Mark Williams - Suicide & Attempted Suicide - Understanding The Cry of Pain-Penguin UK (2002)
Mark Williams - Suicide & Attempted Suicide - Understanding The Cry of Pain-Penguin UK (2002)
ATTEMPTED SUICIDE
MARK WILLIAMS
Professor Williams is married and has one son and two daughters.
0041486617
Suicide and
Attempted
Suicide
Understanding the Cry o f Pain
MARK W I L L I A M S
/CD
P E N G U IN BOOKS v>
To Phyllis, Rob, Jennie and Annie.
P E N G U IN B O O K S
Penguin B ooks India (P) Ltd, 11 C o m m u n ity C en tre, Panchsheel Park, N ew Delhi - n o 017, India
Penguin B ooks ( N Z ) Ltd, C n r R osedale and A irborn e R oads, A lbany, A uckland, N e w Z ealan d
P enguin B ooks (South Africa) (Pty) Ltd, 24 Sturdee A ven u e, R o seban k 2 19 6 , Sou th A frica
The first edition o f this book, Cry of Pain, aimed to summarize what is
known about suicide and attempted suicide, and to understand the
biological, social and psychological factors that come together to
increase suicide risk. Since then, there have been some changes in the
statistics, and several new research studies. In the United Kingdom
during the mid-1990s, for example, it looked as if the number o f young
men attempting suicide was increasing so fast that it would catch up or
exceed the number o f wom en. This did not occur, and the situation
looks as if it is settling back into a more familiar pattern whereby
w om en (especially young wom en) are more likely to do themselves
harm that is not fatal, whereas men remain more likely to harm
themselves in ways that are fatal. Other studies have confirmed a
number o f findings that appeared to be emerging then: the close link
between attempted suicide and a history o f child sexual abuse in wom en;
the w ay suicidal behaviour is affected by stories in the media; the
problem o f suicide in rural areas, especially among the farming com
munities; and the greater suicide risk o f people, w ho have been in
psychiatric hospital, during the first few weeks after discharge. Since
then the United Kingdom has also seen the outpouring o f public
emotion that occurred following the death o f a princess, an event that
caused an increase in the number o f suicides and attempted suicides,
and further studies have confirmed that the availability o f the means to
harm oneself remains a large factor in many suicides.
The first edition was written with two sorts o f reader in mind. First,
there are health professionals w ho regularly meet people who are, or
Preface to the New Edition
might become, suicidal. But there are also many general readers who
simply want to know more about what makes someone harm them
selves. It is clear that suicidal feelings and behaviour are much broader
than the world o f professional counselling and therapy. There are few
people who have not, at some point in their lives, been touched by
suicide, either by their own suicidal feelings, or by suicidal urges and
behaviour in family or friends. Central to the argument o f Cry o f Pain
was the idea that suicidal behaviour represents a response to a feeling
o f being defeated combined with a feeling o f not being able to escape
the consequences o f defeat. Like an animal caught in a trap, the struggle
to be free is followed by defeat and hopelessness. Social, psychological
and biological /genetic studies are increasingly confirming this view.
People who are suicidal find themselves seeing evidence o f defeat and
rejection everywhere; periods o f high activity (struggle and protest)
are punctuated by periods o f inactivity and despair. This slide into
hopelessness can occur irrespective o f psychiatric diagnosis, for the
feelings o f inescapability arise as much from the failure to control one's
own internal symptoms (thoughts, feelings and behaviour) as they do
from any failure to deal with 'external' problems in living (family, job
and relationships).
In writing this edition, I am grateful to a number o f colleagues,
especially Andy MacLeod, Leslie Pollock, Paul Gilbert, Caroline
Creasey, Bruce Napier, Simant W estley and Keith Hawton. Thanks also
to Stefan McGrath and Caroline Pretty o f Penguin Books, for help,
advice and encouragement throughout the project. I hope that this
book will continue to be helpful, both to the professionals and to all
those who want to understand more about the enigma o f suicide.
MARK WILLIAMS
February 2001
Acknowledgements
xi
Acknowledgements
MARK WILLIAMS
April 1996
Introduction
There is much about suicide that presents a real and often tragic puzzle
to be solved by family and friends, by physicians and other professionals
involved with som eone who has committed or attempted suicide. At
the level o f the individual event the questions are most often: ‘W hy did
they do it?' ‘W hy did they not see there was help available?’ At the
larger level, the questions remain whether a biological, sociological or
psychological explanation will provide the best clue to suicidal
behaviour. Each level o f explanation, seen in isolation, has its drawbacks,
so w e need to see how each type o f explanation m ay fit together with
the others to give a more complete picture.
The large growth in interest during recent years in the subject o f
suicide and suicidal behaviour m ay be due, first, to the rapid increase
since the mid-1970s in suicides by young people (particularly young
men). Although the same period has seen a decrease in the numbers o f
older people committing suicide (the traditionally most vulnerable
group), the change in rates has meant that many more suicide victims
have taken their lives ‘at the prime o f life5, challenging society’s view
o f itself as successful.
Second, there has been an increased wish within the World Health
Organization to challenge health care systems to meet targets to
improve the health o f their populations. Suicide rates are seen as a
visible and quantifiable aspect o f the mental health o f a nation.
Third, interest in suicide has been fuelled by a renewed fascination,
especially in the United States, with euthanasia, medically assisted
suicide and rational suicide (self-deliverance). W hen Derek H um phry’s
Introduction
book Final Exit was published in 1991, with its matter-of-fact, detailed
discussion o f the best w ay o f ending life, it sold more than 500,000
copies in its first year. These are issues and debates in which both public
and professionals are interested.
In the main, tw o types o f book are written in response to this
increased interest. The first type is written for the mental health pro
fessionals. The second type is written for suicidal people, their families,
or for suicide ‘survivors’.1 O f all the books which have attempted to be
useful to both the professional and lay communities, Erwin Stengel’s
Suicide and Attempted Suicide2 became a classic. It was written originally
in the early 1960s, and reprinted several times. It gave the main facts
and figures together with sufficient interpretation to help the reader
understand something o f the mind o f those who feel suicidal. The book
kept apart suicide and attempted suicide, but in its revision Stengel
included an extra discussion about the 'appeal function’ o f attempted
suicide, which he believed had been misunderstood, and more on the
separation o f completed from non-fatal suicide attempts, which was
controversial in its day.
There have been many developments in the field since Stengel’s
death. These include changes in the socio-demographic pattern o f
suicide deaths; knowledge about which social factors are most likely to
produce changes in rates o f suicide; a greater awareness o f the strategies
to prevent suicide (and their limits); more focus on the impact o f the
media on suicidal behaviour; and an understanding o f the w ay in which
depression and hopelessness may act as ‘final com m on pathways’ to
suicidal behaviour. The present book aims to deal with these issues in
such a w ay that a lay audience m ay come to know what are the main
facts and theories, but also so that health and mental health professionals
may have a digest o f a subject which will be constantly in the background
in their practice.
For this reason, the book’s subtitle reflects not the puzzle about
suicidal behaviour, but instead summarizes one possible answer. Suicidal
behaviour is most often a ‘cry o f pain’. This perspective is intended to
capture the w ay in which behaviour can have a communication outcome
without communication being the main motive. It is analogous to the
animal caught in a trap, which cries with pain. The cry is brought about
xiv
Introduction
xv
Introduction
xvi
1
1
Suicide and Attempted Suicide
It is often thought that in Greek and Rom an times there was a tolerance
for suicide. H owever, many philosophers condemned self-destruction.
Pythagoras, for example, compared suicides with soldiers who deserted
their posts. Aristotle said a person should not com m it suicide since such
an act cancelled unilaterally his or her obligations to the state, the
contract between a person and society. On the other hand, Stoic and
Epicurean philosophers believed that suicide could be the right course
o f action in some circumstances, e.g. where there was terminal illness
or unremitting pain. They also allowed that in certain circumstances it
might be an act o f nobility where it was an expression o f political rights
or values.
In the fifth and sixth centuries, attitudes to suicide shifted gradually
away from more permissive Roman philosophical ideals and became
more punitive. St Augustine thought the philosophical support o f sui
cide abhorrent. H owever, the Christian Church has always had to
contend with the difficulty that suicide is nowhere explicitly condemned
2
A Brief History of Suicide
3
Suicide and Attempted Suicide
naked and with a wooden stake through the body. The hole was filled
in, sometimes with the stake showing above the earth so that passers-by
might be reminded o f the awfulness o f the circumstances o f this death.
The clergy did not attend the ceremony. It was carried out by officials
o f the parish, including the church wardens and their assistants.
As part o f the posthumous trial for self-murder, the coroner's court
had to decide whether it was really murder or not. The only mitigating
circumstance was if it was found the person was insane. If someone
killed themselves when mad or mentally incompetent in some w ay,
they were not convicted. Instead o f returning a verdict o f felo de se (a
felon o f himself), the person was deemed to be non compos mentis (not
o f sound mind).
Despite all these popular feelings and the fact that committing suicide
was considered a crime, during the Middle Ages very few juries actually
brought in verdicts o f felo de se. It was not that suicide was rare, but
that juries, often consisting o f local people w ho had sympathy with the
family, were reluctant to see them become paupers. The Crow n had
little control over these local juries, and although from time to time
attempts were made to tighten the enforcement o f the law, this was
rather sporadic. It was not until Tudor times that this aspect o f social
and legal life was more rigorously controlled.
4
A Brief History of Suicide
continued, however. Local officials would often declare that the suicide
victim had very little possessions. In one extreme case the total value
o f a person's possessions was declared to be precisely the value (to the
last ha’penny) o f the value o f the person's debts. In this way, local
coroners' juries attempted still to protect their fellow-villagers and
townsfolk from the excesses o f central government.
The results o f all the governm ent reforms were reflected in what
appears to be a dramatic increase in suicides, or at least those reported
to the King's Bench. The average number o f suicide inquisitions in 1500
was 61 ( felo de se)y and 1 non compos mentis. By 1600 this number had
risen to 873 felo de se and 7 non compos mentis verdicts.
The Tudor revolution in governm ent was reinforced by the attitude
o f the churchmen, w ho continued to emphasize their belief that self-
murder was an expression o f despair brought about by the devil.
Both Calvinists and non-Calvinists were equally vehement in their
condemnation. For example, George Abbot, one time Archbishop o f
Canterbury, declared that suicide was ‘a sin so grievous that scant any
is more heinous unto the Lord'. He was Calvinist, but the anti-Calvinist
Lancelot Andrewes was similarly explicit: ‘It is worse than beastly to
kill or drown or make aw ay with ourselves; the very swine would not
have run into the sea but that they were carried by the Devil' - a
reference to the story o f the Gadarene swine in the Bible (Mark 5,
Luke 8).
In popular stories and in sermons the idea o f Satan playing on a
man's guilt and luring him to his death by abject despair is prominent.
In morality plays, a point is often reached where the devil casts Man
into a despair or ‘wan-hope'. In John Skelton's Magnyficence, a character
is persuaded by Despair that his sins are so bad that God will not forgive
them. Despair urges him to suicide with the words ‘ryd thy selfe rather
than this lyfe for to lede'. The character Mischief appears, offering the
instruments o f self-murder: ‘Lo here is thy knyfe and a halter, and all
were go ferther, spare not thyself, but boldly thee murder.' The central
character is about to use the knife against him self when Hope suddenly
enters and urges him not to kill h im se lf‘against Nature and Kynde'.3
In such a culture, suicide meant the struggle had been lost, the devil
had won. In Christopher M arlow e’s Dr Faustus, the devil instigates
5
Suicide and Attempted Suicide
suicide more explicitly perhaps than in many other allegories. The Good
Angel and the Evil Angel try to persuade Faustus to repent on the one
hand and to despair on the other. Faustus cries, ‘My heart’s so harden’d,
I cannot repent.’ W hen Mephistopheles offers Faustus a dagger after he
has yielded to despair, Faustus cries, 'D am n’d art thou, Faustus, damn’d;
despair and die!’
O f course there is little doubt that m any o f the people who tried to
kill themselves or actually succeeded were extremely depressed. The
fact that non compos mentis verdicts were not brought in for them cannot
hide this fact. In many cases the depression was clearly o f psychotic
proportions. A young Puritan, Nehemiah Wallington, imagined him self
'provoked by the D evil’ to suicide and made n attempts. According to
Nehemiah’s own account, Satan showed him self in various manifes
tations: as a crow, as his sister, as a minister and as a disembodied voice.
The disembodied voice immediately suggests auditory hallucinations
o f the psychotic patient. But the pull to life turned out to be stronger
than the pull to death, despite ‘the temptation o f Satan’ .
Then Satan temted me again and I resisted him again. Then he temted me a
third time, and I yielded unto him and pulled out my knife and put it neere my
throat. Then God o f his goodness caused me to consider what would follow if
I should do so. With that I felle out a weaping and I flong away my knife.
6
A Brief History of Suicide
7
Suicide and Attempted Suicide
no one could be certain that, should a suicide occur, there would not
be a felo de se verdict followed by the requirement to forfeit the person’s
goods. This is well illustrated in Samuel Pepys’s Diaries.
A businessman called Anthony Joyce, kinsman to Pepys, tried to
commit suicide by throwing him self into a pond. Pepys takes up the
story in his diary entry o f 21 January 1667:
Comes news from Kate Joyce that, if I would see her husband alive, I must
come presently. So I to him, and find his breath rattled in the throat; and they
did lay pigeons to his feet, and all despair o f him. It seems, on Thursday last,
he went, sober and quiet, to Islington, and behind one o f the inns, the White
Lion, did fling himself into a pond: was spied by a poor woman, and got out
by some people, and set on his head and got to life: and so his wife and friends
sent for.
Joyce explained to Pepys that he had been led by the devil and that he
had ‘forgot to serve God as he ought’. H owever, it is interesting to note
that Pepys him self believed the real reason for his suicidal behaviour
was that his business had failed after losses sustained in the Great Fire.
Here we see evidence o f the increased secularization o f attitudes. But
side by side with it another theme emerges.
Pepys goes on to give a first-hand account o f the panic a suicide
instilled in the bereaved family: panic that their goods would be forfeit
(even though technically Joyce survived the suicide attempt a few days).
Pepys even agrees to hide some o f the fam ily’s goods to prevent their
being seized and some o f the panic transfers to him. Finally, he uses his
contacts in high places to avert the crisis:
The friends that were there, being now in fear that the goods and estate would
be seized on, though he lived all this while, because o f endeavouring to drown
himself, my cosen did endeavour to remove what she could o f the plate out o f
the house, and desired me to take my flagons; which I did, but in great fear all
the way o f being seized; though there were no reason for it, he not being dead.
So, with Sir D. Gauden, to Guild Hall, to advise with the Towne-Clerke about
the practice o f the City and nation in this case; and he thinks it cannot be found
selfe-murder; but if it be, it will fall, all the estate, to the King. So I to my cosen's
again; where I no sooner come but find that her husband was departed. So at
8
A Brief History of Suicide
their entreaty, I presently to White Hall, and there find Sir W. Coventry: and
he carried me to the King, the Duke o f York being with him, and there told my
story which I had told him: and the King, without more ado, granted that, if it
was found, the estate should be to the widow and children. I presently to each
Secretary’s office, and there left caveats, and so away back to my cosen’s. When
I come thither, I find her all in sorrow, but she and the rest mightily pleased
with my doing this for them; and which, indeed, was a very great courtesy, for
people are looking out for the estate.4
9
Suicide and Attempted Suicide
suicides were permissible and poured scorn on the view that these acts
were carried out by the secret command o f God. This, Donne says, is
mere supposition and cannot be considered a rational argument. Indeed
Donne goes further at one point to argue that one could see the death
o f Jesus o f Nazareth as suicide.
Donne's treatise constituted an extremely rare set o f arguments
since most defendants o f suicide o f the time relied on classical literature.
Even Thom as More, who had written Utopia envisaging that suicide
would be permissible for humanists in that society, had not envisaged
that suicide would be permissible for Christians.
By contrast, the Stoic view was that death ‘unlocked the chains o f
suffering' and no law could forbid it. As Montaigne put it, ‘as I offend
not the laws made against thieves when I cut my own purse, so am I
nothing tied unto laws made against murderers, if I deprive m yself o f
m y own life'.5 Far from imagining that suicide was against the natural
law, writers relying on Stoic arguments argued that no one was obliged
to live when his or her death might promote some greater good. In
fact, they thought that suicide in certain circumstances was consonant
with the natural law.
Yet John Donne was so worried about how Biathanatos would be
received that he refused to permit it to be published in his own lifetime.
In presenting the manuscript to Sir Robert Ker, he commented that it
was a book ‘written by Jack Donne and not by Dr Donne'. It was finally
published in 1647. It appears that Donne him self shied away from his
own conclusions - not an uncommon occurrence. It was one thing to
use arguments for the justification o f suicide to undermine the state’s
wish to punish severely a person who took their own life, another to
use arguments in such a w ay as to give the impression that suicide was
an option implying that life could be ended lightly as an acceptable way
o f dealing with life’s difficulties. W hy such reluctance to publish? Donne
was a poet, but also, as Dean o f St Paul’s, a pastor. As a pastor he may
have had to deal with people bereaved by suicide, and m ay well have
felt it anathema to take a punitive line. But the complexity o f what is
said as part o f a private pastoral relationship does not transfer well into
public pronouncement or public policy.
10
A Brief History of Suicide
Madness as illness
By the end o f the 1700s, the felo de se verdicts were as rare as the non
compos mentis verdicts had been in the early Tudor period. The prevailing
belief had become that suicide was either a product o f rational choice
or a medical calamity.
Once again a case-by-case leniency exercised by individual courts
confirmed the change. The coroner's ju ry became increasingly reluctant
to enforce penalties for self-murder. T w o trends drove the change:
increasing resistance to a law which seemed too draconian, and increas
ing willingness to see suicide as the product o f an unbalanced mind.
Assisting the first trend against property forfeit was the increased belief,
expressed by John Locke, that no governm ent should interfere with the
cult o f private property. Governm ent should not ‘take to themselves
the whole or any part o f the subjects property without their own
consent'. Such beliefs had been confirmed and emphasized by the
English Revolution o f 1688.
Assisting the second trend towards mental illness as an explanation
was the rejection o f religious enthusiasm, whether Puritan or Catholic.
Although it is likely that popular belief in supernatural forces continued
(and still continues), the intellectual elite gradually, during the Enlight
enment, came to reject almost absolutely the belief in supernatural
intervention in the natural world. This gave w ay, in the case o f suicide,
to a medical interpretation which greatly influenced the middle-ranking
men w ho served on coroners' juries. N ow , evidence o f melancholy,
previously evidence that the person had given in to the devil, was found
sufficient proof that a person’s balance o f mind was disturbed.
The suicide verdicts reflect the trends. Less than 7 per cent o f the
suicides reported to the King’s Bench were declared to have been non
compos mentis in the middle o f the seventeenth century. By the 1690s,
around 30 per cent o f suicides were thought to be non compos mentis,
and this proportion steadily increased over the next few decades into
the eighteenth century. Forty per cent o f verdicts were non compos
mentis in 1710, and by the last third o f the eighteenth century this had
become the usual verdict.
11
Suicide and Attempted Suicide
Indeed, in Norwich the figures between 1670 and 1799 show the
percentage non compos mentis at the early part o f that period as being
quite low, below 30 per cent; but by 1770-1800100 per cent o f the verdicts
were non compos mentis. The non compos mentis verdict itself implicitly
rejected the folklore and religious interpretations o f suicide as a super
natural intervention by the devil. Instead it substituted a medical exp
lanation which declared explicitly that the suicide was excusable.
The increased tolerance for self-murder can be seen in various writers
o f the time. William Ram esey in The Gentlemen's Companion (1672)
suggested that those who killed themselves, because they were fre
quently the victims o f mental illness,
One consequence o f this change was that no more were non compos
mentis verdicts used simply as a device for protecting some and not
others. During the medieval and Tudor periods a person’s social stand
ing, his personality, his relationship with his neighbours, and his sur
vivors’ claims on the sympathy o f the local community all played a part
in determining which verdict was brought in by the coroners’ courts
(as in the Pepys example). Families o f the rich who committed suicide
were able to bribe the coroner or the jury, or perhaps more com m only
could afford to bring in an expert medical opinion to certify that
the person had undoubtedly been unbalanced at the time o f death.
Gradually, as coroners’ juries became more reluctant to find the suicide
guilty o ffelo de se, and more likely to bring a non compos mentis verdict,
these differences and influences became less and less significant.
