36 Suicide in military settings: Combatants and veterans

Suicide rates in military units are lower than in civilian populations and differ considerably from nation to nation. The processes that may influence suicide rates within the armed forces can be different from that of the civilian life, especially when armed forces are under reformation, downsizing and economic pressure.

Risk factors vary between groups and settings, such as active duty versus reservist/veteran or war versus peacekeeping mission. However, common risk factors are: easy access to firearms, exposure to traumatic stress, lack of social support, and the military life style with frequent relocations. Two subgroups are considered as equally important to target for suicide prevention: young conscripts and war veterans. In the first case, screening and crisis intervention are in focus, and in the second, treatments for post-traumatic stress syndrome, depression and substance abuse. Leadership interventions and changes in firearm regulations are other preventive measures.

Combatants have committed suicide throughout the history of armed conflicts. Soldiers sometimes killed themselves in order to avoid capture and slavery, commanders would do so rather than accept defeat. Officers have committed suicide to avoid revealing secrets under interrogation or torture. Today, armies are usually not involved in large battles, and the nature of conflicts has changed a lot, e.g. guerilla warfare. The military is often regarded as a profession, not only for males, since the number of females in service is growing, especially in the Western world. When suicide rates in the military are compared with those of other professional groups, it is found that they are generally lower in the military than in civilian populations of the same gender and age (Wasserman 1992). This is usually seen as a so called ‘healthy worker’ effect—a result of ruling out a variety of psychiatric conditions on the stage of admittance to the army, as well as ongoing control throughout the appointment. However, suicide rates in military units may differ considerably from nation to nation and in different types of military units. In this chapter, we will discuss results and experiences from recent studies addressing the problem of suicide in the military and provide some recommendations for future preventive strategies and policies.

Recently, we have tried to outline some common features of suicide within the military environment (Rozanov et al. 2002). The following issues may be associated with an increased risk of suicide in the military context:

In spite of the existence of some risk factors specific to the military environment that must be added to risk factors in the general population, suicide rates in military populations remain, in most cases, lower than the civilian population of men of the same age. This could possibly be attributed to the existence of protective factors that may balance the situation. We shall discuss these factors later, after a short review of epidemiological situation data on suicide from military populations in different parts of the world in diverse cultural and socio-economic contexts.

When addressing military systems in different countries, it is necessary to take into consideration the military forces' level of involvement in various types of operations and actions, and the result of such actions. In view of this, the situation of the US Army is distinctive.

Regardless of involvement in military actions, for members of the US Army, suicide is the second leading cause of death after accidents, unintentional injuries or combat loss (Ritchie et al. 2003). This is distinctly different from the general population, where suicide is the second or third leading cause of death in the age group 15–34 years; and in older groups, it is the fifth–sixth leading cause of death (National Strategy for Suicide Prevention 2001). Suicide rates in the US Army declined from 14.8 per 100,000 in 1995 to 9.1 in 2001. With deployment to several missions in recent years (Iraq 1991, Afghanistan, Iraq 2003) large military contingents have been exposed to combat stress, as well as to complicated geographical, climatic and unfamiliar sociocultural situations. While the Iraq 1991 war and Afghanistan missions did not seem to cause any major change in suicide rates in the US Army, the painful Iraq 2003 war seems to have had a distinct impact on these rates. In a review of soldier suicides (2003), the Mental Health Advisory Team reports that the suicide rate for soldiers deployed to Operation Iraqi Freedom II (OIF-II, starting in mid-March 2003) from January to October 2003 was higher than recent Army historical rates (15.6 per 100,000 compared to the average annual rate of 11.9 for the period 1995–2002). Compared to historical army suicide rates, the OIF-II suicide rates were higher for active component male and female soldiers and lower compared to reserve component soldiers. Firearms were the predominant method of suicide. The majority of suicides were committed by young males (Annex D 2003).

