Chapter 7

High Risk Groups


There are groups of people who are more at risk than others. This chapter reviews those who are known to be at greater risk and the factors that contribute to this outcome.


Rich and Pitts (1979) found that while suicide rates of physicians in the U.S. were twice as high as the general population, when corrected for age, differences disappeared. Thus, white males over 25 had a rate of 34.6 at the same time that male physicians showed a rate of 35.7.

Female physician rates present a different result. While the general population of females over age 25 shows in 11.4 rate, female physicians attained a rate of 40.7 Explanations based on objective data are lacking, although speculations are usually offered based on the changing roles for women.

Blachly (1969) identifies psychiatry, anesthesiology, and ophthalmology as specialties at greater risk for suicide. Pediatrics are the least. Rich & Pitts (1980) report that psychiatrists kill themselves twice as often as other specialties and this finding is constant and statistically significant for the years reviewed. They note as evidence that white psychiatrist suicides reach 12% of the total number of physician commits when they represent only 7% of the total.

Murphey (1975) notes that victims in the general population have had contact with a physician hours to months before killing themselves and that over half who died from overdose did so with prescriptions issued the previous week. He concludes that the principal opportunity for prevention occurs within the medical office when high risk people come for one last effort to obtain help. The physicians at risk for suicide are like the general victim, which includes an indirect call for help to their colleagues, and especially their personal physicians. However, doctors are notoriously poor in compliance with their own medical care (Simon, 1986), and especially if this is aggravated by previous impairments or addiction.

Werner Simon (1986) reviews the topic of physician suicide and lists issues that represent greater risk and therefore warning signs for occasions of greater danger. These are :

  1. transition from residency to independent practice
  2. approach of retirement
  3. when parents become terminal
  4. when children leave home
  5. chronic marital discord
  6. intense competition from younger colleagues
  7. already impaired physicians with new losses

Simon along with Murphy underscore the physicians office as the principal opportunity to deter physician suicides.

Interventions can consist of one or more of the following options:

  1. sympathetic listening, non judgmental demeanor and supportive efforts;
  2. when alerted to the issues of death wishes, or self injury assessment section in this manual;
  3. If the wish to die is elicited, formulate a direct prevention plan with the client-physician's cooperation;
  4. Schedule return visits in a week or less to reassess status and compliance with initial prevention plan;
  5. If medications are used, give in non-lethal quantities; samples are optimal ways to prescribe since they are small and represent professional courtesy;
  6. Refer to appropriate mental health resource whenever tolerated by the client;
  7. Referral to the committee on disabled physicians as soon as possible, in the absence of cooperation;
  8. If substance abuse or dependencies are detected, the interventions must be more aggressive and immediate;
  9. Confidentiality is often raised as a reason for avoiding aggressive treatment or hospitalization. While desirable and necessary to seek whenever possible, it remains illusory since associates always know about the addiction or impairment and recognize this earlier than the victim. The necessity to initiate detoxification especially in the face of self injury risks should not be delayed to protect privacy.

Special problems exist for the psychiatrist whose patients engage in self injury behavior. The client-physician and indeed any, psychotherapist, suffer special distress. For the general medical physician whose client is a psychiatrist with a recent (1-12 month) patient suicide, there are some special treatment issues. These are :

  1. they need an opportunity to cartharse their feelings; simply listening and accepting is the single most effective service and is the kind of attention most difficult to obtain (Goldstein & Buongiorno, 1974).
  2. reframing their distress as a form of professional grief for the losses suffered when a client dies.
  3. affirming that suicide is currently unpredictable, and that the base rate for high risk clients imply a number of suicidal deaths per year per treating unit or therapist. It is only personal vanity that supposes a 0.0 suicide rate.
  4. help the client to form a collegial support group for survivors of suicide attempts or commits. There are national and regional associations of professionals for suicide prevention who may be able to fulfill this role. There are in addition support groups for survivors of suicide who are not identified with any profession (see resource section).
  5. Suicide prevention is everybody's responsibility, and not only the treating professional. No one person can prevent, nor cause another to engage in self injury behavior. Everyone can do more to deter others.
Top Menu Main Menu Bottom Menu


Accident and suicide in the 15-24 age class is the major causes of death because fatal diseases are relatively infrequently. Since 1950 the suicide rates for all American youths have been rising. The male rates have increased from 6.7 to 20.5 between 1950 and 1980. The group at greatest risk is white males between 20-24 with a rate of 27.8 which is nearly double that of the 15-19 class with a rate of 15.0. Black males between 20-24 show the highest rates 20.9 for their race, but are less than whites. black and white females of the same age classes have remained low and consistently below 6.0 over the last 30 years. The method of choice has shifted to firm arms for both sexes in the younger age classes.

Seiden (1984) reviewed the issues in a Public Affairs Report (1984) in which he stressed the role of home and family, alcohol and other drugs, social relations & stress, academic environment and employment among other factors. He notes the tendency to omit replacement of the father in divorce or loss. Paffenberger et al (1966) and Stanley & Barter, (1970) have documented the relation to adult suicide in boys suffering loss of fathers. Others have reported that currently only 38% of children are with their natural parents. Seiden notes the increased role of alcohol. S. Blumenthal in her testimony (subcomm, 1985, 1985 p.53) noted that alcohol is present in 80% of the victims who attempt suicide. Seiden also stresses the role of loaded guns in the American homes as a facilitator of impulsive self injury actions in youth.

Seiden concludes that "the serious problem of youth suicide has many contributing causes, including some that will be hard to change." He ends optimistically because "comparatively low cost but effective techniques of interventions can save significant number of lives as well as avert much grief and trauma" (1984).

Smith & Crawford (1986) reports a suicide attempt rate of 8.4% for Kansas High Schools and offers data that nearly 90% did not seek medical treatment afterwards. Their data seem to reconcile the differences reported in rates versus percentages of attempted suicide. He confirms that conservative estimates of attempt to commit ratios are above 100 to 1.

The range of attempt rates for the 15-19 age classs varies from a low of 3.9% in Los Angeles (Mintz 1970). In between is 13% from Northern California (Ross, 1986), and 10 in New York (Klagsbrun 1976). College student rates have received more scientific and media attention than the young person in the general population (Farnsworth, 1966). An additional subgrouping receiving more scrutiny is the veteran, especially those from the Viet Nam era 1964-1972 (Farberow & Williams, 1982). Bruyn & Seiden (1965) computed the student suicide rate at UC Berkeley and controlled.

For population at risk in the state of California in 1952-1961. They concluded that suicide rates of the college population was greater. Peck and Schrut (1971) who had access to all coroner cases in LA country found that the college rate for the 1960-66 period was 5.1 per 100,000 live students per year. This was well below the general population rate. The data from these two studies were from different times and places. These two carefully conducted investigations document the difficulty of identifying causative factors. Peck and Schrut suggest that larger universities usually located in urban centers tend to have higher rates because they attract more older men than the smaller colleges, and point towards life styles and cultural forces in addtion to personal vulnerability. They minimize the roles of drugs or substance abuse as explanatory factors. Peck & Schrutoffera typical profile of the male victims they studied; "sensitive, lonely, unhappy, ... who lacked close meaningful relations....."

Paffenberger, King, & Wing (1969) in a follow-up study of Harward and University of Pennsylvania undergraduates during 1926-1959 identified predictors of later suicide in physicians with data collected as students. The average age at death was 38. The profile they constructed based on significant differences from cohorts is limited to the times and places, but may be useful to researchers seeking to recognize high risk students. The fathersb were either dead or separated, college educated and professional. The students had come from boarding schools, gave histories of smoking, using alcohol, and dropping out of college. Current complaints at the time of data collection were insomnia, worries, self consciousness and cyclothymic character.

Farberow and Williams (1982) report that Viet Nam era veterans have the highest rates of all periods of service and that the under 35 age class from 1970-1980 have experienced the highest rates. Together they account for 34% of all veteran suicides in this same period. This pattern reverse the usual age distribution where the highest rates are associated with the oldest veterans. The veteran population with the highest rates at the young and older age classes, but dipping lower in the 35-40 ages. Farberow and Williams note that the rates for the under 35 veterans are trending down starting in 1979-80.

