1 Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland, and
2 Department of Psychiatry, University of Helsinki, Helsinki, Finland
Received 17 February 1999; in revised form 1 June 1999; accepted 25 June 1999
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
SUBJECTS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The database included 1397 suicide cases. Information from next-of-kin interviews was available for 1155 cases, of which 12 were excluded because of incomplete interview forms. Among the remaining 1143 cases there was reliable and structured information about the use of alcohol for 997 victims, who formed the present study sample. These cases did not differ from the excluded victims in age or sex, but were more often married or cohabiting (58.1% vs 40.4%).
Alcohol misuse
The victim was recorded as having misused alcohol if reported by the informant to have been in an obvious state of drunkenness at least once or twice a week during the last year (Heikkinen et al., 1995a,b
). A total of 349 cases were considered as misusers: 41.2% (n = 317/769) of the males and 14.0% (n = 32/228) of the females. The remaining 648 cases were thus treated as non-misusers.
To validate the concept of misuse, we investigated the agreement between informant-based misuse data and the respective DSM-III-R (American Psychiatric Association, 1987) best-estimate consensus alcohol dependence diagnoses. These had been previously assigned in a diagnostic study of a systematic random sample of 229 suicide cases, and based on all available information, including multiple interviews and health care records. In the diagnostic procedure, DSM-III-R diagnoses were assessed by two independent reviewers, and a consensus meeting was held in cases of diagnostic disagreement (Henriksson et al., 1993
). Estimates of possible alcohol misuse were available for 173 of these 229 cases: 35% (60/173) were misusers and 65% (113/173) non-misusers. Misuse was assessed in 79% (n = 48/61) of the alcohol-dependent subjects, and 80% (n = 48/60) of misusers had alcohol dependence (kappa for inter-rater reliability between misuse and alcohol dependence was 0.68). For a supplementary analysis we also examined these 173 misusers and non-misusers belonging to the diagnostic sample with regard to some diagnostic and other characteristics which had been individually assessed according to all available information.
In the total population of suicides, we previously reported frequent alcohol misuse among males living alone (Heikkinen et al., 1995a), and that alcohol misusers who were currently employed were more likely to commit suicide during weekends than the unemployed (Pirkola et al., 1997
).
Sociodemographic factors
In addition to five-class categorized data on occupational status, a dichotomy was created according to whether the victim had (regardless of current employment status) regularly worked as an entrepreneur, upper or lower level employee or manual worker. Current employment and marital status were recorded, and also whether the victim had had children.
Recent life events
The interview with the next-of-kin included a separate life-event questionnaire based on the Recent Life Change Questionnaire instrument by Rahe (1977), with some modifications from the list by Paykel et al. (1969). The 32 items covering events within the 3 months prior to death were classified on logical grounds into either (1) events independent of the victim's own behaviour; or (2) events possibly dependent on the victim's own behaviour. The independent events included death or severe illness of a family member, while examples of possibly dependent events were separation, serious financial setback or job problems. Included items were combined into larger categories in which more than one event was counted only once for that category. Details of the procedure have been published (Isometsä et al., 1995)
Previous suicidality and psychiatric treatment
Whether or not the victim had previously communicated or attempted suicide was recorded, as well as lifetime psychiatric inpatient treatment or psychiatric treatment overall.
Blood-alcohol concentration and suicide method
Blood-alcohol concentrations were determined at forensic autopsy for 95.1% (n = 948/997) of the victims, and a concentration of 0.10% (100 mg/dl) at the time of suicide was considered a state of alcohol intoxication. Hanging, shooting, cutting, jumping from a high place or in front of a vehicle, and deliberately crashing a vehicle were considered violent suicide methods. Fatal overdose using a solid or liquid (other than alcohol) substance was also recorded.
Statistical methods
In comparisons between groups, the 2-test with Yates' correction and Student's t-test were performed, both two-tailed.
