SEARCH

SEARCH BY CITATION

Keywords:

  • depression;
  • risk factor;
  • substance use disorder;
  • suicide

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Aim:  The aim of this study was to identify risk factors for suicide in Japanese substance use disorder (SUD) patients, adjusting for age and sex, and to examine sex differences in suicide risk among these patients.

Methods:  A self-reporting questionnaire on age, sex, types of abused substances, current depression, and suicidality was administered to 1420 SUD patients who consecutively visited seven hospitals specializing in SUD treatment during the month of December 2009. Unadjusted/adjusted odds ratios of factors associated with suicidality were calculated for each sex.

Results:  The multivariate analysis using the total sample identified younger age, female sex, and current depression as risk factors for severe suicidality in SUD patients. The multivariate analysis by each sex demonstrated that younger age and current depression were associated with severe suicidality in male SUD patients. Only current depression was associated with severe suicidality in female patients.

Conclusion:  Current depression is a risk factor for suicide in SUD patients common in both Western countries and Japan, although in Japanese SUD patients both younger age and female sex were more closely associated with severe suicidality than aspects of SUD. Additionally, young male SUD patients are speculated to have psychosocial features associated with suicidality in common with female SUD patients.

SUBSTANCE USE DISORDER (SUD) is an important risk factor associated with suicide. Many psychological autopsy studies have estimated that SUD is the second most common mental disorder in suicide-completers, following depressive disorder.1,2 DeLeo and Evans3 proposed that SUD may increase a patient's suicide risk through both indirect and direct mechanisms. An indirect mechanism operates when SUD causes unemployment, divorce, separation from family, imprisonment, and/or psychiatric problems, worsening one's economic and psychosocial situations. A direct mechanism operates when the pharmacological effects of the abused substance disinhibit impulsivity, promoting aggression toward oneself.

Several Western studies4–6 have identified risk factors for suicidal behavior in SUD patients. A meta-analysis by Harris and Barraclough4 compared standardized mortality rates for suicide among SUD patients according to substances abused, reporting that patients who abused prescribed psychotropic drugs, opiates, or multiple substances had higher suicide mortality rates than those with alcohol or cannabis use disorder. An epidemiologic study using national survey data5 estimated that individuals abusing alcohol, inhalants, and opiates tend to attempt suicide more frequently, though suicidality is more closely associated with abusing multiple substances than with abusing any particular type of substance. A cohort study conducted by Davis et al.6 indicated that the co-occurrence of depression might heighten the suicide risk of individuals with SUD.

However, it is unclear whether risk factors identified through Western studies also apply to Japanese SUD patients because the types of substances commonly abused in Japan are different from those commonly abused in Western countries. For example, few SUD patients in Japan abuse opiates unlike in Western countries, while methamphetamine has been the most serious drug-related problem since the 1950s in Japan.7 It is the most abused substance other than alcohol among Japanese SUD patients.8

In this context, we previously revealed that Japanese patients with methamphetamine use disorders or who abused prescribed psychotropic drugs were more likely to have a history of attempted suicide and the presence of suicide ideation than those with alcohol use disorder.8 However, that study has two important shortcomings. First, SUD patients simultaneously abusing more than two types of substances were not considered because they were excluded from the subjects to be analyzed. Second, the comparison was not adjusted for age and sex. In our previous study,8 the methamphetamine abusers were younger than alcohol abusers, and those who abused prescribed psychotropic drugs included a larger proportion of female subjects. These findings suggest that severe suicidality of SUD patients may not be associated with types of abused substances, but with age and sex. To our knowledge, there have been no Japanese studies that have addressed these shortcomings.

The purpose of the present study was to identify risk factors for suicide in Japanese SUD patients adjusting for age and sex, and to examine sex differences in suicide risk among these patients.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Subjects

A pool of 1650 outpatients with a DSM-IV-TR diagnosis of psychoactive substance use disorders who consecutively visited seven hospitals specializing in SUD treatment during the month of December 2009 was identified. These hospitals were representative medical facilities for SUD treatment located in five areas of Japan: three in the Kanto area (including metropolitan Tokyo) and one each in the Tohoku/Hokkaido, Tokai/Hokuriku, Chugoku/Shikoku, and Kyushu/Okinawa areas. A total of 1420 SUD patients (86.1%; 1113 men and 307 women; mean age ± SD: total, 50.5 ± 13.3 years; men, 52.3 ± 12.9 years; women, 43.0 ± 11.7 years) consented to participate.