In another extensive document, focusing on the mental health and well-being of soldiers experiencing numerous combat stressors, acute and post-traumatic stress is the top mental health concern (Walter Reed Army Institute of Research 2005). The extent of the problem may be seen from the following figures: a survey of 1700 soldiers and marines serving in Iraq since 2003 revealed that 94.5 per cent saw dead bodies or human remains, 92 per cent reported being attacked or ambushed, 86.2 per cent knew someone killed or seriously wounded, and 55.7 per cent caused the death of an enemy combatant (Walter Reed Army Institute of Research 2005). The percentage of study subjects whose responses met the screening criteria for major depression, generalised anxiety or post-traumatic stress disorder (PTSD) was significantly higher after duty in Iraq (15.6 to 17.1 percent) than after duty in Afghanistan (11.2 percent) or before deployment to Iraq (9.3 percent); the largest difference was in the rate of PTSD. For example, PTSD (strict definition) was found in approximately 12.5 per cent of the personnel in army and marine contingents (Hoge et al. 2004), while the estimated lifetime prevalence of PTSD among adult Americans is 7.8 per cent. Women (10.4 per cent) are twice as likely as men (5 per cent) to have PTSD at some point in their lives (Kessler et al. 1995). Knowing the link between PTSD, depression and suicide, there is no surprise that suicides are increasing (Mehlum 2005). On the other hand, those who are in the army still demonstrate high levels of adaptation, supported by the following data.

Despite the fact that personnel having served in the US Army in Iraq have higher suicide rates than the average US Army rates, both the Iraq contingent rate and the overall army rate remain below the civilian male population (13.5 per 100,000 troops compared with 17.5 per 100.000 of population or approximately 25 per 100,000 for men aged 20–55) (National Strategy for Suicide Prevention 2001; Nelson 2004). Lower suicide rates in the army have been reported for decades (Rothberg et al. 1990; Senteil et al. 1997; Nelson 2004). Most recent reports confirm that military rates for the period 1990 to 2000 were approximately 20 per cent lower than the civilian rate (Eaton et al. 2006). On the other hand, those who had been in the army (veterans) have almost twice the suicide risk of non-veterans in the general population, as reported by Kaplan and colleagues (2007). In addition, some authors argue that some of the suicides in the army remain underestimated, the classification errors may account for about 21 per cent of additional suicides (Carr et al. 2004). In general, among the 1.4 million active duty US military service members, 6 per cent receive outpatient treatment for mental health disorders each year (Hoge et al. 2003).

In every army, there may be differences in suicide rates in different types of forces, which may depend on exposure to stress, traditions, policies and other factors. This is also the case in the US Army, and official sources provide extensive information regarding this subject. Regular suicide statistics from the US military testify that suicide rates in the Navy are the lowest, followed by the Air Force, Army and Marine Corps, which are at highest risk according to Department of Defence statistics (Allen et al. 2005). In the Air Force, suicide accounted for 23 per cent of all deaths and was the second leading cause of death (MMWR 1999). If speaking only about recruits, from 276 recruit deaths in the US military from 1997 through 2001, 28 per cent (77 deaths) were classified as traumatic (suicide, unintentional injury, homicide), and after age-adjustment, traumatic death rates were highest in the army (four times higher than in the Navy and Air Force, and 80 per cent higher than in the Marine Corps). The majority (60 per cent) of traumatic deaths was due to suicide, followed by unintentional injuries (35 per cent), and homicide (5 per cent) (Scoville et al. 2004).

Sociocultural, economical and other societal peculiarities, as well as structural particularities, traditions and policies may influence such phenomenon as suicide in the military. Several reports were published recently in France (Desjeux et al. 2001, 2004). In a short review covering suicides and suicide attempts in the French armed forces during the course of the year 1998, the authors found a very typical pattern: from 145 records surveyed, 40 were suicides and 105 suicide attempts, suicide cases were exclusively males, average age 36, while attempts were distributed evenly between males and females, average age 30. In cases of completed suicide, main methods were firearms and hanging, while in attempted suicides—drug overdose and self-cuttings—previous suicide attempts were found in 21 per cent of attempters and 10 per cent of completers. The rate of completed suicides was reported, to be 14 per 100,000 of military troops (compared with the male French general population level, about 22 per 100,000) (Desjeux et al. 2001; European Health for all Database 2008). In a later report, the same authors have analysed the situation from the year 1997 to 2000 inclusive. During that period, 230 suicides occurred among 315,934 persons, making the overall annual suicide rate 18.2 per 100,000 of active duty personnel. In comparison to national rates for men of similar age categories, the rate in the army is lower. The main suicide methods used were firearms (51 per cent) and hanging (28 per cent). The incidence rate in the gendarmerie, a military body in charge of police duties among civilians, was twice as high as in the land forces. Men under 25 and aged 40–44 were at the highest risk (Desjeux et al. 2004). The overall conclusion of these reports is that despite global lower risk than general populations, gendarmerie personnel and younger people need specific surveillance measures.

