Data sources and their quality

The quality of data on suicide mortality is to a considerable degree associated with the procedures underlying the ascertainment of a suicide. This is a two-stage process: first, the death has to be recognized as being due to other than natural causes; second, it has to be established that it was caused by a deliberate act of the deceased with the intention of ending his or her own life. The procedures and the responsibility for determination of suicide as the cause of death vary from country to country. In many countries, death can only be recognized as suicide by a coroner after a judicial enquiry; in others, the decision is made by an officially designated medical doctor. In most countries, the investigation of the circumstances in which sudden death occurred is carried out by the police (WHO,1974). As a result, some inter-country variation in suicide mortality is undoubtedly due to the differences in certification practices. In addition, there is no doubt that the incidence of suicide is always underestimated by the official statistics, because in most countries the verdict of suicide requires legal proof beyond reasonable doubt that the deceased intended to end his/her life; intention cannot be presumed on the basis of circumstantial evidence alone or on the balance of probabilities.

Table 1. Age-standardized suicide mortality rates in selected Asian societies

Source: WHO, World Health Statistics Annual, 1995.

* Annual standardized rate per 100,000 persons according to WHO world population standard.

Data on deaths by suicide are available in the eight countries and areas of the region listed in table 1. However, in the Republic of Korea death registration and cause of death certification are incomplete (Kim, 1990:52-56). In that country, the cause of death is certified by a physician in only about 30 per cent of all deaths (Suk, 1992). Deaths due to external causes, i.e. accidental injuries, poisonings and suicides, are investigated by the police; thus, more reliable data about the incidence of suicide may be obtained from the records of the National Police Headquarters. As expected, the proportion of registered deaths certified by a physician is higher in urban than in rural areas of the country; a large proportion of deaths with medically certified cause occurred in hospitals.
In the remaining countries and areas examined here, registration of death is considered complete according to the United Nations definition, although in Sri Lanka the certification of the cause of death is not very reliable. In 1983, for instance, 29 per cent of male and 36 per cent of female deaths were attributed to "senility, signs, symptoms and other ill-defined conditions". However, at least some of them should undoubtedly have been assigned to specific diseases and external causes, e.g. accidental deaths and suicides. Moreover, because of the country's civil disturbances, the registration of deaths has been incomplete because parts of the country are outside the administrative control of the government.

Cause of death registration, according to the principles of the World Health Organization (WHO) International Classification of Diseases and Causes of Death, was introduced in China in the 1980s and annual returns have been supplied to WHO since 1987. The registration of causes of death is not carried out on a nation-wide basis. Instead it is carried out in samples which together comprise about 10 per cent of China's population. There are 61.4 million people in the urban sample, consisting of Beijing and 36 other major cities, and 53.3 million people in the sample taken from rural areas in the eastern provinces.

In India and Bangladesh, fragmentary information is available on causes of death based on certification of symptoms by lay persons. In both countries the investigations are carried out in samples of rural populations.

In India, the Registrar General in 1965 introduced a "model" Rural Registration Scheme. Under the scheme, the paramedical staff of the Primary Health Centre gather information on deaths occurring in the villages under the Centre's responsibility and, from the reporting of symptoms and conditions prior to death, they assess the likely cause of death. The system has many limitations. In particular: the data are not representative even for the rural areas, not to say for India as a whole; the determination of cause of death from symptoms and conditions prior to death is fraught with problems. Deaths of women and children are under-reported in the survey.

Another source of data on suicide in India is the Civil Registration System (CRS). However, only about one quarter of all deaths in India are registered and, of them, only about 10 per cent are medically certified. Deaths from homicide and suicide are particularly under-reported. It was partly to overcome the shortcomings of the CRS that the Rural Registration Scheme was instituted.

According to the Indian Criminal Procedure Code, all unnatural deaths or deaths under suspicious circumstances have to be reported to the nearest police station or magistrate. Their cause is then determined by a medical examiner. The statistics are compiled by the Bureau of Police Research and Development in New Delhi and are published annually in Accidents and Suicides in India and Crime in India.

