics compared with suicides by other methods. & 2014 Elsevier B.V. All rights reserved.
Keywords: Coal briquette Copycat Epidemic Intentional poisoning Mortality
1. Introduction The suicide rate in South Korea for 2011 was the highest (33.3 per 100,000 population) among the Organization for Economic Cooperation and Development (OECD) countries (OECD, 2013). The rapid increase in suicide from carbon monoxide (CO) poisoning, mostly by burning coal briquette or barbecue charcoal, in some East/Southeast Asian countries (Chang et al., 2014), and the increased number of carbon monoxide deaths after a celebrity suicide by burning coal briquettes in South Korea in September 2008 (i.e. the actor Ahn Jae-Hwan) (Chen et al., 2014) have been the subject of considerable interest among public health professionals in South Korea. Identifying the magnitude and trends of carbon monoxide suicide rates and different epidemiologic
n
Corresponding author. Tel.: þ 82 2 2286 1413; fax: þ 82 2 927 7220. E-mail address: [email protected] (W.J. Lee).
http://dx.doi.org/10.1016/j.jad.2014.06.026 0165-0327/& 2014 Elsevier B.V. All rights reserved.
characteristics between suicides by carbon monoxide poisoning and other methods at the national level would provide scientific evidence for suicide prevention strategies. The purpose of this study, therefore, was to examine the trends in carbon monoxide suicides and their epidemiologic features based on national death registration data.
2. Methods Suicides by carbon monoxide poisoning were obtained from registered death data from Statistics Korea in the period between 2006 and 2012. Registered deaths were classified according to the underlying cause of death using the International Classification of Diseases and Related Health Problems, 10th Revision (ICD-10). Carbon monoxide suicides were identified using the codes X67 (intentional self-poisoning by other gases and vapors) and T58 (carbon monoxide poisoning effect). Other types of suicides included were hanging
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
(X70), pesticide poisoning (X68), and jumping (X80). Demographic data included age, gender, area of residence, educational level, month of death, marital status, and occupation. Areas of residence were grouped as metropolitan, city, and rural area based on the addresses in death certificates. Occupations were categorized into non-manual, manual, and unemployed groups. Mortality rates were directly standardized based on 5-year age bands using the World Standard Population (Ahmad et al., 2001) as the standard population. Age-specific population registration data between 2006 and 2012 were obtained from Statistics Korea. The annual rates of suicide mortality were calculated as number of death cases per 100,000 population. Odds ratios (ORs) and 95% confidence intervals (CIs) for each socio-demographic group were estimated using logistic regression models to investigate the differences in socio-demographic profiles between carbon monoxide suicide and suicide by other methods (non-CO suicide). For this analysis, suicide from carbon monoxide poisoning or not was coded as a binary outcome variable, and the study period was divided into the pre-epidemic period (January 2006–August 2008) and the post-epidemic period (September 2008–December 2012), based on the previous finding (Chen et al., 2014). In the multivariable logistic regression models all the following variables were included: age, sex, area of residence, marital status, educational level, and occupation. All statistical tests were performed via Stata, version 12.0 (StataCorp, College Station, Texas, USA).
3. Results Table 1 shows numbers and age-standardized mortality rates according to suicide methods and calendar year in South Korea between 2006 and 2012. The number of carbon monoxide suicides was only 34 in 2006 but increased rapidly to 267 in 2008 and 1125 in 2012; in 2006 carbon monoxide suicide accounted for only 0.3% of total suicides but the proportion rose to 7.9% in 2012. Carbon monoxide suicides became the fourth most common suicide method following hanging, jumping, and pesticide poisoning in South Korea in 2012. The age-standardized suicide rates from carbon monoxide poisoning increased substantially (3,183%) from 0.06 to 1.97 per 100,000 population in 2006–2012, whereas those of hanging, pesticide poisoning, and jumping changed 34.3% (8.43–11.32), 38.8% (4.95–3.03), and 45.8%, (2.73–3.98), respectively. Table 2 presents numbers and odds ratios of carbon monoxide suicide according to demographic characteristics and the pre- and post-epidemic periods between 2006 and 2012. The majority of the carbon monoxide suicides involved men, those aged under 50, and those residing in metropolis or cities. Compared with suicides by other methods, carbon monoxide suicides were more likely to be men, young age groups, single or the divorced, and those with high education and non-manual jobs. Compared with carbon monoxide
323
suicides in the pre-epidemic period, cases in the post-epidemic period showed greater ORs for men and young age groups and a tendency towards smaller ORs for rural area, single or the divorced, and spring season. In the post-epidemic period there were greater odds of carbon monoxide suicide in the winter season compared with other seasons.
