UNIVERSITY OF OULU P .O. Box 8000 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

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

University Lecturer Anu Soikkeli

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0983-8 (Paperback)ISBN 978-952-62-0984-5 (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 1322

ACTA

Tiina T

irkkonen

OULU 2015

D 1322

Tiina Tirkkonen

EARLY ATTACHMENT, MENTAL WELL-BEING AND DEVELOPMENT OF FINNISH CHILDREN AT PRESCHOOL AGETWINSHIP – RISK OR OPPORTUNITY?

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF MEDICINE

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 3 2 2

TIINA TIRKKONEN

EARLY ATTACHMENT, MENTAL WELL-BEING AND DEVELOPMENT OF FINNISH CHILDREN AT PRESCHOOL AGETwinship – risk or opportunity?

Academic dissertation to be presented with the assent ofthe Doctora l Train ing Committee of Health andBiosciences of the University of Oulu for public defence inthe Leena Palotie auditorium (101A) of the Faculty ofMedicine (Aapistie 5 A), on 4 December 2015, at 12 noon

UNIVERSITY OF OULU, OULU 2015

Copyright © 2015Acta Univ. Oul. D 1322, 2015

Supervised byProfessor Irma MoilanenProfessor Hanna Ebeling

Reviewed byProfessor Anne Mari TorgersenDocent Ilona Luoma

ISBN 978-952-62-0983-8 (Paperback)ISBN 978-952-62-0984-5 (PDF)

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

Cover DesignRaimo Ahonen

JUVENES PRINTTAMPERE 2015

OpponentDocent Jari Sinkkonen

Tirkkonen, Tiina, Early attachment, mental well-being and development ofFinnish children at preschool age. Twinship – risk or opportunity?University of Oulu Graduate School; University of Oulu, Faculty of MedicineActa Univ. Oul. D 1322, 2015University of Oulu, P.O. Box 8000, FI-90014 University of Oulu, Finland

Abstract

This study focuses on early attachment and its effects on later cognitive development and mentalwell-being in Finnish twins and singletons. Altogether 84 infants, including both singletons (N =27) and twins (N = 57), were assessed at 18 months of age to determine their infant–mother andinfant–father attachment using the Preschool Assessment of Attachment (PAA), a modification ofthe Ainsworth infant classification system. In the whole sample, approximately one third (37%)of the children were assessed as using avoidant/defended insecure (Type A), one third (35%)secure/balanced (Type B), one fifth (20%) resistant/coercive insecure (Type C) and the restinsecure other (IO, 8%) attachment strategy with their mothers, while the correspondingpercentages of attachment strategies with fathers were 36%, 33%, 28% and 3%, respectively. Theresults yielded a difference in the maternal attachment between twins and singletons; the twinswere more often Type B attached than the singletons.

At 36 months the children were tested with the Stanford-Binet test for measuring theirintelligence level. Children with Type A or Type B attachment pattern attained the highest IQ.High socio-economic status of the family and mature pregnancy (≥ 37 weeks) were alsosignificantly linked with higher intelligence levels.

At the age of 48 months, the children’s mental well-being was assessed by using CBCL (ChildBehavior Checklist, Achenbach) questionnaires completed by the mothers of 22 singletons and 51twins and the fathers of 20 singletons and 60 twins. Type A attachment strategy with the motheramong singleton toddlers was significantly associated with higher CBCL scores reported bymothers, concerning withdrawal, somatic problems and total internalizing symptoms, whereasamong twins there were no such correlations. In addition, on the basis of the parental reportssingletons had significantly more behavioural and emotional symptoms than the twins.

Keywords: attachment, externalizing problems, internalizing problems, mental well-being, twins

Tirkkonen, Tiina, Leikki-ikäisten suomalaislasten varhainen kiintymyssuhde,kognitiivinen kehitys ja psyykkinen hyvinvointi. Kaksosuus - riski vaimahdollisuus?Oulun yliopiston tutkijakoulu; Oulun yliopisto, Lääketieteellinen tiedekuntaActa Univ. Oul. D 1322, 2015Oulun yliopisto, PL 8000, 90014 Oulun yliopisto

Tiivistelmä

Tutkimuksessa tarkastellaan suomalaisten yksös- ja kaksoslasten varhaisen kiintymyssuhteenvaikutusta myöhempään lapsuusiän kehitykseen ja psyykkiseen hyvinvointiin. Yhteensä 84 lap-sen, 27 yksösen ja 57 kaksosen, kiintymyssuhdestrategia arviointiin 18 kuukauden iässä äitiin jaisään PAA – menetelmällä (Preschool Assessment of Attachment), joka on modifikaatio Ains-worthin pikkulasten kiintymyssuhteen luokitusjärjestelmästä. Lapsista noin kolmanneksella(37%) oli suhteessa äitiin välttelevä kiintymyssuhdestrategia (A-tyyppi), toisella kolmanneksel-la (35%) turvallinen (B-tyyppi), viidenneksellä (20%) vastustava (C-tyyppi) ja loppuosalla muuturvaton (IO, 8%). Vastaavat osuudet lasten ja isien välisissä kiintymyssuhteissa ryhmittäin (A,B, C ja IO) olivat 36%, 33%, 28% ja 3%. Kaksosten kiintymyssuhdestrategia suhteessa äitiin oliyksösiä useammin turvallinen.

Lasten kognitiivinen taso arvioitiin Stanford-Binet –testillä 36 kuukauden iässä. Lapset, joi-den kiintymyssuhde äitiin tai isään oli joko A- tai B-tyyppinen, menestyivät testissä parhaiten.Korkea sosioekonominen status ja täysiaikaisuus (≥ 37 raskausviikkoa) olivat yhteydessä mer-kitsevästi parempaan suoriutumiseen testissä.

Lasten psyykkistä hyvinvointia arvioitiin 48 kuukauden iässä CBCL-lomakkeella (ChildBehavior Checklist), jonka täytti 51 kaksosen ja 22 yksösen äitiä sekä 60 kaksosen ja 20 yksösenisää. Yksösillä, joilla oli A-tyyppinen kiintymyssuhde äitiin, oli äitien raportoimana merkitse-västi enemmän vetäytymistä, somaattisia oireita ja ylipäätään internalisoivia (tunne-elämän)oireita. Lisäksi vanhemmat raportoivat yksösillä olevan merkitsevästi enemmän sekä käyttäyty-misen että tunne-elämän oireita kaksosiin verrattuna.

