ABCDEFG

UNIVERSITY OF OULU P .O. B 00 F I -90014 UNIVERSITY OF OULU FINLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

S E R I E S E D I T O R S

SCIENTIAE RERUM NATURALIUM

HUMANIORA

TECHNICA

MEDICA

SCIENTIAE RERUM SOCIALIUM

SCRIPTA ACADEMICA

OECONOMICA

EDITOR IN CHIEF

PUBLICATIONS EDITOR

Professor Esa Hohtola

University Lecturer Santeri Palviainen

Postdoctoral research fellow Sanna Taskila

Professor Olli Vuolteenaho

University Lecturer Veli-Matti Ulvinen

Director Sinikka Eskelinen

Professor Jari Juga

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0556-4 (Paperback)ISBN 978-952-62-0557-1 (PDF)ISSN 0355-3221 (Print)ISSN 1796-2234 (Online)

U N I V E R S I TAT I S O U L U E N S I S

MEDICA

ACTAD

D 1257

ACTA

Anniina Lahti

OULU 2014

D 1257

Anniina Lahti

EPIDEMIOLOGICAL STUDYON TRENDS AND CHARACTERISTICS OF SUICIDE AMONG CHILDREN AND ADOLESCENTS IN FINLAND

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU, FACULTY OF MEDICINE,INSTITUTE OF CLINICAL MEDICINE, DEPARTMENT OF PSYCHIATRY;OULU UNIVERSITY HOSPITAL, DEPARTMENT OF PSYCHIATRY;HELSINKI UNIVERSITY CENTRAL HOSPITAL,DEPARTMENT OF PSYCHIATRY, DIVISION OF ADOLESCENT PSYCHIATRY

A C T A U N I V E R S I T A T I S O U L U E N S I SD M e d i c a 1 2 5 7

ANNIINA LAHTI

EPIDEMIOLOGICAL STUDY ON TRENDS AND CHARACTERISTICS OF SUICIDE AMONG CHILDREN AND ADOLESCENTS IN FINLAND

Academic Dissertation to be presented with the assentof the Doctoral Training Committee of Health andBiosciences of the University of Oulu for public defencein Auditorium 1, Building PT1 of the Department ofPsychiatry (Peltolantie 17), on 10 October 2014, at 12noon

UNIVERSITY OF OULU, OULU 2014

Copyright © 2014Acta Univ. Oul. D 1257, 2014

Supervised byProfessor Pirkko RäsänenDocent Kaisa RialaDoctor Helinä Hakko

Reviewed byProfessor Eila LaukkanenDocent Markus Henriksson

ISBN 978-952-62-0556-4 (Paperback)ISBN 978-952-62-0557-1 (PDF)

ISSN 0355-3221 (Printed)ISSN 1796-2234 (Online)

Cover DesignRaimo Ahonen

JUVENES PRINTTAMPERE 2014

OpponentProfessor Jouko Lönnqvist

Lahti, Anniina, Epidemiological study on trends and characteristics of suicideamong children and adolescents in Finland.University of Oulu Graduate School; University of Oulu, Faculty of Medicine, Institute ofClinical Medicine, Department of Psychiatry; Oulu University Hospital, Department ofPsychiatry; Helsinki University Central Hospital, Department of Psychiatry, Division ofAdolescent PsychiatryActa Univ. Oul. D 1257, 2014University of Oulu, P.O. Box 8000, FI-90014 University of Oulu, Finland

Abstract

Finnish youth suicide mortality is exceptionally high in international comparisons. This studyinvestigated the epidemiology of child and adolescent (< 18 years) suicides in Finland, with aspecial focus on gender differences.

Two data sets were employed. Data from the national Finnish Cause of Death Register wasused to examine trends in the rates and methods of child and adolescent suicides in Finland in1969–2012. The characteristics of 58 child and adolescent suicides, which occurred in theprovince of Oulu in Northern Finland between 1988 and 2012, were explored based on individual-level data extracted from documents pertaining to establishment of the cause of death inmedicolegal autopsy investigations. The primary source of data was death certificates. Othersources included documents such as police investigation reports and the results of toxicologicalinvestigations. In addition, the suicide data was linked with the Finnish Hospital DischargeRegister.

After 1990, suicide rates generally decreased for males, but increased for females. Amongfemales, hanging exceeded poisoning as the most common suicide method after 1990, whereasfirearms were the most common method among males, until traffic suicides took the leadingposition for both genders during 2008–2012. Violent suicide methods of all types increased amongfemales after 1990. Male firearm suicide rates were nearly three times higher in Northern Finlandthan in Southern Finland, while there was no regional difference in rates of suicide by othermethods.

15% of male and 17% of female suicide victims in the province of Oulu had a history ofprevious psychiatric hospitalization. The discharge diagnoses were heterogeneous. Previoussuicidality and self-cutting were more common among females than males. Half of all adolescentswere under the influence of alcohol at the time of their death. The majority of the intoxicatedadolescents committed suicide at night, during descending blood alcohol concentrations. Anotable fall peak was observed in male firearm suicides.

The increase in violent, i.e. more lethal, suicide methods among females is alarming, asfemales are known to have high rates of attempted suicide. The special characteristics of malefirearm suicides in Northern Finland suggest the need for region-specific preventive measures.The high frequency of acute alcohol use in child and adolescent suicides warrants attention.

Keywords: adolescent, alcohol, epidemiologic studies, firearms, gender, mentaldisorders, seasonality, suicide

Lahti, Anniina, Epidemiologinen tutkimus lasten ja nuorten itsemurhienerityispiirteistä Suomessa. Oulun yliopiston tutkijakoulu; Oulun yliopisto, Lääketieteellinen tiedekunta, Kliinisenlääketieteen laitos, Psykiatria; Oulun yliopistollinen sairaala, Psykiatrian klinikka; HYKS,Psykiatrian tulosyksikkö, Nuorisopsykiatrian klinikkaryhmäActa Univ. Oul. D 1257, 2014Oulun yliopisto, PL 8000, 90014 Oulun yliopisto

Tiivistelmä

Suomalaisnuorten itsemurhakuolleisuus on kansainvälisessä vertailussa korkea. Tässä epidemio-logisessa tutkimuksessa selvitettiin alle 18-vuotiaiden suomalaislasten ja -nuorten itsemurhienerityispiirteitä.

Tutkimuksessa käytettiin kahta aineistoa. Kansallisen kuolemansyyrekisterin avulla selvitet-tiin alaikäisten itsemurhakuolleisuutta ja itsemurhien tekotapojen muutoksia Suomessa vuosina1969–2012. Empiirisen aineiston asiakirjat koskivat entisen Oulun läänin alueella vuosina1988–2012 itsemurhan tehneiden 58 alaikäisen oikeuslääketieteellistä kuolemansyyn selvittä-mistä. Aineiston perustiedot koottiin kuolintodistuksista. Lisäksi tietoja poimittiin muista kuole-mansyyn selvittämisasiakirjoista, kuten poliisin tutkintapöytäkirjoista ja oikeuskemiallisista tut-kimustuloksista. Aineistoon liitettiin myös hoitoilmoitusrekisteritietoja.

1990-luvun alusta lähtien poikien itsemurhat ovat vähentyneet ja tyttöjen lisääntyneet. Tytöil-lä hirttäytyminen ohitti myrkyttäytymisen yleisimpänä itsemurhan tekotapana vuoden 1990 jäl-keen. Pojilla ampuminen oli yleisin itsemurhamenetelmä koko tutkimusjakson ajan, kunnes lii-kenneitsemurhat nousivat yleisimmäksi tekotavaksi molemmilla sukupuolilla vuosina2008–2012. Väkivaltaisten itsemurhan tekotapojen käyttö on lisääntynyt tytöillä vuodesta 1990lähtien. Poikien ampumisitsemurhakuolleisuus oli Pohjois-Suomessa lähes kolminkertainenmuuhun Suomeen verrattuna. Vastaavaa alueellista eroa ei ollut muissa tekotavoissa.

Oulun läänin alueella itsemurhan tehneistä pojista 15 % ja tytöistä 17 % oli ollut psykiatri-sessa osastohoidossa ennen itsemurhaa. Diagnoosit olivat heterogeenisiä. Aiempi itsetuhoisuusja viiltely oli yleisempää tytöillä kuin pojilla. Puolet alaikäisistä oli ollut humalassa itsemurha-hetkellä. Suurin osa päihtyneenä tehdyistä itsemurhista tapahtui laskuhumalassa yöaikaan. Poiki-en ampumisitsemurhat olivat yleisempiä syksyllä kuin muina vuodenaikoina.

Väkivaltaisten eli tappavampien itsemurhamenetelmien lisääntyminen tytöillä on huolestutta-vaa, koska tyttöjen itsemurhayritysten tiedetään olevan yleisiä. Itsemurhien ehkäisyssä tulisihuomioida alueelliset erityispiirteet, joita tässä tutkimuksessa havaittiin poikien ampumisitse-murhissa Pohjois-Suomessa. Päihtyneenä tehtyjen itsemurhien suuri osuus alaikäisillä korostaanuorten alkoholinkäytön vähentämisen merkitystä.

Asiasanat: alkoholi, aseet, epidemiologiset tutkimukset, itsemurha,mielenterveyshäiriöt, nuoret, sukupuoli, vuodenaikaisvaihtelu

7

Acknowledgements

This work was carried out at the Department of Psychiatry of the University of

Oulu and the Department of Psychiatry of Oulu University Hospital. I wish to

express my gratitude to Professor (emeritus) Matti Isohanni, Professor Juha

Veijola and Docent Outi Saarento for providing excellent facilities in which to

work and for their support during my work.

I am most grateful to my supervisors Professor Pirkko Räsänen, Docent Kaisa

Riala and Dr Helinä Hakko. I feel privileged to have been tutored and inspired by

this brilliant trio. Pirkko, I especially value your intelligence, inventiveness and

optimistic outlook. Kaisa, I always looked forward to the special monthly “Kaisa

days” when you visited Oulu and I had the opportunity to hear your clinical

viewpoints and enjoy your sense of proportion. Helinä, I truly appreciate your

helpfulness, patience and wisdom, and of course your expertise in statistics.

I would like to sincerely thank all of my co-authors for their valuable

contribution to the original publications. Professor (emeritus) Erkki Vuori is

gratefully acknowledged for sharing his expertise in forensic toxicology. In

addition, I would like to acknowledge the Department of Forensic Medicine of

the University of Oulu and, for their more recent contribution, the National

Institute for Health and Welfare for their kind collaboration in providing the study

data.

I greatly appreciate the pre-examiners of this thesis, Professor Eila

Laukkanen and Docent Markus Henriksson, for their advice and constructive

comments. The professional language experts are warmly thanked for revising the

English language of the original publications and the final form of this thesis.

M.A. Olli Romppanen is acknowledged for his technical assistance with the

thesis.

I owe my thanks to Ms. Pirkko Kaan, Ms. Niina Keränen, Ms. Anja

Kylmänen, Mr. Jani Moisala and other staff at the Department of Psychiatry of

Oulu University Hospital and the Department of Psychiatry of the University of

Oulu for their help during the work. The Research & Development Unit (TUKE)

is acknowledged for its financial and practical support for this study.

I wish to express my loving thanks to all of my friends for just being there

and for encouraging me throughout the years. Finally, I reserve my deepest

gratitude for all of my family: my parents and siblings and their families. I cannot

thank you enough for all the support and unconditional love you have given me

throughout my life.

8

Financial support received for this work from the Department of Psychiatry

of Oulu University Hospital, The Alma and K. A. Snellman Foundation, The Emil

Aaltonen Foundation, The Finnish Cultural Foundation, The Finnish Medical

Foundation, The Finnish Psychiatric Research Foundation, The Jalmari and

Rauha Ahokas Foundation, The Otto A. Malm Foundation, The Päivikki and

Sakari Sohlberg Foundation, and The Yrjö Jahnsson Foundation is gratefully

acknowledged.

Oulu, August 2014 Anniina Lahti

9

Abbreviations

APC Annual percent change

BAC Blood alcohol concentration

BMI Body mass index

CI Confidence interval

FHDR Finnish Hospital Discharge Register

ICD-8 International Statistical Classification of Diseases, Injuries and

Causes of Death, 8th edition

ICD-9 International Statistical Classification of Diseases, Injuries and

Causes of Death, 9th edition

ICD-10 International Statistical Classification of Diseases, Injuries and

Causes of Death, 10th edition

SD Standard deviation

UAC Urine alcohol concentration

WHO World Health Organization

10

11

List of original publications

This thesis is based on the following publications, which are referred throughout

the text by their Roman numerals:

I Lahti A, Räsänen P, Riala K, Keränen S & Hakko H (2011) Youth suicide trends in Finland, 1969–2008. Journal of Child Psychology and Psychiatry 52: 984–991.

II Lahti A, Keränen S, Hakko H, Riala K & Räsänen P (2014) Northern excess in adolescent male firearm suicides: A register-based regional study from Finland, 1972–2009. European Child & Adolescent Psychiatry 23: 45–52.

III Lahti A, Harju A, Hakko H, Riala K & Räsänen P (2014) Suicide in children and young adolescents: A 25-year database on suicides from Northern Finland. Journal of Psychiatric Research 58: 123–128.

IV Lahti A, Räsänen P, Karvonen K, Särkioja T, Meyer-Rochow VB & Hakko H (2006) Autumn peak in shooting suicides of children and adolescents from Northern Finland. Neuropsychobiology 54: 140–146.

Some unpublished data are also presented in this thesis.

12

13

Contents

Abstract

Tiivistelmä

Acknowledgements 7 

Abbreviations 9 

List of original publications 11 

Contents 13 

1  Introduction 15 

2  Review of the literature 17 

2.1  Terms and definitions .............................................................................. 17 

2.1.1  Suicide .......................................................................................... 17 

2.1.2  Suicidal and nonsuicidal self-injury, self-harm and

suicidality ..................................................................................... 18 

2.1.3  Childhood and adolescence .......................................................... 19 

2.2  Suicide research methodology ................................................................ 20 

2.2.1  Macro approach ............................................................................ 21 

2.2.2  Micro approach ............................................................................. 21 

2.3  Epidemiology of child and adolescent suicides ...................................... 22 

2.3.1  General characteristics .................................................................. 22 

2.3.2  Suicide rates and secular trends .................................................... 23 

2.3.3  Methods of suicide ....................................................................... 25 

2.4  Risk factors for child and adolescent suicides ........................................ 25 

2.4.1  Mental disorders, substance abuse and alcohol

intoxication ................................................................................... 26 

2.4.2  Previous suicide attempt and self-harm ........................................ 29 

2.4.3  Familial and psychosocial risk factors .......................................... 29 

2.4.4  Availability of lethal methods ....................................................... 30 

2.4.5  Suicide contagion and clusters ..................................................... 31 

2.5  Seasonal and other temporal patterns of suicide ..................................... 31 

2.6  Child and adolescent suicide prevention ................................................. 32 

2.6.1  Universal suicide prevention ........................................................ 33 

2.6.2  Selected suicide prevention .......................................................... 35 

2.6.3  Indicated suicide prevention ......................................................... 36 

3  Aims of the study 39 

14

4  Material and methods 41 

4.1  Suicide data ............................................................................................. 41 

4.1.1  Macro approach (I, II) .................................................................. 41 

4.1.2  Micro approach (III, IV) ............................................................... 45 

4.2  Statistical methods .................................................................................. 49 

4.3  Ethical considerations and personal involvement ................................... 51 

5  Results 53 

5.1  Macro approach (I, II) ............................................................................. 53 

5.1.1  Child and adolescent suicide trends in Finland (I) ....................... 53 

5.1.2  Child and adolescent male firearm suicides (II) ........................... 55 

5.2  Micro approach (III, IV).......................................................................... 56 

5.2.1  Demographic features of child and adolescent suicides

(III) ............................................................................................... 56 

5.2.2  Suicide methods (III) .................................................................... 57 

5.2.3  Mental health (III) ........................................................................ 57 

5.2.4  Circumstances of the suicides (III, IV) ......................................... 60 

5.2.5  Toxicology findings (III) .............................................................. 63 

5.2.6  Somatic findings from the medicolegal autopsies (III) ................ 66 

6  Discussion 67 

6.1  Overview of the results ........................................................................... 67 

6.2  Discussion of the findings ....................................................................... 68 

6.2.1  Trends in child and adolescent suicides (I) ................................... 68 

6.2.2  Regional differences in child and adolescent male firearm

suicides (II) ................................................................................... 70 

6.2.3  Characteristics of child and adolescent suicides in the

province of Oulu (III, IV) ............................................................. 72 

6.3  Strengths and limitations ......................................................................... 76 

6.3.1  Strengths of the study ................................................................... 76 

6.3.2  Limitations of the study ................................................................ 77 

7  Conclusions 79 

7.1  Main conclusions .................................................................................... 79 

7.2  Implications of the study ......................................................................... 80 

7.3  Implications for further research ............................................................. 81 

References 83 

Appendix 97 

Original publications 99 

15

1 Introduction

Suicide in a child or adolescent is a painful and devastating event that has a

widespread effect on families, friends, and the community. To many, the very idea

of young people dying by their own hands is incomprehensible. Yet, it is

estimated that global annual suicide fatalities among children and adolescents

may approach 100,000 (Greydanus & Calles 2007), making suicide one of the

major causes of adolescent mortality throughout the world (Patton et al. 2009).

Suicidal behavior in the young population has long been recognized as a

serious public health issue. Completed suicide in children and adolescents is the

tip of the iceberg considering the large continuum of suicidality. According to

some estimates, there are more than 100 suicide attempts to every completed

suicide among adolescents (Maris 2002). Approximately 2–10% of adolescents

report having attempted suicide, females more often than males (Bridge et al. 2006, Evans et al. 2005). Thoughts about death and suicide are common in

adolescence. Around 10% to 30% of adolescents have had suicidal thoughts at

some point in their lives (Evans et al. 2005, Nock et al. 2013).

Finnish adolescent suicide mortality is among the highest in Europe

(European Detailed Mortality Database 2014). In early 21st century Finland,

suicide has been the second leading cause of death among 15–19-year-old

adolescent males and the third for females of the same age (Statistics Finland

2014a). The United Nations Committee on the Rights of the Child has recently

expressed its concern about the high rate of depression and high number of

suicides among children in Finland, and about inadequate access to mental health

services (United Nations Committee on the Rights of the Child 2011). The

committee recommended the strengthening of mental health services for children

and the intensification of suicide prevention measures.

Gender is an important factor in child and adolescent suicides. Whereas the

rate of attempted suicides is higher among females, males predominate in

completed suicides, particularly in Western societies (Hawton et al. 2012b).

Further, specific psychiatric disorders, a history of previous suicidality, and the

distribution of different suicide methods among adolescent suicide victims have

all been shown to vary by gender (Brent et al. 1999, Grøholt et al. 1999,

Marttunen 1994, Shaffer et al. 1996). Despite the clinical importance of these

differences, not all child and adolescent suicide studies have reported their

findings for males and females separately, which may reduce the utilizability of

the results.

16

The ultimate aim of suicide research is to increase our knowledge of suicidal

behaviors in order to form a basis for effective preventive measures.

Epidemiological studies have an important contribution to make to the field of

suicidology, as they provide insights regarding the different determinants of

suicide. Accordingly, this study sought to investigate the epidemiology of child

and adolescent suicides in Finland, with a special focus on gender differences.

Finland’s reliable national registers and cause-of-death certification practices

form a good basis for performing such research.

17

2 Review of the literature

2.1 Terms and definitions

2.1.1 Suicide

The concept of suicide is much more complex than is conveyed by the words

“killing oneself”, which originate in the Latin terms “sui” (of oneself) and

“caedere” (to kill). In defining suicide, it has been emphasized that suicide is a

behavior – not a disease, disorder or diagnosis – and that all behavior is

multidimensional and determined by multiple factors (Silverman et al. 2007).

Numerous definitions of suicide have been proposed, a selection of which is

presented in Table 1.

Table 1. A selection of definitions of suicide.

Definition of suicide Source

The act of killing oneself deliberately initiated and performed by the person

concerned in the full knowledge, or expectation, of its fatal outcome.

World Health

Organization 1998

Self-inflicted death with evidence (either explicit or implicit) that the person

intended to die.

American Psychiatric

Association 2003

An act with a fatal outcome, which the deceased, knowing or expecting a

potentially fatal outcome, has initiated and carried out with the purpose of

bringing about wanted changes.

De Leo et al. 2006

Self-injurious behavior that resulted in a fatality and was associated with

at least some intent to die as a result of the act.

Posner et al. 2007

Death caused by self-directed injurious behavior with any intent to die

as a result of the behavior.

Crosby et al. 2011

Death, i.e. a fatal outcome, is the most easily agreed element between different

definitions. Other key aspects include the agency of the act (self-inflicted), the

intention to die or stop living, and consciousness/awareness of the outcome (De

Leo et al. 2006, Silverman 2006). With regard to the intention to die, because of

the generally ambivalent attitude towards living or dying in the case of an

individual contemplating suicide, post mortem determination of the presence and

degree of suicidal intent is often difficult (Silverman 2006). A central question

posed by De Leo et al. (2006) is, “how much more must the suicidal subject wish

for death rather than life, in order for their death to be classified as a suicide”.

Establishing intent and consciousness/awareness of the outcome can be

18

particularly problematic regarding suicide in children and young adolescents.

According to Mishara (1999), coroners and medical examiners may not classify a

pediatric death as suicide due to the commonly held view that children have an

insufficient understanding of the finality and irreversibility of death. However,

Mishara (1999) found that most 5–7-year-olds and almost all children aged 8–12

know that suicide results in death and understand that death is permanent and

final.

For the purposes of this thesis, suicide is defined as deaths that have been

officially classified as suicides by medical examiners, based on a medicolegal

death investigation practice.