Psychological Factors in College Suicide

Rook (1959) concluded that the higher rates among college students were associated with rigorous academic selections for Oxford and Cambridge Universities in England. This earlier report is limited by time and location, although it supports the more conventional argument that academic achievement is a major stress causing the increasing suicide rates. However other factors have been identified which play an equal if not greater role in explaining rising suicide rates. Minnea and Rush (1981) conclude that death anxiety is associated with preference for the right to commit and more varied justifications for choice of suicide in people who have acknowledged self injury behavior in the past. Boor (1976) correlated internal - external locus of control scores with increasing national suicide rates in the US. The younger age classes, more comparable to college age students, showed more external scores while the older classes scored higher on internal locus of control.

Lester (1967) studied Buss-Durkee test scores which measure seven different types of hostility in college students acknowledging suicidal considerations, threats, attempts and no self injury ideation or behavior. Major findings were presence of irritability and resentment score elevations in the suicidal groups. The author concludes that there is no support for the hostility to the self hypothesis. However, the elevated scores suggest a displaced or repressed type of hostilities in other areas. Aside from occurrence rates 30% of college students acknowledged suicidal thoughts in the current year when queried by Craig and Senter (1972). Eleven Percent of counselees at UC Berkeley came for suicidal ideation (Bruyn & Seiden, 1965). Mishra (1982) reports that college students have greater contact with victims before death by suicide and have therefore the greatest opportunities to deter peer from self injury action. His sample reported 15% prior attempts in their own histories.

Richman (1979) lists 14 points of intervention in family systems approaches to older children and adults at high risk. These are :

  1. an intolerance of separation.
  2. a symbiosis without empathy.
  3. a fixation upon childhood patterns.
  4. a fixation upon earlier social roles.
  5. a closed family system.
  6. a particular pattern of dealing with aggression.
  7. scapegoating.
  8. sadomasochistic relationships
  9. double blind relationships.
  10. acting out of the bad self of the family.
  11. a quality of family fragility.
  12. a family depression.
  13. communication disturbances.
  14. an intolerance of crises.

Domino, Gibson, Poling and Westlake (1980) reported students attitudes toward suicide. Some of their findings: 72% feel that prevention efforts are required: 83% see suicide as a cry for help; 76% endorse the idea that everyone thinks of suicide; only 11% see suicidal behavior as normal; Many individuals endorsed commonly held misconceptions known as myths about suicide. e.g. people who talk about suicide don't commit or propel who kill themselves have made a firm decision to die. Despite all the research and documentation of increasing suicidal behavior, a survey of 90 California college deans or administrators by Hendrickson and Cameron (1976) yielded a consensus that there was not a need to increase resources or deal with this problem.

Schott & Clum (1982) studied a sample of 175 students enrolled in an introductory psychology course at Virginia Poly Tecyh Institute. They report support for the stress-problem solving model. Students with poor problem solving and higher life stress are at greater risk for depression, hopelessness and self injury behavior. Carmen & Blaine investigated 69 Harward and Radoliff students previous suicide attempts found these followed upon the failure of defensive and coping systems in the face of specific frustration. Important influences were loss by death or separation in the past, versus the absence of meaningful relations in the present.

A list of significant factors is reported by Pfeffer (1982) for family therapy with children ages 6-12, at risk of suicide: family inhibition of change; lack of generational boundaries, severity of spouse conflict, projection of inappropriate parental feelings onto children, and symbiotic parent child interaction. The American Psychiatric Association has issued a parent handout title "Facts about Teen suicide in which they urge they list five behaviors, any two of which is sufficient to require talking to a trusted adult or a mental health professional. These are :

* I am sleeping much later that I used to;
- I'm not sleeping well and wake up early in the morning;
- I'm beginning to take a lot of naps.
* My appetite has changed, and I've noticeably lost or gained weight.
* I feel restless
* I have withdrawn from friends and family
* I can't concentrate very well

The pamphlet notes that these should persist for two or more weeks before construing them as signs of early depression. The latter is said to suffer a rate of suicide equal to 15% of the victims.

While clearly intended to provide some helpful assistance, the clues suffer from the usual problems of poor prediction. The use of this check list will yield many false positives and cause latrogenic stigma or stress by identifying subjects who are not at risk for self injury behavior.

Checklist of possible interventions for any school setting

1. Parents and administrators need to acknowledge that a problem exists in order to open communication about a taboo topic. A similar failure existed at the start of the drug era which occurred at the same time that increased suicide rates of the young became apparent.

2.Education is helpful for adults and youth. It should focus on:

3. Identify high risk students by reviewing histories, self reports or appropriate testing in areas described below.

4. Students identified as at greater risk need access to more support.

The increased availability of Support groups for friends and family victims makes the postvention effort a little easier. See the support group section for list of names and addresses.

5. Continuing education for staff with existing material on a weekly basis until some level of proficiency is achieved and then monthly follow-ups. One professional goal is to develop faculty capability of providing death education courses, and lead ongoing education for the students. However, effective suicide prevention is every employee's responsibility; conversely all employees have some degree of grief reaction to any death. Special debriefing sessions should be scheduled for all staff following an incident as part of continuing education.

In this context the recent availability of a wide number of audio visual material is expressive of the public concern and the effort being expended to develop more and earlier educational efforts to reach into the schools prophylactically. See the reference to audio visuals for a complete list in this manual.

6. Psychological autopsies for staff following a suicidal incident is helpful (attempts or verbal threats should be included). Everything else being equal it is appropriate to include one or more students in these conferences especially if they play a prominent role as observers or a significant others. Selected autopsies can provide powerful educational material for the entire student body, especially if confidentiality issues can be resolved.

7. Peers, are in the strongest position in deter and influence other students before they reach a point of self injury. In order to mobilize these effectively, there needs to be a well defined faculty supervised, but peer led support program. This requires an ongoing weekly review of student leaders who are in resource positions such as resident counselors.

8. A suicide prevention committee is an optimal resource for developing faculty consensus, advising the administration and providing support to employees before during and after self injury incidents.

9. Help seeking attitudes need to be encouraged early in the year an during the first year of academic residence. Peer support and role modeling for these behaviors can be encouraged by orientation sessions focused on health care and stress management co led by senior students.

Top Menu Main Menu Bottom Menu


People over 65 kill themselves more often then any other age group. This and other information was made available by McIntosh (1986) in a remarkably succinct one page summary distributed at the joint meeting at AAS & IASP in San Francisco May 25-30, 1987. In 1983, the old were 11.7% of the population, but 18.7% of all suicides. On average, 14 old people per day killed themselves. It is estimated there are 200 attempts for every young commit, but only four attempts for every old suicidal death. The older victims are more likely to die because they use more lethal methods, have greater intentions to die, are less likely to be rescued, and have poorer recuperative powers when interrupted. Common features are depression which manifest differently in the older person (Zung, 1980), physical health problems Osgood (1982) and preoccupation somatic issues (Zemore & Eames 1979). Factors that seem to contribute are losses of significant others, lessened importance of their roles and the effects of retirement (Osgood, 1982). The latter include reduced income, less status, purposelessness, loss of personal control (Seligman, 1976) living alone, and increased self dislike.

Osgood (1982) summarizes the above by providing a list of warning signs. These are: change in sleep patterns with insomnia noted; loss of appetite and weight decline; extreme fatigue; increased concern with bodily functions; increased alcohol intake; mood shifts towards apathy, anger, irritability, sad, or withdrawn, manifest, anxiety; loss of emotional control e.g. screaming throwing; sudden interest in church and religion or affective expressions; scheduling appointment with physician for no apparent physical reason or shortly after a recent visit; loss of physical skills, general confusion, poor comprehension, deteriorating judgment and memory.

In a separate study, McIntosh and Santos (1981) report increased suicide rates for elderly non white minorities, especially Chinese, Japanese, and Filipino-American. McIntosh goes on to urge aggressive outreach because the elderly rarely seek available in most communities. When coupled with their greater risk and lethality, there is more justification for uninvited intervention. He goes on to recommend a networking approach in which all social resources are mobilized; e.g. neighbors, family friends, clergy, etc. To trigger this he exhorts all parties to be aware of clinical signs or early warning signals, and to ask directly about suicidal thinking or planning.