Micklewright recently reported on deliberate self-harm in personnel of the Royal Navy (Micklewright 2005) and concluded that these acts should be viewed in the context of the environment that often imposes psychological, emotional and social pressures on servicemen. In the UK, detailed information on suicide in the regular Armed Forces are published by the Defence Analytical Services Agency (DASA). The report from 1984 to 2006 includes a comparison with the UK general population. For the 23-year period, suicide rates in the Army ranged from 12 to 20 per 100,000, suicide rates in the naval service ranged from 6 to 14, and from 3 to 15 per 100,000 in the Royal Air Force. Rates were rather high in the end of 1980s and first half of the 1990s, and from 1999, rates are decreasing in all of the armed forces (DASA 2007). Overall, suicide rates in the armed forces were lower than the general population (around 18 per 100,000 males of all ages) and changed over time in a very similar pattern, as the rates in the general population. However, there is an exception—young army members, aged 20 years and under showed periodically higher suicide rates than their civilian counterparts (DASA 2007).

In an Irish defence forces study (Mahon et al. 2005) of all regular duty personnel for the period from 1970 to 2002, the average annual suicide rate was found to be 15.3 per 100,000 (rates in the general population for men in this period in Ireland increased from 5 to 22 per 100,000) (European Health for all Database 2008). Firearm suicides accounted for 53 per cent of all cases. A history of previous episodes of deliberate self-harm, morning duty and a recent medical downgrading were identified as independent risk factors predicting suicide in the military (Mahon et al. 2005).

Several studies on risk factors for suicidal behaviour among military personnel have been conducted in Norway. Engelstad's (1968) and Hytten's (1985) epidemiological studies document a fourfold increase in suicide rates of young soldiers—reaching 13.6/100,000) during the period 1977–1984. This was, however, still considerably lower than in the general population of young males (28.1) in the corresponding time period. Clinical and descriptive studies from Norway (Mehlum 1990, 1992, 1994, 1998) revealed that suicidal behaviour in young soldiers was in many cases impulsive, and about one third of the suicide attempts were made under the influence of alcohol (Mehlum 1990). In cases of completed suicide, the self-destructive method was very often the use of firearms (Hytten 1985), whereas suicide attempters usually ingested drugs or poisons and/or cut their wrists (Mehlum 1990). Similar results were obtained in Finland, where suicide in the military is also lower than the general population. One of the major triggering factors for suicide among soldiers was situational stress, and alcohol was identified as a factor in many suicides (Marttunen et al. 1997).

Several reports have been published recently regarding suicide in the Italian military environment (Mancinelli et al. 2001, 2003). For the period 1986 to 1998, the authors have revealed 122 suicides and 136 suicide attempts, subjects age ranged from 17 to 60 years, and the most frequent ages both for suicides and suicide attempts were 19–22 years. The predominance of suicides in the Italian military was also found to be lower than in the general population. Authors attribute these figures to existing screening procedures of military personnel, excluding the mentally disturbed at an early age, and the development of a positive feeling of belonging to a group among young soldiers.

The military cannot be separated from the rest of society, and many internal problems in the military may originate from current social problems in the surrounding larger society. With this in mind, it is interesting to look at the situation in the armies of the Russian Federation and the Ukraine. These two post-Soviet countries have high suicide rates, reflected also in the military setting. A report by Litvintsev et al. (2003) shows that suicide rates in the Russian Federation Army increased from 14.0 to 32.1/100,000 in the period 1993 to 1999. Suicide rates in the military, as in other countries, were well below the rates in the general population of males during the same period (increased from 59.3 to 74.3). On the other hand, changes in rates over time had their own peculiarities. It is well known that in most of the former USSR countries, since the perestroika in 1986, there was a sharp reduction in suicide rates until 1991, when a dramatic rise occurred (Värnik 1997; Wasserman et al. 1998) peaking in 1995–1996. After this, a slow and steady decrease was observed until the present (Rozanov 2007). These changes, especially shortly after the perestroika, were explained by the lowering of alcohol consumption due to a strong anti-alcohol campaign started by Gorbachev at that time (Värnik 1997; Wasserman et al. 1998). However, while general population suicide rates in the Russian Federation were reduced from 42.4 to 35.5 per 100,000 during the period 1994 to 1998, suicide rates in the army showed the exact opposite ten-dency, increasing from 14 to 32–30 during the period 1994 to 1997–1998, and substantially declining in 1999 (Litvintsev et al. 2003).