Despite these changes, the law o f self-murder remained on the
statute books. It was deemed necessary to discourage people from
suicide (it was thought suicide would cause a disintegration o f society)
and offered a means o f condemning men and wom en who had com m it
ted suicide to escape punishment after being found guilty o f an offence.
In such cases, the rituals associated with self-murder were used as a
12
A Brief History of Suicide
In the early days o f the colonies in America, most colonies adopted the
traditional English punishments though not all enforced the laws o f
forfeiture. Pennsylvania and Delaware explicitly abolished forfeiture in
1701. After the Revolution, Maryland and N ew Jersey decriminalized
suicide as part o f their constitutions. Thom as Jefferson attempted to
abolish forfeiture and argued against punishing suicide. He also pointed
out that juries disapproved o f such severity and it was therefore sensible
to get rid o f the severe rules altogether.
The enforcement o f laws on suicide in Germany, Spain and Italy is
an unknown quantity. Few records remain. H ow ever, there is some
evidence from France and parts o f Switzerland. French law consisted
o f a collection o f regional customs and codes, and different punishments
were stipulated in different regions o f the country. In some places
penalties were worse than those imposed in England, in others they
were mild. The m ajor difference, however, was that throughout France
the laws against self-murder were applied very infrequently. In 1670
they were standardized in a law decreeing that every convicted suicide
should forfeit his goods and be drawn on a hurdle to a profane grave.
But interestingly, after this standardization, the customs and laws were
13
Suicide and Attempted Suicide
enforced even less often. By the time it took place, the growth in Enlight
enment philosophies had begun to undermine the case for such punitive
handling o f suicide. In 1770 in France, the Rites o f Desecration were abol
ished. Suicide was decriminalized completely 21 years later in 1791.
Switzerland, or at least the Swiss city states, Geneva and Zurich,
were more rigorous in attempts to enforce regulations against self-
murder. Their Calvinist regimes attempted to impose strict religious
rules and customs. H ow ever, while suicides were often convicted, they
seldom gave rise to severe punishment. As in France, by the time the
authorities tried to reform the punishment o f suicide, opposing trends
had already m oved towards the secularization o f suicide. The last
example in Switzerland o f dragging the corpse o f a suicide victim
through the streets occurred in 1732, and although the old law remained
in force, as it did in France, until the early 1790s, the city officials in
Geneva, for example, declared in 1735 that all suicides w ere insane and
were to be spared punishment.
14
A Brief History of Suicide
Parliament did not abolish the religious penalties for self-murder until
1823, and did not abolish the secular punishments until 1870. W hy it
took so long to change these laws remains something o f a mystery. The
main reason seems to have been the relation between the law on suicide
and other criminal laws.
The problem o f deterrence was the main issue. It is somewhat
bizarre now to recall that divines such as John W esley called for victims
o f suicide to be gibbeted and for their corpses to be left to rot. In 1823,
however, Parliament acted to repeal the custom o f profane burial. Sir
Jam es Macintosh in the House o f Com m ons declared that punishing
suicide was ‘an act o f brutal folly’. Although the Com m ons defeated
Macintosh’s resolution, as soon as it had failed in its passage through
the House T. B. Lennard introduced a separate Bill to abolish the ritual
penalties for felo de se and substitute the milder sanction o f night-time
burial. He called the Rites o f Desecration ‘an odious and disgusting
cerem ony’ . Such practices were useless as regards the dead and ‘only
tortured the living*. The Bill passed with only one amendment to
make it clear that this change in the law did not mean that clergymen
15
Suicide and Attempted Suicide
were required to perform the usual rites over the bodies o f suicides.
Instead the services they provided were left to their own devising.
Some conservative clergymen resisted pressure to read the burial
rites, and there is evidence that in the dying years o f the nineteenth
century the controversy am ong the Anglican clergy raged as fiercely
as ever.
The provisions for the forfeiture o f property for self-murder were
repealed in 1870 as part o f an omnibus Bill that covered felonies o f every
kind. Felo de se is mentioned only once in the Bill itself, and suicide was
not discussed in the debates on the Bill, presumably because forfeiture
for suicide was such a rare event and no longer a current topic.
One interesting development which arguably allowed the law on
suicide to change was the gradual emergence o f a common-law crime
o f attempted suicide. In earlier centuries, suicide had been against the
law, but it had not been a crime to try and kill oneself. During the 1700s,
people who ‘attempted suicide' were sent to madhouses, gaols or
workhouses to prevent them killing themselves. H owever, during the
nineteenth century there is evidence that, in London at least, people
increasingly began to be arrested for ‘attempting suicide’ . By 1850 this
crime was recognized by the courts, and by the later part o f the century
the courts all over England were arresting and occasionally trying
people for attempted suicide.
In part this trend arose from the increase in police forces throughout
the country. Previously there would have been very few officials who
could have enforced such a law had it existed. But the rise in prosecutions
for attempted suicide was also an answer to those who complained that
the law o f suicide punished the innocent (the family) and spared the guilty
(the one who committed suicide). After about 1850, the law could be used
as a measure to punish (and perhaps rehabilitate) these would-be suicides.
The criminal status o f suicidal behaviour did not end for the United
Kingdom until 1961, when Parliament repealed the com m on-law felony
o f self-murder. Coroners and physicians subsequently no longer needed
to attest that someone who killed him or herself had done so when ‘the
balance o f the mind was disturbed’ .
16
A Brief History of Suicide
Concluding remarks
17
2
Suicide is the most individual o f acts, yet it happens too often for us to
ignore the possibility that there m ay be trends within the data when
these individual acts are examined together. Some questions can only
be answered by taking account o f entire populations. Is the number o f
suicides on the increase or decrease - and in which groups o f people?
H ow is the figure related to gender, and to social class? W hat means do
most people use? Are any trends limited to a single country - do suicidal
trends respect national, cultural or ethnic boundaries? H ow is suicide
related to socio-economic conditions? And what about unemployment
and other such sources o f stress - can we say definitively whether they
affect the suicide rate, and if not, w hy not? This chapter aims to address
each question as essential background to later chapters which ask for
explanations.
Durkheim defines suicide as ‘the termination o f an individual's life
resulting directly from a positive or negative act o f the victim him self
which he knows will produce this fatal result'.1 But are such definitions
useful in the real world o f the inquest? Interpreting suicide statistics is
often said to be difficult because they are inherently unreliable ‘Suicide’
is a legal definition, so official statistics inevitably underestimate the
number o f suicides. The authorities have to be certain beyond reason
able doubt that death was caused by self-inflicted injury and that the
deceased intended to kill him or herself One result, for example, is that
passive methods o f dying are less likely to be called suicide than active
methods. Thus drowning is called suicide in 54 per cent o f cases whereas
hanging is called suicide in 98 per cent. Another result o f definition
18
Suicide: Facts and Figures
19
Suicide and Attempted Suicide
20
o
o
CD
O
O
r<S
IT\
O
O
O
O
O
of
ITN
O
O
O
of
o
o
LTJ
sipeap jo J9quinj\i
o
o
O
M
o
O
in
O
21
Suicide: Facts and Figures
Figure 2,1 Number of suicide deaths by gender, 1911-99, England and Wales.
Source : Office of National Statistics.
Suicide and Attempted Suicide
40
Male
35
Rate per 100,000 population
30
25
20
15
10
35-44 —+ 75-84
45-54 — 85 and over
Year
Rate per 100,000 population
Year
22
Suicide: Facts and Figures
and older adult rates for suicides have been converging over the past 30
or 40 years. Currently over 40 per cent o f all suicides occur in males
between the ages o f 15-44 years, and this differs from previous times
when the rates were much higher for older people. The situation by
the end o f the 1990s was that for men between 25 and 75 there was little
difference in age-related suicide risk.
W hy this steep increase in young male suicides? One clue lies in the
changing pattern o f marriage and divorce. Divorced men have higher
rates o f suicide, so if the divorce rate rises, the suicide rate will rise
with it.
Table 2.1 shows the suicide rates for several different countries reporting
to the W orld Health Organization. The rates for many different coun
tries follow the patterns seen in the United Kingdom in many respects.
In every country but China, the suicide rate is low er for wom en than
for men. These statistics do not show the trends over time, which differ
between men and wom en. In many European countries, for example,
the suicide rate in wom en has remained flat or decreased while the rate
for men has increased, most markedly in Ireland, Spain, Netherlands,
N orw ay, Belgium, France, Luxem bourg, Denmark, Finland and North
ern Ireland. A notable exception to international trends is in West
Germany, where there has been a substantial decline in both male and
female suicide.
Are the trends within the suicide statistics from North America
similar to European trends in general and U K trends in particular?7
During the past century, the rate in the United States has averaged a
rate rather similar to that in the United Kingdom: 12.5/100,000. A high
rate o f 17.4 was observed during the Depression in 1932 and a low point
o f 9.8 in 1957. This rose again steadily to a peak o f 13.3 in 1977, then
dipped to a rate o f 12.7 in 1987. The US data since 1950 are shown in
Figure 2.3. These rates are also very similar to those o f Canada, where
the 1995 rate was 13.5 per 100,000 (21.5 per 100,000 in men and 5.4 per
100,000 in women).
23
Suicide and Attempted Suicide
All Ages
24
Suicide: Facts and Figures
Methods of suicide
25
26
Suicide and Attempted Suicide
Figure 2.3 Comparison of suicide rates for men and women for different ethnic groups in the United States,
1950-96.
Source: The National Center for Health Statistics, Vital Statistics of the United States
Suicide: Facts and Figures
Figure 2.4 Distribution of recorded suicides by gender, year of death and method, England and Wales.
0
}00
00
ir\
00
C1O
00
0
00
00
r\
m rC
1
r\
o
7'
00
VO
Vin
O
cu , ////\
<
1
Jo
VO
0
00
1
00
r\
1n
£
o
c V'
<u 'm m m 00
'O
in
V
T
O
------------- ---------- ro
VO
, mu,’ v . ; -!— . >.t , ’ K • ■» ; j " 7 -
O
^ ,/v -^ . " ^ 5 x -* . •*
a____________ _ __________ 00
in
7*^7
o
______________________I 0T0
a
o o o o o o o o
o
<s
00 V
O (S o 00 o (S
27
Suicide and Attempted Suicide
hanging rose from 31 per cent in 1990 to 52 per cent in 1998. In addition,
the increase in hanging continued for wom en too, with 31 per cent
using this method in 1998. (Note that these proportions are higher
than estimates based on combining suicides with open verdicts, where
proportions for hanging o f 31 and 15 per cent for men and wom en
respectively were reported for 1996 to 1998. This difference is due to
the fact that o f all methods o f death, hanging is the most likely to result
in a verdict o f suicide, whereas other methods, such as self-poisoning
or drowning often end in an open verdict. The data for 1998 given here
are for suicides alone, since the data in Figure 2.4 are based on verdicts
o f suicides only.)
Table 2.2 shows the distribution o f methods included in the ‘other’
category. As can be seen the largest proportion o f other methods is
drowning. The table also shows that there has been a recent decrease
in the use o f firearms for both men and wom en. H owever, jum ping
from a height or in front o f m oving vehicles remains a relatively high
category o f ‘other methods’ . Over the last 45 years there has been a
reduction in the proportion o f drowning for both men and wom en.
The apparent increase during the mid 1900s for men m ay have been
temporary, since latest figures (for 1998) show a reduced proportion o f
12 per cent.
When methods o f suicide are examined for the US population, there
are clear differences from the United Kingdom. Males still choose more
violent means than females, but use many m ore firearms (64 per cent).
As in the United Kingdom, females are more likely to choose non-violent
means, e.g. drug overdose (38 per cent). (See Figure 2.5.)
Gender
In almost all countries, the rates for men exceed those for wom en by a
factor o f between 2:1 (Denmark, Netherlands and Sweden) and 5:1
(Finland).8 This sex ratio has existed from the earliest days. Historical
data show that men have committed suicide at least twice as often as
wom en in almost all samples o f data that survive. For example, the sex
28
Suicide: Facts and Figures
Men
Drowning 35 14 20
Firearms 21 27 16
Cutting 20 10 8
Jumping 9 16 22
Other 15 33 34
Women
Drowning 60 33 33
Firearms 3 5 1
Cutting 10 9 4
Jumping 15 23 19
Other 13 30 43
ratio in the King’s Bench data for 1485-17149 was 1.8 men for every 1
wom an; for Norwich in the eighteenth century it was 2.3:1; and for
greater London it was 2:1. Completed suicide seems always to have
been an act more associated with men than women.
As we have seen in the data for England and Wales (Figure 2.1), the
pattern during most o f the twentieth century, in which wom en and
m en’s suicide rates followed each other fairly closely (albeit at different
overall levels), has now changed. The rates for wom en fell during the
1970s and 1980s, while those for men rose. The net effect was to reduce
the overall figures for wom en over this period, but to leave the rates
for men (taking all ages together) rising steadily.
29
Suicide method
Q
3
u
toJO
C3 p
u
C rP
*-* C
O OJ
T3 C/)
+-»
<L) <v
C/3
C/3 C/3
30
X
g
c03
Suicide and Attempted Suicide
'So
W)
£
6
it
C/)
03
o
o
o
o
O
o
<s
CO
a
c
ctJ
v
<
u
4->
uu
<u
'bJO
Figure 2.5 Number and percentages of suicide method by gender, US rate 1996.
Source: The National Center for Health Statistics, Vital Statistics of the United States
Suicide: Facts and Figures
Ethnic group
The suicide rate for different ethnic groupings is different: whites in the
United States are approximately twice as likely to commit suicide as
non-whites (see Figure 2.3). This difference is particularly pronounced
for the older age group. In the United States, native American young
people have a very high rate o f suicide, the highest rate being in those
tribes undergoing the greatest and fastest cultural assimilation. In the
United Kingdom, there have been reports that the suicide rate among
Asian w om en is almost double that for other wom en o f the same age,
and although research has not yet established the reason, difficulties in
resolving cultural clashes is a clear possibility.
Sexual orientation
A 1989 report from the US Department o f Health and Human Services,
the Secretary's Task Force Report on Youth Suicide, suggested that gay
youth were tw o or three times more likely to complete suicide than
other young people.10 It suggested that lesbian and gay young people
m ay comprise up to 30 per cent o f youth suicides annually. However,
this part o f the report was not backed by any evidence. Elsewhere it
admitted that this was a neglected area, with virtually no research on
youth suicide having taken sexual orientation into account. The two
studies that have been published (both US surveys) suggest much more
conservative figures: that between 2.5 and 5 per cent o f suicides are gay,
and that these figures are not significantly different from the proportions
one would expect by chance. One study11 surveyed 283 adolescent and
adult suicides, finding that 5 per cent were gay. Another study, using
psychological autopsy techniques, studied 120 consecutive suicides
under the age o f 20 years, and 147 community age-, sex- and ethnic-
matched controls living in the Greater N ew York City area.12 Three
(3.5 per cent) o f the 95 teenage suicides in this sample (2.5 per cent o f
the total sample o f 120) were reported to have had a homosexual
experience. Both studies concluded that psychiatric disturbance and/or
substance abuse were critical predisposing factors regardless o f sexual
orientation.13
31
Suicide and Attempted Suicide
Religious affiliation
Rates vary according to religious affiliations. A number o f studies have
found that Protestants are more likely to commit suicide than Catholics
or Jew s. The Catholic countries o f Spain, Italy, Portugal, Greece, Ireland
and Poland have fairly low suicide rates. Yet there are other Catholic
parts o f Europe, such as France, where the suicide rate is relatively
high, though the increased secularization o f France, and reduced church-
going, may be responsible. In one study in N ew York, rates for Prot
estants per 100,000 were 31.4, those for Catholics 10.9 and those for Jew s
15.5. H owever, many studies have also found that churchgoing in
general is associated with low er suicide risk, presumably because o f the
social network and sense o f meaning in times o f suffering it can som e
times provide.
Social class
Table 2.3 shows the proportional mortality rates calculated for different
social classes in England and Wales. Social class I is defined as pro
fessional; II as intermediate (lower professional and executive); III as
skilled (manual and non-manual); IV as partly skilled; V as unskilled.
The proportional mortality ratio (PM R) is a ratio which enables the
impact o f a disease upon an exposed population to be examined.
Table 2.3 shows that the risk o f suicide is greater in social class I and
social class V, with the highest risk o f all carried by those with no
occupation at the time o f the census.
I 116 1 319
II 99 5 202
IIIN 102 3 083
HIM 87 9 004
IV 100 5 649
V hi 3 582
‘Unoccupied’ 126 3885
32
Suicide: Facts and Figures
Unemployment
Economic conditions have been studied many times, with some conflict
ing results. Studies on unemployment in suicide have found that men
who are unemployed and seeking w ork are at two or three times greater
risk o f suicide death than the average. Suicide is thought by many investi
gators to be one o f the possible consequences o f unemployment, and yet,
in the United Kingdom at least, there seems little relationship between
change in unemployment levels and the suicide rate over tim e.14
National unemployment rates in the United Kingdom increased
substantially at the end o f the 1970s, stayed high until the mid 1980s,
started to decline until the late 1980s and then began to rise again. In
the 1981-90 period male suicide rates were at their highest at the time
when unemployment rates were lowest. H owever, the relationship is
a complicated one. Most unemployment is short term (under six
months) and unemployment totals are net figures taken at a fixed point
in time. This means that unemployment rates never give the whole
picture o f the distress caused by unemployment. Stephen Platt has
pointed out that the impact o f unemployment depends on whether an
entire firm has been made unemployed in a community (in which case
there may be some increase in social support - an ameliorating factor
in suicide risks). By contrast, if an individual is made unemployed,
independent o f other people becoming unemployed, then this may be
more o f a stress factor.
In fact, other studies have found that the United Kingdom is an
exception in not finding a connection between unemployment and
33
Suicide and Attempted Suicide
suicide. Boor (in 1980)15 compared the unemployment rates with the
suicide rates in the following eight countries: United States, Canada,
Sweden, France, Great Britain, Germ any, Italy and Japan. The suicide
and unemployment rate correlated positively in all but two countries,
Italy and Great Britain. Boor also found that the relationship between
unemployment and suicide held for both men and wom en. The rate o f
unemployment among suicide victims is 50 per cent across a number
o f studies.
Rural areas
Geographical region also plays a part in different suicide rates across
countries. As in the United Kingdom, where rural areas have the highest
rates, so in the United States the mountain regions have the highest
overall rate, Nevada ranking highest at 17.6 per 100,000. In a study
with m y colleagues Leslie Pollock, John Hollis and Patrick Vesey, we
examined the suicide rates in Powys, Wales, one o f the least densely
populated areas o f Europe. W e found a suicide rate greatly in excess o f
the U K average (around 20 per 100,000). Looking closely at the data for
a three-year period, w e found that farmers or farm labourers accounted
for 22 per cent o f this figure, though they only represented 11 per cent
o f the population. W e also examined whether it was the lack o f access
to primary care services in rural areas that determined the high rates,
but could find no evidence that these individuals had visited their
doctor any less frequently than happens in more urban areas. Rather, it
appeared to be the greater availability o f lethal weapons that accounted
for the high rates, with firearms used more com m only than the national
average. A study by Keith Hawton and colleagues reinforces this con
clusion.16 Between 1981 and 1993 there were 719 suicides or open verdicts
o f farmers in England and Wales. Firearms were used in 40 per cent o f
cases (compared with the national average o f 4.3 per cent), and hanging
in 30 per cent (compared with the national average o f 23 per cent).
Hawton pointed out how the easy availability o f such lethal means was
a critical factor in the high suicide rate within this community.
34
Suicide: Facts and Figures
Imprisonment
The number o f prison suicides a year has risen considerably since the
early 1960s. In 1961 there were 15 suicides in the United Kingdom, and
this figure had increased to 67 per year by 1997. Most o f these are in the
15-44 age group o f men. This increase represents 4 per cent o f the
increase in the young male suicide rate over this period. The increase
has given cause for concern, especially since most o f the suicides o f the
under-45 age group are male prisoners on remand, not yet found guilty
o f any crime. D ooley examined 300 prison deaths between 1972 and
198717 and analysed the reasons for each suicide. Forty per cent o f
suicides could be attributable to the prison environment (overcrowding,
etc.). Fifteen per cent were attributable to outside pressures, 12 per cent
to guilt feelings for the offence and 22 per cent to already diagnosed
mental illness.
Medical illness
There are several medical conditions that increase the risk o f suicide.
