In This Issue

Abstract

Ed Shneidman reports on his 1971 analysis of the suicides that had occurred in the Terman Gifted Children Study. Begun in the 1920s at Stanford University, Professor Lewis Terman identified 1,528 high-IQ California students and, over the years, they have been continuously followed. Today, the remaining subjects (about 100) have a mean age of 93. In his uniquely Shneiderian manner, Ed relates how, in 1971, he utilized a psychological autopsy approach to accurately predict the suicides of a small subsample of the 20 suicides that had occurred to date among the 857 male participants. Almost 35 years later, he shares with us his slightly revised view of those factors that he believes play significant predictable roles in the eventuality of suicide. Jeremy Kisch et al. report on analyses from the Spring 2000 National College Health Assessment Survey (NCHA), sponsored by the American College Health Association. This is the largest and most comprehensive survey to date (15,977 college students) that provides a replication of the CDC's 1995 National College Health Risk Behavior Survey (NCHRBS). This study investigates the relationship between suicidal behavior and depressed mood, as well as other risk factors which increase vulnerability to suicidal behavior. Of particular note is that less than 20% of college students reporting suicidal ideation or attempts were receiving treatment (psychotherapy and/or medications). Adolescent suicide continues to be a major focus of much research as evidenced by the number and range of studies published in SLTB over the years. Researchers have appropriately focused on identifying critical risk factors associated with suicidal behaviors. Thompson et al. explored the roles of anxiety, depression, and hopelessness as mediators between known risk factors and suicidal behaviors among 1,287 potential high school dropouts. As a step toward theory development, a model was tested that posited the relationships among these variables and their effects on suicidal behaviors. The results showed direct effects of depression and hopelessness on suicidal behaviors for males, and direct effects of hopelessness, but not depression, for females. For both males and females, anxiety was directly linked to depression and hopelessness; drug involvement had both direct and indirect effects on suicidal behavior. Lack of family support showed indirect influences on suicidal behaviors through anxiety for both males and females as hypothesized. As the authors point out, for mental health professionals, school personnel, and prevention scientists working with at-risk or suicidal youth, the findings make evident the need to address multiple co-occurring problem behaviors such as drug involvement, emotional distress, and suicidal behaviors. Over the years, SLTB has published data emanating from the CDC's Youth Risk Behavior Survey (YRBS), which is conducted biannually. (see SLTB 30: 304–312; 32: 321–323). The four questions that measure suicidal ideation and attempts are ordered along an implicit continuum of development and severity from seriously considering suicide, to planning, to actually attempting suicide, and, finally, to requiring medical attention for a suicide attempt. As presented by Victor Perez, this process assumes a straightforward sequence of suicidal thought and action and is the underpinning of the content and order of the YRBS questions. However, the assumed ordinal properties of the four questions had never been empirically tested. Brener, Krug, and Simon (SLTB 30: 304–312) found discrepancies in the trends of responses to the questions measuring suicidal activity in their analysis of the YRBS from 1991–1997. Specifically, they showed that although the percentage of students who reported having seriously considered or attempted suicide in the past 12 months had declined steadily since 1991, the percentage of students requiring medical attention for an injurious suicide attempt had increased. They concluded that injurious suicide attempts might not be linked to seriously considering or planning a suicide attempt. Perez examines the assumed ordinal relationship in the 1999 Youth Risk Behavior Survey by constructing a trajectory that identifies all possible response patterns among the four questions measuring suicidal activity. Significant differences between means of dependent variables at each level of the normative trajectory supported the hypothesis that frequency of risk behaviors increases monotonically with successive suicidal thought and behavior. Research on the relationship between schizophrenia and suicidal behaviors has languished until recently (see SLTB 30: 34–49; 34: 66–76; 34: 76–86; 34: 311–319). Camilla Haw et al. identify the risk factors for deliberate self-harm (DSH) in schizophrenia. They provide an explanation and reference to their use of the broader (European) term DSH, rather than attempted suicide. Although there are several published review articles on risk factors for suicide and suicidal behavior in schizophrenia, there are few reviews which make any mention of risk factors for DSH, and there is an assumption that the risk factors for DSH and suicide are the same. Haw et al. examined cohort and case-control studies of patients with schizophrenia or related diagnoses that reported DSH as an outcome. Five variables (past or recent suicidal ideation, previous DSH, past depressive episode, drug abuse or dependence, and higher mean number of psychiatric admissions) were associated with an increased risk of DSH, while one (unemployment) was associated with a reduced risk. In a continuing series of studies (see SLTB 27: 153–163), Eric Blaauw et al. focus on the prevention of suicide in jails and prisons. This current study aims to identify combinations of characteristics (demographic, psychiatric, and criminal) that are capable of identifying potential suicide victims. Characteristics of 95 suicide victims in the Dutch prison system were compared with those of a random sample of 247 inmates in ten jails. Combinations of indicators for suicide risk were also tested for their capability of identifying 209 suicides in U.S. jails and 279 prison suicides in England and Wales. A combination of two demographic characteristics (age over 40, homelessness), two criminal characteristics (one prior incarceration, violent offense), and two indicators of psychiatric problems (history of psychiatric care, history of hard drug abuse) proved capable of identifying 82 percent of the suicide victims in the Netherlands at a specificity of .82 in the general inmate population. Less powerful combinations correctly classified 53% of the U.S. suicides and 47% of the U.K. suicides. Blaauw and colleagues conclude that a small set of demographic and criminal characteristics and indicators of psychiatric problems