2.1.2 Suicidal and nonsuicidal self-injury, self-harm and suicidality

Not only the concept of suicide itself, but the overall terminology, nomenclature

and classification systems of suicidology are complex and have been a topic of

considerable international debate (De Leo et al. 2006, Skegg 2005). From one

viewpoint, death is not always the aim of self-injurious behaviors, i.e. there is no

intent to die. The presence or absence of the intent to die can therefore be used as

a premise for dichotomizing self-injurious behaviors into suicidal (when there is

evidence of suicidal intent) and nonsuicidal (no evidence of suicidal intent)

behavior (Crosby et al. 2011, Kapur et al. 2013). On the other hand, “self-harm”

is a term used to encompass all self-injurious behaviors regardless of suicidal

intent (Hawton et al. 2012b, Skegg 2005). The reasoning behind the latter view is

that suicidal intent is fluid, it exists on a continuum rather than being a categorical

construct, and its evaluation is difficult and unreliable (Brunner et al. 2014,

Ougrin 2012). Furthermore, multiple motivations for self-harm may exist

simultaneously, and even the persons themselves may be equivocal as to the

underlying reasons of their self-harming behavior (Kapur et al. 2013).

The term “suicidality” is a broad concept that refers to all suicide-related

thoughts and behaviors, including suicidal ideation, suicide attempts and

completed suicides (Bridge et al. 2006). Suicidal ideation has been defined as

thoughts about suicide that may include the planning of suicide attempts

(Waldvogel et al. 2008), a wish to or threat to die (Pelkonen & Marttunen 2003),

or thoughts of harming or killing oneself (Bridge et al. 2006). According to a

definition provided by Crosby et al. (2011), a suicide attempt is a non-fatal self-

directed potentially injurious behavior, with any intent to die as a result of such

behavior. However, in common language “suicide attempt” is often used as an

19

umbrella term for any self-injurious behaviors regardless of intent, a view also

shared by some investigators (Kerkhof 2000).

Differences of opinion also exist as to the term “nonsuicidal self-injury”

(NSSI), defined as intentional self-inflicted damage to the surface of the body

without conscious suicidal intent (American Psychiatric Association 2013). Here,

the focus of controversy again rests on the question of whether a clear-cut

distinction can be made between “non-suicidal” and “suicidal” self-injury –

especially given the strong association that exists between NSSI and suicidal

behavior (Kapur et al. 2013). Nevertheless, it is of importance in terms of this

thesis that the most common method of NSSI is cutting or carving oneself with a

sharp implement (Nock 2010). In the research literature, “self-mutilation” and

“self-cutting” are some of the terms used to refer to such behavior (although often

with no precise definition regarding the suicidal intent of the individual). In this

thesis, the term “self-cutting” will be used to describe the above-mentioned

behavior, regardless of the suicidal intent.

2.1.3 Childhood and adolescence

One dictionary definition of childhood refers to it as the period of life before

puberty (HarperCollins 2014). In the Convention on the Rights of the Child,

children are defined as persons up to the age of eighteen years (United Nations

1989). In legal terms, children are referred to as those under the age of majority,

the limit for which is 18 years in many countries, including Finland. The World

Health Organization (WHO) defines adolescence as the period in human growth

and development that occurs after childhood and before adulthood, from the ages

of 10 to 19 years (World Health Organization 1986). Some other definitions of

adolescence are not based solely on chronological age. For example, it has been

stated that adolescence begins with the onset of physiologically normal puberty,

and ends when an adult identity and behavior are accepted (Sacks et al. 2003).

As a consequence of the various definitions of childhood and adolescence,

studies of child and adolescent suicides tend to be marked by high variability in

age groupings. Numerous suicide research papers focus on “young people”

(individuals aged 15–24 years), thereby including adolescents in this broader age

group and excluding children and younger adolescents. Since the developmental

and sociodemographic characteristics of children and adolescents are in many

ways different from those of young adults, the main emphasis of this literature

20

review is on studies investigating suicides in children and adolescents under 20

years of age.

In this thesis, children and adolescents are defined as those under 18 years of

age. In addition to being the age of majority, this age limit is of importance in

child and adolescent mental health services and social services in Finland. The

Finnish Child Welfare Act (2007) refers to children as anyone under 18 years of

age. The Finnish Mental Health Act (1990) requires that patients under 18 years

of age – referred to as “minors” – in need of inpatient psychiatric treatment must

be treated separately from adults. Further, regarding involuntary commitment, the

conditions for ordering treatment are broader for minors than for adults (Mental

Health Act 1990). It is also worth noting that the Finnish law defines the time

limits within which patients must be given access to health care services, and

regarding access to mental health care the treatment of children and young people

(<23 years of age) must begin within three months, while the limit is six months

for adults (Health Care Act 2010).

2.2 Suicide research methodology

Suicide is a complex research subject. By definition, the first obstacle concerns

the unavailability of the people who it is most important to understand – the

deceased themselves (Berman et al. 2006). Another challenge is the relative rarity

(low base rate) of completed suicide amongst the population. This means that in

prospective studies, the generally preferred research strategy, very large sample

sizes would be required in order to obtain meaningful results (Nock 2009, Robins

& Kulbok 1986). In addition to clinical and methodological difficulties, potential

ethical restraints are of importance (Berman et al. 2006). It is therefore not

surprising that studies of adolescent suicidality have to a great extent focused on

living suicidal individuals, such as suicide attempters and individuals with a

history of self-harm. However, although adolescent suicide attempters and

completers share many characteristics, they also differ from each other in many

important ways (Moskos et al. 2004). Results obtained from studies on the former

group may not therefore directly apply to the latter.

As outlined by Berman et al. (2006), there are two main methodological

strategies for studying completed suicides. The “macro approach” makes use of

large samples of aggregate data to study the characteristics of population groups.

The “micro approach”, on the other hand, investigates individual suicide cases

with the aim of identifying more specific clinical characteristics and risk

21

variables. In both approaches, suicide data may be linked to various registers,

such as national health registers, in order to obtain additional relevant

information.

2.2.1 Macro approach

The macro approach, i.e. the ecological research method, is based on population

data rather than focusing on the characteristics of individual suicides (Stack

1987). Suicide counts and rates are compared between regions, genders, or age

groups, for example. Temporal trends in the suicide mortality of different

populations can also be examined. (Robins & Kulbok 1986)

2.2.2 Micro approach

In the micro approach, i.e. case study strategy, individuals serve as units of

analysis (Stack 1987). Sources of data include existing records, and/or interview

data (of relatives, physicians, etc.) generated for research purposes. Compared

with the macro approach, case studies involve the prospect of finding more

specific predictors for suicide. (Berman et al. 2006) Two main types of micro

approach strategies, i.e. psychological autopsies and record review studies,

account for most of our current knowledge of the clinical characteristics and risk

factors of adolescent suicides.

Psychological autopsies

In psychological autopsy studies, interviews of family members, other close

intimates and health professionals are used in conjunction with different kinds of

record data to reconstruct the social and psychological features of deaths by

suicide (Marttunen et al. 1993). This technique has been regarded as particularly

useful when studying suicides among young people, because many informants,

such as parents and other family members, friends, and teachers, are available in

most cases (Marttunen et al. 1993). Nevertheless, there are surprisingly few

psychological autopsy studies on child and adolescent suicides. Since the 1980s,

only around ten research groups have published results of psychological autopsies

conducted on children and adolescents under 20 years of age (Brent et al. 1999,

Eisele et al. 1987, Freuchen et al. 2012, Marttunen et al. 1991, Portzky et al. 2009, Renaud et al. 2008, Rich et al. 1990, Shaffer et al. 1996, Shafii et al. 1985).

22

Record review studies

Record review (case series) studies make use of the available record data. Such

information is typically extracted from suicide case files compiled by medical

examiners or other relevant authorities responsible for establishing the cause of

death. Case files usually include death certificates in combination with

investigatory records containing more detailed information (Berman et al. 2006).

2.3 Epidemiology of child and adolescent suicides

2.3.1 General characteristics

There is a great deal of international variability in adolescent suicide rates

(Wasserman et al. 2005). A universal characteristic of suicide epidemiology,

however, is the rarity of completed suicide before puberty and the subsequent

increase in frequency throughout adolescence (Gould et al. 2003). Explanations

proffered for the low rates of suicide in children and young adolescents as

compared to older adolescents include fewer risk factors, lower suicide intent and

cognitive immaturity in the planning and execution of a lethal act (Brent et al. 1999, Grøholt et al. 1998).

The male-female disparity in rates of completed suicides is another almost

universal pattern in adolescent suicide epidemiology, with higher rates among

adolescent males than females being identified throughout Western countries

(Wasserman et al. 2005). In contrast, rates of self-harm, suicide attempts, suicidal

ideation and self-cutting are generally more common among female than male

adolescents (Brunner et al. 2014, Laukkanen et al. 2009, Lewinsohn et al. 1996,

Lewinsohn et al. 2001, Moran et al. 2012, Wunderlich et al. 2001). This

discrepancy between males and females has been called the “gender paradox of

suicidal behavior” (Canetto & Sakinofsky 1998, Schrijvers et al. 2012). The

reasons for the phenomenon are by no means clear. Suggested explanations

include gender differences in accuracy and honesty in the reporting of suicidal

behavior (Canetto & Sakinofsky 1998, Lewinsohn et al. 2001), the influence of

cultural factors (suicide being regarded as a masculine and attempting suicide as

feminine behavior) (Canetto 1997), disparity in help-seeking behaviors (Beautrais

2002) and differences in methods of suicide (Beautrais 2003) and

psychopathology (Kaess et al. 2011, Schrijvers et al. 2012).

23

2.3.2 Suicide rates and secular trends

The age groupings most commonly utilized when reporting rates of child and

adolescent suicides are those of 10–14 years and 15–19 years, which is likely due

to the statistical practices of the World Health Organization. As regards the

younger group, worldwide suicide rates (per 100,000 population) among 10–14-

year-olds range between 0.0 and 8.5 for males and between 0.0 and 6.5 for

females (Dervic et al. 2008, Kolves & De Leo 2014). Recent reports on secular

trends in suicide rates in this age group are scarce, and where published data

exists, the findings are inconsistent. Decreasing suicide trends in 10–14-year-old

males in Austria between 1970–2001 (Dervic et al. 2006) and a sporadic increase

among females in Canada from the 1980s until 2008 (Skinner & McFaull 2012)

have been reported. At the same time, no significant changes in suicide rates

occurred among Austrian females (Dervic et al. 2006), Canadian males (Skinner

& McFaull 2012), or among either gender in England and Wales in the first

decade of the 21st century (Windfuhr et al. 2013).

Table 2 presents the rates of suicide (per 100,000 population) among 15–19-

year-old adolescents by gender in selected European countries. Eastern European

countries, Finland and Ireland lead the rates among male adolescents, and the

situation is largely similar among females.

24

Table 2. Suicide death rate (per 100,000 population) among 15–19-year-old males and

females in selected European countries. (European Detailed Mortality Database 2014)

Males Females

Country and year 1 Suicide death rate Country and year 1 Suicide death rate

Russian Federation 2010 24.7 Russian Federation 2010 7.2

Belarus 2009 21.3 Lithuania 2010 6.2

Lithuania 2010 20.2 Belarus 2009 6.1

Finland 2011 20.2 Norway 2012 5.7

Ukraine 2012 18.1 Finland 2011 5.6

Ireland 2010 17.6 Austria 2011 4.6

Latvia 2012 14.8 Ukraine 2012 4.0

Poland 2011 13.4 Sweden 2010 3.9

Czech Republic 2012 10.4 Belgium 2010 3.8

Sweden 2010 9.6 Ireland 2010 3.7

Estonia 2012 8.8 Switzerland 2010 3.6

Hungary 2012 8.4 Netherlands 2011 2.9

Belgium 2010 8.4 Poland 2011 2.4

Iceland 2009 8.2 Denmark 2011 2.3

Norway 2012 7.7 France 2010 2.3

France 2010 7.2 Germany 2012 2.3

Austria 2011 7.1 United Kingdom 2010 2.1

Switzerland 2010 6.9 Greece 2011 1.8

Germany 2012 6.6 Hungary 2012 1.8

Netherlands 2011 5.1 Czech Republic 2012 1.6

Denmark 2011 4.9 Portugal 2011 1.5

United Kingdom 2010 4.4 Spain 2011 1.1

Spain 2011 2.9 Italy 2010 0.8

Italy 2010 2.8 Estonia 2012 0.0

Greece 2011 2.1 Iceland 2009 0.0

Portugal 2011 1.4 Latvia 2012 0.0 1 Latest year available.

With regard to changes over time, an upward trend in suicides among males aged

15–19 years was reported in many countries between the 1960s and 1990s (Rutz

& Wasserman 2004, Wasserman et al. 2005). More recently, declining trends in

male adolescent suicides have been reported in the United States (US) (Bridge et al. 2008), Canada (Skinner & McFaull 2012) and the United Kingdom (UK)

(Windfuhr et al. 2008, Windfuhr et al. 2013), for example. Opposing trends have

also emerged, for example in South Korea (Park et al. 2014). Among 15–19-year-

old females, suicide rates have generally remained more stable in many countries

(Madge 1999, Rutz & Wasserman 2004, Wasserman et al. 2005). In recent times,

25

it has been reported that adolescent female suicide rates are stable in the UK

(Windfuhr et al. 2013), and decreasing in the US (Bridge et al. 2008), but are

increasing in Canada (Skinner & McFaull 2012). To conclude, secular trends in

adolescent suicides have been diverse between different countries and genders.

2.3.3 Methods of suicide

As well as suicide rates and trends, the distribution of suicide methods among

adolescents varies between nations and genders. Use of firearms is the leading

method of suicide among 15–19-year-old males in Finland (Statistics Finland

2014a) and the US (Lubell et al. 2007), as well as in Switzerland for all <20-year-

old males (Hepp et al. 2012). Firearms have also been the most common suicide

method among 15–19-year-old males in Canada, but hanging surpassed firearms

at the beginning of the 1990s (Skinner & McFaull 2012). At the same time, in

Canada the distribution of suicide methods also changed among adolescent

females, with hanging overtaking poisoning as the leading method of suicide

(Skinner & McFaull 2012). In a similar manner, in the US at the beginning of the

2000s hanging surpassed firearms as the most common suicide method among

adolescent females (Lubell et al. 2007).

In contrast to the use of firearms among adolescent males, hanging is the

most common method of suicide among 15–19-year-old females in Finland

(Statistics Finland 2014a), while traffic (railway suicides) is the most common

method among Swiss female adolescents (Hepp et al. 2012). In Italy, hanging is

the leading method among males, whereas jumping from a high place accounts

for most female adolescent suicides (Pompili et al. 2009). Otherwise, recent

reports on the distribution of suicide methods among European adolescents (<20

years) are scarce. However, a study of 15 European countries examined suicide

rates in young people (15–24 years) and found that hanging was the most

common method of suicide among males in 12 countries and among females in

all but one country (Värnik et al. 2009).

2.4 Risk factors for child and adolescent suicides

Suicide is a highly complex and multidetermined outcome. No single cause can

be shown to underlie any individual suicide, and no single path leads to this end.

Nevertheless, research from several previous decades has consistently identified

26

factors that are associated with an increased likelihood of completed suicide in

adolescence – that is, risk factors for adolescent suicides.

2.4.1 Mental disorders, substance abuse and alcohol intoxication

Psychological autopsy studies have identified at least one psychiatric disorder in

approximately 90% of adolescent suicide victims (Brent et al. 1993b, Marttunen

et al. 1991, Portzky et al. 2005, Renaud et al. 2008, Shaffer et al. 1996).

However, Brent et al. (1999) found that only 60% of under 16-year-old suicide

victims had a diagnosable psychiatric disorder, and similar results have also been

found in other studies (Grøholt et al. 1998, Shaffer et al. 1996).

In psychological autopsy studies, mood disorder has been the most prevalent

mental disorder among adolescent suicide victims (Brent et al. 1993b, Grøholt et al. 1999, Renaud et al. 2008, Shaffer et al. 1996, Shafii et al. 1988). Gender

difference is a consistent finding across studies: more female (approximately

70%) than male (approximately 40–60%) adolescent suicide victims have

suffered from a mood disorder (Brent et al. 1999, Marttunen et al. 1991, Shaffer

et al. 1996). Substance abuse or dependence (alcohol or drugs) is another

important risk factor (Brent et al. 1999, Marttunen et al. 1991, Portzky et al. 2005, Renaud et al. 2008, Shaffer et al. 1996). As regards gender differences, one

study from the US (Shaffer et al. 1996) found higher rates of substance abuse

among male adolescents (42% in males vs. 12% in females), whereas another

(Brent et al. 1999) reported fairly similar rates (35% vs. 24%) between genders.

In a Finnish study (Marttunen et al. 1991), the rates of alcohol abuse/dependence

were higher among females (23% vs. 44%), although note should be taken of the

small number of adolescent females (n=9) in that study. Findings regarding age

differences in terms of the prevalence of substance abuse among adolescent

suicide victims are more consistent, the rates being significantly higher among

older adolescents (Brent et al. 1999, Grøholt et al. 1998, Shaffer et al. 1996).

Substance abuse is often comorbid with mood disorders (Brent et al. 1993b,

Shaffer et al. 1996), and psychiatric comorbidity is common in adolescent

suicides in general, as up to 70% of adolescent suicide victims have had more

than one psychiatric disorder (Bridge et al. 2006). Apart from mood disorders and

substance abuse, disruptive disorders (such as oppositional disorder) are among

the mental disorders that have been shown to increase the risk of adolescent

suicide, especially among males (Brent et al. 1999, Renaud et al. 2008, Shaffer et al. 1996).

27

Whereas alcohol abuse, i.e. chronic alcohol use has consistently been shown

to be a risk factor in adolescent suicide, empirical research on the role of acute

alcohol use in increasing the risk of suicide among adolescents is scarce. On a

general level, it has been suggested that acute alcohol consumption is a potent risk

factor for completed suicides (Lamis & Malone 2012). Proposed mechanisms

include increased psychological distress (particularly, it is suggested, during

descending blood alcohol concentration, which is often associated with depressed

and sedative-like effects (Holdstock & de Wit 1998)), increased aggressiveness,

and the impairment of cognitive processes (Hufford 2001). In relation to cognitive

functioning, Hufford (2001) argues that alcohol use impedes alternative coping

strategies and narrows the range of options that the individual perceives as being

available. It is also well-known that alcohol impairs inhibitory control (Field et al. 2010) and that there is a strong association between impulsivity and alcohol use

(Lamis & Malone 2012).

Table 3 presents results from previous studies that included data, based on the

results of forensic toxicology examinations, on adolescent suicide victims who

were under the influence of alcohol at the time of their death. Where information

is available, the results are shown by gender. The overall figures for acute alcohol

use among adolescent suicide victims fall between 10% and 50%, between 20%

and 50% for males and between 10% and 40% for females, although an

exceptionally high proportion (56%) of intoxicated female adolescent suicide

victims was reported in a Finnish study by Marttunen et al. (1992). In addition to

the information provided in Table 3, a study from Belgium (Portzky et al. 2005)

reported that 37% of adolescent suicide victims were known, based on informant

interviews, to have consumed alcohol during the day of their death (no results of

toxicology examinations were available). Regarding age differences, studies have

shown that being under the influence of alcohol at the time of death is rare among

<15-year-old adolescents (Brent et al. 1999, Freuchen et al. 2012, Grøholt et al. 1998).

28

Ta

ble

3. A

do

les

cen

t v

icti

ms

un

de

r th

e in

flu

en

ce o

f a

lco

ho

l a

t th

e t

ime

of

de

ath

, b

ase

d o

n f

ore

ns

ic t

ox

ico

log

y e

xam

inati

on

s.

Stu

dy

Countr

y, t

ime s

pa

n

Ag

e g

rou

p

(ye

ars

)

To

tal n

(m

ale

s %

) A

lcohol t

oxi

colo

gy

scre

ens,

%

Under

the in

fluence

of alc

ohol (

posi

tive

case

s am

on

g those

test

ed)

To

tal,

%

Ma

les,

%

Fe

ma

les,

%

Eis

ele

et al.

19

87

U

S,

19

83

–1

98

4

13

–1

9

25

(7

6%

) 8

8%

1

8%

n

ot

rep

ort

ed

n

ot

rep

ort

ed

Ho

be

rma

n &

Ga

rfin

kel 1

98

8

US

, 1

97

5–

19

85

<

20

2

29

(8

0%

) n

ot

rep

ort

ed

2

8%

1

no

t re

po

rte

d 2

n

ot

rep

ort

ed

2

Ald

rid

ge

& S

t Jo

hn

19

91

C

an

ad

a,

19

77

–1

98

8

10

–1

9

63

(8

4%

) n

ot

rep

ort

ed

2

1%

n

ot

rep

ort

ed

n

ot

rep

ort

ed

Ma

rttu

ne

n e

t al.

19

92

F

inla

nd

, 1

98

7–

19

88

1

3–

19

5

3 (

83

%)

94

%

51

%

50

%

56

%

Grø

ho

lt et al.

19

97

, 1

99

9

No

rwa

y, 1

99

0–

19

92

8

–1

9

12

9 (

77

%)

79

%

45

%

ap

pr.

45

%

ap

pr.

40

%

Bre

nt

et al.

19

99

U

S,

19

84

–1

99

4

13

–1

9

14

0 (

85

%)

no

t re

po

rte

d

29

% 3

3

2%

1

5%

Wein

berg

er

et

al.

20

01

U

S,

19

96

–1

99

7

11

–1

6

46

(7

4%

) 1

00

%

11

%

no

t re

po

rte

d

no

t re

po

rte

d

Johanss

on e

t al.

20

05

S

we

de

n,

19

81

–2

00

0

13

–1

9

88

(7

5%

) a

pp

r. 9

7%

3

5%

a

pp

r. 4

0%

a

pp

r. 2

0%

Shie

lds

et al.

20

06

U

S,

19

93

–2

00

2

11

–1

7

10

8 (

89

%)

94

%

17

%

no

t re

po

rte

d

no

t re

po

rte

d

Sin

gh

& L

ath

rop

20

08

U

S,

19

79

–2

00

5

9–

17

4

33

(7

9%

) n

ot

rep

ort

ed

2

0%

2

3%

3

9%

3

Ka

rch

et al.

2013

U

S,

2005–2008

10–17

1046 (

not

report

ed

) appr.

75

%

ap

pr.

10

%

no

t re

po

rte

d

no

t re

po

rte

d

1 E

viden

ce o

f pro

bable

alc

oh

ol c

on

sum

ptio

n (

base

d o

n o

bse

rvers

’ report

s) in

anoth

er

17%

. 2 M

ale

s and fem

ale

s equally

like

ly to

have

been u

nder

the in

fluence

of alc

ohol;

figure

s n

ot pro

vided.

3 F

igure

s ca

lcula

ted b

ase

d o

n in

form

atio

n in

the a

rtic

le t

ext

.