This strategy is reasonable, up to a point. The reservation that needs to be considered is the motivational differences between suicide in the elderly when compared to the younger or the middle aged. In people over 65, suicide gets confounded with appropriate dying, death with dignity, and the right to commit suicide, all of which impedes suicide prevention with the elderly. For further discussion of these issues see the section on cancer and death with dignity.

It may be helpful to distinguish high risk from acutely self injurious people. High risk people are typically those in between suicide attempts. Rates of those over 60 have traditionally been high. In the last 10 years, the under 25 age group has become higher, but the senior rate remains at the same high rate. Acute covers all ages and implies people on the verge of action. The older group engages in more lethal methods (Weiss, 1968). In the elderly the meaning and choice of method tends to be more specific.

High risk criteria for the elderly include people who live alone and are male. Frazer (1985) reporting on metropolitan trends in Cayahoga county (Ohio) among female suicide victims, notes higher rates for adolescent and senior citizens. While the rates for single and married were proportional to the general population, widowed were higher than divorced. Patients with an early diagnosis of life threatening illness are also at greater risk. Cancer patients exhibit greater rates of suicide than the general population at risk (Marshall, Burnett, & Brasure, 1982).

A primary suicide prevention tactic for the elderly is the process of coming to terms with death (Cutter, 1978) which yields affirmation of past and present life. This also enhances a sense of integrity (Ericson, 1968) which supports continuing, while permitting a readiness to die. These concepts are elusive in clinical practice but individuals can be allowed the privilege of acting on the assumption that death is near whether next week, next month or next year; and that constructive use of remaining time can be made for those who have come in terms with the inevitability of death (Cutter, 1978).

Given the universality of death, appropriate dying is strangely esoteric. The current zeitgeist regards birth as normal and death as always due to a pathological cause. The process of "growing down" and coming to terms with death is no longer recognized as the usual experience of the elderly. These growth processes have enabled the elderly of all generations to approach death with more equanimity have happens currently (Cutter, 1978). This change is associated with the medical alterations of death starting in the post world war II era. Death to-day comes slowly occurs out of the home, and in unfamiliar places. The social process of dying itself tends to be over determined by expectations and habits of the victim's previous life style. As a result, the past adaptations to death and dying have been disrupted by medical interventions and changes; none of which permit a proper role for the victim in order to accomplish the "work of dying".

Broomberg & Cassell (1983) call attention to the paternalism of care given and requests greater autonomy for the patient. Georgotas et al (1983) exhort more attention for affective disorders in the elderly. However, it is not clear how ones diagnosis can be differentiated from appropriate anticipatory grief in the elderly efforts to achieve an appropriate death.

People in the terminal stages of any disease have unfinished business to complete. This work consists of saying good-by, dealing with the choices of last things, helping the survivors to continue after significant loss, and expressing sentiments. An illustration of the latter is the example of Jacob on his death bed giving a blessing to his children. The biblical model assumes a knowledge that death is at hand and that the person is ready to die.

Such a process is comforting while liberating in anyone approaching the end stages of life threatening illness. The occurrence of a readiness to die is not the same as wishing for death, nor does it necessarily have to end in self injury actions. Such a readiness can facilitate patient comfort in continuing existence either with, or without, life support efforts. Readiness for death also permits the patient to maintain his or her dignity and carry it over to a point of actual death. All of which document that the nature of self injury in the elderly is different and its appropriate prevention by health care professionals must also be different.

Health care professionals can deal with the management of self injury behavior, by starting with the meaning of current options in the broader context of the patient's history, values and relations to survivors. Such a review is inherently life affirming in that the relation works both ways; the potential survivor and the victim can share in this review and the subsequent decisions. By doing so, they all facilitate optimism in the work that each must do to cope with inevitable loss. The patient always has the right to refuse treatment if deemed competent to make that choice. Any patient can also choose to act self injuriously, regardless of rationality or competence since the act is usually under way before discovery. The resolution of the right to suicide controversy suggested for the elderly is to accept the premise that it is equally erroneous to delay death as well as to facilitate it. The recent development of societies supporting the idea of assisted dealt are in response to the differences noted above, and limit their scope to those with fatal diseases (Battin, 1982). The California natural death act empowers health care personnel to respond to the prior documented preferences of elderly patients with deteriorating conditions. The process of completing this document facilitates a greater sense of dignity in preparing for death. It also deters overt, impulsive, self injury behavior.

Allowing nature to take its course is not equivalent to clinical suicide but merely another life affirming choice that can come out of the existing relations and life styles of the people involved. The search for death with dignity can be supported by health professionals. The distress motivated impulse to engage in solidarity, secretive self injury requires professional help.

Cancer And Death With Dignity

The current literature seems to confirm the earlier report by Farberow, Shneidman, & Leonard (1963). Commits were higher than expected in patients with a diagnosis of malignant neoplasms. Whitlock (1978) looked for tumors at autopsy in people age 50 and over. Those who died by suicide mode yielded a higher occurrence of tumors than matched controls dying in other violent modes. Louhivuori & Hakama (1979) report 1.3 and 1.9 times higher rates for suicide in male and female cancer patients respectively, when compared to the general population in Finland. They note that gastrointestinal cancer patients and those with non localized tumors were at higher risk. Graves and thomas (1981) describe results of a longitudinal study with medical students and the Rorschach test (Maryland). They report equivalent profiles for those who eventually succumb to cancer, suicide, and mental illness in contrast to those wh remain healthy or develop cardiovascular adjustment in 80 cancer patients versus 80 suicide later study (Bukberg, Penman & Holland, 1983) on the observed prevalence of depressive states in randomly selected cancer patients with good to excellent social support. Subjects were 32 male and 30 female aged 23-70, hospitalized for treatment. Forty four percent were coping well without manifest evidence of depression, while 24% showed severe depression. The balance of the patients with good to excellent social support. Subjects were 32 male and 30 female aged 23-70, hospitalized for treatment. Forty four percent were coping well without manifest evidence of depression, while 24% showed severe depression. The balance of the patients showed moderate (18%) and some (14%). Marshall and Burnett (1983) note that the probability of suicide in 5009 cancer patients compared to 17,064 controls cases ranged from 50 to 100 percent higher (New York).

A somewhat finding cases from Kinlen (1983) who followed 3440 veterinary surgeons from graduation in 1949-1953 until 1975 (United Kingdom). He reports a roughly two fold increase in mortality from suicide and a decreased mortality from respiratory illness. There was no excess number of leukemia or other cancers as might have been expected from professionals exposed to oncogenic viruses in their work. Rutqvist (1984) studied Swedish inter current mortality in 3857 breast cancer victims from 1961-63 and 1971-73. He notes a 21% increase over what would be expected with excess risk of death from circulatory disease, other tumors, accidents, suicide, injuries and infections.

These recent studies document an increased risk for suicide in patients with cancer and especially in the first 90 days following the achievement of a diagnosis (Olafsen, 1981). While some have argued for a "right to suicide" under these circumstances (Humphrey, 1987) and others for euthanasia (Euthanasia Foundation; Blachly, 1971; Wallace & Eser, 1981) those professionals committed to preventing self injury behavior can be encouraged to seek the middle ground in assisting any patient to achieve an appropriate death providing assistance with the real and perceived threats to well being in the context of accepting treatment for a serious illness.

The achievement of death and dignity

The point of departure for mental health professionals is the denial of death found in most people. This is related in large part to the inability of anyone to experience the moment of death and go on to report its nature. It is only possible to observe others dying or to imagine ones own termination. The belief systems about dying, and continuation are necessarily personal and derived in an intellectual process. Like all beliefs it is also amenable in an intellectual process. Like all beliefs it is also amenable to change of efforts.

Young-Brockopp (1978) evaluated 61 adult terminal cancer patients in the middle phase of their illness (Buffalo NY). She reports the need for hope, honestly, and information as important, but found individual differences in personal need to discuss feelings and aspects of death. All health disciplines have a role to play in helping patients with serious illness achieve death with dignity, when cure is not possible. In between they can facilitate quality of life in accepting the impairments of a continuing illness.