Accordingly, there are three different sources of information on mortality from violence in India; each of them provides some information about the incidence of suicide, but none of the information is of a high standard (Bhat, 1991). The discrepancies between the various sources of data on suicide in India can be gleaned from table 2. At least one-third of suicides are missed in police records. Medically certified deaths from the vital registration system identify merely about 15 per cent of deaths by suicide, largely because the system omits non-institutional deaths, but partly also in order to avoid making the judgement of suicide.

Table 2. Estimates of suicide mortality in India from different sources

Source: Bhat, Mari (1991). Mortality from Accidents and Violence in India and China. Research Report 91-06-1 Center for Population Analysis and Policy, Humphrey Institute of Public Affairs, University of Minnesota, Minneapolis, United States.

In Bangladesh, the only information about causes of death is available in the records of the Matlab field station of the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B). As with the Indian Rural Registration System, the certification of the cause of death is based on lay reporting of symptoms. The data are collected in a rural area inhabited by about 210,000 people. The quality of the cause-of-death data is not high (Zimicki and others, 1985); moreover, as a Muslim country, there may be a tendency in Bangladesh to disguise suicide as accidental death (table 3) since suicide is socially and religiously unacceptable.

Table 3. Suicide mortality in the Matlab field station of the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B): 1980-1996 annual rates per 100,000

Source: ICDDR,B annual reports.

To obtain data on suicide in the populations of the Pacific islands, one has to draw on police records and data of health authorities; only exceptionally do vital registration systems exist. Further, the data vary in completeness and quality. As in other countries, under-reporting of suicides is due to concealment as well as to uncertainty of intent to end one's life. Data which are considered comparatively reliable are shown in table 4 (Booth, forthcoming).

Table 4. Age-standardized suicide mortality rates in selected Pacific island societies

Source: Booth, H. (forthcoming). "Pacific island suicide in comparative perspective" Journal of Biosocial Science.

* WHO world standard: average annual rate per 100,000 persons.

For the majority of the countries and areas in the ESCAP region, there is no information about the incidence of suicide at all or only from small-scale studies reporting analysis of data on hospital admissions (for instance, Ladrido-Ignacio and Gensaya, 1992; Choprapawon and Visalyaputra, 1992).
It should be pointed out that only a small proportion of suicidal acts, however, result in death. In developed countries, it is estimated that there are about 8-10 attempted suicides (para-suicides) for every registered fatal case (WHO, 1968:31; Chesnais, 1976). Some estimates put this ratio as high as 20 attempts per suicide achieved (United Nations, 1996). The suicide-attempt-to-suicide-achieved ratio is also dependent on age and sex. In Australia in the period 1995/96, hospital discharge statistics show that, among young men aged 15-24, there are about six attempted suicides for each fatal case. Among young women of that age, however, there were as many as 55 attempted suicides for each registered self-inflicted death. Consequently, suicide incidence based on reported deaths may be considered only the proverbial tip of the iceberg.

One of the differences between suicides and para-suicides is the seriousness of intent to end one's life. Not all suicidal acts are intended to be fatal -- some are meant as a warning or "call for help". One of the pioneers of modern epidemiology, S. Peller, established that those who commit and those who attempt suicide differ statistically by age and sex, the latter being predominantly female and in the younger age groups (Peller, 1932). Suicides and attempted suicides also differ in the selection of method used, which is partly associated with the seriousness of the intent to die. Women in their suicidal acts most often use methods that leave, fortunately, quite a good chance to intervene and save the person's life: various therapeutic drugs, sedatives or pain killers are typical examples; cutting one's wrist is another. In contrast, suicidal acts by men are dominated by hanging, use of firearms or, lately, motor vehicle exhaust gas, all methods that have a very high case-fatality rate and leave little possibility for effective intervention.

It has been established that individuals who threaten suicide or attempt suicide are at high risk of actually committing the act and, possibly, dying as a result (Stengel, 1973).

Statistics on para-suicides are compiled either from the records of the police or from hospital admissions; as for completeness, they are even more deficient than the statistics on suicide.