29

2.4.2 Previous suicide attempt and self-harm

A history involving a prior suicide attempt has consistently been shown to be one

of the strongest risk factors in adolescent completed suicide (Brent et al. 1999,

Grøholt et al. 1997, Shaffer et al. 1996). More female (36–67%) than male (12–

37%) adolescent suicide victims have previously attempted suicide (Brent et al. 1999, Grøholt et al. 1999, Hoberman & Garfinkel 1988, Karch et al. 2013,

Marttunen et al. 1992, Marttunen et al. 1995, Shaffer et al. 1996, Soor et al. 2012,

Weinberger et al. 2001), which is similar to the situation amongst the general

adolescent population (Evans et al. 2005, Hultén et al. 2001, Lewinsohn et al. 2001). It has been suggested that a previous suicide attempt increases the risk of

suicide among male adolescents in particular, the predictive value being less

important in females (American Academy of Child and Adolescent Psychiatry

2001, Gould et al. 2003), but this interpretation does not seem justifiable when

we examine original data from case-control studies that employ the psychological

autopsy method (Brent et al. 1999, Shaffer et al. 1996).

Self-cutting in adolescents is often regarded as low-risk behavior. However, a

study from the United Kingdom explored the long-term outcomes of self-harm

among <19-year-old children and adolescents and found that self-cutting was

strongly associated with the risk of suicide. Importantly, the study showed that

self-cutting conveyed a greater risk of suicide than self-poisoning. (Hawton et al. 2012a)

2.4.3 Familial and psychosocial risk factors

Parental psychopathology (most commonly, depression and substance use) and a

family history of suicidal behavior are risk factors in adolescent suicide (Brent et al. 1996, Gould et al. 1996). Case-control studies have shown that parental

divorce/separation is more common in the families of suicide victims than in

those of controls, but the association decreases when parental psychopathology is

accounted for (Gould et al. 2003).

Both in the longer term and shortly before a suicide, stressful life events, such

as interpersonal losses and conflicts, discord, and legal and disciplinary problems

are more common among adolescent suicide victims than among controls (Bridge

et al. 2006, Gould et al. 2003). In the latter case, the stressors are referred to as

precipitating factors or precipitants. Marttunen et al. (1993) described a

precipitant as a stressor that occurred during the month preceding the suicide and

30

that is believed to have directly contributed to it. An argument with one’s parents

is more common among child and young adolescent suicide victims, while

arguments or end of a relationship with a boy/girlfriend is associated with

suicides among older adolescents (Brent et al. 1999, Grøholt et al. 1998). A

precipitating factor was reported in 45% of male and 50% of female adolescent

suicides in a Norwegian study (Grøholt et al. 1999) and in 70% and 64% of all

adolescent suicide cases in studies conducted in Finland and the US, respectively

(Marttunen et al. 1993, Rich et al. 1990). One study from the US reported that a

precipitating event occurred on the day of the suicide in the case of 33% of

suicide victims under 18 years old (Barber 2005).

2.4.4 Availability of lethal methods

Much of the literature on the association between adolescent suicides and the

availability of lethal methods focuses on the use of firearms and largely originates

in the US, where use of firearms is the leading method of suicide among

adolescent males (Lubell et al. 2007). Results of ecological studies have

suggested that areas with a higher availability of firearms have higher rates of

suicide (Birckmayer & Hemenway 2001, Miller & Hemenway 1999, Miller et al. 2002, Miller et al. 2007). Further, studies indicate that this association is

relatively more important in suicides among adolescents and young people than in

the adult population (Birckmayer & Hemenway 2001, Miller et al. 2002, Miller et al. 2007, Sloan et al. 1990). More robust empirical evidence concerning the

association between firearms availability and adolescent suicide can be found in

case-control studies (Miller & Hemenway 1999). Such studies have shown that

the presence of a firearm in the home is associated with a higher risk of

adolescent suicide (Brent et al. 1988, Brent et al. 1991, Brent et al. 1993c, Shah

et al. 2000). With regard to age differences, Brent et al. (1999) have suggested

that availability of lethal means may play a greater role among younger than older

adolescents. Furthermore, there is some evidence that a firearm in the home is

associated with a particularly high risk of suicide among adolescents with no

discernible psychopathology (Brent et al. 1999). The role of impulsivity has been

emphasized in this respect (Brent & Bridge 2003).

31

2.4.5 Suicide contagion and clusters

It is recognized that some suicides occur in clusters, supporting the view that

suicide can be contagious behavior (Niedzwiedz et al. 2014). There are two main

patterns of suicide clusters: point clusters (localized in space and time, therefore

also known as space-time clusters) and mass clusters (media related, localized in

time only) (Haw et al. 2013, Joiner 1999, Mesoudi 2009). Regarding mass

clusters, the evidence indicates that the number of suicides increases following

mass media reporting of suicide – so-called copycat behavior or the Werther

effect (Mesoudi 2009, Niederkrotenthaler et al. 2010). Mesoudi (2009) defines

point cluster as a temporary increase in the frequency of suicides within a small

community or institution, relative to both the baseline suicide rate before and after

the point cluster and the suicide rate in neighboring areas. According to a recent

systematic review study, the evidence suggests a greater risk of point suicide

clustering among adolescents compared with older age groups (Niedzwiedz et al. 2014). However, point clustering has been reported as being a factor in less than

5% of suicides among adolescents and young people (Gould et al. 1990, Jones et al. 2013b).

The internet’s role in adolescent suicide contagion is currently a subject of

interest (Olson 2013, Robertson et al. 2012), but little research has been done on

the topic. A systematic review concluded that internet use may have both positive

effects, such as functioning as a support network, and negative effects, such as

normalizing self-harm and increasing suicidal ideation and hopelessness, on

young people who are at risk of self-harm or suicide (Daine et al. 2013).

2.5 Seasonal and other temporal patterns of suicide

Seasonal variation in suicide occurrence is a classic subject of interest in

psychiatric research. Conventionally, seasonality is defined as a “cyclical change

related to seasons of the year”, although nowadays the term is sometimes used to

refer to periodic fluctuations of any kind (Ajdacic-Gross et al. 2010a). Apart from

seasonal patterns, shorter-term temporal variations have been reported in

accordance with the day of the week and time of day (Law & De Leo 2013).

Seasonal variation in suicides among the general population, with peaks

during the spring and early summer, has been well documented in the medical

literature (Ajdacic-Gross et al. 2010a, Christodoulou et al. 2012). This general

pattern is known to vary between subpopulations according to the suicide method,

32

gender, and age (Ajdacic-Gross et al. 2010a, Christodoulou et al. 2012).

However, seasonality of suicides among adolescents is a poorly studied area, the

current, inconsistent findings being based on just a few published reports. Studies

from the US (Hoberman & Garfinkel 1988), Canada (Aldridge & St John 1991)

and Turkey (Uzun et al. 2009) identified no seasonality in <20-year-old

adolescent suicides. Another study from Canada (Thompson 1987) revealed a fall

peak among <21-year-old males. In the US, a September peak has been

demonstrated among 11–17-year-old children and adolescents (Shields et al. 2006) and fall and winter peaks among <16-year-old males (McCleary et al. 1991). Dervic et al. (2006) identified discrete suicide peaks in the spring and fall,

i.e. a bimodal seasonal trend, for <15-year-old Austrian children and adolescents.

In Norway, 63% of <16-year-old males and 75% of females of the same age who

committed suicide, did so “during the dark season”, i.e. between September and

February (Freuchen et al. 2012).

With regard to shorter-term fluctuations in the occurrence of suicide, an

excess of suicides on Mondays amongst the general population has been reported

(Maldonado & Kraus 1991, Miller et al. 2012), although an increased risk over

the weekend days was observed in a Finnish study (Partonen et al. 2004).

Findings from adolescent suicides have rarely been reported. In one study, a peak

on Saturday (30% of all suicides) was observed (Aldridge & St John 1991).

Another study found that 44% of adolescent suicide-related deaths occurred

between Friday and Sunday (Johansson et al. 2005) – a figure very close to what

would be expected if the suicides were distributed evenly throughout the days of

the week. 55% of suicides among <16-year-old Norwegian females were

committed on Monday, and only 17% of male suicides occurred between Friday

and Sunday (Freuchen et al. 2012). In terms of the time of day, a peak in suicides

during the late morning hours has been demonstrated amongst the general

population (Gallerani et al. 1996, Preti & Miotto 2001), whereas the peak for

younger people (14–24 years) occurred in the late afternoon (Preti & Miotto

2001). Similarly, a preponderance of suicides in the afternoon or evening has been

reported among <20-year-old (Hoberman & Garfinkel 1988) and <16-year-old

(Freuchen et al. 2012) suicide victims.

2.6 Child and adolescent suicide prevention

It is a widely accepted view that, in order to be effective, suicide prevention

efforts need to be comprehensive, multimodal and integrated (Lönnqvist et al.

33

2011, Mann et al. 2005, Schwartz-Lifshitz et al. 2012). Suicide prevention

strategies can be conceptualized using a model based on Universal, Selected and

Indicated preventive approaches (the USI model) (Dumon & Portzky 2013a).

While universal strategies target the entire population (or large subgroups such as

adolescents), selective strategies address subgroups of the total population that

have an above average risk of suicide. Indicated prevention strategies target

individuals who are at specifically high risk of suicidal behaviors, such as those

with suicidal thoughts. (Dumon & Portzky 2013a, Nordentoft 2011, U.S.

Department of Health and Human Services (HHS) Office of the Surgeon General

and National Action Alliance for Suicide Prevention 2012)

The low base rate of suicides entails methodological problems in

demonstrating the efficacy of various suicide prevention measures, since very

large studies with long follow-up periods would be needed in order to conduct

research with completed suicide as an outcome (Robinson et al. 2011). Due to

these practical considerations, not to mention ethical ones, in this field it is

practically impossible to perform randomized controlled trials, with ecological or

individually based naturalistic studies being carried out instead (Nordentoft

2011). Robust scientific evidence on the efficacy of most suicide prevention

approaches is therefore largely lacking. The evidence used in suicide prevention

may be imperfect from the scientific perspective, but given the significance of

suicide as a serious public health concern preventive acts and recommendations

must rely on the best available knowledge (Campo 2009, Nordentoft 2011).

In the following review of suicide prevention approaches, the emphasis is on

the strategies considered most relevant to adolescents.

2.6.1 Universal suicide prevention

School-based programs

Compared to suicide prevention measures amongst the general population, a

specific dimension of child and adolescent suicide prevention lies in the fact that

most young people attend school. A school setting therefore seems a natural

standpoint from which to target virtually all individuals in this age group.

Suicide awareness programs number among the most commonly applied

adolescent suicide prevention programs (Portzky & van Heeringen 2006). Among

the main goals of these programs are enhancing help-seeking behavior and

34

providing adolescents with the information they need to identify at-risk peers,

followed by confiding in responsible adults (since suicidal adolescents are much

less likely to talk to adults than to peers in times of distress) (Berman et al. 2006,

Brown et al. 2007, Gould et al. 2003). However, no firm evidence exists on the

positive impact of these programs on reducing suicidal behavior (van der Feltz-

Cornelis et al. 2011). As reviewed by Gould et al. (2003), Mann et al. (2005) and

Pelkonen and Marttunen (2003), whereas positive effects have sometimes been

observed, indifferent and even unfavorable effects have been reported, such as

negative reactions among suicidal adolescents and a decreased likelihood of

recommending mental health evaluations to a suicidal peer. It is also

acknowledged that changes in behavior are not necessarily brought about by

changes in knowledge and attitudes (Berman et al. 2006). The emphasis has

therefore been shifted toward other school-based strategies, including skills

training programs focusing e.g. on the problem-solving and cognitive skills of the

adolescents in question (Bridge et al. 2006, Gould et al. 2003). Although no direct

causal relationship can be established between skills training programs and

reduced suicide rates, improvements in other factors such as distress coping skills

are regarded as being of importance in general.

Media education

Since considerable evidence exists of suicide contagion, a plausible intervention

option would involve influencing media reporting on suicides. Accordingly,

media guidelines have been published (Dumon & Portzky 2013b, World Health

Organization 2008), including descriptions of factors that should be avoided (such

as portraying suicides in a glamorizing or romanticizing manner, or disclosing the

method and location of suicide) and suggestions for enhancing the positive role

played by the media in suicide prevention efforts (such as providing information

on the help available and stressing the multifactorial nature of suicidal acts).

Method restriction

Restricting access to lethal means of suicide is a key and evidence-based suicide

prevention strategy at the population level (Florentine & Crane 2010, Mann et al. 2005, Yip et al. 2012). The background rationale for this approach lies in the idea

that ambivalence about life and death is common among suicidal individuals, the

suicidal intent “waxes and wanes”, and periods of high risk are often of short

35

duration. Restricting access to suicide methods during this short period may

therefore have the potential to prevent suicides (Florentine & Crane 2010). At the

same time, it is acknowledged that such a suicide prevention approach does not

address the underlying causes of suicide (Hawton 2005).

The substitution hypothesis constitutes a potential dilemma in relation to the

method restriction approach. This hypothesis states that if access to one method

of suicide is restricted, another will be used instead (Florentine & Crane 2010).

However, there is some evidence that suicidal individuals have a preference for a

specific method of suicide and tend not to switch easily from one to the other

(Daigle 2005).

One of the most intensely studied areas of the method restriction strategy is

controlling access to firearms. Not only has it been shown that the availability of

firearms increases the risk of adolescent suicide by such means, but several

studies have also concluded that adolescent firearm suicide rates fall following

stricter regulation of such access (Beautrais et al. 2006, Cheung & Dewa 2005,

Lubin et al. 2010, Niederkrotenthaler et al. 2010, Webster et al. 2004). Examples

of other method restriction approaches shown to reduce suicide rates among the

general population include packing analgesics in blister packets and limiting

packet size, detoxification of domestic gas, restrictions on pesticides, and the

construction of barriers at jumping sites (Nordentoft 2011, van der Feltz-Cornelis

et al. 2011).

2.6.2 Selected suicide prevention

The aim of this approach is to prevent the onset of suicidal behaviors among

specific above-average risk subpopulations (Dumon & Portzky 2013a). Since

psychiatric disorder is a salient risk factor in adolescent suicides, those with

mental illnesses are the most important risk group with respect to prospective use

of the selected suicide prevention approach.

Screening and gatekeeper education

A screening strategy involves the use of self-report and individual interviews to

identify at-risk adolescents, with subsequent referral for treatment (Gould et al. 2006). A potential problem in this approach is the relatively high rate of false

positives. Although even non-suicidal adolescents would probably benefit from

interventions, the downside is the large burden imposed on limited mental health

36

resources (Horowitz et al. 2009). Suicide screening programs in schools are

popular in the US, yet a review study concluded that, given the weak evidence

base demonstrating the effect of this approach in reducing suicidal behaviors,

“efforts to use such programs should be regarded as investigational in nature”

(Peña & Caine 2006).

The rationale for gatekeeper education is that by educating adults (such as

school personnel) who meet adolescents on an everyday basis about suicide risk

factors, the identification of at-risk adolescents could be improved. In their

review, Gould et al. (2003) concluded that the research findings are encouraging,

pointing to improvements in e.g. knowledge, attitudes and referral practices

among school personnel, while the amount of research is limited.

Treatment of psychiatric disorders

In reducing suicidal behaviors, the effectiveness of treatment of adolescent

psychiatric disorders provides the fundamental grounds for high-risk suicide

prevention strategies. Because mood disorders are the most common diagnostic

risk factors involved in adolescent suicide, most research has focused on the

treatment of depression (Pelkonen & Marttunen 2003). A review of adolescent

depression treatment approaches would be beyond the scope of this literature

review, but it should be mentioned that psychotherapy and use of antidepressants

are commonly accepted evidence-based treatments (Thapar et al. 2010).

Regarding antidepressant treatment and suicidal behavior in children and

adolescents, during the first decade of the 21st century concerns emerged about

the possible adverse effect of antidepressants in provoking suicidal behavior

among adolescents (Hammad et al. 2006). Subsequently, “black box warnings”

were issued by the US Food and Drug Administration (FDA) (Leslie et al. 2005).

However, the subject remains controversial (Cooper et al. 2014, Isacsson & Rich

2014).

2.6.3 Indicated suicide prevention

The indicated suicide prevention approach focuses on the high-risk group of

adolescents who have already attempted suicide, or who are at specifically high

risk of suicidal behaviors, such as those with suicidal thoughts (Dumon & Portzky

2013a). As previously noted, a history of prior suicide attempts is one of the most

potent risk factors in terms of adolescent suicides. It should be noted, however,

37

that only a minority of adolescents who engage in self-harm and attempt suicide

seek treatment or present to hospitals (Bridge et al. 2006, Hawton et al. 2012b). A

large proportion of these high-risk individuals may therefore remain unknown to

health care professionals.

In cases of suicidal crisis, guidelines commonly recommend that adolescents

be referred to inpatient care (American Academy of Child and Adolescent

Psychiatry 2001). Although intensive treatment based on skilled observation and

support are offered in hospital settings, no empirical evidence exists on the

effectiveness of this treatment in reducing rates of completed suicides (Gould et al. 2003).

A recent review (Robinson et al. 2011) studied randomized, controlled trials

testing interventions in the case of adolescents and young adults who had attended

a clinical setting due to suicidal behavior or deliberate self-harm. The outcome

measures were suicidal behavior, ideation, and/or deliberate self-harm. The

review concluded that no differences were found between treatment and control

groups, except for one study in which favorable outcomes emerged in an

individual cognitive therapy group compared to treatment as usual (i.e. standard

treatment used in clinical practice) received by a control group. The authors of the

review remarked that the sample sizes were far too small to enable a study of the

effectiveness of treatments in preventing completed suicides.

38

39

3 Aims of the study

The purpose of this work was to investigate the epidemiology of child and

adolescent suicides in Finland, with a special focus on gender differences. The

specific aims were:

1. To examine national trends in rates and methods of child and adolescent

suicides in 1969–2012 in Finland (I).

2. To analyze the regional difference between Northern and Southern Finland in

rates of firearm suicides among male children and adolescents in 1972–2009

(II).

3. To explore the characteristics of child and adolescent suicides, including

demographic, psychosocial and situational factors, use of alcohol, previous

suicidality and the psychiatric inpatient treatment histories of the suicide

victims, and seasonal pattern of the suicides in the province of Oulu between

the years 1988–2012 (III, IV).

Hereafter, the Roman numerals I–IV refer to the original publications.

40

41

4 Material and methods

4.1 Suicide data

The data presented in this thesis are based on information from official documents

concerning deaths classified as suicides by medical examiners. Regarding death

certification practices, Finnish law requires that the police conduct an

investigation into the cause of death and request a medicolegal autopsy 1) when a

person’s death is due or is suspected of being due to a criminal offence, an

accident, suicide, intoxication, an occupational disease or medical treatment, 2)

when a person has died from an unknown disease or did not receive medical

treatment during their last illness, and 3) when the death was otherwise

unexpected (Act on the Inquest into the Cause of Death 1973).

As stated above, in Finland medicolegal autopsy is performed in every case

in which the actual or suspected cause of death was suicide. Medicolegal

autopsies are performed by official medical examiners who are forensic

pathologists. They make the final decision on the cause and manner of the death

and issue the death certificate once the information acquired from the autopsy is

complete and the necessary background information, such as medical history and

recent life events, has been gathered on the deceased. The data included in this

thesis refer only to deaths classified as suicides, whereas deaths classified as

being of undetermined cause did not qualify for the study.

4.1.1 Macro approach (I, II)

The data for the macro-level studies (I, II) were obtained from the Finnish Cause

of Death register, which is produced by Statistics Finland (Statistics Finland

2013). Statistics on the cause of death cover persons who have died in Finland or

abroad and were permanently resident in Finland at the time of their death. The

Cause of Death Register is based on data from death certificates and includes

information on the cause and manner of death, the date and place of death, and

the person’s exact birth date, personal identity code, age, gender and place of

residence. Since 1996, causes of death have been coded according to the 10th

revision of the International Statistical Classification of Diseases and Related

Health Problems (ICD-10). In 1969–1986, such statistics were compiled

according to ICD-8, and ICD-9 was used in 1987–1995.

42

For the purposes of this study, deaths classified as suicides were extracted

from the Causes of Death Register. The diagnostic codes used for suicides were

E950–959 of ICD-8 and ICD-9 in 1969–1995, and codes X60–X84 and Y87.0 of

ICD-10 since 1996. The classification of specific suicide methods in various

revisions of ICD is illustrated in Table 4.

Table 4. Classification of suicide methods in ICD-8, ICD-9 and ICD-10.

Method of suicide ICD-8

(used in 1969–1986)

ICD-9

(used in 1987–1995)

ICD-10

(used since 1996)

Poisoning E950 E950 X60–X65, X68–X69

Gassing E952 E951–E952 X66–X67

Hanging E953 E953 X70

Drowning E954 E954 X71

Use of firearms E955 E955 X72–X75

Jumping from a high place E957 E957 X80

Other methods E956, E958–E959 E956, E959 X76–X79, X81–X84, Y87.0

Traffic suicides not specified E959A–E959B X81–X82

Other/Unspecified E956, E958–E959 E956, E959 X76–X79, X83–X84, Y87.0

Year-end population estimates for specific age groups were also obtained from

Statistics Finland, in order to calculate age-standardized suicide mortality rates,

i.e. annual suicide rates per 100,000 population. An age limit of 12–17 years was

chosen for the adolescent population, because suicides among persons younger

than 12 years of age are extremely rare.

Adolescent suicide trends in Finland (I)

The data derived from the original publication I cover all suicides committed by

under 18-year-old children and adolescents (hereafter referred to as

“adolescents”) between 1st January 1969 and 31st December 2008 in Finland. In

this study, trends in overall and method-specific suicide rates among male and

female adolescents were investigated. The suicide methods presented in Table 4

were further classified into violent methods (use of firearms, hanging, jumping

from a high place, drowning), non-violent ones (poisoning and gassing) and

others.

At the time the data was purchased, Statistics Finland classified suicide

methods according to ICD-9 (from 1987 until 1997). Nowadays, the causes of

43

death classification follows the year limits used in the official classification of

diseases described in Table 4.

This thesis also includes unpublished results for updated suicide trends. The

updated data includes the years 2009–2012.

