Ugandan Men’s Perceptions of What Causes and What Prevents Suicide

Birthe Loa Knizek, Eugene Kinyanda, Vicki Owens and Heidi Hjelmeland

Suicidal behavior is illegal in Uganda. There are no reliable public suicide statistics, but studies indicate that the rates of both suicide and nonfatal suicidal behavior are higher for men than for women. This study examined Ugandan men's perceptions of what causes and what prevents suicide as well as their attitudes towards suicide and suicide prevention, including religiosity. Knowledge about this is important as the country is now planning suicide prevention strategies. A majority (26 percent) of men believed that suicide was a response to illness/disease and problems with relationships (24 percent), followed by perceived pressure (10 percent), lack of control (9 percent) and economic hardship (8 percent). These men also endorsed negative attitudes towards suicide, including when suicide is precipitated by an incurable disease. Most men believed that suicide could be prevented via health care services (45 percent) and education (22 percent), while only a few explicitly mentioned an improvement of the socioeconomic conditions. These findings are interpreted in light of the ideological demands of being a man and the socioeconomic reality making this difficult or impossible.

Suicidal behavior is a continuous challenge around the world and the World Health Organization (2001) estimates that about one million people kill themselves every year. However, this challenge has not yet been taken up in all parts of the world. For instance, most parts of Africa lack official suicide preventive initiatives. As in many other developing countries there are no public statistics on suicidal behavior in Uganda and the scope of the problem is thus uncertain. The Support to the Health Sector Strategic Plan Project estimated a 15.5 percent lifetime prevalence of nonfatal suicidal behavior in the 14 districts covered by the study (Kinyanda et al., 2004). However, big differences between the districts were observed, with figures varying from 4.9 percent in Yumbe to 16.1 percent in Adjumani. The authors assumed that underlying ecological factors operating at district level might explain these variations. Other studies have found much higher rates. Bolla (2002) estimated a suicide rate of 99/100 000 and a suicide attempt rate of 518/100,000 in Adjumani [page 5] district. Ovuga et al. (2005) found a suicide rate of 16.7/100,000 in Adjumani District whereas Kinyanda et al. (2009) found a suicide rate of 15-20/100,000 among a post-conflict population in Northern Uganda. From 1986 to 2006 this part of the country was severely affected by civil conflict between Uganda Government armed forces and rebel groups (Dolan, 2009). Underlying ecological factors must therefore be taken into consideration and we would therefore expect differences in suicidal behavior between the conflict area of Adjumani and the capital Kampala, where the informants of the present study come from, differences that might affect men's perceptions and attitudes. In a recent study from Kampala, 8 percent of Ugandan psychology students reported having experienced suicide within their family, 53 percent knew of someone outside their family having killed themselves, and 24 and 61 percent, respectively, knew someone in or outside their family who had engaged in suicidal behavior (Hjelmeland et al., 2008). Other than this, only limited research has recently been published on suicidal behavior in Uganda (Hjelmeland et al., 2006; Hjelmeland et al., 2008; Kinyanda et al., 2004, 2005a, 2005b, 2005c; Ovuga et al., 2005). However, the studies conducted indicate that suicidal behavior is a considerable public health problem within this country. It is, however, impossible to elaborate on trends over time due to the lack of baseline data. Since there are no reliable suicide statistics in Uganda, the exact sex ratio of suicide is unknown. However, based on the studies cited above there is reason to believe that more men than women engage in suicidal behavior in this country. For completed suicide, a male:female ratio of 4.4:1 (Kinyanda et al., 2009)and for nonfatal suicidal behavior 1.7:1 has been reported (Kinyanda et al., 2004).

Despite the growing recognition of suicide as a severe health problem there is a paucity of literature on attitudes towards suicide in Africa in general and Uganda in particular. Only few studies exist on attitudes towards suicide in Africa (e.g., Lester & Adebowale, 2001; Peltzer et al., 2000; Eshun, 2006). These studies are all quantitative, which limits explorations of the decisive social and ideological context (Denzin & Lincoln, 2005; Gergen & Graumann, 1996). Some qualitative studies on attitudes towards suicide in Uganda (Mugisha et al., in press) and Ghana (Osafo et al., in press) are, however, underway. To the best of our knowledge, theoretical reflections on suicide in Africa are non-existing. Being highly context-dependent, it does not seem meaningful to transfer theoretical models developed to fit other cultural settings.