For example, the risk o f suicide among those with epilepsy is four times
that o f normal controls, and for temporal-lobe epilepsy it is 25 times
greater than controls. Cancer has been associated with increased risk o f
suicide, the risk highest immediately following diagnosis, and in those
receiving chemotherapy. (Those with a severe and morbid fear o f
cancer are also at increased risk.) People with peptic ulceration have a
high risk, probably because o f the prevalence o f alcoholism as a cause
o f ulceration. With Huntington’s chorea there is a sixfold increase in
the risk o f suicide compared with the general population, both in those
suffering from the disease and in family members who may also be at
risk but as yet have no overt symptoms. People undergoing renal
dialysis have been said to have a greater incidence o f suicide compared
with the general population, but there are not enough reliable studies
to estimate the risk exactly. There is an increased risk in people who
have spinal-cord injuries (a fivefold increase compared with the general
population), and in those who have multiple sclerosis (a threefold
increase for men and a twofold increase for women). The risk o f suicide
in those who have Aids was some 36 times greater than the general
35
Suicide and Attempted Suicide
population in the 1980s, but this figure appeared to reduce during the
1980s to a sevenfold risk.18
In many cases the co-occurrence o f depression with physical illness
combines to increase the suicide risk. Medical illness can precipitate
severe depression and also produce an organic mental disorder leading
to perceptual, cognitive and m ood changes. In other cases, the suicide
appears to be a rational act where the person decides that the prospect o f
suffering and loss o f dignity is intolerable. H owever, research evidence
suggests that suicide in the physically ill very rarely occurs in the absence
o f psychiatric disorder.
It is important, therefore, for the physician and other health pro
fessionals not to ignore associated psychiatric problems that co-occur
with physical illness. For example, depression associated with terminal
cancer is often ignored or thought to be ‘normal’ . It therefore goes
untreated despite the fact that psychological treatment is often very
helpful in alleviating the depression. Depression in such patients goes
beyond sadness and includes a sense o f worthlessness and failure which
is not justified on the basis o f the person’s experience.19
Lifespan issues
36
Suicide: Facts and Figures
opportunities. These are just the groups that have suffered the greatest
rise in suicide rate over the past 20 or 30 years.
Different rates o f medical illness across the lifespan might also contrib
ute to differences between different ages. Illness is associated with 50
per cent o f adult completed suicides, contributing more prominently
for older adults. Medical illness in older adults contributes to suicidal
risk in a number o f possible ways. Depression m ay be worsened by the
fact that the physical illness puts an increased burden on the fewer
social supports that exist. There are also cognitive impairments often
associated with ageing in general and with physical illness in the aged
in particular. Finally, the use o f certain medications to alleviate the
physical illness sometimes lowers m ood and impairs judgement.
Do different psychiatric disorders increase the risk o f suicide at differ
ent stages in life? W e shall see in the next chapter that the individuals
most at risk are those suffering from depression and schizophrenia.
Right across the lifespan, if these conditions co-occur with particular
personality traits such as impulsivity and aggression, and with alcohol
and substance abuse, the risk is increased. H owever, there are changes
over the lifespan in the particular combinations o f these disorders.
Adolescents who commit suicide tend to show a combination o f
depression and conduct disorders, though the depression is often only
recognized after the event, as research by Alan Apter and Israel Orbach
shows.20
It is possible that it is psychiatric disorder per se which is the important
issue. Different psychiatric disorders m ay appear to raise the suicide
risk at different points in the lifespan simply because they are more
prevalent at that point. Some have argued that it is changes in the action
o f the major neurotransmitters over the lifespan which are implicated.
For example, brain serotonin metabolism is believed to change during
the adult lifespan. This remains speculative.
37
Suicide and Attempted Suicide
deaths am ong men aged 15-24, and a fivefold increase for males aged
25-44. There are similar trends am ong wom en, but the rates are
generally much lower. The data for alcohol show a similar pattern,
with men aged 25-44 showing the largest increase in rates (more than
fourfold between 1968 and 1990) followed by men aged 15-24. Once
again, rates for younger w om en rose as well, but the levels were very
low compared with those for men.
Alcohol and substance abuse represent major risk factors for suicide
right across the lifespan. The prevalence o f alcohol and drug abuse
increases steadily from 15 until about 45, then declines. The number o f
years somebody who completes suicide has typically been abusing
alcohol is between 20 and 25, possibly because, as alcoholism progresses,
it destroys those factors known to protect against suicide. First, it
destroys social supports because o f the alcoholism-related guilt and
anger. Second, it destroys intellectual function through brain damage
and brings about reduced health and increased incidence o f medical
complications. Third, chronic alcoholism reduces personal control and
increases helplessness. Since, however, 60 to 70 per cent o f patients with
a diagnosis o f alcohol problems have additional psychiatric diagnoses, it
is possible that the substance abuse exacerbates the course o f any
psychiatric illness. George M urphy’s studies show that alcohol and drug
abuse are most lethal when they occur alongside an affective disorder
(such as depression).21
H ow ever, alcohol and substance abuse raises the risk o f suicide as
soon as the abuse starts. Substance abuse has been diagnosed in over
one third o f young people committing suicide.22
In the 1990s the United States found itself with one o f the highest suicide
rates for young men in the world, even exceeding Japan and Sweden.
The question o f w hy there had been such an increase in youth suicide
became very pressing indeed. It appeared to parallel the increase in risk
factors associated with suicide in young people: depression, conduct
disorders and substance misuse. In addition there are population effects
38
Suicide: Facts and Figures
A cohort effect?
Studies by Murphy and W etzel for the United States, by Solomon and
Hellon in Alberta, Canada, and by Goldney and Katsikitis for Australia
followed through the suicide rates for people bom within a few years
o f each other - each ‘cohort's’ suicide rate could then be compared
with cohorts bom at other times.23 Each study found that if a cohort
had a high suicide rate at a young age, then the increased rate continued
throughout the life o f that cohort. Is this also true o f England and
Wales?
Cohort data published by the Office o f Population Census and
Surveys in 199224 confirm the existence o f a cohort effect for both men
and wom en for the suicide rate in England and Wales. For men, more
39
Suicide and Attempted Suicide
recent cohorts have higher age-for-age mortality than earlier ones. That
is people bom between 1962 and 1971 have a higher risk o f suicide than
those bom between 1952 and 1961, who in turn have a higher risk o f
suicide than those bom between 1942 and 1951. The rates for the more
recent cohorts may not yet have peaked, in which case we may expect
a continued rise in suicide rates by younger men for some years to
come (as the high-risk cohorts m ove through the different age ranges).
Recent data show that the cohort bom around 1971 has a still higher
rate.
For wom en over 45 years, successive cohorts show lower age-for-age
suicide risk, people bom in the 1910s, 1920s, 1930s and 1940s showing
gradually reducing rates. If this trend continues, we m ay expect that
rates for wom en will remain low.
But how could a cohort effect come about? Some studies have
pointed to the size o f a birth cohort. Where there is a large number o f
babies bom during the same period, there is increased competition for
resources (as occurred at the turn o f the twentieth century). This
results in higher unemployment and decreased access to educational
opportunities for that cohort.
A period effect?
It is also possible that the recent trend for increases in young male
suicides does not represent a cohort effect (i.e. people bom recently
m ay not have a greater disposition to suicide), but might rather be a
period effect (something in current society is pushing up the rates).
Taking the data for the 1936 cohort for wom en, when they were 25-29,
their curve was rising sharply at this time, but it turned out to result
from the 1960s period effect. On that occasion, the same period effect
was evident for older age groups too.
Although the recent increase in young male suicides appears not to
be reflected in older males, it may nevertheless be a period effect
combining with a cohort effect. W hatever current stress factors there
may be, they may be differentially affecting young males. This is
increasingly likely the larger the size o f the birth cohort. Throughout
their lives, more people in the larger cohort will be chasing whatever
resources are available. The effect o f the cohort size will depend upon
40
Suicide: Facts and Figures
the amount o f resources. A large birth cohort will not always have this
effect, if, by chance, there are sufficient resources to sustain members
o f the cohort throughout their lives.
H owever, if the number o f jobs in a country falls during a recession,
then the effects are likely to fall disproportionately on those in the
population who are members o f the large cohort, whatever age the
cohort has reached. Suddenly too many people o f about the same age
will be chasing too few resources.25 This is consistent with the sugges
tion that the reduced status o f men caused by changes in employment
prospects and family cohesion, together with reduced anticipation o f
long-term roles as husband and parent, has produced a rise o f alcohol
and drug abuse in this subpopulation that contributes to the increased
suicide rate. Is there evidence to support this particular causal pathway?
41
Suicide and Attempted Suicide
42
Suicide: Facts and Figures
Concluding remarks
43
Suicide and Attempted Suicide
in the United Kingdom, however, those who commit suicide are very
much more likely to have been unemployed at the time than a person
matched for age.
44
3
Psychiatric and Social. Factors
in Suicide
45
Suicide and Attempted Suicide
46
Psychiatric and Social Factors in Suicide
47
Suicide and Attempted Suicide
48
Psychiatric and Social Factors in Suicide
Neuroses 2-5%
49
Suicide and Attempted Suicide
50
Psychiatric and Social Factors in Suicide
51
Suicide and Attempted Suicide
52
Psychiatric and Social Factors in Suicide
or others. The person sees a knife, or walks near a cliff-edge, and the
idea o f suicide comes suddenly into mind. If the patient survives, they
m ay say afterwards that they were not aware o f or cannot remember
any reason for their action.
In summarizing the psychiatric suicides, Durkheim notes that all
such suicides are either ‘devoid o f any m otive' (obsessive and impulsive
suicides) or are determined by ‘purely imaginary m otives’ (maniacal
and melancholy suicides). W here psychiatric disturbance itself explains
the motivation, it does so by indicating how the illness acts to produce
the behaviour either in the absence o f m otive or through the w ay it
distorts the w ay the person views their world. Durkheim points out
that ‘many voluntary deaths fall into neither category; the majority
have motives, and motives not unfounded in reality. Not every suicide
can therefore be considered insane, without doing violence to language/
53
Suicide and Attempted Suicide
54
Psychiatric and Social Factors in Suicide
• older age;
• gravity o f previous attempt;
• living alone;
• psychosis;
• left: a suicide note;
• unemployed or retired;
• from a broken home;
• men more than wom en;
• multiple attempts;
• not married;
• poor physical health;
• lethal method used before; and
• infrequent use o f health agencies.
55
Suicide and Attempted Suicide
56
Psychiatric and Social Factors in Suicide
% unemployed (+)
% o f population under 15 years (—)
% wom en employed (+)
divorce rate (+)
homicide rate (+)
change in alcohol use (+)
change in church affiliation (+)
Durkheim has been criticized from various quarters. First, some have
pointed out that he needed a better and more operational definition o f
social integration. Later sociologists have come up with clearer defi
nitions o f terms, e.g. o f ‘status' (a category o f people with clearly defined
roles). Thus a person may have a variety o f statuses and roles: within
society (male, white), within the family (father, husband), within w ork
(teacher, counsellor). In this w ay it is possible to calculate the frequency
with which any person's combination o f statuses conforms to the
combinations most com m on in the society in which they live. In similar
vein, sociologists have much better definitions o f ‘social support' than
those proposed by Durkheim.9
Second, he based his conclusions on studies o f a W estern society,
and there is evidence that in other societies risk factors are different.
For example, although the association between divorce rates and suicide
holds for W estern nations (the correlation for the United States has
57
Suicide and Attempted Suicide
been found to be 0.78 for the 48 continental states), it does not hold for
Taiwan. The correlation between the divorce rate and suicide rate
among the sixteen counties o f Taiw an is an insignificant 0.05.
Third, Durkheim too readily separated melancholic from other
depressive suicides. He allowed that many normal persons w ho kill
themselves m ay also be depressed and dejected, but maintained that
they did not fall thereby into the category o f melancholic mental
illness. The difference, he suggested, is that in melancholy the person's
depression was unrelated to their external circumstances, whereas in
those he wished to call ‘norm al’ the state o f depression and the act o f
suicide had an objective cause. In this respect, Durkheim ’s analysis was
wrong. There is no clear-cut difference between different types o f
depression, based on some having been preceded by negative life events
and others coming ‘out o f the blue’ .
Whereas it used to be thought that some depressions were ‘reactive’ ,
caused by life circumstances, and others were ‘endogenous’, caused by
biological factors (‘endogenous’ means ‘originating from the inside’),
research has shown that negative events and circumstances precede all
types o f depression equally often, and all types o f depression have some
biologically driven features. Even in depressions which follow directly
after m ajor loss or disappointment, there is evidence to show that, as
the depression deepens, so certain neurochemical pathways in the brain
undergo a change, affecting a person’s eating and sleeping patterns,
energy levels and capacity to enjoy previously enjoyed hobbies, interests
and social contacts. Such changes sometimes self-correct in time, and
where they do not, antidepressant medication or types o f psychotherapy
that provide a structure within which the person can regain control
over their moods have been found to help.
Nevertheless, the central thrust o f Durkheim ’s argument remains
true: we need to look for factors other than mental illness to explain
w hy risks o f suicide differ between individuals and between societies.
The importance o f social support cannot be overestimated in moderat
ing the impact o f other stresses. Not only is it important at times o f
stress, but people who have social support tend to adhere to the
treatment suggested, and it is therefore no surprise that they respond
better. With social support, people are more likely to take the opportun
58
Psychiatric and Social Factors in Suicide
ities given to begin to solve the problems in their life, with or without
the help o f mental-health professionals.
H ow ever, understanding the social facts needs to be combined with
understanding the individual circumstances in causing particular suicidal
acts. If this were not so, suicidal behaviour would be much more
com m on than it is.
59
Suicide and Attempted Suicide
60
Psychiatric and Social Factors in Suicide
61
Suicide and Attempted Suicide
62
4
Attempted Suicide: Facts and
Figures
I had been collecting the things for months obsessionally, like Green Stamps,
from doctors on both sides o f the Atlantic . . . hoarding them in preparation for
the time I knew was coming. When it finally arrived, a box was waiting stuffed
with pills o f all colors . . . I gobbled the lot.
My wife got back at noon, took one look and called the ambulance. When
they got me to the hospital I was, the report says, ‘deeply unconscious, slightly
cyanosed, vomit in mouth, pulse rapid, poor volume’ .
I was still unconscious the next day and most o f the day after th at. . . During
the afternoon o f the third day, December 28, I came to . . . In a fog I saw my
wife smiling hesitandy, and in tears. It was all very vague. I slept . . .
At some point the police came, since in those days suicide was still a criminal
offense. They sat heavily but rather sympathetically by my bed and asked me
63
Suicide and Attempted Suicide
questions they clearly didn’t want me to answer. When I tried to explain, they
shushed me quietly. ‘It was an accident, wasn’t it, sir?’ Dimly, I agreed. They
went away.
Definitions
64
Attempted Suicide: Facts and Figures
clinicians and researchers have, over the past twenty years, adopted the
term ‘parasuicide’ or ‘deliberate self-harm’. Parasuicide was, for a time,
the most widely used term for all such self-harm (whatever the explicit
or implicit intention). Most clinicians, however, now use the terms
‘deliberate self-harm’ and ‘attempted suicide’ interchangeably (and I
shall follow that convention in this book, without making any assump
tions about whether the behaviour was intended to be fatal). In any
event, the definition is the same:
65
Suicide and Attempted Suicide
There are perhaps a further 30,000 (of all ages) who never reach hospital,
either telling no one about their self-harm, or seeing their family doctor,
who m ay not refer them to the hospital. Another slight fall at the end
o f the 1980s was followed by a rise at the beginning o f the 1990s. In the
United Kingdom this represents a rate o f 264 per 100,000 for men, and
368 per 100,000 for w om en.3 This ratio o f female:male o f 1.4:1 declined
from a ratio o f 2.1:1 in 1976, and 1.9:1 in 1984. During the 1990s the ratio
declined again, so that men almost caught up with w om en in attempted
suicide numbers. By the end o f the 1990s and the beginning o f the 2000s,
however, the ratio had begun to revert. Many more wom en than men
were again attempting suicide.
Until the 1990s, our understanding o f the problem o f attempted
suicide had been based on a pattern o f statistics that had not changed
for three decades. Overall numbers were rising, but the general charac
teristics o f the data remained relatively stable. W hereas suicide appeared
more a problem o f older men, attempted suicide was more a problem
o f younger wom en. T wo-thirds o f attempted suicide cases were wom en,
and two-thirds were under 35. Around 90 per cent o f such cases harmed
themselves by taking an overdose, with younger people more likely to
take analgesics such as aspirin or paracetamol, and older people more
likely to take sleeping pills or antidepressants. (This pattern o f age-
related overdose by different substances reflects the fact that all ages
take whatever is available to them. Younger people are less likely to
have sleeping pills and antidepressants prescribed, so take whatever else
comes to hand. If they do take psychotropic medication, it is often
because they have taken the medication prescribed for someone else in
the family.)
H ow far is this pattern o f data still true? T o answer this, w e need to
examine the statistics in more detail.
66
Attempted Suicide: Facts and Figures
the trends are remarkably similar both for total numbers, balance
between the sexes, socio-economic status data and methods used. It is
therefore useful to examine the Oxford data to chart the U K trends,
since the Oxford Centre for Suicide Research has data from the 1970s
to the present. The picture o f suicidal behaviour begins in the mid
1970s, at a point where the rates (increased fourfold in the ten years to
1973) began to decline (see Figure 4.1). The decline was more marked in
wom en than men, corresponding to the earlier greater increase in
wom en. Young people are at most risk o f attempted suicide. The data
for 1999 are shown in Figure 4.2 where it can be seen that the peak age
for w om en remains at 15-19 years, whereas the peak age for men lies
at 25-29 years.
The data for marital and employment status help fill out the picture.
Most people who attempt suicide are people who are single or divorced
(56 per cent in 1999). H owever, despite the population o f those who
attempt suicide being weighted towards those who are single or di
vorced, the majority are living with parents or with a partner. Only 35
per cent o f men and 26 per cent o f wom en who attempted suicide in
1999 in Oxford lived alone. This is consistent with research that shows
that most non-fatal suicidal behaviour occurs in the context o f problems
in relationships.
The data also reveal that the majority are from low er socio-economic
status backgrounds. The rate for social classes I and II is around 50 per
100,000 for wom en and 20 per 100,000 for men. This rate goes up over
eight times for wom en and over 12 times for men in social classes III—V.
W hat proportion will harm themselves again? Estimates over the
years and from various centres have varied between 10 and 25 per cent,
but settled down to around 15 per cent. This is reflected in the Oxford
data, where 15.5 per cent o f men and 12.8 per cent o f wom en repeated
the attempt within the first year. Some go on to make further repeat
attempts, so that in any sample o f cases coming to hospital, around 44
per cent o f both men and wom en have a previous episode o f deliberate
self-harm.
In terms o f methods used in self-harm, although 9.5 per cent cut
themselves, the vast majority use self-poisoning (86 per cent medicines,
and 3 per cent other chemical substances), a pattern unchanged over
67
0001
oo
0\
68
00 h-*
oo oo oo
oo
Suicide and Attempted Suicide
saposidg
oo
oo
o
CNo
o o
o
Figure 4.1 Episodes of attempted suicide referred to the John Radcliffe Hospital, Oxford, 1976-99.
Males
J9quinf\i
69
Attempted Suicide: Facts and Figures
<
3
H
u
O
G,
cbJD
bJO
Figure 4,2 Oxford City attempted suicides, 1999, by age and gender.
Suicide and Attempted Suicide
the years. H ow ever, the pattern o f substances used has changed (see
Figure 4.3). Tranquillizers and sedatives used to be the most common
method, but their use has declined from around 40 per cent o f all
overdoses to around 20 per cent, most likely because o f the marked
decline in prescribing these drugs over the period. Their place has been
taken by an increase in paracetamol (from around 14 to 42 per cent).
Often the tablets are either taken with alcohol (32 per cent o f men and
20 per cent o f wom en) or within a few hours o f consumption o f alcohol
(56 per cent o f men have taken alcohol in the previous six hours, as
have 37 per cent o f women).