Adolescent male firearm suicides in Northern and Southern Finland (II)

Data derived from the original publication II cover all male adolescent suicides

committed with firearms in Northern and Southern Finland between 1st January

1972 and 31st December 2009. As reference groups, the data also includes firearm

suicides by males aged 18–24 years (hereafter referred to as “young adults”) and

25–44 years (“adults”). Rates of adolescent male firearm suicides in Northern

Finland and Southern Finland are compared. The study regions are illustrated in

Figure 1.

44

Fig. 1. Study regions: Northern and Southern Finland and the province of Oulu.

In ICD-8 and ICD-9, only one diagnostic code (E955) was used for firearm and

explosive suicides. In ICD-10, firearm suicides were classified as follows:

handgun discharge (X72); rifle, shotgun and larger firearm discharge (X73); other

and unspecified firearm discharge (X74); explosive material (X75). Since

information on specific types of firearm was not recorded before 1996, type-

specific firearm results can only be reported from that year onwards.

The characteristics of Northern and Southern Finland are presented in Table 1

of the original publication II. A major difference between the study regions lies in

the greater number of firearms per person in the north (50/100 persons in

45

Northern Finland vs. 27/100 persons in Southern Finland) (The Finnish Police,

unpublished data).

4.1.2 Micro approach (III, IV)

Data from the “Suicides in the province of Oulu” research project

Data on micro-level studies originate in the “Suicides in the province of Oulu”

research project, which was initiated in 1988. This includes all suicides

committed in the province of Oulu since 1st April 1988. Such data is collected by

the Department of Psychiatry, University of Oulu, in collaboration with the

competent authority in charge of forensic medicine. In the province of Oulu, this

competent authority was the Department of Forensic Medicine of Oulu University

until 2009 and the National Institute for Health and Welfare from 2010.

The province of Oulu, situated in Northern Finland, is shown in Figure 1. The

area will be referred to under that name in this thesis, although as of the beginning

of 2010 the Finnish provinces were abolished and the former area of the province

of Oulu is now administered by the Regional State Administrative Agency for

Northern Finland. The province of Oulu is a geographically extensive area

covering 19% of Finland’s land area, while the population accounted for 9%

(n=480 000) of the total population of Finland in 2014 (Statistics Finland 2014b).

The data used in the “Suicides in the province of Oulu” research project are

based on documentation related to establishing the cause of death during

medicolegal autopsy investigations. Related sources of information are illustrated

in Table 5. Primary source of data consists of death certificates in which suicide

has been classified as the manner of death by official medical examiners. Data

extracted from death certificates include age, gender, method of suicide, date and

place of death, contributory cause(s) of death (defined as reasons adversely

affecting the development of the condition leading to death and hence

contributing to it) (Statistics Finland 2013), and the presence of alcohol in the

blood at the time of death (use of alcohol coded as the contributory cause of

death). In addition to their structured data fields, the death certificates include a

blank text field for event description, i.e. a short case history written in the form

of free text (Lahti & Penttilä 2001). Such a case history usually includes

background information on issues such as the previous suicidality of the

deceased, or critical life events prior to suicide. The type of firearm used in

46

firearm suicides is always noted in the event description. Appendix 1 presents a

copy of the Finnish death certificate form.

Table 5. Sources of information for the “Suicides in the province of Oulu” research

project.

Information source Specific information source Original

publication

Documents pertaining to

establishing the cause of death

Primary source Death certificates III, IV

Other sources Autopsy referrals III

Police investigation reports

Findings of the external and internal examinations of

the medicolegal autopsies

Microscopic examinations

Toxicological examinations

Final medicolegal autopsy reports

Register linkage data Finnish Hospital Discharge Register III

Other sources of information are also presented in Table 5. Autopsy referrals are

written by primary care physicians and often include information on the

deceased’s prior state of health. Most police investigation reports include

information based on interviews with next-of-kin or others who knew the

deceased, and copies or mentions of possible suicide notes. Detailed external and

internal examinations of the body are conducted by official medical examiners,

who also take tissue samples for microscopic examination. Toxicological

examinations include, for example, the determination of alcohol and drug

concentrations in the blood and urine. All toxicological analyses are performed in

the Forensic Toxicology Laboratory at the Department of Forensic Medicine Hjelt

Institute of the University of Helsinki. Final medicolegal autopsy reports are

written by medical examiners based on all of the above information.

National health care register information was also used as a source of data. In

Finland, it is possible to make register linkages using the personal identity codes

with which every Finnish citizen is provided. The suicide data was linked to the

Finnish Hospital Discharge Register (FHDR, renamed as the Care Register for

Health Care in 1994), a nationwide health register covering all inpatient hospital

admissions at primary and specialized level care in Finland since 1969 (Sund

2012). The FHDR data contain primary diagnoses and comorbidities upon

47

discharge, as well as the length of stay and discharge dates for all treatment

episodes. Such data is provided by the National Institute for Health and Welfare.

Adolescent suicide data (III, IV)

The study sample in the original publication III comprised all 58 adolescents

below 18 years of age who committed suicide in the province of Oulu between 1st

April 1988 and 31st December 2012. Among the documents pertaining to

establishing the cause of death, death certificates were available for all subjects.

In terms of other documents, these were available for 54 (93%) adolescents and,

due to administrative obstacles, unavailable for four male subjects. Among the 54

suicide victims, the medical examiners’ suicide case files included documents of

all other kinds, except for autopsy referrals (or equivalent documents) which were

available for 56% of the subjects. In addition, while the content of toxicological

examinations varied, the blood alcohol concentration (BAC) was available for all

subjects.

For the purpose of this research, besides data derived from the death

certificates, information extracted primarily from the other documents included

blood and urine alcohol concentrations, the presence and concentrations of

substances other than alcohol in the blood and urine, the time of suicide, the

presence of a suicide note, the family circumstances and living arrangements of

the adolescents, precipitating factors (defined as a recent stressor believed to have

directly contributed to the suicide), the psychological and behavioral

characteristics of the adolescent, previous suicidality (expression of suicidal ideas

or a suicide attempt) and self-cutting. Self-cutting was defined as distinctive scars

detected by a medical examiner in an external examination of the body. A

previous suicide attempt was considered to be present if “suicide attempt” was

mentioned in any of the documents. As described in section 4.1.1., suicide

methods were categorized as violent or non-violent.

Blood alcohol concentrations are reported in line with the Finnish standards

on the issue, i.e. as mass of alcohol per mass of blood (1 mg/g, indicating 1 per

mil (‰)), in contrast to many other countries which use mass of alcohol per

volume of blood (e.g. 1 g/L) as the unit of measurement (Jones 2006). A cut-off

concentration of 0.20 g/L (i.e. 0.19 mg/g, or 0.19‰) was used when reporting

positive blood alcohol results (Jones et al. 2013a), due to the risk of small

amounts of alcohol being generated in the body after death.

48

In the original publication IV, the seasonal distribution of adolescent suicides

(n=42) was examined for the period 1988–2005. Adult suicides (≥18 years) were

used as a comparison group. Since the seasonality of suicide differs by method

and the use of firearms was the most common suicide method among the

adolescents studied, only the seasonality of firearm suicides was investigated.

Seasonal patterns among adolescent and adult firearm suicides were compared

with one another and with the uniform distribution of suicide cases over the year.

Additionally, the seasonal pattern of firearm suicides within a subgroup of young

adults aged 18–23 years was compared with that of the rest of the adult

population (≥24 years).

The distribution of suicides was investigated by both month and season. The

seasons were defined as follows: winter (November to January), spring (February

to April), summer (May to July) and fall (August to October). As such, they

follow the occurrence of solstices and equinoxes and are based on the amount of

light, which in Northern Finland peaks in June and troughs in December. Monthly

mean sunshine hours during the period 1971–2000 in the province of Oulu were

obtained from official publications by the Finnish Meteorological Institute (2004).

This thesis also includes unpublished results based on updated suicide

seasonality data related to adolescent suicides up to the year 2012. For male

adolescents, the updated results are presented as overall data, and separately for

firearm suicides and for suicides by all other methods. With respect to adolescent

females, due to the small number of cases the updated results are presented for

total data only.

Further, this thesis presents unpublished data based on the results of forensic

toxicology examinations, which include results related to alcohol concentrations

in the blood (BAC) and urine (UAC). The stage of alcohol metabolism in suicide

victims at the time of death, i.e. whether they committed suicide during the

ascending or descending phase of blood alcohol concentration, can be assessed by

calculating the ratio of UAC and BAC. The UAC/BAC ratio also serves as an

indicator of the time which elapsed between the point at which drinking ended

and death (Jones 2006).

As regards alcohol metabolism, alcohol is absorbed into the circulatory

system from the stomach and gut, whereby the BAC begins to increase. The

alcohol is then distributed throughout the body (Kugelberg & Jones 2007). The

UAC begins to increase, although at a lower rate than the BAC, when alcohol

passes into the kidneys from blood carried by the renal artery (Jones 2006).

During the initial 60–120 minutes after alcohol consumption, the UAC is less

49

than or equal to the BAC during the absorptive phase, as the BAC increases

(ascending phase of the blood alcohol concentration), and slightly higher than the

BAC during the plateau phase (an “intermediate period” denoting that alcohol is

still being absorbed, but no increase in BAC is occurring due to the elimination

process simultaneously occurring in the liver). In the post-absorptive phase, i.e.

approximately >120 minutes after drinking, as the BAC is decreasing due to

alcohol metabolism in the liver (the descending phase of the blood alcohol

concentration), the UAC is higher than the BAC due to the higher water content

of the urine (~100%) compared with that of the blood (~80%). (Jones 2006, Jones

& Kugelberg 2010) Following alcohol consumption, the UAC/BAC ratio

therefore changes as a function of time. A relatively low UAC/BAC ratio (<1.0)

suggests that the absorptive phase was under way at the time of death, indicating

that the individual had been drinking alcohol fairly recently, around one hour

before death. A UAC/BAC ratio between 1.0 and 1.2 suggests that the decedent

was in the plateau phase, while a ratio greater than 1.2 is indicative of the

complete absorption and distribution of alcohol (post-absorptive phase), implying

that most of the alcohol was consumed several hours before death. (Jones 2006,

Kugelberg & Jones 2007, Levine & Smialek 2000)

UAC/BAC ratios were calculated for all adolescent suicide victims who were

under the influence of alcohol at the time of their death, except for one subject

whose forensic toxicology results were not available (but use of alcohol was

recorded on the death certificate). The proportion of adolescents in the ascending

phase (UAC/BAC ratio ≤1.2) and in the descending phase (UAC/BAC ratio >1.2)

of blood alcohol concentrations at the time of suicide was determined. To

calculate the UAC/BAC ratios, BAC units of 1 mg/g were converted into 1 g/L to

enable use of the aforementioned cut-off value of 1.2, which is based on mass per

volume measurements (Levine & Smialek 2000). Since the water content of urine

is close to 100%, i.e. its density is almost 1.0, conversion of units is not required

for UAC values (Jones 2006).

4.2 Statistical methods

The statistical significances of the group differences in categorical variables were

assessed using Pearson’s chi-squared test or Fisher’s exact test, and in continuous

variables using Student’s t-test. In addition, the specific statistical analyses used

in the original publications are presented below.

50

Original publication I. For the analysis of suicide trends, a Joinpoint

regression model was employed in order to identify changes in suicide rates over

time. A grid-search model was used to fit the regression function to unknown

joinpoints. In addition, the annual percent change (APC) was used to characterize

trends in suicide rates over time and the model suggested by the Joinpoint

regression analysis (Kim et al. 2000) was used as the best fit for the adolescent

suicide rate data per 100,000 population. The Joinpoint regression Program,

version 3.4.2 (http://srab.cancer.gov/joinpoint/) served as the software. Owing to

the relatively small number of adolescent suicides, three-year moving averages

were used for statistical modeling.

In analyzing trends in specific suicide methods, trends from the year 1975

onwards were examined. This was because during the time of study completion

method-specific suicide data were unavailable for 1969–1974. Since it is well

established that the year 1990 marked a notable turning point in overall suicide

rates in Finland, i.e. the rising trend turned to a decrease (Hiltunen et al. 2009),

these method-specific trends were examined separately for the years 1975–1989

and 1990–2008. From 1987, the “other methods” were split between traffic

suicides and other/unspecified methods, since due to a change in the statistical

practices used within Statistics Finland, a separate code was assigned to traffic

suicides (E959; X81–X82). When calculating the trend in violent suicides for the

period 1990–2008, traffic suicides were included in violent suicide methods.

Original publication II. Suicide rates were visualized as 3-year moving

average rates. Differences in suicide rates between Northern and Southern

Finland were analyzed using the Independent Samples t-test. Regression analysis

was used to test the linear trend in suicides.

Original publication III. Because this study was descriptive in nature, no

sophisticated analyses were performed. The data were analyzed by gender and

age group (<15 years and 15–17 years), where such an approach was considered

relevant.

Original publication IV. Overall seasonality was assessed using the chi-

square test for multinomials and the ratio of the observed to the expected number

of suicides was calculated with 95% confidence intervals (Wonnacott &

Wonnacott 1990). Expected numbers of suicides were calculated based on the

uniform distribution of suicide cases over the year. Adjustment for equal month

length was performed and account was taken of the effect of leap years.

51

All statistical tests were two-tailed, and a limit for statistical significance was

set at p<0.05. The statistical software used in the analyses was SPSS Statistics

(IBM Corp. 2013).

4.3 Ethical considerations and personal involvement

The research plan for the “Suicides in the province of Oulu” project, of which the

micro approach section of the present research forms part, was approved by the

Ethics committee of the Faculty of Medicine, University of Oulu on 8th November

1999, and by the Ministry of Social Affairs and Health on 22nd May 2000 (Dnro

14/07/2000). The data used in this study are solely based on administrative

documents and information culled from Finnish national registers. The topic of

this doctoral thesis was approved by the Postgraduate Research Committee of the

Faculty of Medicine at the University of Oulu on 16th December 2008.

Suicide in children and adolescents is a particularly sensitive topic, and

careful consideration must be given to the publication and presentation of study

results related to this subject. The results of this study have therefore been

reported while maintaining the confidentiality of deceased children and

adolescents. Individual cases cannot be identified in any of the published results.

No family members or other people close to the suicide victims were contacted

and no interviews were performed at any stage of the research.

The author of this thesis has been participating as a researcher in the

“Suicides in the province of Oulu” study project since 2004. In addition, the

author has made a major contribution to all of the original publications, is named

as the first author in each of them, and was the corresponding author of

manuscripts I–III. She was involved in designing the study, collecting and

analyzing the data and reporting the results. Also, in consultation with a

professional statistician, she participated in the statistical analyses of all of the

original studies. She wrote the first draft of each manuscript and was responsible

for the final form of each paper as it was submitted.

52

53

5 Results

5.1 Macro approach (I, II)

5.1.1 Child and adolescent suicide trends in Finland (I)

In the original publication I, trends in overall and method-specific suicide rates

among Finnish male and female adolescents were explored. Over a period of 40

years between 1969 and 2008, there were 901 suicides among children and

adolescents under the age of 18 years in Finland. Of the suicide victims, 78%

(n=701) were male and 22% (n=200) female, giving a male-to-female ratio of

3.6:1.

Analysis based on the updated data showed that, over the period 1969–2012,

the number of adolescent suicides increased to 964, 77% of which (n=746) were

male and 23% (n=218) female suicides (unpublished data). In the additional

period 2009–2012, there were 45 (71%) male and 18 (29%) female suicides. The

male-to-female ratio for the years 2009–2012 was thus 2.5:1.

Overall trends in adolescent suicides

Figure 1 of the original publication I shows gender-specific adolescent suicide

trends between the years 1969 and 2008. Male adolescent suicide rates increased

significantly (p<0.001) in 1969–1989 and decreased after 1990 (p<0.001). Rates

among females were inconsistent during the earlier period, turning into a

significant increase (p<0.001) in the latter period.

Adolescent suicide trends were recalculated with updated data covering the

period up to the end of 2012 (unpublished data). As shown in Figure 2, the results

remained the same among adolescent males, except for the turn from a decreasing

trend towards an increase during the most recent years (change p=0.013). Female

adolescents also show similar results for the earlier years, but the rate has begun

to decrease slightly in more recent times (change p=0.007).

54

Fig. 2. Child and adolescent suicide trends in Finland in 1969–2012.

Trends in methods of suicide

Among male adolescents, use of firearms was the most common method of

suicide, as it was involved in 49% of all suicides in 1975–1989 and in 49% of all

suicides in 1990–2008 (I: Figure 2). The second most common suicide method

was hanging (35% in the earlier period and 26% in the latter period). The

proportion of all non-violent suicide methods among adolescent males was low

throughout the study period, as they accounted for 3% of all suicides in the earlier

period and 5% in the latter. As shown in Table 1 of the original publication I, no

significant changes occurred in rates of different methods of suicide among male

adolescents in the period 1975–1989. In 1990–2008, a significant decrease was

observed in suicides by the use of firearms, by poisoning and by all violent

methods. The decrease in suicides by all violent methods including traffic

suicides was likewise significant. There was no significant change in the rate of

all non-violent methods.

For female adolescents, the most common method of suicide was poisoning

(26%) in 1969–1989 and hanging (32%) in 1990–2008 (I: Figure 2). Hanging was

the second most common method (22%) in the earlier period, while other

55

methods (25%) were in the latter. Traffic suicides accounted for 96% of the other

methods. Among females, poisoning was the only non-violent method of suicide,

accounting for 26% of all suicides in the earlier period and 21% in the latter. As

presented in Table 2 in the original publication I, no significant changes in the

rates of any specific female suicide methods were observed during the period

1975–1989. In 1990–2008, a significant increase was observed in suicides by

“other methods”, especially traffic suicides. A significant increase was also

observed in violent suicides when traffic suicides were included. Rates of non-

violent suicides remained unchanged.

In a further analysis using updated data, the distribution of adolescent suicide

methods was examined for the most recent 5-year period (2008–2012)

(unpublished results). During this period, traffic suicides were the most common

method of suicide among both genders, as they accounted for 34% of all suicides

among males and for 35% among females. For males, the second and third most

common suicide methods were hanging (28%) and use of firearms (26%), and for

females hanging (27%) and jumping (19%). Non-violent methods accounted for

2% of all male and 12% of all female adolescent suicides.

5.1.2 Child and adolescent male firearm suicides (II)

In the original publication II, the regional difference in the rates of male

adolescent firearm suicides between Northern and Southern Finland was

examined. Of the total of 654 adolescent (<18 years) male suicides in Finland

between 1972 and 2009, 47% (n=305) were committed through the use of

firearms. In the reference groups of young adult (18–24 years) and adult (25–44

years) males, 36% (n=1,530) of the 4,252 young adult male suicides and 25%

(n=3,588) of the 14,674 adult male suicides were committed using firearms.

Regional differences in firearm suicide rates

Throughout the study period, 1972–2009, the mean rate of Finnish adolescent

male firearm suicides (per 100,000 population aged 12–17 years) was

significantly higher in Northern Finland than in Southern Finland (8.7 vs. 3.1,

p<0.001) (II: Figure 3a). The difference in rates between these regions was 2.8-

fold. A similar regional difference in firearm suicides was also observed among

young adult males, as the regional difference was 2.6-fold (33.8 vs. 13.2,

56

p>0.001) (II: Figure 3b) and, in adult males, firearm suicide rates involved a 1.9-

fold (22.0 vs. 11.4, p<0.001) difference between regions (II: Figure 3c).

Regarding the rates of suicide by other methods, no regional differences were

observed among adolescent males (4.8 vs. 4.3, p=0.176) or young adult males

(26.8 vs. 28.7, p=0.086). A significant difference was observed among adult

males (43.2 vs. 38.7, p=0.004).

Proportion of suicides involving firearms

Among adolescent males, the proportion of suicides involving firearms was

significantly higher in Northern Finland than Southern Finland (64% vs. 41%,

p<0.001) (II: Figure 4). A similar regional difference was also found among

young adult males (55% vs. 31%, p<0.001) and adult males (34% vs. 23%,

p<0.001). The highest proportion of suicides involving firearms was observed

among adolescent males in Northern Finland (64%) and the lowest among adult

males in Southern Finland (23%).

Types of firearms

The exact type of firearm involved was included in the diagnostic codes for

suicides from the year 1996, when ICD-10 was implemented in Finland. Type-

specific firearm results are therefore reported only for the years 1996–2009. A

shotgun or rifle (i.e. a hunting gun) was the most commonly used weapon in male

firearm suicides in all three age groups, both in Northern and Southern Finland

(II: Table 2). Among adolescent males, a hunting gun was involved in 88% of all

firearm suicides in Northern Finland, whereas the proportion was 67% in

Southern Finland (p=0.060). The regional difference was similar among young

adults (85% vs. 67%, p<0.001) and adult males (79% vs. 56%, p<0.001).

5.2 Micro approach (III, IV)

5.2.1 Demographic features of child and adolescent suicides (III)

The original publication III presented the characteristics of child and adolescent

suicides, with a special focus on gender differences. Of the total of 58 adolescent

suicide victims in the province of Oulu during 1988–2012, 79% (n=46) were male

57

and 21% (n=12) female. The male-to-female ratio was thus 3.8:1. Approximately

one fifth of both male (n=9) and female (n=2) suicide victims were under 15

years of age (III: Table 1) and the youngest suicide victim was 11 years of age. In

our sample, 17-year-old adolescents comprised almost half of all suicide victims,

the proportions being 43% (n=20) among males and 58% among females (n=7) in

(p=0.518).

With regard to the family circumstances and living arrangements of these

adolescents at the time of suicide, most had been living with at least one

biological parent (III: Table 1). Approximately one in ten of the males were

housed by child welfare services (institutions, relatives, foster family). Of the

adolescents living with at least one biological parent, at least half of the males and

two thirds of the females were living with both biological parents.

5.2.2 Suicide methods (III)

Use of firearms was the most common method of suicide in both genders, with

63% (n=29) of males and 33% (n=4) of females employing this method (III:

Table 1). Hanging was the second most common suicide method among the

males, accounting for 20% (n=9) of all male adolescent suicides. In addition, 9%

(n=4) of the males jumped from a height, 7% (n=3) committed suicide in traffic,

and 2% (n=1) committed suicide using exhaust gas. Of the adolescent females,

25% (n=3) committed suicide in traffic (railway suicides in all cases), 17% (n=2)

took a drug overdose, and hanging, a jump from a height and drowning were each

employed by one female. There was a significant difference in the distribution of

suicide methods between males and females (p=0.009).