This discussion reviews the issues and options from the perspective of those wish to assist people with life threatening illness. Americans tend to deny death in the form of alternate belief systems that support the idea of continuing indefinitely which is exaggerated by the more recent avoidance of natural death topics and an increased preoccupation with violent modes of death (Cutter 1978). This is most easily observed in the public information media which over report homicide, suicide and accident while under reporting natural deaths. Such special attention to violent deaths has been called the pornography of death (Gorer 1965) and in direct analogy with the sexual repressions can induce a perverse interest in deviant sex. Such extreme feelings about death in the US tend to get focused on cancer, even though heart disease fatalities occur about twice as often.

Individual life styles shape the quality of death existence and in turn influence the reactions to serious illness. These give rise to different adjustments. Given equal severity of disease, some patients make the effects of illness worse, while others make it better through the quality of compliance and presence of positive thinking. The observable denial of serious illness and especially of its life threatening aspects is a partial explanation of how some patients aggravate their health. Such behavior has been labeled as suicidal because of the later abuse of foods known to impair health. The patients so observed are desirous of life, wish to stay healthy and are compliant most of the time. However, periodically they express strong feelings and do the forbidden such as eating food high in potassium, sodium, excess fluids, in kidney failure or stopping insulin and increasing sugar intake with diabetic conditions.

Oncologists are the health professionals who most often encounter patients with a life threatening illness who fail to cooperate or frankly resist a treatment plan that is considered optimal. The unexpected response occurs universally, although in small numbers. Some of these reluctant clients, go on to seek unproven methods such as laetril, coffee enemas, vitamins only, and still other procedures. The physician encountering these clients is perplexed, and often distressed by such lack of medical cooperation.

Understanding of this behavior can be increased if the concept of denial is viewed in its original usual. Denial is meant to describe consciously chosen avoidance of the unacceptable. Unconscious avoidance is called repression. The difference is degree of control that the subject retains in choosing to be aware or not aware of a topic. Patients who seem to deny the life threatening nature of their cancer sometimes deliberately choose to avoid this unpleasant topic. Other times they repress awareness of the idea. Often the two overlap with a subsequent loss of affect which gets expressed unpredictably. At less conscious levels there observable and measurable awareness of he difference between denial and repression is detectable in the residual ability to tolerate a frank discussion to bring up the topic of death directly when it is relevant, the client can be inferred to be repressing awareness. This assumes the professional is also able to deal with death directly.

The denial of threat becomes problematic when it increases behavior that impedes good health care. These are

  1. choosing no treatment,
  2. choosing unproven methods,
  3. manifesting a panic state,
  4. over reactions in terms of health habits,
  5. excess pain behavior,
  6. frankly psychotic behavior,
  7. suicidal threats or self injury behavior.

This range of reactions can and does occur, but fortunately in smaller numbers. Unfortunately, these responses to diagnosis and treatment of tumors are frequent enough to present continuous challenge to all health professionals. The suggested care of each of these reactions when attributable to breakdown of denial systems, will be discussed later. All of them require increased professional attention, best initiated by the attending physician, but requiring the skills provided by psychotherapists.

People in conflict need assistance in accepting and managing threateing information. When denial becomes repression, a logical or direct presentation will not get through, or if it does, can induce a panic reaction. Patients on the border betweendenial and repression are more difficult since they often encourage the physician to be direct, but later they may exhibit panic and need emergency support.

The patient with poorly managed denial is in conflict with him or her self and is thus more vulnerable to excess stress. Some cancer patients focus their feelings on specific objects or processes of care which then develop into phobic reactions to their treatment; e.g. the chemotherapy room, IV drop mechanism, the radiation therapy machine, the waiting room, etc. Others react to somatic aspects such as sleep, pain, nausea, vomiting, itching. and fatigue. Such distress is attributable to the medical conditions but is often a patient over or under reactions to physical sensations when tolerance levels decrease with experienced stress.

The general suggestion is to recognize that the patient with the most distress is not necessarily the one with the most disease or objective impairments. Where the denial system does not create management or unexpected behavior there is no need to change anything. However, where personal problems, life styles, or compliance issues begin to manifest, there is a need to address the patient's awareness of the life threatening aspects. The more the patient is able to review these in a neutral supportive atmosphere, the more control can return for personal perceptions that are troubling. Neither the diagnosis nor treatment needs to change, but rather the reactions of the client. This takes time, and special training of those in the psycho social areas, but the physician can appropriately start the process and make a referral. This allows a standard or optimal medical regime to continue.

Patients repressing awareness when threats to life occur in medical care rather than simple denial, require special assistance. The difference may be difficult to detect, and time consuming, but the tip of the iceberg can be notice when patients behave unusually in response to standard medical procedures. These range from avoidance of treatment, non compliance, or extreme reactions to changes in health habits.


In general medical practice, failure to follow advice occurs in all practice and is probably due to a variety of reasons. Since most patients, most of the time do follow professional advice, the issue of non compliance is usually noted, but not confronted. Psychological the failure to comply with recommendations represent an inability to change habits and resistance to new demands. In some medical conditions and with some non compliance the resistance itself aggravates health. This has been observed most in the management of chronic conditions much as TB in the past and diabetes today. Psychological attention in the form of support, alternative ways of delivering the prescriptions, and reframing of patient definitions are helpful. Those who are not responsive demonstrate a persistence in a choice contrary to medical advice. When time permits it is usually more constructive to accept the actual right to refuse therapy than to break off treatment relation. This general practice gets tested in the case of serious illness where choice issues are exaggerated by the omnipresence of a life threatening diagnosis.

The failure to accept treatment is most poignant in the management of cancer, when a patient explicitly refuses an optimal treatment plan. With truly informed consent this is simply a patient's right. If patients' find the physician demands of therapy to excessive, they will want to choose no therapy. Too often though, patients are influenced by stress reactions rather than the objective sensations of illness and treatment.

The optimal approach is to help patients deal with their understanding of issues by first resolving strong feelings, and later sorting out personal priorities. The psychological goal here is to help patient keep control of choice by steps that minimize the disruptive of diagnosis, treatment, and adaptation to serious illness. For some patients the greater preference may be to opt for nooo treatment at all. This can be positive for the overall comfort of the patient if the following conditions are present.

  1. the patient exhibits a great deal of effort in acquainting him or her self with the relevant information and developing informed consent.
  2. The prior life style, and decision making is consistent with current reactions to medical care.
  3. The need for control and choice are high priority items for the given individual in contrast to the more traditional patient who accepts uncritically, the recommendations of a trusted physician.

The positive and reassurance issue here, is keeping the patient comfortably as the primary obligation even in the context of less than optimal treatment.


Some patients who reject treatment options, especially in the absence of informed consent or knowledge, go on to seek unproven methods. These people, or their spouses, tend to be action oriented and impose great effort or change on their environments in order to achieve significant goals before becoming ill. In addition they seem to combine a brittle denial system with an active problem solving approach to life. They need to be vigorous participant controllers of their care. They may or may not be knowledgeable about treatment procedures. The major consideration is a straw of hope that they can invest with positive attitudes. The physician faced with this client has limited options. The optimal one is to set a time frame with in which standard medical care can be tried, before opting for the unproven approach. In doing so the client's freedom to choose and wisdom should be supported fully.


Some patients exhibit a panic state at the first ominous sign. Others manage well, until later, when diagnostic or treatment procedures are attempted. While all clients will suffer increased stress the majority find ways to manage, and do not manifest acute distress. This paragraph address those people who show an acute, anxiety attack which represent a dramatic change of usual demeanor. This usually manifests suddenly either all at once; reaction. Either way the client is overwhelmed by the idea of death, and dying in the immediate future. It is irrelevant that the objective facts do not support this real fear.

Factual reassurance does not go far enough unless it is coordinated with more cathartic opportunity to express the ultimate fears of death in a sympathetic and classically non judgmental manner. This is an especially effective attitude for the patients who are also self critical and need approval. With this professional support the patient is ready to hear positive facts. Often reframing the patient's perceptions of "I'm dying" into "Living one day at a time" is extremely reassuring. Actually, such panics are widows of opportunity to provide effective assistance to clients who are normally to proud to admit ned of what they define as weakness. These same attitudes persist and often impede behavioral health follow up because of the reluctance and often impede behavioral health follow because of the reluctance to avoid further reminders of the panic states and "weakness".