This tendency can be attributed to specific risk factors within the army, and especially to financial limitations, low salaries and poor social protection of those who were leaving the army, especially officers and warrants. Russian specialists consider socio-economic factors and the lowering of prestige of the military service to be the main explanatory factors for the relatively high suicide rates in the Russian Federation army as compared with armies of Western countries (Litvintsev et al. 2003). The authors state that 80 per cent of those who committed suicide had no diagnosis of severe psychiatric disorder, whereas 60 per cent seemed to have stress-related disturbances, and in most cases, unfavourable social and psychological factors were present: 65.8 per cent of the suicides were committed by soldiers and sergeants drafted on a conscription basis (two-thirds of suicides occurred during the first year of service). Altogether, 16.2 per cent of the suicides were committed by officers, 9.8 per cent by ensigns, and 8.2 per cent by other categories (Litvintsev et al. 2003).

A comparison of mental problems in a group of military officers who were engaged on a contract basis and completed suicide, with a sample of suicide completers from the general population, gave the following results: among 10 per cent of the officers, mental disorders were present, whereas 15–25 per cent were reported as having mental disorders in the general population; in 20–25 per cent of the officers, borderline neurotic disorders were found, compared to 35–40 per cent in the general population; and about 70 per cent of the officers were without any disorder compared to the drastically lower figure of 40–50 per cent in the general population (Litvintsev et al. 2002). In the army of the Russian Federation, suicides account for one-third to one-fifth of all deaths. Regarding different kinds of forces, the navy personnel were at biggest risk, followed by the air force, strategic missile forces and landing forces. The predominate methods of suicide were those with firearms as well as hanging, and in about 16 per cent of suicides, a farewell letter was found (Litvintsev et al. 2001a, b).

In the navy, the time course of suicides also differed distinctly from the general population: while in the whole USSR there was lowering of suicides in 1986 (see above for the suicide preventive effect of the perestroika), in the navy there was a rise in the percentage of suicides in general death structure during the period 1986 to 1995 (period of serious economical problems, fleet downsizing, lowering of the prestige of the marine professions). Only from 1998 to 2000 did the percentage of suicides start to diminish. In the navy, two-thirds of all suicides occur among conscripted personnel during their first year of service. In 65 per cent of the cases, the method of suicide was hanging, 20 per cent firearms, 5.5 per cent intoxications, 5.2 per cent self-cutting, 2.2 per cent jumping and finally, 1.8 per cent drowning (Sharaevskiy et al. 2002).

In the Russian Federation army, contingents in the Northern Caucasus region have been involved in the conflict in Chechnya, and the situation among these troops gives us an idea of how the involvement in action affects suicides and mental health in general. In an observation by Bogachenko et al. (2003) of 4953 soldiers in the Northern Caucasus military district, mental health instability was found in about 10.5 per cent of the cases. In 1999, 732 persons were relieved due to mental health problems, 457 of these had once attempted suicide, in 68.8 per cent of the cases, suicide attempts were provoked by interpersonal conflicts due to bullying and harassment. More than half of all suicide attempts occurred during the first year of service. These results are confirmed in another paper analysing soldiers with neurotic reactions that were admitted to the hospital. It was found that neurotic reactions in soldiers during two year periods of service had several distinct exacerbations in the third, sixth, twelfth and eighteenth months of service; suicidal behaviour showed the same temporal profile. The authors suggest an explanation for each peak of neurotic reactions, starting with poor adaptation to the military environment, with later shifts towards interpersonal conflicts, and a combination of poor adaptation, conflicts and family problems. It is important to note that suicidal behaviours have the same time profile: the general tendency is a lowering of risk with time, with the last 4–5 months of service remaining almost free of suicidal tendencies (Fadeev et al. 2001).

Many authors from the Russian Federation underline bullying as an important stressful or triggering event in the suicidal behaviour of the soldiers. An important factor is the process of pre-enrolment screening of the conscripts, which often fails to prevent the conscription of soldiers with existing suicidal tendencies.