The aim of the present study was to examine Ugandan men's perceptions of what causes and what prevents suicide as well as their attitudes towards suicide and suicide prevention. Knowledge about this is important as the country is now planning suicide prevention strategies and men seem to be a particularly vulnerable group. We decided to study men only in order to focus on the inherent logic in men's beliefs rather than to compare them with women. The social and ideological context for men and women seems to be different (Kinyanda et al., 2005) and needs to be taken into account both for comprehending the internal logic in their understanding and for planning prevention efforts. When action is to be taken, attitudes become central since it is generally assumed that attitudes towards suicide are of great importance. Such attitudes affect the will of people as well as health care staff, helping persons in a suicidal crisis or those who have deliberately harmed themselves (Bagley & Ramsey, 1989). Suicidal behavior is illegal in Uganda. It is also [page 6] considered a bad omen for the clan, necessitating cleansing rituals. Hence, suicide carries enormous stigma (Hjelmeland et al., 2008). Uganda is also considered to be a very religious country (Uganda Bureau of Statistics, 2002). This is important as we know that religious people are considerably more intolerant towards suicide than less religious people (see Koenig et al., 2003, for an overview). Suicide is thus perceived negatively in African countries (Lester & Akande, 1994; Peltzer et al. 1998), which also was found in a study in Uganda and Ghana by Hjelmeland et al. (2008). However, it also has been shown that among religious people there is a greater belief that suicide should be prevented (Bascue et al., 1982).

To understand the men's perceptions and attitudes it also is necessary to look at the social conditions in Uganda since they constitute the framework for their statements. With recent rebel activity, the HIV/AIDS pandemic (which has led to the death of more than 1 million people in more than 20 years of the epidemic), and a large number of deaths caused by malaria annually, Uganda has considerable challenges, not only on the economic and political scene, but also in the daily lives of its population. Ugandans are squeezed by poverty, unemployment, high rates of premature death, and insecurity regarding prospects for the future. According to the World Health Report (WHO, 2001), people in East Africa are some of the poorest in the world. Almost every Ugandan is affected by the situation of family instability and/or poverty and struggles for a decent living. This also affects Ugandan men as many of them have problems in finding adequate jobs and maintaining their traditional position as the breadwinners of the family (Dolan, 2002; Kinyanda et al., 2005). Barker and Ricardo (2005) point at the same problem when they underline that young men perceive multiple and sometimes conflicting ideas about what it means to be a man and generally perceive that they are constantly judged and evaluated for their actions as men. These pressures—arising from the clash of ideologies, Westernization trends, socioeconomic change and the challenges to traditional masculinity—may lead to feelings of humiliation, both in a man's sense of self, as well as in his sense of how he is perceived by others (Dolan, 2002) and might impact on Ugandan men's suicidal behavior and attitudes towards suicide.

Given the complexity of the phenomenon and the fact that little is known about suicide and Ugandan masculinity, we chose to mainly use qualitative methodology and focus only on men. With such an approach we try to handle the problem that theories are often gender biased with male suicide generally perceived as a rational choice and a reaction to external, impersonal factors, whereas female suicide is connected to emotional, personal factors (Canetto, 1997). Whether this is true across cultural settings is an important research question.

Method

Responses to the two open-ended questions: "What is the most important cause of suicide?" and, "What do you think can be done to prevent suicide?" were analyzed qualitatively. Also, some quantitative analyses were conducted in order to illuminate the qualitative analyses further where relevant. This is described in detail below.

Sample

Altogether, 329 men from the following groups participated in the study: medical students (n = 46), psychology students (n = 116), social work students (n = 79), [page 7] nursing students (n = 76) and Psychiatric Clinical Officers (PCOs; n = 12). Their mean age was 24.9 years (SD = 6.3) ranging from 18-66 although the majority were in their 20s. To assure anonymity, only age group is presented with quotations. Eighty (25.3 percent) of the men were living alone, whereas the others were living together with someone in various family constellations. The vast majority of the men were Christian (90 percent; n = 291) whereas 8 percent (n = 26) were Muslim.

Instrument

The two main questions analyzed in this study (see above) were from the Attitudes Towards Suicide questionnaire (ATTS) developed in Sweden by Salander Renberg and Jacobsson (2003). This questionnaire mainly consists of items to be scored quantitatively (for instance, on a 5-point Likert scale). On the two open-ended questions, however, the participants could describe their personal thoughts in their own words, which makes this part of the questionnaire less culturally dependent than the quantitative part. The psychometric properties of the instrument as such are described by Salander Renberg and Jacobsson (2003) and with relevance to Uganda by Hjelmeland et al. (2006).