Paracetamol
Other analgesics
(e.g. aspirin)
Tranquillizers
and sedatives
Antidepressants
1, , 1 1999
1 _ J 1976
• ';
•
Other
10 20 30 40 50
Percentage o f all overdoses
70
Attempted Suicide: Facts and Figures
Italy (Padua)
Italy (Emilia-Romagna)
—
___
Spain (Guipuzcoa)
Germ any (Wurzburg) B ll Female
.... ■
Austria (Innsbruck) Male
Netherlands (Leiden)
Sweden (Umla)
Switzerland (Berne)
France (Bordeaux)
Sweden (Stockholm)
IM M M H im ili
Norway (Sor-Trondelag)
Denmark (Odense)
Hungary (Szeged)
UK (Oxford)
Finland (Helsinki)
300 400
71
Suicide and Attempted Suicide
Across Europe for the period up to the 1990s, the number o f wom en
attempting suicide exceeded that o f men in most countries. Only in
72
Attempted Suicide: Facts and Figures
73
Suicide and Attempted Suicide
There is no one single predictor, but the most com m only found
factors that predict repetition o f attempted suicide are:
Those patients who scored 3 or less had a repetition rate averaging 4.9
per cent (5.1 per cent for females, 4.6 per cent for males). Those with a
score between 4 and 7 had a repetition rate o f 20.5 per cent (18 per cent
for females, 22.9 per cent for males). Those with a score o f 8+ had a
repetition rate o f 41.5 per cent (36.9 per cent for females, 46 per cent for
74
Attempted Suicide: Facts and Figures
males). So that almost 60 per cent o f those with a score o f 8+ did not
attempt suicide again. Isaac Sakinofsky has pointed out that factors that
predict suicidal behaviour in one sub-group o f the population may not
generalize to another. Therefore, we must be cautious in making
predictions o f w ho is most at risk.10
Suicidal intent
75
Suicide and Attempted Suicide
%
endorsement
I wanted to die 61
I wanted to make people sorry for the way they have treated
me 9
Note : Each person w as allow ed to endorse as m any item s as they felt fitted th eir case, so
th e n u m bers do n o t add up to 100.
76
Attempted Suicide: Facts and Figures
(b) Self-report
1. Did they believe what they did would kill them? People vary in how
much they know about the lethality o f suicidal acts, particularly over
doses. The amount o f drugs taken has been found to correlate with
suicide intent if a large enough sample is taken,13 although other studies
have not found this to be the case.14 Medical lethality is useful for
judging seriousness o f intent where it is known the person who has
taken the overdose is aware o f the relative lethality o f drugs. Otherwise,
77
Suicide and Attempted Suicide
the person's own report about what they thought the outcome would
be is the most important factor.
2. Do they say they wanted to die? Many studies show that over half o f
people who deliberate self-harm say they did not want to die. High
suicide intent is associated with a clear indication that death was the
intended outcome.
3. How premeditated was the actf Two-thirds o f patients have not thought
about it for more than an hour beforehand. The longer the idea o f
suicide had been in the mind, the greater the suicidal intent.
4. Is the patient glad (or sorry) to have recovered? Clinicians are always
concerned about the person who has harmed themselves, and after
wards says they are sorry to be still alive.
Research has found that high suicidality, defined in this w ay, predicts
future suicidal behaviour and future suicide.15 One study found that 21
per cent o f patients with high suicide intent later committed suicide.16
Another study, by Aaron Beck and colleagues,17 compared the Suicide
Intent Scale scores o f 194 patients who eventually completed suicide
against data from 231 other suicide attempters. They found that the
completers had higher scores on the 'Objective Circumstances’ items
o f the Suicide Intent Scale. Further, those attempters who had another
suicide attempt within a year o f discharge (n = 19) had a higher total
intent score for this last episode. Further studies from the same research
group18 have found that the ‘Precautions against Intervention’ factor
o f the Suicide Intent Scale predicted eventual suicide am ong suicide
attempters and that intent was higher just before completion o f suicide.
O f particular interest is how suicide intent relates to age, and whether
intent is related to the use o f alcohol, either in general or specifically
around the time o f the attempt. Results have shown that people who
harm themselves with high suicide intent are more likely to have had
previous episodes o f self-harm, to be single or divorced and to live
alone. However, they are no more likely to be depressed or personality
disordered, no more likely to have had psychiatric treatment in the past
and show no difference in the type o f drug used in overdose or in
whether alcohol was taken. By far the largest predictor is a chronic
78
Attempted Suicide: Facts and Figures
problem with alcohol abuse. Since chronic alcohol abuse itself puts a
person at high risk o f suicide, the joint influence o f abusing alcohol and
harming oneself with high suicide intent should be taken as a very
serious indicator that a person is vulnerable to suicide.
Alvarez revisited
79
Suicide and Attempted Suicide
80
5
The Causes of Attempted Suicide
\
81
Suicide and Attempted Suicide
Parenting
Although there is an association between loss o f parents (through death
or divorce) and such suicidal behaviour, this does not imply that all
cases have this in their background. Nevertheless, several studies suggest
that poor parenting occurs unusually frequently. A study by Robert
Goldney3 in Adelaide examined the type o f care that young (18-30)
female overdose patients reported having received as children. They
used the Parental Bonding Instrument,4 a questionnaire that measures
how much people perceive their parents cared for them versus being
indifferent or rejecting. It also assesses how much parents w ere overpro-
tective versus encouraging independence. Several research studies have
established the reliability and validity o f this instrument.
Results showed that suicidal patients reported their parents as having
been more rejecting and more overprotective. It seems that these young
wom en, through their experience o f deficient parenting, perceived
themselves as less deserving o f care. They had poor self-esteem and
placed a low er value on their own life. The curious combination o f a
more rejecting style and overprotection is significant. Overprotection
undermines the grow ing child's sense o f autonomy and sense o f being
in control o f events in its own life.
Sexual abuse
Several authors have suggested that a high proportion o f suicide
attempters have an even more acute disruption in early social relation
ships: they suffer sexual abuse. Only recently has this been studied
systematically. A 1993 study by M aijan van Egm ond and colleagues5
from the Department o f Clinical and Health Psychology, University o f
Leiden, Netherlands, examined the extent o f sexual abuse in a sample
o f 158 female suicide attempters aged 20 years or older. The degree o f
sexual abuse was established by asking two questions: 'H ave you ever
in your life been forced by anyone to have sexual intercourse?' and,
‘Has anyone ever forced you to perform or allow sexual contacts other
than sexual intercourse?' Fifty per cent o f subjects (79) reported having
been sexually abused at some time in the past. The sexually abused
wom en made their first suicide attempt earlier than the non-abused
wom en: when they w ere 27 years old, on average, compared to an
82
The Causes of Attempted Suicide
average age o f 36 years. The abused w om en also had almost double the
number o f previous suicide attempts.
The 79 w om en had often been abused many times: 57 by an unknown
man or boy, 30 by a friend or acquaintance, 27 by relatives other than a
stepfather or father and 12 by a father or stepfather. Four had been
abused by a professional care giver. Only 20 per cent o f these wom en
had been abused once. The vast majority had been abused repeatedly
by the same abuser or by multiple abusers. For 75 o f the 79, the first
suicide attempt took place after sexual abuse.
As other research has found, once a person has been abused they
often find themselves being victimized again by other men in later
childhood or adulthood. One wom an had been raped and physically
abused by six boys at the age o f 17 years. A year later she was raped and
abused by two unknown men. She made her first suicide attempt when
21 years, but found herself hardly able to talk about the rapes, except to
say that on the second occasion her body was mutilated by the men
using cigarettes. The suicide attempt was precipitated by an innocent
remark by another m em ber o f her family about an article o f their
clothing being destroyed by burning cigarettes. The memories came
rushing back and shortly afterwards she gathered all the medicines she
could find at home and swallowed them.
Particularly w orrying in van Egm ond’s study is the finding that
w om en who have been abused have a much greater probability o f
multiple suicide attempts later. They followed up the w om en for a
year, and found that significantly more sexually abused w om en (48 per
cent) had made further suicide attempts during this period than wom en
with no history o f sexual abuse (29 per cent). They found no difference
within the group o f sexually abused w om en between those w ho had
been the victims o f child sexual abuse (under 16 years) and those first
abused when 16 years or older. Similarly, and somewhat surprisingly,
there was no difference in the pattern o f suicidal behaviour between
those who had one sexual abuse experience and those who had suffered
multiple experiences. Neither did the history o f sexual abuse affect the
characteristics o f the suicide attempt: for example, the method used, or
the reasons to explain the suicide attempt, or even the suicidal intent.
This is an important finding. It undermines the argument that a
83
Suicide and Attempted Suicide
84
The Causes of Attempted Suicide
Employment status
A large study in Edinburgh, conducted by Stephen Platt,8 examined a
two-year cohort o f male attempted suicide cases, comparing the
employed (n = 158) with the unemployed (n = 199). The unemployed
men were less likely to be married, and less likely to live with their
family; they were more likely to be o f a low er social class, to have been
given a diagnosis o f abnormal personality, to misuse drugs, to be in
trouble with the police and to have a criminal record. This conclusion
has important implications for studies that have found a relationship
between unemployment and deliberate self-harm. Unemployment is
85
Suicide and Attempted Suicide
Substance abuse
Keith Hawton, in studying how many o f the cases o f deliberate self-
harm, which his team see in Oxford City, also abuse alcohol or drugs,9
found that 41 per cent o f men and 21 per cent o f w om en were abusing
alcohol. Sixteen per cent o f men and 6 per cent o f w om en were
habitually abusing drugs. In an earlier study o f predictors o f suicide in
15-24-year-olds, substance abuse emerged as a key predictor o f suicide
following an earlier attempted suicide.10 Habitual abuse o f alcohol and
drugs provides the person with a readily available means o f overdosing;
it also decreases the sense o f risk in doing so. Finally, the abuse can
affect judgem ent so that normal ability to solve problems (which may
already be suspect in this group) becomes even more impaired.
Precipitating factors
Precipitating factors are those events which occur in the few days prior
to the attempt. Disharm ony with 'key other' people in the person's life
is the most com m on event; disharmony with relatives, anxiety about
work/em ploym ent, financial difficulties and physical pain or illness are
other reasons. One study11 found that almost half their female patients
had had a quarrel with their spouse or boyfriend in the w eek (mostly
48 hours) prior to the attempts. Some clinicians say that, for those who
harm themselves repeatedly, different types o f events precede self-
harm o f differing levels o f lethality. This suggestion warrants further
research.
Special dates
Days in the calendar that are special for some are likely to be the
most difficult for others. W hen everyone else appears to be enjoying
themselves, those who are depressed and hopeless are at their most
vulnerable. W e have already seen how suicidal feelings are sensitive to
86
The Causes of Attempted Suicide
Public Events
87
Suicide and Attempted Suicide
conditions fraught with difficulties. Yet, when such people come for
help, they often blame themselves. It sometimes seems as though the
very act o f seeking help confirms to them that everything is their fault.
W hy else, they think, would they be needing help?
Someone w ho has undergone long-term stress m ay suffer from a
number o f psychological difficulties. N ot all who have suffered such
stress will experience these, and there m ay be some who have these
difficulties without their having been brought about by a difficult past.
But the association between such difficulties and suicidal behaviour is
well-established and m ay need very little added stress to precipitate
suicidal ideas and behaviour.14
Interpersonal problem-solving
88
The Causes of Attempted Suicide
89
Suicide and Attempted Suicide
self-harmed. One such study in T exas19 found that 40 per cent o f the
suicide attempters had a family m em ber w ho had made a suicide
attempt (as opposed to 8 per cent o f the control group). The deliberate
self-harm patients w ere also more likely to have been arrested at some
time, and more likely to have had a recent break-up with a girlfriend or
boyfriend (38 per cent v. 10 per cent for both comparisons).
W hat is the basis for poorer problem-solving by suicidal individuals?
A possible clue comes from research on problems such people have in
their memories o f events from their past (the main topic o f Chapter 10,
but relevant to mention here briefly). Patients who have recently taken
overdoses tend to remember their past in a summarized, over-general
way. For example, in response to a cue word such as ‘happy', an
overdose patient might say, ‘W hen I'm out with friends'; that is, their
m em ory response does not single out a particular event. In contrast,
non-suicidal people retrieve specific, datable events, such as ‘Last Friday
when I w ent out for a meal with friends.' Attempted suicide patients
prefer to recall a general description o f a class o f events. They stop
short o f retrieving a specific m em ory contained within that general
description, though after further prompting often recall such an event.
As w e will see in Chapter 10, getting ‘stuck' at the stage o f recalling a
general category o f events has implications for problem-solving.
90
The Causes of Attempted Suicide
Some years ago Marsha Linehan and her colleagues developed a Reasons
for Living Inventory.24 Compared to both the general population and
psychiatric controls, attempted suicide patients endorse few er important
reasons for living. Some o f the main items from her scale can be seen
91
Suicide and Attempted Suicide
in Table 5.1. The importance o f knowing about such reasons for living
is that individuals m ay be very motivated to commit or attempt suicide,
yet not do so because they have reasons for staying alive. Reasons for
dying m ay vary independently from reasons for living, and both need
to be taken into account. There is particular danger when someone
who has long been suicidal suddenly finds he or she has lost their
reasons for living (e.g. their children, partner, religious faith).
W hy do suicidal people have problems thinking about positive things
that might happen in the future, which might give them a reason for
living? An obvious answer is that poorer circumstances and reduced
opportunities mean they actually have less to look forward to. Poorer
upbringing, stressful events, marital and family disputes, poverty and
unemployment all feature in the lives o f many such people. A poorer
outlook for the future m ay be realistic in many cases.
But our research has found that hopelessness adds significantly to
the burden. N ot everyone w ho has had such stress in their lives is
suicidal. People differ widely in the extent to which their mind translates
specific stress events into the general feeling that nothing can be done,
a sense o f helplessness that goes much wider than the original situation.
It is the over-generalization from one situation (that m ay indeed have
been impossible to do anything about) to other situations (that might
be solvable) which is at the root o f hopelessness. If a person is hopeless
that anything can be done about the second situation, they either do
nothing or they give up too early. There is then a self-fulfilling prophecy
('I said nothing w ould help, and nothing has') which serves only to
increase the sense o f helplessness about any new situations.
The result is that such people m ay disengage from thinking about
the future. This is adaptive in some ways. It m ay reduce how upset
they feel right now, but at a cost. They are less likely to become aware
o f any future possibilities for happiness or to make plans which could
bring about positive events. So whereas negative events appear to play
a major role in the onset o f suicidal feelings and hopelessness, whether
people are able to anticipate positive events is important in determining
how quickly they can recover from hopelessness.
92
The Causes of Attempted Suicide
Table 5.1 Sample of items from Marsha Linehan's Reasons for Living
Inventory
Responsibility to family
It would hurt my family too much and I would not want them to
suffer.
I have a responsibility and commitment to my family.
Child-related concerns
It would not be fair to leave the children for others to take care of.
I want to watch my children as they grow.
Fear of suicide
I am afraid o f the actual ‘act’ o f killing m yself (the pain, blood,
violence).
I am afraid o f the unknown.
Moral objections
My religious beliefs forbid it.
I am afraid o f going to hell.
93
Suicide and Attempted Suicide
Emotional experience
94
The Causes of Attempted Suicide
95
Suicide and Attempted Suicide
One adolescent girl said it felt as if her head would explode unless she
did something to stop it. She knew that cutting would stop it and
put her back in touch with herself and her body. W hat explains the
catastrophic escalation o f emotion that often precedes self-cutting? T o
understand this w e need to understand that negative emotion often
results from breaking the rules one has set for oneself.
Consider the situation in which a person has been punished in the
past for the expression o f emotion. In relatively mild cases, this occurs
every time a parent tells a child to ‘pull yourself together'. The child
m ay learn the lesson, 'It is shameful to display your em otions/ Or, in
the most extreme case, a similar thing m ay happen in sexual abuse by
an adult, where a child has been threatened with punishment if he or
she becomes upset. In both the mild and the severe case, there is conflict
between the expression o f emotion and the suppression o f emotion to
avoid future punishment.
In later life, when the person feels emotional, the very fact o f feeling
emotion is breaking one o f their rules. But breaking the rules causes
more emotion, which then leads to attempts to suppress it. The result
is a rapid escalation o f negative emotion arising from such a feedback
loop: the expression o f emotion itself violates a goal (‘do not feel or
show emotion'), but the consequence o f violating any goal is increased
emotion. This leads to patients feeling upset but telling themselves they
should not feel upset (one o f the ‘invalidation strategies' that maintains
the very emotion it is designed to abolish, identified by Marsha Linehan
in her treatment for people who suffer from ‘borderline’ symptoms).
Linehan has also developed a number o f specific strategies for dealing,
in a non-harmful w ay, with the escalation o f emotion. As part o f her
therapy, she m ay advise a patient to hold an ice cube in each hand until
it melts. The pain o f the cold ice is often sufficient substitute for
self-cutting, but without the physical damage.
96
The Causes of Attempted Suicide
Concluding remarks
97
6
Rational Suicide, Euthanasia
and Martyrdom
In the last few days o f 1994, the State o f Oregon was blocked in its bid
to change the law to allow assisted suicide. From 8 Decem ber 1994, the
terminally ill w ere to be allowed to ask their doctors to prescribe a
lethal dose o f medication. The m ove, which followed a state referendum
held just over a year before, was always likely to cause controversy. Up
to the last moment, it looked as if Oregon's decision in favour o f assisted
suicide would stand. H ow ever, reports said that many doctors were
unprepared for the changes.
Agreeing with the m ove towards such a change, a 47-year-old man
suffering from Aids said, 'W hose life is it anyway? It should be between
me, m y God and m y doctor/ Elsewhere, a wom an w ho tw o years
previously had been told she would not live, argued against the measure.
She was now well again, but said that at the time she was so depressed
she would have asked a doctor to end her life.
The Catholic Church in the state vigorously opposed the change in
the law, and continue to campaign against it. They argued that it would
put the w eak and the vulnerable at risk. At least six other states
were considering similar legislation and were looking carefully at what
happened in Oregon.
In Britain it remains illegal for anyone to aid someone to take their
own life under any circumstances, but the courts are sometimes lenient.
In December 1994, the courts decided not to prosecute a man who had
administered a lethal dose o f morphine to his terminally ill wife. In this
case the Voluntary Euthanasia Society (Exit) were careful to distance
themselves from supporting the man. They pointed out that there was
98
Rational Suicide, Euthanasia and Martyrdom
no evidence that the wife had agreed to her life being ended in this
way.
Developments in this area are occurring at some speed. Euthanasia
was legalized in the Northern Territories o f Australia in 1996. Assisted
suicide was legalized in N ew York in April 1996 when two federal
appeals courts lifted the ban on doctors helping their patients to die.
Esther Fein o f The New York Times posed a stark question: 'W ill the
right to die becom e the duty to die?" Let us look more closely at the
arguments for and against.
The w ord euthanasia (from the Greek words eu and thanatos) literally
means 'easy* or 'gentle death’ . It was declared unethical by the W orld
Medical Association in 1950, and to understand w hy w e have to see
its background. T w o forms o f euthanasia are normally distinguished.
The first is compulsory euthanasia, or 'm ercy killing’ , done without the
consent o f the person concerned, and mostly applied to grossly
deformed children. For example, a Dutch gynaecologist in 1994 admitted
killing a three-day-old girl at her parents’ request. Her brain was only
partly developed, she had spina bifida and partial paralysis, her limbs
were malformed, and the doctor judged she faced a life o f constant
pain.
The usually cited problem with this is that it constitutes a 'slippery
slope’ and violates the fundamental right to life, this being backed
up by religious people’s belief in life as God-given. H owever,
a distinction is often made between an analgesic and lethal dosage o f
pain relief. An analgesic dose m ay have the unintended effect o f
shortening life, but a lethal dose has the direct intention o f shortening
life (the ethical doctrine o f 'double effect’ allows the first but not the
second).
The second form o f euthanasia is voluntary euthanasia. Here, indi
viduals in sound mind ask that their life should be ended in the event
o f their becom ing victims o f irreversible illness. Advocates (e.g. the
H em lock Society in the United States or members o f Exit) maintain
that, with safeguards, the law should permit it and people should be
supplied with the means to take their own life or a doctor should
be authorized to end their life provided the request is made before
witnesses.
99
Suicide and Attempted Suicide
The difficulty here is the assumption that w e have the right to decide
over our own life or death absolutely. The argument from these groups
suggests that w e ow n5 our own lives. This argument has a long history.