The situation in the Ukrainian army is rather similar to that of the Russian Federation. Army suicide rates were once reported to be alarmingly high (340 per 100,000 in 1996), though it seems to be a result of unchecked information or short-term evaluations (Chuprikov et al. 1998). Later studies have revealed that military suicide rates, as in many other countries, are below the rates for men in the general population (46.7 for men of all ages, 52.5 for men aged 20–55, average for 1991–1999) and in a given military unit (the air forces, unit comprising almost half of the air forces rank and file) constituted 32.6 per 100,000 (average for 1991–1999) (Rozanov et al. 2002). The most at-risk categories of military in this unit appeared to be soldiers by conscription and warrants (ensigns), whilst officers were at lowest risk. In a study evaluating 66 cases of completed suicides of the soldiers by conscription, it was registered that in 42.5 per cent of the cases, the method of suicide was hanging, and in 46.9 per cent of the cases, firearms were chosen (Gichun 2000). In Ukrainian resources, suicides in the militia is much more extensively studied than those in the regular army. This contingent (which resembles the military very much) also has lower suicide rates than the general population (20 per 100,000 as compared with 28–30 per 100,000 in 1995–1998). The period of the most alarming rise in suicides during the last two decades was from 1990 to 1998, after which a lowering of the number of suicides has been observed (Chorny 2001). The change in rates was highly similar to that observed in the general population. Alcohol consumption was associated with about half of the suicides: in 63 per cent of the cases, the method used was hanging, and in 25 per cent, firearms (Chorny 2002).

Many authors pay special attention to the screening process of conscripts, especially on conscription occasions. In every country, screening instruments may vary, but are generally based on psychological testing (cognitive style, psychological performance) and clinical interviews. Of course, screening procedures for risk of suicidal behaviour need to be introduced with caution in such settings where there could be a real risk of suicide contagion. In these circumstances, it seems advisable to screen explicitly for depressive symptoms (including suicidal ideation and hopelessness), alcohol and drug misuse, and signs of reduced coping. In the search for further possible dimensions to utilize in screening procedures, several authors have studied the Sense of Coherence (SOC), a dimension developed by Antonovsky (Antonovsky 1993), which is potentially very important in military units and companies. In a study by Mehlum (Mehlum 1998), a low SOC score was found to be a strong predictor of suicidal behaviour in 663 male Norwegian conscripts. The same was found in Greece. A study based on interviewing 1098 young conscripts revealed that subgroups with suicidal ideation and behaviour showed a significantly lower sense of coherence compared with the whole sample (Giotakos 2003). Similar results obtained in a Finnish study determined that a sense of coherence was impaired among those conscripts who committed suicide as well as those with mental health problems, alcohol and drug abuse (Ristkani et al. 2005).

In Sweden, a study examining the relation between self-rated health, risk factors in youth and adolescence and mortality among 49,321 young men participating in a nationwide military conscription survey was performed. It was found that poor self-rated health at conscription was associated with increased mortality during a 27-year follow-up (Larsson et al. 2002). In another study from Sweden, utilizing the Swedish military service conscription register and examining possible correlation between intelligence tests results at the age of 18 and suicides, it was found that there was a distinct correlation. The risk of suicide was two to three times higher in those who had the lowest scores. The greatest risk was seen among poorly performing offspring of well-educated parents (Gunnel et al. 2005). A certain predictive potential can be derived from the data of Jiang, Rasmussen and Wasserman (1999) who have followed more than 150 thousand young people in Sweden born in 1973–1975 and found that short stature and poor psychological performance (logic test) were significantly inversely associated with the risk of attempted suicide. These results were confirmed later on in a larger contingent (more than 1 million males conscripted from 1968 to 1999) showing that the risk of suicide decreases by 15 per cent for every five points of the body mass index (Magnusson et al. 2006).

The problem of prediction of performance (in an inverse way, with regards to pre-enrolment suicidal tendencies) during military service was studied in Israel. The authors paid special attention to adolescents who had records of attempted suicide prior to being drafted into the army. These showed much poorer performance during the military service (but not much difference regarding cognitive/educational abilities) compared with those who had no history of suicide attempt (Farbstein et al. 2002). Turning to other Israeli resources, it is interesting to mention a study which deals with greater susceptibility of men to contextual and situational factors, which contributes to completed suicide in the armed forces. This is based on the finding that suicide in the armed forces frequently occurs on the first working day, especially among men (Israeli soldiers can spend weekends at home in peacetime) (Weinberg et al. 2002).