The differences between the groups were analysed via univariate tables, but to control for confounding factors and to summarize the results, a backwards stepwise (likelihood ratio test) logistic model was created including the major categories of characteristics studied. Alcohol misuse was the dependent variable, and the independent variables among all major characteristics studied were: male sex, age (continuous), marital status (divorced or separated), any lifetime work, suicide communication, overdose as the suicide method, lifetime psychiatric hospitalization, recent (within 3 months) unemployment, financial problems or interpersonal loss or conflict, and alcohol detected in blood at autopsy 0.10%. To create a model with a large number of variables, the probability for removal was set at P > 0.10.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Life events
More of the misusers had experienced any life events within the last 3 months (87 vs 76%, 2 = 15.66, df = 1, P < 0.0001); four-fifths had had life events possibly dependent on their own behaviour compared to only half of the non-misusers (Table 1
). Within the last 3 months, more of the male misusers compared with non-misusers had experienced interpersonal loss (38 vs 18%,
2 = 32.53, df = 1, P < 0.0001), loss or conflict (55 vs 27%,
2 = 52.03, df = 1, P < 0.0001), financial problems (27 vs 13%,
2 = 20.03, df = 1, P < 0.0001), or unemployment (27 vs 11%,
2 = 31.03, df = 1, P < 0.0001). Among females, more of the misusers had experienced financial problems (29 vs 9%,
2 = 7.84, df = 1, P = 0.005) or unemployment (29 vs 3%,
2 = 21.45, df = 1, P < 0.0001).
Psychiatric contacts
The misusers had psychiatric hospitalizations and any psychiatric treatment less frequently than the non-misusers, but previous suicide attempts and suicide communication more often (Table 1). Among both groups, those without any lifetime psychiatric contacts were more likely to be intoxicated at the time of suicide than those with any psychiatric treatment (66 vs 48%,
2 = 10.60, df = 1, P = 0.001, 19 vs 9%,
2 = 10.91, df = 1, P = 0.001).
Blood-alcohol concentration and suicide method
Suicide methods and the proportions of victims with alcohol in the blood appear in Table 2.
|
Psychiatric comorbidity among misusers and non-misusers in the diagnostic sample
Within the diagnostic sample, some differences were found between the misusers (n = 60) and non-misusers (n = 113). A DSM-III-R depressive disorder not otherwise specified was more common among the misusers (35%, n = 21/60 vs 19%, n = 21/113, 2 = 4.89, df = 1, P = 0.027), whereas major depression (MD) was more common among the non-misusers (42%, n = 47/113 vs 22%, n = 13/60,
2 = 6.02, df = 1, P = 0.014). Misusers were more likely to have been diagnosed with any DSM-III-R personality disorder (43%, n = 26/60 vs 26%, n = 29/113,
2 = 5.45, P = 0.02) and less often with a non-organic psychosis (psychotic MD included) (13%, n = 8/60 vs 37%, n = 42/113,
2 = 9.71, df = 1, P = 0.001). Regular treatment with antipsychotic drugs had been prescribed in 8% (n = 5/60) of the misusers and 30% (n = 34/113) of the non-misusers (
2 = 9.41, df = 1, P = 0.0022), the respective figures for benzodiazepine treatment being 8% (n = 5/60) and 28% (n = 32/113) (
2 = 8.16, df = 1, P = 0.004) and for antidepressant medication 7% (n = 4/60) and 20% (n = 22/113) (
2 = 4.08, P = 0.04).
The logistic model
The logistic model (Table 3) predicted alcohol misuser status in 78%. Male sex, being divorced or separated, suicide communication, overdose as the method, any lifetime work, unemployment, financial problems or interpersonal loss or conflict within 3 months, and alcohol intoxication at the time of suicide, were partially associated with alcohol misuse (Table 3
).
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Alcohol misusers
Compared to the non-misusing suicide victims, the alcohol misusers tended to be younger, male, more frequently divorced or separated from a marriage and to have children. They were more likely to have worked, usually in manual jobs, whereas the non-misusers were very often retired. In accordance with observations by Hayward et al. (1992), the misusers had received psychiatric treatment less often. These findings suggest a somewhat higher psychosocial status among the misusers, and perhaps a smaller burden of long-standing psychiatric morbidity and psychosocial stress. However, they were also more likely to have experienced disruption in their lives and committed suicide while drunk. The problems of the misusers in the areas of finance, work and personal relationships were doubtless partly a consequence of their excessive drinking, although perhaps also related to their more prevalent personality disorders. These factors are generally known predictors of poor outcome (Heikkinen et al., 1995b; Pirkola et al., 1999
). Moreover, the corrosive influence of habitual misuse may have heightened the vulnerability to suicide in cases with recent adversity. As an example, 75% had regularly worked at some time, but only half were currently employed, a quarter having experienced unemployment within the last 3 months. These findings in the suicidal process of an alcohol misuser imply a relatively smaller loading of long-standing contributing factors other than problematic alcohol use per se with its both short- and long-term consequences.