is useful for the identification of suicide risk in jails and prisons. They point out that the characteristics can easily be incorporated in a screening device that can be administered during the intake process for new inmates. Some risk factors for suicidal behavior are similar to those often studied in the field of criminology. The link between suicide and crime is most apparent when there is a homicide followed by suicide. People who kill others rarely kill themselves afterward. When they do, they are more likely to have killed someone with whom they are intimately involved—an intimate partner or a child. A review of the literature on this phenomenon in Australia, Canada, and the United States showed that, in all three countries, the majority of those who commit suicide after a homicide are male partners or ex-partners of female victims. Using data on over 700 intimate femicides, Myrna Dawson examined the role of premeditation in cases of intimate femicide-suicide compared to killings that do not culminate in a suicide. Her results show that premeditation is more likely to occur in cases involving the offenders' suicide, but that evidence of premeditation varies depending on the type of suicidal killer. Michel Préville et al. present the results of a psychological autopsy investigation of 101 adults aged 60 years and older who died by suicide in Quebec in 1998–1999. The study looks at a number of variables from proxy survivor interview data, including health- and mental health-related behaviors, mental health diagnoses, and social and demographic variables. In this study, 42.6% of the suicide cases presented with mental disorders at the time of their death (mainly depression), and nearly 44% of the suicide cases had no current or pre-existing psychiatric condition. Only 27.7% of the cases did not express any idea of death during the 6-month period preceding their suicidal death. Interestingly, 53.5% of the suicide cases consulted a general practitioner or specialist during the 2-week period preceding their death. The authors suggest that family members and friends could play an important role in preventing elderly suicide attempts by encouraging their parents to discuss their suicidal thoughts with their general practitioner. The results show some similar and some disparate results compared to other studies of elderly suicides, indicating the need for research that examines the complex causal nature of the relationship between mental disorders and suicide among the elderly. Adding to his studies of homeless veterans who abuse substances (SLTB 33:430–432), Brent Benda studied 315 male and 310 female homeless military veterans in a V.A. inpatient program designed to treat substance abusers, many of whom also suffer psychiatric disorders. The study examined gender differences in factors associated with the odds of having suicidal thoughts, and of attempting suicide, in comparison to being nonsuicidal. Childhood and current sexual and physical abuses, depression, fearfulness, relationship problems, limited social support, and low self-esteem were more strongly associated with suicidal thoughts and attempts for women than for men veterans. Extent of alcohol and other drug abuse, aggression, resilience, self-efficacy, combat exposure, combat-related PTSD, and work problems were more strongly associated with suicidal thoughts and attempts for men than for women. While anticipating the release of the CDC's final 2002 national suicide death numbers and rates, there have been different explanations offered for the apparent slow, but steady decline in national suicide rates over the last years. In 1996 the rate was 11.52/100,000; 11.23 in 1997; 11.13 in 1998; 10.47 in 1999; 10.43 in 2000, and 10.69 in 2001. By the time you receive this issue, we will know whether the trend is heading downward or beginning to climb back up. Recently, my attention was drawn to an Editorial originally published on July 30, 1904, in the Journal of the American Medical Association, entitled, “The Increase in Suicide.” I have excerpted portions as follows: The increase of suicide has come to be such a marked feature of social statistics in this country that physicians must be made to realize the possibilities there may be of bringing about a decrease in this unfortunate matter by more care and prevision. Suicides are somewhat more than twice as frequent now as they were ten years ago. … Carefully collected statistics show that there was a constant increase from 3,531 suicides in 1891 to 6,600 in 1897, then a drop in 1898 to 5,920, and in 1899 to 5,340. In 1900 there were 6,755 suicides, an increase of over 150 above the figures for 1897, the highest previous number, and there has been a constant increase since, in 1903 the number of suicides being very close to 8,600. During the last thirteen years—that is, since 1891, there have been altogether 77,617 cases of suicide reported in the newspapers of this country. The decrease in the number of suicides during 1898 and 1899 is not surprising, if we remember the conditions that prevailed in the commercial world at that time. After a period of hard times there was the wave of prosperity and a decided reaction in men's feelings that made the future look bright enough for everyone. Curiously enough, the statistics, however, do not show that city life is so much harder on the people than country life. During the last ten years, the suicide rate in fifty cities of this country has about doubled. That is, however, only in proportion to the suicide rate throughout all the rest of the country and does not especially condemn the high pressure of large city life as a disturbing factor of mentality. The most serious thing about the statistics is the fact that, though suicides are more common among men than women in all countries, the difference is gradually growing less, and in recent years, this has been quite marked. (43: 333–334) The 1890 census data from some states was partially destroyed by a fire in 1921; however, the federal records indicate an 1890 U.S. census of 62,116,811. Using this number, the rates are as follows: 5.68/100,000 in 1891, 10.63 in 1897, 9.53 in 1898, and 8.60 in 1899. The official U.S. census in 1900 was 74,607,225. In 1900, there were 6,755 suicides, yielding a rate of 9.05/100,000. In 1903, there was a significant increase in the number of suicides to approximately 8,600. Using the 1900 census, the suicide rate rose to 11.53/100,000. During the 13-year period between 1891–1903 there were 77,617 suicides, or approximately 5,971/year. This yields a roughly estimated rate of 9.61/100,000 for this 13-year period. Taken as a whole for the years surrounding the turn of the 20th century, the rate hovered around 9–10/100,000. One hundred years later, we continue to grapple with similar rates and similar risk factors.