Cultural and legal attitudes towards suicide

In addition to the formal restraints on passing a verdict of suicide, there are social and cultural impediments in operation. Most often they are anchored on the attitude of the dominant religion towards self-destructive behaviour. The Christian churches, in particular the Roman Catholic Church, traditionally have considered a person who has committed suicide to have died in mortal sin. As a temporal punishment, no religious rites are to be conducted at the tomb of such a person, no Mass should be said for the repose of that person's soul, and the body should not be treated with respect or buried in consecrated ground. The Roman Catholic Church has never altered its unequivocal position on the subject; however, in practice, the rigid provisions of Canon Law are often mitigated by individual members of the clergy. The Reformation did not relax the Church's position with respect to suicide but the reformed denominations adopted a more lenient attitude, in particular with respect to the burial of a suicide (Dublin, 1963).

In the Chinese cultural tradition, Confucianism stresses the virtue of family affiliation, teaching that one should not injure one's own body because it is given to a person by her/his parents. Implicitly, attempting suicide is prohibited except in cases where it promotes loyalty to the family. Chinese Buddhists, on the other hand, stress that human life is primarily one of suffering and one should learn to cope with it from childhood to maturity. Suicide happens through lack of one's ability to tolerate stress. Contemporary Chinese thinking generally does not look on the act of suicide as heroic or brave; on the other hand, it would not condemn it either (Rin, 1975; Wai-Hoi, 1992).

In Indian Hindu culture, various religious and philosophic writings have enunciated views on suicide. Some permitted suicide on religious grounds (the best sacrifice was man's life itself); others censured suicide. Sati -- the self-immolation of a widow -- received the sanction of religion: by her self-destruction on the funeral pyre of her husband, the widow would atone for the sins of her husband, free him from punishment, and open the gates of paradise for him.

In general, Hindu philosophy held rather tolerant attitudes towards suicide based on the view that life does not end with death; death, after all, only leads to rebirth in another body or form (Rao, 1975; Keith, 1912).

With regard to Islam, there is no specific interdiction of suicide in the Koran but according to historical tradition suicide is a violation of a divine command contained in the holy books of Islam. An acceptance of life's events is the cardinal factor in obedience to God; suicide, consequently, is an act of revolt against God and the perpetrator of such an act risks His wrath.

Suicide statistics are more difficult to obtain in Muslim societies than in other countries; where available, they show a very low incidence of suicide (Patton, 1912). This is illustrated in Singapore, for instance, by the marked differences in suicide mortality among ethnic Chinese, Indians and Malays: between 1968 and 1971 the suicide rate was 11.2 per 100,000 for Chinese, 10.8 per 100,00 for Indians, and 7.2 per 100,000 for "other" races (mainly Eurasians and Europeans). The annual average suicide rate for Malays (predominantly Muslims) was only 1.4 per 100,000 during the same period (Hassan, 1983). In Kuala Lumpur, in the period 1985/86 the percentage distribution by ethnicity of suicides and attempted suicides recorded by University Hospital was in stark contrast to the distribution of the population:

Both suicides and attempted suicides were heavily concentrated among the Indian and Chinese ethnic groups in Kuala Lumpur (Ong and Leng, 1992).
In some societies, suicide is considered a criminal offence punishable by law. Thus, the verdict of suicide has often been avoided whenever possible in order to spare both the reputation of the person who comitted suicide and his or her family from further embarrassment and indignity. This is, for instance, the case in Malaysia (Ong and Leng, 1992). Even where suicide is not considered a punishable offence, complicity in another's suicide remains a felony. Such provisions of law apply in many countries; as one side-effect, they open physicians assisting in euthanasia to legal prosecution.

Suicide is often a response to "problem" situations; the problems may be emotional (e.g. unhappy love affair), family problems, bereavement, the "unbearable feeling of loneliness" (see Halbwachs, 1930), financial ruin or poverty, chronic painful disease, or mental disorder. It seems that, even when the incidence of suicide is attributed to mental disorders, social influences are implicated, at least partly. According to Halbwachs (1930) suicide does not occur as a result of factors purely "internal" to the individual -- there is always an interaction between the individual and his/her social environment.

In theory, marriage is often thought to "protect" against suicide (Durkheim, 1897; Halbwachs, 1930; Ruzicka and Choi, 1993; Booth, forthcoming). However, in societies in which arranged marriage prevails and where divorce is difficult to obtain or is socially unacceptable, the only solution some unhappy wife sees to disharmony with her husband and/or in-laws is in taking her own life (Wai-Hoi, 1992; Bhat, 1991; Peng, 1992). Childlessness has been identified as a specific element adding to or provoking marital problems. The continuing low status of women is another component in the societal environment contributing to high female suicide rates.

