La littérature scientifique sur le suicide est très importante. Elle est aussi très intéressante sur les liens entre suicide et troubles bipolaires.

De nombreux auteurs ont travaillé sur ce sujet, et certains ont décidé de faire une vaste synthèse de tous les saviors acquis sur cette thématique.

Le suicide chez les bipolaires, ce sont 55 articles scientifiques et 13 études sur le suicide.

Selon certaines etudes, le suicide ne relève pas de caractéristiques familiales, sociales ou professionnelles, mais ces différents facteurs de stress démultiplieraient le risque suicidaire chez les bipolaires. Etre sans emploi est un facteur accrû de risque selon certaines études, mais d’autres etudes de trouvent pas ce meme résultat, insistant plutôt sur le fait que cela n’est pas retrouvé chez les patients. Les célibataires se suicident plus que les personnes en couple, mais cela n’est pas retrouvé dans toutes les etudes. Les femmes sont plus à risque dans certaines études, les hommes dans d’autres, ce qui rend ce critère difficile à appréhender.

Il n’y aurait pas de relation entre suicide et consommation d’alcool, drogues ou troubles psychotiques pour certaines etudes. D’autres etudes trouvent l’inverse.

Les personnes souffrant de troubles bipolaires ont un très haut risque de suicide à cause de leur maladie, mais il n’y aurait pas de différence de gravité entre les formes I et II des troubles bipolaires.

Les bipolaires de type I ayant déjà fait des tentatives de suicide et ayant dans leur famille un antécédent de suicide, sont plus à risque.

Les bipolaires se suicident soit en depression, soit en état mixte, soit après une phase maniaque avec éléments psychotiques. Ce sont les trois phases les plus souvent décrites. Le nombre d’épisodes dépressifs passé et l’âge précoce du début des troubles semble un indicateur supplémentaire de risque suicidaire.

Idées noires et tentatives de suicide passées sont un indicateur de risque suicidaire reel. La composante dominante dans le passage à l’acte suicidaire serait l’impulsivité, bien plus que le désespoir. Certaines études insistent néanmoins sur le fait que le désespoir est significativement mentionné dans les tentatives de suicide et les suicides accomplish.

Les personnes bipolaires ayant subi des abus sexuels sont plus sujettes au passage à l’acte suicidaire.

Sythèse dʼarticles mondiaux

Une longue synthèse des différents articles mondiaux vous est donnée ici :

Risk factors of suicidal behaviour in bipolar disorder

Hawton and colleagues recently made a systematic review of suicide and attempted suicide in bipolar disorder (Hawton et al., 2005). Studies were selected for inclusion in their review if they met the following criteria: an international classification of disease, at least 90% of the subjects were aged 16 years or over, the design of the study was a cohort study with a minimum follow-up period of one year, a case-control or a cross-sectional study, an outcome of suicide or attempted suicide was reported, specific risk factors for suicide or attempted suicide were investigated. They found 55 articles to fulfil these criteria. The included articles reported on 13 studies of suicide: the results for three studies for bipolar patients were supplied by author. The systematical review included 23 studies of attempted suicide, of which, only one was a prospective cohort study.

Risk factors for completed suicide

Hawton and co-workers (2005) found in their meta-analysis that male gender and history of  attempted suicide and hopelessness at the index episode were associated with suicide.
Marital status or employment was not related to suicide (Hawton et al., 2005). There were no statistically significant associations of suicide with personal, social or family history characteristics (Hawton et al., 2005). Completed suicide was not related to alcohol or drug abuse, rapid cycling, psychotic symptoms or expression of suicidal ideas (Hawton et al., 2005). Furthermore, suicide risk did not seem to vary according to whether the individual had a diagnosis of bipolar I or bipolar II disorder (Hawton et al., 2005). Ösby et al. (2001) found the standardized mortality ratio for suicide was especially high for younger patients during the first years after the diagnosis. Høyer et al. (2000) found the risk of suicide was high both immediately after admission and immediately following discharge. Tsai and colleagues (2002) found in their logistic regression model that those bipolar I inpatients who have a first-degree family history of suicide, and who have had more suicide attempts (at least once in seven years of illness), are more likely to commit suicide (Tsai et al., 2002).
Isometsä and co-workers (1994) examined suicides of persons with bipolar disorder. Most suicides of persons with bipolar disorder occurred during a major depressive episode (79%), but in some cases it occurred during a mixed state (11%) or even during or immediately after remission of psychotic mania (11%) (Isometsä et al., 1994; Isometsä, 2005). Isometsä et al. (1995) have also reported that adverse life events: serious conflicts with family, substantial financial deterioration, job problems, somatic illness, residence change are related to suicide among bipolar patients.