This study was approved by the ethics committee of the National Center of Neurology and Psychiatry.

Variables

Age, sex, and types of substances abused

A self-reporting questionnaire comprising questions on sex, age, and types of substances abused was administered to all participants.

The question used to identify types of abused substances was ‘Which types of substances caused the problems for which you received hospital treatment?’ Participants were requested to select all types of substance abuse applicable from four categories: alcohol abuse, methamphetamine abuse, other illicit drug abuse (cannabis, opiates, organic solvents, and various psychedelics, including lysergic acid diethylamide and 3,4-methylenedioxy-N-methylamphetamine), and abuse of prescribed psychotropic drugs (hypnotics and anti-anxiety drugs). If participants selected more than two substances, the category used was multi-substance abuse.

Current co-occurrence of depressive disorder

The Kessler 10 (K10), a brief self-reporting questionnaire that screens for depressive disorder,9 was employed to assess current co-occurrence of depressive disorder. This scale consists of 10 items used to identify symptoms of depression experienced within the last week. The validity and reliability of the Japanese version have been established, and a score of more than 25 points indicates the presence of DSM-IV major depressive disorder.10 We defined a score greater than 25 as current co-occurrence of depression in this study.

Suicidality

The suicide risk subsection of the Japanese version 5.0.0 of the Mini International Neuropsychiatric Interview (M.I.N.I),11 an established, structured interview schedule that screens for psychiatric disorders,12 was employed to assess suicidality. This subsection consists of six items that identify any suicide-related episodes or phenomena, including suicidal ideation and suicide attempts, within the last month. In the Japanese version of the ‘M.I.N.I.’ booklet, written by Sheehan and Lecrubier, and translated by Otsubo et al.,13 the scores for each answer are weighted according to their importance in predicting future suicide (e.g. lifetime histories of attempting suicide = 4; presence of having suicidal ideation within a month = 6; planning or attempting suicide within a month = 10), although the validity of this scoring system has not been completely established.11,12 According to Sheehan and Lecrubier,13 a total score of more than 10 points out of the possible 33 indicates a high risk for suicide, while a score from 1 to 5 and from 6 to 9 indicates a low and moderate risk, respectively.

In the present study, we used each of these six items in the self-reporting questionnaire, and defined a high risk (M.I.N.I. suicide risk ≥10) as severe suicidality. Their internal consistency was established in our sample (Cronbach's α was 0.772).

Statistical analyses

The self-reporting questionnaire was completed anonymously and collected immediately. To determine which factors were associated with severe suicidality, logistic regression analysis was employed for the seven variables (age, current depression, alcohol abuse, methamphetamine abuse, other illicit drug abuse, abuse of prescribed psychotropic drugs, and multi-substance abuse), and the unadjusted and adjusted odds ratios (OR) were calculated for both sexes and for each sex separately. P-values less than 0.05 were accepted as indicating significance, and all were two-tailed. All statistical analyses were performed using spss software for Windows (version 17.0; spss, Chicago, IL, USA).

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Of the 1420 SUD patients, 1118 (78.7%) reported alcohol abuse, 190 (13.4%) reported methamphetamine abuse, 62 (4.7%) reported abuse of prescribed psychotropic drugs, 25 (1.8%) reported other illicit drug abuse, and 171 (12.0%) met the conditions for multi-substance abuse. The mean score (SD) of the K10 was 22.6 (10.3) in the total sample (men, 21.3 [9.6]; women, 27.2 [11.1]), and that of M.I.N.I. suicide risk subsection was 9.6 (12.1) (men, 8.0; women, 15.3 [13.3]). Additionally, 534 (37.6%) scored above the limit for depression on the K10, and 495 (34.9%) showed severe suicidality on the M.I.N.I. suicide risk subsection.