The responses to the two open-ended questions were analyzed by thematic analysis (Boyatsis, 1998) where the responses were categorized in main categories. The analyses were first conducted by the first author (who is Danish) and a category system was developed. The analysis was then presented to the rest of the research group (one Ugandan, one American who has lived in Uganda for the last 25 years, and one Norwegian) and discussed. If an answer encompassed several causes, each of those was categorized, but none of them was coded twice in different categories. Subsequently, subcategories for each of these main categories were developed, resulting in a branched structure. This branched structure developed into a hierarchical categorization system covering all the responses and reflecting an internal logic. The same method of analysis was employed in previous publications by this research group where similar branched structures were found using data from several European countries (Hjelmeland & Knizek, 2004; Knizek et al., 2008) as well as from Ghana (Knizek et al., in press). Descriptive statistics are presented for the respondents' own suicidal ideation and behavior, their experiences of suicidal behavior in their surroundings, and for some of the attitude items.

Procedure

The vast majority of the participants in this study were students and the questionnaire was handed out in class at all levels of the studies involved. The participants were informed of the study in writing and in person and the voluntary nature of participation was emphasized. Participants were especially requested to seek help from a qualified counselor if the questionnaire caused them any discomfort or if they had some related issues they would like to discuss (contact details were given). None of the participants made such contact during, immediately after, or later following the data collection. The questionnaires were only distributed to those students present in the different classes targeted. The classes chosen for data collection were either those with compulsory attendance or those with a traditionally high attendance rate. There is no reason to believe that there were any systematic differences between those students attending class and [page 8] those who did not on the day of the study, since the study was not announced beforehand. All students handed in the questionnaires. A sample of convenience of PCOs (n = 12) attending work at the psychiatric hospital on the day of the study also filled in the questionnaire.

The English version of the questionnaire was used. English is Uganda's official language and also the language used in schools so the students are very familiar with it. The data were collected in 2002. The study was approved by a research ethics committee in Norway and by the relevant bodies at Makerere University in Uganda.

Analyses and Discussion

Participants' Experience with Suicidal Ideation and Behavior

Thirty-four percent (n = 112) of the participants expressed having had suicidal ideation (3.3 percent often, 14.0 percent sometimes and 16.7 percent hardly ever) and 22.5 percent had made a suicide plan during the last year (4.3 percent often, 8.8 percent sometimes and 9.4 percent hardly ever). Thirty-eight percent (n = 125) had thought about suicide (3.6 percent often, 17.6 percent sometimes and 16.7 percent hardly ever) and 28 percent had made a suicide plan earlier in life (3.3 percent often, 13.4 percent sometimes and 10 percent hardly ever). In an earlier study we found that the Ugandan students expressed more life-weariness last year compared to students in Ghana and Norway, whereas the Ugandan and Norwegian students expressed more life-weariness earlier in life compared to Ghanaian students (Hjelmeland et al., 2008). In the present study 4 percent (n = 12) had engaged in a suicidal act last year, whereas 7 percent (n = 22) had engaged in suicidal behavior earlier in life. Hjelmeland et al. (2008) found that Ghanaian and Ugandan students more often than the Norwegians had made at least one suicide attempt during the last year, but the effect size of this difference was very low. Nine percent (n = 28) of the men had experienced suicide and 21.5 percent (n = 65) had experienced nonfatal suicidal behavior in the family. The corresponding figures for experiences of suicide and nonfatal suicidal behavior among others were and 59.7 percent (n = 187) and 65.7 percent (n = 201), respectively. In Hjelmeland et al. (2008) the Ugandan students had more often experienced both suicides and suicide attempts in their family than the Ghanaian and Norwegian students. The Norwegian students had experienced suicide attempts among others more often than the Ugandan, who in turn had experienced this more often than the Ghanaian. However, the Norwegian and the Ugandan had experienced suicide among others more often than the Ghanaian students.