In regard to violent and non-violent methods of suicide, with the exception of

one gassing suicide all male adolescent suicides were committed employing a

violent method (98%, n=45). Two females committed suicide by drug overdose,

meaning that the majority (83%, n=10) of adolescent females had used violent

suicide methods. No significant difference was observed in the choice between

violent and non-violent methods of suicide between the genders (p=0.106).

5.2.3 Mental health (III)

Table 6 summarizes the findings concerning the mental health status of adolescent

suicide victims. Evidence of mental health problems – defined as lifetime

inpatient hospital care due to a mental disorder or self-poisoning, a mental

58

disorder as a contributory cause of death in the death certificate, previous

suicidality (expression of suicidal ideas or a suicide attempt), self-cutting, or

nonspecific symptoms of mental health problems – was found in the majority of

adolescent suicide victims in both genders: 63% of males and 83% of females.

Table 6. Evidence of mental health problems among adolescent suicide victims, by

gender.

Evidence of mental health problems Males Females p-value1

n % n %

History of inpatient psychiatric hospitalization 7 15 2 17 1.000

History of inpatient somatic hospitalization due to

self-poisoning

– – 2 17 0.040

Mental disorder as a contributory cause of death

in the death certificate

4 9 3 25 0.147

Previous suicidality 14 30 9 75 0.020

Expression of suicidal ideas 13 28 7 58 0.086

Suicide attempt 2 4 3 25 0.055

Self-cutting 3 7 4 33 0.028

None of the above, but nonspecific symptoms of

mental health problems

7 15 1 8 1.000

Any of the above 29 63 10 83 0.302

No mental health problems by parental report 6 13 – – 0.328

Not reported 11 24 2 17 0.716 1 Fisher’s exact test, two-tailed significance

Lifetime inpatient psychiatric care

A history of inpatient psychiatric hospitalization was infrequent, with only seven

males and two females having been admitted to hospital during their lifetime. Of

these adolescents, five males and one female had only one treatment episode with

one discharge diagnosis. One male had three different discharge diagnoses within

one treatment episode and another had several long treatment episodes due to

autism, whereas one female had six discharge diagnoses within five inpatient

treatment periods. Self-poisoning was diagnosed in two males and one female.

Depression was diagnosed in one male and one female. Similarly, a substance

abuse diagnosis was given in the case of one male and one female. The remaining

discharge diagnoses were diverse and each occurred only once. Examples include

conduct disorder, adjustment disorder and eating disorder. Child psychiatric

diagnoses, such as autism and encopresis, were seen only among males.

59

The duration of previous psychiatric inpatient treatment periods varied from 1

day to 3 months (median 4 days, excluding one male subject with autism). Of

seven males with a history of psychiatric hospitalization, three had been

discharged within 6 months to 1 year before committing suicide, one within 1–2

years, and three more than two years before their suicide. One female subject was

on home leave from a psychiatric hospital at the time of her suicide, and another

had been discharged within 1–2 years before her suicide. Of the two females with

a history of admission to medical wards due to self-poisoning (i.e. a suicide

attempt), one had been discharged within 6 months and another within 1–2 years

before committing suicide. None of the adolescents with a history of lifetime

inpatient hospitalization due to mental disorder or self-poisoning were under the

age of 15 at the time of their deaths.

Mental disorder as a contributory cause of death

A psychiatric diagnosis, depression in all cases, was recorded on the death

certificates by the official medical examiners for four male and three female

suicides. Two males and two females had previously been hospitalized because of

a psychiatric disorder.

Previous suicidality and self-cutting

A significantly greater proportion of female than male adolescents (75% vs. 30%)

had shown evidence of previous suicidality, defined as the expression of suicidal

ideas or suicide attempts (Table 6). In a separate analysis of the expression of

suicidal ideas and previous suicide attempts, gender differences were shown to be

marginally significant. Significantly more of the female than male adolescents

had a history of self-cutting (33% of females vs. 7% of males).

Nonspecific symptoms of mental health problems

15% of the males and 8% of the females had shown nonspecific symptoms of

mental health problems, but did not have a history of inpatient hospitalization due

to a mental disorder or self-poisoning, previous suicidality, self-cutting, or a

mental disorder as a contributory cause of death. Information on nonspecific

symptoms of mental health problems was based on parental reports, which

described the recent behavioral and psychological characteristics of the

60

adolescent, for example signs of being depressed, withdrawn, nervous, or

uncommunicative. Evidence of mental health problems was also found in sources

such as the adolescents’ diary entries and suicide notes, revealing that they had

been contemplating suicide for a long period.

Other mental health related information

Information on the adolescents’ previous mental health treatment in outpatient

settings was rarely found among the documents establishing the cause of death. A

known history of outpatient psychiatric treatment was reported for only one male

and one female, and at the time of suicide, one female was currently receiving

such treatment.

Information concerning the mental health of the suicide victims was absent

for 24% of the males and 17% of the females. In 13% of the male suicides, the

documents stated explicitly that, according to parental reports (and in some cases

the primary care physician), the adolescent was physically and mentally healthy

and his behavior had been normal prior to the suicide. No precipitating events

were reported for these male adolescents. The method of suicide was firearm use

in all such cases, one of which involved a boy under 15 years of age.

5.2.4 Circumstances of the suicides (III, IV)

An adolescent’s home and its surroundings provided the most common location

for suicides among adolescents. A total of 59% (n=27) of the males and 50%

(n=6) of the females committed suicide in the home, in an outbuilding next to

their home, or outside in the near vicinity of their home (p=0.746). In terms of

suicide notes, 37% (n=17) of the males and 40% (n=5) of the females left a

written or recorded note (p=0.752). A precipitating event was reported in 24%

(n=11) of the males and 17% (n=2) of the females (p=0.716). The reported

precipitating events among males were an argument or the end of a relationship

with a girlfriend (45%, n=5), an argument with their parents (27%, n=3), an

argument with friends (9%, n=1), an argument with a girlfriend and friends (9%,

n=1) and shoplifting (9%, n=1). Among the females, the precipitating events were

an argument with friends in one case, and a recent abortion in another.

61

Seasonal pattern (IV) and time of suicides (III)

In the original publication IV, an analysis of seasonal variation was performed for

25 adolescent firearm suicides that had occurred in the province of Oulu in 1988–

2005. Adult (≥18 years) firearm suicides, n=554, committed in the province of

Oulu during the same time period were used as a comparison group. A significant

seasonal pattern (p<0.001) with a notable peak in the fall (ratio 2.70, 95% CI:

1.97–3.42) was found in adolescent firearm suicides (IV: Figure 1). 68% of these

firearm suicides occurred in the fall (August–October), 14% in the winter

(November–January), 21% in the spring (February–April) and none in the

summer (May–July). Among adults, however, firearm suicides peaked in the late

spring/early summer (p=0.05; ratio 1.19, 95% CI: 1.05–1.35). The seasonal

pattern of firearm suicides differed significantly between these two age groups

(p<0.001). In addition, the seasonal pattern of firearm suicides involving the

subgroup of young adults aged 18–23 years was similar to that of adult victims

(p=0.140).

By using updated data, this thesis presents unpublished results on adolescent

suicide seasonality by gender, up to the year 2012. Figure 3 shows the seasonal

distributions of adolescent male suicides within the updated data (1988–2012).

There was a significant fall peak (p=0.004; ratio 1.81, 95% CI 1.24–2.38) in all

male adolescent suicides, which was accounted for by the fall peak in firearm

suicides (p<0.001; ratio 2.60, 95% CI 1.91–3.29). In suicides by other methods,

there was a nonsignificant peak in the winter (p=0.188; ratio 1.87, 95% CI 0.93–

2.81). Very few male adolescent suicides were committed in the summer.

62

Fig. 3. Seasonal distribution of all suicides, firearm suicides and suicides by all other

methods among male adolescents in the province of Oulu, 1988–2012.

Due to the small number of female cases, the seasonal distribution of adolescent

female suicides was investigated for total data only. There was no discernible

seasonal pattern in the suicides of female adolescents, as 25% occurred in the

spring and 25% in the summer, 33% in the fall, and 17% in the winter (p=0.882).

In the original publication III, regarding the day of the week and time of day,

the exact time of death was known for 89% (n=41) of the male and 92% (n=11) of

the female adolescents (III: Figure 2). The variation in suicides in terms of days

of the week was rather small, with 48% of male suicides and 33% of female

suicides occurring on the weekends between Friday and Sunday. Among the

males, half of the suicides were committed during the day (between 06.00 and

18.00) and half at night (between 18.00 and 06.00) (III: Figure 2). Among the

females, one third of the suicides were committed in daytime and two thirds

during the night.

20%n=9

21%n=6

18%n=39%

n=4 0%n=0

24%n=4

46%n=21

66%n=19

47%n=2

26%n=12

14%n=4

47%n=8

0%

25%

50%

75%

100%

Adolescent males, all suicides(n=46)

Adolescent males, firearmsuicides (n=29)

Adolescent males, suicides  byother methods (n=17)

Seasonal distribution of suicides (%)

Spring Summer Fall Winter

63

5.2.5 Toxicology findings (III)

Alcohol

48% (n=22) of the males and 58% (n=7) of the females were under the influence

of alcohol at the time of their suicide (p=0.747). Among the males, 11% (n=1) of

those under 15 years of age and 57% (n=21) of those aged 15–17 years were

under the influence of alcohol at their time of death (p=0.023). The corresponding

figures for the females were 0% and 70% (n=7) (p=0.152).

Blood alcohol concentration (BAC) data were available for 95% (n=21) of

the males and 100% (n=7) of the females. For the males, the mean (±SD; min–

max) BAC value was 1.34‰ (±0.47; 0.45–2.1) and for the females it was 1.19 ‰

(±0.52; 0.52–2.0) (p=0.476). Urine alcohol concentration (UAC) data

(unpublished results) were available for all of these adolescents. The mean UAC

value for the males was 1.96‰ (±0.72; 0.38–3.1) and for the females it was

1.71‰ (±0.64; 1.20–2.8) (p=0.419).

Figure 4 presents the UAC/BAC ratios of 21 male and 7 female adolescents

who were under the influence of alcohol at the time of suicide (unpublished

results). The reference line is the UAC/BAC ratio of 1.2, which indicates the

beginning of the post-absorptive phase of the BAC, i.e. descending blood alcohol

concentration. A total of 90% (n=19) of the males and 86% (n=6) of the females

were in the post-absorptive phase (UAC/BAC ratio >1.2) at the time of their

deaths. The mean (±SD; min–max) UAC/BAC ratio was 1.37 (±0.19; 0.79–1.59)

for the males and 1.45 (±0.37; 1.03–2.18) for the females (p=0.497).

64

Fig. 4. The UAC/BAC ratio of adolescents who were under the influence of alcohol at

the time of their suicide, by gender.

Males who committed suicide during the night were significantly more often

under the influence of alcohol at their time of death than males who committed

suicide in the daytime (86% (18/21) vs. 15% (3/20), p<0.001) (III: Figure 2).

Among females, 86% (6/7) of those who committed suicide at night were under

the influence of alcohol, compared with 25% (1/4) of those who committed

suicide during the day, and a trend towards statistical significance was found

(p=0.088).

Substances other than alcohol

Toxicology screens for substances other than alcohol had been performed for 79%

(n=33) of the 42 males for whom information on toxicological investigations was

available and for 92% (n=11) of the females. Among them, 21% (7/33) of the

males and 36% (4/11) of the females tested positive for the presence of illegal or

65

prescription drugs. Among the females, a drug overdose was the method of

suicide in two such deaths. The specific substances found in the toxicology

screens are shown in Table 7. Benzodiazepines and benzodiazepine-like drugs

(diazepam, temazepam, unspecified benzodiazepine and zolpidem) and

antidepressants (mianserin, doxepin, mirtazapine and paroxetine) were the most

commonly detected prescription drugs. Antidepressants had been prescribed for

the two females who screened positive for such medication, but in other cases

there was no indication that the drugs had been prescribed for the adolescents.

Cannabis was found in one male. No positive results were found in cases where

subjects were tested for the presence of so-called hard drugs such as

amphetamine, cocaine or buprenorphine. Four males and three females were also

under the influence of alcohol at the time of death. None of the adolescents who

tested positive for the presence of illegal or prescription drugs were under 15

years of age at the time of their death.

Table 7. Substances other than alcohol detected in the blood or urine of adolescent

suicide victims.

Cases Substances in blood Substances in urine

Males

Alcohol present

Case 1 diazepam 1

Case 2 zolpidem 1

Case 3 benzodiazepine (unspecified)

Case 4 codein, salicylates

No alcohol present

Case 1 mianserin 1

Case 2 methyl tert-butyl ether 1 2 doxepin

Case 3 cannabis 1 3 cannabis 1 3

Females

Alcohol present

Case 1 mirtazapine, ibuprofen

Case 2 temazepam temazepam

Case 3 4 codein, paracetamol, ibuprofen codein, paracetamol, ibuprofen

No alcohol present

Case 1 4 paroxetine, levomepromazine 1 Recorded as contributing cause of death in death certificate. 2 Had drunk gasoline. 3 Also detected in hair. 4 Drug overdose as a method of suicide.

66

5.2.6 Somatic findings from the medicolegal autopsies (III)

Below are presented unpublished findings, based on data from external and

internal examinations and microscopic examinations conducted during the

medicolegal autopsies of adolescent suicide victims. In relation to the somatic

diseases observed during the autopsies, there were early signs of atherosclerosis

in 33% (n=15) of the males and in 25% (n=3) of the females, mild fatty

degeneration of the liver in three cases, one case of gastritis, one case of mild

pneumonia, and one case of thyroid cancer.

Information on the body mass index (BMI) was available for 89% (n=41) of

the males and 100% (n=12) of the females. The average BMI was 22.7 kg/m2 for

the males and 21.2 kg/m2 for the females. 13% (n=6) of the males and none of the

females were overweight or obese (BMI >25 kg/m2). 2% (n=1) of the males and

8% (n=1) of the females were underweight (BMI <18.5 kg/m2).

Of the three males with self-cutting scars, the scars were located on the thighs

in the case of two males and the thighs and arms in the case of one male. Among

females, the location of the scars was the arms for all four suicide victims.

67

6 Discussion

6.1 Overview of the results

Following the continuously increasing trend in male adolescent suicides in

Finland from the late 1960s onwards, rates began to decline after the early 1990s.

However, this favorable development has not continued in recent years, as slight

increases in suicide rates have been observed among male adolescents. Among

adolescent females, following earlier inconsistent trends suicide rates began to

increase in the early 1990s. This alarming upward trend was followed by a

slightly decreasing trend from the early 2000s.

From the 1970s, use of firearms was the most common suicide method

among Finnish male adolescents, until traffic suicides took the lead position in

recent years. Among females, poisoning used to be the most common suicide

method before 1990, but hanging surpassed poisoning after that time. In line with

the situation for males, traffic suicides have recently exceeded hanging to become

the most common method of suicide among female adolescents. On a more

general level, there has been a decrease in the use of all violent methods among

males, and an increase for females since the early 1990s.

Between the years 1972–2009, the adolescent male firearm suicide rate in

Northern Finland was almost three times higher than that observed in Southern

Finland, while there was no difference in rates of suicide by other methods. A

northern excess in firearm suicide rates was also found among young adult and

adult males. Hunting guns were the most common type of firearms employed in

the case of young male suicides, and their use was particularly common in

Northern Finland.

In a comprehensive sample of adolescent suicides that occurred in the

province of Oulu during a period of 25 years, only a minority of the suicide

victims had a previous history of psychiatric hospitalization. The discharge

diagnoses were heterogeneous. A higher proportion of females than males had a

known history of previous suicidality. Half of the adolescents were under the

influence of alcohol at the time of death. Of these, the vast majority committed

suicide during the night and during the descending phase of the blood alcohol

concentration. One fifth of adolescents screened positive for substances other than

alcohol. Violent methods of suicide predominated (males 98%, females 83%).

Use of firearms was the most common suicide method among both genders. A

68

significant seasonal pattern was discernible among male adolescent firearm

suicides, with a peak occurring in the fall.

6.2 Discussion of the findings

6.2.1 Trends in child and adolescent suicides (I)

The results of the study of trends in the suicide rates of Finnish adolescents aged

less than 18 years showed that male adolescent suicide rates increased from 1969

to 1989, while there were no consistent changes in the rates for adolescent

females during that period. The transition from the 1980s to the 1990s marked a

turning point in adolescent suicides, since from that time onwards the adolescent

male suicide rate began to decline, while the female suicide rate followed an

upward trend. It is worth noting that more recent data up to the year 2012 again

revealed shifts in suicide trends among both genders. The previous favorable

decline in male adolescent suicides gave way to a slight increase in the mid-

2000s, while the previous increase in female adolescent suicide rates has ceased

and a slightly declining trend has emerged since the early 2000s. However,

despite these trend changes, current male adolescent suicide rates are still

remarkably lower, and the corresponding rates in females higher, compared to the

rates in the early 1990s.

The general decreasing trend in suicide rates among adolescent males from

the 1990s is in line with the decline reported in the general Finnish male

population, whose suicide rate fell by half between 1990 and 2012. Whereas the

rate was as high as 50.1 per 100,000 male population in 1990, it had fallen to 23.6

by 2012 (Statistics Finland 2014a). As regards females in the general population,

a declining suicide trend has also been observed in their rates of suicide between

1990 and 2012 (from 12.1 to 7.5 per 100,000 population). Therefore, during the

1990s the rising suicide trend among adolescent females was the opposite to that

among Finland’s general female population.

It is remarkable that, despite the long history of suicide research, explanations

of changing trends in suicide rates have remained largely speculative. In Finland,

it has been suggested that decreasing suicide rates among the general Finnish

population are related to better depression treatments, with special regard being

paid to the increased use of SSRI antidepressants (Korkeila et al. 2007, Lönnqvist

2009, Partonen 2012). In addition, the multidimensional effects of the national

69

suicide prevention project in Finland, established in the late 1980s (Lönnqvist et al. 1993), and positive advances in society in a broader sense, have been proposed

as possible contributors to this decreasing trend (Hiltunen et al. 2009, Lönnqvist

2007). With regard to the use of adolescent outpatient psychiatric services in

Finland, the number of 13–22-year-old patients has increased by two thirds in the

2000s and it has been suggested that the previously hidden need for psychiatric

services has given way to active demand (Pylkkänen 2013). Also researchers

from various other countries have suggested that improvements in antidepressant

treatment and government suicide prevention strategies may underlie decreasing

suicide rates among young people (Bridge et al. 2005, McKeown et al. 2006,

Morrell et al. 2007). Other authors seeking explanations for changes in adolescent

suicide rates have focused on factors such as rates of alcohol consumption, access

to lethal means and the influence of the internet (Biddle et al. 2008, Brent et al. 1987, Bridge et al. 2008, McClure 2001). On the other hand, it has been

suggested that a combination of socio-economic and clinical factors accounts for

the changes (Windfuhr et al. 2008), which is consistent with the view that suicide

is a multidetermined event.

Better treatment of depression and the favorable societal changes that have

occurred in Finland may play a role in explaining the decrease in adolescent male

suicides since 1990, but the important question arises of why the rate for

adolescent females has followed the opposite trend at the same time. Notably, the

increase has specifically occurred in the rates (and proportion) of violent suicide

methods, whereas the rate of non-violent methods has remained unchanged.

Similarly to the present findings, a shift towards the use of more violent suicide

methods among young (20–24 years) females had been observed in Finland

before the 1990s (Öhberg et al. 1996), and more recently in several other

European countries (Värnik et al. 2009). It has been suggested that converging

gender roles, i.e. as more similar roles have begun to replace traditional

masculine/feminine roles, are contributing to this phenomenon (Grøholt et al. 1999, Pompili et al. 2009, Värnik et al. 2009). This is alarming, since females are

known to have higher rates of attempted suicide than males and the use of violent

suicide methods is more likely to end in a fatal outcome.

It is also of interest to compare the distributions of violent and non-violent

suicide methods among Finnish adolescents to those among the adult population

in Finland. Between 1998 and 2012, violent methods were employed in 76% of

male and 46% of female adult (≥25 years) suicides (Statistics Finland 2014a),

which illustrates the traditional gender difference in the choice of violent and non-

70

violent methods among adult suicide victims. As the results of the present study

show, the proportions of violent suicide methods among adolescents of both

genders are notably higher than the corresponding proportions for adults.

Correspondingly, Hepp et al. (2012) have reported that suicides by intoxication

are less frequent among Swiss adolescents than among adults. They suggest that

this may be due to adolescents having more difficulty accessing potentially lethal

drugs and lacking sufficient knowledge of the toxicity and lethal doses involved.

Apart from considerations of method availability, another possible explanation for

the large proportion of violent suicide methods among adolescents is related to

the sociocultural acceptability of different suicide methods, with norms and

traditions having an effect on the choice of suicide method (Cantor & Baume

1998).

Recent changes in the types of suicide methods among adolescents in Finland

also merit consideration. Use of firearms has been the predominant method for

Finnish adolescent males – accounting for half of all suicides – since at least the

1970s. The most recent suicide statistics from the years 2008–2012 showed,

however, that traffic suicides have overtaken the use of firearms and hanging

to become the leading suicide method among male adolescents. Similarly, traffic

suicides have assumed the leading position in suicide methods among adolescent

females and the shift towards the increasing use of this method had already begun

among females in the 1990s. Again, this suggests that sociocultural acceptability

may be of importance to the choice of method and that traditional patterns of

suicide among adolescents are subject to change.

6.2.2 Regional differences in child and adolescent male firearm suicides (II)

The results of this study demonstrated a marked regional difference in firearm

suicide rates among adolescent males, with significantly higher rates in Northern

Finland compared to Southern Finland. By contrast, rates of suicide by all other

methods did not show such a regional difference. Use of firearms evidently

accounts for a major part of the northern excess in suicide rates among Finnish

male adolescents.