4. Discussion Results of this study showed that the number of carbon monoxide suicides has skyrocketed in a short period (2006–2012) and carbon monoxide poisoning has emerged as an important means of suicide in South Korea. The percent change in suicide rates by carbon monoxide poisoning was massive (a 3,183% increase between 2006 and 2012) compared with those for hanging, pesticide poisoning, and jumping. While trends in suicides by other methods showed various patterns including a fluctuation (for hanging), a reduction (pesticide poisoning), or a plateauing (jumping) between 2006 and 2012, suicides by carbon monoxide poisoning showed an exponential increase. Carbon monoxide suicide was more prevalent among men, the young aged, single or the divorced, those with high education, and non-manual groups compared to suicides by other methods. Our findings indicate the urgency of developing national prevention strategies to reduce suicides from carbon monoxide poisoning in South Korea. The epidemic of carbon monoxide suicide in South Korea that we observed is similar to those in Hong Kong and Taiwan in late 1990s and early 2000s (Liu et al., 2007), albeit with a smaller magnitude as of 2012 (Chang et al., 2014). Potential contributors of this epidemic may include an imitation effect following increasing or extensive media reporting of carbon monoxide suicide (Chen et al., 2010, 2013). In South Korea, the influence of mass media reporting on carbon monoxide suicide has been reportedly as important (Lee et al., 2007; Huh et al., 2009; Heo et al., 2009). A recent Korean study presented that the rise in carbon monoxide suicides after Ahn Jae-Hwan's death in September 2008 was the greatest in the demographic groups most closely matching the celebrity and in the location where the celebrity took his life (Chen et al., 2014). Although media effects on carbon monoxide suicide were not statistically significant in previous studies (Fu and Chan, 2013; Suh et al., 2014), it should be noted that these prior investigations heavily relied on the number of media reports but could not incorporate the celebrity's popularity and the ways of presentation for suicide which may be also important factors of the media effect. In South Korea, a significant copycat effect on suicides has been found (Jeong et al., 2012; Ju et al., 2014). In addition, the celebrity effect on increases in suicides was reported to be higher in South Korea compared to the effect in other countries (Niederkrotenthaler et al., 2012). Because media effect
Table 1 Age-standardized mortality rates per 100,000 population by suicide method and calendar year in South Korea, 2006–2012. Year (Population)
2006 2007 2008 2009 2010 2011 2012 a b
(48,887,027) (49,130,354) (49,404,648) (49,656,756) (49,879,812) (50,111,476) (50,345,325)
CO poisoning (X67 and T58)
Hanging (X70)
Cases (%)a
ASRb
Cases (%)
34 63 267 713 695 1165 1125
0.06 0.11 0.48 1.25 1.21 2.03 1.97
4669 (43.8) 6052 (49.7) 6441 (50.1) 8259 (53.6) 8549 (53.9) 8347 (53.0) 7149 (50.5)
(0.3) (0.5) (2.1) (4.6) (4.5) (7.3) (7.9)
Pesticide poisoning (X68)
Jumping (X80)
ASR
Cases (%)
ASR
Cases (%)
ASR
Cases (%)
ASR
8.43 10.87 11.25 14.08 14.20 13.58 11.32
2747 2881 2800 2743 2719 2580 2103
4.95 5.03 4.70 4.44 4.23 3.87 3.03
1454 1708 1795 2122 2159 2328 2345
2.73 3.17 3.27 3.84 3.80 4.00 3.98
10,653 12,174 12,858 15,402 15,558 15,906 14,159
19.34 21.78 22.40 26.31 25.87 25.92 22.58
(25.8) (23.7) (21.8) (17.8) (17.5) (16.2) (14.9)
Proportion of suicide deaths by the suicide method among total suicide deaths. Age-standardized mortality rates per 100,000 population using World Standard Population.
(13.6) (14.0) (14.0) (13.8) (13.9) (14.6) (16.6)
Total (X60-X84)
(100) (100) (100) (100) (100) (100) (100)
324
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
Table 2 Odds ratios (ORs) and 95% confidence intervals (CIs) of carbon monoxide suicide by demographic characteristics and study periods in South Korea, 2006–2012. Characteristics
Pre-epidemic (January 2006–August 2008)
Post-epidemic (September 2008–December 2012)
Casesa (%)
Cases (%)
ORb (95% CI)
Total Sex Female Male
34 (22.2) 119 (77.8)
1.00 2.12 (1.42–3.18)
707 (18.1) 3202 (81.9)
1.00 2.69 (2.47–2.94)
Age 0–29 30–39 40–49 50–59 Z60
32 (20.9) 49 (32.0) 34 (22.2) 14 ( 9.2) 24 (15.7)
2.64 (1.24–5.61) 3.71 (1.94–7.09) 2.06 (1.11–3.79) 1.10 (0.54–2.23) 1.00
832 (21.3) 1287 (32.9) 965 (24.7) 516 (13.2) 309 (7.9)
6.07 (5.12–7.20) 7.24 (6.21–8.45) 4.37 (3.78–5.05) 2.55 (2.19–2.96) 1.00
Area Metropolitan City Rural
59 (38.6) 75 (49.0) 19 (12.4)
1.00 1.31 (0.92–1.86) 1.21 (0.70–2.10)
1677 (42.9) 1860 (47.6) 372 ( 9.5)
1.00 1.08 (1.00–1.15) 0.92 (0.82–1.04)
Marital status Married Unmarried Divorced, widowed
47 (30.7) 66 (43.1) 39 (25.5)
1.00 1.49 (0.92–2.41) 2.07 (1.33–3.21)
1342 (34.4) 1763 (45.2) 798 (20.5)
1.00 1.17 (1.07–1.29) 1.51 (1.37–1.65)
Educational level rElementary Middle High ZUniversity
24 24 60 42
(16.0) (16.0) (40.0) (28.0)
1.00 1.37 (0.73–2.57) 1.36 (0.76–2.43) 1.84 (0.98–3.44)
374 ( 9.7) 484 (12.6) 1820 (47.2) 1180 (30.6)
1.00 1.21 (1.04–1.40) 1.49 (1.31–1.70) 1.62 (1.41–1.87)
Occupation Unemployed Manual Non-manual
96 (62.8) 15 ( 9.8) 42 (27.5)
1.00 0.69 (0.39–1.23) 1.24 (0.84–1.84)
2091 (53.5) 460 (11.8) 1358 (34.7)
1.00 0.97 (0.87–1.09) 1.34 (1.24–1.45)
Season Spring Summer Fall Winter
67 29 20 37
1.29 (0.86–1.93) 0.63 (0.39–1.02) 0.69 (0.40–1.20) 1.00
1035 (26.5) 803 (20.5) 1083 (27.7) 988 (25.3)
0.85 (0.77–0.93) 0.65 (0.58–0.71) 0.78 (0.71–0.86) 1.00
a b
153 (100)
OR (95% CI)
(43.8) (19.0) (13.1) (24.2)
3909 (100)
Numbers may not sum to total due to missing information. Adjusted for sex, age, area, marital status, educational level, and occupation.