Asiasanat: eksternalisoivat ongelmat, henkinen hyvinvointi, internalisoivat ongelmat,kaksosuus, kiintymyssuhde

To my Family

8

9

Acknowledgements

First I want to thank the parents who patiently and faithfully took part with their

children in the studies of this dissertation. You and your children are the stars of

this project.

I am very grateful to my supervisor and mentor, emerita Professor Irma

Moilanen for introducing me to the area of research on child attachment and twins

and for her skilful guidance throughout this long project. Her optimism and

courage have been an inspiration to me. I am likewise grateful to my other

supervisor professor Hanna Ebeling for her wise comments.

Very special thanks are due to my friend, co-writer and colleague Anne Kunelius,

with whom I videotaped the SST-processes and who has encouraged me with this

project over these years.

Warm thanks go to Leena Joskitt for her patience, kindness and time in

sharing her statistical experience. Many thanks are due to psychologist Terttu

Tapio for sharing her skills and experience in this project, as well as to the other

co-writers Marika Huhtaniska, Heidi Tuomikoski, and Miia Pylkkönen.

I owe my respectful gratitude to the official reviewers of this dissertation,

docent Ilona Luoma and emerita professor Anne Mari Torgersen, for their careful

revision of this thesis. Your comments helped me to improve the final version of

this thesis.

I extend my warm thanks to secretary Meeri Jämsä for her kind arrangements

and other important practical help during these years. Warm thanks also go to

Teuvo Ryynänen for proficient and precise help with the editing of this book.

Furthermore, I sincerely thank Michael Bailey, for his invaluable help in checking the

English language of this dissertation.

I extend special thanks to Dr Patricia Crittenden, PhD for training me in the

attachment theory and PAA–method. I am grateful to Dr Nathan Szajnberg, PhD

for guiding me in the attachment theory and SST-classification.

I warmly thank my friends together with my colleagues and my co-workers

in Medifamilia and in other offices and clinics for their encouragement during

this project.

My warm and respectful thanks go out to my dear parents Raija and Timo for

their support and encouragement to take up different things in my life, relying on

my wings to bear me up. I am grateful to my dear sister Kaisa-Reetta for her

practical advice and encouragement and for being a brilliant example of

10

perseverance, as well as to my dear brother Pekka and my dear mother-in-law

Ritva for support with this project.

Finally, I give my warmest thanks to my dear family, Veli-Pekka, Lauri and Elias.

My darling husband, Veli-Pekka, your help and encouragement made it possible for

me to bring this project to completion. You have a special talent to encourage in the

dark and rejoice in the light moments. My dear children, Lauri and Elias, you are the

sunshine of my life, thank you for being there.

The studies of this dissertation were financially supported by the Research

Academy of Finland, the Finnish Child Psychiatric Research Foundation and the

Alma and KA Snellman Foundation, Oulu, Finland.

Oulu, August 2015 Tiina Tirkkonen

11

Abbreviations

ADHD Attention deficit and hyperactivity disorder

ART Assisted reproductive techniques

CBCL Child Behavior Checklist

DMM Dynamic Maturational Model

DZ Dizygotic

GA Gestational age

IO Insecure other

IQ Intelligence quotient

IUGR Intrauterine growth restriction

IVF In Vitro Fertilization

MZ Monozygotic

PAA Preschool Assessment of Attachment

SB Stanford - Binet intelligence scale

SB IV Stanford - Binet intelligence scale, 4th version

SES Socioeconomic status

SF Singleton families

TF Twin families

THL Terveyden ja hyvinvoinnin laitos, National institute for Health and

Welfare

SSP Strange Situation Procedure

SST Strange Situation Test

Type A Avoidant/Defended insecure attachment strategy

Type B Secure/Balanced attachment strategy

Type C Resistant/Coercive insecure attachment strategy

UCH University Central Hospital

YSR Youth Self Report

12

13

List of original publications

This thesis is based on the following publications, which are referred to in the text

by their Roman numerals. In addition, the thesis includes some unpublished data.

I Moilanen I, Kunelius A, Tirkkonen T & Crittenden P (2000) Attachment in Finnish twins. In Crittenden P & Claussen A (eds) The Organization of Attachment Relationships. Cambridge University Press: 125–40.

II Tirkkonen T, Kunelius A, Ebeling H, Pylkkönen M, Tuomikoski H & Moilanen I (2008) Attachment in Finnish singletons and twins to both parents. Psychiatria Fennica 39: 89–98.

III Tirkkonen T, Joskitt L, Kunelius A, Huhtaniska M & Moilanen I (2015) Twinship as a protective factor against behavioural and emotional problems at preschool age. Early Child Development and Care. DOI: 10.1080/03004430.2015.1066783.

14

15

Table of contents

Abstract Tiivistelmä Acknowledgements 9 Abbreviations 11 List of original publications 13 Table of contents 15 1  Introduction 17 2  Review of the literature 19 

2.1  Attachment theory ................................................................................... 19 2.1.1  Attachment Classification ............................................................ 21 2.1.2  The Preschool Assessment of Attachment (PAA) ......................... 24 

2.2  Concordance of attachment in a family................................................... 26 2.3  Attachment patterns in different countries .............................................. 27 2.4  Internalizing and externalizing problems in childhood ........................... 27 2.5  Attachment and cognitive development .................................................. 29 2.6  Attachment and mental well-being ......................................................... 30 2.7  The context of twins ................................................................................ 31 

3  Aims and hypotheses of the study 35 4  Materials and methods 37 

4.1  Participants .............................................................................................. 37 4.2  Procedure ................................................................................................ 38 4.3  Methods ................................................................................................... 40 

4.3.1  Parental Questionnaire at 18 months ............................................ 40 4.3.2  Strange Situation Procedure and Preschool Assessment of

Attachment (PAA) ........................................................................ 40 4.3.3  Stanford-Binet Intelligence Scale (SB) ........................................ 41 4.3.4  Child Behavior Checklist (CBCL) ............................................... 41 4.3.5  Statistical methods ........................................................................ 42 

5  Results 43 5.1  Attachment of singletons and twins to both parents (I, II) ...................... 43 5.2  Attachment and cognitive development (Previously unpublished

data)......................................................................................................... 49 5.3  Attachment and mental well-being (III) .................................................. 51 

6  Discussion 61 6.1  Attachment of singletons and twins to both parents................................ 61 6.2  Attachment and cognitive development .................................................. 62 

16

6.3  Attachment and mental well-being in singletons and twins .................... 63 6.4  Limitations .............................................................................................. 65 

7  Conclusions: Implications and future research 67 References 69 Original publications 79 

17

1 Introduction

John Bowlby's attachment theory is an evolutionary one, in that babies are seen as

having a biological drive to seek proximity from a protective adult, usually the

primary caregiver, to cope with danger. The goal of this drive for closeness is to feel

safe, secure and protected. This leads to a range of proximity-promoting signals to the

caregiver, attachment behaviours in the first months of life, these signals are repeated

numerous times. When the baby is hungry or lonely the sensitive and responsive

caregiver will both recognize and respond rapidly enough to meet the baby's needs.