An important issue is the so-called ‘psychological portrait’ of the suicidal soldier, and other (contextual) factors that may predict suicide patterns in the military. Background factors such as the presence of a mental disorder, accumulation of negative life events, developing the state of hopelessness, and emotional pain are the same for suicide in the general population and in soldiers. As mentioned before, the vulnerability to negative life events may be higher in young soldiers due to impairment of social support (Mehlum and Schwebs 2001). In Greece, Botsis and co-authors (1999) have studied psychological correlates of suicidal ideation in conscripts. It was found that those who had suicidal thoughts (17 per cent of the sample of 528 persons) had significantly higher scores for depression and hostility. It was also revealed that they had suffered much more stressful life events prior to conscription. From a case report, authors have concluded that a psychological complex of basic inferiority, low educational levels together with family problems and poor integration into a military unit may be strong determinants of soldier suicide (Cabarcapa and Pania 2004). One of the more recent papers from Israel compared psychological characteristics (from army records) of combatant and non-combatant soldiers who committed suicide, with others who did not commit suicide. It was found that combatant soldiers who committed suicide showed proof of greater behavioural adjustment, motivation to serve and a higher sense of duty. Those who were involved in combat had fewer referrals for psychological evaluation and fewer unit changes. This may reflect the tendency of perfectionism in these soldiers. The authors come to the conclusion that excessive motivation, and the tendency to be autonomous and independent may account for suicide in combatant soldiers, while in non-combatant soldiers the main predisposition for suicidal behaviour may be personality weakness (Bodner et al. 2006).

Whereas suicide in the military has been known for centuries, the awareness of the problem increased substantially after the introduction of post-traumatic stress disorder (PTSD) as a new diagnosis in the DSM-III system (American Psychiatric Association 1980) in the wake of the Vietnam war. Studies of veterans from this war have shown that there is a direct causal link between traumatic stress exposure and subsequent PTSD (Davidson 2000). The Vietnam veterans have also been shown to have a significantly increased standard mortality ratio (the ratio of the number of deaths observed in the study population to the number that would be expected if the study population had the same rate as the standard population) for suicide; particularly, those who had a diagnosis of PTSD (Bullman and Kang 1994) or had been wounded (Bullman and Kang 1996). In other studies, Vietnam veterans have also been shown to have increased levels of suicidal ideation and history of suicide attempts, and again these phenomena seem to be highly correlated with a diagnosis of PTSD. With additional diagnoses, particularly depression, there is an even stronger association (Hendin and Haas 1991; Kramer et al. 1994). Recently, Lester has again drawn attention to Vietnam veterans' suicides, highlighting a possible undercounting of suicides in this contingent (Lester 2005). Kaplan and colleagues in a prospective population-based study, compared about 104,000 veterans of different ages from the First and Second World Wars, Korean, Vietnam and post-Vietnam conflicts with more than 200,000 non-veterans. The comparison showed that suicide risk among veterans is twice as high as in non-veterans (Kaplan et al. 2007).

The Vietnam experiences led to new research efforts into the long-term psychological consequences of war and traumatic stress in general. Very soon, it became obvious that problems of war veterans are very similar across cultures notwithstanding social and economical differences. Though rates are not often calculated and compared to general population, suicides were reported among Russian (Afghanistan and Chechnya) veterans, British Falkland conflict veterans and US 1991 Iraq mission veterans. The extent of the problem can also be seen from the 1982 British-Argentine Falkland conflict where 256 British soldiers were killed: since then, 264 veterans have committed suicide, thus, suicides have taken more lives than the conflict itself (Spooner 2002).

Today, war veterans' mental health problems and suicide, in particular, are under even more intense investigation. For instance, Vietnam veterans continue to experience higher mortality rates due to external causes of death in comparison to non-veterans (Tegan et al. 2004). In the longitudinal model study of male Vietnam veterans in the USA, with a history of drug abuse it was found that PTSD, drug-dependence, non-fatal attempted suicides and suicidal ideation showed strong continuity over time (Price et al. 2004). Suicide attempters among veterans had higher psychiatric comorbidity, and more severe substance abuse, characteristics. In men, these problems were more significant than in women (Benda 2003, 2005). It is emphasized that auto-aggressive behaviour in veterans more often occurs in cases where PTSD is present (Begic and Jokic-Begic 2001). Different addictions, including gambling, are associated with suicide attempts. About 40 per cent of the veterans that are pathological gamblers have had attempted suicide (Kausch 2003). PTSD veterans appeared to own four times as many firearms as other subjects (Freeman et al. 2003). In general, there seems that starting with Second World War veterans to the Korean and Vietnam war veterans, suicide rates are growing (Lester 2005). Unfortunately, there is no data of veterans of other war conflicts. Conclusively, the problem of suicide in veterans is mainly a problem of the most socially unprotected contingent, or those with PTSD and different addictions.