Alcohol and the method of suicide
With regard to the final act of suicide, alcohol has been suggested to impair judgement, cause impulsivity and contribute to the choice of method (Welte et al., 1988; Hayward et al., 1992
; Öhberg et al., 1996
). In spite of our finding that the misusers of both sexes had far less often been prescribed psychopharmacological medication, death by intoxication with a liquid or solid substance tended to characterize their suicides. The possible potentiating effect of alcohol on the lethality of another substance used for suicide may partly explain these intoxication suicides as well as the high proportion of misusers with alcohol in the blood.
The role of recent life events
The analysis within the random sample showed that the alcohol misusers in this study resembled alcohol-dependent suicide victims studied earlier (Murphy and Robins, 1967; Murphy, 1992
; Heikkinen et al., 1994
), and thus partly confirmed these findings in an unselected nationwide population. Recent interpersonal and employment problems were associated with alcohol misuse, which also accords with previous studies. The misusers had more events possibly dependent on their own behaviour, probably relating to psychosocial disruption in the structure of their personal life caused by excessive drinking, and maladaptive functioning in cases of comorbid personality disorders. Moreover, a specific vulnerability and current drinking may well be important contributors to the final act (Murphy et al., 1992
).
Studies on alcohol consumption and suicide rate
According to Mäkelä (1996), the suicide rate of younger Finnish males is associated with alcohol consumption, and similar findings are reported elsewhere (Caces and Harford, 1998). In Finland, heavy drinking and intoxication account for a large proportion of overall alcohol consumption, and alcohol-related problems concentrate in a heavy drinking minority (Simpura, 1987
). The alcohol-misusing suicide victims in our nationwide psychological autopsy study were almost exclusively male and three-quarters were aged <50 years. While it is tempting to see them as a key group with regard to temporal variation in suicide rates, they had other psychiatric morbidity (Henriksson et al., 1993
; Pirkola et al., 1999
), recent adverse changes and disintegration in their life structure, besides a history of abundant drinking. In general, when combined with psychiatric disorders, alcohol misuse probably predisposes an individual to a particular type of suicide process characterized by disturbances in life course. In this sense, alcohol consumption via alcohol-related problems among individuals at risk may have the potential to cause variation in suicide rates, and thus to explain part of the international differences in them.
The non-misusing victims
It is likely that the population of non-misusers in this study consisted of heterogeneous, but basically more psychiatrically ill, victims than the misusers. The supplementary analysis within the diagnosed random sample supports this, revealing more major depressive disorders and non-organic psychotic disorders, as well as more antipsychotic and benzodiazepine treatment among the non-misusers. The association of recent unemployment and alcohol intoxication among the non-misusers may reflect reactive drinking to adversity among some individuals, which is interesting in the light of general reports on unemployment and suicide (Pritchard, 1992).
Methodological issues
Some methodological limitations need to be noted. First, there were no living controls from the general population. Second, by necessity, only retrospective and second-hand information is available in psychological autopsy studies. In the evaluation of life events, recall bias, a tendency to explain suicide (effort after meaning) and the limits of the questionnaire may have caused either under- or over-reporting of events. Furthermore, although the classification of possibly dependent and independent events was made on logical grounds, the true intentionality of dependent events is beyond reach. These problems have been discussed earlier (Miller et al., 1986; Heikkinen et al., 1994
; Isometsä et al., 1995
). Third, a number of cases had to be excluded from the total suicide population on the basis of missing, incomplete or unreliable interviews. In addition, some cases among those with interviews had insufficient reliable and structured information about the victim's use of alcohol, and were also excluded. However, these individuals did not differ from the included cases in terms of sex, age, or current employment, though they tended to be married or cohabiting less often. Fourth, the size of the study population precluded detailed examination of individual alcohol use in the light of all available information, so the informant's estimate of the victim's pattern of heavy drinking was the sole source used to categorize their alcohol use, which may have led to under-reporting of any problems with alcohol. Finally, due to the relatively high number of univariate tests, the possibility of spurious associations cannot be fully excluded. Nevertheless, we think that the size of the population and use of a logistic model offered some protection against this possibility. Furthermore, the convergence of the findings with what is already known supports their reliability.
![]() |
GENERAL CONCLUSIONS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Caces, F. E. and Harford, T. (1998) Time series analysis of alcohol consumption and suicide mortality in the United States, 19341987. Journal of Studies on Alcohol 59, 455461.[ISI][Medline]
Clark, D. and Horton-Deutsch, S. (1992) Assessment in absentia: The value of the psychological autopsy method for studying antecedents of suicide and predicting future suicides. In Assessment and Prediction of Suicide, Maris, R., Berman, A., Maltsberger, J. and Yufit, R. eds, pp. 144182. Guilford Press, New York.
