40 Somatic diseases and suicidal behaviour

The association between chronic somatic disorders and the risk of suicide has been examined in many studies. Common features in the studies are a large variation in quality and choice of study method. The studies performed, in more recent decades, have substantially improved knowledge not only on the extent of risk, but also on factors influencing the risk. Most of the studies on completed suicides have been made in European countries, the United States of America and Australia, one single study is from Japan (Whitlock 1985; Allgulander and Fisher 1990; Stenager and Stenager 1992; Harris and Barraclough 1994; Stenager and Stenager 1997; Ruzicka et al. 2005).

The majority of studies on the association between chronic somatic disorders and suicidal behaviour, including suicidal thoughts and suicide attempts, studies have demonstrated a correlation. The studies have been performed in Europe (DeLeo et al. 1999; Pajonk et al. 2002), the US (Druss and Pincuss 2000) and Australia (Lawrence et al. 2000).

This chapter is a review on present knowledge on suicide and suicidal behaviour in selected somatic disorders and pain syndromes, with focus on studies from different parts of the world, and whether or not this reflects variation in the estimated risk of suicidal behaviour.

Studies on the association between somatic disorders and suicide are qualitatively better than those investigating the association between somatic disorders and suicide attempts. Three types of studys have been performed: first, frequency and diagnosis of somatic disorders in patients who have made suicide attempts (Kontaxakis et al. 1988; Nielsen et al. 1990; Dietzfelbinger et al. 1991; Öjehagen et al. 1991; Wedler 1991; Stenager et al. 1992). Next, follow-up studies on patients with suicide attempts and somatic disorders, where the risk of repetition of suicidal behaviour has been estimated (Nielsen et al. 1990; DeLeo et al. 1999). Thirdly, studies on the risk of suicide attempts in defined populations and studies in defined populations in comparison to a control group (Hawton et al. 1980). The above mentioned studies have shown a frequency of 27–50 per cent of somatic disorders differing between in- and outpatients and depending on the definition of the somatic disorder. Painful disorders and disorders involved with an increased risk of depression are most common in suicide attempters. Furthermore, somatic disorders in older people are of significant importance.

Suicidal behaviour in a human being can be considered, as a consequence of life, being unbearable. Reasons for this are manifold and may be of physical, psychological and social character. The threshold for suicidal behaviour in individuals varies, and a single cause may only be a precipitating factor. The hard task is to predict when a given person reaches the threshold and, before that occurs, provide relevant help. In this context, knowledge about which disorders might entail increased risks of suicide may be helpful.

Somatic disorder involves troubles of physical, psychological, and social character. The somatic disorder may imply pains, handicaps, as well as distress about whether or not it is life-threatening. Complicating, simultaneously present, psychic disorders give limitations in social performance, loss of the capability to work and need for public social services. The risk of suicide may change during the course of a disorder. There may be periods during the course where the risk may be increased, for instance, before or after the diagnosis is made, or when the patient has had the disease for a long period. Age may also be a parameter of importance (Juurlink et al. 2004). Other factors influencing the risk of suicide could be depression, cognitive deficits, anxiety, and medically induced abuse of medication due to pains, psychosis and organic mental disorders, e.g. vascular or Alzheimer's dementia. Attention should be given to psychosomatic disorders that have been increasingly recognized during the last decade.

Studies concerned with the association between pain and suicide come across considerable methodological problems, such as the definition of pain and the selection of a study group. A controlled study in England of 6569 persons with pains found a fivefold increased risk of suicide, accidents and violence. The study involved persons with extensive pain, among which many were cancer patients (Macfarlane et al. 2001). A review of eighteen studies on chronic pain and suicidal behaviour concluded that suicidal behaviour is frequent in patients with pain. It was recommended to be aware of this aspect in a psychiatric setting (Fishbain 1999). A British study of 1665 suicide attempters showed that in 4 per cent of the cases, pain was of importance (Theodoulou et al. 2005). The patients differed from other studies in that they were older, had a higher suicidal intention score and were rarely seen in a psychiatric setting.

As many patients with cancer have pain, an increased risk of suicide can be expected in these patients.

A number of conditions need to be considered when designing studies of this sort. The most important is the presence of, and access to, adequate information about the health care given to the studied population. This puts a limitation on the number of communities where such studies can be performed. Furthermore, this implies that the disorder should be well characterized with accepted diagnostic criteria. The course of the disorder should be well described in order to be able to ascertain during which periods the risk of suicide increases. It is of great interest to know whether it is a genetic disorder, and if it is possible to screen for the disorder.