Recently, with the wide role of peacekeeping duties worldwide, several researchers have studied suicide in former peacekeepers (Hall 1996; Ponteva et al. 2000; Wong et al. 2001; Thoresen et al. 2003; Thoresen and Mehlum 2004). The peacekeeping soldier is not expected to be engaged in regular war activities, but rather to act as a buffer between hostile parties. As a result, he or she has a more complex role, and in some crucial respects, a task completely different from soldiers traditionally trained for combat (Mehlum and Weisaeth 2002). Though, not involved directly in war, these contingents suffer great tension and psycho-emotional problems in the regions of ethnical and other conflicts. Recently, Thoresen et al. (2003) have reported the significant increase of suicide by firearms by Norwegian peacekeepers.

Some reports cannot give distinct evaluation of relative suicide risk in peacekeepers and concentrate on the most risky periods of the duty. For instance, four cases of suicide were reported among 4,000 Danish soldiers who took part in the UN mandated forces, two of them committed suicide less then one month before deployment, and two within a year after discharge from a mission (Hansen-Schwartz et al. 2002). On the other hand, a Canadian study does not confirm the higher risk of suicides among UN peacekeepers (Wong et al. 2001). In a Swedish study, 39,768 former peacekeepers were compared to the general population, and again, a lower number of suicides was found in this contingent. The authors focus mostly on military lifestyles and interpersonal relationship strains as possible factors contributing to suicide among peacekeepers. This research area has unique characteristics possibly revealing specific situations in each cultural and social context of the country of origin and the country of mission.

Protective factors

Since suicides in the military setting are lower than in the general population, there are obviously some protective features in the military that may play a role in prevention. These may be:

The majority of publications on military suicide discuss their results from the point of view of how findings can help to identify suicidal persons, and how the results may be implemented in organizing suicide prevention programmes and strategies. However, few studies have been published regarding suicide prevention in the armies and evaluations of such preventive measures. Strategic approaches to suicide prevention are well described from the United States and Norwegian armies. In Norway, the army suicide prevention programme has been a part of the National Strategy for Suicide Prevention launched in 1994 (Mehlum and Schwebs 2001). It has primary, secondary and tertiary preventive components and includes five main domains: leadership interventions, information/education, medical interventions, welfare, and 24-hour crisis telephone service. Much attention is paid to the role of commanders in military units, which is why a lot is invested in such leaders' competence. Several training packets extending from two hours to two days have been developed and disseminated, from the level of the private soldier to commanders. In recent years, the Norwegian army has also emphasized restricting access to firearms in private homes of officers, members of the home guard, and the military reserve; this initiative has been accompanied by a strong reduction in the number of military firearm suicides.

The US Army suicide prevention strategy (The Army Suicide Prevention Program) is based on the same principles as the Norwegian. They both focus on a proactive suicide prevention programme, which is fundamental in averting the needless tragedy of suicide in the military. Suicide is preventable, and leaders must play an active and sensitive role in showing care and concern for their soldiers. Positive leadership, careful listening and deep concern for soldiers are crucial for suicide prevention in the military. It is important for leaders to know their soldiers and their concerns, and never hesitate to obtain professional help for a soldier in need. The key to preventing suicide in the unit is to respond quickly to any verbal, behavioural or situational clues. Soldiers need to be taught to take any suicidal statement by a fellow soldier seriously, and to inform the chain of command immediately. Prevention efforts must also focus on the personal responsibility of commanders and leaders to care for the soldiers under their charge.