Table 1 shows the results of logistic regression analyses for severe suicidality adjusted by sexes. Bivariate analysis demonstrated that younger age, female sex, current depression, methamphetamine abuse, abuse of prescribed psychotropic drugs, and multi-substance abuse were significantly positively associated and alcohol abuse was significantly negatively associated with severe suicidality in SUD patients. Multivariate analysis identified four of these variables, younger age, female sex, current depression, and multi-substance abuse, as being significantly associated with severe suicidality.

Table 1. Logistic regression analyses for severe suicidality of outpatients with substance use disorder (n = 1420)
Dependent variablesIndependent variablesSevere suicidality (n = 495)Non-severe suicidality (n = 925)Bivariate analysisMultivariate analysis
MeanSDMeanSDBUnadjusted OR95%CIBAdjusted OR95%CI
  1. *P < 0.05, **P < 0.01, ***P < 0.001. M.I.N.I., Mini International Neuropsychiatric Interview; OR, odds ratio, CI, confidence interval.

Severe suicidality (M.I.N.I. suicide risk ≥ 10)Age (years)45.111.953.013.0−0.0490.952***0.943–0.961−0.0270.973***0.962–0.985
  n % n %      
Sex (female percentage: male = 0, female = 1)161/49532.5%149/92016.2%0.9382.555***1.974–3.3080.4521.571**1.150–2.146
Current depression (K10 score ≥25; absence = 0, presence = 1)346/49270.3%188/91320.6%2.2139.139***7.108–11.7501.9847.271***5.580–9.476
Alcohol abuse (absence = 0, presence = 1)350/49570.7%768/92083.5%−0.7390.478***0.368–0.6200.1001.1060.726–1.684
Methamphetamine abuse (absence = 0, presence = 1)94/49519.0%96/92010.4%0.6992.012**1.478–2.3790.1401.1510.720–1.839
Other illicit drug abuse (absence = 0, presence = 1)8/4951.6%17/9201.8%−0.1360.8730.374–2.036−0.5210.5940.211–1.674
Prescribed-psychotropic drug abuse (absence = 0, presence = 1)34/4956.9%28/9203.0%0.8542.350**1.407–3.9230.0951.1000.579–2.091
Multi-substance abuse (absence = 0, presence = 1)99/49520.0%72/9207.8%1.0802.944***2.125–4.0800.4901.632*1.107–2.405

Table 2 shows the results of logistic regression analyses for severe suicidality in male SUD patients. Bivariate analysis demonstrated that younger age, current depression, methamphetamine abuse, abuse of prescribed psychotropic drugs, and multi-substance abuse were significantly positively associated and alcohol abuse was significantly negatively associated with severe suicidality. Multivariate analysis identified two of these variables, younger age and current depression, as being significantly associated with severe suicidality in male subjects.

Table 2. Logistic regression analyses for severe suicidality of male outpatients with substance use disorder (n = 1113)
Dependent variablesIndependent variablesSevere suicidality (n = 334)Non-severe suicidality (n = 779)Bivariate analysisMultivariate analysis
MeanSDMeanSDBUnadjusted OR95%CIBAdjusted OR95%CI
  1. **P < 0.01, ***P < 0.001. M.I.N.I., Mini International Neuropsychiatric Interview; OR, odds ratio, CI, confidence interval.

Severe suicidality (M.I.N.I. suicide risk ≥ 10)Age (years)47.611.854.412.8−0.0440.957***0.947–0.968−0.2800.972***0.959–0.985
  n % n %      
Current depression (K10 score ≥25; absence = 0, presence = 1)227/33168.6%141/76918.3%2.2659.628***7.195–12.8832.1078.227***6.039–11.207
Alcohol abuse (absence = 0, presence = 1)249/33474.6%660/77485.3%−0.6810.506***0.369–0.6940.2311.2600.748–2.121
Methamphetamine abuse (absence = 0, presence = 1)57/33417.1%70/7749.0%0.7352.086***1.433–3.0360.2371.2680.710–2.263
Other illicit drug abuse (absence = 0, presence = 1)6/3341.8%16/7742.1%−0.1340.8750.336–2.255−0.5420.5820.184–1.841
Prescribed-psychotropic drug abuse (absence = 0, presence = 1)21/3346.3%19/7742.5%0.9892.690**1.426–5.0720.4881.6290.736–3.603
Multi-substance abuse (absence = 0, presence = 1)61/33418.3%55/7747.1%1.0722.921***1.977–4.3150.4041.4980.939–2.390