When studies are evaluated regarding the quality of design, the following three parameters should be taken into consideration. In the choice of study population, one should consider whether the selected population is at a high or low risk of suicide. Regarding the registration of suicide, one should bear in mind how the suicide data is validated. Finally, it is important to reflect upon how the control group is selected. When studying the risk of suicide in somatic disorder, there are quite a few prerequisites. A disorder with an adequate diagnostic criterion is of prime importance as well as large and representative groups to study. Furthermore, the control group should include a complete background population. Usually the normal population in the study area are used, i.e the Danish population. Only validated registry of death causes should be used and statistical analyses with calculation of standard mortality ratio (SMR) controlling for at least sex and age and preferably with survival curves. SMR is a statistical measure for the relative risk adjusted for age and sex. An SMR of 2 means a doubled suicide risk.

A number of early studies were based on autopsy studies. Such studies are not useful if the purpose is to estimate the risk of suicide, but are valuable when studying characteristics of patients and risk factors in patients who have committed suicide.

Follow-up studies in well-defined groups of patients have frequently been performed. The studies follow the patients with a defined disorder for a certain period and, by comparing with controls, estimate the number of suicides. The studies are often part of studies on mortality of the disorder. Such studies can be biased by a number of factors: first, a lack of defined diagnostic criteria; also, in the selection process of the patients, if, for example, only inpatients are included. Further, there may be great variation of follow-up periods, not well-defined controls and far from optimal statistic sources. From a methodological point of view, studies based on regional or nationwide registries on disorders with well-defined diagnostic criteria, compared to the background population using the sex- and age-standardized mortality ratio (SMR), are advantageous. The benefits of such studies are many: they deal with well-defined patients on a regional/national basis found in registries of causes of death during a defined period. Moreover, they compare risks of suicide in patients and the background population and calculate a standardized SMR for sex and age. Generally speaking, the quality of the studies has increased in recent years, concurrently with the improvement of registries on specific disorders and causes of death and improved statistics in many parts of the world. Many studies are now based on validated registries and large populations in contrast to earlier studies, which were based on case stories or small populations.

The diagnosis of cancer can be associated with troublesome treatments, pains, bad prognosis, economical troubles and, as a possible consequence, of depression and crisis reaction. Many cancers are ultimately lethal. Thus, it is not surprising that a diagnosis of cancer is associated with an increased risk of suicide. Numerous studies have dealt with the association of cancer and risk of suicide. They include autopsy studies and register-based studies in large populations, particularly from Scandinavia and the US (Louhivouri and Hakama 1979; Fox et al. 1982; Allebeck et al. 1989; Levi et al. 1991; Storm et al. 1992). Studies from Japan also demonstrate an increased risk of suicide in cancer patients compared to the background population (Tanaka et al. 1999; Akechi et al. 2002). These methodologically well-performed studies estimate the SMR in males between 1.9 and 2.8, while there is disagreement whether or not women have an increased risk. There is no significant difference cross-nationally in the results. Few studies found an elevated risk of suicide in certain types of cancer. In patients with cancer in the oesophagus, an increased suicide risk of 35× was found (Innos et al. 2003). The risk was especially increased in the period after the diagnosis was made. A Swedish study found an increase of suicide by 16× in all males with cancer in the first year after diagnosis (Allebeck et al. 1989). However, Norwegian and Danish studies could not confirm this elevated risk during the first year after the diagnosis (Storm et al. 1992; Hem et al. 2004; Yousaf et al. 2005).

Next to cancer, neurological disorders are the most thoroughly investigated disorders. This is obvious considering that many neurological disorders have an increased risk of psychic disorders such as depression. Some of the neurological disorders do not always have well-defined diagnostic criteria, or have not been studied in populations with good registries. As a consequence, not all disorders have been subjected to well-designed studies. Most of the studies have been conducted in Europe, Australia, the US and Japan. Despite the different geographic areas investigated, the results are fairly consistent. A study from Japan (Kishi et al. 2001) examined the extent of suicidal thoughts in patients with stroke, traumatic brain injury, myocardial infarction and spinal cord injury. A total of 7.3 per cent had suicidal thoughts, and among patients with depression, the figure was 25 per cent. The study stressed the importance of being aware of, and carefully treating, depression in these patients as a preventive measure.