The more common paradoxical behavior for the attending physician is the client who reports difficulties with appetite, sleep, weakness, itching, pain, bowel, and other body processes. This becomes especially difficult to assess when these same effect are reported as common reactions to treatment. However, sometimes this becomes obvious as when patients describe sensations following the second or third treatment by radiation, which usually needs one or two weeks to manifest.

Such patients require more indirect support. The denial of life threatening implications cannot be confronted. If patients are asked by support staff how serious illness seems to be, they describe symptoms mentioned above, but say nothing about the threat to life. For these clients it is more productive to offer reassurance in the form of instructions to cope with disrupted health habits. For example, sleep disturbances may be an over reaction to changed patterns of aging. Providing advice and follow up to manage sleep also provides support and reassurance which can be delivered as a message that life is continuing. Other unexpected reactions can be traced to repression of life threatening aspects of serious illness.


Most patients with life threatening illness are normal people having no history of psychiatric disorders. While some may be undiagnosed, the majority are simply confronting an illness with ominous implications exaggerated by the pornography of death. These can be treated but at the cost of some disruptions to personal life style. In this context patients often encounter sensations that they label as pain. The atypical patient may exhibit continuing and extreme pain. It is important to note that the degree of pain is not the same for different patients with the same illness.

Pain thresholds vary from person to person and from time to time in the same person. These are further influenced by loneliness, anxiety, depression and other medications. On the opposite side, analgesics have a diminishing effect with the development tolerance, or a clouding of consciousness, and a consequent reduction of quality of life, e.g. apathy, poor judgment, discouragement, withdrawal.

The physician can enhance the management of pain by recognizing that it is perceived and is therefore real, and also psychological in the sense that awareness is needed to experience it. A combined approach is suggested that emphasizes any behavioral activities within the capabilities of the patient, with doses of medication that reduce pain without clouding awareness. Where this is not possible, it should be frankly acknowledged that other factors are present besides the illness which require acceptance of the patient's preference.


People with histories of mental illness can also suffer from cancer. Before the age of tranquilizers, it was observed that schizophreniacs often had temporary remissions of psychoses with physical illness such as appendectomies, tonsillectomies etc. Such recoveries were short lived and relapses were universal when patients were discharged from acute physical care or entered a convalescent stage. A similar phenomenon seems to be present when schizophreniacs receive a diagnosis of cancer. Their bizarre or irrational behavior diminishes and they become indistinguishable from medical patients. The explanatory rationale is that the anxiety gets focused onto body symptoms or sensations and thus bound, allows the psychotic person to function more like an ordinary patient. A secondary observation is that the kindness and attention provided by health care personnel for the psycho social needs of all cancer patients is exactly what mentally ill people need, and it is this aspect that allows for remission of psychiatric symptoms.

Such observations are suggestive for the medical management of encology patients. The patients with histories of mental illness or psychotherapy are usually no greater treatment problem than general client. It also suggests that mental health consultation for the oncology patient is not limited to those with psychoses. Behavioral methods are needed to manage the general medical patient when they exhibit specific symptoms of stress such as fear of death and somatic complaints, failures to comply, nightmares, and depressions. Conversely, such psychological impediments to good care, usually remit or diminish with two to four sessions of psycho social support.


Fortunately suicide is rare. Self injury behavior is more frequent which is usually on the order of ten to 400 times the number of observed commits. However, the rare occurrence of suicide demands attention for two reasons. Potential victims usually visit their physicians in the last six months of life in what seems to be an indirect search for some kind of help with psycho social concerns.

The second reason is that some victims of neo plastic disease prefer suicide as a solution to the perceived inevitability of pain, suffering, and certain death. This subjective perception of cancer is usually based on memory of loved ones or experiences with selected others who have suffered unusually rather than the whole range of human response to cancer. Patients who manage illness well are not usually noticed by the lay person or current patients.

Patients who react to diagnostic or treatment procedures with continuing and verbalizations about self injury in their questions, or fantasies are appropriate candidates for a direct question about their "wish to die". This relatively benign phrase opens the door to exploring ways in which the patient may have thought about planning for a self injury act. Such a discussion is usually productive because it tends to delay action by the patient. Implicit here s the notion of mixed feelings. Many suicidal people have simultaneous wish to live and to die rather than an all or none attitude. Any delaying tactic is life saving because it permits people time to opt for continuing life. This practice is especially urgent because such people are not psychiatric casualties, nor otherwise available for prevention and support.

These suggestions are intended to give health professionals supportive points of view when compliance issues arise in normal patients being treated for serious illness, and especially cancer. The overall principle of keeping the patient comfortable while nature takes its course gets impeded by mixed feelings about pain medication. The clouding of consciousness has already been mentioned. There is in addition the inadvertent and negative message that death is close, and going to sleep is all that a patient has left to do. Such a message aggravates pain, discourages hope and is contrary to the usual optimism expressed by health care personnel. There is a remarkable variety of time intervals between the point when health staff give up on treatment and expiration occurs. Encouraging patients to live one day at a time even in these extreme conditions is positive, supportive and too often contradicted by the attitude expressed in the administration of pain medications to a point of continuous drowsiness.

The alternative is to consider the possibility of supporting an active role in the achievement of death with dignity through constructive tasks such as taking care of last things, saying good bye to family, expressing feelings of love and regret to be leaving, and perhaps even giving a positive statement to survivors in the form of the old biblical blessing.

The preceding may not solve the controversy, but it does provide the health professional and family members a more neutral and constructive role in the face of life threatening illness in victims whom they wish to assist.


The relation of alcohol and substance abuse to suicide behavior has been reported in earlier reviews going back to Charles Moore (1790). Recent evidence comes from two kinds of surveys. The first is of alcoholics and substance abuse patients who go on to commit suicide. The second is retrospective studies of commits for history of alcohol and substance abuse. The author asked the SIEC to survey all articles with the key words substance abuse and suicide. On 10/21/87, 148 citations were found. The author classified these as either surveys of substance abuse populations (19) or suicide and suicide attempt populations (9). The contents of each generally supported the notion that their was increased risk of self injury behavior with substance abuse populations who come to the attention of treatment agencies. Conversely, there was a general finding that self injury victims who come to the attention of health professionals are more likely to be involved in substance abuse prior to the self injury event.

It can be conceded that alcohol and substance abuse figure prominently in the risk for violent death in general and suicide in particular. According to Litman (1986) it is second only to age as a predictive factor. Moore et al (1978) used the index of potential suicide. They were able to classify correctly, 87% of the methadone patients and 98% of normal controls. Chynoweth (1980) describes 135 consecutive commits in the Brisbane, Australia area and notes that 34% were drug dependent. Ryser (1983) reports a significant increase in drug overdoses between a 1974 and 1980 samples in students treated at emergency rooms. Fifty and 1980 samples in students treated at emergency rooms. Fifty percent of attempters used drugs to commit suicide or as a gesture. Clark &Compagnari (1985) studied the causative role of death in 150 active duty service personnel in the San Diego area between January 1983 and June 1984. Forty percent of all deaths reflected January 1983 and June 1984. Forty percent of all deaths reflected ethahol involvement. Suicide accounted for 18% and 30% of these were legally intoxicated at the time of death.

The most recent and definitive review of alcoholism and suicide is reported by Ray and Linnoila (1986). The array 21 studies from 1935 to 1094 and tabulate the number of deaths n alcoholics followed by the number and percent of suicides among these deaths. Where available they also give the number of years of follow-up. They note that the range varies from 2-56 percent. Since this confounds the variable follow-up periods, the present writer used the longest years given the reduced the total percent to an annual basis. This yielded a smaller range; 0.23% to 11.2% commits per year in follow up of alcoholic patients. The small number of deaths in five samples (ranging from N=6) to N=42) suggested their remaining 9 studies with follow up periods yielded a range from 0.23% to 3.5% with a median of 1.0% per year for suicidal deaths among all alcoholic fatalities. The smallest rate 0.23% was found in a 22 year follow up period, and is probably not typical. One percent per year may be regarded as high, but it is less than the 1-2% per year reported for victims with previous attempts. The lesser occurrence of suicidal death rate for alcoholics is consistent with lesser commit rate for alcoholics with previous attempts when compared to non alcoholics are examples of indirect suicide (Farberow, 1980) and seem to have a lesser rate of commit suicide than the other high risk groups (Motto, 1980), but probably have a lesser life expectancy if incident, homicide and natural causes of death are added.