The social and cultural structures of the society may permit suicide or even suggest it as the appropriate or honourable form of action in particular circumstances. For instance, many Japanese believe that death brings forgiveness for any transgressions committed during one's life. Therefore, suicide may be opted for as a way of solving one's problems. It may be an honourable death committed to maintain the honour or the hierarchy of a family, state and nation (Ong and Leng, 1992; Yoshimatsu, 1992): note the suicides during this year of several Japanese businessmen and bankers bankrupted by the downturn in the economy.

Levels of suicide mortality

Durkheim (1897) argued that each country has a collective tendency towards suicide such that, providing the circumstances of the country did not change in any essential way, its suicide rate also would not change. Taking this position as given, Durkheim's followers have related suicide rates to components of the social structure or culture of the groups to which the rates refer.

The suicide rates set out in tables 1-4 provide only very fragmentary information about the incidence of suicides in selected countries and areas of the ESCAP region. Despite their limitations, the data demonstrate the enormous variation in the propensity to take one's own life between the peoples of the region. Among men, the highest incidence of suicide appears to be in Chuuk State of the Federated States of Micronesia, Samoa and Sri Lanka, and among Fijian Indians. Guam in Micronesia and rural areas of China have about the same level of incidence of fatal suicides. Among women, suicide rates in rural areas of China and among Fijian Indian women exceed markedly those of any other society; suicide incidence among women in Sri Lanka and Samoa, although considerably lower, is still high by international standards.

If allowance is made for under-reporting of women's deaths in the Indian rural Cause of Death Survey (CDS) in general and for under-enumeration of suicides in particular, India's rural men and women appear to commit suicide about as often as those in urban China or the Republic of Korea.

In most societies, women die less often by suicide than do men. This is particularly manifest in Micronesia where the risk of suicide is 6-14 times higher for males than females. In most countries and areas of the region, the ratio is between two and four; however, it is lower than that in Fiji both among ethnic Indians and Fijians, and in Hong Kong China and Singapore, both having predominantly Chinese populations. In urban China the risk of suicide is about the same for men and women while in rural China suicide rates of women exceed those of men.

Age patterns of suicide mortality

In most societies, stressful situations that may precipitate suicidal behaviour are more frequent in old age. Hence, in most societies the incidence of suicide among older men and women is higher than among younger ones. A typical example of such an age pattern of suicide mortality is furnished by the data for the Hong Kong China, Republic of Korea and Singapore in figures 1A and 1B.

Figure 1A. Age pattern of suicide mortality among males in Singapore, Hong Kong, China and Republic of Korea

Figure 1B. Age pattern of suicide mortality among females in Singapore, Hong Kong, China and Republic of Korea

A similar pattern of steadily rising suicide mortality rates with increasing age has also been recorded in China and Japan. The incidence of fatal suicides in rural China exceeds by a considerable factor the incidence in urban areas. This is true for both men and women. In rural China, suicide rates of women below age 45 years exceed those of men in the same age group. This is a rather rare phenomenon, although there is some suspicion that a similar sex difference in suicide mortality may exist in rural India (Bhat, 1991) (see figures 2A-C).

Figure 2A. Age pattern of suicide mortality among males in China (urban and rural), 1994

Figure 2B. Age pattern of suicide mortality among females in China (urban and rural), 1994

Figure 2C. Average annual suicide rates in Japan, 1992

However, the pattern of an increasing incidence of suicide with age is not universal. In a few countries, the incidence of fatal suicides has been rising among young people. Since about the end of the Second World War and, in particular, since the early 1970s in Australia and New Zealand suicide rates among young adolescents have been steadily increasing. As a result, in the early 1990s, the incidence of suicide among men took up a bi-modal age pattern. In New Zealand especially, the probability of death by suicide peaks at ages 15-24 years; at ages 25-34 years, the rate is only marginally lower than at the highest age group 75 years and older. In Australia, the early peak appears to be at ages 25-34 years, followed by a decline to a trough at 55-64 years; thereafter, suicide rates increase with age, reaching the highest incidence at 75 years and older (figure 3A). Among women, suicide mortality is much lower than among men and the age pattern is less pronounced (figure 3B).