The largest study on the frequency of suicide in patients with multiple sclerosis (MS) is from the Danish MS Registry and based on approximately 5000 patients (Stenager et al. 1992). The SMR for males diagnosed before the age of 40 years was 3.12, and the SMR for females diagnosed before the age of 40 years was 2.12. There was no increased risk of suicide for patients diagnosed after the age of 40 years. The risk was largest in the first five years after diagnosis. The cumulated lifetime risk of suicide was approximately twice that of background population. There was a follow-up study from the MS Registry (Brønnum-Hansen et al. 2005) comprising 10,174 patients with MS diagnosed in the period 1953 to 1996 and a follow-up the 1 January 1999. This study confirmed a doubled risk of suicide in MS patients, as well as new findings that an increased risk of suicide was also found 20 years after the diagnosis was made. A study from Sweden (Frederikson et al. 2003), based on 122,834 persons followed from 1969 to 1996, confirmed the Danish results with a SMR of 2.3 in MS patients, and an increased risk within the first 5 years after the diagnosis was made. However, a Canadian study (Sadovnick et al. 1985) found a substantially increased risk of suicide by a factor of 7.5 compared to the Scandinavian results, yet this study did not calculate the SMR or standardize for sex and age, which is a bias as the sex distribution is skewed in MS patients. Yet, another study from the US (Feinstein et al. 2002) examined suicidal intention in 140 MS patients and found that it was associated with depression, abuse of alcohol and social isolation.

Most studies demonstrate an increased risk of suicide in patients with Huntington's chorea, but all studies have methodological problems and, subsequently, the conclusions on the size of the risk should be taken with caution (Schoenfeld et al. 1984; Farrer 1986). A study from The Huntington study Group Database from the US (Paulsen et al. 2005) examined the suicidal thoughts of 4171 patients, who were in different stages of the disease, from healthy carriers to severely disabled. Among the healthy carriers, i.e. with normal neurological signs, 9.1 per cent had suicidal thoughts. In the group with slight neurological signs, 19.8 per cent had suicidal thoughts, and in the group with signs of a possible disease, the figure was 23.5 per cent. In the group where the disease was active with certainty, patients in the early stages had a high risk, while the risk decreased with the progression of the disease. The study showed that the risk of suicidal thoughts was the largest immediately before the diagnosis was made, and when the disorder resulted in loss of capability of taking care of oneself. Similar results were found in a study from Hungary (Baliko et al. 2004), which found more suicides in the early stages of the disease compared to the later stages.

Huntington's chorea is an autosomal dominant disorder that can be diagnosed in healthy family members. A Canadian study (Almquist et al. 1999) examined 4527 persons who had been genetically tested with the purpose of estimating whether or not the test results made a difference in the number of persons who committed suicide, made suicide attempts or were admitted to a psychiatric ward. The study concluded that the behaviour of the tested persons was not dependent on whether or not they tested positive on the diagnosis. Instead, suicidal behaviour was dependent on whether the tested persons had previous psychiatric disorders or were unemployed.

Spinal cord lesions are most frequently the result of accidents in young males, and may result in life-long dependence on a wheelchair. Many studies have been performed on the risk of suicide in patients with spinal cord lesions. Most studies have methodological problems, but a well conducted North American study found a 4.9 increased risk of suicide (DeVivo et al. 1991).

Epilepsy is one of the somatic disorders, in which the association of risk of suicide has been examined most thoroughly. A review (Barraclough 1987) of the existing literature on suicidal risk in different groups of epilepsies found that patients with epileptic seizures generated in the temporal lobes had a five times increased risk of suicide, and patients with difficult treatable epilepsy, an increased risk of 25 per cent.

A methodologically adequate British study (White et al. 1979) followed 2000 patients admitted between 1931 and 1971 and on anticonvulsant treatment to 1977. In estimating the mortality, consideration of sex, age and period of risk was made. The study found an increased risk of suicide of 5.4. A mortality study from Sweden (Nilsson et al. 1997), based on 9061 patients diagnosed with epilepsy, found an SMR of 3.6. When including death due to injuries and poisoning, conceivably hidden suicides, the SMR was 5.6. A follow-up study (Nilsson et al. 2002) examined risk factors for suicide and suicide attempts in 6880 patients with epilepsy. The risk of suicide was increased 9× times in patients with a psychiatric disorder. The risk was increased 10× if the patients were also on an anti-psychotic treatment. In patients with onset of epilepsy before the age of 18, the risk was increased 16× compared to patients with onset after the age of 18 years.