Historical background
100
Rational Suicide, Euthanasia and Martyrdom
101
Suicide and Attempted Suicide
In our own day, there are many people who, standing in the tradition
o f David Hume, take a more liberal view than the original Hippocratic
oath envisaged. Jack Kevorkian is such a person. He began to assist
terminally ill people with suicide in the United States in 1990. The
court’s reaction at first was to prosecute for murder, but the ju ry refused
to return such a verdict. The Michigan state senate then passed a
temporary Bill making it illegal to help anyone commit suicide. Sub
102
Rational Suicide, Euthanasia and Martyrdom
The Netherlands is one country where doctors are given greater free
dom to decide. In 1995, television audiences throughout the world were
able to watch a Dutch doctor featured in a television documentary,
Death on Request. Dr Wilfred van Oyjen carries out this ‘good death’
103
Suicide and Attempted Suicide
three or four times a year. The film featured a man who had developed
Lou Gehrig's disease, which brings about incurable degenerative wast
ing o f the muscles. His feet and legs became paralysed first, then his
right shoulder and arm, then his face. He and his family realized that,
unless some decisive action was taken, he faced death by suffocation,
as the weakening muscles o f his chest finally collapsed.
His requests for euthanasia were repeated as the guidelines in H ol
land prescribe (a voluntary, well-considered and lasting request to die).
Dr van Oyjen is one o f those w ho feels it w ould be letting his patient
down to refuse his request. The guidelines also prescribe that there
must be no other solution acceptable to the patient, that the time and
manner o f the death must not cause unnecessary suffering to others
(such as next o f kin) and the doctor must prescribe and administer the
right drugs.
This contrasts with the situation in the United Kingdom, where it is
accepted that doctors often use the 'doctrine o f double effect' to deal
with situations o f unbearable suffering. This permits doctors to use
drugs in sufficient quantities to relieve suffering, even if that hastens
death, so long as they do not intend to kill. According to the Voluntary
Euthanasia Society (Exit), one survey o f British doctors showed that 50
per cent had been asked by a patient for help to die. O f these, one-third
o f the doctors said they had complied.
This is not ethical, according to D r van Oyjen. Such surreptitious
euthanasia can take several weeks, and is dishonest. It also means that
the patient can die in delirium brought about by huge doses o f m or
phine. 'I am giving people the possibility to make choices. W hat kind
o f quality o f life, and death, do they want? Death is not always awful.
With a good doctor, death can be faithful, like a good friend.' Perhaps
most importantly, the doctor can choose the right moment, when he
judges that the person and his family are most at peace with the ending
o f life. Wilfred van Oyjen will take hours, days if necessary, waiting for
the right moment after the decision has been taken. He feels it essential
to get close to the fam ily if a 'good death' is to be brought about.
W hen the arguments are expressed in this way. they seem very
compelling. Yet many have continued to maintain that assisted suicide
is either intrinsically w rong or socially undesirable.
104
Rational Suicide, Euthanasia and Martyrdom
Consent
One argument against euthanasia is that a terminally ill patient may
not be in a state, at the point near death, to be able to withdraw consent.
Suppose someone was to agree, before witnesses, family and doctor
that, in the event o f breakdown o f bodily function, active steps should
be taken to end their life. Once they have deteriorated, how could they
ever withdraw that consent should they change their mind? The right
(at any time) to withdraw consent from or change one's mind about a
medical procedure is seen by many to be paramount. Yet, with the very
frail, this right m ay be denied because o f their frailty.
On the other hand, this raises the question o f the prolongation o f
life, and with it a number o f imprecise issues. Should we withhold or
continue efforts to prolong life, for example by using m odem life-
support equipment? Under these circumstances, ethical theories usually
refer to the ordinary/extraordinary principles. That is, there is an obliga
tion to use ordinary means but no strict obligation to use extraordinary
means to prolong life. Extraordinary means refer to those means which
do not offer a reasonable hope o f cure or remission or those which
require excessive hardship to obtain. So most ethical theories assume
there are limits to preserving life, but that there is a presumption o f a
duty to preserve life in most circumstances. Ethical principles need to
be brought to bear.
The first principle cited is usually that o f autonom y and individual
conscience. The patient's attitude is primary, but if the patient cannot
be consulted, there is a secondary role for others on the grounds o f
principles o f beneficence and social justice. Proxy decisions are therefore
possible, indeed necessary in the case o f life-support machines. In these
cases there is a rank order o f people to be consulted, the family, the
physicians, a designated committee and the courts (in that order). These
make the decisions based on ‘substitutive judgem ent’ standards, which
involve asking what the patients themselves would have decided, based
if at all possible on the patients’ own expressed wishes.
105
Suicide and Attempted Suicide
106
Rational Suicide, Euthanasia and Martyrdom
No man is an island
Another argument against ‘rational suicide', assisted or not, is the ‘no
man is an island' argument (it is ironic that the quote comes from a
meditation by John Donne, who believed that suicide was justifiable
and argued so strongly in Biathanatos).
One version o f this argument is that by ending our life w e deny our
family and society the opportunity to fulfil their duty o f care to us.
Another version is m ore simple: those w ho would justify committing
suicide on the grounds that their life is their own property are failing to
take into account that each o f our lives is shared by parents, brothers
and sisters, partners, children, neighbours and friends, colleagues at
w ork and so on. Not to take account o f their reaction to our death, or
worse, to make assumptions (e.g. that they would be relieved o f a
burden) without checking, is hardly to take a ‘rational' decision.
In what w ay does suicide affect others? Research shows how suicide
affects family, friends and community. It can be the worst o f all deaths
in its impact on survivors. It causes grief they m ay never resolve, guilt
in a w ay no other death does, even raising the risk o f suicide in others.
A truly rational approach to self-killing or assisted suicide must take
account o f all these things. If a rational person says, ‘But no one would
mind if I ended m y life\ then it will be inappropriate to call it rational
107
Suicide and Attempted Suicide
108
Rational Suicide, Euthanasia and Martyrdom
109
Suicide and Attempted Suicide
natural and therefore goes untreated. Yet deep sadness is natural, while
depression is more than sadness. It brings with it a sense o f guilt, o f
being a burden, o f worthlessness and a feeling o f failure. Further, the
person often believes that no one will want to see them in their
condition, so there is no point in contacting anybody. They m ay feel
guilty about things they have done, or about situations or relationships
that remain unresolved, and the depression will ensure ‘it is not worth'
trying to resolve them.
Is there a role for therapy here? The evidence suggests there is.
People can, in fact, overcom e their depression even within the context
o f a terminal illness, evidence the hospice m ovem ent has been citing for
some years.8 Do people seeking assisted suicide need a psychotherapist?
Proponents o f rational suicide discourage people from seeking help on
the grounds they are not mentally ill. This is probably right, but you
don't have to be crazy to see a therapist any more than you have to be
stupid to go to school.9
All wishes to die represent in some degree a problem perceived as
insoluble except through death. In that situation one needs to respect a
person's autonomy. But personal autonom y should not be confused
with isolation and loss o f social cohesion. H ow can we judge whether
or not someone is in control o f their own actions? Depression leaves
little room for argument. W e know it affects chemical pathways in the
brain and results in lack o f energy, lack o f pleasure and lack o f interest
in social activities and social contacts. It also results in recurrent thoughts
o f death or suicide. Com pounding this is the sense o f hopelessness that
occurs in many depressions: a foreshortening o f the future and an
inability to see anything to look forward to. Depression therefore leaves
little choice. It pushes the person, impels them to certain actions. It is
not open to reasonable argument.
Proponents o f rational suicide say that people have chosen this
option and are not depressed. In that case they must expect to be able
to discuss rationally the pros and cons o f their actions. They must also
expect that their reasons will be questioned. The major presupposition,
that each o f us has an overall right to decide on our ow n life, can be
questioned, and only where the full social and family context is taken
into account can it ever be justified deliberately to shorten a life.
110
Rational Suicide, Euthanasia and Martyrdom
Martyrdom
The definition o f martyrdom is closely tied to the group that shares the
belief o f the one w ho dies. A suicide-bombing by an Islamic fundamen
talist m ay be called martyrdom by his or her own group, but is to the
non-Muslim m erely terrorism. Each group, religious or political, has its
martyrs. They differ in a number o f respects from the types o f suicide
so far considered.
Most people who commit suicide are depressed when they do so;
they see death as the end to their suffering. One o f tw o feelings usually
predominates in the mind o f the person w ho is suicidal in this depressive
sense, both stemming from hopelessness. The first is that they have
been abandoned by everyone; the second that they are a burden to
everyone, especially to those they love. Contrast this with the martyr.
They see hope, and believe in a cause. Although there are m any different
contexts for martyrdom, all martyrs believe that by their death they are
bringing about some combination o f the following gains:
111
Suicide and Attempted Suicide
The kamikaze
Learning about the mind o f the m artyr seems impossible since they are
not here to study. H ow ever, there are some 'wartim e martyrs' who,
though they volunteered to die, escaped death: Japanese kamikaze
112
Rational Suicide, Euthanasia and Martyrdom
(literally ‘divine tempest') pilots who survived when their planes were
shot down on their w ay to a target in the Pacific and were awaiting
recall to another suicide mission when the w ar ended.
By the last year o f the war, the US N avy thought itself invincible in
the Pacific, but the kamikaze shook its confidence. So keen were the
young Japanese pilots to participate in this ‘spirit o f the Samurai' that
some signed their papers in their own blood to try and increase chances
o f selection. By the end o f 1944, over 500 kamikaze missions had been
flown against the US Fleet, but Japan was still losing. A new elite,
T h e Thunder Gods’ , was formed to pilot aircraft equipped with large
missiles. These turned out to be too cumbersome and the pilots were
shot down before they could reach their targets. The Thunder Gods
were therefore asked to undertake a different sort o f mission: to take
ordinary planes and crash them into the fleet.
On 1 April 1945 the United States invaded Okinawa, and on the 6th,
after an ominous silence, the US Fleet was attacked by up to 350 planes
at a time. By the end o f June, after about 2,000 kamikaze missions, 36
Allied ships had been sunk, ten times as many damaged; 500 Allied
personnel were dead and 4,800 wounded.
What was the motivation o f the kamikaze pilots? By the accounts o f
those who remain, and their officers who assigned them, they had an
attitude to death characteristic o f many martyrs. First, they believed
that obedience gave meaning to self-sacrifice. They had come to the
point when to give up one’s life for one’s country was the highest
honour. Second, they believed they were not going to die at all. Their
souls would go to Yasakuni, a special place for those who die fighting
for their country; a special place also reserved for the Emperor. As one
said, ‘Even if you did die, you felt you would still be alive to describe
it.’ The widows o f these men paid their respects and went to honour
their ‘god-like’ husbands.
But it wasn’t always quite so straightforward for the pilots. People
who interviewed them subsequently found that a different, more
sombre mood sometimes emerged. Some appear to have brooded
about their death, complaining that the waiting went on for ever.
Others said it felt as if they were being sentenced to death. Still others
questioned whether they had been in their right mind when they
113
Suicide and Attempted Suicide
volunteered. They rejoiced for the camera, but harboured doubts off
camera. They dealt with such feelings by reattributing any emotions to
things other than the fear o f dying. One pilot said that at the time he’d
find him self crying in the night, but not because he was frightened. He
couldn’t say he really ever knew w hy he was crying. Perhaps it was
because most kamikaze pilots never reached the age o f 20 years.
Concluding remarks
114
7
Psychodynamics, Biology
and Genetics
Psychoanalytic perspectives
115
Suicide and Attempted Suicide
Since outwardly directed reproaches towards the lost object are not
possible, they turn into self-reproaches and the wish to harm oneself.
W hen the rage is sufficiently intense, it will lead to strong urges towards
self-destruction.
If the love for the object - a love which cannot be given up though the object
itself is given up - takes refuge in narcissistic identification, then the hate comes
into operation on this substitutive object, abusing it, debasing it, making it
suffer and deriving sadistic satisfaction from its suffering.
It is this sadism alone that solves the riddle o f the tendency to suicide which
makes melancholia so interesting - and so dangerous . . . No neurotic harbours
thoughts o f suicide which he has not turned back upon himself from murderous
impulses against others . . . The ego can kill itself. . . if it is able to direct against
itself the hostility which related to an object and which represents the ego’s
original reaction to objects in the external world.2
116
Psychodynamics, Biology and Genetics
We find that the excessively strong super-ego which has obtained a hold upon
consciousness rages against the ego with merciless violence, as if it had taken
possession o f the whole o f the sadism available in the person concerned.
Following our view o f sadism, we should say that the destructive component
had entrenched itself in the super-ego and turned against the ego. What is now
holding sway in the super-ego is, as it were, a pure culture o f the death instinct,
and in fact it often enough succeeds in driving the ego into death.3
117
Suicide and Attempted Suicide
feel they have retained control over. Some believe they will still, in
some sense, be present after their death to ‘see what happens next’,
and, in a spirit o f revenge that is sometimes present as part o f suicidal
thoughts, able to experience pleasure in the distress their death might
cause others. The inconsistency between death as escape into oblivion
on the one hand and a continued existence in which the pain o f others
can be experienced on the other is not apparent to the suicidal person
in their confusion.4
Object-relations approach
Freud had suggested that suicide was an attack on the love-object, now
lost, that had been internalized. Melanie Klein extended these ideas,
suggesting that the person was motivated by the wish to preserve the
good aspects o f the internalized object (now a valued part o f the self ).5
The inwardly directed attack was therefore aimed mainly at the bad
part o f the object. Since aggression mainly arose from the death instinct,
there was a danger that such destructive forces might destroy a good
object. Guntrip developed these ideas further, distinguishing between
depressive suicide and schizoid suicide.6 Depressive suicide comes about
as the result o f hatred redirected towards the self from the hated and
loved object (as Freud had argued). By contrast, schizoid suicide comes
about when the person finds themselves facing the loss o f self, and
begins to fantasize about death as a path to rebirth.
Early in a child's development, it is in a symbiotic relationship with
its mother and the child’s idea o f self and o f mother are not yet
differentiated. Later there will come inevitable separation and individu
ation, where self and object are differentiated. W hat happens if this
transition is not completed satisfactorily? One outcome might be that
the person will tend to become involved in relationships where the
other party is not treated as a separate and unique individual, but as
part o f the self. This choice o f friend or partner has been called ‘symbiotic
object-choice' corresponding to the ‘narcissistic object-choice’ in Freud’s
writings.7 Suicide, according to this analysis, involves not only ridding
118
Psychodynamics, Biology and Genetics
self o f bad internal objects, but the fusion with the earlier 'symbiotic
mother' o f pre-individuation infancy.
The hypotheses about suicidal behaviour that invokes fantasies about
return to infantile dependence on the mother have been extensively
developed by attachment theorists. According to Bow lby,8 infants
respond to even brief separations with distress, and if such separations
and returns are not predictable or controllable, there develops a style
o f anxious and insecure attachments. These styles, resulting from the
breakdown o f the reciprocal interaction between child and primary
care-giver, m ay persist, and can even affect the w ay the child, when
grown up, relates to his or her own children.
Secure attachments in infancy are the primary means by which the
child learns to regulate his or her own emotions. In the absence o f
mature self-regulatory structures, such children come to depend too
much on others, or ideas about others that have been internalized, for
comfort. Such a person will become over-sensitive to abandonment;
vulnerable to crises o f aloneness which give w ay to self-contempt and
murderous rage. Suicidal behaviour in this context can be seen as an
interpersonal act, an (albeit maladaptive) means o f procuring attach
ment, signalling distress to others in the environment and punishing
them for their actual or perceived rejection o f the person.
John Maltsberger9 summarizes the sorts o f ideas about death that
might be apparent in the suicidal person:
119
Suicide and Attempted Suicide
An underlying theme is the idea that one’s body is a prison from which
suicide promises an escape - particularly true o f those who suffer
physical pain. Maltsberger advises therapists to assess the extent to
which their suicidal clients have lost the capacity to tell whether or not
their own body is a part o f themselves. 'Does he feel at home in it and
take it for granted as an integrated self aspect, or does he experience it
as an alien cage in which he is confined, a cage belonging to, or even
identified with, someone else?' This aspect o f the psychodynamics o f
suicide takes us a long w ay from simply seeing such death as an
expression o f anger turned towards the self. Importantly, it raises the
theme o f escape. The notion o f escape is important because it builds
bridges to other approaches to suicide. M eanwhile inescapable stress
has biological consequences, which might be triggered by stress that is
believed to be inescapable (which introduces psychological mechanisms).
Biological approaches
From the brain, and from the brain only, arise our pleasures, joys, laughters
and jests, as well as our sorrows, pain, grief and tears . . . It is the brain which
makes us mad or delirious; inspires us with dread and fear, whether by night
or day; brings sleeplessness, mistakes, anxieties, absent-mindedness, acts that
are contrary to our normal habits. These things that we suffer from all come
from the brain, including madness.
120
Psychodynamics, Biology and Genetics
121
Suicide and Attempted Suicide
122
Psychodynamics, Biology and Genetics
Genetic influences
123
Suicide and Attempted Suicide
124
Psychodynamics, Biology and Genetics
they all carry an increased risk o f self-harm behaviour. But the fact that
suicide does not discriminate between the diagnoses throws doubt on
this possibility. I have suggested that it was not the diagnosis per se but an
individual's perception o f their symptoms as aversive and uncontrollable
that was important.
The next generation o f research on serotonin is likely to have to
address several pressing issues. W hat is the effect o f chronic stress on
the serotonergic system and on impulsivity? W hat is the relationship
between alcohol and impulsivity? Alcohol disinhibits a person and
makes them act in a more dangerous way, but also lowers serotonin. It
remains unclear whether this disinhibition comes from the serotonergic
effect rather than the disinhibiting effect o f alcohol itself. It is possible
that alcohol’s effect on impulsiveness is simply short-term, whereas that
o f serotonin is longer-term.
Progress in this area is unlikely to be fast. First, changes in serotonin
function m ay only be a marker for another causal factor, and even if it
were decisively established that it was linked to certain forms o f suicide,
both suicide and altered serotonin function may be caused by this third
variable. Second, decreased serotonin function may represent the brain's
attempt to compensate in one pathway for an increase in serotonin
activity elsewhere, or to compensate for a problem in another neuro
transmitter system. Third, since all the data on serotonin function come
from correlations between known suicidal behaviour and biological
function, we do not yet know whether there are any changes in
serotonin function that precede such behaviour. Therefore no infer
ences can yet be drawn about the predictive value o f knowing about
serotonin function in an individual prior to suicidal behaviour.
A final problem is that biological studies have to date been done in
isolation. They have been concerned to establish links between brain
systems and behavioural tendencies such as violent or impulsive
behaviour, without asking what the psychological (e.g. information-
processing) and social (e.g. acute and chronic stress) factors involved in
such behaviour might be, or what the neuropsychological mediators o f
such behaviours are.
Following a review o f this material, Seym our Kety concludes:
125
Suicide and Attempted Suicide
W e cannot dismiss the possibility that the genetic factor in suicide is an inability
to control impulsive behavior, while depression and other mental illness, as
well as overwhelming environmental stress, serve as potentiating mechanisms
that foster or trigger the impulsive behavior, directing it toward a suicidal
outcome. In any case, suicide illustrates better than any o f the mental illnesses
. . . the very crucial and important interactions between genetic and environ
mental influences.19
Concluding remarks
126
8
The Effect of the Media
My teenage son and I were clearing up the back garden. I was clearing
parts o f a tree which had been trimmed some time before, and he was
sawing them into logs. He remarked that it was exactly a year ago since
he had last been sawing logs. W hen I expressed surprise at his accurate
m em ory, he said he remembered the date clearly because it was the
anniversary o f Kurt Cobain’s death by suicide. Kurt Cobain was the
lead singer o f the band Nirvana. He had harmed him self in the past and
finally shot himself, having left a long suicide note. When such a
high-profile suicide occurs, people want to know what effects the death
has on all the fans.
The debate about whether accounts o f suicides (whether fictional
stories or actual suicide) cause people to imitate them has a long history.
In the eighteenth century Goethe was accused o f encouraging suicidal
behaviour by the publication o f his novel The Sorrows o f Young Werther
(1774), in which a young man ends his life by shooting him self in the head.
The book had a tremendous influence on attitudes to romantic
suicide and coroners all over Europe began to believe that the W erther
effect was influential in bringing about suicides by lovelorn young
people. The book was accused o f sentimentalizing self-destruction, but
whether or not it produced imitative suicides, it certainly produced
imitative poetry and literature. The sentimentalization o f death which
followed such literature inspired pity for those who committed suicide
and helped to erode the revulsion suicide had form erly inspired. The
result was that many became concerned that suicide would spread by
imitation.
127
Suicide and Attempted Suicide
128
The Effect of the Media
Fictional portrayals
129
Suicide and Attempted Suicide
130
The Effect of the Media
not only compared the number o f 'imitative' cases with the numbers
o f people attending in control years, but looked for any trends which
might have contributed to an apparently large number in the target
period. For both hospitals, overdose attenders in the 14 days after the
broadcast were higher than those in the corresponding period for
the tw o control years, but the rise in numbers began well before the
programme was shown, in December in one hospital, January in the
other. Analysis o f daily data showed no evidence o f unusual short-term
changes within the fortnight after the programme. W e concluded there
had been no copycat effect.
