One of the most well-established «truths» about suicide is that mental disorders play a significant role in at least 90 % of cases. Therein also lies an assumption about a causal relationship. However, the evidence base for this «truth» is doubtful because it is derived from studies that do not stand up to examination. This article discusses some serious implications of this.
It is constantly repeated, both in the academic literature and in the media, that mental disorders play a key role in nine out of ten suicides (1 – 3). The evidence base for this «90 % truth» consists mainly of a long series of psychological autopsy studies (PA studies) (3), in which psychiatric diagnoses of the deceased have been made based on interviews with the bereaved, often many years after the suicide. Such studies are, however, fraught with serious methodological problems (4), in particular connected with the diagnostic process in which someone (the deceased) is diagnosed based on interviews with others (the bereaved) (5). Nevertheless, it is claimed that so long as standardised diagnostic instruments are used, findings from such studies are both reliable and valid (4, 6).
If the questions included in these types of standardised diagnostic instruments are examined more closely, it becomes evident, however, that many of them cannot be reliably answered by anyone other than the person to be diagnosed (see 5 for examples). And – if the questions cannot be answered reliably, the diagnoses made cannot be valid (5).
It might appear that the strong association between mental disorders and suicide has acquired the status of fact based simply on a large number of studies of the same type. However, it does not help that a finding is repeated in innumerable studies, if the studies are not suited to answering the research question (5).
The way you ask dictates the answer you receive?
If following a suicide the bereaved are interviewed based on a diagnostic questionnaire, as in traditional PA studies, there is a relatively strong possibility of ending up with a diagnosis for the deceased. This emerges clearly from the review study by Cavanagh et al. (3). However, such diagnoses are based on the subjective perceptions, emotions and experiences of the bereaved, or on speculations about questions which, in many cases, they cannot possibly answer with certainty on behalf of the deceased (5). On the other hand, if the bereaved are allowed to speak freely about what they think was central to the suicide, the picture is entirely different.
A clear example of this is a PA study from England (7, 8). In the first part of this study the informants answered diagnostic questions. 68 % of the deceased were then found to qualify for a psychiatric diagnosis (7). When the narrative part of the interviews with the same informants was analysed qualitatively, however, it transpired that very few spoke of psychiatric disorders as being central to the suicide (8).
Two qualitative PA studies have now been conducted in Norway, in which several bereaved persons associated with each suicide have been able to tell their story of what led to the suicide. In her study of suicide among the elderly (65 – 90 years) Kjølseth found that suicide had to do with who the deceased had been, how they had lived their lives given their circumstances and what their experiences meant when confronted with old age (9). Many had lived in very difficult circumstances, both when growing up (for example, with the loss of important carers, illness and poverty) and later in life (serious losses and challenges). The informants described the deceased as conscientious, action-oriented and skilled in their work (10). They were also described as emotionally distant, obstinate and with a need for control, which had contributed to creating conflicts in close relationships. It had been difficult for them to accept help. This would entail relinquishing control, which was contrary to their self-image/identity (10). It might thereby appear that their strength and ability to deal with difficulties throughout their lives were what made them vulnerable to suicide in old age, because they would not or could not adapt to age-related loss by developing new coping strategies or accepting necessary assistance (9, 10). Age-related loss of function resulted in a feeling of having lost themselves, since they could no longer do what they wanted. Life was thus perceived as a burden. They had a realistic view of the future, a future they did not want, and therefore made an existential choice to take their own lives. In this way they regained control. The title of Kjølseth’s PhD thesis is hence also «Control in life – and in death: an understanding of suicide among the elderly» (9).
In Rasmussen’s study of suicide among young men (18 – 30 years) (11) only men who had not been in contact with mental health care, and who had made no previous suicide attempts were included. In-depth interviews with mothers, fathers, siblings, girlfriends and friends, as well as the deceased’s suicide notes made it possible to analyse data from both a developmental and a relational perspective. As in Kjølseth’s study, the young men were also described as diligent and achievement-oriented at their work and studies. Several were described as perfectionists. Despite the fact that they appeared successful, however, the analysis showed that early in life they had developed a fragile, performance-based self-esteem that made them vulnerable even when encountering small setbacks. The study reveals a particular vulnerability to experiencing themselves as unsuccessful and rejected when they failed to live up to their ideal of performance and how they had envisioned that life should be. Instead of lowering their performance expectations in such situations, they were overwhelmed by strong emotions, particularly shame and anger, which they were unable to either regulate or deal with. Their previous strategy of compensating by continuously improving their performance no longer worked, and suicide became the way out of a state of intolerable mental pain (12).
An interesting finding of both these studies is that the informants placed little emphasis on mental disorders in their narratives about what was central to the deceased’s suicide. Few informants had seen signs of serious mental illness (11), and many explicitly stated that the deceased had not been depressed (11, 13). This is in strong contrast to the conclusion drawn in most quantitative PA studies, namely that almost everyone who has taken his/her own life was found to have one or more mental disorders (3), with accompanying causal implications. The findings therefore challenge the established notion that suicide is mainly a symptom of a mental disorder.
Does it really matter?
The question, then, is whether it matters that it is constantly claimed that mental disorders play a significant role in nine out of ten suicides. It is of course important to treat mental disorders, also with the intent to prevent suicide, but an exaggerated focus on «the 90 % truth» can have unfortunate consequences.
Dyregrov (14) has previously described this as «a dangerous discourse». One unfortunate consequence is the possible propagation of the notion that provided there are no signs of mental illness, there is no danger afoot (14). As shown in the study by Rasmussen et al. (12), this may be disastrous.
Another consequence is that when the focus is on mental disorders as «the main cause» of suicide, implicit in this is that the most important thing to be done to prevent suicide is to diagnose and treat mental disorders. This is often also stated explicitly (3), and mental health care is described as «our most important tool for suicide prevention» (15). The spotlight is thereby directed at the individual, since suicidality is regarded as something that lies within the individual, while the importance of context and relationships receives less attention (16). This in turn can lead to a perception that one almost needs to be a psychiatrist or psychologist to be able to help prevent suicide. This is unfortunate.
What, then, about all those who are not in contact with the health services prior to a suicide (11)? To quote the International Association for Suicide Prevention: «Prevention of suicide is everybody’s business» (17). Everyone can contribute to preventing suicide. This is an important message to highlight, and the constant repetition of «the 90 % truth» by influential professionals is a hindrance to this.
We wish to emphasise that we are in no way claiming to have «disproved» that there is a relationship between mental disorders and suicide. Many of those who take their own lives probably have a mental disorder. Our point is that there is no valid research evidence to claim that this applies to virtually all.
It is also important to highlight that even in some of the cases where a mental disorder clearly existed, it is by no means certain that this in itself was decisive for the suicide. Many people have been mentally ill for a long time before a suicide. This, together with the fact that by far the majority who have one or more mental disorders do not take their own lives, makes it obvious that suicide is about far more than mental disorders, and maybe about something quite different (16). It may thus be counterproductive to constantly point to «the 90 % truth», especially in contexts where there is an attempt to present a more nuanced picture, such as in the Aftenposten newspaper colour supplement «A-magasinet» of 4 April (2, 18).
In suicide prevention it is high time to focus more on the complexity that always lies behind a suicide. The biomedical illness model falls short when it comes to suicide prevention. We also need to incorporate the contextual and relational in a life course perspective if we wish to understand the nature of suicide (16).