Description and Discussion of the Thematic Analysis of the Open-ended Questions

The most important cause of suicide. Three hundred and sixteen men responded to the question: "What do you think is the most important cause of suicide?" (45 medical students, 69 nursing students, 112 psychology students 79 social work students, and 12 PCOs). The responses ranged from one word to a few sentences. As each person could give more than one response, there were in total 710 statements of which 660 (93 percent) were specific and 50 (7 percent) were unspecific. An example of an unspecific response was: "Want to stop the pain in the rat race of life" (medical student in his 20s), where no specific cause was presented. Only the [page 9] specific responses were analyzed further. The thematic analysis revealed three main categories of causes: intra-personal, inter-personal and extra-personal causes (Figure 1). We have found the same structure in previous analysis of data from Europe (Hjelmeland & Knizek, 2004; Knizek et al., 2008) and in Ghana (Knizek et al., in press). The majority of the responses (60 percent) fell into the category of intra-personal causes (n = 390) whereas 185 (28 percent) could be categorized as inter-personal, and 85 (13 percent) as extra-personal causes. The subcategories for each of these main categories are presented in the following:

Figure 1: Categories of responses to the question: What is the main cause of suicide? Number of statements in parentheses (only the specific suggestions are included in the figure).

Intra-personal causes. Most of the statements in the group of intra-personal causes pointed towards illness/disease (43 percent) and about a third of these statements concerned mental disorders or problems, but only a quarter of these mentioned depression explicitly. Drugs and alcohol abuse were mentioned 12 times as the most important cause of suicide. However, the majority of statements in the illness/disease category were statements about HIV/AIDS and cancer. The quantitative data showed a negative attitude towards the right to kill oneself even in relation to incurable disease: The mean score on this factor was 2.2 (SD = 0.83) indicating that the men did not agree that people have the right to kill themselves, even in relation to severe, incurable illness. The score on the single item "People should have the right to take their own lives" was 1.9 (SD = 1.2).

Only one man mentioned impotence even though virility and sexual performance is one of the crucial arenas for the social construction of masculinity (Silberschmidt, 2005). Fallers and Fallers (1960) reported impotence as a significant contributor to suicide in a study they undertook in Eastern Uganda. More recently, Kinyanda et al. (2005c) also observed that sexual problems were a significant contributor to repetition of suicidal behavior in a study in urban Uganda. If sexual performance is that important, it is striking that it is not mentioned more often. This could indicate that this arena might have lost its priority or that it is taboo to talk [page 10] about this issue. Another explanation might be that this sample is relatively young and thus has not felt the pressure of this issue yet.

The second largest group within the intra-personal causes was about frustration (n = 31) and stress (n = 34) and could be named "perceived pressure." This category has a lot of similarity with the Shweder et al. (1997) description of emergent causal ontology. This category was one of the causal ontologies of suffering they found in folk psychologies and "rooted in the metaphors of external 'stress', 'pressure', and 'environmental risk factors'" (p. 122). These statements were equally quite unspecific, mainly expressing that "something" must be too much for these people: "The main reason could be severe stress." Hopelessness (n = 29) and loneliness (n = 26) were also mentioned relatively often and together with the statements about low self-esteem (n = 6). This might describe a situation with lack of power/control exemplified by a statement of a young medical student: "People commit suicide because they have been pushed to breaking point. They are in a tight corner up against the wall and see no possible way out of that corner (can be a problem or anything). They've tried all possible routes and failed."

When a person is entrapped in a situation like this, he either must find a way out or accept the situation through a different way of coping. Since Uganda is a religious country (Uganda Bureau of Statistics, 2002) spiritual or religious coping might be a possibility. Geyer and Baumeister (2005) emphasize how religion and morality in general as a central and explicit goal have to restrain the self and override people's tendency to act out of self-interested motives. This means that even though the short-sighted goal is to put an end to life's misery, religion here might contribute to the individual's ability to endure. Faith thus could provide a possibility to accept a difficult situation as has been shown in a range of previous studies (see Klaassen et al., 2006, for references).

