Late-life suicidal behavior: Complexity, descriptors, level of risk

Abstract

Suicide is disproportionately fatal for older persons compared to other cohorts. Presently, there is an epidemic among those, mostly men, who are 85 years and older, where the rate of suicide has reached an all-time high of 81.6/100,000 resident population ((for white men it is 85.9/100,000) NCHS 1996). The present research builds on previous work (Meisel (Weinreich) 1994) to identify older persons at risk, and incorporates Chaos and Complexity Theory in an effort to address the intractability of the suicidal act. This research sought to formulate a consensus among practitioners and observers who experienced the suicide of an older person. A Q-sort was developed and a Principle Components Analysis (PCA) performed. Confirmation of the existence of a consensus on the descriptors of the suicidal older person was determined. The PCA yielded a five component and a two-component solution. Both solutions met Kaiser's criteria of eigenvalues >1. However, the 5-component solution was not supported by a manual scree test. The 2-component solution was discussed. Component 1, mental health/emotional health, described older persons suffering with some form of emotional distress. Many subjects were clinically diagnosed with major depression or bi-polar disorder. Component 2, physical health, described older persons who, while mentally healthy, possessed mitigating physical ailments that seemed to make life less worth living. Three of the four constructs proposed in this research were supported, namely: subjective physical decline, subjective mental decline, and hopelessness. No support was found for the fourth construct: intolerance for the normal aging process. This research also yielded intriguing support for the possibility to differentiate between young (those under age 55) and old (those 55 and older) persons committing the suicidal act. Several Q-sorts were rejected for the present analysis because subjects were too young. None-the-less, these Q-sorts were loaded specifically to see if respondents describing younger persons would load distinctly differently. This in fact occurred. Although there were too few respondents to be definitive, the Q-sorts produced two additional components. It is believed that, had there been a larger representation of younger persons, strong, separate components may have emerged with distinctly different groupings of terms for younger persons committing suicide.