In Finland, particularly in the northern areas of the country, hunting is a

traditional sport among males of all age groups, including young adolescents. The

prevalence of firearms in general, and hunting guns in particular, is substantially

higher in Northern Finland compared to Southern Finland (The Finnish Police,

71

unpublished data). The association between a greater number of firearms and

higher rate of firearm suicides within any specific region has been demonstrated

in several studies (Ajdacic-Gross et al. 2010b, Johnson et al. 2000, Miller et al. 2007). The results of the present study showed that a regional difference in rates

of firearm suicide was present among all <45-year-old Finnish males, but that the

difference in rates was more prominent among adolescent and young adult males

(2.8- and 2.6-fold differences, respectively) than in adult males (1.9-fold

difference). Earlier studies have reported that the association between greater

availability of firearms and higher rates of firearm suicides is more evident in

young people than adult populations (Birckmayer & Hemenway 2001, Sloan et al. 1990). It has been suggested that this is because young persons are more

impulsive. This impulsiveness when combined with easy access to a highly lethal

method then leads to greater susceptibility to suicide among younger people,

compared to older people (Miller et al. 2007).

With regard to suicide prevention, the evidence strongly suggests that means

restriction is among the most effective available policies (Mann et al. 2005, Yip et al. 2012). Correspondingly, stricter firearm control as a means of suicide

prevention has been emphasized in Finland (Wahlbeck 2011). Finland actually

implemented stricter firearms legislation back in 1998 and again in 2011

(Firearms Act 1998), and a legislative proposal has very recently been presented

to the Finnish Parliament proposing the further restriction of e.g. firearm storage

practices (HE 20/2014). Further, from 2011 physicians have been duty-bound to

notify the police of any persons they consider unfit to possess a firearm on the

basis of the person’s health or behavior, as stated in amendment 11.2.2011/124 of

the Firearms Act (1998). Physicians’ duty to notify the police is further specified

in the ongoing legislative proposal (HE 20/2014). Those opposed to means

restriction strategies argue that, according to the substitution hypothesis, if one

method of suicide becomes unavailable, a suicidal person will seek an alternative

method. However, it has been suggested that individuals have a preference for a

specific means of suicide, which makes method substitution unlikely (Daigle

2005). To conclude, the findings of this study underscore the importance of

further research on relations between firearm availability and suicide, and

adolescents should be included in these studies.

72

6.2.3 Characteristics of child and adolescent suicides in the province of Oulu (III, IV)

Various characteristics of adolescent suicides were scrutinized in order to broaden

the understanding of what may lie behind suicides among adolescents from the

province of Oulu in Northern Finland. The basic demographic characteristics of

adolescent suicide victims in this case series study are in line with previous

research, as the number of male suicides was almost four times that of females,

and suicide rates increased with age among both genders. Violent methods of

suicide were almost always used (in 98% of cases) by adolescent males, and also

predominated among females (83%); this is in accordance with the distribution of

violent and non-violent suicide methods among Finnish adolescents in the country

as a whole. The present study reveals that use of firearms was the most common

method of suicide among both genders. This contrasts with the method

distribution among adolescent females in the whole of Finland, since less than

10% of these suicides were committed using firearms during 1990–2008. As

discussed earlier, cultural factors may play a role in the choice of specific suicide

methods, which may also be of importance in explaining the present finding on

the relatively greater use of firearms among adolescent females in the province of

Oulu.

A minority of the suicide victims (15% of males and 17% of females) had a

lifetime history of psychiatric inpatient hospitalization, which coheres with

previous findings based on international studies. Brent et al. (1999) reported that

15% of adolescent suicide victims had received inpatient care, and other

researchers have reported figures as low as between 2–5% (Grøholt et al. 1997,

Portzky et al. 2005). The figures reported for lifetime mental health care contacts

among adolescent suicide victims have varied between 20 and 50% (Grøholt et al. 1999, Karch et al. 2013, Marttunen et al. 1992, Portzky et al. 2005, Shaffer et al. 1996). In the present study, explicit information on psychiatric outpatient

treatment was scarce. Only one male and two females were known to have

received such care in their lifetimes. However, the clinical impression which

emerged based on all of the available information suggested that the majority of

the adolescent suicide victims had never been in contact with mental health

services. Current psychiatric treatment (at the time of death of the adolescent) was

also rare in this study, again a finding consistent with previous studies (Brent et al. 1988, Grøholt et al. 1997, Rich et al. 1990, Shafii et al. 1988, Windfuhr et al. 2008).

73

The infrequency of psychiatric care among adolescent suicide victims may

have several explanations. Outward signs of mental health problems among

adolescents may not be sufficiently severe to be recognized by others, and the

adolescents may conceal or deny the extent of their problems. Further, even

where problems are recognized, there may be reluctance on the part of the family,

friends and the adolescents themselves to seek help from health care

professionals, possibly due to the stigma associated with mental illness (Moskos

et al. 2007). Young people’s reluctance to seek professional help for mental health

problems is a well-known phenomenon and applies to young males in particular

(Rickwood et al. 2007).

It is also possible that some of these adolescents had no need of psychiatric

care since mental health problems were not involved. Based on the parental

reports, 13% of the male adolescents covered in this study had been healthy,

normally behaving adolescents whose suicides were completely unexpected. This

finding seems consistent with the results of psychological autopsy studies, which

have reported that approximately 10% of adolescent suicide victims (usually

male) had no diagnosable psychiatric disorder (Brent et al. 1993a, Marttunen et al. 1991, Renaud et al. 2008, Shaffer et al. 1996). Suicides of this kind have often

been explained in terms of impulsiveness. However, Brent et al. (1993b) found

that adolescent suicide victims without an apparent psychiatric disorder were

more likely to have some psychiatric risk factors, such as familial psychiatric

disorders, compared to community controls. With regard to the present study, data

on the mental health characteristics of the adolescents in question lacked

sufficient detail for firm conclusions to be drawn in this respect.

In the present study, 25% of the females and just 4% of males were known to

have previously attempted suicide, or the reverse, with 75% of the females and

96% of the males having died in their first suicide attempt. Previous studies have

shown that a prior suicide attempt is a key risk factor in completed suicides

among adolescents of both genders (Brent et al. 1999, Grøholt et al. 1997, Shaffer

et al. 1996). However, in the current study a history of previous suicide attempt(s)

seemed to serve as a potential predictor of suicide among females, but not among

males.

The results of the present study showed that self-cutting scars were much

more prevalent in adolescent female than male suicide victims (33% vs. 7%).

Compared with these results, the life-time prevalence of self-cutting amongst the

general Finnish adolescent population aged between 13 and 18 years is lower

(17%) for females and the same (7%) among males (Laukkanen et al. 2009).

74

From a predictive viewpoint, self-cutting may therefore be an indicator of a

potential suicide risk among female, but not male, adolescents. On the other hand,

Laukkanen et al. (2013) have shown that adolescents who cut themselves on parts

of the body other than just the arms present with psychiatric problems of greater

severity, such as being more likely to have suicidal thoughts and attempt suicide,

than those with self-cutting on the arms alone. Interestingly, in the present study

self-cutting scars were located only on the arms among all four females, while all

three males had cut themselves on the thighs (and also on the arms in one case).

In the normal population study by Laukkanen et al. (2013) 11% of male

adolescents who had a history of self-cutting had cut themselves on parts of the

body other than the arms, compared with 100% of male suicide victims with self-

cutting in the present study. The question therefore remains whether the location

of self-cutting elsewhere than only in the arms serves as an indicator of suicide

risk among male adolescents.

It is of great importance that this study found that 48% of male and 58% of

female adolescent suicide victims had committed suicide while under the

influence of alcohol. These figures, especially with respect to females, are notably

high compared to findings from other countries. In the US, approximately 20–

30% of adolescent male and 10–15% of adolescent female suicide victims were

intoxicated with alcohol at their time of death (Brent et al. 1999, Singh & Lathrop

2008). The corresponding figures were approximately 45% and 40% in Norway

(Grøholt et al. 1999) and 40% and 20% in Sweden (Johansson et al. 2005).

Compared with previous findings from Finland, in agreement with the present

results a study by Marttunen et al. (1992) reported that 50% of male and 56% of

female adolescents had been intoxicated at the time of suicide. However, a recent

report on all deaths among Finnish children in 2009–2011 found a substantially

lower proportion, 25%, of intoxicated <18 year-old suicide victims

(Onnettomuustutkintakeskus 2014). No gender-specific results were presented.

An intriguing finding arising from the present study concerns the stage of

alcohol metabolism among intoxicated adolescent suicides. Alcohol has a well-

known biphasic effect: stimulation and elation are typical of the phase during

which the level of blood alcohol concentration (BAC) is still increasing, whereas

sedation and feelings of depression characterize the declining phase of the BAC

(Jung 2009). In the present study, most intoxicated suicide victims were in the

declining phase of BAC at the time of death. Similar findings have been reported

in studies on adult suicides (Hayward et al. 1992, James 1966), with the majority

of suicide victims having died during the descending phase of BAC. Both studies

75

concluded that acute alcohol use seems to lower the threshold for suicidal

behavior, probably through its depressant and disinhibiting effects, rather than

functioning as “Dutch courage” (i.e. first deciding to commit suicide and then

drinking alcohol to gain sufficient determination to complete the act). The current

study also found that being under the influence of alcohol was related to

committing suicide at night. Whereas 86% of both male and female adolescents

who committed suicide during the night were intoxicated, in the daytime the

proportion of suicides committed while under the influence of alcohol was only

15% among males and 25% among females. It may be that tiredness-related

adverse psychological changes, such as loss of inhibitions and impairment of

judgment, further intensify the similar adverse effects of acute alcohol use –

especially during the BAC’s descending phase, when the depressant action of

alcohol prevails.

The results of toxicological analyses revealed that, with regard to substances

other than alcohol, only one suicide victim tested positive for cannabis. No

positive findings presented in cases where so-called hard drugs, such as

amphetamine and cocaine, were tested. Prescription drugs were detected in just

under one fifth of the cases in which toxicology screens had been performed,

excluding victims who had used drug overdose as the method of suicide. The

most commonly detected drugs were anxiolytes. With two exceptions (two

females taking antidepressants), there was no indication that the prescription

drugs detected had been prescribed to the adolescents, so these drugs seemed to

have been used for intoxication purposes.

Unfortunately, no information on previous alcohol or drug use patterns

among the adolescent suicide victims was available within the study data. This

precluded an elucidation of the possible role played by more chronic substance

abuse. Previously, Pirkola et al. (1999) have shown that, compared with other

adolescent suicide victims, Finnish adolescent (13–22 years) suicide victims who

misuse alcohol suffered from more severe and long-standing psychiatric

morbidity. They also had a greater tendency to be intoxicated at the time of

suicide.

Analysis of seasonal distributions in adolescent suicides revealed a

noteworthy finding of the present study, i.e. that male adolescent firearm suicides

show a peak incidence in the fall. The seasonal pattern was remarkably different

to that in the adult population, which showed a traditional spring/early summer

peak. It was of note that the seasonal pattern of firearm suicides among young

adults (18–23 years) was similar to that observed in other adults, which indicates

76

that the fall peak is a specific phenomenon among the youngest age group. Other

authors who have observed a fall peak in young people’s suicides have suggested

that the intensification of social life after the summer vacations exposes young

people to stressful situations and increases the risk of suicide (Chew & McCleary

1994, Näyhä 1982, Shields et al. 2006). Notably, in the present study the fall peak

was found only in firearm suicides among adolescent males, whereas no specific

pattern could be found among suicides by all other methods among males and all

methods among females. This suggests that firearms play a specific role in

explaining the seasonal distribution. One explanation for this may be easy access

to hunting guns in homes in the fall, due to the beginning of the hunting season.

6.3 Strengths and limitations

6.3.1 Strengths of the study

The national register of causes of death, from which the data for the macro-level

studies in this thesis were drawn, has proven a reliable source of information for

scientific purposes (Lahti & Penttilä 2001). In Finland, the effort is to apply

methods of medicolegal death investigation in all cases in which the cause of

death may be other than a disease, and therefore the proportion of autopsies is

high among such cases (Ministry of Social Affairs and Health 2007). Overall, the

rate of medicolegal autopsies is high in this country. For example, in the year

2012 a medicolegal autopsy was performed in 19% of all deaths (Statistics

Finland 2014a). Further, the suicide statistics cover the entire country with equal

reliability. It is unlikely that determination of death as suicide differs between

medical examiners from different regions.

The study sample in micro-level studies was comprehensive, including all

consecutive child and adolescent suicides committed in the province of Oulu

during the 25 study years. The completeness and accuracy of the nationwide

health register (FHDR), which was used as one source of information, is generally

good (Sund 2012). Information collected from FHDR and the death certificates

was systematically available for all adolescent suicide victims. In addition, police

investigation reports, findings from external and internal examinations of the

medicolegal autopsies, and the final medicolegal autopsy reports were

systematically present in all 93% of cases in which the suicide case files were

77

available. Blood alcohol concentrations had been investigated in all cases and

other toxicological examinations had been performed in most.

6.3.2 Limitations of the study

A number of limitations in this study should be acknowledged. Regarding macro-

level studies, as the data consist of suicide counts, it was not possible to analyze

any biopsychosocial factors, such as the psychiatric histories of individual

victims. Some suicides may have been misclassified as accidental or

undetermined deaths (Öhberg & Lönnqvist 1998). However, the proportion of

undetermined deaths has remained stable and it is unlikely that changes in suicide

rates can be attributed to misclassification (Chishti et al. 2003).

In the micro-level studies involved in this thesis, the documents used to

obtain the data were not compiled for research purposes, but in order to establish

the cause of death. Information was not therefore systematically and consistently

available on every variable in every case. In particular, information on psychiatric

symptoms and mental health status recorded in the documents was either scarce

or had simply not been documented in any form. Conclusions on this issue must

therefore be drawn with great caution. Furthermore, unlike psychological autopsy

studies, record review studies such as the present one do not allow for psychiatric

diagnostic assessment. Gender comparisons may be biased and underpowered due

to missing data. Although all self-cutting scars identified in external examinations

were reported, in some cases there may have been a history of superficial self-

cutting which leaves no permanent scarring on the skin. When seeking potential

predictors for suicide, case-control design would allow the calculation of the

suicide risk in relation to the various background and clinical characteristics of

suicide victims. However, no meaningful controls were available for the purposes

of this study.

The small numbers of cases may have caused a lack of statistical power in the

analyses of subgroups, particularly in the micro-level studies. Some potential

findings may therefore have remained statistically non-significant (type II error).

Although the main findings are statistically robust, several comparisons have

been performed and the risk of spurious findings (i.e. type I errors) cannot be

discounted.

78

79

7 Conclusions

7.1 Main conclusions

Both favorable and unfavorable changes have occurred in suicide rates among

children and adolescents in Finland, with those of males decreasing since the

beginning of 1990s, while rates among females have increased. Although

indications of opposite trends have been observed more recently in both genders,

on a general level rates of male suicides are still lower, and those of females

higher, than they used to be two decades ago. The Finnish male adolescent suicide

rate was higher than that of females throughout the 44-year study period covered

in this research. However, the gender difference in overall suicide rates has

narrowed following the changes in suicide rates.

Finnish female adolescents now have a greater tendency to choose a violent

suicide method. The increase in overall suicide rates among female adolescents

stems specifically from the increase in violent methods of suicide, the rate of

nonviolent methods having remained unchanged. This phenomenon is of

particular concern, as females are known to make a large number of suicide

attempts, and the use of violent suicide methods is more likely to end in death.

With further regard to suicide methods, use of firearms has predominated among

Finnish adolescent males since the 1970s. In female adolescents, hanging has

exceeded poisoning as the leading method of suicide since 1990. In recent years,

however, traffic suicides have taken the leading position as the most common

method of suicide for both genders, which may indicate that the sociocultural

acceptability of this method of suicide has increased among Finland’s young

population.

Between 1972–2009, the adolescent male firearm suicide rate was almost

three times higher in Northern Finland than in Southern Finland, while no such

regional difference appeared in rates of suicide by all other methods. This

suggests that, compared with Southern Finland, use of firearms plays a major role

in explaining the higher male child and adolescent suicide rate in Northern

Finland. Although a causal relationship cannot be inferred from this, it is

proposed that the greater prevalence of firearms in Northern Finland compared to

Southern Finland is a contributing factor.

As regards the similarities and differences in adolescent male and female

suicides, apart from those observed in the rates and methods of suicide, the

80

empirically based study from the province of Oulu investigated more specific

characteristics associated with the suicide victims. Regardless of gender, evidence

of mental health problems was common, but only a minority had a history of

psychiatric hospitalization. Previous suicidality and self-cutting were more

common in female than male adolescents. Whereas being under the influence of

alcohol at the time of death was infrequent among child and young adolescent

suicide victims, more than half of the 15–17-year-old male and female

adolescents had been intoxicated with alcohol at the time of their death, which

highlights the role of acute alcohol use in adolescent suicides in both genders.

Most of the intoxicated adolescents were in the descending phase of blood alcohol

concentration at the time of their suicide, which suggests that the depressant

effect of alcohol use may have contributed to the decision to commit suicide. This

study also found that two thirds of all adolescent male firearm suicides were

committed between August and October, indicating that the fall is a high-risk

period for such suicides.

7.2 Implications of the study

The general upward trend in Finnish female adolescent suicide rates warrants

attention. As suicide is a multifactorial process, no conclusive explanations can be

offered for the changes in suicide trends, but the increase in violent methods of

suicide seems to contribute to the phenomenon. In terms of suicide prevention

efforts, method restriction is considered one of the most effective preventive

strategies. The predominance of hanging suicides and, in recent times, traffic

suicides in particular, impose limitations on the potential for employing the

method restriction strategy, since both of these methods are difficult to restrict.

The rate of adolescent male firearm suicides was substantially higher in

Northern than Southern Finland. Another special characteristic of firearm suicides

among adolescent males in Northern Finland was the notable seasonal pattern,

with a peak incidence of firearm suicides occurring in the fall. These findings

emphasize the need to develop suicide prevention in line with specific regional

characteristics.

Acute alcohol use was shown to play a central role in adolescent suicides,

regardless of gender. Although no information on patterns of previous alcohol use

among adolescents was available for the present study, previous research has

shown that alcohol abuse is an important risk factor in adolescent suicide.

81

Reducing alcohol use and drunkenness-oriented drinking among underage

adolescents is highly important from a preventative viewpoint.

Psychiatric disorders are thought to be present in the majority of adolescent

suicides, but these disorders have often gone unrecognized and untreated. Many

adolescents, especially males, tend not to seek help for their mental health

problems. However, the role of health care professionals in suicide prevention

(most importantly, identifying and treating psychiatric disorders) is limited to

individuals who come into contact with health services. The role of family,

friends, and other close people can therefore be vital in recognizing possible signs

of distress in adolescents. Nevertheless, when an adolescent presents to the health

care system and his or her general health is being evaluated, consideration should

always be taken of the patient’s mental health status. Finally, functioning, low-

threshold mental health services providing easy access to care, preferably

integrated with the school health care system, and a reduction in the stigma

associated with mental health problems, could be helpful in encouraging

vulnerable children and adolescents to seek and obtain help.

7.3 Implications for further research

This epidemiological study on the trends and characteristics of suicide among

children and adolescents in Finland provides the following insights for future

research. Firstly, the increased use of violent suicide methods among female

adolescents is alarming. Further studies are needed in order to investigate the

possible underlying factors at individual level and to explore whether the

phenomenon is universal amongst this young age group. Secondly, studies on the

seasonality of child and adolescent suicides are rare and the findings are

heterogeneous. Additional studies are therefore needed, which should also take

account of the effect of biopsychosocial factors in order to identify their putative

influences on seasonal variations in suicides. Thirdly, it remains to be seen

whether the novel finding on the descending phase of blood alcohol

concentrations at the time of suicide among children and adolescents can be

confirmed in future studies.

82

83

References

Act on the Inquest into the Cause of Death (1973) No. 459/1973. URI: http://www.finlex.fi/fi/laki/ajantasa/1973/19730459 [in Finnish]. Cited 2014/05/07.

Ajdacic-Gross V, Bopp M, Ring M, Gutzwiller F & Rossler W (2010a) Seasonality in suicide – a review and search of new concepts for explaining the heterogeneous phenomena. Soc Sci Med 71(4): 657–666.

Ajdacic-Gross V, Killias M, Hepp U, Haymoz S, Bopp M, Gutzwiller F & Rossler W (2010b) Firearm suicides and availability of firearms: the Swiss experience. Eur Psychiatry 25(7): 432–434.

Aldridge D & St John K (1991) Adolescent and pre-adolescent suicide in Newfoundland and Labrador. Can J Psychiatry 36(6): 432–436.

American Academy of Child and Adolescent Psychiatry (2001) Practice parameter for the assessment and treatment of children and adolescents with suicidal behavior. American Academy of Child and Adolescent Psychiatry. J Am Acad Child Adolesc Psychiatry 40(7 Suppl): 24S–51S.

American Psychiatric Association (2003) Practice guideline for the assessment and treatment of patients with suicidal behaviors. Am J Psychiatry 160(11 Suppl): 1–60.

American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders, 5th edition. Arlington, VA, American Psychiatric Association.

Barber C (2005) Exploring Differences in Suicide Rates in the Northeast. A Research Summary from the Harvard Injury Control Research Center. URI: http://www.hsph.harvard.edu/wp-content/uploads/sites/1264/2013/01/Differences_in_ Northeast_State_Suicide_Rates.pdf. Cited 2014/06/16.

Beautrais AL (2002) Gender issues in youth suicidal behaviour. Emerg Med (Fremantle) 14(1): 35–42.

Beautrais AL (2003) Suicide and serious suicide attempts in youth: a multiple-group comparison study. Am J Psychiatry 160(6): 1093–1099.

Beautrais AL, Fergusson DM & Horwood LJ (2006) Firearms legislation and reductions in firearm-related suicide deaths in New Zealand. Aust N Z J Psychiatry 40(3): 253–259.

Berman AL, Jobes DA & Silverman MM (2006) Adolescent Suicide: Assessment and Intervention. Washington, DC, American Psychological Association.

Biddle L, Brock A, Brookes ST & Gunnell D (2008) Suicide rates in young men in England and Wales in the 21st century: time trend study. BMJ 336(7643): 539–542.

Birckmayer J & Hemenway D (2001) Suicide and firearm prevalence: are youth disproportionately affected? Suicide Life Threat Behav 31(3): 303–310.

Brent DA, Baugher M, Bridge J, Chen T & Chiappetta L (1999) Age- and sex-related risk factors for adolescent suicide. J Am Acad Child Adolesc Psychiatry 38(12): 1497–1505.