Figure 3A. Age pattern of suicide mortality among males in Australia and New Zealand

Figure 3B. Age pattern of suicide mortality among females in Australia and New Zealand

The U-shaped age pattern of suicide mortality rates has been even more evident in Sri Lanka. There young women appear to take their life almost as often as young men and only with rising age does this tendency decline noticeably (figure 4).

Figure 4. Average annual suicide rates in Sri Lanka, 1983-1986

The very high incidence of suicide among young men and women is also reported in the Pacific island populations (Booth, forthcoming 1998). In some of them, the incidence of suicide among 15-24-year-old men and women is the highest in the world: the suicide rate of young men in Chuuk State of Micronesia, i.e. 182 per 100,000, is three times higher than the next highest rates on record, namely in Samoa (64 per 100,000) and among Fijian Indians (57 per 100,000). In another of the Micronesian islands, Guam, the suicide rate of young men, at 49 per 100,000, is comparable to the highest level recorded in Europe, that of the Russian Federation (table 5).

Table 5. Suicide mortality of young adults aged 15-24 in selected countries and areas

Source: See tables 1 and 3.

Among women aged 15-24 years, the risk of suicide is highest in Samoa at 70 deaths per 100,000; it is closely followed by Fijian Indians at 60 per 100,000. The young women of Chuuk State and Guam have considerably lower risk of suicide at 12 and 10 per 100,000, respectively.
In rural China, as noted previously, suicide mortality is higher among women than men at ages below 45 years; in rural Bangladesh, Samoa and among Fijian Indians suicide incidence is higher among women than men at ages 15-24.

Although the statistics on causes of death are of poor quality in Peninsular Malaysia and Thailand, a similar sex difference in suicide mortality may be noted. In 1985, in Peninsular Malaysia, despite the very low incidence of suicides, mortality among young women aged younger than 25 years was 1.3 per 100,000, which was higher than among men in the same age group. Higher female than male suicide rates were also recorded at ages 45-54 years: i.e. 1.6 and 1.1 per 100,000, respectively. In Thailand, in 1984 suicide mortality among young women aged 15-19 years was, at 11.6 per 100,000, almost twice that of young men in that age group. In the next age group, 20-24 years, the incidence of fatal suicides among men slightly exceeded that of women: 13.8 and 12.5 per 100,000, respectively (Choprapawon and Visalyaputra, 1992).

Discussion

Given its incidence, suicide is obviously a serious social as well as public health problem in several countries and areas of the ESCAP region. In some societies, specifically on the Indian subcontinent, those of Sri Lanka, China and some Pacific islands, high suicide mortality among women under 25 years of age has been widely noted, because of, among other factors, the method of suicide used. In these societies, suicidal acts often involve the use of highly toxic pesticides: for example, ingestion of the herbicide paraquat (dipyridilium), for which there is no antidote, is associated with fatality rates in excess of 50 per cent among Fijian Indians and Samoans (Booth, forthcoming). In contrast, young women in Australia, Japan and New Zealand most often use medicinal drugs; their fatality rates are low -- considerably less than 5 per cent -- and thus timely intervention is usually possible and death may be prevented.

In recent years, in several countries of the region the rising incidence of death by suicide among young adult males has become a matter of public concern. Economic conditions and social change are often identified as the main cause of suicides, or are at least considered as contributing factors. In some of the industrialized countries of the region, youth unemployment has reached levels never previously recorded. Moreover, the average period without a job is getting longer for each new generation of school leavers. In other societies, the transition from a traditional to a modern way of life and coinciding changes in social obligations and structures bring about inter-generational conflicts and pressures which appear to influence the propensity to commit suicide, especially among young males. An aggravating factor is the recent deterioration of the economic situation in some ESCAP countries. The serious economic and financial crises experienced in Indonesia, Malaysia and Thailand, for instance, will undoubtedly result in increasing proportions of individuals living below the poverty level and some of them will be at increased risk of taking their life.