A Russian study (Kalinin and Polianski 2003) on risk factors for suicidal behaviour in patients with epilepsy confirmed that patients with organic affective disorders, personality changes and cognitive deterioration had substantially increased risk of suicide.

An British study examined the frequency of epilepsy in patients admitted due to a suicide attempt during a two-year period. Compared to the prevalence of epilepsy in the background population, the number of suicide attempts in patients with epilepsy increased 5×. Patients with epilepsy had more frequently been in psychiatric treatment and had more suicide attempts compared to the background population. A US study (Mendez et al. 1989) comparing suicide attempts in matched groups, with and without epilepsy, concluded that patients with epilepsy more frequently had borderline personality, psychotic disorders and previous suicide attempts.

A US study (Breslau et al. 1991) compared diagnosed migraine patients, with and without aura, and found an increased risk of suicide attempts with an odds ratio of 3.0 in patients with migraine with aura.

All studies on suicidal risk in patients with brain injuries are based on studies on wounded soldiers from the Second World War. These studies do not fulfil the present methodological demands on studies in suicidology. However, they report an increased risk of suicide.

Oquendo et al. (2004) have found that traumatic brain injury is associated with psychiatric illness, suicidal ideation, suicide attempts and completed suicide. The study was retrospective and has to be confirmed in a prospective design.

An Australian study (Simpson and Tate 2005) has examined the importance of demographic, clinical and other parameters in suicidal behaviour in 172 patients with traumatic brain injury. They found that patients with a previous psychiatric disorder and history of abuse had a 21× increased risk of suicidal behaviour compared to patients with no previous psychic disorders or abuse.

Few studies on suicidal risk in Parkinson's disease (PD) have been performed. In a Danish study (Stenager et al. 1994) on 485 patients followed in an outpatient clinic for a little less than 20 years, a reduced risk of suicide was found in males with PD, while the risk in females was the same as in the background population. The low risk was explained as the result of late age at onset and good possibilities for treatment. Only a small number of patients participated, so the results have to be treated with caution. In the study from USA of 144,364 patients with PD were found to be at a 10× reduced risk of suicide, that was reduced 10× compared to the background population (Myslobodsky et al. 2001). This study, thus, confirms the results previously found.

In a Danish study (Stenager et al. 1998) of 37,869 patients with stroke admitted to hospital and followed up to 17 years, the risk of suicide in females below 60 years of age was increased 13× and in males 6×. Patients older than 60 years had an increased risk of 1.5–2×. Part of the increased risk could be explained by an increased risk of depression in females after a stroke. A recent Danish study (Teasdale and Engberg 2001) confirms the doubled risk of suicide in stroke patients. In patients below 50 years of age, the SMR was 2.85, i.e. lower than in the first study. The risk was the largest in the first five years after diagnosis and in patients discharged after a short admittance.

A single study (Bak et al. 1994) did not find any increased risk of suicide in motor neuron disease. Only 116 patients participated.

In recent years, studies on mental retardation have been published. A Swedish study (Gunnell et al. 2005) on 987,308 males examined for military service showed that males with low intelligence scores in the psychological tests had 2–3× increased risk of suicide compared to males with high intelligence scores.

On the other hand, a Finnish study (Patja et al. 2001) in a population of persons with mental retardation found that females had a risk comparable to the total Finnish population, while males had a lower risk than the background population. Persons who committed suicide had only mild retardation and a psychic disorder. The mental retardation being of different magnitude in these two studies may explain the difference in results.

Recent studies on the risk of suicide in patients with cardiac diseases have dealt with whether or not cholesterol-reducing medication increases the risk of suicide and other types of violent deaths in these patients. A US study (Neaton et al. 1992) of 350,000 males followed for 12 years found a 1.6× increased risk of suicide in males with low cholesterol count.

In Jacobs' et al. (1992) meta-analysis, a similar conclusion was reached. People with low cholesterol level had an increased risk of dying from reasons other than heart disorders, including suicide. Other studies (La Rosa et al. 1995) did not find any association between low cholesterol level and mortality from other disorders apart from heart diseases. More studies are needed to reach a reliable conclusion.