Do these findings mean that no imitation ever takes place? Clearly
that cannot be concluded on the basis o f this study. There remains
enough prima facie evidence to suggest caution when editors decide
what to print or to show. Other studies, which have provided a more
rigorous 'natural experiment’, indicate that fiction can be powerful.
Arnold Schmidtke showed how a fictional portrayal o f a young man’s
suicide on a railway line could have an imitation effect on view ers.9 He
studied the number o f railway suicides in the 70 days after the fictional
episode was broadcast and found an increase. The suggestion that this
was indeed imitation was reinforced by the finding that the suicides
most often occurred in people o f the same age and sex as the fictional
character. Furthermore, the numbers o f suicides correlated with the
audience figures. The broadcast was repeated by the television author
ities and Schmidtke found, once again, an increase in suicide. In total,
his report estimated that an extra 60 suicide deaths occurred as a result
o f this fictional episode.
Similarly, in 1999, Keith H awton and colleagues found a marked
increase in self-poisoning cases following a television programme show
ing a paracetamol overdose. Forty-nine accident departments around
the United Kingdom were put on alert, and monitored their attendances
before and after a fictional account o f an overdose on the B B C ’s Casualty
drama. Overdoses increased by 17 per cent in the w eek following the
broadcast. The use o f paracetamol in the overdoses doubled in those
who had actually seen the program m e.10
131
Suicide and Attempted Suicide
132
The Effect of the Media
Mechanisms of imitation
133
Suicide and Attempted Suicide
Concluding remarks
134
The Effect of the Media
135
9
The Cry of Pain
At the time, several commentators suggested that her case should signal
a danger that the education system had become too dominated by a
league-table culture that emphasized academic performance as the sole
136
The Cry of Pain
137
Suicide and Attempted Suicide
If the case comes to court, the verdict will often reflect the judgem ent
that the person is suffering a psychological problem that needs
psychological intervention rather than punishment, but many commit
suicide before the case comes to court.
Interpersonal disputes can involve combinations o f the same em o
tions. A wom an in her mid-forties took a massive overdose soon after she
discovered that the man she planned to m arry in a few weeks (having lived
with him for several years) had unilaterally cancelled the arrangements.
Without her knowledge, he had sent notes to all the guests to say the
wedding was cancelled. He came to visit her in hospital, but only to give
her ring back, and tell her to collect her stuff, which he had left in suitcases
on the front doorstep. Her overdose was not lethal, and she lived. But her
crisis threatened to last for some time, and while it did so, she remained
at serious risk o f further suicidal behaviour.
Biological, psychodynamic and social aspects m ay be combined in
explaining such behaviour. Individuals are sensitive to signals from their
social environment o f threats both to their rank within the group and
their acceptance as part o f the group. Events signalling threat to rank or
group similarity are seen, by the person, as signs that they are a ‘loser’.
Like the wom an whose wedding was cancelled, such people have lost a
battle. Such ‘loser’ status we know as reduced self-esteem, and it triggers
evolutionarily ‘old’ biological patterns in the brain. This affects m ood and
produces further hypersensitivity to social information, resulting in a
vicious circle that appears to confirm the individual as a ‘loser’ . In the early
stages a ‘reactance’ pattern is shown, in which the individual ‘protests'
against the threatened loss. At this stage, ‘low intent’ suicidal behaviour
m ay occur. In the later stages o f response, despair and apathy m ay be
seen, and serious attempts at suicide may occur. Occasionally, however,
a person may experience a catastrophic increase o f suicidal impulse in
response to loss or failure, and an unexpected suicide occurs.
Knowledge o f brain function is not all we need to explain such
behaviour. W e need rather to understand what environmental and
psychological ‘signals’ bring about biological changes in brain function
(see pages 120-6), and, in turn, what effects such biological changes will
have on the w ay we think and act. For example, psychosocial stress is
a critical factor in bringing about biological changes. Long-term stress
138
The Cry of Pain
These can become distorted in catastrophic w ays that cause the indi
vidual to believe he or she is trapped, and to begin to think o f ways o f
escaping. Suicidal behaviour is best seen as a cry o f pain - a response
elicited by this situation o f entrapment - and only secondarily as an
attempt to communicate or change people or things in the environment.
Conservation-withdrawal
The Australian psychiatrist Robert Goldney was one o f the first to take
an ethological approach to suicidal behaviour, employing Engel’s ideas2
about ‘conservation-withdrawal’.3 This refers to a biological system
which responds to stress by reducing activity, raising barriers against
stimulation and conserving energy: a sort o f hibernation. This pattern
m ay follow a prolonged period o f heightened arousal where the
139
Suicide and Attempted Suicide
Evolutionary approach
140
The Cry of Pain
An important aspect is rank and status within a social group that shares
the same territory. High-ranking individuals explore more, have more
erect posture and are less timid than low-ranking individuals across a
number o f species.
Changes in rank can bring about changes in biological systems (e.g.
in levels o f testosterone), but more importantly, for understanding
141
Suicide and Attempted Suicide
142
The Cry of Pain
o f people during the 1980s and 1990s. Players are allocated characteristics
in various dimensions: size, strength, skill, previous success, weapons,
allies, etc. The outcome o f encounters between competitors is deter
mined by their ability to use and to conserve these ‘attributes’.
Social biologists refer collectively to these attributes, when observed
in animals, as Resource Holding Potential (RH P). Ritual encounters
between two members o f the same species result in comparison o f
relative R H P , and the loser will back off. As the hierarchy becomes
established, the amount o f fighting reduces. Gilbert suggests that the
human equivalent o f R H P is self-esteem. Self-esteem is an estimate o f
one’s ability to secure important goals such as a desired job, or partner.
It is damaged by unemployment, by failure in love and by aggression
from others, as in school by bullies or in the home by family members
who use psychological or physical violence.
Paul Gilbert suggests that dominance hierarchies have evolved
through stages. First, they were territorial, where fighting was designed
to create space for territorial ownerships. Subsequently, territories gradu
ally dissolved and success in the dominance hierarchy became associated
with other forms o f social success; it was a dispenser o f useful resources.
Dominance, how ever established, creates inhibition in subordinates.
Since that time there have been large changes in the importance o f
affiliative behaviour, particularly alliance building (see next section),
which requires approach rather than inhibition. This results in the fact
that, today in humans, rank is often (though not always) determined
by things such as perceived popularity, beauty, talent, etc. Gilbert main
tains that when individuals lose rank, then the defensive systems o f
escape, etc., become activated. The means o f acquiring rank and status
m ay have changed, but the feeling when one loses has not. Submission
in humans continues to involve giving up claims on resources.
143
Suicide and Attempted Suicide
144
The Cry of Pain
run deep. It is the comparative effect, not the absolute effect, that
determines the amount o f stress or depression. During the Thatcher/
Reagan years, politicians used to answer the accusation that a large gap
had opened up between rich and poor by saying that, although the rich
had indeed become richer, the poor had little cause to complain since
they had not become much poorer, if at all. H owever, this takes no
account o f the fact that people's m ood and their psychological (and
sometimes physical) health are determined by social comparison. Abso
lute levels o f deprivation are higher in war, but cases o f suicide and
depression fall because all are perceived to be suffering together - social
cohesion goes up. Similarly, absolute levels o f deprivation are higher in
underdeveloped countries than in most Western countries, but there is
no excess o f suicides and depression in these poorer countries.
This explains some o f the discrepancy in the debate about suicide
and unemployment. There is little doubt that being unemployed is an
extra risk factor for suicidal behaviour, but overall rates are sometimes
found not to fluctuate with the unemployment rate as one might expect.
Research by Stephen Platt and colleagues in Edinburgh shows that this
confusion arises partly from the fact that unemployment has less
capacity to cause depression and suicide if the person perceives everyone
else to be in the same boat.7 It is not absolute levels o f poverty and
employment which are important in physical and mental health, but
distribution o f inequality.8 It is the conclusion that, compared to others,
one is a failure, unwanted or powerless, which increases vulnerability
to emotional distress.
Involuntary subordination
145
Suicide and Attempted Suicide
146
The Cry of Pain
147
Suicide and Attempted Suicide
feelings about the will to live and the will to die. As w e know, even
within the general population, many people have at some time felt
suicidal. Given that the lifetime risk for clinical depression is 20 per
cent, and that suicidal feelings are a com m on accompaniment to such
depression, it makes little sense to dismiss those who actually harm
themselves as not serious, or to dismiss the pressure to escape that the
person feels.
For these reasons, the cry for help, which many have misinterpreted
as a lack o f genuineness, is better seen as a cry o f pain. Suicidal behaviour
can have a communication outcome without communication being
the main motive. The behaviour is elicited by a situation in which the
person feels trapped. As with the animal in a trap that cries in pain, the
fact that the behaviour affects the behaviour o f other members o f
the species does not mean that the only motive for the cry was to seek
help. Suicidal behaviour m ay be overtly communicative in a minority
o f cases, but mainly it is ‘elicited’ by the pain o f a situation with which
the person cannot cope - it is a cry o f pain first, and only after that a
cry for help.
The ‘cry for help’ idea, though originally intended to be a neutral
theory about suicidal behaviour, has outlived its usefulness. It has
become limited to non-fatal suicide attempts and thus contributed to a
widening o f the gap between fatal suicide and attempted suicide. Also
it is almost always used pejoratively, or at least to imply that a certain
suicidal act was not so serious, but ‘m erely’ a cry for help. Such
behaviour is never ‘m erely’ anything.
Some self-harm m ay not be motivated by a wish to die, but most
shares with suicide the wish to escape from an otherwise unbearable
situation. The difficulty many authors have got into is to define com
pleted suicide as the core behaviour that needs to be explained, and
attempted suicide as its pale reflection. Instead, if one defines entrapment
and helplessness in theface of actual or threatened loss as the basic dimension,
it becomes easier to see that all self-harm falls som ewhere along the
dimension, w hatever the outcome. Such feelings o f being trapped are
fuelled and maintained by biological and psychological changes. They
appear completely impervious to intellectual argument, as Alvarez
found:
148
The Cry of Pain
Entrapment
149
Suicide and Attempted Suicide
the case o f the A level student with which the chapter began, there was
an unpredicted and sudden surge o f such feelings o f ‘no escape'. In
most cases, however, the entrapment is a longer-term state, first because
external causes o f stress are themselves long-term (e.g. the person is
trapped by bullying peers or partners). But it m ay be prolonged also
because the person has grown up in an environment where he or she
could exercise very little control, where others had all the power, so
the person has learned over a long time period that the only possible
response option is to submit. The person is thus extra-sensitive to social
threats, and their world constantly seems to present fewer alternatives
for action, whatever the reality. The result is long-term demobilization,
a biological state involving chemical changes.
The depressed person effectively ‘takes themselves off” the list" o f
those who might threaten the more powerful. This is akin to the
low-risk low-gain strategy o f some animals, a strategy that involves
internal inhibition o f any behaviour that would appear to challenge
those higher in the dominance hierarchy. Gilbert suggests that the
downturn into depression may involve the activation o f these inhibitory
mechanisms o f ‘no challenge'. At the extreme, they are associated with
not only reduced aspiration but also abnormally low aspiration: feelings
o f worthlessness, uselessness, powerlessness; and lack o f interest in
engaging in any social behaviour that might involve taking even min
imal risks. Depression is the biological assignment o f ‘loser’ status. It
signals to oneself and others that one is not prepared to take on
challenges or fight for resources.
The benefit in the animal kingdom is that the risk o f losing fights is
minimized, but the cost is that the chance o f dominating others is lost.
In the case o f bullying in schools or armed services or prisons, however,
the competitive nature o f the institution appears to encourage the bully
to continue to torment the subordinate long after the weaker one has
given up. The inhibitory mechanisms designed to protect the loser from
further losses fail to work. At such times the risk o f suicide is increased.
150
The Cry of Pain
151
Suicide and Attempted Suicide
152
The Cry of Pain
as closely associated with suicide. W e can now see how such anomie,
a sociological concept, relates to the psychological concept o f social
comparison, which itself relates directly to evolutionary old biological
sub-systems, each having a generally adaptive function, but sometimes
producing a maladaptive outcome. Social comparison has two import
ant elements: ranking (upwards v. downwards) and similarity to others
(like v. unlike). Those who are depressed and 'anomic' feel both inferior
to and also different from others.
Thus low self-esteem is a biologically old tendency that m ay be
activated relatively early in life, and reactivated at times when indi
viduals are, or imagine themselves to be, defeated, powerless and
failing to meet the challenges o f the world o f w ork or o f interpersonal
relationships. H ow is the gender effect to be explained within the
'entrapment' model? Historically, w om en have been given roles in
society in which they have less control, e.g. in earning power, control
over resources, in jobs even when working. In Western society, wom en
are likely to attribute their failures to themselves, and attribute their
successes to luck or to other people or circumstances. W om en have
thus explored the boundaries o f ‘no control' more than have men.
W hen the trap begins to close, they are more likely to see it closing.
The result is that they take action earlier than men. The cry o f pain
comes earlier in the entrapment process. At this early stage, people are
m ore ambivalent about dying, and use less lethal methods. The historic
tendency for w om en to be more likely to attempt suicide, but less likely
to commit suicide, is explained by this 'time course' model. In the past,
men did not show self-punitive behaviour at this stage. H owever,
having not reacted to threatened loss and entrapment early on in the
sequence, they miss out on the important benefits that earlier cries o f
pain may bring, and are more likely to m ove, in the extreme cases, to
helplessness and social isolation. W hen the trap closes, the effect is
more catastrophic.
For men, with less opportunity for receiving care (fewer social
support networks) and less opportunity for care giving (fewer caring
roles), the expectation o f no escape feeds directly into a sense o f
hopelessness and the possibility o f suicidal behaviour as the escape
route. As men become increasingly marginalized in the job market,
153
Suicide and Attempted Suicide
Concluding remarks
154
10
Memory Traps
155
Suicide and Attempted Suicide
biased or faulty, then our predictions are also likely to be biased and
faulty. Our self-esteem also depends on m emory. Self-esteem is based
upon our past successes and failures, and how successfully we have
navigated a path through the world so far.
Depression is one o f the major vulnerability factors for suicidal
behaviour and depression does not only affect mood, it affects m em ory
as well. Memory provides the key to understanding how, when som e
one feels under pressure from their life circumstances, they begin to
feel trapped in a mental cage from which they appear unable to escape.
First, m em ory can be biased so that it tends to retrieve only negative
events. Second, m em ory can be over-general. Instead o f recording
specific events, a person tends to lump together events o f the same type
or category without distinguishing between them. The story o f how
this m em ory problem was discovered is interesting in its own right.
156
Memory Traps
horse race. W hichever event gets to the line first wins, and when it
wins, that is the event that comes to mind.
It is now known that such a ‘horse race' occurs between positive
events and negative events from the past. One can imagine positive
memories as a white horse, and negative memories as a black horse.
My seeing the man and his dog activated a pleasant m em ory o f long
ago; the white horse won. But one effect o f depressed mood is to bias
m em ory so that the black horse wins much o f the time. Had I been
depressed, the sight o f the man and the dog might have brought to
mind the day I had to take m y dog to the vet to be put down, when he
was very old and grey and ailing. Further, the depression would have
suppressed any happy memories o f subsequent dogs, thus allowing me
to conclude, unhappily, that I have never had a dog as wonderful again.
Some o f the first experiments on depressed m ood m em ory bias
phenomena were performed by Professor Alwyn Lishman at the Insti
tute o f Psychiatry in London and by John Teasdale and colleagues at
the W am eford Hospital in Oxford. Lishman concluded that patients
who were more depressed in m ood found it easier to recall negative
events, but a problem with this early study was that the apparent mood
m em ory ‘bias’ might have arisen because the people who were more
depressed had fewer positive events in their lives.
T o avoid this problem, John Teasdale took student volunteers who
were not depressed and experimentally manipulated their mood using
a Mood Induction Procedure. All students started by being randomly
allocated to different m ood groups, so that results could not be explained
by differences in the number o f positive and negative events in their
lives. The earlier effect was replicated, with the additional finding that
negative m ood did not so much speed negative events as slow down
the recall o f positive events.1
Taking the analogy o f a horse race, the finding that depression does
not seem to w ork by speeding up the black horse is important. In
psychotherapy for depression, it will not be sufficient for a client to
w ork to make negative events less accessible. The therapy will need
also to increase the availability o f more positive aspects o f the past.
There is abundant evidence that depressed people have suffered a great
deal o f genuinely negative life events and chronic difficulties. Given the
157
Suicide and Attempted Suicide
158
Memory Traps
159
Suicide and Attempted Suicide
Cue-word valence
Figure 1 0 .1 Time taken to recall specific autobiographical memories
(from Williams and Broadbent, 'Autobiographical Memory in
Attempted Suicide Patients', Journal o f Abnormal Psychology , 1986).
160
Memory Traps
Overdose patients
Control participants
161
Suicide and Attempted Suicide
162
Memory Traps
Intermediate
163
Suicide and Attempted Suicide
Table 10.2 Scenarios used to cue memories (from Moore, Watts and
Williams, The Specificity of Personal Memories in Depression', British
Journal of Clinical Psychology, 1988)
164
Memory Traps
165
Suicide and Attempted Suicide
166
Memory Traps
167
Suicide and Attempted Suicide
1. Someone who loves his/her partner very much, but they have
many arguments, after one o f which the partner leaves.
5. Someone who comes home after shopping and finds that he /she
had lost a watch.
168
Memory Traps
In order to illustrate the scoring for effectiveness, and to illustrate the range o f
this dimension, two examples representing the extreme poles o f the scale,
from
169
Suicide and Attempted Suicide
170
Memory Traps
171
Suicide and Attempted Suicide
172
Memory Traps
Concluding remarks
173
11
The Prevention of Suicidal
Behaviour
In assessing what scope exists for preventing suicide, one o f the most
obvious measures would be the more sensitive assessment o f suicide
by health professionals. Depression is a frequently occurring psychiatric
disorder and most depressed patients are treated in general practice.
Training primary-care professionals to recognize depressive symptoms
and treat them appropriately is a potentially important prevention
strategy. Evidence suggests that 50 per cent o f sufferers are not recog
nized by their general practitioner as suffering m ajor depression. A
further 10 per cent are subsequently recognized, and o f the 40 per cent
not recognized half will remain depressed. Tw enty per cent o f the
original sample, therefore, are still depressed and not recognized six
months later.
There are a number o f reasons w hy depression is missed in the
context o f a G P ’s clinic. The patient m ay present somatic symptoms
and physical problems; m ay feel there is a stigma about presenting
psychological problems, compounded by a belief that the doctor does
not have time to listen to their psychological problems. There is also
some evidence that depression is more likely to be missed if it is o f
recent origin, where the symptoms are atypical, the depressed m ood is
less severe and the patient has less insight into their own symptoms. A
picture emerges o f a complex interplay o f factors which result in the
primary-care worker not spotting depressed symptoms.
Depression and hopelessness are clearly associated with suicidal
ideation and behaviour. If depression is going unrecognized, then sui
cidal ideation is also likely to go unrecognized. Implicit blame is often
174
The Prevention of Suicidal Behaviour
175
Suicide and Attempted Suicide
all around them elicits talk about their physical symptoms rather than
their psychological difficulties.
Another indication o f how the interpersonal context determines
what takes place between health professional and patient is a finding
about how the age o f the patient determined outcome. Patients o f 55
or older were not asked at all about suicidal thinking. Given that these
patients were the most vulnerable for completed suicide, it seemed a
surprising result. One possible explanation is that older people are more
likely to express problems through physical symptoms. Another is that
care-givers feel more reluctant to ask about psychological symptoms o f
people older than themselves. Social mores that dictate respect for older
people and their privacy m ay be a more powerful determinant o f
professionals' behaviour than previously thought.
176
The Prevention of Suicidal Behaviour
177
Suicide and Attempted Suicide
178
The Prevention of Suicidal Behaviour
Reply to criticism
In fact, the writer o f the letter had made a simple but critical error in
his reanalysis o f the Gotland data. Instead o f examining the suicide rate
year by year, he had calculated a five-year m oving average. Taking a
m oving average o f such data is a well-recognized method o f ‘smoothing
a graph so that trends m ay be seen more clearly. In the procedure, a
‘w indow ' o f consecutive points in a series (between two or five points
usually) is taken, and the average used as a new data point. So, a
five-year m oving average takes data points from years i, 2, 3, 4 and 5
for the first data point; years 2, 3, 4, 5 and 6 for the second; 3, 4, 5, 6 and
7 for the third, and so on. The five-year ‘w indow ' is m oved across the
data and the graph is smoothed.