This experience of a secure base calms the baby and reduces his or her anxiety, which

in turn allows him or her to play and explore. When the baby is relaxed and playful,

the caregiver will share and reinforce this mood. Through the process of attachment

behaviours being responded to sufficiently quickly and appropriately, the baby's

survival is ensured, and additionally his or her emotional, social and physical

development is supported in the context of the relationship. (Bowlby 1969/1982,

1970, 1980).

Thus the ways in which children modify their behaviour and strategies for

achieving safety in relationships are influenced primarily by their caregiving

history and their temperament (Cassidy 1994). The strategies have been termed

“patterns of attachment” and can be reliably measured in a standardised

laboratory procedure, Strange Situation Procedure, SSP (Ainsworth et al. 1978).

This has made it possible to study and compare the effects of attachment pattern

on an individual’s development and well-being world-wide. Previous Finnish

studies of attachment using the SSP, have focused on ELBW (Extremely Low Birth

Weight) infants and their mothers (Sajaniemi et al. 2001) and on children, who live

with their foster parents (Kalland & Sinkkonen 2013). Thus, as far as we know, the

first original publication in 2000 of this thesis brought Finnish attachment

research into international awareness, also analyzing the results within the

contexts of twins and of Finnish society.

The primary attachment bond is formed with the person most involved in rapidly

responding to an infant’s cries and who initiates social play during the first year of

his/her life. The person who measures up second-best will usually become the most

important secondary attachment figure. For most people the longest attachment

relationship they may ever have in their life is with a sibling. A sibling might be a

secondary attachment figure, but sometimes a primary one. Well-developed

secondary attached relationships can provide considerable safety if the primary

attachment figure is, for one reason or another, not emotionally or physically

18

available. In particular the case in point is twins, with a unique life-long bond. In

addition to the biological risks involved in twin pregnancy and delivery, twins

might have another obstacle to favourable attachment and development: they

have to share their parents’ attention, and the parents’ response to proximity-

promoting signals may take longer compared to singletons. How do these factors

effect on early attachment? On the other hand, is it possible that twins can provide

the co-twin with both emotional and physical safety even in early childhood? The

author searched for the answer to these questions in the original publications of this

thesis by analyzing preschool attachment to mother and father in singletons and twins,

and the possible consequences of living as a twin up to the age of four years.

19

2 Review of the literature

2.1 Attachment theory

Attachment theory was developed in collaboration between John Bowlby and

Mary Ainsworth and describes how an infant’s relationship with the primary

caregiver lays the base for later social-emotional development. John Bowlby

formulated the basic dogmas of the theory by drawing on concepts from ethology,

developmental psychology, information processing and psychoanalysis; Mary

Ainsworth’s role was to provide innovative methodology making it possible to

test some of Bowlby’s ideas empirically. In addition, Ainsworth contributed the

concept of the attachment figure as a secure base from which an infant can

explore the world. She also formulated the concept of maternal sensitivity to

infant signals and its role in the development of infant-mother attachment patterns.

(Ainsworth & Bowlby 1991)

Bowlby (1969/1982) described the goal of the child to be not an object, but

rather to maintain a state of desired distance from the mother, depending on the

circumstances. He presented children as wanting to maintain a certain proximity

to their mothers. When a separation lasts too long or the distance is too great, the

attachment system is activated and when adequate proximity has been achieved,

the attachment system is terminated. A balance between the need for exploratory

behaviour and the attachment behavioural system must be achieved that is

complementary and yet mutually inhibiting or stimulating. The nature of this

balance is thought to have developed to ensure that although the child is protected

by maintaining proximity to the attachment figure, at the same time he or she

nonetheless learns about the environment through exploration (Cassidy 1999). A

balance between attachment behaviour and environmental exploration is shown in

the use of an attachment figure as a secure base from which to explore, and which

provides optimal development of independence (Ainsworth 1971).

Bowlby (1979) highlighted that emotions are intimately associated with

behavioural systems. He proposed that during human evolution, genetic selection

favoured attachment behaviours because they increased the probability of

protection and provided a survival advantage (Bowlby 1969/1982). Many

predictable outcomes favourable to the child are thought to result from the child’s

proximity to the parent, such as social interaction, feeding, learning about the

environment, and the basic biological function of attachment behaviour,

20

protection from predators (Cassidy 1999). In Bowlby’s attempt to integrate

psychology and psychiatry with behavioural biology, he speculated that the

attachment system, an evolutionary mechanism common to both humans and

animals, would ultimately be located in specific areas of the brain (Bowlby

1969/1982). Updated models of attachment theory that emphasize both emotional

and social functions and neurobiological structures are now interfacing with

developmental neuroscience to generate a wide body of interdisciplinary studies

(Schore 2005).

In expanding the biological emphasis of Bowlby`s initial theorizing, Main

(1990) proposed that the biologically based human tendency to become attached

is paralleled by a biologically based ability to be flexible towards a range of

caregiving environments. This flexibility is thought to promote variations

associated with quality of attachment (Cassidy 1999). Thus even if nearly all

children become attached, not all of them are securely attached. Secure

attachment Type occurs when a child has a mental representation of the

attachment figure as being available and responsive when needed.

Emotional regulation, defined as characteristic strategies and behaviours used

to modulate emotional arousal, is closely connected to quality of attachment

(Contreas et al. 2000) and may serve as one of the mechanisms through which

attachment security affects later socioemotional outcomes (Braungart-Rieker et

al. 2001). Emotional regulation involves both intrinsic processes and those

extrinsic to the child, particularly the child’s relationship with his or her parent.

Parents play an important role in emotional development and it is largely within

the parent-child relationship that children learn about emotional regulation in the

service of attaining their goals. (Thompson 1999). Emotional regulation is

thought to be adaptive in relation to attachment (Main 1990). When an infant is

faced with a particular type of caregiving and learns from these experiences,

he/she is able to tailor his/her own behaviour accordingly; this is described as a

strategy (Cassidy 1994). Working models, developed during the infant’s first year,

are based on repeated daily experiences with the parent (Bowlby 1969/1982).