Recently, the US Army introduced a new suicide prevention campaign plan focusing on standardized suicide prevention training (Army News Service 2001a). The US Army suicide prevention model focuses on four major areas: developing soldiers' life-coping skills, encouraging help-seeking behaviour, raising vigilance through suicide awareness, integrating and synchronizing unit and community programs (Army News Service 2001b). Special teams are assessing the situation and present it to authorities (Jontz 2004). Psychological autopsies are gathered and used for lessons learned after suicide; in addition, special policy and training courses are developed (Ritchie and Gelles 2002). Introducing anger management intervention (as far as anger appeared to be a predominating emotion) in Bagram, Afghanistan, secured the absence of suicide cases in the 7000 contingent during a half-year period in 2002 (Reyes and Hicklin 2005). Special prevention programs exist in specific branches of service, such as the air force and the navy (Patterson et al. 2001; Stander et al. 2004).

The programme of the air force was recently thoroughly evaluated by Knox et al. (2003). The investigation was designed as a cohort study with a quasi-experimental approach and analysis of cohorts before (1990–1996) and after (1997–2002) the intervention. The intervention was targeted at reducing risk and enhancing protective factors and consisted of measures aimed to encourage personnel to seek help for mental health, psychological or relationship problems, and enhancing general understandings of mental health issues as well as changing policies and social norms. One of the important aspects of the intervention was training (education) in suicide prevention. It was found that implementation of the programme was associated with a sustained decline in the rate of suicides and other adverse outcomes (homicide, family violence). A 33 per cent relative risk reduction was observed for suicide after the intervention, reduction for other outcomes ranged from 18–54 per cent (Knox et al. 2003).

In the Ukraine, suicide preventive intervention targeting air force units with about 10,000 personnel was performed (Rozanov et al. 2002). This intervention was based on the US model of gatekeepers' education and included education/information of the responsible personnel, military psychologists and medical staff. Distribution of educational material to the distant subunits and yearly educational seminars collecting all relevant commanders and staff proved to be a good model of prevention. By introducing suicide prevention education, the rate of suicides quickly decreased, giving rise to enthusiasm and satisfaction among commanders. Nevertheless, it was soon revealed that only constant education and consultations of specialists will ensure success, an important experience similar to the one made in the Gotland study (Rutz et al. 1995).

In many armed forces across the world, suicide preventive efforts have been developed in recent years and some of these have been made part of a national programmes for suicide prevention: in the US, UK, Norway, Sweden and Germany to mention a few. There are also reports about existing prevention programmes in the Hellenic navy (Polychronidis 1999), in the army of Serbia and Montenegro (Dedic et al. 2006; Gordana and Milivoje 2007), and in the Turkish air force (Gerede 2006). There are no reports regarding evaluations of these programmes, some of which have just been launched. To our knowledge, the evaluation performed by Knox et al. (2003) remains the only example.

Introduction of suicide prevention measures in the military environment sometimes is not an easy thing to do. Part of military self-esteem is the ability to stand straight while facing different threats. Therefore, training military staff to ask for help may not be easily accepted, in that such help-seeking behaviour, as well as referral to mental health facilities, can be seen as the end of a career. Even screening for psychological illness in the military is evaluated in a specific way (Rona et al. 2005).

In most countries, a lot more could be gained through more coordinated action between the military and civil sectors, and through more focused strategies within the armed forces. An obstacle to such a development has probably been the downsizing of military forces seen nearly everywhere, depleting the military of resources in terms of experts, manpower and money.

Suicide in military settings has been a subject of research in many countries. The problem is most excessively covered in countries with big armies (USA, Russian Federation), and those involved in peacekeeping missions (Norway, Sweden). Almost all EU countries, most of them members of NATO, are implementing research and prevention regarding military suicides. There are reports also from Canada, Israel, Poland, Serbia and Montenegro and Ukraine. In the resources available, we could not find any information regarding the armies of China, Pakistan or countries of the Middle East. Prevention is an important topic in all studies. Nevertheless, stringent studies regarding prevention measures and their evaluation are not very numerous. Suicide in military settings is generally lower than in the general population of the same gender and age (some exceptions can be seen when specific age groups are evaluated). There are some specific risk factors for suicide in the army, such as exposure to traumatic stress (in war and peacekeeping duty), the military lifestyle with frequent relocations and break-up of protective social structures, and the easy access to firearms and other dangerous equipment. The risk factors may sometimes be balanced by protective factors such as social support, medical support and additional possibilities for prevention. Many authors state that during pre-enrolment and regular medical examination of the mental health of conscripts, suicidal tendencies are often not identified successfully, pointing to the need for better screening at an early stage. Educational initiatives directed at military leaders and medical personnel within the armed forces are other recommended measures.