H owever, if an intervention is introduced, the m oving w indow must
stop at the last data point before the intervention. Otherwise the window
is gradually contaminated by more and more data points from the
intervention phase. T o illustrate this, consider the dum my data in
Figure 11.2 which clearly show the effect o f an intervention. Figure 11.3,
however, shows what happens to this graph if a five-year m oving
average is applied which does not stop when the intervention starts.
The baseline data are contaminated increasingly by the data from the
intervention phase, giving the w rong impression that the rate o f the
behaviour in question started to decline before the intervention -
precisely the mistake made in the critical letter.
Figure 11.4 shows the smoothed means o f the Gotland data retaining
a five-year m oving average but stopping the window prior to the
intervention. As can be seen, the intervention follows three successive
drops in the smoothed graph (o f 1.9 ,1.2 and 2.7 per 100,000 respectively).
In this sense, the author o f the critical letter was right to point out some
reduction in the suicide rate prior to the intervention phase starting,
but his error made this reduction seem much greater than it was. The
first post-intervention data point is 7.0 per 100,000, low er again than the
last point prior to the intervention - a much larger drop than that
between any two prior data points.
179
Suicide and Attempted Suicide
30 1-
20 -
10 -
o i 1 1 1 ___ 1_____1__ 1_____ 1_1_____ 1____ 1_________ 1_1_1____1----- 1-------- 1--1----- 1------1----- 1—
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Time
Figure 11.2 Dummy data illustrating effect of hypothetical
intervention to reduce death-rate.
Time
180
The Prevention of Suicidal Behaviour
Years
Figure 11.4 Graph showing the smoothed means of actual Gotland
data retaining a five-year moving average but stopping the window
prior to intervention.
181
Suicide and Attempted Suicide
•
182
The Prevention of Suicidal Behaviour
rate for ‘true positives', predicting suicide and getting suicide). It must
also have adequate specificity, predicting the true negatives (predict no
suicide and get no suicide).8 The issue o f specificity is particularly
significant because o f the low base rate o f suicidal behaviour. Thus,
even in a highly selected group, there will be more people falsely
identified as at risk (false positives) than correctly identified as at risk
(hits). W hen trying to predict suicide the problem is even greater.
Pokom y, in a prospective study,9 followed up almost 5,000 in
patients. The predictive model identified 35 out o f the 67 subsequent
suicides but at the cost o f over 1,000 false positives. W hichever analysis
Pokom y used, he found the result was a false positive rate o f 25-30
per cent. O ver 1,000 people would have been put on the vulnerable list,
but not have committed suicide. Perhaps more worrying was the 44
per cent false negative rate. This means that o f the people who did
com m it suicide, 44 per cent would have been allocated to the ‘low risk'
category.
This pessimistic message is reinforced by the suicidologist, Ronald
Maris.10 He points out that so-called 'high risk' groups are not, after all,
at that high a risk o f suicide. Take the statistic that 15 per cent o f those
who have been patients in psychiatric units with a diagnosis o f major
depression will commit suicide. They are at a greatly increased risk
compared with the general population. Yet even they will commit
suicide at a group rate o f some 1 per cent per year over a period o f
about 30-35 years. For every 100 patients in the ‘suicide high risk'
category, only one will actually commit suicide in any one year, and
we cannot be sure which one, or when. The question o f ‘when’ is one
o f the most difficult to answer.
Timing
183
Suicide and Attempted Suicide
184
The Prevention of Suicidal Behaviour
Availability of means
186
The Prevention of Suicidal Behaviour
Clinicians and researchers have had little doubt for some time that
if help and protection are available during a period o f suicidal crisis, and
lethal methods are not to hand, the crisis m ay pass and the person not
commit suicide. It m ay not even be help that is needed, but simply
distraction at a critical moment. One person who put the muzzle o f a
revolver into his mouth and was feeling for the trigger, suddenly heard
his children laughing and running through the hall. 'That snapped me
out o f it. The suicide impulse lasted only a moment - but that's all it
takes,' he reported later.
The basis for expecting that availability o f lethal methods makes a
difference is this: if the preferred method is not there, the motivation
to search for an alternative m ay not be high enough to prompt such a
search. The suicidal feelings m ay pass without being acted upon. If this
seems difficult to understand, we only need recall that suicide occurs in
the context o f hopelessness and despair. If a single opportunity appears
to present an escape, yet that escape route is blocked, the despair may
turn into hopelessness about suicide as an effective solution. The
‘Russian roulette' aspect o f suicidal motivation produces a sense o f not
caring ‘whether I live or die'. If a person has effectively ‘allowed the
Fates to decide', then their verdict in favour o f staying alive may be
passively accepted. But if, at that moment, the person has access to
lethal methods, the outcome will be suicide.
O f course, a person determined to kill him or herself may take many
steps to ensure success, including taking themselves away to a place
where they will not be interrupted. Even where this does not occur,
the family or friends o f a suicidal person cannot be on hand 24 hours a
day. T o try to be constantly present in this w ay would put so much
pressure on a relationship it would be unsustainable. Nevertheless it is
possible to try to ensure that lethal means o f suicide are unavailable for
those attempts that are more impulsive. The problem is knowing which
threats to remove.
In Savage God, Alvarez quotes the views o f Seneca, who said that the
means o f committing suicide are everywhere: each precipice and river,
each branch o f each tree, every vein in the body will set a person free.
Alvarez disagreed: ‘No one is promiscuous in his w ay o f dying. A man
w ho has decided to hang him self will never jum p in front o f a train.
187
Suicide and Attempted Suicide
And the more sophisticated and painless the method, the greater the
chance o f failure: I can vouch, at least, for that.’ If true, this means that
a health professional may ask someone who is suicidal what they have
thought o f doing, and try to ensure that these means are removed.
Research evidence
The preventative effect o f rem oving such lethal means has been demon
strated many times. The most com m only cited example is the fall in
suicide rates during the 1960s and 1970s in the United Kingdom as
domestic gas was detoxified. In 1948-50, poisoning by domestic gas
accounted for 41 per cent o f male suicides and 60 per cent o f female
suicides. By 1970, only 16 per cent o f males and 9 per cent o f females
used domestic gas in suicide. Death by this method had completely
disappeared by 1990. It has been estimated that the detoxification o f
domestic gas has prevented approximately 6,700 deaths by suicide.
Similarly, there was a decline in the suicide rate in Australia in the
late 1960s and early 1970s, an effect directly attributable to legislation to
reduce amounts o f barbiturate and other sleeping pills. Before the mid
1960s, drug overdose was the most common form o f suicide in Australia.
Up to this time, 100-300 tablets or capsules o f sedative sleeping pills,
such as barbiturates, had been available, making suicide by their use
extremely easy. In Ju ly 1967 it became illegal to prescribe sedative
hypnotic drugs, particularly barbiturates, in greater quantities than
25 tablets or capsules. Statistics for Australia showed that restricting
barbiturates in this w ay made the suicide rate fall in ensuing years.
H owever, the reduction in overall rate from the unavailability o f
one method may gradually bottom out and alternative methods become
more common. In England and W ales this occurred with the use o f car
exhaust replacing domestic gas. Car exhaust now accounts for 35 per
cent o f male suicides in England and Wales, compared with 31 per cent
w ho hang themselves and 14 per cent who take poisons. The pattern
for females is slightly different, with self-poisoning remaining the most
common method (44 per cent), followed by hanging (23 per cent) and
vehicle exhaust (13 per cent).
The increase in vehicle exhaust deaths parallels the increase in m otor
vehicle use in the past twenty years. In fact it slightly exceeds the rate o f
188
The Prevention of Suicidal Behaviour
increase in car usage, but this may be owing to the fact that within the
car market there has been an increase in the proportion o f hatchback
models, which makes suicide by this method easier. Some have advo
cated changes to exhaust-pipe design, to make it more difficult to attach
a hose, though ultimately it m ay be more useful to install safety cut-out
switches in cars to switch o ff the engine in response to high levels o f
fumes. In any event, as emission controls on car exhausts are adopted
by more and more countries, we can expect to see a reduction in suicide
using car exhaust in the coming years.
Dangerous weapons
Suicide rates using firearms are, o f course, another indicator that avail
ability o f means is important. Guns are used in over 50 per cent o f cases
in the United States compared to only 3 per cent in Great Britain. In the
United States the impact o f availability has focused on the strictness
o f gun legislation from state to state. Lester (1989) found significant
correlations between ‘gun control statute strictness' and rates o f suicide
over the 48 continental states.
Although stricter gun laws correlated with few suicides by firearms
(suggesting that making the method less available will reduce fatalities),
there was also an association between low er firearm availability and
the use o f alternative methods (though not poisons or hanging). Despite
this, there are reasons to think that availability is important. The positive
correlation between strict gun laws and alternative methods does not
tell us the overall level o f suicide by these other methods. Indeed, other
w ork by Lester16 has shown that the total suicide rate is lower in states
with strict gun control laws. It appears Alvarez is right: only a few
suicidal people switch to an alternative method for suicide. People seem
to have a preferred method they would use to kill themselves, and are
unlikely to deviate from this when actually suicidal. This is important
clinically, for depressed patients, when not very suicidal, may agree to
get rid o f the means o f killing themselves that they know they might
use later when feeling worse.
189
Suicide and Attempted Suicide
Dangerous medication
Other evidence o f a link between suicide and availability o f method is
shown by the findings o f a marked correlation between the number o f
prescriptions given out for sedatives, sleeping pills and tranquillizers,
and the rate o f suicide by self-poisoning.
Prescribed mood-altering drugs, such as tricyclic antidepressants,
account for 15 per cent o f all suicides. Some suicides may therefore be
prevented by the prescribing o f relatively non-toxic antidepressants.
One problem with prescribing safer antidepressants is that they are
more expensive, and difficulty in predicting who is most vulnerable to
suicide at what point in time means there might have to be a general
switch to these newer substances to bring about a reduction in the
suicide rate. Such a switch would mean that the cost per life saved
would be very high indeed. On the other hand, use o f these safer
antidepressants in those known to have a higher risk o f suicide (psychi
atric patients, those with a history o f self-harm, etc.) appears an im por
tant potential development.
A clear relationship between prescribing patterns and the deaths by
these drugs can be found. The clear implication is the need to give
patients at risk medication for a few days only and/or give the medi
cation to another family m em ber to keep. Neither o f these preventative
strategies has been systematically studied.
There are even more deaths attributable to analgesics, anti-fever
and anti-rheumatic medications, many o f which are available without
prescription. Ten per cent o f all suicide deaths by overdose are caused
by an overdose o f paracetamol and 5 per cent from aspirin. H ow ever,
nearly all patients (94 per cent) who take an overdose take the first drug
they can obtain. Few (only 20 per cent)17 are found to know about the
toxicity o f a drug or what quantity would be lethal. Many wrongly
believe that aspirin and paracetamol, because freely available from
the pharmacist, are relatively non-toxic. In France paracetamol is not
allowed to be sold in greater quantities than 8 grams (16 x 500 milligram
tablets) and there are very few fatal paracetamol overdoses. Similar
limits have now been introduced in the United Kingdom.
190
The Prevention of Suicidal Behaviour
Dangerous jo b s
A final piece o f evidence that availability o f lethal means is important
comes from the sort o f jobs people do. The only thing the professions
most vulnerable to suicide (veterinary surgeon, dental practitioner,
pharmacist, farmer and medical practitioner) have in common is access
to lethal means o f self-killing. Recent evidence suggests that the nursing
profession is also vulnerable. Between 1988 and 1992, nurses accounted
for a greater number o f suicides than any other occupation and over
5 per cent o f all female suicides.18 This is w holly consistent with the
‘dangerous jobs' explanation, since nurses know about the lethality o f
drugs and have ready access to them.
Between 4 and 7 per cent o f people who commit suicide in the United
Kingdom have had past contact with the Samaritans, but how recent
these contacts are is not clear. For many years there has been a debate
about whether suicide prevention centres and organizations significantly
affect the suicide rate. W hat is not in dispute is that there is a great need
for such centres and organizations, as shown by the extensive use made
o f them. H owever, even though people know where to turn for help,
when they are very suicidal they m ay not act on this knowledge. One
study conducted by Greer and Alderson in 197919 found that o f those
patients who attempted suicide, 72 per cent had sufficient knowledge o f
the Samaritans to be able to contact them yet less than 2 per cent had
actually sought their help on that occasion. Barraclough and colleagues
found that those towns which had established a Samaritans branch did
not differ in suicide rate from those towns which had not.20
A more encouraging report came from an analysis o f the changes of
the suicide rates in 226 cities in the United States between 1968 and 1973.
Some had developed suicide prevention centres over the period, others
had not. There was a reduction in suicide rates am ong white females
below the age o f 25 years in those cities that had introduced such
centres. Given that this was just the sort o f client who most frequently
used those centres, here indeed was an encouraging finding.
191
Suicide and Attempted Suicide
School-based intervention
192
The Prevention of Suicidal Behaviour
Concluding remarks
Given all this information about who is most vulnerable and what are
the most dangerous circumstances, can we estimate the possible effects
o f making changes? Gunnell lists all the potential effects o f suicide
prevention strategies for the United Kingdom.24 Most strategies, he
concludes, would have uncertain effects. H ow ever, he estimates that a
4 per cent reduction in suicide rates could be achieved by G P s prescrib
ing safer antidepressants; that increased care around the time o f
193
Suicide and Attempted Suicide
194
12
Therapy for Suicidal Feelings and
Behaviour
195
Suicide and Attempted Suicide
(e) how stable is the client’s life situation (both objectively and
subjectively to the client), especially their interpersonal relation
ships (by far the most com m on precipitant)?;
( f ) H ow impulsive is the client? Given that two thirds o f episodes
o f self-harm are contemplated for less than an hour beforehand,
impulsiveness m ay be considered an important vulnerability
factor.
The client may arrive at the conclusion that life is intolerable through
combinations o f the errors in reasoning described in depression. In
dichotomous thinking (black/white, all-or-nothing thinking), there is no
perceived middle path - just the extremes. Selective abstraction consists
o f the selecting out o f small parts o f a situation and ignoring others,
e.g. a tutor’s report on an essay gives much praise, but mentions at one
point that the introduction is too long: ‘He doesn’t like m y essay’ would
be selective abstraction. In arbitrary inference, a conclusion is inferred
from irrelevant evidence, e.g. an individual phones a boy/girlfriend and
no one answers. The conclusion: ‘(S)he’s probably out with another
partner’ would be an arbitrary inference (if inferred on those grounds
alone). Over-generalization is concluding from one specific negative event
that other negative events are therefore more likely, e.g. failure at
maths means failure at everything. Finally, catastrophizing is to think
the very worst o f a situation.
Many o f the errors o f logic which underlie such beliefs and assump
tions are exactly those found in depression. The danger with suicidal
clients is that they act so decisively and violently on their beliefs. Often
this process takes place in only a few minutes. First, their thinking about
their problems is dominated by the distortions described; then they
react to these thoughts as inescapable facts and take proportionately
196
Therapy for Suicidal Feelings and Behaviour
197
Suicide and Attempted Suicide
198
Therapy for Suicidal Feelings and Behaviour
( f ) whether they expect to put their plan into action, and how
imminently;
(g) whether anyone would care if they carried out such a plan.
199
Suicide and Attempted Suicide
evidence for and against each reason for living and for dying; and
discussing w hy some reasons are weighted in the client's mind more or
less heavily than others. There is no evidence that explicit discussion o f
such issues increases suicidal intent. Indeed, clinical experience suggests
the converse.
Problem-solving therapy
200
Therapy for Suicidal Feelings and Behaviour
The first step is probably the most important. During a suicidal crisis,
everything in a client’s life can seem overwhelm ing to them and to
their therapist. It m ay be difficult therefore to collaborate in articulating
as accurately as possible all the things that are going wrong. Yet the
aim remains to draw up a problem list. On the list, agreed between
201
Suicide and Attempted Suicide
client and therapist, each problem is clearly described. Often this will
be very difficult. People's problems may not seem easy to describe
clearly. The therapist will then need to help the client be more specific
(e.g. ‘W hat is it about x which is difficult?' ‘Could you go into more
detail about the problem you are having with yY). Only if sufficient
specificity is achieved at this stage can concrete strategies to cope with
the problem be identified.
Hawton and Kirk offer the following helpful check-list o f potential
problem areas:
202
Therapy for Suicidal Feelings and Behaviour
During the next session, client and therapist make use o f the information
gained from the last hom ew ork to decide on the next step. This clearly
depends on what progress was made, and will lead to further agreement
about subsequent tasks. Further goal(s) m ay need to be agreed, or
problems redefined for further loops around the sequence.
There are a number o f other techniques used in problem-solving
therapy, including some that are common to cognitive therapy. These
include:
203
Suicide and Attempted Suicide
204
Therapy for Suicidal Feelings and Behaviour
Hawton and Kirk point out that problem-solving may not be appropriate
in every case, or at every stage in the crisis:
Early studies using these types o f approach have used both out-patient
treatment sessions and sessions conducted in the client’s own home. In
these early studies, repetition rates were not reduced. O f course, other
benefits o f such treatments have emerged. One such study5 used task-
centred case-work (nine sessions over three months) and compared it
with normal psychiatric out-patient visits. At four months’ follow-up,
the psycho-social treatment group showed significantly fewer social
problems and were better at handling personal relationships and social
transitions than the other group. In addition, they were less lonely and
better at coping with their own emotional distress as well as with
practical difficulties in their lives.
H ow ever, this study, and others like it, was not able to show an effect
on repetition. This is for two reasons. First, the treatment produces most
205
Suicide and Attempted Suicide
206
Therapy for Suicidal Feelings and Behaviour
207
Suicide and Attempted Suicide
208
Therapy for Suicidal Feelings and Behaviour
159
15
DBT
14
TAU
13
12
11
10
Phase o f treatment
209
Suicide and Attempted Suicide
r "V""-
_ J DBT
TAU
2.1
1.06
O.I
The promising data o f the 1991 report only provide details o f the effect
o f treatment while it continued. Remarkable as these treatment gains
are, it was still an open question whether any o f them would be
maintained once the therapy finished. The follow-up study was there
fore o f great interest.8 It was a naturalistic follow-up, and therefore did
not prevent people from returning to treatment if they wished. Control
clients were able to continue to receive psychotherapy in the com
munity. D B T clients, although required to take a ‘two-month vacation'
from their individual D B T therapists, could continue in psychotherapy
thereafter if the therapist was agreeable. Seven subjects (35 per cent)
took this option. This contrasts with 55 per cent who continued with
psychotherapy in the control group.
Follow-up assessments were made at 6 months and 12 months
210
Therapy for Suicidal Feelings and Behaviour
16
___: DBT
14
TAU
12
10
o
0-6 months 7-12 months
Post-treatment period
211
Suicide and Attempted Suicide
o. 8
m D BT
TAU
0.6 0.56
0.2
0.1
0.1 0.05
o
0-6 months 7-12 months
Post-treatment period
212
Therapy for Suicidal Feelings and Behaviour
213
Suicide and Attempted Suicide
214
Therapy for Suicidal Feelings and Behaviour
contact with the psychiatric service over the year, the majority by
telephone. Some got in contact more than once, but even the total
number o f contacts was only 19, o f which 4 involved face-to-face
interviews. In half such contacts, a single episode o f discussion was
enough to deal with the problem. Only in one case was there a request
for readmission to hospital.
There needs to be some caution in generalizing these results. W e
have seen (page 72) that the pattern o f non-fatal suicidal behaviour
changed during the 1990s, with more men being admitted following
self-harm than ever before. W hether a Green Card will be effective for
this new group o f suicidal clients remains to be seen. More recent
studies o f the Green Card have been more uncertain o f its effects.
H ow ever, the studies referred to in the previous chapter, by Motto
(1976) and De Leo (1995), are more optimistic. M erely keeping in touch
with vulnerable people by correspondence or by telephone appears to
be able to reduce the suicide rate, at least for the time that the contact
is kept up.14
215
Suicide and Attempted Suicide
216
13
Final Thoughts
217
Suicide and Attempted Suicide
such feelings are, whether there have been models among family or
friends or in the media, whether a suitable method is readily available,
how violent or impulsive the person is, and whether drugs or alcohol
are available which reduce fear o f death and impair judgem ent.