Through these socialization practices, among other things, children acquire

emotional models that guide their predictions of the consequences of expressing

various emotions in certain states (Cassidy 1994). For example, the way in which

caregivers respond to their children’s negative moods and their availability to

their child when he or she is upset, will affect what types of strategies children

adopt to regulate their emotional states (Contreas et al. 2000).

21

2.1.1 Attachment Classification

Attachment classification is based on the behaviour of the young toddler (12 to 20

months of age) in a Strange Situation Test (SST), which is a laboratory procedure

designed by Ainsworth and her colleagues (1978). It assesses the balance of

attachment and exploratory behaviour under increasing though moderate stress

(Ainsworth et al. 1978). Differences in quality of attachment relationships are

divided into two main categories: secure and insecure. The terms secure and

insecure describe the infant’s apparent perception of the availability of the

caregiver if a need for protection or comfort should arise, and the organization of

the infant’s response to the caregiver in the light of these perceptions of

availability (Weinfeld et al. 1999).

Security in the relationship with an attachment figure indicates that an infant

is able to rely on that caregiver being available when he/she needs protection or

comfort. The attachment figure is consistently sensitive, accepting and

cooperative (Ainsworth et al. 1978). The flexibility in emotion expression is

thought to arise in part from experiences with a caregiver who responds

sensitively much of the time to a range of infants’ affective signals, without

selective ignoring (Bretherton 1980). Ainsworth et al. (1978) propounded that

infants whose mothers respond sensitively to their signals are more likely to be

securely attached (Type B). Secure relationships encourage infants’ exploration of

the world and expand their mastery of even unsettling environments; they can

rely on their caregivers to be there and alleviate their fears (Bowlby 1973).

Securely attached infants seek proximity and contact with little or no avoidance

or angry resistance towards their caregivers. The contact and comfort they receive

are effective in calming them, allowing them to play following distress (Lyons-

Ruth & Jacobvitz 1999). Securely attached toddlers appear to be more

enthusiastic, persistent, affectively positive and compliant in problem solving

(Matas et al. 1978) and have better social skills than insecurely attached (Waters

et al. 1979). A recent (Groh et al. 2014) meta-analytic review focused on the

association between attachment during early life and social competence with peers

during childhood. Based on eighty independent samples (N = 4 441), the association

between security and peer competence was significant and not moderated by the age

at which peer competence was assessed. At preschool-age securely attached

children exhibit affectional, confident interaction with their caregiver following a

separation (Main et al. 1985). Securely attached children exhibit less anxiety and

22

aggression, and more empathy and overall affect regulation than insecurely

attached children (Weinfeld et al. 1999).

Avoidantly attached (Type A) infants learn to ignore the caregiver and reject

his/her attention in unfamiliar situations rather than using him or her as a secure

base from which to explore the environment (Solomon & George 1999). The

caregivers of avoidantly attached infants have been described as intrusive,

controlling, rejecting, and insensitive (Ainsworth et al. 1978). In the SST, Type A

attached children appear to minimize emotional expression. They seem to be

largely neutral in affect, showing neither overt distress on separation from the

caregiver, nor pleasure on reunion (Ainsworth et al. 1978). Infants develop a

reasonable strategy of minimizing attention to the attachment relationship,

minimizing the importance of the caregiver as a source of comfort and hence their

apparent need for him/her (Main & Solomon 1986). Inhibiting negative emotions

such as sadness, anger, or distress is thought to be strategically useful (Cassidy

1994). It is this minimizing that actually permits the infant to maintain proximity

to the caregiver in order to ensure protection. Avoidance and masking of negative

emotion decrease the infant’s arousal level (Bowlby 1980) and prevent the direct

expression of anger towards the attachment figure. At preschool-age, infants with

avoidant attachment strategy show minimal displays of affect (Lyons-Ruth &

Jacobvitz 1999) and ignore the parent, responding minimally when addressed and

moving to a distance from parent when reunited following separation (Main et al.

1985). As pre-schoolers they have been described as hostile, socially isolated

and/or disconnected in the preschool setting, which represents the defensive

posturing one would predict for a child with an attachment figure who is

rejecting, emotionally unavailable, or perhaps depressed. Preschool teachers

reported these avoidantly attached children as being more withdrawn, likely to

give up easily and more exhibitionistic and impulsive than other children.

(Erikson et al. 1985).

Resistantly attached (Type C) infants display angry, resistant behaviour

towards the parent in the SST, exhibiting extreme distress on separation and

difficulties in calming on reunion. Attachment-related experiences that promote

this strategy of heightened emotions and increasing demand for attention are

thought to involve a minimally or inconsistently available parent (Ainsworth et

al. 1978). By increasing dependence on the attachment figure, the resistantly

attached infant successfully draws the attention of the parent (Main & Salomon

1986). They are conspicuously unable to use the caregiver as a secure base for

exploration of new scenarios (Ainsworth et al. 1978). It is typical in this

23

classification in most cases that the child will maintain contact aggressively once

it is achieved (Weinfeld et al. 1999). The negative emotionality of the Type C

attached child may be extravagant and chronic because the child experiences that

to relax and allow him/herself to be soothed by the presence of the attachment

figure is to run the risk of then losing contact with the inconsistently available

caregiver (Cassidy 1994). Caregivers of resistantly attached infants were found to

be insensitive to signals but not notably rejecting (Ainsworth et al. 1971).

According to Bowlby (1973) such a strategy can be dysfunctional if the negative

emotionality is so pervasive that it threatens the existence of the relationship and

other developmental tasks such as social interaction and exploration. Resistantly

attached infants are more likely in later childhood to have problems with anxiety,

possibly as a result of the constant alertness they have developed during their

early attachment relationship (Weinfeld et al. 1999). They have been rated by

observers in preschool as being less confident and assertive and as having poorer

social skills than securely attached infants (Bates & Bayles 1988).