The Ray and Linnoila (1986) summary reflects international and probably public supported health centers which may be less applicable to clients in the US who are treated in privately funded health care resources. These data show more men than women on the order of 87.4% according to the authors. Again this reflects the sociology of who is treated in public facilities. Women tend to drink at whom with the direct and indirect support of family or significant others until physical termination occurs thus minimizing the percent of suicidal deaths.

The authors go on to review biological factors in suicide and alcoholism. For more information on this point the reader should refer to the section on biology in this manual. They start with cerebro-spinal fluids (csf). Lowered serotonin in the brain is the principle finding. They concluded with the self medicating hypothesis in which the victim's addition to alcohol is an unconscious effort to correct the imbalance. Abstinence or withdrawal aggravates serotonin levels even further.

Depressed patients with family histories of depression had lower csf levels of both 5-HIAA and the Notepinephrine metabolites 3-MHPG (3Methy-4hydroxyphenyglycol).

Thyroid stimulating hormone (TSH) response to Throtropic releasing hormone (TRH) stimuli in 52 depressed females with history of violent suicide attempts showed a lowered TSH hen compared to depressed females without violent suicide attempts or depressed only. At five year follow up their were four commits who had shown no TSH response earlier. Ray & Linnoila note a negative correlation between CSF 5-HIAA and TSH response to TRH. While interesting and potentially significant, the small samples require replication for serious credibility.

MAO is aggravated by alcohol itself and is usually observed in severe withdrawal manifested usually by older males with the ravages of chronic alcoholism.

The small samples given make all conclusions extremely tentative. Their review of the current status of alcoholism and suicide converges on the research in "biological markers for suicide". The Ray and Linnoila review (1986) concludes "there is a great need for further research int the determinants of suicide among alcoholics".

The relative effectiveness of the biochemical determinants and subsequent medications in the management of schizophrenia over the last 30 years is a model of success for those who study depression, suicide and alcoholism. Unfortunately the diagnostic criteria for self injury behavior is less reliable than with the other forms of mental disorder. As detailed elsewhere in this manual there is a need to use more agreed upon criteria for risk estimation, lethality of methods, deliberate self harm, loss of hope, wish to die (reason for living, intentionality, satisfaction, role of significant losses, etc) rather than the more casual "depression" or suicide attempt language.

Biological markers for suicide will be no more successful than psychological predictors unless more effort is applied to precise outcome criteria for self injury behavior as reviewed in the prediction section.


The study and treatment of alcoholism suffers from the same problems as suicide. They both occur universally from antiquity to the present and despite much effort there has been little gain in prevention of cure. The definition and causes are obscure and yet consequences are far reaching. It is quite obvious that some alcoholics commit suicide and some suicides are alcoholic. Both populations have generated massive scientific literature which has not as yet led to significant improvement in prevention or care. One sign of this is the minimal attention found in textbooks of abnormal psychology or psychiatry. Considering the universal occurrence of suicidal behavior and alcohol addiction in clients seen by mental health services, only one or two percent of the textbook pages are devoted to these topics.

In the clinical settings of the past, both syndromes provoked professional anxiety, avoidance behavior, and outright hostility when either kind of patient appeared for care. Except for specialized alcohol treatment or suicide prevention centers the same reactions persist. Since the Suicide Prevention Triangle (1987) manual largely focused on suicide, short review of alcoholism may be helpful.

The early models of alcoholism assumed anxiety as the motivating source and following psychoanalytical thinking, effort was concentrated on developing insight, or resolving conflict as prerequisites to control of alcohol abuse. In this view, dependence and addicting were consequence of inadequate management of anxiety by the victim. Control would come with psychological cure. An opposite perspective is implied in Jellinek's (1946) progressive disease model. He starts with descriptions of behavior problems related to continued alcohol abuse through the life cycle of the victim. His array is visible in the drinking history scale of 52 items starting with universal behaviors such as the first drink and culminating in more extreme or rare events such as continuous benders or blackouts. Allergy or biological vulnerability is the assumed causality in the AA model.

Both formulations were derived from different populations of alcohol addiction. The anxiety model was largely derived from, and applied to, candidates for psychotherapy with presenting neurotic syndromes and ability to cooperate in the process. The Jellinek model was mainly based upon and applied to the public inebriate. The relation between the two when addressed was thought to be one of time, or degree; early for the anxiety and late for the Jellinek model. The more appropriate dimension here would be mild, moderate or severe alcoholism.

An addictive model is also implied in Jellinek's thinking, and comes in terms of habituation, dependence and tolerance. Anxiety is often relieved by alcohol in take. Social approval permits habituation. Time allows dependence. In the career of the alcoholic, a point of tolerance occurs. It takes more and more alcohol intake to achieve less and less degrees of release from anxiety. Eventually, the victim and alcohol dependence can be called addicted because of entrapment in a down hill health and social process. If left unchecked the addict is only amenable when one of many possible consequences becomes apparent. These are impaired health, alienated family, deteriorating work performance, reduced income, conflict with the law, and thinking disorders which may stimulate mental illness not usually relate to alcohol dependency. In end stages, the victim is unable to care for details of living; food, shelter, clothing and cleanliness are neglected.

This progression seemed to require a 5-15 year period in previous generations (Bacon 1973). Today the alcoholic career progresses over a one to five year cycle. Drinking patterns are more extreme and goals are to achieve total intoxication or "black out". There seems to be some connection with the expectations associated with the drug era, psychedelic values, and hedonistic goals of the period when today's youth was raised.

It is now generally recognized that psychotherapy cannot begin until substance use stops. It is less well known that often weeks or months of treatment are necessary and deficiencies in thought, both of which make minimal problem solving difficult enough to provoke renewed drinking.

The health professionals faced with patients addicted to alcohol and substances with high risk for self injury behavior tends to err on the side of suicide prevention by giving higher priorities to issues that may facilitate suicide. This writer would advocate a higher priority to treatment for addiction and the optimal management of the clinical sequels of the recovery cycle which may extend through 6 to 12 months after withdrawal. Suicide risk while continuously present is not usually acute during the observation and support needed for effective treatment of addition. It is in the laster months of the recovery cycle that the stressors of abstinence may provoke self injury events.

Top Menu Main Menu Bottom Menu


Almost all the articles addressing self injury behavior in detention, report a higher than expected level of suicide phenomenon. Whether it be rates, percents, frequencies of attempts, commits or need for prevention, there seems to be consensus. Table 8 gives the reader a sense of the range and degree of suicidal phenomena for victims in detention.

Table 8

Sources, descriptions and parameters of self injury behaviour in detention centers.

Source Parameter Date Frequency
Topp - 1979 3x Gen Pop
Hayes - 1981 16x Gen Pop
Burtsch - 1981 -
Flaherty - 1983 -
Rakis Urban detention staff 1984 Need for resources
Stone - 1984 "Very S"
Copeland - 1984 2nd most common cause for medical attention
CA Legislators Revolution to force more prevention 1984 6 incid.
Kennedy Survey of facilities long vs short stay detention centers 1984 Rates of 16 vs 53 100,000
Woolf 10-17 year old delinguents 1985 7% of all incidents

The one glaring exception was a report for the NY City woman's house of detention (Lombardi, 1979) that reported no suicides. The study was a thesis comparing four male correctional facilities in the city of NY to determine the significance of social climate influenced by correctional workers. However, other studies document female rates are also elevated. Depression is twice as high as in the general population according to Martin & Clonninger (1985). They note that 28% of their sample committed suicide. Climent, Plutchik, Ervin & Rollins (1977) identify significant loss of father before the age of 10 is more highly correlated with depression than the loss of a mother. They studied 95 female prison inmates in Mass. Women who had made a suicide attempt had a higher overall depression based on self report items before institutionalization. Significant items in descending order of significance are : insomnia, sadness, overall depression, unworthiness, emptiness, hopelessness, suicidal thoughts. Suicide attempts were more likely in women who exhibited more violent behavior. Violence was judged by self report, observations of correctional officers, the nature of the crime, length of sentence, and MMPI profile (Climent & Ervin, 1972). Women judged as violent had lost both parents at an earlier age.