was reported to be particularly important during the early period of the epidemic in Taiwan (Chen et al., 2013), special caution when reporting suicide should be exercised to prevent a further increase in carbon monoxide suicide in South Korea. Although South Korea has media guidelines on reporting suicide, it has been shown that the guidelines have not been properly abided by (Kim, 2005). New media may also facilitate the spread of carbon monoxide poisoning as a suicide method. Online information about the technical details of the method and internet forums which provide a space to share pro-suicide attitude and message may have contributed to the increase in suicide pacts by carbon monoxide poisoning in Japan (Ozawa-de Silva, 2008). Our data showed greater odds ratios for carbon monoxide suicide among men, the young aged, those with high education, and non-manual groups compared with those of other suicide methods. These sociodemographic groups might have had greater access to the internet and news media and might have been more influenced by information or reporting related to the method than their counterparts. In South Korea, the average proportion of internet users within the latest one month was 82%, and reached almost 100% in the 10s–30s age groups (Korea Internet and Security Agency, 2013). These groups may also take up the technical aspect of the new method more readily than other groups. A prior study (Pitman et al., 2012) suggested that several risk factors for suicide in young men such as single marital status, psychiatric illness, and lower socioeconomic status may trigger the use of carbon monoxide as a suicide method because this group
may prefer using highly lethal methods and may be more likely to be attracted by new methods than others. It is also suggested that changes in the labor market and job insecurity may have contributed to the rise in suicide rates among young people in South Korea (Kwon et al., 2009). Another possible reason for the rapid increase in carbon monoxide suicide may be the easy access to coal briquettes in South Korea. Although the use of coal briquette in the households showed a rapid decline over the last two decades in South Korea (Korea Energy Statistics Institute, Yearbook of energy statistics, 2012), they are still used as a heating source in some households and widely used in restaurants for cooking; therefore, they can be easily purchased at local shops in South Korea. Although coal briquettes may be more commonly used in socioeconomically disadvantaged households, our data showed that carbon monoxide suicides had a higher educational level and were more likely to hold non-manual jobs than suicides by other methods. One explanation for the pattern is that people from a higher socioeconomic position are more likely to learn about and accept the new method than those from a lower socioeconomic position, as they are more exposed to media including internet and are more receptive to the new method. The historically widespread Confucian thought in Asia may also partially explain the preference for self-poisoning using carbon monoxide since it has been perceived as a non-violent way to end one's life without injury to the body (Pan et al., 2010). Our findings of some distinct socio-demographic characteristics of carbon monoxide suicide compared with other suicide methods
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
were similar to study findings from Hong Kong and Taiwan (Lin and Lu., 2008; Chang et al., 2010; Pan et al., 2010; Chen and Yip, 2011). The different profiles of carbon monoxide suicide compared to suicide by other methods suggest that carbon monoxide poisoning may attract a new group of vulnerable people who would not have died from suicide if they were unaware of the method (Chen and Yip, 2011). This would have important public health implications because it suggests that preventive measures such as responsible media reporting may be effective, as these carbon monoxide suicides are not people who have simply shifted from other methods to carbon monoxide poisoning (i.e., method substitution). This study has limitations. Validity of death certificates for carbon monoxide poisoning may be a potential concern; however, death certificates are a reliable source of information on external causes of death. In South Korea, death registration is reportedly complete and 98.3% of all deaths in the national registered death data in 2012 were confirmed by physicians' certification (Statistics Korea, 2013). Although some suicides might have been miscoded as deaths from other causes in their death certificates, Statistics Korea has corrected some of these errors using the National Police Agency data which are based on medico-legal investigations, or using other data sources from emergency departments or military data (Statistics Korea, 2013). These efforts by Statistics Korea might have reduced the magnitude of underreporting or misclassification of suicides. The lack of a specific ICD code for carbon monoxide suicide is also a limitation. We have used two codes, X67 and T58, to enhance the specificity of identifying carbon monoxide suicides. There is no information on the sources of carbon monoxide. However, it is believed that the majority of post-epidemic cases have used coal briquettes, as used in the celebrity suicide, in South Korea; to our knowledge there is no marked increase in suicide using carbon monoxide from other sources such as car exhaust gas. Further studies of suicide attempters may shed light on this issue. In conclusion, carbon monoxide poisoning has emerged as an important suicide method in South Korea and suicides using this method showed some demographic and seasonal differences relative to suicides by other methods. Further studies are needed to investigate the characteristics and proximal risk factors of carbon monoxide suicide. Urgent national intervention efforts such as stricter regulation of sensational media reporting and dangerous online information should be considered to prevent a further increase in carbon monoxide suicides in South Korea.
Conflict of interest None.
Role of funding source None.
Acknowledgment None.