Disorganized attachment (type D) was added by Main and Solomon (1986)

as a fourth attachment classification to the Ainsworth classification system which

they named “disorganized/disorientated” to describe the diverse group of

previously unrecognized fearful, odd, disorganized or overtly conflicting

behaviours exhibited during the SST. Disorganized infants exhibit such

behaviours as appearing apprehensive, crying and huddling to the floor in

response to their parents’ return following a short separation, turning in circles

while simultaneously approaching their parents, and a freezing of movement

while exhibiting a trance-like expression (Main & Solomon 1986, Lyon-Ruth &

Jacobvitz 1999). Main and Hesse (1990) proposed that such behaviours are a

result of the infant being exposed to inexplicably frightening and/or frightened

behaviour by a caregiver. The frightening attachment figure poses an inherent

conflict for the infant: fear activates the attachment system (Bowlby 1969/1982)

and the infant feels compelled to seek proximity; however, through seeking

proximity to the caregiver, the source of fright increases the infant’s fear, leading

to contradictory behaviour (Main & Hesse 1990). Competing tendencies to move

towards and away from the caregiver, who is both a source of alarm and the basis

for safety, intensifies the infant’s fear. This leads to the collapse of behavioural

and regulatory strategies (Lyons-Ruth & Jacobvitz 1999). In contrast to moderate

levels of anxiety and anger that may serve to maintain closeness in anxious

avoidant and anxious resistant attachment relationships, the concurrent activation

of fear and the attachment behavioural system produces strong conflicting

24

motivations for the child exhibiting disorganized behaviours (Carlson 1998).

Infants classified as disorganized in the SST exhibit behaviours that appear to

lack an observable goal or explanation. Most characteristic of these infants is their

lack of a coherent attachment strategy. Disorganized children are those who have

an underlying attachment strategy like A, B, C or A/C, but who are unable to

implement their strategy smoothly and who appear uncertain regarding how the

parent will respond (Crittenden 1995). Longitudinal studies of children classified

as Type D attached in infancy showed associations with defensive aggression

towards peers at age 4 (Jacobvitz & Hazen 1999) and hostile, aggressive

behaviour towards peers at age 5 (Lyons-Ruth et al. 1993). Children with

disorganized attachment history are at particular risk of psychological symptoms

in later childhood. (Carlson 1998).

Research on children's emotional understanding has revealed that insecure

attachment is linked to relative difficulties in interpreting and understanding emotions

(Main et al. 1985, Steele et al. 2008). In their early longitudinal study, Main, Kaplan

and Cassidy (1985) asked 6-year-old children to indicate how a hypothetical child

would respond to an expected 2-week separation from his or her parents. Children

with Type A attachment had difficulty discussing the child's response to the separation

and often refused to discuss possible ways in which children would cope with the

separation. Children with insecure-disorganized attachment tended to become

distressed, silent, self-destructive, and/or irrational when discussing the hypothetical

separations. However, children who had been classified as secure in infancy tended to

give a variety of positive responses, such as that the child should engage in fun

activities to pass the time or should find an alternative attachment figure with whom

to stay. De Rosnay and Harris (2002) reported similar data indicating that lower

attachment security was linked to children's inability to understand emotions in

different situations, including child–parent separations.

2.1.2 The Preschool Assessment of Attachment (PAA)

Crittenden (1992a) argued that the transition from infancy to the preschool-age period

involves concomitant elaboration and differentiation of internal working models of

attachment. Developmentally based change beginning around this time is first the

advent of more complex language, which allows more varied attachment negotiations

and richer, more direct communication about needs and wishes, and second

perspective taking, which enables children to take into account another person's

response and may encourage them to hide their true feelings or thoughts in the service

25

of maintaining tolerable attachment relationships. One system for coding attachment

in children after around 20 months of age is Crittenden's Preschool Assessment of

Attachment (PAA1992a). Crittenden (1992a) described preschool children’s

attachment needs as revolving around the need to communicate rather than simply to

monitor, to have more control, and to be able to negotiate and plan with their

caregivers. The PAA relies on the Strange Situation Test to elicit attachment

behaviour, but Crittenden's classification criteria are in keeping with her proposal

that the preschool child is more sophisticated than the infant at evolving coping

strategies, and that these strategies will both reflect a more complex interpretation

of caregiver behaviour and include a greater diversity of attachment-related

responses. Preschoolers have at their disposal a greater range of behaviours,

including inhibition of feigned affect and behaviour that can be used to achieve a

particular goal; therefore attachment coding must address and focus on the

function of behaviours rather than on the presence or absence of a specific

behaviour.

The PAA classificatory system includes all of the categories of the infant

procedure (i.e. A1-2, B1-4, C1-2, A/C), plus several patterns that develop during or

after the pre-operational shift at the end of the second year of life (Figure1).

Specifically, Ainsworth's three basic strategies for negotiating interpersonal

relationships are modified to fit preschoolers' more sophisticated mental skills and

organization of behaviour. Thus, the patterns are renamed as secure/balanced (Type

B), defended (Type A) and coercive (Type C). (Crittenden 1992b).

26

Fig. 1. Dynamic maturational model (DMM) of patterns of attachment adapted from

Crittenden’s homepages (with permission of the author).

2.2 Concordance of attachment in a family

A meta-analysis of 11 studies evaluating the concordance of infants’ attachment

strategy with their mothers and fathers revealed the pattern of maternal and

paternal attachment to be interdependent, so that security with one parent would

predict security with the other, and correspondingly insecurity with one parent

would predict insecurity with the other (Fox et al. 1991). About half of the

children were assessed as Type B attached to both parents when the evaluations

were made at the age of 11–12 months. In an overview of siblings’ attachment

with their mothers (van Ijzendoorn et al. 2000), sibling relationships were found

to be concordant when classified as dichotomized secure or insecure (62%

concordance) but not when further sub-classified. Maternal insensitivity to both

siblings was related to a concordance in sibling insecurity. Siblings of the same

gender were more likely to form concordant relationships with their mother (68%

concordance). The results of a twin study of attachment in preschool children

indicated that the degree of sibling similarity in attachment was substantial, with

an overall concordance rate of 67% at the secure/insecure level (O’Connor &

A dynamic‐maturational modelof patterns of attachment in the preschool years

True cognition

Distorted cognition &Omitted negative affect

False positive affect

Integrated true information

True negative affect

Distorted negative affect &omitted cognition

False cognition

A3‐4CompulsivelyCaregiving/Compliant

A1‐2Inhibated/Socially facile

B1‐2Reserved

B3Comfortable

B4‐5Reactive

C1‐2Threating/Disarming

C3‐4Aggressive/

Feigned Helpless

Copyright: Patricia M. Crittenden, 2014

A/C

27

Croft 2001). However, these studies have not given attention to parental

preference of infants, a phenomenon that is often seen in twin families (Moilanen

& Pennanen 1997, Trias et al. 2006).

2.3 Attachment patterns in different countries

Earlier research on attachment has focused on normative samples from North

America and has later expanded to other cultural contexts. Although Type B is

usually the most common type of attachment, there are many types of insecurity.

