Suicide risk factors are not generalisable
medwireNews: Canadian researchers have identified differences in suicide risk factors between people with and without bipolar disorder, highlighting the hazard of generalising suicide risk factors to all populations.
Compared with people without bipolar disorder (n=2716), those who died by suicide and had bipolar disorder (n=170) were 1.8 times more likely to be female, twice as likely to have previously attempted suicide and 1.6 times more likely to have had recent contact with psychiatric or emergency services. But bipolar patients who died by suicide were two-thirds less likely than controls to have had recent medical stressors.
The researchers, led by Ayal Schaffer (Sunnybrook Health Sciences Centre, Toronto, Ontario), also observed significant differences in suicide method between the bipolar and non-bipolar individuals, who were all residents of Toronto and died between 1998 and 2010.
Lethal self-poisoning was the cause of death in 33.5% of those with bipolar disorder compared with 17.4% of those without bipolar disorder, whereas hanging was more common in the non-bipolar group than in the bipolar group (29.7 vs 17.1% of deaths).
These results reinforce “the research and clinical imperative to better understand suicide within and across specific groups of people, including those with [bipolar disorder]”, write Schaffer and co-authors in Bipolar Disorders.
To further understand the risk factors specific to bipolar disorder, the researchers conducted a cluster analysis within the bipolar disorder suicide group.
They identified five clusters that showed differences primarily in pre-suicide variables such as substance abuse, past suicide attempts, gender, living circumstances, presence of an employment or financial stressor, presence of a police or legal stressor, interpersonal stressors and age.
For example, cluster 1 (the largest, n=56) comprised all men, most of whom were unmarried and living alone, none of whom had a past suicide attempt and who had the highest rates of suicide by fall or jump.
By contrast, cluster 5 (the smallest, n=23) comprised all women, and was the youngest group, with nearly all living with others. It had the highest rate of recent employment or financial stressors and the lowest rate of contact with mental health services.
Cluster 2 (n=32) was characterised by, among other things, a 100% rate of past suicide attempts. Cluster 3 (n=30) comprised mostly older women who lived alone, while cluster 4 (n=29) was characterised by a 100% rate of substance abuse, high rates of past suicide attempts and police/legal stressors, and the highest rate of self-poisoning as method of suicide.
Schaffer et al conclude that if these clusters can be replicated they “should be incorporated into the design and analysis of future studies of suicide in [bipolar disorder].”
medwireNews (www.medwirenews.com) is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2014
By Laura Cowen, medwireNews Reporter