Risk factors for attempted suicide in bipolar disorder

Hawton and co-workers (2005) found in their meta-analysis that attempted suicide was significantly more common in single individuals. Furthermore, they reported suicide attempts were associated with a family history of suicide, the earlier onset of bipolar disorder, admission to hospital resulting from depression, severity of episodes, mixed affective states at presentation, rapid cycling, comorbid anxiety disorder and alcohol and drug abuse (Hawton et al., 2005). However, there are only a few studies which have assessed a wide range of risk factors for attempted suicide.

Previous suicidal behaviour

History of attempted suicide is a strong indicator of a future suicide attempt (Bottlender et al., 2000; Nordström et al., 1995; Möller, 2003; Fawcett et al., 1990; Oquendo et al., 2000; 2004; 2006; Leverich et al., 2003; Harris and Barraclough, 1997). Suicidal ideation appears also to be an important marker for identifying patients at risk for suicide attempt (Mann, 1999). Bipolar patients with a history of suicide attempt had more severe suicidal ideation prior to admission (measured by SSI) (Oquendo et al., 2000; Galfalvy et al., 2006).

Sociodemographic characteristics

Women attempt suicide generally more often in the general population. The majority of studies have not found an association with suicide attempt and gender among bipolar patients (Galfalvy et al., 2006; Tondo et al., 1999; Tsai et al., 1999; Lopez et al., 2001; Dalton et al., 2003; Leverich et al., 2003). Oquendo and co-workers (2000) found that among hospitalized bipolar I patients men attempted suicide more often than women, whereas MacKinnon et al. (2003) reported female gender was associated with suicide attempts. The majority of previous studies have not found an association with suicide attempt and age (Tondo et al., 1999; Tsai et al., 1999; Oquendo et al., 2000). However,
Galfalvy et al. (2006) recently reported younger age is related to suicide attempt among depressed bipolar patients.
Previous studies have not found an association between suicide attempt and marital status among bipolar patients (Galfalvy et al., 2006; Tsai et al., 1999; Lopez et al., 2001; Oquendo et al., 2000), instead the Stanley Foundation Bipolar Network Study found that those who attempted suicide were significantly more commonly single individuals (Leverich et al., 2003). Oquendo et al. (2000) could not find evidence that having children protected against suicidal behaviour.
The Stanley Foundation Bipolar Network Study found that patients with a history of a suicide attempt had less college education and less income (below $ 20 000/year) (Leverich et al., 2003), whereas other studies have not found an association between suicide attempt and education (Oquendo et al., 2000; Galfalvy et al., 2006). Unemployment is similarly not associated with suicide attempt among bipolar patients (Tsai et al., 1999).
The age of onset of BD is associated with suicide attempt among bipolar disorder patients in many (Tsai et al., 1999; Lopez et al., 2001; Leverich et al., 2003; MacKinnon et al., 2003; Galfalvy et al., 2006), but not all studies (Oquendo et al., 2000; Dalton et al., 2003). Of sociodemographic factors, only the earlier onset of BD has independently predicted suicide attempt (Tsai et al., 1999).