There seems to be further findings (Kennedy, 1984) that the inmates at greatest risk are those housed in short term centers with rates of 57.5 versus 16-17 per 100,00 live inmates. The greatest period of risk appears to be within the first ten days of incarceration. Topp (1979) reports highest risk with sentences of 18 months. He also notes the period of greatest risk is within the first few weeks of custody.

High frequency of sexual offenses and lower number of affective disorders have been observed in victims of self injury who are incarcerated in Alaska (Sperber & Parlour, 1984). The co-existence of these behaviors must b viewed as co-related to some larger disorder that induces the observable symptoms.

Deheer & Schweitzer (1985) used the victims profile reported by the National Center for Institutions and Alternatives (Hayes, 1981) to compare two samples of suicidal behavior. The first was a suicide watch subgroup of 48 non attempters most of whom had verbalized a suicidal intention. A second subgroup of 20 attempters who had engaged in self injury actions while in custody and who had verbalized a suicide intention accompanied with a plan of action. The authors conclude that two fold comparison with the NICA profile held up with several small exceptions. The consensus of findings are:

  1. suicidal behaviors occurs within the first 24 hours,
  2. the high risk inmate is a white, male in his twenties who would use hanging,
  3. their would be a higher likelihood of alcohol and drug abuse intoxication would be present and their would be a history of both,
  4. a history of prior suicide attempts would be present, but no previous mental illness
  5. the nature of the charge would not be a good indicator of suicide risk.

Hankoff (1980) identifies the elements of suicide prevention in institutions including prisons;

  1. systematic assessment
  2. use of an institution wide committee to study every self injury event;
  3. training of staff;
  4. extended treatment for recently diagnosed affective disorders.

Complicating Factors

Factors which may complicate prevention efforts within custody are reviewed text. Malingering is as common as the self injury problem with significant effort expended to differentiate "serious" from "mild" degrees of self injury; a distinction not fully tenable as noted by Deheer & Schweitzer (1985) and confirmed by (Haycock, 1985). Risk factors must be evaluated for the distinction to be practical (Suicide on the inside, 1976). Much of the controversy can be eliminated by evaluating degree of lethality in the preferred method for any given victim, as discussed in the section under measures of lethality.

Alcohol and drug abuse is a prominent concomitant of self injury behavior (Nat'l Study of Jail Suicide, 1981). A recapitulation of these findings is given in Jail Suicide Update (1986). 76.6% of arrests are for non violent crimes. 30% of these involve alcohol or substance abuse charges. Almost 60% were intoxicated at the time of arrest. Fifty percent of all suicides died in the first 24 hours of incarceration. 27% died in the first three hours. 88% of those under the influence at the time of arrest died by suicide in the first 48 hours. Over 50% died in the first three hours. Two out of three inmates who died from suicide were being held in isolation.

These findings emphasize at least two significant issues for suicide prevention. The first is the delayed effects of substance abuse on the newly confined. Since inmates are presumptively not ingesting contraband, the role of withdrawal symptoms exerts a larger force in subsequent depression and or self injury impulses in the first 48 hours of incarceration. While delirium tremens or dts may not be visible, increased suggestibility and depression are common clinical observations, usually countered by more sympathetic nursing and/or withdrawal medications. The second is the aggravating role of isolation in facilitating suicides. The traditional use of isolation is to deter suicide with removal of methods and provision of supervision. Detention centers do not usually provide the necessary supervision nor the sympathetic nursing, coupled with the greater vulnerability the combination can explain increased suicides.

Cox, Paulus, & McCain (1984) reviewed archival records from four state prisons in Illinois, Mississippi, Oklahoma, and Texas from 1952-1980. They report increased pathology with increased crowding. They cite increased violent and non violent deaths, psychiatric commitments, attempted suicide, and self mutilation. They conclude that the primary causes of the observed negative effects related to crowding may be due to cognitive strain, anxiety, fear, or frustration.

Four factors were identified by Gaston (1981) as occurring more often in inmates of a prison. These are addiction, depression, poor reality testing, and poor impulse control. Jail Suicide (Hayes, 1987) adds: talking about suicide, previous attempts, hopelessness, helplessness, unrealistic plans for the future, preoccupied with the past, ineffectual with the present, difficult with others. Many of these are associated with recovery behavior during the first 30 days following detoxification.

Voluntarism (Straffer 1983) is the process of giving up existing "rights" in order to hasten early execution for criminals convicted of homicides. This is relatively rare, but yet occurs often enough to raise questions about its role in contributing to homicide. The author construes this behavior as suicidal.

There is also an ethnic factor to consider for those who commit or attempt suicide in prisons. McCain (1983) reports that medical deaths in those under 45 occur more often in black and non-hispanics compared to the same groups in the community. Copeland (1984) looked at inmates of Dada County Florida detention centers. He reports that most suicides occur in white, younger inmates and tend to be by hanging. The author relates the occurrence of suicide to the degree of supervision provided by the correctional staff. Anson & Cole (1984) confirm. Anson (1983) notes ethnicity and suicide trends in the state prison system of the US. He argues for the use of percentages of ethnic populations to the total in assessing suicide rates or percentages in the incarcerated. He does this by using data available in the 1981 source book of Criminal Justice Statistics published by the U S Department of Justice. He found the number of inmates classified by ethnic status in 51prison systems, including the District of Columbia. With 84 completed suicides he computed rates of commits by all inmates for each prison and compared these to ethnic percentages of the total population confined. Anson computed correlation coefficients for both but analyzed them separately by institutions for high, medium and low numbers of inmate populations.

Anson found that for high population prisons:

  1. an inverse relation between the proportion of black inmates and the rate of suicide,
  2. a direct relation between the proportion of white inmates and the rate of suicide,
  3. no relation for other minorities (Indian, Orientals, & Hispanics).

These relation do not obtain in smaller prisons, and the black relation reverses in the medium size prisons.

The literature of the last ten years continues a negative view of the effects of incarceration on suicide risk. Confinement seems to provide significant stress on vulnerable victims. The situation aggravates the dangers by inadequate psychological care in a context of priority for detention and the use of isolation for suicidal inmates.


There is a general expectation in the public perception that self injury behavior occurs more often among the homosexual population. Rofes (1983) discusses the issues under a provocative title that reflects contemporary attitudes to homosexuality. However, data on this idea is lacking, or when it is elicited suffers from highly selected or volunteer samples which do not reflect all homosexually oriented people. Precision is also lacking as to what is implied by homosexuality. The range of possibilities goes from a one time sexual encounter to exclusive same sex orgasms. In between are men and women who have tried both sexes one or more times (Kinsey et al 1949). Still others belong to the homophile population with none or minimal sexual activities (Motto, 1985). The research situation resembles the same difficulties as with suicide.

Homosexuality has existed from antiquity with variations in tolerance by the larger community. While there is no immediate threat to life, there is stigmatization of people perceived as engaging in homosexual behavior. One significant change is the recent evaluation from a state of abnormally to one of relatively normal choice in sexual orientation (APA, 1987).

Saunders and Valente (1987) reviewed the risk factors in lesbian and gay men for suicidal behavior. They conclude that male and female homosexuals are linked to increased rates. They based this on several sources. They cite "three large well designed studies that found a significant link between suicide attempts and gay and lesbians". They cite Saghir, et al (1970a & b) who are described more fully below. Saunders & Valence note that subsequent researchers (Bell & Weinberg, 1978; Jay & Young, 1979) report a increased presence of high risk factors including 30% more alcohol abuse. Saunders & Valente site a significantly higher rate of alcohol abuse, suicide behavior, and interrupted social ties, all of which are linked to suicide mortality. They note that these studies are not based on direct evaluation of gay men and lesbians and caution that membership in a high risk group does not make every individual a suicidal victim.

Martin (1985) followed 500 psychiatric outpatients for six to twelve years. Death from natural causes was 1 1/3 times more likely than expected but not significant. Unnatural death was 3 1/2 times more likely than expected. Suicide and homicide were particularly high. Initial diagnoses predictive of unnatural death included alcoholism, antisocial personalities, drug addiction, and homosexuality.

Harry (1983) studied four samples of homosexual men and women in San Francisco. The data supported the hypotheses of gender role non conformity during childhood associated with social isolation. The latter is related to subsequent suicidal feelings and attempts. However, men were more vulnerable than women in subsequent suicidality.