These may indicate non-normative relationship patterns but are not necessarily

pathological. Within one cultural context, insecure attachment may signify

insecurity, whereas in others it may indicate behaviours that are appropriate and

expected. Intercultural differences are also seen within the different expressions

of caregiver sensitivity, which may affect children’s expressions of security and

insecurity (Mizuta et al. 1996). In a meta-analysis of attachment patterns (van

Ijzerndoorn & Kroonenberg 1988) elicited in almost 2000 “Strange Situations” in

eight different countries, 65% of the children were classified as Type B, 20% as

Type A, and 14% as Type C. The data suggested a pattern of cross-cultural

differences, in which Type-A classifications emerged as relatively more prevalent

in the Western European countries and Type-C classifications as relatively more

frequent in Israel and Japan. Analysis of attachment distribution in a Swedish

sample revealed that 62% of infants were Type-B-attached, 16.5% were classified

as insecure-avoidant and 21% as insecure-ambivalent (Bohlin & Hagekull 2000).

In Finland there have been no studies of healthy children from ordinary families

before this current study.

2.4 Internalizing and externalizing problems in childhood

The majority of child development and mental health researchers agree on a

classification of behavioural problems that distinguishes between internalizing and

externalizing manifestations of dysfunction (Cicchetti & Toth 1991). Internalizing

behaviour problems defined as an over-control of emotions include social withdrawal,

low demand for attention, feelings of worthlessness or inferiority, and dependency

(Achenbach & Edelbrock 1978, McCulloch et al. 2000). Conversely, externalizing

behaviour problems characterized by an under-control of emotions include difficulties

with interpersonal relationships and rule breaking as well as displays of irritability

and belligerence (Achenbach & Edelbrock 1978).

28

There is evidence that preschool externalizing problems can predict later

internalizing problems (Lavigne et al. 1998, Mesman et al. 2001). In a Finnish

population-based birth cohort (Pihlakoski et al. 2006), children’s emotional and

behavioural problems were assessed at age 3 using the Child Behavior Checklist

(CBCL, Achenbach 1991a) and at age 12 by parents using the CBCL and by the

children themselves with the Youth Self Report (YSR, Achenbach 1991b).

Behavioural ratings were obtained from 800 subjects at both time points.

Problems at age 3 predicted both externalizing and internalizing problems among

both genders in preadolescence. In girls, internalizing behaviour at age 3

predicted parent-reported internalizing behaviour, and correlated negatively with

externalizing behaviour in self-reports at age 12. Aggressive behaviour was found

to be remarkably stable, and both aggressive and destructive behaviour

independently predicted a wide range of later externalizing and internalizing

problems. Parent ratings of withdrawal showed stability from early childhood to

preadolescence in both genders. Somatic complaints of 3-year-old girls

independently predicted both externalizing and internalizing problems according

to parent and self-reports. (Pihlakoski et al. 2006).

In a study conducted in the United States the level of parent-reported

internalizing problems increased from ages 5 to 10 years for both boys and girls (Leve

et al. 2005). Similarly in a study conducted in the Netherlands, internalizing problems

increased from age 4 years until early adolescence, with the number of internalizing

problems being similar for boys and girls (Bongers et al. 2003). Gilliom and Shaw

(2004) found increasing levels of internalizing problems in children from ages 2 to 6

years. Conversely, Keiley et al. (2000) reported stable levels of problems from ages 5

to 13 years for both boys and girls according to mothers and teachers. One recent

study, in which trajectories of children’s internalizing symptoms were examined as

predicted by interactions among maternal internalizing symptoms and children’s

physiological regulation, showed that trajectories decreased from ages 8 to10 years

(Wetter & El-Sheikh 2012).

Several child and family characteristics have been found to be associated with

internalizing problems in childhood, such as parental divorce (Amato 2000,

Hetherington & Stanley-Hagan 1999), children’s shy temperament and behavioural

inhibition (Feng et al. 2008, Hirshfeld-Becker et al. 2008, Muris et al. 2011),

maternal depression (Feng et al. 2008, Sterba et al. 2007), unfavourable parenting

styles (Luyckx et al. 2011) and environmental stress (Grant et al. 2003, van Oort

2010).

29

Boys appear to have higher mean levels of externalizing problems than girls

(Bongers et al. 2003, Keiley et al. 2000, Leve et al. 2005, Miner & Clarke-Stewart

2008, Stanger et al. 1997). Leve et al. (2005) showed that parent-rated externalizing

problems increased from age 5 to 17 years for both boys and girls. Miner and Clarke-

Stewart (2008) in turn showed that the level of parent- and teacher-rated externalizing

problems declined between the ages of 2 and 9 years. In a study by Keiley et al.

(2000), maternal-rated externalizing problems decreased, whereas teacher-rated

externalizing problems increased from age 5 until 13 years.

Family characteristics that are associated with externalizing problems include

parental divorce (Amato 2000, Hetherington & Stanley-Hagan 1999) and

unfavourable parenting styles such as rejection (Luyckx et al. 2011, Shaw et al. 2003,

Shaw et al. 2005).

2.5 Attachment and cognitive development

Securely attached children are expected to perform better in problem-solving

situations compared to insecurely attached children. Longitudinal studies have

shown that infants who are securely attached to their mothers are more curious

and more active than insecurely attached children in exploring the environment as

toddlers (Hazen & Durret 1982). They are also more eager to learn (Wilson 1974)

and show better self-esteem (Jacobsen & Hofmann 1997) than children with

insecure attachment. There is also evidence that maternal attachment is related to

better children's language development (Klann-Delius & Hofmeister 1997, van

Ijzendoorn et al. 1995, Belsky & Fearon 2002), preferable cognitive development

(De Ruiter & van Ijzendoorn 1993, van Ijzendoorn et al. 1995), and better

emergent literacy (Bus & van Ijzendoorn 1988). West, Mathews and Kerns (2013)

reviewed the literature and suggested that attachment may also be related to academic

performance and IQ in middle childhood. They investigated relations of early

mother–child attachment and cognitive performance in middle childhood. Mother–

child attachment was assessed at 15, 24, and 36 months, and child grades and IQ were

assessed at grades 3 and 4. Attachment patterns at 15 months and avoidant attachment

at 36 months were not related to school performance or IQ in middle childhood.

Securely attached children at 24 or 36 months had better school performance and

higher IQs in middle childhood compared to insecurely attached children.