There is considerable diversity in the ways in which societal transformation leads to increased suicide rates. In China, for instance, high suicide rates in rural areas are sometimes attributed to social stress resulting from inability to adapt to the new competitive economic environment. The break-up of collectives, which previously had held almost total control over the lives of peasants, has seriously weakened the influence of ideology and law and other institutions and has lead not only to a resurgence in religious and supernatural belief systems but also to economic uncertainty (Phillips, 1998).

The diversity of responses to social and economic change is partly due to socio-cultural factors, both pre-existing in the traditional societies and developing in the modern structures. Among them are the influence of religion and its attitude towards self-inflicted death, the social stigma attached to suicide, often affecting the whole family of the person who committed suicide, and the legal considerations which may make an attempted suicide a punishable offence.

The comparatively high suicide incidence estimated to prevail in some of the countries and areas of the region points to the need for greater attention to be paid to the gravity of the situation. There is need to improve the completeness and quality of the data on suicide mortality by improving the death registration procedures and cause of death certification. In all societies, investigation of the circumstances leading to sudden death involves the police and special judicial officers, or coroners. It is important that adequate training be provided to the investigating officers to identify self-inflicted injury or intentional poisoning. In addition to cause of death data, hospital admission or discharge statistics with at least information on the cause of admission, gender and age of the patient and his/her marital status are needed to establish the extent of the problem of suicide in the society. To start with, such data could be obtained from teaching hospitals and the reporting gradually expanded to cover all public hospitals.

There is also a need in some countries to formulate policies in areas that have been established as having an association with levels of suicidal behaviour: one such area is the low status of women; another may be the adverse impact of economic transition and development. Yet another is easy access to popular methods of self-destruction, such as drugs or firearms. Thus, for instance, Sri Lanka's Government, in an attempt to curb high suicide mortality, introduced in 1980 a bill to regulate the distribution and sale of pesticides (Miller and Kearney, 1988). When introducing the Bill, the cabinet minister concerned asserted:

"This Bill has become very necessary because every day when you open the newspapers ... you read about suicides. Owing to some little thing, a person gets angry with the wife or husband or mother, rushes to a boutique, buys insecticide and swallows it". (Miller and Kearney, 1988:115)

Suicide is a personal act in response to a problem. What is considered to be the "problem" and why taking one's life is seen as the only solution is the individual's conceptualization of the situation. Herein is the rationale for suicide prevention and intervention. Although effective suicide prevention is still an elusive goal in all societies, some programmes and actions, in particular those involving non-governmental organizations, such as Lifeline and the Samaritans, have been useful in counselling and assisting people in stressful situations. In most countries of the region, however, training in counselling and the provision of such services are poorly developed. Professional psychiatric health services are rarely available even in urban areas and totally lacking in rural areas. As a result, general practitioners may have to be called upon to play a more prominent role in suicide prevention. It has been observed that some potential suicides seek consultation with their physician for depression or uneasiness of mind. In such situations the physician's counselling may prevent a suicidal act even in situations where referal to a psychiatrist is not feasible.

The fact that in the ESCAP region each year more than half a million people kill themselves and more than 5 million attempt suicide is by itself a strong argument for action to achieve a real and sustained reduction in the overall incidence of suicidal behaviour. Such action has to be not only at the governmental policy level but also has to consider the role and potential participation by non-governmental voluntary organizations of concerned individuals.

References

Bhat, Mari (1991). "Mortality from accidents and violence in India and China". Research Report 91-06-1, Center for Population Analysis and Policy, Humphrey Institute of Public Affairs, University of Minnesota, Minneapolis, MI, United States.

Booth, H. (forthcoming). "Pacific island suicide in comparative perspective" Journal of Biosocial Science.

__________ (1998). "Casualties of development? Child and youth suicide in the Pacific" Development Bulletin, 44 (January): 48-50.

Choprapawon, C. and S. Visalyaputra (1992). "Suicidal behaviour in Thailand", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 127-143.

Chesnais, J.C. (1976). Les morts violentes en France depuis 1826: comparaisons internationales. Cahiers Travaux et Documents no. 75. Institut National d'Etudes Demographiques, Paris.