Existential and spiritual causes were also mentioned (10 percent) as illustrated by this statement from a medical student in his 20s: "People commit suicide because of the meaninglessness of life at that particular instant in their lives." Life seems meaningless, maybe as a result of extraordinary life circumstances such as death of loved ones, as some men suggested. The significance of reflecting on the meaning of life for the men in this study was also underlined by the quantitative analysis. On the question "How often have you thought of the meaning of your life?" the mean score was 1.4 (SD = 0.7) for last year and 1.9 for earlier in life (scored on a 4-point scale from 1 = often to 4 = never) indicating a clear tendency of existential reasoning. For some, this existential reasoning seemed intertwined with spiritual reflections; nine men explicitly mentioned loss of faith as a possible cause for suicide because then all problems seem to be overwhelming, as there is no hope for a better future, neither in this life nor in the after-life: "Frustrations accompanied with little trust in Jesus Christ the Saviour of all" (psychology student in his 30s). Here, faith is believed to give hope under unbearable circumstances; confer the concept of existential coping mentioned above (Wong & MacDonald, 2002). The meaninglessness of life might be intensified by not being a believer as this coping mechanism will not be accessible: "Everybody gets into a difficult situation, but the question is what value do they attach to life. And greater than this question DO THEY FEAR GOD. If one fears God, then even if he should think of committing suicide, he may have to fear to even attempt" (medical student in his 20s). However, fear of [page 11] what happens after life here seems to be more important in preventing suicide than religion's ability providing hope and thus giving power to deal with life's adversities. This emphasis on negative reinforcement and prohibition as a suicide prevention strategy can perhaps be described as a moral of justice (Kohlberg, 1983) regulated by both spiritual and secular legislation. Suicide being a criminal offence in Uganda already defines it as an act violating rules and thus the welfare/good of the community that will be punished here in life; seeing it as a sin adds a spiritual offence as God is considered to be the only one that is entitled to give and take life. As many men put it, God is the "author of life" and thus to take one's life is to compete with God, an unforgivable sin that will be punished after death. While a secular person would be able to escape this world's atrocity through death, this is no option for the religious person, who believes in an afterlife with divine legislation.

However, aside from devastating circumstances, cognitive coping can also be problematic, not only by lack of faith in superior powers, but also by having "wrong" attitudes, which limits cognitive coping. Thirty statements (8 percent) pointed at the cause for suicide being a wrong attitude: "Most people fail to comprehend situations so they are forced to think that life with challenges is the wrong life. Yet the ideal life of no problems practically is not there" (medical student in his 20s). The statements in this group are quite different, but point to a cognitive inability to see an actual situation in a wider perspective or to lack creativity/flexibility in looking for solutions. This might be what has been described as tunnel vision as a characteristic of a suicidal process (Shneidman, 1985). There is in this category a wide connotative difference in the statements; part of them are expressions with a judgmental valor blaming individuals engaging in suicidal behavior, whereas others are mere descriptions of an ongoing process. Fifteen statements in this group point to an underlying egoism or weakness and pull in the direction of judgmental opinions. It was mentioned that a cause for suicide could be "personal satisfaction" (PCO, in his 40s), which indicates that the individual only pursues a selfish goal of peace for himself instead of fulfilling the obligations towards others in an ethics of community (Miller, 2007). This would be in contrast with what is expected in a society where people in general and men especially have many heavy and prescribed obligations arising from the actual economic situation and the numerous demands of the extended family system.

More surprising, however, were other statements like "to be remembered as a hero" (psychology student in his 30s), or "martyrdom; heroism" (social work student in his 20s). Especially the notion of heroism seems to contradict the relatively negative attitude endorsed in this study in general. These statements might reflect a notion of suicide as an act of self-sacrifice for a higher cause and thus not as selfish but rather as the opposite: a subordination of individual wishes to societal ones thus again indicating an underlying ethics of community. This is in keeping with Durkheim's (1981) "altruistic suicide." The notion of heroism also seems to fit into an ideology of masculinity that Canetto (1992-93) has discussed in relation to a Western population:

The low number of statements going in this direction could indicate that this myth, although existing, is not widely accepted among our informants and mainly could be interpreted as ironical or sarcastic statements; and therefore are categorized under wrong attitudes.

Some (5 percent) pointed to failed expectations as a possible cause for suicide. This was most often not further explicated; they just pointed out that goals in life were not achieved. Sometimes these failed expectations were coupled with disappointment being a consequence. These failed expectations might be another expression of perceived pressure that they cannot fulfill and an acknowledgement of the overwhelming duties/obligations that people (men) have.

The last, small sub-category under intra-personal causes was about a genetic or hereditary disposition to suicide (n = 4).

Inter-personal causes. From the 185 statements under the inter-personal causes, 87 percent dealt with disharmonic relationships. Problems in marriage, with family or friends as well as broken hearts, were frequently mentioned.