325
References Ahmad, O.B., Boschi-Pinto, C., Lopez, A.D., Murray, C.J.L., Lozano, R., Inoue, M., 2001. Age standardization of rates: A new WHO Standard. GPE Discussion Paper series: No. 31. World Health Organization, Geneva. Chang, S.S., Gunnell, D., Wheeler, B.W., Yip, P.S., Sterne, J.A., 2010. The evolution of the epidemic of charcoal-burning suicide in Taiwan: a spatial and temporal analysis. PLos Med. 7, E1000212. Chang, S.S., Chen, Y.Y., Yip, P.S., Lee, W.J., Hagihara, A., Gunnell, D., 2014. Regional changes in charcoal-burning suicide rates in East/Southeast Asia from 1995 to 2011: a time trend analysis. PLoS Med. 11 (4), e1001622. Chen, Y.Y., Tsai, P.C., Chen, P.H., Fan, C.C., Hung, G., Cheng, A., 2010. Effect of media reporting of the suicide of a singer in Taiwan: the case of Ivy Li. Soc. Psychiatry Psychiatr. Epidemiol. 45, 363–369. Chen, Y.Y., Yip, P.S., 2011. Suicide sex ratios after the inception of charcoal-burning suicide in Taiwan and Hong Kong. J. Clin. Psychiatry 72, 566–567. Chen, Y.Y., Chen, F., Gunnell, D., Yip, P.S., 2013. The impact of media reporting on the emergence of charcoal burning suicide in Taiwan. PLoS One 8, e55000. Chen, Y.Y., Yip, P.S., Chan, C.H., Fu, K.W., Chang, S.S., Lee, W.J., Gunnel, L.D., 2014. The impact of a celebrity's suicide on the introduction and establishment of a new method of suicide in South Korea. Arch. Suicide Res. Fu, K.W., Chan, C.H., 2013. A study of the impact of thirteen celebrity suicides on subsequent suicide rates in South Korea from 2005 to 2009. PLoS One 8, E53870. Heo, I.Y., Choi, S.C., Lee, C.A., Ahn, J.H., Min, Y.G., Jung, Y.S., Cho, J.P., Kim, J.S., 2009. Influence of the Werther effect: an increase of intentional carbon monoxide poisoning. J. Korean Soc. Clin. Toixcol. 7, 143–149. Huh, G.Y., Jo, G.R., Kim, K.H., Ahn, Y.W., Lee, S.Y., 2009. Imitative suicide by burning charcoal in the southeastern region of Korea: the influence of mass media reporting. Leg. Med. (Tokyo) 11, S563–S564. Jeong, J., Shin, S.D., Kim, H., Hong, Y.C., Hwang, S.S., Lee, E.J., 2012. The effects of celebrity suicide on copycat suicide attempt: a multi-center observational study. Soc. Psychiatry Psychiatr. Epidemiol. 47, 957–965. Ju, J.N., Young, L.W., Seok, N.M., Yip, P.S., 2014. The impact of indiscriminate media coverage of a celebrity suicide on a society with a high suicide rate: epidemiological findings on copycat suicides from South Korea. J. Affect. Disord. 156, 56–61. Kim, Y.J., 2005. Reporting on suicide before and after the suicide guidelines presented. The Korean Society for Journalism & Communication Studies Workshop, vol. 49, pp. 140–165. Korea Energy Statistics Institute, Yearbook of energy statistics. 2012. Gyeonggi-do, Korea Energy Statistics Institute, Republic of Korea. Korea Internet & Security Agency, 2013. A survey of internet use in 2013, Seoul. Available from: 〈http://isis.kisa.or.kr/eng/〉 (accessed 04.04.14). Kwon, J.W., Chun, H., Cho, S.I., 2009. A closer look at the increase in suicide rates in South Korea from 1986-2005. BMC Public Health 9, 72. Lin, J.J., Lu, T.,H., 2008. High-risk groups for charcoal-burning suicide in Taiwan, 2001–2005. J. Clin. Psychiatry 69 (9), 1499–1501. Liu, K.Y., Beautrais, A., Caine, E., Chan, K., Chao, A., Conwell, Y., Law, C., Lee, D., Li, P., Yip, P.S., 2007. Charcoal burning suicides in Hong Kong and urban Taiwan: an illustration of the impact of a novel suicide method on overall regional rates. J. Epidemiol. Community Health 61, 248–253. Lee, S.Y., Huh, G.Y., Kim, K.H., Eun, S.S., Kim, S.J., 2007. Fatal carbon monoxide poisoning from incomplete combustion of charcoal briquettes – effect on public health of forensic medical expert's announcement through the mass media. Korean J. Leg. Med. 31, 117–120. Niederkrotenthaler, T., Fu, K.W., Yip, P.S., Fong, D.Y., Stack, S., Cheng, Q., Pirkis, J., 2012. Changes in suicide rates following media reports on celebrity suicide: a meta-analysis. J. Epidemiol. Community Health 66, 1037–1042. OECD, 2013. OECD Health Statistics 2013. Causes of mortality, Suicides, deaths per 100,000 population. Available from: 〈http://www.oecd.org/health/health-sys tems/oecdhealthdata.htm〉 (accessed 04.04.14). Ozawa-de Silva, C., 2008. Too lonely to die alone: internet suicide pacts and existential suffering in Japan. Cult. Med. Psychiatry 32 (4), 516–551. Pan, Y.J., Liao, S.C., Lee, M.B., 2010. Suicide by charcoal burning in Taiwan, 1995– 2006. J. Affect. Disord. 120, 254–257. Pitman, A., Krysinska, K., Osborn, D., King, M., 2012. Suicide in young men. Lancet 379, 2383–2392. Statistics Korea, 2013. The 2012 Annual Report on the Cause of Death Statistics. Statistics Korea, Daejeon (in Korean). Suh, S., Chang, Y., Kim, N., 2014. Quantitative exponential modelling of copycat suicides: association with mass media effect in South Korea. Epidemiol. Psychiatr. Sci. (Epub ahead of print).
