TORONTO— Several key risk factors distinguish persons who have attempted suicide from those with suicidal ideation only, according to a study conducted by Stephen B. Woolley, MPH, DSc, John W. Goethe, MD, and colleagues. People who had attempted suicide were much more likely to have a history of trauma or abuse, substance abuse, and living alone, compared with individuals who had only thought about suicide. In contrast, family history of substance abuse was significantly less prevalent among those who had attempted suicide.

Between suicide attempts and ideation, no statistically significant associations were observed regarding age, gender, violence, anger, arrests, homelessness, feelings of isolation, personal substance abuse, and family history of attempted or completed suicide. Dr. Woolley presented his group’s findings at the 159th Annual Meeting of the American Psychiatric Association.

The researchers based their results on interviews with persons who had been recently admitted to the Institute of Living, Burlingame Center for Psychiatric Research and Education, in Hartford, Connecticut, in an adult day program. That information was corroborated by checking patients’ medical records to ensure that any suicide attempts were "serious," noted Dr. Woolley, a Senior Scientist at the Burlingame Center. The goal was to develop a predictive model for people at risk of attempting suicide that could be used in a clinical setting.

"We cast a wide net to find information to evaluate this risk," said Dr. Woolley. The rationale for attempting to predict suicide "is really based on the association that has been shown for many years in many studies that current suicidality is strongly associated with recurrent and future suicidality. Hopefully, there are factors [in our study] that actually are associated with future suicidality."

AT RISK FOR SUICIDE

A total of 117 patients were interviewed during a six-month period in 2002. Of this group, 106 (91%) who had previous suicidal experiences were analyzed—77% reported a suicide attempt, and the remaining 23% acknowledged a history of suicidal ideation. Eleven communicated no ideation or attempts and were excluded. The mean age was 38 (range, 18 to 62), and the majority of patients were female. A quarter of patients were married, and 20% had full-time employment. Participants offered "an enormous list" of lethal means for completing suicide, noted Dr. Woolley.

The researchers obtained information on patients’ history of medical treatment and illness; abuse and trauma; violence, anger, and arrest; headache and treatment; depression and other psychiatric diagnoses; and other major life events. They also looked at parent and family history of substance abuse, mental illness, and suicidal behaviors, as well as patients’ social support systems.

More than half of patients had three or more lifetime admissions to an inpatient service for a mental health reason. About 60% of patients had a history of trauma or abuse "that we rated as significant," said Dr. Woolley. More than half of subjects reported anger, violence, and arrest, and 71% of patients had a significant physical illness. Nine percent had a medical disability, and about 23% had a psychiatric disability.

Nearly two thirds of the cohort indicated that they had a disrupted family, 12% were homeless, 37% lived alone, and more than half indicated that they felt socially isolated from family and friends. "On the other hand, 84% could list someone that they could count on in acute situations such as hospitalizations or even for emergency care," noted Dr. Woolley.

A majority of patients had a history of either family or parental substance abuse, and "this was not trivial substance abuse," emphasized Dr. Woolley. In addition, 30% had parents who required treatment for mental illness, 14% had an extended family member who had completed suicide, and 7% had an immediate family member who completed suicide.

Patients were also evaluated using the Beck Hopelessness Scale. "We were a little bit surprised, given the self-reporting … in this population, that the average score actually was right on the cusp between mild and moderate—8.1," said Dr. Woolley. "And there was no significant difference between ideators alone and those who had ideation and attempted."

PRELIMINARY ANALYSIS

The investigators calculated odds ratios for attempted suicide versus no attempt for each predictive variable that they examined. Initial pairwise analysis revealed that trauma and abuse were twice as prevalent among those with suicide attempts versus those with ideation only, according to Dr. Woolley. "Interestingly—anger, violence, and arrest—if you look at these individually, they statistically were not very strong," he said. "So that surprised us. There was perhaps a suggestible mild association with headaches. Disability was associated with a doubling of the odds of attempts versus ideation, and that was driven primarily by psychiatric disability. And I want to point out that these confidence intervals show clearly that some of these estimates are unstable, driven by [the fact] that only 23% of people were in our ideation group."

Parental substance abuse was associated with a 71% increase in odds of attempts, and parental mental illness was associated with more than a 2.5-fold increase in odds of attempts. Living alone increased the risk of attempts by 175%. "Homelessness had no effect, no association," said Dr. Woolley. "This is counterintuitive to me." As for the effect of marital status/changes, "We looked in-depth at this … and nothing really came of that."

LOGISTIC REGRESSION FINDINGS

The researchers then used logistic regression analysis to develop a predictive model that "we feel is a good starting point for future research for us," said Dr. Woolley. After controlling for demographic factors, the investigators found that living alone (odds ratio, 3.5), trauma or abuse (odds ratio, 3.1), and substance abuse (odds ratio, 3.5) were associated with an increased risk of suicide attempts versus ideation.

"Living alone was the strongest factor in this model—over a three and a half–fold increased risk of attempts versus ideation," said Dr. Woolley. "Initially, one hypothesis that we had was that [among] people who said that they were living alone, many of them were actually homeless but didn’t want to acknowledge it."

A history of family substance abuse was actually protective of suicide attempts, which was a reversal of the pairwise analysis finding, Dr. Woolley pointed out. This is "an interesting factor," he said. "Perhaps resilience issues are involved in this." Family disruption also appeared to be preventive or protective, "but it is a very unstable" estimate, commented Dr. Woolley. On the other hand, parental mental illness increased the odds of attempts, "as we expected," as did psychiatric disability.

Among the factors that were not associated with an increased risk of suicide attempts, several others were unexpected, according to Dr. Woolley. One was the finding on marital status/changes. "We were surprised at that, since there were a great number of very disruptive marital status or partner status changes in this population," he said. "Interestingly, in the end, physical illness, although disability was a factor, … was not associated with attempts versus ideation. Finally, hopelessness was not a player, so to speak, in this analysis." Taking medication for headaches was also associated with decreased odds of suicide attempts./FONT>