Hendin (1982) reviewed the scientific literature in a chapter on homosexuality and also suggests that there does exist evidence for an increased risk of self injury behavior, both attempts and commits. He bases this on earlier studies that are generally small samples, with black and white races of both sexes. The one exception he cites is the Saghir et al study (1970).

Saghir, Robins, Walbran, & Gentry (1970a & b) are cited by the previous authors and are worthy of a fuller description even though their research occurred twenty years ago. Their subjects were cooperating homophiles in the St Louis area. These subjects were homosexual male (n=89) and female (n=57) who were compared to heterosexual male (n=35) and female (n=45). All subjects were single volunteers. The percentage of suicide attempts reported for females were 23% and 5%for homosexual and homosexual orientations, respectively. The difference between homophile and heterosexuals were statistically significant for the females but not for the males. Emotional disorders were slightly greater than in the heterosexuals, but alcohol, and substance abuse histories were significantly greater in the homophiles. The prior attempts were of minimal lethality and do not reflect the higher lethality attempts or commits.

The current status of the research with homosexuals and suicide confirms that they represent a high risk group, but the motives, causality and predictability of future self injuries remains as enigmatic as with other identified higher risk populations.

Black Suicide

The literature between 1977 through 1987 reflects a continuing concern about the ways in which black suicide is different than white and factors that might explain the lack of equivalence. The peak years for black suicide in men continues to be the 20s (Davis, 1070) in contrast in white suicide which increased directly with age. Rates in those under 35 are now equal for blacks and whites men. Female blacks show lower rates in all regions of the US while black men show an increase in the NE and South (Davis, 1979a). Loss of Love object (Datel & Jones 1982) or relations (Bush 1978) and weakening of family and communical ties are reported as contributing factors in black suicide. The situations for black youth continues to grow worse judged by factors correlated with suicide rates. Thus rates of unemployment, delinquency, substance abuse, teen-age pregnancy, and suicides are higher (Gibbs, 1984).

There is little variation by race but wide differences in socioeconomic variables for a suicide potential score devised by Wenz (1978) using census tract infrormation for a northern city. Sotuh (1984) tracked post world war two differences in black and white suicide rates. These were related to the declining racial income inequality. He concludes that the changes are correlated with an explanation based on non white suicide rates moving higher, i.e. towards the level of white suicide rates.

There was support for the hypothesis that suicide attempts would occur more often in a population where black consciousness would be lower (Kirk & Bucher, 1979). There was partial support for additional hypotheses that group cohesiveness was lower and depression higher in the same sample. The sample was based on inner city, young adult black males with suicide attempts in the previous six months and matched controls.

Vargas (1982) studied bereavement in survivors where victims died in one of four modes of death. Subjects were recruited from the metropolitan Los Angeles and consisted of white, black and hispanics females, with approximately 25% additional males. Data was analyzed for the total and for females only. After careful screening for comparability 66 white, 70 black and 65 Hispanic survivors were studied. The author concludes that mode of death and ethnicity of survivor influenced the degree of bereavement. More specifically, while all exhibited emotions o bereavement Hispanic females suffered the greatest measurable grief from homicide victims, and second most from accidental deaths. The lest grief was observed in black and white females who survived victims of natural and suicidal deaths. These findings suggest that the emotions of bereavement are more than a situational depresson and are shaped by the culture of the ethnic group survived.

Martin (1984) reports an inverse relation between suicide rates for race and church attendance by ethnic subpopulations in 1972 to 1948. The annual suicide rates of black, white, male and female subpopulations moved inversely to the same annual figures for church attendance. The author construes this as evidence for the prevention effects of religiosity.

Howze (1977) made in depth clinical study of 13 black women who attempted suicide and attributed their self injury behavior to early and irreversible losses to self esteem and security when they were children. As adults they felt frustrated, guilty, and acted out aggressive impulses.

Convulsive disorders

The scientific literature before 1980 was reviewed by Mathews and Barabas in 1981 and supports the estimate epileptic suicide rates being about four times as high as the general population (1981). The last ten years of suicidology literature continues to report higher suicide rates (Barraclough, 1981), and the use of anti-seizure medications for suicide attempts (McKay, 1979, Hawton, Fagg, & Marsack 1980). Brent (1986) reports data that supports

  1. an increased risk for suicidal behavior in children with epilepsy,
  2. greater lethality of suicide attempts with greater intent present, and
  3. these attempts are associated with the use of phenobarbital as an anticonvulsant.

An overview in the British Medical Journal notes that the suicide rate is four, five or seven times that of the general population depending on the study (1980). Higher risk is attributed to difficulties with housing, schooling, employment, social relations, psychiatric disturbances, and anticonvulsant drugs. The risk can be lessened by education, watching for psychiatric disturbances and monitoring the amounts of drugs prescribed for epileptics according to this articles.

Mittan, Locke & Gatica (1983) report more suicidal thoughts and impulses and higher risk factors for suicide in three samples of epileptics studied in Metropolitan Los Angeles. They tested 193 urban whites, 116 urban blacks, and 89 urban Latinos who were being treated at the Sepulveda VAMC, Martin L King Hospital, and Harbor General Hospital. They conclude that the combnation of depression and hostility was more prevalent in their samples than in normal individuals. An unexpected finding was the significant role played by patient fears about seizures in actual suicide attempts and reasons for their suicidal impulses and fantasies. Mathews and Barabas (1981) reviewed the literature and while confirming the preceding added the following:
Robertson & Trimble also reviewed the literature on depressive illness & epilepsy (1983).

Incarceration suicide among epileptics is especially high and specifically in combination with depressive diagnoses. Gunn (1974) obtained social class, occupational, and family background data from 158 epileptic prisoners, 66 hospitalized epileptics, and 180 non-epileptic prisoners. The epileptic prisoners exhibited a greater degree of depressive and suicidal pathology, especially in terms of drinking behavior.

Roy (1977) compared 17 patients with hysterical fits, previously diagnosed and treated by neurologists as epileptics, and a matched control group of 17 epileptics. They were differentiated by five factors: a family history and personal history of psychiatric disorders, attempted suicide, sexual maladjustment, and a current affective syndrome.

Stewart and Lovitt (1982) addressed the differential diagnosis of hysterical seizures by a double blind methodology using psychological tests with three defined groups: organic (neurological) seizures alone, (N=11), organic (neurogenic) and psychogenic (hysterical), those with purely psychogenic seizures (N=13). The tests used were: the schedule for affective disorders and schizophrenia-life time version (SADS-L), in conjunction with the DSMIIR and the RDC to establish psychiatric diagnoses. Other tests were the WAIS, the Bender-Gestalt, the MMPI and the Rorschach.

Significant differences were found. Patients with neurogenic (organic) seizures were found to have alcoholism, anxiety disorder, and minor affective disorders. Patients with mixed and psychogenic seizures had more severe psychopathology including major affective disorders and major character pathology. Patients with mixed and psychogenic seizures also had a markedly higher incidence of suicide attempts and past history psychiatric treatment. The authors excluded temporal lobe epilepsy patients.

Solomon and Solomon (1982) note the difficulties of differential diagnosis between multiple personality, temporal lobe epilepsy, schizophrenia and other dissociative disorders.

The literature includes treatment approaches as different as art therapy (Naitove, 1983) and partial amygdalectomy (Mempel, 1971) with beneficial effects reported by both authors on relatively small samples.

The review is somewhat disappointing because of the lack of publications dealing with this topic. This may be an artefact of underfunding or lesser priority in editorial policies. Either way it documents an insufficient attention to a significant source of self injury behavior.


This chapter has reviewed the subgroups known to be at higher risk for death suicide. The group at highest risk seems to be those over 65. Youth suicide has increased dramatically since 1958, but is not as extreme as the elderly. In between are the deaths of the new incarcerated, especially the short term prisoners in detention centers. Along the way, are the equally tragic, but less frequent subgroups of female physicians, young blacks, homosexuals, convulsive disorders, victims of alcohol dependence, cancer, aids, borderline personality, and the native Americans.

Other Chapters...

Top Menu Main Menu

Copyright© FRED CUTTER, Phd.