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Brief report
Suicide from carbon monoxide poisoning in South Korea: 2006–2012 Young-Rim Choi a, Eun Shil Cha a, Shu-Sen Chang b,c, Young-Ho Khang d, Won Jin Lee a,n a
Department of Preventive Medicine, Korea University College of Medicine, 73, Inchon-ro, Seongbuk-gu, Seoul 136-705, South Korea Hong Kong Jockey Club Centre for Suicide Research and Prevention, The University of Hong Kong, Pokfulam, Hong Kong SAR, China Department of Social Work and Social Administration, The University of Hong Kong, Pokfulam, Hong Kong SAR, China d Institute of Health Policy and Management, Seoul National University College of Medicine, Seoul, South Korea b c
art ic l e i nf o
a b s t r a c t
Article history: Received 10 April 2014 Received in revised form 13 June 2014 Accepted 13 June 2014 Available online 24 June 2014
Backgrounds: Suicide from carbon monoxide poisoning by burning coal briquette or barbecue charcoal increased rapidly in some East Asian countries in the recent decade. The purpose of this study was to examine trends in suicides from carbon monoxide poisoning in South Korea and their epidemiologic characteristics. Methods: We presented age-standardized mortality rates of carbon monoxide suicide and compared them with those of suicide by other methods using registered death data from Statistics Korea (South Korea) from 2006 to 2012. Logistic regression analysis was conducted to estimate odds ratios of carbon monoxide suicide by socio-demographic characteristics before and after the marked increase in carbon monoxide suicide in September 2008. Results: The number of carbon monoxide suicides in South Korea was only 34 in 2006 but rapidly increased to 267 in 2008 and was 1125 in 2012, with the age-standardized rates of 0.06 (2006), 0.48 (2008), and 1.97 (2012) per 100,000 population respectively (a striking 3,183% increase in 2006–2012). Suicide by carbon monoxide poisoning showed greater odds ratios among men, younger age groups, single or the divorced, and those with high education and non-manual jobs compared with suicides by other methods. Limitations: This study only used data for fatal self-poisoning by carbon monoxide (non-fatal cases not included) and had no information on the sources of carbon monoxide. Conclusions: Carbon monoxide suicides substantially increased in South Korea over the relatively short study period and showed some distinct socio-demographic characteristics compared with suicides by other methods. & 2014 Elsevier B.V. All rights reserved.
Keywords: Coal briquette Copycat Epidemic Intentional poisoning Mortality
1. Introduction The suicide rate in South Korea for 2011 was the highest (33.3 per 100,000 population) among the Organization for Economic Cooperation and Development (OECD) countries (OECD, 2013). The rapid increase in suicide from carbon monoxide (CO) poisoning, mostly by burning coal briquette or barbecue charcoal, in some East/Southeast Asian countries (Chang et al., 2014), and the increased number of carbon monoxide deaths after a celebrity suicide by burning coal briquettes in South Korea in September 2008 (i.e. the actor Ahn Jae-Hwan) (Chen et al., 2014) have been the subject of considerable interest among public health professionals in South Korea. Identifying the magnitude and trends of carbon monoxide suicide rates and different epidemiologic
n
Corresponding author. Tel.: þ 82 2 2286 1413; fax: þ 82 2 927 7220. E-mail address: [email protected] (W.J. Lee).
http://dx.doi.org/10.1016/j.jad.2014.06.026 0165-0327/& 2014 Elsevier B.V. All rights reserved.
characteristics between suicides by carbon monoxide poisoning and other methods at the national level would provide scientific evidence for suicide prevention strategies. The purpose of this study, therefore, was to examine the trends in carbon monoxide suicides and their epidemiologic features based on national death registration data.
2. Methods Suicides by carbon monoxide poisoning were obtained from registered death data from Statistics Korea in the period between 2006 and 2012. Registered deaths were classified according to the underlying cause of death using the International Classification of Diseases and Related Health Problems, 10th Revision (ICD-10). Carbon monoxide suicides were identified using the codes X67 (intentional self-poisoning by other gases and vapors) and T58 (carbon monoxide poisoning effect). Other types of suicides included were hanging
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
(X70), pesticide poisoning (X68), and jumping (X80). Demographic data included age, gender, area of residence, educational level, month of death, marital status, and occupation. Areas of residence were grouped as metropolitan, city, and rural area based on the addresses in death certificates. Occupations were categorized into non-manual, manual, and unemployed groups. Mortality rates were directly standardized based on 5-year age bands using the World Standard Population (Ahmad et al., 2001) as the standard population. Age-specific population registration data between 2006 and 2012 were obtained from Statistics Korea. The annual rates of suicide mortality were calculated as number of death cases per 100,000 population. Odds ratios (ORs) and 95% confidence intervals (CIs) for each socio-demographic group were estimated using logistic regression models to investigate the differences in socio-demographic profiles between carbon monoxide suicide and suicide by other methods (non-CO suicide). For this analysis, suicide from carbon monoxide poisoning or not was coded as a binary outcome variable, and the study period was divided into the pre-epidemic period (January 2006–August 2008) and the post-epidemic period (September 2008–December 2012), based on the previous finding (Chen et al., 2014). In the multivariable logistic regression models all the following variables were included: age, sex, area of residence, marital status, educational level, and occupation. All statistical tests were performed via Stata, version 12.0 (StataCorp, College Station, Texas, USA).
3. Results Table 1 shows numbers and age-standardized mortality rates according to suicide methods and calendar year in South Korea between 2006 and 2012. The number of carbon monoxide suicides was only 34 in 2006 but increased rapidly to 267 in 2008 and 1125 in 2012; in 2006 carbon monoxide suicide accounted for only 0.3% of total suicides but the proportion rose to 7.9% in 2012. Carbon monoxide suicides became the fourth most common suicide method following hanging, jumping, and pesticide poisoning in South Korea in 2012. The age-standardized suicide rates from carbon monoxide poisoning increased substantially (3,183%) from 0.06 to 1.97 per 100,000 population in 2006–2012, whereas those of hanging, pesticide poisoning, and jumping changed 34.3% (8.43–11.32), 38.8% (4.95–3.03), and 45.8%, (2.73–3.98), respectively. Table 2 presents numbers and odds ratios of carbon monoxide suicide according to demographic characteristics and the pre- and post-epidemic periods between 2006 and 2012. The majority of the carbon monoxide suicides involved men, those aged under 50, and those residing in metropolis or cities. Compared with suicides by other methods, carbon monoxide suicides were more likely to be men, young age groups, single or the divorced, and those with high education and non-manual jobs. Compared with carbon monoxide
323
suicides in the pre-epidemic period, cases in the post-epidemic period showed greater ORs for men and young age groups and a tendency towards smaller ORs for rural area, single or the divorced, and spring season. In the post-epidemic period there were greater odds of carbon monoxide suicide in the winter season compared with other seasons.