Personal, social and family characteristics

Lifetime levels of aggression have been found to be higher among bipolar suicide attempters compared with non-attempters, measured by both the Buss-Durkee Scale and the Brown Goodwin Aggression History Scale (Oquendo et al., 2000; Galfalvy et al., 2006). However, hospitalized bipolar I attempters did not differ from non-attempters on the measure of impulsivity, the Barret Impulsive scale (Oquendo et al., 2000), whereas Galfalvy found baseline attempters manifested more impulsive traits reflected in evaluated impulsivity scores (Galfalvy et al., 2006).
In recent studies, attempted suicide is associated with a history of early physical abuse, early sexual abuse and lack of a confidant prior to illness onset (Leverich et al., 2002; 2003; Leverich and Post, 2006; Galfalvy et al., 2006). Early sexual abuse, as well as lack of a confidant prior to illness onset, remained independent risk factors for attempted suicide in multinomial statistics in the Stanley Foundation Bipolar Network Study (Leverich et al., 2003). Tsai and colleagues (1999) also reported that interpersonal problems and occupational problems predicted suicide attempt in their logistic regression model.
In the few previous studies, some aspects of family characteristics are associated with suicide attempt among bipolar patients. Leverich et al. (2003) found that a family history of substance abuse and a family history of suicide were associated with suicide attempt in bipolar patients, whereas some studies did not find any statistically significant association with family history of affective disorder or family history of suicide attempt (Dalton et al., 2003; Galfalvy et al., 2006). Lopez et al. (2001) reported family history of affective disorders to predict suicide attempt among bipolar disorder patients in their multinomial statistics (Lopez et al., 2001). Of personal, social and family characteristics, early sexual abuse, lack of a confidant prior to illness onset as well as both interpersonal and occupational problems and family history of affective disorders, has independently predicted suicide attempt.

Subtype of bipolar disorder

The prevalence of suicide attempt between bipolar I and II disorders is a controversial issue (Lester, 1993; Vieta et al., 1997; Rihmer and Pestality, 1999). Some studies (Stallone et al., 1980; Bulik et al., 1990; Tondo et al., 1999; Balazs et al., 2003) have reported higher rates of suicide attempt for bipolar II disorder, whereas the Stanley Foundation Bipolar Network Study (Leverich et al., 2003) and some other studies (Endicott et al., 1985; Coryell et al., 1987; Dalton et al., 2003) found no difference. In theory, bipolar II could carry a higher risk if the longitudinal symptomatic course of BD II is dominated more by the depressive phase of illness (Judd et al., 2003a), implying more time at risk for suicidal acts. Alternatively, possible differences related to illness episodes, such as different severity of depression, lability of mood, level of hopelessness or other characteristics, such as comorbidity, could result in risk disparities between bipolar I and II.

Hopelessness

Hopelessness, which has been researched widely as a risk factor for suicide attempt in suicidology in general, has seldom been studied among BD patients. For instance, the comprehensive meta-analysis by Hawton did not mentioned hopelessness as a risk factor of attempted suicide. One study involving hospitalised bipolar I patients showed a trend towards higher levels of hopelessness (Oquendo et al., 2000), whereas in a recent study the hopelessness was not related to suicide attempts (Galfalvy et al., 2006). Fawcett and co-workers found in their prospective study of unipolar and bipolar patients that hopelessness was a risk factor of attempted and completed suicide (Fawcett et al., 1990).

Clinical state

Depressive aspects of illness have been related to suicide attempts: higher number of prior major depressive episodes (Oquendo et al., 2000; Fagiolini et al., 2004) higher levels of depression measured by subjective ratings (BDI) or objective ratings (HAM-D) prior to admission (Oquendo et al., 2000; Fagiolini et al., 2004; Galfalvy et al., 2006), history of hospitalisation during depressive episodes (Lopez et al., 2001) and current depressive or mixed episodes (Oquendo et al., 2000; Tondo et al., 1999), whereas psychotic features have not found to be related to suicide attempts (Lopez et al., 2001). Rapid cycling is related to suicide attempt in some studies (Dalton et al., 2003; MacKinnon et
al., 2003; 2005), but not in all (Wu and Dunner, 1993; Serretti et al., 2002). Of clinical characteristics, history of hospitalization during depressive episodes has independently predicted suicide attempts.