Two US studies have estimated the risk of suicide in asthma patients. One study examined suicidal thoughts in 1285 young American patients in the age interval of 9–17 years of age. Patients with asthma had an increased risk of 3× for suicidal thoughts. Controlling for the comorbidity of a psychiatric disorder did not change the result. A study (Goodwin and Eaton 2005) on the association of suicidal thoughts and suicide attempts in patients with asthma, controlling for psychiatric disorders, confirmed the result. Hence, the studies indicate that asthma may increase the risk for suicidal behaviour.

Disorders like Crohn's disease and ulcerative colitis often affect young people and are associated with pains, operations, and discomfort. Crohn's disease is also associated with the risk of depression. An Italian study (Palli et al. 1998) showed a non-significant tendency to increased mortality due to suicide in patients with Crohn's disease and ulcerative colitis. A Danish study (Winther et al. 2003) found an increased risk of suicide in females with ulcerative colitis. Previous studies found an increased risk of suicide in females with Crohn's disease in the UK (Prior et al. 1981). Another English study (Cooke et al. 1980) demonstrated an increased risk in both males and females. In conclusion, patients with bowel disorders seem to have an increased risk of suicide.

Liver transplantation implies much strain on the patients, it is also expensive and the number of available organs is small. Candidates for treatment usually display symptoms of depression, anxiety, cognitive disorders and fear of the future. The reported number of suicides has been rare in candidates for liver transplantation. On the other hand, case reports (Riether and Mahler 1994) have indicated an increasing number of patients attempting or committing suicide. Thus, it is important to be aware of the risk for suicide in candidates for liver transplantation. However, the small number of transplantations, and an even smaller number of suicides makes difficult to draw any conclusion as to whether or not patients with liver transplantations have an increased risk of suicide.

The risk of suicide in patients with kidney disorders, and especially in patients with renal failure and kidney transplants, has been studied. Patients with renal failure have an increased risk of suicide (Haenel et al. 1980). Methodological difficulties in the studies do not allow an exact estimation of the size of the risk. The studies have not found any difference in the risk in transplanted patients compared to non transplanted patients. Part of the explanation may be that transplanted patients need lifelong medication and they have a risk of renal failure in the transplant.

A number of studies have examined the risk of suicide in diabetes mellitus, but only one (Kyvik et al. 1994) has dealt satisfactorily with the methodological demands previously described. This is a Danish study of 1682 male patients with diabetes mellitus treated with insulin. Twelve had committed suicide, and the study found that males in the age interval of 20–24 years of age had an increased risk with a SMR of 2.98. The study concluded that the number of suicides may have been underestimated as a large number of patients died from unknown reasons. This problem was later discussed in a German study on mortality in diabetes mellitus (Mulhauser et al. 2002).

A number of studies on mortality in diabetes have been performed, primarily European ones (Sartor and Dahlquist 1995; Swerdlow and Jones 1996; Warner et al. 1998; Podar et al. 2000). The SMR for suicide has not been calculated, but an increased mortality, especially in young diabetics with bad control of the disease, could indicate that some of these deaths are due to suicide.

The frequency of suicide attempts in diabetes mellitus has been discussed frequently. In this disorder, abuse of insulin through self-destructive behaviour can result in an insulin chock. A review of the literature on the abuse of insulin (Kaminer and Robbins 1989) revealed 17 cases of suicide and 80 cases of attempted suicide. The sex ratio was 1:1, the age distribution among the attempters was even, and about 50 per cent of the attempters were repeaters. A study of overdoses of insulin (Gale 1980) found that 4 of 204 episodes were suicide attempts, and holds that suicidal overdoses of insulin are not uncommon. A British study (Jefferys and Volans 1983) of self-poisoning in diabetic patients referred to the National Poisons Service in London in the period 1978 to 1979 found that 64 among 386 diabetic patients used hypoglycaemic drugs. The rest had used other drugs. Among those taking hypoglycaemic drugs, ten died from brain damage. The authors concluded that self-poisoning was common in diabetic patients and suggested toxicological screening in patients with prolonged keto-acidotic coma. A study from the United States has concluded that suicide attempts and suicide with insulin in diabetics are not uncommon (Arem and Zoghbi 1985). However, one must bear in mind that, in certain cases, an overdose of insulin might be a mistake.