4. Discussion Results of this study showed that the number of carbon monoxide suicides has skyrocketed in a short period (2006–2012) and carbon monoxide poisoning has emerged as an important means of suicide in South Korea. The percent change in suicide rates by carbon monoxide poisoning was massive (a 3,183% increase between 2006 and 2012) compared with those for hanging, pesticide poisoning, and jumping. While trends in suicides by other methods showed various patterns including a fluctuation (for hanging), a reduction (pesticide poisoning), or a plateauing (jumping) between 2006 and 2012, suicides by carbon monoxide poisoning showed an exponential increase. Carbon monoxide suicide was more prevalent among men, the young aged, single or the divorced, those with high education, and non-manual groups compared to suicides by other methods. Our findings indicate the urgency of developing national prevention strategies to reduce suicides from carbon monoxide poisoning in South Korea. The epidemic of carbon monoxide suicide in South Korea that we observed is similar to those in Hong Kong and Taiwan in late 1990s and early 2000s (Liu et al., 2007), albeit with a smaller magnitude as of 2012 (Chang et al., 2014). Potential contributors of this epidemic may include an imitation effect following increasing or extensive media reporting of carbon monoxide suicide (Chen et al., 2010, 2013). In South Korea, the influence of mass media reporting on carbon monoxide suicide has been reportedly as important (Lee et al., 2007; Huh et al., 2009; Heo et al., 2009). A recent Korean study presented that the rise in carbon monoxide suicides after Ahn Jae-Hwan's death in September 2008 was the greatest in the demographic groups most closely matching the celebrity and in the location where the celebrity took his life (Chen et al., 2014). Although media effects on carbon monoxide suicide were not statistically significant in previous studies (Fu and Chan, 2013; Suh et al., 2014), it should be noted that these prior investigations heavily relied on the number of media reports but could not incorporate the celebrity's popularity and the ways of presentation for suicide which may be also important factors of the media effect. In South Korea, a significant copycat effect on suicides has been found (Jeong et al., 2012; Ju et al., 2014). In addition, the celebrity effect on increases in suicides was reported to be higher in South Korea compared to the effect in other countries (Niederkrotenthaler et al., 2012). Because media effect
Table 1 Age-standardized mortality rates per 100,000 population by suicide method and calendar year in South Korea, 2006–2012. Year (Population)
2006 2007 2008 2009 2010 2011 2012 a b
(48,887,027) (49,130,354) (49,404,648) (49,656,756) (49,879,812) (50,111,476) (50,345,325)
CO poisoning (X67 and T58)
Hanging (X70)
Cases (%)a
ASRb
Cases (%)
34 63 267 713 695 1165 1125
0.06 0.11 0.48 1.25 1.21 2.03 1.97
4669 (43.8) 6052 (49.7) 6441 (50.1) 8259 (53.6) 8549 (53.9) 8347 (53.0) 7149 (50.5)
(0.3) (0.5) (2.1) (4.6) (4.5) (7.3) (7.9)
Pesticide poisoning (X68)
Jumping (X80)
ASR
Cases (%)
ASR
Cases (%)
ASR
Cases (%)
ASR
8.43 10.87 11.25 14.08 14.20 13.58 11.32
2747 2881 2800 2743 2719 2580 2103
4.95 5.03 4.70 4.44 4.23 3.87 3.03
1454 1708 1795 2122 2159 2328 2345
2.73 3.17 3.27 3.84 3.80 4.00 3.98
10,653 12,174 12,858 15,402 15,558 15,906 14,159
19.34 21.78 22.40 26.31 25.87 25.92 22.58
(25.8) (23.7) (21.8) (17.8) (17.5) (16.2) (14.9)
Proportion of suicide deaths by the suicide method among total suicide deaths. Age-standardized mortality rates per 100,000 population using World Standard Population.
(13.6) (14.0) (14.0) (13.8) (13.9) (14.6) (16.6)
Total (X60-X84)
(100) (100) (100) (100) (100) (100) (100)
324
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
Table 2 Odds ratios (ORs) and 95% confidence intervals (CIs) of carbon monoxide suicide by demographic characteristics and study periods in South Korea, 2006–2012. Characteristics
Pre-epidemic (January 2006–August 2008)
Post-epidemic (September 2008–December 2012)
Casesa (%)
Cases (%)
ORb (95% CI)
Total Sex Female Male
34 (22.2) 119 (77.8)
1.00 2.12 (1.42–3.18)
707 (18.1) 3202 (81.9)
1.00 2.69 (2.47–2.94)
Age 0–29 30–39 40–49 50–59 Z60
32 (20.9) 49 (32.0) 34 (22.2) 14 ( 9.2) 24 (15.7)
2.64 (1.24–5.61) 3.71 (1.94–7.09) 2.06 (1.11–3.79) 1.10 (0.54–2.23) 1.00
832 (21.3) 1287 (32.9) 965 (24.7) 516 (13.2) 309 (7.9)
6.07 (5.12–7.20) 7.24 (6.21–8.45) 4.37 (3.78–5.05) 2.55 (2.19–2.96) 1.00
Area Metropolitan City Rural
59 (38.6) 75 (49.0) 19 (12.4)
1.00 1.31 (0.92–1.86) 1.21 (0.70–2.10)
1677 (42.9) 1860 (47.6) 372 ( 9.5)
1.00 1.08 (1.00–1.15) 0.92 (0.82–1.04)
Marital status Married Unmarried Divorced, widowed
47 (30.7) 66 (43.1) 39 (25.5)
1.00 1.49 (0.92–2.41) 2.07 (1.33–3.21)
1342 (34.4) 1763 (45.2) 798 (20.5)
1.00 1.17 (1.07–1.29) 1.51 (1.37–1.65)
Educational level rElementary Middle High ZUniversity
24 24 60 42
(16.0) (16.0) (40.0) (28.0)
1.00 1.37 (0.73–2.57) 1.36 (0.76–2.43) 1.84 (0.98–3.44)
374 ( 9.7) 484 (12.6) 1820 (47.2) 1180 (30.6)
1.00 1.21 (1.04–1.40) 1.49 (1.31–1.70) 1.62 (1.41–1.87)
Occupation Unemployed Manual Non-manual
96 (62.8) 15 ( 9.8) 42 (27.5)
1.00 0.69 (0.39–1.23) 1.24 (0.84–1.84)
2091 (53.5) 460 (11.8) 1358 (34.7)
1.00 0.97 (0.87–1.09) 1.34 (1.24–1.45)
Season Spring Summer Fall Winter
67 29 20 37
1.29 (0.86–1.93) 0.63 (0.39–1.02) 0.69 (0.40–1.20) 1.00
1035 (26.5) 803 (20.5) 1083 (27.7) 988 (25.3)
0.85 (0.77–0.93) 0.65 (0.58–0.71) 0.78 (0.71–0.86) 1.00
a b
153 (100)
OR (95% CI)
(43.8) (19.0) (13.1) (24.2)
3909 (100)
Numbers may not sum to total due to missing information. Adjusted for sex, age, area, marital status, educational level, and occupation.