Harris, E. C. and Barraclough, B. (1997) Suicide as an outcome for mental disorders. A meta-analysis. British Journal of Psychiatry 170, 205228.[Abstract]
Hayward, L., Zubrick, S. and Silburn, S. (1992) Blood alcohol levels in suicide cases. Journal of Epidemiology and Community Health 46, 256260.[Abstract]
Heikkinen, M. E., Aro, H. M., Henriksson, M. M. et al. (1994) Differences in recent life events between alcoholic and depressive nonalcoholic suicides. Alcoholism: Clinical and Experimental Research 18, 11431149.[ISI][Medline]
Heikkinen, M. E., Isometsä, E. T., Marttunen, M. J. et al. (1995a) Social factors in suicide. British Journal of Psychiatry 167, 747753.[Abstract]
Heikkinen, M. E., Isometsä, E. T., Aro, H. M. et al. (1995b) Age-related variation in recent life-events preceding suicide. Journal of Nervous and Mental Disease 183, 325331.[ISI][Medline]
Henriksson, M. M., Aro, H. M., Marttunen, M. J. et al. (1993) Mental disorders and comorbidity in suicide. American Journal of Psychiatry 150, 935940.[Abstract]
Inskip, H. M., Harris, E. C. and Barraclough, B. (1998) Lifetime risk of suicide for affective disorder, alcoholism and schizophrenia. British Journal of Psychiatry 172, 3537.[Abstract]
Isometsä, E. T., Heikkinen, M. E., Henriksson, M. M. et al. (1995) Recent life events and completed suicide in bipolar affective disorder. A comparison with major depressive suicides. Journal of Affective Disorders 33, 99106.[ISI][Medline]
Mäkelä, P. (1996) Alcohol consumption and suicide mortality by age among Finnish men, 19501991. Addiction 91, 101112.[ISI][Medline]
Miller, Mc, C., Dean, P., Ingham, J. G. et al. (1986) The epidemiology of life events and difficulties, with some reflections on the concept of independence. British Journal of Psychiatry 148, 686696.[Abstract]
Murphy, G. E. (1992) Suicide in Alcoholism. Oxford University Press, New York.
Murphy, G. E. and Robins, E. (1967) Social factors in suicide. Journal of the American Medical Association 199, 8186.
Murphy, G. E., Wetzel, R. D., Robins, E. et al. (1992) Multiple risk factors predict suicide in alcoholism. Archives of General Psychiatry 49, 459463.[Abstract]
Öhberg, A., Vuori, E., Ojanperä, I. et al. (1996) Alcohol and drugs in suicides. British Journal of Psychiatry 169, 7580.[Abstract]
Paykel, E. S., Myers, J. K., Dienelt, M. N. et al. (1969) Life events and depression: a controlled study. Archives of General Psychiatry 21, 753760.[ISI][Medline]
Pirkola, S., Isometsä, E., Heikkinen, M. et al. (1997) Employment status influences the weekly pattern of suicide among alcohol misusers. Alcoholism: Clinical and Experimental Research 21, 17041706.[ISI][Medline]
Pirkola, S., Isometsä, E., Heikkinen, M. et al. (1999) Female psychoactive substance dependent suicide victims differ from male. Results from a nationwide psychological autopsy study. Comprehensive Psychiatry 40, 101107.[ISI][Medline]
Pritchard, C. (1992) Is there a link between suicide in young men and unemployment? A comparison of the UK with other European Community countries. British Journal of Psychiatry 160, 750756.[Abstract]
Rahe, R. H. (1977) Epidemiological studies of life change and illness. In Psychosomatic Medicine. Current Trends and Clinical Applications, Lipowski, Z. J. et al. eds, pp. 421434. Oxford University Press, New York.
Simpura, J. (1987) Finland's annual alcohol consumption and its distribution. In Finnish Drinking Habits. Results from Interview Surveys Held in 1968, 1976 and 1984, Simpura, J. ed., pp. 5577. The Finnish Foundation for Alcohol Studies, Jyväskylä.
Welte, J., Abel, E. and Wieczorek, W. (1988) The role of alcohol in suicides in Erie County, NY 197284. Public Health Reports 103, 648652.[ISI][Medline]