Completed suicide and non-fatal attempted suicide can be under
stood as different responses to these circumstances, occurring at differ
ent points in the downward spiral into hopelessness. For m any years it
was thought that suicide and attempted suicide had to be qualitatively
different behaviours, partly because many w ho harm themselves say
they do not want to die. Their motivation seemed much more complex
than a simple wish to die. But motivation for completed suicide is also
complex. Because the tragedy o f a death by suicide is so extreme, we
tend to assume that, when death is the outcome, death has been the
predominant motive.
Yet I have also suggested that the predominant motivation in suicidal
behaviour is escape. The person feels trapped. They can see no w ay
out o f their prison, and take little account o f the possibility that some
o f their feeling o f entrapment comes from a biased view o f their own
past life that feeds into their hopelessness about the future.
Getting aw ay from seeing completed suicide as motivated only by a
wish to die frees us from a punitive view o f attempted suicide. In the
past, we have allowed the question o f how much a person wished to
die to define the w ay in which w e understand all who harm themselves.
W e ask ourselves, ‘H ow suicidal was this behaviour really?' If we decide
it was not, then w e are inclined to dismiss it, get angry about it, see it
as ‘manipulative', and so on. (O f course, the truth is that those whom
we call manipulative have actually failed to manipulate us. For w e all
manipulate each other all the time. W e are simply so good at it that no
one notices. It is only when w e meet someone in w hom such attempts
are more obvious that we notice it. It is ironic that we accuse them
o f being manipulative when, in fact, they have failed adequately to
hide the social performance in which we all engage.) The problem with
the idea o f the ‘cry for help’ is that it fed this dismissive view o f
self-harm.
O f course, there were other reasons for thinking that suicide and
attempted suicide were different things. The most important reason
218
Final Thoughts
was that those who completed suicide on the one hand, and those who
attempted suicide on the other, seemed to differ in a number o f respects.
For example, while suicide, until recently, was predominantly a feature
o f older males, attempted suicide seemed predominantly a feature o f
younger females. Furthermore, while suicide rates were coming down
in the 1960s in the United Kingdom, attempted suicide was rising. Such
differences between suicide and attempted suicide seemed to suggest
that different explanations for each were needed. Yet, if we are to see
age and sex differences as pointing to different underlying causes, what
are we to make o f the recent data from the United States and the United
Kingdom that show male suicide is rising while female suicide is falling?
Here w e have the same outcome, suicide deaths, m oving in different
directions in different sub-groups o f the same population.
Instead it is possible that the same underlying set o f causes might
produce the age and sex differences in suicidal behaviour that are
sometimes observed. If suicidal behaviour is best seen as a cry o f pain
- a response to feeling trapped by uncontrollable external circumstances
and uncontrollable internal anguish - then different types o f people,
older or younger, male or female, take action on these feelings at
different points in the sequence o f events as the trap is perceived to
close. Internal and external stresses give rise to differences in perceived
escape potential (i.e. their ability to see a w ay out). This, in turn, gives
rise to stronger or weaker wishes to die.
At one end o f the continuum, a weaker wish to die may not be
expressed as a wish to die at all, but rather be seen as seeking temporary
oblivion. But such people are best seen as one end o f a dimension o f
lethality that cannot be ignored. The protest and anger that most often
produce non-fatal suicidal behaviour represent a response early in the
sequence o f events when escape still seems possible. The despair and
apathy that produce more lethal suicidal behaviour represent the
response to loss that comes later when the person sees no hope in their
situation at all. Even at this stage, however, the presence o f social
support can ameliorate the intensity o f feelings o f hopelessness. Men
differ from w om en in the extent to which they perceive or make use
o f such supports. This is not simply a question o f biology, but o f the
different ways in which boys and girls are taught to view emotional
219
Suicide and Attempted Suicide
220
Final Thoughts
depressed, even if they have talked o f suicide, the actual death comes
as a huge shock. And, o f course, suicide sometimes comes right out o f
the blue.
The huge trauma o f any sudden death is compounded in a suicide
by a number o f factors: many still feel a social stigma about suicide, and
there is often a ‘conspiracy o f silence’ . This is especially true where
children are involved, and children whose parent or brother or sister
has committed suicide may thereby have to deal, with little support,
with their own misconceptions and irrational guilt about what has
happened. The outcome for children when their parents have separated
before one parent commits suicide is worse than for those whose
parents are still together.2
Even when social support is available, the suicide survivor is less
likely to use it than those bereaved by other kinds o f sudden death,
221
Suicide and Attempted Suicide
222
Final Thoughts
world, including one’s own actions. Small children develop the ability
to take full account o f other people’s feelings and beliefs as part o f their
developing language and social communication in the second and third
years o f life.
I believe that the radical loss o f empathy that accompanies suicidal
impulses in some people results from impairment o f these same psycho
logical mechanisms. Suicidal despair switches o ff the processes respon
sible for understanding the beliefs, intentions and feelings o f others.
Though this will often appear to be a callous disregard for others’
feelings (especially close family and friends), it is more likely to be
something over which the person has little control.
The quotation from Primo Levi with which this short chapter began is
taken from his last book, The Drowned and the Saved. Levi had survived
Auschwitz, and had written o f these experiences in his books, I f This Is
a Man and The Truce. He was found dead at the bottom o f the stairwell
o f his house in Turin in 1987, and is presumed to have committed
suicide. In the quotation, Levi speaks o f an anguish sometimes so deep
that only the specialist, the analyst o f souls, knows how to exhume it.
But what help can an analyst o f souls provide? W e cannot opt out
o f trying to answer such questions, no matter how much we would
wish to remain silent. W hen therapies bring hope to those in despair,
what is it that they are doing? O f course, they are offering support.
They are helping the person to see that perhaps some o f their problems
are solvable. They are helping the person to gain some distance from
the constant propaganda o f the mind that would persuade them they
are a failure. But I believe these are helpful only to the extent that they
allow the person to give up wanting things to be different. As Marsha
Linehan points out in her therapy (see page 208), it is about balancing
acceptance and change. For when people are in such distress, they
become trapped by the idea that, if things were different, then all would
be well: if their partner was different; if their job was different; if their
house was different; if they themselves were different.
223
Suicide and Attempted Suicide
It’s one that cannot be resolved by trying to make the world into a different
place - which tends to be the normal approach. To make peace with despair is
a matter o f understanding not just where difficulties such as sickness and
violence arise, but also how the feeling o f being bound to and oppressed by
those problems occurs.4
Therapies that are helpful are those which allow the person to see their
moods as normal, rather than as evidence o f their inherent deficiency
as a person. They encourage the person to ask not, 'H ow can I make
everything different?' but rather, 'H ow can I take care o f m yself right
now?' In short, they encourage the person to be gentle with themselves.
224
Notes and References
Introduction
225
Suicide and Attempted Suicide
226
Notes and References
227
Suicide and Attempted Suicide
23. G. E. Murphy and R. D. Wetzel, 'Suicide Risk by Birth Cohort in the United
States, 1949-1974’, Archives of General Psychiatry, 37 (1980), pp. 519-2.3; M. J.
Solomon and C. P. Hellon, ‘Suicide and Age in Alberta, Canada, 1951-1977: a
Cohort Analysis’, Archives o f General Psychiatry, 37 (1980), pp. 511-13; R. D.
Goldney and M. Katsikitis, ‘Suicide Rates in Australia’, Archives of General
Psychiatry, 40 (1983), pp. 71-4.
24. Charlton, Kelly, Dunnell et a i, ‘Trends in Suicide Deaths in England and
Wales’ (op. cit. see note 6).
25. See U. Bille-Brahe, ‘Sociology and Suicidal Behaviour’ , in K. Hawton and K.
van Heeringen, (eds) (op. cit.), pp. 193-208. Bille-Brahe discusses evidence
suggesting that cohort size is not always associated with suicide, but only in
those countries that have poorer welfare provision so cannot make up for
limited resources when a large cohort competes for them (pp. 24-5).
26. Charlton, Kelly, Dunnell et al., ‘Trends in Suicide Deaths in England and
W ales’ (op. cit., see note 6).
27. N. Kreitman, V. Carstairs and J. Duffy, ‘Association o f Age and Social Class
with Suicide among Men in Great Britain’,Journal o f Epidemiology and Community
Health, 45 (1991), pp. 195-202.
28. For the Cross National Collaborative Group reference, see ‘The Changing
Rate o f Major Depression: Cross-national Comparisons’, Journal of the American
Medical Association, 268 (1992), pp. 355-73.
228
Notes and References
1. A. Alvarez, Savage God, A Study of Suicide, Random House, New York, 1972.
2. Although true in the individual case, studies o f larger numbers have found a
correlation between suicide intent and the number o f pills taken. More than
twenty pills has been found to indicate higher suicide intent. See R. D. Goldney,
‘Attempted Suicide in Young Women: Correlates o f Lethality’, British Journal
o f Psychiatry, 147 (1981), pp. 382-90; K. G. Power, D. J. Cooke andj. S. Gibbons,
‘Life Stress, Medical Lethality, and Suicidal Intent’, British Journal of Psychiatry,
147(1985), pp. 655-9; and D. J. Pallis, J. S. Gibbons and D. W. Pierce, ‘Estimating
Suicide Risk among Attempted Suicides’, British Journal o f Psychiatry, 144 (1984),
pp. 139-48.
3. K. Hawton, J. Fagg, S. Simkin, E. Bale and A. Bond, ‘Attempted Suicide in
Oxford, 1994’, unpublished report from University Department o f Psychiatry,
Oxford.
4. K Michel, P. Ballinari, U. Bille-Brahe et a l, ‘Methods Used for Parasuicide:
Results o f the W H O /E U R O Multicentre Study o f Parasuicide’, Social Psychiatry
and Psychiatric Epidemiology, 35 (2000), pp. 156-63.
5. For details o f the Canada data, see I. Sakinofsky, ‘The Epidemiology o f Suicide
in Canada’, in A. Leenaars, S. Wenckstem, I. Sakinofsky, M. Krai, R. Dyck and
R. Bland (eds), Suicide in Canada, University o f Toronto, 1996.
229
Suicide and Attempted Suicide
230
Notes and References
231
Suicide and Attempted Suicide
232
Notes and References
1. D. Humphry, ‘Rational Suicide among the Elderly’, Suicide and Life Threatening
Behavior, 22 (1992), pp. 125-9, see p. 127; D. Humphry, Final Exit: the Practicalities
o f Self-deliverance and Assisted Suicide, Carol Publishing, Secaucus, NJ, 1991.
2. Humphry, ‘Rational Suicide among the Elderly’ (op. cit., see note 1).
3.J. H. Groenewoud, P .J. van der Maas, G. van der Wal eta i, ‘Physician-Assisted
Death in Psychiatric Practice in the Netherlands’, New England Journal of
Medicine, 336 (1997), pp. 1795-801.
4. Humphry, ‘Rational Suicide among the Elderly’, (op. cit.), p. 126.
5. A. Koestler, Arrow in the Blue, Collins and Hamish Hamilton, London, 1952.
6. R. D. Goldney, ‘Arthur Koestler: Was His Suicide Rational?’ , Crisis, 7 (1986),
pp- 33- 53-
7. J. Richman, ‘A Rational Approach to Rational Suicide’ , Suicide and Life
Threatening Behavior, 22 (1992), pp. 130-41.
8. Scott, ‘Cancer Patients’, in Scott, Williams and Beck (eds), Cognitive Theory
in Clinical Practice, (op. cit., see note 19, Chapter 2).
9. Richman, ‘A Rational Approach to Rational Suicide’ , (op. cit., see note 7).
1. S. Freud, Mourning and Melancholia (1917), Standard Edition, vol. 14, Hogarth
Press, London, 1957.
2. Ibid.
3. S. Freud, The Ego and the Id (1923), Standard Edition, vol. 19, Hogarth Press,
London, 1961.
4. For psychoanalytic observations on suicide as a ‘reversible’ and ‘magical’ act,
see K A. Menninger, ‘Psychoanalytic Aspects o f Suicide’, International Journal
o f Psychoanalysis, 14 (1933), p. 376; C. W. Wahl, ‘Suicide as a Magical Act’, Bulletin
o f the Menninger Clinic, 21 (1957), p. 91.
5. M. Klein, ‘A Contribution to the Psychogenesis o f Manic-depressive States’
(i935), in Contributions to Psycho-Analysis 1921-1945: Melanie Klein, Hogarth Press,
London.
6. H. Guntrip, Schizoid Phenomena, Object Relations and the Self, Hogarth Press,
London, 1968.
7. S. Asch, ‘Suicide and the Hidden Executioner’, International Review of Psycho
analysis, 7 (1980), pp. 51-60.
8. J. Bowlby, Maternal Care and Mental Health, Columbia University Press, 1951.
233
Suicide and Attempted Suicide
234
Notes and References
235
Suicide and Attempted Suicide
236
Notes and References
and London, 1989. See also J. S. Price and L. Sloman, ‘Depression as Yielding
Behaviour: An Animal Model Based on Schjelerup-Ebb's Pecking Order',
Ethology and Sociobiology, 8 (1987), pp. 85-98.
17. See P. Gilbert and S. Allan, ‘The Role o f Defeat and Entrapment (Arrested
Flight) in Depression: An Exploration o f an Evolutionary View', Psychological
Medicine, 28 (1998), pp. 585-98. There is now increasing evidence that biological
processes are triggered by the perception and expectation o f inescapability and
that a vicious circle involving biological, social and psychological factors can
occur. See K. van Heeringen, K. Hawton an d j. M. G. Williams, ‘Pathways to
Suicide: An Integrative Approach', in K. Hawton and K. van Heeringen (eds)
(op cit.), pp. 223-36.
18. MacDonald and Murphy, Sleepless Souls (op. cit., see note 2, Chapter 1).
10 Memory Traps
237
Suicide and Attempted Suicide
Approaches to the Study o f Memory, Cambridge University Press, 1988, pp. 244-76.
8. J. Morton, T h e Development o f Event M em ory’, Psychologist, 1 (1990),
pp. 3-10.
9. Although no studies have yet been designed to answer questions about non
specificity in children's memory, there are indications that early attachment
problems are seen in biases in memory in children as young as three years old.
See J. Belsky, B. Spritz and K. Cmic, ‘Infant Attachment Security and Affective-
cognitive Information Processing at Age 3’ , Psychological Science, 7 (1996),
pp. 111-14 .
10. W. Kuyken and C. R. Brewin, ‘Autobiographical Memory Functioning in
Depression and Reports o f Early Abuse’, Journal o f Abnormal Psychology, 104
(1995), pp. 585-91.
11. J. Evans, J. M. G. Williams, S. O’Loughlin and K. Howells, ‘Autobiographical
Memory and Problem Solving Strategies o f Parasuicide Patients’, Psychological
Medicine, 22 (1992), pp. 399 - 405 *
12. E. M. Marx, J. M. G. Williams and G. S. Claridge, ‘Depression and Social
Problem-solving’, Journal of Abnormal Psychology, 101 (1992), pp. 78-86.
13. G. L. Sidley, K. Whitaker, R. M. Calam and A. Wells, T h e Relationship
Between Problem Solving and Autobiographical Memory in Parasuicide
Patients’, Behavioural and Cognitive Psychotherapy, 25 (1997), pp. 195-202.
14. L. R. Pollock andj. M. G. Williams, ‘Effective Problem Solving Depends on
Specific Autobiographical Recall’, Suicide and Life Threatening Behaviour (in
press).
15. For the work that led to these conclusions, see J. M. G. Williams, N.
Ellis, C. Tyers, H. Healy, G. Rose and A. K. MacLeod, T h e Specificity of
Autobiographical Memory and Imageability o f the Future’, Memory and Cog
nition, 24 (1996), pp. 116-25.
16. C. B. Traux and R. R. Carkhuff, ‘Concreteness: a Neglected Variable in
Research in Psychotherapy’, unpublished MS, Psychotherapy Research Pro
gram, Universities o f Kentucky and Wisconsin, 1967. For a similar observation,
made in studies o f young mothers who were having problems in relationships
with their children, see R. G. Wahler and A. D. Afton, ‘Attentional Processes
in Insular and Non-insular Mothers: Some Differences in Their Summary
Reports about Child Problem Behaviours’, Behaviour Therapy, 2 (1980), pp. 25-41.
17. A. D. Brittlebank, J. Scott, J. M. G. Williams and I. N. Ferrier, ‘Autobiographi
cal Memory in Depression; State or Trait Marker?’ , British Journal o f Psychiatry,
162 (1993), pp. 118-21.
18. A. G. Harvey, R. A. Bryant and S. T. Dang (1998), ‘Autobiographical Memory
in Acute Stress Disorder’, Journal of Consulting and Clinical Psychology, 66,
pp. 500-506.
238
Notes and References
239
Suicide and Attempted Suicide
240
Notes and References
241
Suicide and Attempted Suicide
13 Final Thoughts
1. P. Levi, The Drowned and the Saved, Michael Joseph, London, 1988.
2. D. M. Shepherd and B. M. Barraclough, ‘The Aftermath o f Parental Suicide
for Children’, British Journal o f Psychiatry, 129 (1976), pp. 267-76.
3. N. L. Farberow, ‘Adult Survivors after Suicide: Research Problems and
Needs’, in A. A. Leenaars (ed.), Life Span Perspectives o f Suicide, Plenum Press,
New York, 1991.
4. Ajahn Sucitto, ‘Making Peace with Despair’, in Peace and Kindness, Amaravati
Publications, Hemel Hempstead, 1990.
242
Index
243
Index
244
Index
245
Index
246
Index
247
Index
France, 13, 14, 23, 24, 34, 71,190 Harvey, Allison, 172
Freud, Sigmund, 115-17, 118 Hawton, Keith, 34, 72, 81, 86, 131,
201-5
gambling, 56 Health Advisory Service, 1
gases and other vapours Hellon, C. P., 39
domestic gas detoxified, 25, 188 helplessness, 60, 61/92, 152, 169
and the state o f the brain at death, Helsinki, Finland, 71, 73
122 Hemlock Society, 99, 103
UK statistics, 25, 27 Hippocrates, 102, 120
US statistics, 30 Hippocratic oath, 102
gender, 28-9 Hollis, John, 34
and attempted suicide, 68, 69 homosexuality, 31
suicide method by, 23, 24, 27,31 hopelessness, xiv, 48, 60, 61, 86,
see also female suicide rates; male 90-91, no, 155, 167, 170-71, 173,
suicide rates 174, 182, 187, 195, 196-7, 199 , 217
genetic influences, 123-6 Hopelessness Scale, 91
Gentlemen's Companion, The humanism, 9, 10
(Ramesey), 12 Hume, David, 100-101, 102, 103
geographical mobility, 39 Humphry, Derek, xiii-xiv, 103, 106
Germany, 13, 23, 24, 34, 71 Hungary, 71
Gilbert, Professor Paul, 140-41, 143, Huntington’s chorea, 35
145-6, 149-50 hypnotics, 177, 188
Gilden, Charles, 100
Gilpin, John, 6 identical twins, 123-4
Goethe, Johann W olfgang von, 127 I f This Is a Man (Levi), 223
Goldney, Professor Robert D., 39, 82, Impact o f Events Scale, 166
109, 139-40 impulsive suicide, 52
Gotland study, 177-9 ,180 impulsivity, 122-3, 125-6, 182, 196
Gould, M., 134-5 insomnia, 48, 181
Greece, 24 internalization, 115
Green Card system, 214-15 interpersonal relationships
Greer, S., 191 and parasuicide, 86-7, 88-90, 138
Guide to Self-deliverance (Exit), 108-9 and suicidal behaviour, 196
Gunnell, D., 193 invalidation strategies, 96
guns see firearms involuntary subordination, 145-6
Guntrip, H., 118 Ireland, 23, 24
Italy, 13, 24, 34, 71
Hackney Hospital, 130
Hamilton Rating Scale, 172 Jack, Raymond, 84
hanging, i8, 25, 27,30,136, 137,151,18 8 Japan, 34, 24, 38,137
248
Index
249
Index
251
Index
252
Index
253
Index
254
Index
Putting forward the theory that suicide is usually a way of expressing one’s
otherwise inexpressible pain, Mark Williams looks at practical ways to help
those at risk; greater vigilance, more sensitive medical assessment and pre
ventative therapies. Informative and clearly written, this wide-ranging book
wilt help anyone seeking a deeper understanding of the enigma r \
(6)
PENGUIN
Get closer to Penguin at
Psychology/Health/Medicine
www.penguin.com Cover photo by Richard Hamilton / Millennium Images