Lack of support was an issue in 37 statements, whereas betrayal was mentioned specifically in nine. Also, the loss of a partner or family members was emphasized through 37 statements. The Ugandan men seem to recognize the crucial meaning of having close relationships for coping with the challenges of life, which is in line with what Canetto and Lester (2002) pointed out in their study on suicide notes, saying that evidence suggests that interpersonal events are dominant precipitants of both female and male suicidal behavior, in contrast to the common gender myth mentioned above.

One sub-category encompassing 12 statements (7 percent) was about suicide as acts to influence someone. The individual tries to punish others, take revenge or simply seeks attention: "Some commit suicide to pay back dear one who have hurt them because they believe those people will never forgive themselves for causing the death" (young medical student), and, "It could be because they want to punish people who have hurt them in life: For example, a man may commit suicide to punish his unfaithful wife" (social work student in his 20s). These statements are surprising in so far as they illustrate individuals without power, which is not what is expected from men in a patriarchal society. Especially astonishing is the example of suicide as punishment of a wife in this society where polygamy and marital violence are frequent (Koenig et al., 2003; Kishor and Johnson, 2006). Among the Banyoro of south-western Uganda, Beattie (1960) describes the phenomenon of "vengeance suicide" where an aggrieved person of usually "less social status" exerts revenge on a "socially superior" antagonist through suicide.

The last group mentioned under inter-personal causes was to avoid shame (7 percent): "Because they are fed up with the wrong they have done because they don't want to be ashamed maybe, or they think the punishment is equal to be dead" (nursing student in his 20s). But also the combination of shame and fear was mentioned in relation to specific acts: "Fear of being caught as criminal especially [page 13] those who murder their wives" (psychology student in his 20s), or, "when they are totally frustrated or embarrassed by their acts e.g. when one has raped, or defiled a young baby" (social work student in his 20s).

Draguns and Tanaka-Matsumi (2003) have pointed at shame being a more powerful emotion in interdependent cultures than in independent ones and this could explain why men see this as an important cause for suicide. However, it may be necessary to further analyze these feelings of shame in future research in order to tease out meaning and identify possible gender-specific connotations.

Extra-personal causes. Of the 85 statements on extra-personal causes that were mentioned, the vast majority (65 percent) were on economic reasons where especially poverty was mentioned. Sixteen statements were on work-related issues as, for example, loss of job. Only 6 percent of statements dealt with the recent war that has been in Uganda, while 8 percent pointed at politics and norms as a possible cause of suicide: "political crisis where torture and detention are involved" (psychology student in his 20s), or "political harassments made onto people like people were witch hunted during recent elections because of supporting reform" (social work student in his 20s). The politics that are pointed at here are mainly oppressive and the persecution of people seems to be anticipated as a possible cause of suicide.

Only two statements were explicitly on extra-natural causes or superstition: "When one is charmed or due to superstition" (social work student in his 20s), or: "Can't rule out witchcraft!" (social work student in his 20s). It is positive that there were only these two statements that explicitly granted witchcraft some power in a country where such beliefs seem to be common (Wlodarczyk, 2004).

Perception of Suicide Prevention

Three hundred and fifteen men had responded to the question: "What do you think should be done to prevent suicide?" (45 medical students, 68 nursing students, 110 psychology students, 80 social work students and 12 PCOs). Only one medical student thought that suicide was not preventable: "I think talking to a person who has decided to commit suicide is a waste of time as the person will not change their mind." Three hundred and fourteen men thus believed that suicide could be prevented. This positive attitude was also supported by quantitative analyses where we found a mean score of 4.5 (SD = 0.82) on the item "Suicide can be prevented" and 4.4 (SD = 1.0) on the item: "You can always help a person with suicidal thoughts." As one person could offer more than one response, there were a total of 465 statements of which 429 were specific suggestions on what could be done. An example of an unspecific statement was "Solve any problem without committing suicide" (medical student in his 30s). Only the specific statements were analyzed. Again the specific efforts suggested could be divided in three sub-categories, namely "Efforts at the individual level," "Efforts at the interpersonal level" and "Efforts at the societal level" (Figure 2). The same categories have also been found in Europe (Hjelmeland & Knizek, 2004; Knizek et al., 2008) and in Ghana (Knizek et al., in press). However, whereas the most important causes pointed at were at the individual level, followed by the interpersonal and then the extra-personal level, the most important preventive efforts were assumed to be at the societal level (n = 393), followed by the individual level (n = 25) and the interpersonal level (n = 11). [page 14]

Figure 2: Categories of responses to the question: What can be done to prevent suicide? Number of statements in parentheses (only the specific statements are included in the figure).