Comorbidity

The Stanley Foundation Bipolar Network Study found suicide attempters to have a greater mean number of Axis I comorbid disorders (Leverich et al., 2003). Some previous studies (Simon et al., 2004; 2007; Leverich et al., 2003), but not all (MacKinnon et al., 2003), have found comorbid anxiety or anxiety symptoms to be related to suicide attempt. Likewise, comorbid alcohol dependence or abuse is associated with suicide attempts in many (Tondo et al., 1999; Potash et al., 2000; Goldberg et al., 2001b; Lopez et al., 2001; Slama et al., 2004), but not all studies (Oquendo et al., 2000; Leverich et al., 2003). Also, drug dependence/abuse (Dalton et al., 2003, Goldberg et al., 2001b; Tondo et al.,
1999) eating comorbidity (Leverich et al., 2003), personality disorder, according SCID-II among bipolar I patients (Ucok et al., 1998), and among bipolar II patients (Vieta et al., 1999), and Axis II comorbidities based on a self-rated questionnaire, (Leverich et al., 2003) have been associated with suicide attempts.

Risk factors of attempted suicide based on prospective studies

In the few published prospective studies, risk factors for suicide attempt have included a history of suicide attempts (Fawcett et al., 1990; Nordström et al., 1995; Oquendo et al., 2004; Galfalvy et al., 2006; Gonzalez-Pinto et al., 2006; Marangell et al., 2006), higher number of suicide attempts (Galfalvy et al., 2006), suicidal ideation (Fawcett et al., 1990; 1995; Oquendo et al., 2004; Galfalvy et al., 2006; Marangell et al., 2006), subjective rating of depression severity (Oquendo et al., 2004; Galfalvy et al., 2006), hopelessness (Fawcett et al., 1990; Oquendo et al., 2004), higher rates of illness recurrence (Gonzalez-Pinto et al., 2006), higher number of previous hospitalisations (Galfalvy et al., 2006; Gonzalez-Pinto et al., 2006), few reported reasons for living (Galfalvy et al., 2006), more days spent depressed, anxious, irritable over the past year (Marangell et al., 2006), rapid cycling (Coryell et al., 2003), comorbid alcohol dependence or abuse (Fawcett et al., 1990), cigarette smoking (Oquendo et al., 2004), borderline personality disorder (Galfalvy et al., 2006), younger age (Galfalvy et al.,
2006; Gonzalez-Pinto et al., 2006), being single (Gonzalez-Pinto et al., 2006), aggression/impulsivity (Oquendo et al., 2004; Galfalvy et al., 2006), less adherence to lithium treatment (Gonzalez-Pinto et al., 2006) and family history of mood disorder (Gonzalez-Pinto et al., 2006).

Limitations of previous prospective studies

However, these prospective studies have mostly (Coryell et al., 2003) or exclusively (Nordström et al., 1995) included inpatients with mixed and often undifferentiated types of affective disorders (Fawcett et al., 1990; Nordström et al., 1995; Oquendo et al., 2004), that have had small bipolar sample sizes (<70) (Nordström et al., 1995; Oquendo et al., 2004; Galfalvy et al., 2006) or are from tertiary care setting (Galfalvy et al., 2006). Thus, the generalizability of these findings e.g. to bipolar patients in outpatient settings, remains somewhat unclear.

Findings from multivariate statistics according prospective studies of attempted suicide

Few previous studies, except the studies by Oquendo et al. (2004), Galfalvy et al. (2006) and Marangell et al. (2006), have assessed the wide range of risk factors for attempted suicide or used multivariate statistics. Oquendo et al. (2004) found history of suicide attempts, subjective rating of the severity of depression and cigarette smoking to independently predict suicide attempt in their prospective study. Further, Galfalvy et al. (2006) found family history of suicidal acts and comorbid borderline personality disorder to predict an early attempt, while younger age, high hostility scores, higher number of past attempts, subjective pessimism as reflected in depression and suicidal ideation and few reported reasons for living, predicted suicidal acts during the whole period. Marangell et al. (2006) reported that percent days depressed in past year and history of suicide attempts independently predicted suicide attempt.