Dublin, L.I. (1963). Suicide: A Sociological and Statistical Study (New York: The Ronald Press).

Durkheim, E. (1897). Le suicide: étude de sociologie English translation by J.A. Spaulding and G.Simpson (1951) Suicide: a Study in Sociology (Glencoe: The Free Press).

Halbwachs, M. (1930). Les causes du suicide Librairie Felix Alcan, Paris. English translation by H.Goldblatt (1978) The Causes of Suicide (London: Routledge and Kegan Paul).

Hassan, R. (1983). A Way of Dying. Suicide in Singapore (Kuala Lumpur: Oxford University Press).

Hastings, J. (1912). Encyclopaedia of Religion and Ethics (Edinburgh: T. and T. Clark) 12:38.

Keith, A. Berriedale (1912). "Suicide (Hindu)", in: James Hastings (ed.) Encyclopaedia of Religion and Ethics (Edinburgh: T. and T. Clark) 12:33-35.

Kim, Tai-Hun (1990). Mortality Transition in Korea 1960-1980 (Seoul: Population and Development Studies Center, Seoul National University).

Ladrido-Ignacio, L. and J.P. Gensaya (1992). "Suicidal behaviour in Manila, Philippines", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 112-126.

Miller, Barbara D. and R.N. Kearney (1988). "Women's suicide in Sri Lanka", in: Patricia Whelehan (ed.) Women and Health: Cross-cultural Perspectives (Boston: Bergin and Garvey Publishers): 110-123.

Murray, C.J.L. and A.D. Lopez (1996). Global Health Statistics: A Compendium of Incidence, Prevalence and Mortality Estimates for Over 200 Conditions, Global Burden of Disease and Injury Series vol. 2, (Boston: Harvard School of Public Health).

Ong, S. and Y.K. Leng (1992). "Suicidal behaviour in Kuala Lumpur, Malaysia", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 144-175.

Patton, Walter M. (1912). "Suicide (Muhammadan)", in: James Hastings (ed.) Encyclopaedia of Religion and Ethics (Edinburgh: T. and T. Clark) 12:38.

Peng, Kok Lee (1992). "Suicidal behaviour in Singapore", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 176-198.

Peller, S. (1932). Zur Statistik der Selbsmordhandlung. Allgemeines Statistisches Archiv 22:343.

Phillips, M.R. (1998). "The transformation of China's mental health services" The China Journal 39:1-38.

Rao, Venkoba (1975). "Suicide in India", in: N.L.Farberow (ed.) Suicide in Different Cultures (Baltimore: University Park Press).

Rin, Hsien (1975). "Suicide in Taiwan" (Province of China), in: N.L. Farberow (ed.) Suicide in Different Cultures (Baltimore: University Park Press).

Ruzicka, L.T. and C.Y. Choi (1993). "Suicide mortality in Australia, 1970-1991" Journal of the Australian Population Association 10(2):101-117.

Stengel, E. (1973). Suicide and Attempted Suicide (Harmondsworth: Penguin Books).

Suk, Jae-Ho (1992). "Suicidal behaviour in Korea", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 41-57.

United Nations (1996). Prevention of Suicide: Guidelines for the Formulation and Implemetation of National Strategies (New York: United Nations Department for Policy Coordination and Sustainable Development) ST/ESA/245.

Wai-Hoi, L. (1992). "Suicidal behaviour in Hong Kong", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 83-111.

WHO (World Health Organization) (1968). Prevention of Suicide. Public Health Papers No. 35 (Geneva: World Health Organization).

__________ (1974). Suicide and Attempted Suicide Public Health Papers No. 58 (Geneva: World Health Organization).

Yoshimatsu, K. (1992). "Suicide behaviour in Japan", in: Kok Lee Peng and Wen-Shing Tseng (eds.) Suicidal Behaviour in the Asia-Pacific Region (Singapore: Singapore University Press): 15-40.

Zimicki, S., L. Nahar, A.M. Sardar and E. D'Souza (1985). Cause of Death Reporting in Matlab: A Source Book of Cause-Specific Mortality Rates, 1975-1981 Scientific Report No. 63 (Dhaka: International Centre for Diarrhoeal Disease Research, Bangladesh).