The "Efforts at the societal level" could be divided into structural (n = 265) and ideological (n = 128) changes. Structural changes dealt with the improvement of the health services, general socio/economic systems, the legal system and improvement of counseling (religious) services. The ideological changes focused on education, attitude changes and media attention.

Under the structural changes the vast majority (73 percent) implied that improvements in, and access to, health services were central: "Even desperate people need serious counseling and help on how to deal with their serious problems and hardships" (medical student in his 20s). Both mental and somatic health services were mentioned. This makes good sense with the men's frequent mentioning of HIV/AIDS or cancer, along with mental health problems as the most important cause of suicide. However, counseling was not always described as support and could have some rather scary aspects: "For suicidal deaths there needs involvement organisation to provide free counselling to the people about suicide, and this should be emphasised with a lot or strong threats and outcomes or solutions (possibles)" (psychology student in his 20s). This statement follows a line of several from different areas in this study, namely to scare and threaten people, either with earthly or heavenly punishment, or to think twice and find a different solution: "Scaring people about suicide's obvious outcomes, death" (medical student in his 20s), or, "Sensitising people about the badness of committing suicide. The Government is doing it by publicly beating any person who commit suicide so as to discourage others from doing the same" (psychology student in his 30s). Some of the statements thus widen the concept of counseling by introducing threats. This is in line with 27 statements on improvements in religious (counseling) services that are also pointing at threats as an important prevention strategy: "People should be told [page 15] that even God will punish them for killing themselves" (social work student in his 20s), or: "Community counselling methods could be started to sensitize people against this immunal and evil act. Religious leaders could their teaching against it. National condemnation through inhuman treatment of dead bodies for suicide" (social work student in his 20s). The men here point at the consciousness of divine legislation both as an amplifier of the governmental legislation and the divine morality as a booster of the morality of community. This way of moral reasoning fits with Verhoef and Michel (1997, p. 405) who have shown that "African morality has quite a distinctly different orientation from the morality developed within western tradition in general." While morality in the West mainly is regarded as highly individualistic and distinct from religion, it seems as if the Western definitions are unable to capture the specific African mixture of spiritual and moral factors or the strong bonds between the individual and his social context. As Mbiti (2006, p. 2) puts it: "There is no separation between concerns of a religious and philosophical nature, but complex interdependence." We find this general relatedness of morality and religion as an ongoing theme in the responses of these men.

Only a minority of the responses (10 percent) emphasized the need for economic and social change, a surprisingly low number taking into account the socio-economic challenges of Uganda. This again underpins the observation that these men seem to be more occupied by the inter- and intra-human consequences than the socio-economic misery. Fifteen statements focused on the legal system: "In case one attempts to do so & fails, should be eliminated from society" (psychology student in his 30s), or: "Involving everybody in the legal prevention system. Serious punitive measure for the attempts" (psychology student in his 40s). The men emphasizing the law did not mention a decriminalization of suicide, but advocated for strong punishment instead. Threats and punishment seem thus to have a prominent status in the prevention strategies according to our sample. Only two statements pointed to the necessity of research on the issue.

The need for education and sensitization was cited in almost three quarters of the responses. Twenty two statements (17 percent) dealt with necessary attitude changes and 10 statements (11 percent) focused on the important role of the media. Under the ideological changes we again see the tendency of negative reinforcement that comes up in all sub-categories as a means of education: "Mass awareness about the evil of suicide" (social work student in his 20s). Education and sensitization are thus often suggested as means of installing or enhancing both communal and divine morality in people, which again mirrors the intimate relatedness of morality and religion.

The efforts at the individual level could be described as a kind of advice to people to be more open (n = 17), more positive (n = 5), to adjust to the circumstances (n = 2) and avoid drugs (n = 1). Efforts at the interpersonal level were statements about everybody's responsibility to help (n = 9) and to give support (n = 2), which again are to be seen in the frame of communal morality, but also in light of the religious demands of caring for your fellow human beings. The low number of statements in this sub-category seems somewhat surprising taking into consideration the attention these men directed towards interpersonal problems as one of the most important causes for suicide. [page 16]

Methodological Considerations

This study mainly focused on a qualitative analysis of responses to two open-ended questions at the end of a mainly quantitative questionnaire where some of the quantitative variables were used to illuminate the qualitative analyses. The quantitative part thus may have colored the answers of the open-ended questions. Another limitation is that our sample consists of relatively well-educated men who presumably have more knowledge of health related issues than other groups. Also the pressure on these men might be different than in districts with high suicide rates. We therefore suggest that in-depth interviews with men from different districts in Uganda should be conducted to get a better overview on men's perceptions and attitudes.