Table 3 shows the results of logistic regression analyses for severe suicidality in female SUD patients. Bivariate analysis demonstrated that younger age, current depression, and multi-substance abuse were significantly positively associated and alcohol abuse was significantly negatively associated with severe suicidality. Multivariate analysis identified only one of these variables, current depression, as being significantly associated with severe suicidality in female subjects.

Table 3. Logistic regression analyses for severe suicidality of female outpatients with substance use disorder (n = 307)
Dependent variablesIndependent variablesSevere suicidality (n = 161)Non-severe suicidality (n = 146)Bivariate analysisMultivariate analysis
MeanSDMeanSDBUnadjusted OR95%CIBAdjusted OR95%CI
  1. *P < 0.05, **P < 0.01, ***P < 0.001. M.I.N.I., Mini International Neuropsychiatric Interview; OR, odds ratio, CI, confidence interval.

Severe suicidality (M.I.N.I. suicide risk ≥10)Age (years)40.410.745.812.1−0.0440.957***0.937–0.977−0.0220.9790.954–1.004
  n % n %      
Current depression (K10 score ≥25; absence = 0, presence = 1)119/16173.9%47/14432.6%1.7585.800***3.568–9.4281.6825.376***3.182–9.082
Alcohol abuse (absence = 0, presence = 1)101/16162.7%108/14674.0%−0.5240.592*0.363–0.965−0.1710.8430.409–1.740
Methamphetamine abuse (absence = 0, presence = 1)37/16123.0%26/14617.8%0.3631.4230.814–2.487−0.0580.9440.423–2.104
Other illicit drug abuse (absence = 0, presence = 1)2/1611.2%1/1460.7%0.6011.8240.164–20.3270.0291.0290.063–16.880
Prescribed psychotropic drug abuse (absence = 0, presence = 1)13/1618.1%9/1466.2%0.2901.3370.554–3.227−0.5530.5750.201–1.649
Multi-substance abuse (absence = 0, presence = 1)38/16123.6%17/14611.6%0.8522.344**1.257–4.3710.6541.9220.946–3.907

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

To the best of our knowledge, this is the first study to identify risk factors for suicide in Japanese SUD patients adjusting for age and sex, and to examine sex differences in suicide risk among these patients. Although previous Western studies have identified several risk factors for suicide, including abusing prescribed psychotropic drugs,4 abusing multiple substances,5 and co-occurrence of depression,6 such findings were obtained from samples that included few methamphetamine abusers.

In the present study, multivariate analysis identified three variables, younger age, female sex, and current depression, as suicide risk factors in Japanese SUD patients, although bivariate analysis demonstrated that, in addition to these, methamphetamine abuse, abuse of prescribed psychotropic drugs, and multi-substance abuse were significantly associated with severe suicide risk, as previous studies4,5,8 have indicated. These findings suggest that suicide risk in SUD patients may be more closely associated with two demographic variables and one psychiatric disorder other than SUD type (e.g. methamphetamine or prescribed psychotropic drugs) and number of abused substances (e.g. simultaneously abusing more than two types of substances).

Although it is not surprising that current depression is an important risk factor for suicide in SUD patients common in both Western countries1,2,14 and Japan,15 it is of interest that both younger age and female sex were more closely associated with severe suicidality than aspects of SUD. We speculate that these demographic factors may indirectly reflect serious psychosocial problems that the SUD patients have been experiencing. As to younger age, Cloninger et al.16 and Brown et al.17 suggested that early onset of SUD may be promoted by childhood episodes of attention-deficit/hyperactivity disorder (ADHD) and other conduct problems, and a birth cohort study by Sourander et al.18 reported that hyperkinetic tendencies and conduct problems at the age of 8 years may be predictive of suicide in men in early adulthood, as well as early onset of substance abuse.