The studies have concentrated on arthritis and amputated patients. Studies on the last mentioned group are old and mainly on soldiers from the Second World War. A study by Dorpat and Ripley (1960) from the US based on forensic material found that patients with arthritis had an increased risk of suicide in the order of 2–3 times compared to the background population. Another forensic study (Whitlock 1985) did not find any association between suicide and arthritis. The study was based on 1000 suicides in England and Wales. The frequency of suicides in patients with arthritis was compared with the number of patients with arthritis in the UK. The method is questionable. On the other hand, a Finnish study (Timonen et al. 2003) found an association between depression, suicide attempts and suicide in patients with rheumatoid arthritis, and thus confirms the result of the forensic study. A British study of 300 patients with lupus erythomatosis disseminatus has indicated that this group of patients may have an increased risk of suicide, especially in patients with neuro-psychiatric symptoms. The SMR was not calculated (Karassa et al. 2003).

Tinnitus involves an increased risk of depression. An American study has found a life-time prevalence of depression of 62 per cent compared to 21 per cent in a control group (Harrop-Griffiths et al. 1987). A study from the UK (Lewis et al. 1994) on 28 patients with tinnitus who committed suicide demonstrated that they were predominantly males, elderly and socially isolated. Most had psychiatric disorders (97 per cent), mainly depression (70 per cent). Forty per cent of the suicides were committed within one year after the onset of tinnitus and 50 per cent within two years, indicating that tinnitus was an important risk factor. It was estimated that the risk of suicide was increased compared to the background population. However, SMR was not calculated.

Almost from the start of the AIDS epidemic, it has been discussed whether or not this disorder resulted in an increased risk of suicide. A US study (Marzuk et al. 1988) found a 36× increased risk of suicide in males with AIDS compared to controls. The possibilities of treatment have improved considerably since then. Moreover, the social stigmatisation has decreased in the United States. This could be part of the explanation of the reduced risk of suicide reflected in another study from US (Coté et al. 1992), which found an increased risk of suicide of 7.4 compared to controls. Another study performed in the US (Marzuk et al. 1997) on suicide risk in the HIV positive in the period 1991–93 demonstrated that HIV infection is associated with other risk factors for suicide, such as abuse of narcotics and other social factors.

A review of literature (Komiti et al. 2001) from studies in the United States of America, Europe and Australia on suicidal risk in patients with HIV/AIDS concluded that the risk of suicide is the same order as in other chronic somatic disorders. In AIDS, there is an additional problem of psychiatric disorders, abuse and social consequences that are difficult to control for. A US study (Perry et al. 1990) on suicide risk in patients tested for HIV showed that both zero-positive and zero-negative patients had suicidal thoughts in the period of waiting for the result of the test, and that the thoughts were resolved in the following two months.

In this chapter, we have shown that a number of somatic disorders are associated with increased risk of suicidal behaviour. Suicidal behaviour can be considered as the final indicator of a life with insurmountable physical, psychological and social problems. There are fairly robust studies on increased risk of suicide in a number of neurological disorders and cancer, while the studies in cardiac, lung, rheumatologic disorders and other somatic disorders are fewer and far less thorough. A usually well accepted perception of disease as being a biological, psychological, and social process could be expected to be reflected in studies on suicidal behaviour. However, this is not the case. Suicidal behaviour is often considered solely as a psychiatric problem, which is a far too simplistic view (Fredriksen et al. 2005). In line with this, studies controlling for psychiatric disorders still show increased risk of suicide in somatic disorders. Signs of depression, anxiety, hopelessness, crisis reaction, pain, previous suicide attempts, suicidal thoughts, complicated social conditions (family situation, work, finance, leisure time) and the waiting time for somatic diagnosis can be signals of increased risk in somatic patients. Prophylactic measures are obviously of importance in somatic disorders. Depression and pain should be treated adequately. Measures to ease the social–medicinal consequences of disease in the patient, involvement of social workers, psychologists and others should take place.

A final point to be made is that despite the fact that the studies have been made on different continents and populations, they do not show significant differences. However, the majority of studies have been conducted in Western Europe, the USA and Australia, which are countries and continents where many similarities as far as social systems, cultural history and religion exist. When more studies emerge from the remaining continents, further differences may be revealed. Today, there is great need for studies from other parts of the world.