30

2.6 Attachment and mental well-being

In the 2010 decade three meta-analyses concerning the connection between early

attachment and later mental health have been published. First, Fearon et al. in

2010 reported studies in which insecure (A, C) and disorganized attachment types

increased the risk of externalizing problems (N = 5 947). In 69 samples the

association between insecurity and externalizing problems was significant. More

pronounced effects were found in boys, clinical samples and observation-based

outcome assessments. In a review of 55 studies Brumariu and Kerns also in 2010

concluded that the link of insecure attachment to internalizing problems is clear

especially to anxiety and depression. Two years later Groh et al. (2012) in a meta-

analytic review examined the association between attachment and internalizing

symptoms during childhood. Based on 42 independent samples (N = 4 614), the

association between insecurity and internalizing symptoms was significant. Type

A attachment strategy was significantly associated with internalizing symptoms.

Insecurity and disorganization more strongly predicted externalizing than

internalizing symptoms. It is important to notice that Brumariu and Kerns (2010)

had a broader focus on attachment assessed by different measures (behavioural

representational and self-report) from early childhood to adolescence, whereas

Groh et al. (2012) included in their review only studies with early behavioural

measures of attachment.

The protective or risk effect can be seen in children and families especially in

situations having contextual risks, such as low socio-economic status and low

social support, maternal depression or poor marital quality. Using maternal CBCL

questionnaires, in 2002 Belsky and Fearon reported a strong step up in

behavioural problems in the avoidant group of preschool-aged children having

two contextual risks. The other attachment groups reached similar levels of

behavioural problems only at higher levels of risk (i.e. 3+). Previously Shaw et al.

(1995) reported on families with socioeconomic risks in the USA and concluded

that children with insecure attachment status at 18 months had higher scores at

age three in externalizing and internalizing symptoms (measured by using CBCL

questionnaires). When the securely attached group was compared separately with

each insecure group, no major effects of attachment were found. Based on the

same sample Shaw et al. (1996) noticed that disorganized attachment related to

internalizing scores at age five years.

31

2.7 The context of twins

In Finland as in other developed countries, the proportion of twins among

newborn babies is increasing; 1.1% of births were twin deliveries in 1987 and

1.4% in 2013 (Official Statistics of Finland, Terveys 2014). The number of twin

births in the United States has increased from 1.9% in 1980 to 3.3% in 2009 - a

relative increase of 76% (Martin et al. 2012). Data from other industrialized

countries, including Australia, Austria, Japan, Norway, Singapore and Sweden,

have also shown increases in the rate of twinning (Imaizumi 1997). There are

multiple causes for these increased rates of twin pregnancies. Increasing maternal

age is positively correlated with the dizygotic (DZ) twin rate, but not with

monozygotic (MZ) twinning (Keith et al. 2000). In Finland in 2001 the mean

maternal age at first delivery was 27.6 years and among all deliveries 29.9. In

2013 the corresponding figures were 28.6 and 30.4 (Official Statistics of Finland,

Väestö 2014). The putative reason for the increased rate of twinning with advanced

maternal age is that under equal ovarian feedback conditions, premenopausal mothers

of DZ twins have hyperstimulation by endogenous follicle-stimulating hormones

caused by neuroendocrine, hypothalamic, or pituitary mechanisms (Lambalk et al.

1998). The other reason for increasing twin deliveries is the use of assisted

reproductive techniques (ARTs) and non-ART procedures (Artificial

insemination�and hormones to stimulate egg production). As ART traditionally

involved ovarian stimulation and replacement of more than one embryo, it

contributed to a significantly higher number of multiple births (Eriksson &

Fellman 2007). The multiple birth rate from 376 971 European IVF treatment

cycles in 2007 was reported as 22.3% (21.3% twins and 1% triplets) (de Mouzon

et al. 2012).

MZ twins are formed when the fertilized egg splits into two, whereas DZ twins

are produced because two eggs are released at the same time and these both become

fertilized. MZ twins share 100% of each other’s genes, whereas DZ twins share 50%.

These two different types of twins allow geneticists to compare the behavioural

resemblance of MZ twins (with a genetic relatedness of 1.0) to that of DZ twins (with

a genetic relatedness of 0.5). This means that if heredity affects a trait or behaviour,

identical twins should show a greater resemblance for that trait than fraternal twins.

However, if the environment affects behaviour, differences should be minimal. This is

measured by concordance (Collins & Sullivan 2013).

As compared to singleton pregnancies, twin pregnancies have contributed

significantly to preterm deliveries. Preterm delivery (< 37 weeks) occurs in well

32

over half of all twin pregnancies (Spellacy et al. 1990, Moise et al. 1998, Ananth

et al. 2005, Chauhan et al. 2010, Burton & Ananth 2010). In Finland in 2013,

4.51% of the 56 878 singletons and 48.91% of the 1 564 twins were preterm

(THL unpublished statistics). In 2009, of the 137 217 twins that were delivered

in the United States, approximately 59% were preterm and 10% were delivered at

< 32 weeks. (National Vital Statistics Reports 2013).

Twins are more often small for their gestational age compared to singletons

(Rydhstroem & Heraib 2001). Increased levels of problems in singleton samples of

preterm children and children with low birth weight have been reported, such as

ADHD (Lahti et al. 2006) and depression (Räikkönen et al. 2008). Low birth weight

and prematurity are unlikely to have the same significance in twins as in singletons,

because of their different aetiology (Phillips et al. 2001). Twins become growth-

retarded as a result of poor early placental development (Bleker et al. 1988), as

indicated by the fact that from about 24 weeks gestational age onwards, placental

weights of twins are lower than those of singletons. Nonetheless after birth, twins

generally grow well compared to singletons, and are almost as tall as singletons by

the age of 5 years (Estourgie-van Burk 2006).

Multiple pregnancies are associated with a higher rate of spontaneous abortion

and intrauterine foetal demise, and are more likely to result in neonatal and infant

death, cerebral palsy, and congenital birth defects (Dhont et al. 1999). Women who

had multiple births are at increased risk of depression and of experiencing higher

parenting stress (Thorpe et al. 1991). Furthermore, the economic effect on society and

the economic and emotional stresses on families that are associated with raising

twins, triplets, and more children are becoming increasingly apparent (FIGO 2001,

Kinzler, Ananth & Vintzileos 2000). These effects could be associated with higher

levels of emotional and problem behaviours in twins than in singletons. Twins are

obliged to share parental attention with their co-twin and to enter into an inter-

twin relationship that affects their personality, especially on account of inter-twin

dependence or dominance-submissiveness in relationships (Moilanen 1987,

Moilanen & Ebeling 1998). Leonard (1961) even suggested that during early

childhood, twins may identify with their co-twin rather than with an adult. The

strength of this twin-pair tie was demonstrated in a study by Fraley and Tancredy

in 2012, in which twin siblings were more likely than non-twin siblings to be

attached to their siblings.