On the other hand, as to female sex, Lacy and Evans19 found that a substantial number of female SUD patients showed self-destructive behaviors, including repetitive self-injury and repeated suicide attempts accompanied by comorbid eating disorders, particularly bulimia nervosa, and conceptualized this clinical subgroup of female SUD patients as ‘multi-impulsive bulimia.’ Majewska20 and Clark et al.21 also revealed that female SUD patients were more likely to have been childhood victims of physical, sexual, and psychological abuse, and neglect, all of which have been identified as important risk factors for suicidal behavior in adulthood.22

Based on these findings, we suppose that the two demographic variables, younger age and female sex, may reflect the SUD patients with early onset of substance abuse impacted by psychosocial problems in childhood and adolescence (e.g. presence of ADHD and the other conduct problems, and ‘multi-impulsive bulimia’, and childhood histories of maltreatment, as risk factors identified by the Western studies16–21). In this context, younger age and female sex may indicate a subgroup with severe suicidality in Japanese SUD patients similar to that in Western SUD patients. Future studies that include detailed information of childhood psychosocial events are necessary to confirm this argument.

The present study also demonstrated sex differences in suicide risk in SUD patients. Although current depression was a common risk factor for suicidality in both male and female SUD patients, younger age was a risk factor for suicidality only in male patients, and no risk factors other than current depression were identified in female patients. Perhaps the female SUD patients had similar psychosocial features to the younger population of male patients given that the female patients were considerably younger than the men (men: 52.3 ± 12.9 years; women: 43.0 ± 11.7, P < 0.001). This leads us to speculate that young male SUD patients may have psychosocial features associated with suicidality in common with female SUD patients.

This study has several limitations. First, sampling bias cannot be excluded, given that subjects were drawn only from patients who consulted one of seven hospitals specializing in SUD treatment, though each of these hospitals is a representative medical institution for its region. Second, data were cross-sectionally acquired via self-reporting questionnaires rather than through structured interviews. Third, the outcome of the present study was ‘severe suicide risk’ as determined by the M.I.N.I, not prognosis of suicide completion. Fourth, the M.I.N.I. was employed as self-reporting questionnaire in this study, although it was originally developed as a semi-structured interview schedule. Fifth, although in the suicidality section of the M.I.N.I. the scores for each answer are weighted according to their importance in predicting future suicide, the validity of this scoring system has not been completely established.11–13 Lastly, comorbid psychiatric disorders other than depressive disorder were not considered, and data on psychosocial and economic variables were insufficient for all patients. Despite these limitations, this report is of value because it is the first to identify suicide risk factors among Japanese SUD patients, adjusting for age and sex differences.

Conclusion

The present study demonstrated that younger age, female sex, and current depression were independent risk factors for severe suicidality in Japanese SUD patients. Although current depression is a risk factor for suicide in SUD patients common in both Western countries and Japan, in Japanese SUD patients both younger and female subjects were more closely associated with severe suicidality than aspects of SUD.

Our study also showed that current depression was a common risk factor for suicidality in both male and female SUD patients while only in male patients, younger age was a risk factor for suicidality. Young male SUD patients are speculated to have a risk of suicide equal to that of female SUD patients.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