was reported to be particularly important during the early period of the epidemic in Taiwan (Chen et al., 2013), special caution when reporting suicide should be exercised to prevent a further increase in carbon monoxide suicide in South Korea. Although South Korea has media guidelines on reporting suicide, it has been shown that the guidelines have not been properly abided by (Kim, 2005). New media may also facilitate the spread of carbon monoxide poisoning as a suicide method. Online information about the technical details of the method and internet forums which provide a space to share pro-suicide attitude and message may have contributed to the increase in suicide pacts by carbon monoxide poisoning in Japan (Ozawa-de Silva, 2008). Our data showed greater odds ratios for carbon monoxide suicide among men, the young aged, those with high education, and non-manual groups compared with those of other suicide methods. These sociodemographic groups might have had greater access to the internet and news media and might have been more influenced by information or reporting related to the method than their counterparts. In South Korea, the average proportion of internet users within the latest one month was 82%, and reached almost 100% in the 10s–30s age groups (Korea Internet and Security Agency, 2013). These groups may also take up the technical aspect of the new method more readily than other groups. A prior study (Pitman et al., 2012) suggested that several risk factors for suicide in young men such as single marital status, psychiatric illness, and lower socioeconomic status may trigger the use of carbon monoxide as a suicide method because this group
may prefer using highly lethal methods and may be more likely to be attracted by new methods than others. It is also suggested that changes in the labor market and job insecurity may have contributed to the rise in suicide rates among young people in South Korea (Kwon et al., 2009). Another possible reason for the rapid increase in carbon monoxide suicide may be the easy access to coal briquettes in South Korea. Although the use of coal briquette in the households showed a rapid decline over the last two decades in South Korea (Korea Energy Statistics Institute, Yearbook of energy statistics, 2012), they are still used as a heating source in some households and widely used in restaurants for cooking; therefore, they can be easily purchased at local shops in South Korea. Although coal briquettes may be more commonly used in socioeconomically disadvantaged households, our data showed that carbon monoxide suicides had a higher educational level and were more likely to hold non-manual jobs than suicides by other methods. One explanation for the pattern is that people from a higher socioeconomic position are more likely to learn about and accept the new method than those from a lower socioeconomic position, as they are more exposed to media including internet and are more receptive to the new method. The historically widespread Confucian thought in Asia may also partially explain the preference for self-poisoning using carbon monoxide since it has been perceived as a non-violent way to end one's life without injury to the body (Pan et al., 2010). Our findings of some distinct socio-demographic characteristics of carbon monoxide suicide compared with other suicide methods
Y.-R. Choi et al. / Journal of Affective Disorders 167 (2014) 322–325
were similar to study findings from Hong Kong and Taiwan (Lin and Lu., 2008; Chang et al., 2010; Pan et al., 2010; Chen and Yip, 2011). The different profiles of carbon monoxide suicide compared to suicide by other methods suggest that carbon monoxide poisoning may attract a new group of vulnerable people who would not have died from suicide if they were unaware of the method (Chen and Yip, 2011). This would have important public health implications because it suggests that preventive measures such as responsible media reporting may be effective, as these carbon monoxide suicides are not people who have simply shifted from other methods to carbon monoxide poisoning (i.e., method substitution). This study has limitations. Validity of death certificates for carbon monoxide poisoning may be a potential concern; however, death certificates are a reliable source of information on external causes of death. In South Korea, death registration is reportedly complete and 98.3% of all deaths in the national registered death data in 2012 were confirmed by physicians' certification (Statistics Korea, 2013). Although some suicides might have been miscoded as deaths from other causes in their death certificates, Statistics Korea has corrected some of these errors using the National Police Agency data which are based on medico-legal investigations, or using other data sources from emergency departments or military data (Statistics Korea, 2013). These efforts by Statistics Korea might have reduced the magnitude of underreporting or misclassification of suicides. The lack of a specific ICD code for carbon monoxide suicide is also a limitation. We have used two codes, X67 and T58, to enhance the specificity of identifying carbon monoxide suicides. There is no information on the sources of carbon monoxide. However, it is believed that the majority of post-epidemic cases have used coal briquettes, as used in the celebrity suicide, in South Korea; to our knowledge there is no marked increase in suicide using carbon monoxide from other sources such as car exhaust gas. Further studies of suicide attempters may shed light on this issue. In conclusion, carbon monoxide poisoning has emerged as an important suicide method in South Korea and suicides using this method showed some demographic and seasonal differences relative to suicides by other methods. Further studies are needed to investigate the characteristics and proximal risk factors of carbon monoxide suicide. Urgent national intervention efforts such as stricter regulation of sensational media reporting and dangerous online information should be considered to prevent a further increase in carbon monoxide suicides in South Korea.
Conflict of interest None.
Role of funding source None.
Acknowledgment None.