Brent DA & Bridge J (2003) Firearms availability and suicide: Evidence, interventions, and future directions. Am Behav Sci 46(9): 1192–1210.

84

Brent DA, Bridge J, Johnson BA & Connolly J (1996) Suicidal behavior runs in families. A controlled family study of adolescent suicide victims. Arch Gen Psychiatry 53(12): 1145–1152.

Brent DA, Perper J, Moritz G, Baugher M & Allman C (1993a) Suicide in adolescents with no apparent psychopathology. J Am Acad Child Adolesc Psychiatry 32(3): 494–500.

Brent DA, Perper JA & Allman CJ (1987) Alcohol, firearms, and suicide among youth. Temporal trends in Allegheny County, Pennsylvania, 1960 to 1983. JAMA 257(24): 3369–3372.

Brent DA, Perper JA, Allman CJ, Moritz GM, Wartella ME & Zelenak JP (1991) The presence and accessibility of firearms in the homes of adolescent suicides. A case-control study. JAMA 266(21): 2989–2995.

Brent DA, Perper JA, Goldstein CE, Kolko DJ, Allan MJ, Allman CJ & Zelenak JP (1988) Risk factors for adolescent suicide. A comparison of adolescent suicide victims with suicidal inpatients. Arch Gen Psychiatry 45(6): 581–588.

Brent DA, Perper JA, Moritz G, Allman C, Friend A, Roth C, Schweers J, Balach L & Baugher M (1993b) Psychiatric risk factors for adolescent suicide: A case-control study. J Am Acad Child Adolesc Psychiatry 32(3): 521–529.

Brent DA, Perper JA, Moritz G, Baugher M, Schweers J & Roth C (1993c) Firearms and adolescent suicide. A community case-control study. Am J Dis Child 147(10): 1066–1071.

Bridge JA, Barbe RP & Brent DA (2005) Recent trends in suicide among U.S. adolescent males, 1992–2001. Psychiatr Serv 56(5): 522.

Bridge JA, Goldstein TR & Brent DA (2006) Adolescent suicide and suicidal behavior. J Child Psychol Psychiatry 47(3–4): 372–394.

Bridge JA, Greenhouse JB, Weldon AH, Campo JV & Kelleher KJ (2008) Suicide trends among youths aged 10 to 19 years in the United States, 1996–2005. JAMA 300(9): 1025–1026.

Brown CH, Wyman PA, Brinales JM & Gibbons RD (2007) The role of randomized trials in testing interventions for the prevention of youth suicide. Int Rev Psychiatry 19(6): 617–631.

Brunner R, Kaess M, Parzer P, Fischer G, Carli V, Hoven CW, Wasserman C, Sarchiapone M, Resch F, Apter A, Balazs J, Barzilay S, Bobes J, Corcoran P, Cosmanm D, Haring C, Iosuec M, Kahn JP, Keeley H, Meszaros G, Nemes B, Podlogar T, Postuvan V, Saiz PA, Sisask M, Tubiana A, Värnik A & Wasserman D (2014) Life-time prevalence and psychosocial correlates of adolescent direct self-injurious behavior: a comparative study of findings in 11 European countries. J Child Psychol Psychiatry 55(4): 337–348.

Campo JV (2009) Youth suicide prevention: does access to care matter? Curr Opin Pediatr 21(5): 628–634.

Canetto SS (1997) Meanings of gender and suicidal behavior during adolescence. Suicide Life Threat Behav 27(4): 339–351.

85

Canetto SS & Sakinofsky I (1998) The gender paradox in suicide. Suicide Life Threat Behav 28(1): 1–23.

Cantor CH & Baume PJ (1998) Access to methods of suicide: what impact? Aust N Z J Psychiatry 32(1): 8–14.

Cheung AH & Dewa CS (2005) Current trends in youth suicide and firearms regulations. Can J Public Health 96(2): 131–135.

Chew KS & McCleary R (1994) A life course theory of suicide risk. Suicide Life Threat Behav 24(3): 234–244.

Child Welfare Act (2007) No. 417/2007. URI: http://www.finlex.fi/fi/laki/kaannokset/ 2007/en20070417.pdf. Cited 2014/05/13.

Chishti P, Stone DH, Corcoran P, Williamson E, Petridou E & EUROSAVE Working Group (2003) Suicide mortality in the European Union. Eur J Public Health 13(2): 108–114.

Christodoulou C, Douzenis A, Papadopoulos FC, Papadopoulou A, Bouras G, Gournellis R & Lykouras L (2012) Suicide and seasonality. Acta Psychiatr Scand 125(2): 127–146.

Cooper WO, Callahan ST, Shintani A, Fuchs DC, Shelton RC, Dudley JA, Graves AJ & Ray WA (2014) Antidepressants and suicide attempts in children. Pediatrics 133(2): 204–210.

Crosby AE, Ortega L & Melanson C (2011) Self-directed Violence Surveillance: Uniform Definitions and Recommended Data Elements, Version 1.0. Atlanta, GA, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control.

Daigle MS (2005) Suicide prevention through means restriction: assessing the risk of substitution. A critical review and synthesis. Accid Anal Prev 37(4): 625–632.

Daine K, Hawton K, Singaravelu V, Stewart A, Simkin S & Montgomery P (2013) The power of the web: A systematic review of studies of the influence of the internet on self-harm and suicide in young people. PLoS One 8(10): e77555.

De Leo D, Burgis S, Bertolote JM, Kerkhof AJ & Bille-Brahe U (2006) Definitions of suicidal behavior: Lessons learned from the WHO/EURO multicentre Study. Crisis 27(1): 4–15.

Dervic K, Brent DA & Oquendo MA (2008) Completed suicide in childhood. Psychiatr Clin North Am 31(2): 271–291.

Dervic K, Friedrich E, Oquendo MA, Voracek M, Friedrich MH & Sonneck G (2006) Suicide in Austrian children and young adolescents aged 14 and younger. Eur Child Adolesc Psychiatry 15(7): 427–434.

Dumon E & Portzky G (2013a) EUREGENAS: European Regions Enforcing Actions Against Suicide. General Guidelines on Suicide Prevention. URI: http://www.euregenas.eu/wp-content/uploads/2014/01/EUREGENAS-General-Guidelines-on-Suicide-Prevention-F.pdf. Cited 2014/05/05.

Dumon E & Portzky G (2013b) EUREGENAS: European Regions Enforcing Actions Against Suicide. Suicide Prevention Toolkit for Media Professionals. URI: http://www.euregenas.eu/wp-content/uploads/2014/04/Euregenas-Suicide-Prevention-Toolkit-for-Media-Professionals.pdf. Cited 2014/05/05.

86

Eisele JW, Frisino J, Haglund W & Reay DT (1987) Teenage suicide in King County, Washington. II. Comparison with adult suicides. Am J Forensic Med Pathol 8(3): 210–216.

European Detailed Mortality Database (DMDB) (2014) Copenhagen, WHO Regional Office for Europe. URI: http://data.euro.who.int/dmdb/. Cited 2014/05/27.

Evans E, Hawton K, Rodham K & Deeks J (2005) The prevalence of suicidal phenomena in adolescents: a systematic review of population-based studies. Suicide Life Threat Behav 35(3): 239–250.

Feltz-Cornelis CM van der, Sarchiapone M, Postuvan V, Volker D, Roskar S, Grum AT, Carli V, McDaid D, O'Connor R, Maxwell M, Ibelshauser A, Van Audenhove C, Scheerder G, Sisask M, Gusmao R & Hegerl U (2011) Best practice elements of multilevel suicide prevention strategies: a review of systematic reviews. Crisis 32(6): 319–333.

Field M, Wiers RW, Christiansen P, Fillmore MT & Verster JC (2010) Acute alcohol effects on inhibitory control and implicit cognition: implications for loss of control over drinking. Alcohol Clin Exp Res 34(8): 1346–1352.

Finnish Meteorological Institute (2004) Ilmastokatsaus 2003;4:5, 2003;5:5, 2003;7:5, 2003;10:5, 2004;1:7. URI: http://ilmatieteenlaitos.fi/ilmastokatsaus-arkisto [in Finnish]. Cited 2014/06/16.

Firearms Act (1998) No. 1/1998. URI: http://www.finlex.fi/fi/laki/ajantasa/1998/19980001 [in Finnish]. Cited 2014/04/16.

Florentine JB & Crane C (2010) Suicide prevention by limiting access to methods: a review of theory and practice. Soc Sci Med 70(10): 1626–1632.

Freuchen A, Kjelsberg E & Grøholt B (2012) Suicide or accident? A psychological autopsy study of suicide in youths under the age of 16 compared to deaths labeled as accidents. Child Adolesc Psychiatry Ment Health 6(1): 30.

Gallerani M, Avato FM, Dal Monte D, Caracciolo S, Fersini C & Manfredini R (1996) The time for suicide. Psychol Med 26(4): 867–870.

Gould MS, Fisher P, Parides M, Flory M & Shaffer D (1996) Psychosocial risk factors of child and adolescent completed suicide. Arch Gen Psychiatry 53(12): 1155–1162.

Gould MS, Greenberg T, Velting DM & Shaffer D (2003) Youth suicide risk and preventive interventions: A review of the past 10 years. J Am Acad Child Adolesc Psychiatry 42(4): 386–405.

Gould MS, Wallenstein S & Kleinman M (1990) Time-space clustering of teenage suicide. Am J Epidemiol 131(1): 71–78.

Greydanus DE & Calles J,Jr (2007) Suicide in children and adolescents. Prim Care 34(2): 259–73; abstract vi.

Grøholt B, Ekeberg Ø, Wichstrøm L & Haldorsen T (1997) Youth suicide in Norway, 1990–1992: a comparison between children and adolescents completing suicide and age- and gender-matched controls. Suicide Life Threat Behav 27(3): 250–263.

Grøholt B, Ekeberg Ø, Wichstrøm L & Haldorsen T (1998) Suicide among children and younger and older adolescents in Norway: A comparative study. J Am Acad Child Adolesc Psychiatry 37(5): 473–481.

87

Grøholt B, Ekeberg Ø, Wichstrøm L & Haldorsen T (1999) Sex differences in adolescent suicides in Norway, 1990–1992. Suicide Life Threat Behav 29(4): 295–308.

Hammad TA, Laughren T & Racoosin J (2006) Suicidality in pediatric patients treated with antidepressant drugs. Arch Gen Psychiatry 63(3): 332–339.

HarperCollins (2014) Online English Dictionary. URI: http://www.collinsdictionary.com/ dictionary/english/childhood. Cited 2014/08/26.

Haw C, Hawton K, Niedzwiedz C & Platt S (2013) Suicide clusters: a review of risk factors and mechanisms. Suicide Life Threat Behav 43(1): 97–108.

Hawton K (2005) Restriction of access to methods of suicide as a means of suicide prevention. In: Hawton K (ed) Prevention and Treatment of Suicidal Behavior: From Science to Practice. Oxford, UK, Oxford University Press: 279–291.

Hawton K, Bergen H, Kapur N, Cooper J, Steeg S, Ness J & Waters K (2012a) Repetition of self-harm and suicide following self-harm in children and adolescents: findings from the Multicentre Study of Self-harm in England. J Child Psychol Psychiatry 53(12): 1212–1219.

Hawton K, Saunders KE & O'Connor RC (2012b) Self-harm and suicide in adolescents. Lancet 379(9834): 2373–2382.

Hayward L, Zubrick SR & Silburn S (1992) Blood alcohol levels in suicide cases. J Epidemiol Community Health 46(3): 256–260.

HE (20/2014). Hallituksen esitys eduskunnalle ampumaratalaiksi sekä laeiksi ampuma-aselain ja eräiden siihen liittyvien lakien muuttamisesta. DOI: http://www.finlex.fi/fi/esitykset/he/2014/20140020 [in Finnish]. Cited 2014/05/23.

Health Care Act (2010) No. 1326/2010. URI: http://www.stm.fi/c/document_library/ get_file?folderId=5064551&name=DLFE-17718.pdf. Cited 2014/08/04.

Hepp U, Stulz N, Unger-Köppel J & Ajdacic-Gross V (2012) Methods of suicide used by children and adolescents. Eur Child Adolesc Psychiatry 21(2): 67–73.

Hiltunen L, Partonen T, Haukka J & Lönnqvist JK (2009) Itsemurhakuolleisuuden käännekohdat Suomessa 1947–2006. Duodecim 125(16): 1802–1806 [in Finnish, summary available in English].

Hoberman HM & Garfinkel BD (1988) Completed suicide in children and adolescents. J Am Acad Child Adolesc Psychiatry 27(6): 689–695.

Holdstock L & de Wit H (1998) Individual differences in the biphasic effects of ethanol. Alcohol Clin Exp Res 22(9): 1903–1911.

Horowitz LM, Ballard ED & Pao M (2009) Suicide screening in schools, primary care and emergency departments. Curr Opin Pediatr 21(5): 620–627.

Hufford MR (2001) Alcohol and suicidal behavior. Clin Psychol Rev 21(5): 797–811. Hultén A, Jiang GX, Wasserman D, Hawton K, Hjelmeland H, De Leo D, Ostamo A,

Salander-Renberg E & Schmidtke A (2001) Repetition of attempted suicide among teenagers in Europe: Frequency, timing and risk factors. Eur Child Adolesc Psychiatry 10(3): 161–169.

IBM Corp. (2013) IBM SPSS Statistics for Windows. Armonk, NY, IBM Corp. Isacsson G & Rich CL (2014) Antidepressant drugs and the risk of suicide in children and

adolescents. Paediatr Drugs 16(2): 115–122.

88

James IP (1966) Blood alcohol levels following successful suicide. Q J Stud Alcohol 27(1): 23–29.

Johansson L, Stenlund H, Lindqvist P & Eriksson A (2005) A survey of teenager unnatural deaths in northern Sweden 1981–2000. Accid Anal Prev 37(2): 253–258.

Johnson GR, Krug EG & Potter LB (2000) Suicide among adolescents and young adults: a cross-national comparison of 34 countries. Suicide Life Threat Behav 30(1): 74–82.

Joiner TE (1999) The clustering and contagion of suicide. Curr Dir Psychol Sci 8(3): 89–92.

Jones AW (2006) Urine as a biological specimen for forensic analysis of alcohol and variability in the urine-to-blood relationship. Toxicol Rev 25(1): 15–35.

Jones AW, Holmgren A & Ahlner J (2013a) Toxicology findings in suicides: concentrations of ethanol and other drugs in femoral blood in victims of hanging and poisoning in relation to age and gender of the deceased. J Forensic Leg Med 20(7): 842–847.

Jones AW & Kugelberg FC (2010) Relationship between blood and urine alcohol concentrations in apprehended drivers who claimed consumption of alcohol after driving with and without supporting evidence. Forensic Sci Int 194(1–3): 97–102.

Jones P, Gunnell D, Platt S, Scourfield J, Lloyd K, Huxley P, John A, Kamran B, Wells C & Dennis M (2013b) Identifying probable suicide clusters in Wales using national mortality data. PLoS One 8(8): e71713.

Jung J (2009) Basic Psychological Processes. In: Jung J (ed) Alcohol, Other Drugs, and Behavior: Psychological Research Perspectives. Thousand Oaks, Sage Publications: 183–212.

Kaess M, Parzer P, Haffner J, Steen R, Roos J, Klett M, Brunner R & Resch F (2011) Explaining gender differences in non-fatal suicidal behaviour among adolescents: A population-based study. BMC Public Health 11: 597.

Kapur N, Cooper J, O'Connor RC & Hawton K (2013) Non-suicidal self-injury v. attempted suicide: new diagnosis or false dichotomy? Br J Psychiatry 202(5): 326–328.

Karch DL, Logan J, McDaniel DD, Floyd CF & Vagi KJ (2013) Precipitating circumstances of suicide among youth aged 10–17 years by sex: Data from the National Violent Death Reporting System, 16 states, 2005–2008. J Adolesc Health 53(1 Suppl): S51–3.

Kerkhof A (2000) Attempted Suicide: Patterns and Trends. In: Hawton K & van Heeringen K (eds) The International Handbook of Suicide and Attempted Suicide. England, John Wiley & Sons Ltd: 49–64.

Kim HJ, Fay MP, Feuer EJ & Midthune DN (2000) Permutation tests for joinpoint regression with applications to cancer rates. Stat Med 19(3): 335–351.

Kolves K & De Leo D (2014) Regions with the highest suicide rates for children and adolescents – some observations. J Child Adolesc Behav 2(2): 1000e104.

Korkeila J, Salminen JK, Hiekkanen H & Salokangas RK (2007) Use of antidepressants and suicide rate in Finland: an ecological study. J Clin Psychiatry 68(4): 505–511.

89

Kugelberg FC & Jones AW (2007) Interpreting results of ethanol analysis in postmortem specimens: a review of the literature. Forensic Sci Int 165(1): 10–29.

Lahti RA & Penttilä A (2001) The validity of death certificates: routine validation of death certification and its effects on mortality statistics. Forensic Sci Int 115(1–2): 15–32.

Lamis DA & Malone PS (2012) Alcohol Use and Suicidal Behaviors among Adults: A Synthesis and Theoretical Model. Suicidol Online 3: 4–23.

Laukkanen E, Rissanen ML, Honkalampi K, Kylmä J, Tolmunen T & Hintikka J (2009) The prevalence of self-cutting and other self-harm among 13- to 18-year-old Finnish adolescents. Soc Psychiatry Psychiatr Epidemiol 44(1): 23–28.

Laukkanen E, Rissanen ML, Tolmunen T, Kylmä J & Hintikka J (2013) Adolescent self-cutting elsewhere than on the arms reveals more serious psychiatric symptoms. Eur Child Adolesc Psychiatry 22: 501–510.

Law CK & De Leo D (2013) Seasonal differences in the day-of-the-week pattern of suicide in Queensland, Australia. Int J Environ Res Public Health 10(7): 2825–2833.

Leslie LK, Newman TB, Chesney PJ & Perrin JM (2005) The Food and Drug Administration's deliberations on antidepressant use in pediatric patients. Pediatrics 116(1): 195–204.

Levine B & Smialek JE (2000) Status of alcohol absorption in drinking drivers killed in traffic accidents. J Forensic Sci 45(1): 3–6.

Lewinsohn PM, Rohde P & Seeley JR (1996) Adolescent suicidal ideation and attempts: Prevalence, risk factors, and clinical implications. Clinical Psychology Science and Practice 3: 25–36.

Lewinsohn PM, Rohde P, Seeley JR & Baldwin CL (2001) Gender differences in suicide attempts from adolescence to young adulthood. J Am Acad Child Adolesc Psychiatry 40(4): 427–434.

Lönnqvist J (2007) Itsemurhien ehkäisy Suomessa valtakunnallisena interventiona. Kansanterveys 7: 5–7 [in Finnish].

Lönnqvist J (2009) Itsemurhat. Sairauksien ehkäisy. Terveyskirjasto. URI: http://www.terveyskirjasto.fi/kotisivut/tk.koti?p_artikkeli=seh00158 [in Finnish]. Cited 2014/08/11.

Lönnqvist J, Henriksson M, Isometsä E & Marttunen M (2011) Itsetuhokäyttäytyminen. In: Lönnqvist J, Henriksson M, Marttunen M & Partonen T (eds) Psykiatria. Helsinki, Duodecim: 501–515 [in Finnish].

Lönnqvist J, Aro H & Marttunen M (1993) Itsemurhat Suomessa 1987 -projekti. Toteutus, aineisto ja tutkimustuloksia. Sosiaali- ja terveysalan tutkimus- ja kehittämiskeskus STAKES. Tutkimuksia 25. Jyväskylä, Gummerus [in Finnish].

Lubell KM, Kegler SR, Crosby AE & Karch D (2007) Suicide trends among youths and young adults aged 10–24 years – United States, 1990–2004. Morb Mortal Weekly Rep 56(35): 905–908.

Lubin G, Werbeloff N, Halperin D, Shmushkevitch M, Weiser M & Knobler HY (2010) Decrease in suicide rates after a change of policy reducing access to firearms in adolescents: a naturalistic epidemiological study. Suicide Life Threat Behav 40(5): 421–424.

90

Madge N (1999) Youth suicide in an international context. Eur Child Adolesc Psychiatry 8(4): 283–291.

Maldonado G & Kraus JF (1991) Variation in suicide occurrence by time of day, day of the week, month, and lunar phase. Suicide Life Threat Behav 21(2): 174–187.

Mann JJ, Apter A, Bertolote J, Beautrais A, Currier D, Haas A, Hegerl U, Lönnqvist J, Malone K, Marusic A, Mehlum L, Patton G, Phillips M, Rutz W, Rihmer Z, Schmidtke A, Shaffer D, Silverman M, Takahashi Y, Värnik A, Wasserman D, Yip P & Hendin H (2005) Suicide prevention strategies: a systematic review. JAMA 294(16): 2064–2074.

Maris RW (2002) Suicide. Lancet 360(9329): 319–326. Marttunen MJ (1994) Adolescent suicide in Finland. Publications of the National Public

Health Institute, KTL A2/1994. Hyvinkää, T-print Ky. Marttunen MJ, Aro HM, Henriksson MM & Lönnqvist JK (1991) Mental disorders in

adolescent suicide. DSM-III-R axes I and II diagnoses in suicides among 13- to 19-year-olds in Finland. Arch Gen Psychiatry 48(9): 834–839.

Marttunen MJ, Aro HM & Lönnqvist JK (1992) Adolescent suicide: Endpoint of long-term difficulties. J Am Acad Child Adolesc Psychiatry 31(4): 649–654.

Marttunen MJ, Aro HM & Lönnqvist JK (1993) Adolescence and suicide: A review of psychological autopsy studies. Eur Child Adolesc Psychiatry 2(1): 10–18.

Marttunen MJ, Henriksson MM, Aro HM, Heikkinen ME, Isometsä ET & Lönnqvist JK (1995) Suicide among female adolescents: Characteristics and comparison with males in the age group 13 to 22 years. J Am Acad Child Adolesc Psychiatry 34(10): 1297–1307.

McCleary R, Chew KS, Hellsten JJ & Flynn-Bransford M (1991) Age- and sex-specific cycles in United States suicides, 1973 to 1985. Am J Public Health 81(11): 1494–1497.

McClure GM (2001) Suicide in children and adolescents in England and Wales 1970–1998. Br J Psychiatry 178: 469–474.