This study has been supported by the Health Labour Sciences Research Grant by the Ministry of Health and Welfare, Comprehensive Research on Disability Health and Welfare, ‘A study on epidemiology of high-risk group for suicide and suicide prevention (Principal Investigator, Itoh H).’ We declare no conflict of interest that may be inherent in this study.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES
  • 1
    Chynoweth R, Tonge JI, Armstrong J. Suicide in Brisbane: A retrospective psychosocial study. Aust. N. Z. J. Psychiatry 1980; 14: 3745.
  • 2
    Lönnqvist JK, Henriksson MM, Isometsä ET et al. Mental disorders and suicide prevention. Psychiatry Clin. Neurosci. 1995; 49: S111S116.
  • 3
    De Leo D, Evans R. The impact of substance abuse policies on suicide mortality. In : De Leo D, Evans R (eds). International Suicide Rates and Prevention Strategies. Hogrefe & Huber, Cambridge, MA, 2004; 101112.
  • 4
    Harris EC, Barraclough B. Suicide as an outcome for mental disorders. A meta-analysis. Br. J. Psychiatry 1997; 170: 205228.
  • 5
    Borges G, Walters EE, Kessler RC. Associations of substance use, abuse, and dependence with subsequent suicidal behavior. Am. J. Epidemiol. 2000; 151: 781789.
  • 6
    Davis LL, Rush JA, Wisniewski SR et al. Substance use disorder comorbidity in major depressive disorder: An exploratory analysis of the Sequenced Treatment Alternatives to Relieve Depression cohort. Compr. Psychiatry 2005; 46: 8189.
  • 7
    Matsumoto T, Kamijo A, Miyakawa T et al. Methamphetamine in Japan: The consequences of methamphetamine abuse as a function of route of administration. Addiction 2002; 97: 809818.
  • 8
    Matsumoto T, Matsushita S, Okudaira K et al. A study on differences of suicide risk in substance use disorder patients by types of an abused substance: A comparison among alcohol, amphetamine, and sedative, hypnotic or anxiolytic use disorder patients. Jpn. J. Alcohol. Drug Stud. 2010; 45: 530542 (in Japanese).
  • 9
    Kessler RC, Barker PR, Colpe LJ et al. Screening for serious mental illness in the general population. Arch. Gen. Psychiatry 2003; 60: 184189.
  • 10
    Furukawa TA, Kessler RC, Slade T, Andrews G. The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychol. Med. 2003; 33: 357362.
  • 11
    Otsubo T, Tanaka K, Koda R et al. Reliability and validity of Japanese version of the Mini International Neuropsychiatric Interview. Psychiatry Clin. Neurosci. 2005; 59: 517526.
  • 12
    Sheehan DV, Lecrubier Y, Sheehan KH et al. The Mini International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 1998; 59: 2233.
  • 13
    Sheehan DV, Lecrubier Y. (Translated by Otsubo T, Miyaoka H, Kamijima K) M.I.N.I. Mini International Neuropsychiatric Interview. Japanese version 5.0.0, Seiwa Shoten, Tokyo, 2003.
  • 14
    Barraclough B, Bunch J, Nelson B, Sainsbury P. A hundred cases of suicide: Clinical aspects. Br. J. Psychiatry 1974; 125: 355373.
  • 15
    Hirokawa S, Matsumoto T, Katsumata T et al. Psychosocial and psychiatric characteristics of suicide completers who consulted a psychiatrist prior to death: A psychological autopsy study. Jpn. Bull. Soc. Psychiatry 2010; 18: 341351 (in Japanese).
  • 16
    Cloninger C, Sigvardsson S, Bohman M. Childhood personality predicts alcohol use in young adults. Alcoholism 1988; 12: 494505.
  • 17
    Brown SA, Cleghorn A, Schuckit MA, Myers MG, Mott MA. Conduct disorder among adolescent alcohol and drug abusers. J. Stud. Alcohol 1996; 57: 314324.
  • 18
    Sourander A, Klomek AB, Niemelä S. Childhood predictors of completed and severe suicide attempts: Findings from the Finnish 1981 Birth Cohort Study. Arch. Gen. Psychiatry 2009; 66: 398406.
  • 19
    Lacey JH, Evans CD. The impulsivist: A multi-impulsive personality disorder. Br. J. Addict. 1986; 81: 641649.
  • 20
    Majewska MD. Cocaine addiction as a neurological disorder: Implications for treatment. NIDA Res. Monogr. 1996; 163: 126.
  • 21
    Clark DB, Lesnick L, Hegedus AM. Traumas and other adverse life events in adolescents with alcohol abuse and dependence. J. Am. Acad. Child Adolesc. Psychiatry 1997; 36: 17441751.
  • 22
    Bruffaerts R, Demyttenaere K, Borges G et al. Childhood adversities as risk factors for onset and persistence of suicidal behavior. Br. J. Psychiatry 2010; 197: 2027.