325
References Ahmad, O.B., Boschi-Pinto, C., Lopez, A.D., Murray, C.J.L., Lozano, R., Inoue, M., 2001. Age standardization of rates: A new WHO Standard. GPE Discussion Paper series: No. 31. World Health Organization, Geneva. Chang, S.S., Gunnell, D., Wheeler, B.W., Yip, P.S., Sterne, J.A., 2010. The evolution of the epidemic of charcoal-burning suicide in Taiwan: a spatial and temporal analysis. PLos Med. 7, E1000212. Chang, S.S., Chen, Y.Y., Yip, P.S., Lee, W.J., Hagihara, A., Gunnell, D., 2014. Regional changes in charcoal-burning suicide rates in East/Southeast Asia from 1995 to 2011: a time trend analysis. PLoS Med. 11 (4), e1001622. Chen, Y.Y., Tsai, P.C., Chen, P.H., Fan, C.C., Hung, G., Cheng, A., 2010. Effect of media reporting of the suicide of a singer in Taiwan: the case of Ivy Li. Soc. Psychiatry Psychiatr. Epidemiol. 45, 363–369. Chen, Y.Y., Yip, P.S., 2011. Suicide sex ratios after the inception of charcoal-burning suicide in Taiwan and Hong Kong. J. Clin. Psychiatry 72, 566–567. Chen, Y.Y., Chen, F., Gunnell, D., Yip, P.S., 2013. The impact of media reporting on the emergence of charcoal burning suicide in Taiwan. PLoS One 8, e55000. Chen, Y.Y., Yip, P.S., Chan, C.H., Fu, K.W., Chang, S.S., Lee, W.J., Gunnel, L.D., 2014. The impact of a celebrity's suicide on the introduction and establishment of a new method of suicide in South Korea. Arch. Suicide Res. Fu, K.W., Chan, C.H., 2013. A study of the impact of thirteen celebrity suicides on subsequent suicide rates in South Korea from 2005 to 2009. PLoS One 8, E53870. Heo, I.Y., Choi, S.C., Lee, C.A., Ahn, J.H., Min, Y.G., Jung, Y.S., Cho, J.P., Kim, J.S., 2009. Influence of the Werther effect: an increase of intentional carbon monoxide poisoning. J. Korean Soc. Clin. Toixcol. 7, 143–149. Huh, G.Y., Jo, G.R., Kim, K.H., Ahn, Y.W., Lee, S.Y., 2009. Imitative suicide by burning charcoal in the southeastern region of Korea: the influence of mass media reporting. Leg. Med. (Tokyo) 11, S563–S564. Jeong, J., Shin, S.D., Kim, H., Hong, Y.C., Hwang, S.S., Lee, E.J., 2012. The effects of celebrity suicide on copycat suicide attempt: a multi-center observational study. Soc. Psychiatry Psychiatr. Epidemiol. 47, 957–965. Ju, J.N., Young, L.W., Seok, N.M., Yip, P.S., 2014. The impact of indiscriminate media coverage of a celebrity suicide on a society with a high suicide rate: epidemiological findings on copycat suicides from South Korea. J. Affect. Disord. 156, 56–61. Kim, Y.J., 2005. Reporting on suicide before and after the suicide guidelines presented. The Korean Society for Journalism & Communication Studies Workshop, vol. 49, pp. 140–165. Korea Energy Statistics Institute, Yearbook of energy statistics. 2012. Gyeonggi-do, Korea Energy Statistics Institute, Republic of Korea. Korea Internet & Security Agency, 2013. A survey of internet use in 2013, Seoul. Available from: 〈http://isis.kisa.or.kr/eng/〉 (accessed 04.04.14). Kwon, J.W., Chun, H., Cho, S.I., 2009. A closer look at the increase in suicide rates in South Korea from 1986-2005. BMC Public Health 9, 72. Lin, J.J., Lu, T.,H., 2008. High-risk groups for charcoal-burning suicide in Taiwan, 2001–2005. J. Clin. Psychiatry 69 (9), 1499–1501. Liu, K.Y., Beautrais, A., Caine, E., Chan, K., Chao, A., Conwell, Y., Law, C., Lee, D., Li, P., Yip, P.S., 2007. Charcoal burning suicides in Hong Kong and urban Taiwan: an illustration of the impact of a novel suicide method on overall regional rates. J. Epidemiol. Community Health 61, 248–253. Lee, S.Y., Huh, G.Y., Kim, K.H., Eun, S.S., Kim, S.J., 2007. Fatal carbon monoxide poisoning from incomplete combustion of charcoal briquettes – effect on public health of forensic medical expert's announcement through the mass media. Korean J. Leg. Med. 31, 117–120. Niederkrotenthaler, T., Fu, K.W., Yip, P.S., Fong, D.Y., Stack, S., Cheng, Q., Pirkis, J., 2012. Changes in suicide rates following media reports on celebrity suicide: a meta-analysis. J. Epidemiol. Community Health 66, 1037–1042. OECD, 2013. OECD Health Statistics 2013. Causes of mortality, Suicides, deaths per 100,000 population. Available from: 〈http://www.oecd.org/health/health-sys tems/oecdhealthdata.htm〉 (accessed 04.04.14). Ozawa-de Silva, C., 2008. Too lonely to die alone: internet suicide pacts and existential suffering in Japan. Cult. Med. Psychiatry 32 (4), 516–551. Pan, Y.J., Liao, S.C., Lee, M.B., 2010. Suicide by charcoal burning in Taiwan, 1995– 2006. J. Affect. Disord. 120, 254–257. Pitman, A., Krysinska, K., Osborn, D., King, M., 2012. Suicide in young men. Lancet 379, 2383–2392. Statistics Korea, 2013. The 2012 Annual Report on the Cause of Death Statistics. Statistics Korea, Daejeon (in Korean). Suh, S., Chang, Y., Kim, N., 2014. Quantitative exponential modelling of copycat suicides: association with mass media effect in South Korea. Epidemiol. Psychiatr. Sci. (Epub ahead of print).