McKeown RE, Cuffe SP & Schulz RM (2006) US suicide rates by age group, 1970–2002: an examination of recent trends. Am J Public Health 96(10): 1744–1751.

Mental Health Act (1990) No. 1116/1990. URI: http://www.finlex.fi/en/laki/ kaannokset/1990/en19901116.pdf. Cited 2014/05/13.

Mesoudi A (2009) The cultural dynamics of copycat suicide. PLoS One 4(9): e7252. Miller M, Azrael D & Hemenway D (2002) Household firearm ownership and suicide

rates in the United States. Epidemiology 13(5): 517–524. Miller M & Hemenway D (1999) The relationship between firearms and suicide: A review

of the literature. Aggr Viol Beh 4(1): 59–75. Miller M, Lippmann SJ, Azrael D & Hemenway D (2007) Household firearm ownership

and rates of suicide across the 50 United States. J Trauma 62(4): 1029–1034; discussion 1034–1035.

Miller TR, Furr-Holden CD, Lawrence BA & Weiss HB (2012) Suicide deaths and nonfatal hospital admissions for deliberate self-harm in the United States. Temporality by day of week and month of year. Crisis 33(3): 169–177.

91

Ministry of Social Affairs and Health (2007) Oikeuslääketieteellisen kuolemansyyn selvittämisen organisointi. Sosiaali- ja terveysministeriön selvityksiä 2007:64. DOI: http://www.stm.fi/c/document_library/get_file?folderId=28707&name=DLFE-3498.pdf&title=Oikeuslaaketieteellisen_kuolemansyyn_selvittamisen_organisointi_fi.pdf [in Finnish, summary available in English]. Cited 2014/05/14.

Mishara BL (1999) Conceptions of death and suicide in children ages 6–12 and their implications for suicide prevention. Suicide Life Threat Behav 29(2): 105–118.

Moran P, Coffey C, Romaniuk H, Olsson C, Borschmann R, Carlin JB & Patton GC (2012) The natural history of self-harm from adolescence to young adulthood: a population-based cohort study. Lancet 379(9812): 236–243.

Morrell S, Page AN & Taylor RJ (2007) The decline in Australian young male suicide. Soc Sci Med 64(3): 747–754.

Moskos MA, Achilles J & Gray D (2004) Adolescent suicide myths in the United States. Crisis 25(4): 176–182.

Moskos MA, Olson L, Halbern SR & Gray D (2007) Utah youth suicide study: Barriers to mental health treatment for adolescents. Suicide Life Threat Behav 37(2): 179–186.

Näyhä S (1982) Autumn incidence of suicides re-examined: data from Finland by sex, age and occupation. Br J Psychiatry 141: 512–517.

Niederkrotenthaler T, Voracek M, Herberth A, Till B, Strauss M, Etzersdorfer E, Eisenwort B & Sonneck G (2010) Role of media reports in completed and prevented suicide: Werther v. Papageno effects. Br J Psychiatry 197(3): 234–243.

Niedzwiedz C, Haw C, Hawton K & Platt S (2014) The Definition and Epidemiology of Clusters of Suicidal Behavior: A Systematic Review. Suicide Life Threat Behav [Epub ahead of print].

Nock MK (2009) Suicidal behavior among adolescents: correlates, confounds, and (the search for) causal mechanisms. J Am Acad Child Adolesc Psychiatry 48(3): 237–239.

Nock MK (2010) Self-injury. Annu Rev Clin Psychol 6: 339–363. Nock MK, Green JG, Hwang I, McLaughlin KA, Sampson NA, Zaslavsky AM & Kessler

RC (2013) Prevalence, correlates, and treatment of lifetime suicidal behavior among adolescents: results from the National Comorbidity Survey Replication Adolescent Supplement. JAMA Psychiatry 70(3): 300–310.

Nordentoft M (2011) Crucial elements in suicide prevention strategies. Prog Neuropsychopharmacol Biol Psychiatry 35(4): 848–853.

Öhberg A & Lönnqvist J (1998) Suicides hidden among undetermined deaths. Acta Psychiatr Scand 98(3): 214–218.

Öhberg A, Lönnqvist J, Sarna S & Vuori E (1996) Violent methods associated with high suicide mortality among the young. J Am Acad Child Adolesc Psychiatry 35(2): 144–153.

Olson R (2013) Centre for Suicide Prevention. InfoExchange 10: Suicide contagion and suicide clusters. URI: http://suicideinfo.ca/LinkClick.aspx?fileticket= WXg70KbEYsA=. Cited 2014/05/22.

92

Onnettomuustutkintakeskus (2014) Deaths among children. URI: http://www.turvallisuustutkinta.fi/material/attachments/otkes/tutkintaselostukset/fi/muutonnettomuudet/2012/AtKr3zBTr/Y2012-S1_Lasten_kuolemat.pdf [in Finnish, summary available in English]. Cited 2014/05/19.

Ougrin D (2012) Commentary: Self-harm in adolescents: the best predictor of death by suicide? – Reflections on Hawton et al. (2012). J Child Psychol Psychiatry 53(12): 1220–1221.

Park S, Cho SC, Kim BN, Kim JW, Yoo HJ & Hong JP (2014) Increased use of lethal methods and annual increase of suicide rates in Korean adolescents: comparison with adolescents in the United States. J Child Psychol Psychiatry 55(3): 258–263.

Partonen T (2012) Itsemurhan ehkäisy. In: Partanen A, Moring J, Nordling E & Bergman V (eds) Kansallinen mielenterveys- ja päihdesuunnitelma 2009–2015. THL – Raportti 46/2012. DOI: http://www.thl.fi/thl-client/pdfs/53837d85-321b-4694-90ad-a2332abaea71 [in Finnish]. Cited 2014/08/08.

Partonen T, Haukka J, Viilo K, Hakko H, Pirkola S, Isometsä E, Lönnqvist J, Särkioja T, Väisänen E & Räsänen P (2004) Cyclic time patterns of death from suicide in northern Finland. J Affect Disord 78(1): 11–19.

Patton GC, Coffey C, Sawyer SM, Viner RM, Haller DM, Bose K, Vos T, Ferguson J & Mathers CD (2009) Global patterns of mortality in young people: A systematic analysis of population health data. Lancet 374(9693): 881–892.

Pelkonen M & Marttunen M (2003) Child and adolescent suicide: epidemiology, risk factors, and approaches to prevention. Paediatr Drugs 5(4): 243–265.

Peña JB & Caine ED (2006) Screening as an approach for adolescent suicide prevention. Suicide Life Threat Behav 36(6): 614–637.

Pirkola SP, Marttunen MJ, Henriksson MM, Isometsä ET, Heikkinen ME & Lönnqvist JK (1999) Alcohol-related problems among adolescent suicides in Finland. Alcohol Alcohol 34(3): 320–329.

Pompili M, Masocco M, Vichi M, Lester D, Innamorati M, Tatarelli R & Vanacore N (2009) Suicide among Italian adolescents: 1970–2002. Eur Child Adolesc Psychiatry 18(9): 525–533.

Portzky G, Audenaert K & van Heeringen K (2005) Suicide among adolescents. A psychological autopsy study of psychiatric, psychosocial and personality-related risk factors. Soc Psychiatry Psychiatr Epidemiol 40(11): 922–930.

Portzky G, Audenaert K & van Heeringen K (2009) Psychosocial and psychiatric factors associated with adolescent suicide: a case-control psychological autopsy study. J Adolesc 32(4): 849–862.

Portzky G & van Heeringen K (2006) Suicide prevention in adolescents: a controlled study of the effectiveness of a school-based psycho-educational program. J Child Psychol Psychiatry 47(9): 910–918.

Posner K, Oquendo MA, Gould M, Stanley B & Davies M (2007) Columbia Classification Algorithm of Suicide Assessment (C-CASA): classification of suicidal events in the FDA's pediatric suicidal risk analysis of antidepressants. Am J Psychiatry 164(7): 1035–1043.

93

Preti A & Miotto P (2001) Diurnal variations in suicide by age and gender in Italy. J Affect Disord 65(3): 253–261.

Pylkkänen K (2013) Nuorisopsykiatrisen avohoidon laatusuositus. NALLE-projektin loppuraportti. Helsinki, Suomen Nuorisopsykiatrinen Yhdistys. URI: http://www.nuorisopsykiatrinen-yhdistys.org/wp-content/uploads/2013/09/ SNPY_laatusuositus_1013.pdf [in Finnish]. Cited 2014/08/04.

Renaud J, Berlim MT, McGirr A, Tousignant M & Turecki G (2008) Current psychiatric morbidity, aggression/impulsivity, and personality dimensions in child and adolescent suicide: A case-control study. J Affect Disord 105(1–3): 221–228.

Rich CL, Sherman M & Fowler RC (1990) San Diego Suicide Study: the adolescents. Adolescence 25(100): 855–865.

Rickwood DJ, Deane FP & Wilson CJ (2007) When and how do young people seek professional help for mental health problems? Med J Aust 187(7 Suppl): S35–9.

Robertson L, Skegg K, Poore M, Williams S & Taylor B (2012) An adolescent suicide cluster and the possible role of electronic communication technology. Crisis 33(4): 239–245.

Robins LN & Kulbok PA (1986) Methodological strategies in suicide. Ann N Y Acad Sci 487: 1–15.

Robinson J, Hetrick SE & Martin C (2011) Preventing suicide in young people: systematic review. Aust N Z J Psychiatry 45(1): 3–26.

Rutz EM & Wasserman D (2004) Trends in adolescent suicide mortality in the WHO European Region. Eur Child Adolesc Psychiatry 13(5): 321–331.

Sacks D, Canadian Pediatric Society & Adolescent Health Committee (2003) Age limits and adolescents. Paediatr Child Health 8(9): 577.

Schrijvers DL, Bollen J & Sabbe BG (2012) The gender paradox in suicidal behavior and its impact on the suicidal process. J Affect Disord 138(1–2): 19–26.

Schwartz-Lifshitz M, Zalsman G, Giner L & Oquendo MA (2012) Can we really prevent suicide? Curr Psychiatry Rep 14(6): 624–633.

Shaffer D, Gould MS, Fisher P, Trautman P, Moreau D, Kleinman M & Flory M (1996) Psychiatric diagnosis in child and adolescent suicide. Arch Gen Psychiatry 53(4): 339–348.

Shafii M, Carrigan S, Whittinghill JR & Derrick A (1985) Psychological autopsy of completed suicide in children and adolescents. Am J Psychiatry 142(9): 1061–1064.

Shafii M, Steltz-Lenarsky J, Derrick AM, Beckner C & Whittinghill JR (1988) Comorbidity of mental disorders in the post-mortem diagnosis of completed suicide in children and adolescents. J Affect Disord 15(3): 227–233.

Shah S, Hoffman RE, Wake L & Marine WM (2000) Adolescent suicide and household access to firearms in Colorado: results of a case-control study. J Adolesc Health 26(3): 157–163.

Shields LB, Hunsaker DM & Hunsaker JC,3rd (2006) Adolescent and young adult suicide: A 10-year retrospective review of Kentucky Medical Examiner cases. J Forensic Sci 51(4): 874–879.

94

Silverman MM (2006) The language of suicidology. Suicide Life Threat Behav 36(5): 519–532.

Silverman MM, Berman AL, Sanddal ND, O'Carroll PW & Joiner TE (2007) Rebuilding the tower of Babel: a revised nomenclature for the study of suicide and suicidal behaviors. Part 1: Background, rationale, and methodology. Suicide Life Threat Behav 37(3): 248–263.

Singh VD & Lathrop SL (2008) Youth suicide in New Mexico: A 26-year retrospective review. J Forensic Sci 53(3): 703–708.

Skegg K (2005) Self-harm. Lancet 366(9495): 1471–1483. Skinner R & McFaull S (2012) Suicide among children and adolescents in Canada: trends

and sex differences, 1980–2008. CMAJ 184(9): 1029–1034. Sloan JH, Rivara FP, Reay DT, Ferris JA & Kellermann AL (1990) Firearm regulations

and rates of suicide. A comparison of two metropolitan areas. N Engl J Med 322(6): 369–373.

Soor GS, Vukin I, Bridgman-Acker K, Marble R, Barnfield P, Edwards J, Cooper B, Alfonsi J, Hunter J, Banayan DJ & Bhalerao S (2012) The effects of gender on adolescent suicide in Ontario, Canada. J Can Acad Child Adolesc Psychiatry 21(3): 179–185.

Stack S (1987) The sociological study of suicide: methodological issues. Suicide Life Threat Behav 17(2): 133–150.

Statistics Finland (2013) Quality Description, Causes of death 2012. URI: http://www.stat.fi/til/ksyyt/2012/ksyyt_2012_2013-12-30_laa_001_en.html. Cited 2014/05/28.

Statistics Finland (2014a) Causes of death [e-publication]. Helsinki: Statistics Finland. URI: http://www.stat.fi/til/ksyyt/tau.html [in Finnish]. Cited 2014/04/28.

Statistics Finland (2014b) Population structure [e-publication]. Helsinki: Statistics Finland. URI: http://www.stat.fi/til/vaerak/tau_en.html. Cited 2014/04/28.

Sund R (2012) Quality of the Finnish Hospital Discharge Register: A systematic review. Scand J Public Health 40(6): 505–515.

Thapar A, Collishaw S, Potter R & Thapar AK (2010) Managing and preventing depression in adolescents. BMJ 340: c209.

Thompson TR (1987) Childhood and adolescent suicide in Manitoba: a demographic study. Can J Psychiatry 32(4): 264–269.

United Nations (1989) Convention on the Rights of the Child. URI: http://www.ohchr.org/Documents/ProfessionalInterest/crc.pdf. Cited 2014/04/30.

United Nations Committee on the Rights of the Child (2011) Consideration of reports submitted by States parties under article 44 of the Convention. Concluding observations: Finland. URI: http://um.fi/public/default.aspx?contentid=223505& contentlan=2&culture=en-US. Cited 2014/05/27.

U.S. Department of Health and Human Services (HHS) Office of the Surgeon General and National Action Alliance for Suicide Prevention (2012) 2012 National Strategy for Suicide Prevention: Goals and Objectives for Action. Washington, DC, U.S. Department of Health and Human Services.

95

Uzun I, Karayel FA, Akyildiz EU, Turan AA, Toprak S & Arpak BB (2009) Suicide among children and adolescents in a province of Turkey. J Forensic Sci 54(5): 1097–1100.

Värnik A, Kõlves K, Allik J, Arensman E, Aromaa E, van Audenhove C, Bouleau JH, van der Feltz-Cornelis CM, Giupponi G, Gusmão R, Kopp M, Marusic A, Maxwell M, Oskarsson H, Palmer A, Pull C, Realo A, Reisch T, Schmidtke A, Pérez Sola V, Wittenburg L & Hegerl U (2009) Gender issues in suicide rates, trends and methods among youths aged 15–24 in 15 European countries. J Affect Disord 113(3): 216–226.

Wahlbeck K (2011) Asekontrolli on kansanterveystyötä. Duodecim 127: 2580–2851 [in Finnish].

Waldvogel JL, Rueter M & Oberg CN (2008) Adolescent suicide: risk factors and prevention strategies. Curr Probl Pediatr Adolesc Health Care 38(4): 110–125.

Wasserman D, Cheng Q & Jiang GX (2005) Global suicide rates among young people aged 15–19. World Psychiatry 4(2): 114–120.

Webster DW, Vernick JS, Zeoli AM & Manganello JA (2004) Association between youth-focused firearm laws and youth suicides. JAMA 292(5): 594–601.

Weinberger LE, Sreenivasan S, Sathyavagiswaran L & Markowitz E (2001) Child and adolescent suicide in a large, urban area: Psychological, demographic, and situational factors. J Forensic Sci 46(4): 902–907.

Windfuhr K, While D, Hunt I, Turnbull P, Lowe R, Burns J, Swinson N, Shaw J, Appleby L, Kapur N & National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (2008) Suicide in juveniles and adolescents in the United Kingdom. J Child Psychol Psychiatry 49(11): 1155–1165.

Windfuhr K, While D, Hunt IM, Shaw J, Appleby L & Kapur N (2013) Suicide and accidental deaths in children and adolescents in England and Wales, 2001–2010. Arch Dis Child 98(12): 945–950.

Wonnacott TH & Wonnacott RJ (1990) Introductory Statistics, ed. 5. Singapore, Wiley. World Health Organization (1998) Primary prevention of mental, neurological and

psychosocial disorders. Geneva, World Health Organization. World Health Organization (2008) Preventing suicide: a resource for media professionals.

URI: http://www.who.int/mental_health/prevention/suicide/resource_media.pdf. Cited 2014/05/22.

World Health Organization (1986) Young people’s health – a challenge for society. Report of a WHO Study Group on Young People and “Health for all by the Year 2000”. Technical Report Series 731. Geneva, World Health Organization.

Wunderlich U, Bronisch T, Wittchen HU & Carter R (2001) Gender differences in adolescents and young adults with suicidal behaviour. Acta Psychiatr Scand 104(5): 332–339.

Yip PS, Caine E, Yousuf S, Chang SS, Wu KC & Chen YY (2012) Means restriction for suicide prevention. Lancet 379(9834): 2393–2399.

96

97

Appendix 1. The Finnish death certificate form

98

99

Original publications

I Lahti A, Räsänen P, Riala K, Keränen S & Hakko H (2011) Youth suicide trends in Finland, 1969–2008. Journal of Child Psychology and Psychiatry 52: 984–991.

II Lahti A, Keränen S, Hakko H, Riala K & Räsänen P (2014) Northern excess in adolescent male firearm suicides: A register-based regional study from Finland, 1972–2009. European Child & Adolescent Psychiatry 23: 45–52.

III Lahti A, Harju A, Hakko H, Riala K & Räsänen P (2014) Suicide in children and young adolescents: A 25-year database on suicides from Northern Finland. Journal of Psychiatric Research 58: 123–128.

IV Lahti A, Räsänen P, Karvonen K, Särkioja T, Meyer-Rochow VB & Hakko H (2006) Autumn peak in shooting suicides of children and adolescents from Northern Finland. Neuropsychobiology 54: 140–146.

Reprinted with the permission of John Wiley and Sons (I), Springer (II), Elsevier

(III) and Karger (IV).

The original articles are not included in the electronic version of the thesis.

100

A C T A U N I V E R S I T A T I S O U L U E N S I S

Book orders:Granum: Virtual book storehttp://granum.uta.fi/granum/

S E R I E S D M E D I C A

1240. Sova, Henri (2014) Oxidative stress in breast and gynaecological carcinogenesis

1241. Tiirinki, Hanna (2014) Näkyvien ja piilotettujen merkitysten rajapinnoilla :terveyskeskukseen liittyvät kulttuurimallit asiakkaan näkökulmasta

1242. Syrjänen, Riikka (2014) TIM family molecules in hematopoiesis

1243. Kauppila, Joonas (2014) Toll-like receptor 9 in alimentary tract cancers

1244. Honkavuori-Toivola, Maria (2014) The prognostic role of matrixmetalloproteinase-2 and -9 and their tissue inhibitor-1 and -2 in endometrialcarcinoma

1245. Pienimäki, Tuula (2014) Factors, complications and health-related quality of lifeassociated with diabetes mellitus developed after midlife in men

1246. Kallio, Miki (2014) Muuttuuko lääketieteen opiskelijoiden käsitys terveydestäperuskoulutuksen aikana : kuusivuotinen seurantatutkimus

1247. Haanpää, Maria (2014) Hereditary predisposition to breast cancer – with a focuson AATF, MRG15, PALB2, and three Fanconi anaemia genes

1248. Alanne, Sami (2014) Musiikkipsykoterapia : teoria ja käytäntö

1249. Nagy, Irina I. (2014) Wnt-11 signaling roles during heart and kidney development

1250. Prunskaite-Hyyryläinen, Renata (2014) Role of Wnt4 signaling in mammalian sexdetermination, ovariogenesis and female sex duct differentiation

1251. Huusko, Johanna (2014) Genetic background of spontaneous preterm birth andlung diseases in preterm infants : studies of potential susceptibility genes andpolymorphisms

1252. Jämsä, Ulla (2014) Kuntoutuksen muutosagentit : tutkimus työelämälähtöisestäoppimisesta ylemmässä ammattikorkeakoulutuksessa

1253. Kaikkonen, Leena (2014) p38 mitogen-activated protein kinase and transcriptionfactor GATA-4 in the regulation of cardiomyocyte function

1254. Finnilä, Mikko A. J. (2014) Bone toxicity of persistent organic pollutants

1255. Starck, Tuomo (2014) Dimensionality, noise separation and full frequency bandperspectives of ICA in resting state fMRI : investigations into ICA in resting statefMRI

1256. Karhu, Jaana (2014) Severe community- acquired pneumonia – studies on imaging,etiology, treatment, and outcome among intensive care patients

ABCDEFG

UNIVERSITY OF OULU P .O. B 00 F I -90014 UNIVERSITY OF OULU FINLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

S E R I E S E D I T O R S

SCIENTIAE RERUM NATURALIUM

HUMANIORA

TECHNICA

MEDICA

SCIENTIAE RERUM SOCIALIUM

SCRIPTA ACADEMICA

OECONOMICA

EDITOR IN CHIEF

PUBLICATIONS EDITOR

Professor Esa Hohtola

University Lecturer Santeri Palviainen

Postdoctoral research fellow Sanna Taskila

Professor Olli Vuolteenaho

University Lecturer Veli-Matti Ulvinen

Director Sinikka Eskelinen

Professor Jari Juga

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0556-4 (Paperback)ISBN 978-952-62-0557-1 (PDF)ISSN 0355-3221 (Print)ISSN 1796-2234 (Online)

U N I V E R S I TAT I S O U L U E N S I S

MEDICA

ACTAD

D 1257

ACTA

Anniina Lahti

OULU 2014

D 1257

Anniina Lahti

EPIDEMIOLOGICAL STUDYON TRENDS AND CHARACTERISTICS OF SUICIDE AMONG CHILDREN AND ADOLESCENTS IN FINLAND

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU, FACULTY OF MEDICINE,INSTITUTE OF CLINICAL MEDICINE, DEPARTMENT OF PSYCHIATRY;OULU UNIVERSITY HOSPITAL, DEPARTMENT OF PSYCHIATRY;HELSINKI UNIVERSITY CENTRAL HOSPITAL,DEPARTMENT OF PSYCHIATRY, DIVISION OF ADOLESCENT PSYCHIATRY