General Discussion

The results of this study reflect a rather negative attitude towards suicide, even in relation to incurable disease, and a strong belief in the possibility to prevent suicide. A negative attitude towards suicide is generally assumed to be predominant in the developing world (Eshun, 2003; Lester & Akande, 1994). It is necessary to understand this against the background of the deep impact of religion onto the lives of these people: "Africans are notoriously religious, and each people has its own religious system with a set of beliefs and practices. Religion permeates into all the departments of life so fully that it is not easy or possible always to isolate it" (Mbiti, 2006, p. 1). This means that "where the individual is, there is his religion, for he is a religious being. It is this that makes Africans so religious: religion is in their whole system of being (ibid, p. 3). As previously shown, religious people are considerably more intolerant towards suicide than less religious people, but it also has been shown that in this group is a greater belief that suicide should be prevented. Our findings thus are in line with this. However, in order to understand the negative attitude as more than just an effect of religiosity it might be relevant to look at the "Big Three" theories of ethics (Shweder et al., 1997) where three broad types of moral orientations are found. The first focuses on autonomy and concerns harm, rights and justice (Kohlberg, 1983). The second concerns responsiveness to the needs of others and can be described as an ethics of community (Gilligan, 1982; Miller, 2007). The third is based on cultural approaches and focuses on the influence of the divine and deals with issues like sin, purity, and so on (Miller, 2007). From the autonomy perspective, the negative attitude of the men can be understood as a conventional reasoning, where the outset is that suicide is a criminal offence (which it is by law in Uganda) and violates justice. From the community perspective, suicide must be seen in relationship to an intricate social structure with an extended family system that in times of hardship (economic, war, AIDS, etc.) means an even heavier load on the shoulders of each individual. Suicide thus must be perceived as a betrayal of others and therefore perceived negatively. Given the high impact of religion on Ugandans (Uganda Bureau of Statistics, 2002, Gifford, 1999), the ethics of divinity also pull towards the negative direction as suicide is understood as a way of competing with God and thus is considered a sin. The overall negative attitude towards suicide thus might be plausible. Still, one third of the participants reported suicidal ideation, and four percent admitted having engaged in suicidal behavior during the last year, and seven percent earlier in life. The majority had experienced [page 17] suicidal behavior in their surroundings. As causes for suicide they mainly pointed at illness/disease and problems with relationships, followed by perceived pressure and economic hardship. In order to understand this, it seems fruitful to see the burden of being a man in an area under stress in addition to the different underlying moral/religious reflections and demands. Here the problem of not being able to fulfill internal and external expectations (Dolan, 2002) comes in, as masculinity in East Africa is supposed to be linked to dignity and self-control (Silberschmidt, 2005):

In a situation where almost all Ugandans are affected by poverty or premature death because of war or disease, these expectations might be difficult or even impossible to meet and thus result in relatively high suicidal behavior even though their attitudes towards it are negative. Or as Dolan (2007) puts it, the "high level of suicide amongst men who cannot attain the masculine norm of marriage, procreation and protection" (p.6). However, these students might be regarded as a special group rather than as representatives for Ugandan men in general, being more educated and having better prospects for their future. Even though this group might be regarded as a privileged group, more than one third of them admitted to often having had suicidal ideation. This calls for concern and targeted preventive measures.

When we turn to prevention, most of the men assumed that suicide can be prevented. Here it is very clearly indicated that they believe the problems must be solved mainly at the societal level, that is, outside the potentially suicidal individuals. Whereas illness/disease was the most often mentioned cause, in terms of prevention, the men point at improvement of the health services as the most effective way to prevent suicide. However, they also point at the necessity of education; that is, giving people some tools to reflect on the situation or to earn a living in the Ugandan context. The main impression is that the men in their suggestions for prevention seem to underline efforts that can restore people's ability to provide for themselves and their dependents. The emphasis on threats as a powerful counseling and educational tool is, however, surprising and needs further studies.

References