Studies comparing levels of behavioural and emotional problems of twins with

those of singletons have had mixed findings. In the National Epidemiological Child

Psychiatric Study in Finland, among 8–9 year-old children, according to parent’s

33

reports (using the Rutter Parent Questionnaire, Rutter et al. 1970), twin girls

compared to singletons scored significantly lower as regards total and emotional

problems reported by parents. In self-reports depressiveness did not differ

between twins and singletons (Moilanen et al. 1999). Fewer externalizing problems

(i.e. disruptive, hyperactive, and aggressive behaviors) were found in 2- and 3-year-

old Dutch twins compared to same aged singletons (van den Oord et al. 1995). In

Norwegian children aged 5 to 15 years the level of internalizing problems (i.e.

anxiety, depression, withdrawn behaviour, and functional somatic complaints) was

also lower in twins than in singletons (Gjone & Novik 1995). Similarly in a recent

American study of 6–12-year-old ART children (N = 124 twins and 176 singletons)

from 206 families, twins had significantly fewer behavioural problems than

singletons, as reported by the parents (Child Behavior Checklist, CBCL). These

results also suggest that full-term American twins had significantly fewer attention

problems than premature twins, full-term singletons and premature singletons. (Kayla

et al. 2014). Gau et al. (1992) in turn reported maternal CBCL ratings for 1 824

American twins, compared with the CBCL normative sample. The results

indicated that twins showed slightly but consistently higher levels of problem

behaviours. These elevations were significant for older children in both

internalizing and externalizing behaviours; for younger children the elevations

were significant for externalizing but not internalizing behaviours. Maternal

CBCL were also obtained for a sample of 9 651 twins from the Netherlands Twin

Register and for a representative general population sample of 1 351 singletons.

The developmental trajectories of externalizing problems showed a linear

decrease over time, and were not significantly different for twins and singletons.

Internalizing problems also appear to develop similarly for twins and singletons

up to age 9. After this age, internalizing symptoms in twins begin to decrease in

comparison to those of singletons, resulting in less internalizing problems than in

singletons by the age of 12 years. (Robbers et al. 2010). In a study among 11- and

12-year-old Finnish twins and singletons (Pulkkinen et al. 2003), there were no

significant differences in externalizing or internalizing problems. Similarly a

Norwegian sample of 5- to 15-year-old children (Gjone & Novik 1995) showed no

differences in externalizing problems between twins and singletons.

In studies from the 1960s the cognitive development of twins has been shown

to be slower than that of singletons during early childhood (Husen 1961, Dales

1969). In studies from the next decades (Alin Åkerman & Fischbein 1991, Hay et

al. 1984, Myrianthopoulos 1976), the mean cognitive performance of twins was

also about five IQ points (one third of a standard deviation) below that of

34

singletons in reasoning tests. In a later report of a large population-based study

comparing intelligence quotients (IQs) of twins and singletons in Scotland (Ronalds

et al. 2005), in which normal birth weight and term infants predominated, 2.6 points

higher rates were obtained for singletons. Webbink et al. (2008) compared

cognitive skills in school-aged singletons and twins in the Netherlands and found

a small but significant cognitive deficit for twins aged 6 and 8. However, the

difference disappeared by the time the children were 12. Similarly in a UK cohort

of 11-year-olds (175 323 singletons and 3 188 twins), there were no significant

differences in cognitive abilities (Calvin et al. 2009). Some studies show no

significant differences in physical and mental development between twins and

singletons except that twins might have a tendency towards slow language

acquisition (Chaudhari et al. 1997, Ozcakar et al. 2003). On average, twins have

scored lower than singletons on a range of tests of verbal competence, and this effect

has been more evident in male twins (Hay et al. 1987, Rutter et al. 2003). Twin

couples may develop their own autonomous or private language that might impede

normal language development (Bakker 1987, Rutter et al. 2003). In addition in the

case of twins, the mother must simultaneously meet the needs of two infants. Hence,

twins almost always receive less individual attention from their parents than singleton

children, which might have an impact on verbal development (Lytton 1977).

Birth weight and gestational age (GA) are the most important covariates that

might confound or modify the prevalence of neurodevelopmental disabilities in twins

and singletons (Lorenz 2012). Gender also has an important effect on

neurodevelopmental outcome, with females in general having better outcomes than

males for both twins and singletons. (Grether et al. 1993, Bonellie et al. 2005).

Intrauterine growth restriction (IUGR) is more common in twins than in singletons

and has been reported to be associated with increased risk of neurodevelopmental

disability in both singletons and twins (Bonellie et al. 2005, Williams & O’Brien

1998).

Parental education and socioeconomic status have major effects on cognitive

outcome (Lorentz 2012). In some studies, parental education and socioeconomic

status are higher in twins than in singletons (Cincotta et al. 2000, Ronalds et al.

2005). Maternal age at delivery is often higher for twins than for singletons and this

has been reported to be associated with better cognitive function in singletons and in

twins (Ronalds et al. 2005). Zazzo et al. (1979) showed the cognitive deficit in twins

compared with singletons to be similar to that in closely spaced singletons, suggesting

that the change in family dynamics caused by twins may contribute to cognitive

discrepancies between twins and singletons.

35

3 Aims and hypotheses of the study

The objectives of this study were as follows:

1. to examine the attachment of 18 months old Finnish singletons and twins to

their mothers and fathers (I and II).

2. to examine cognitive skills in the same cohort at 36 months of age and to

compare the results with the recorded attachment type at 18 months

(unpublished data).

3. at 48 months to relate mental well-being of the children as reported by their

parents to the 18-month attachment result (III).

The hypotheses of this study were as follows:

1. Because of Finland’s Northern European cultural heritage, its sparse

population and the policy of maternal employment, we expected normative

Finnish children to display a distributional bias toward Type A attachment

strategy (I and II). Because of egalitarian values of Finland, we expected no

gender differences (I and II).

2. We expected fewer Type B attached twins than singletons, because twins are

obliged to share parental attention with their co-twin (I and II), except for

cases in which twins also experienced preterm delivery and neonatal

morbidity with hospital treatment, which increase the probability for secure

attachment because of the need for special parental attention and support (I)

3. Because parents of Type A attached children are likely to guide their children

towards play supporting academic skills, we expected that Type A attachment

pattern would lead to better results in the intelligence test compared to Type

C or IO attached toddlers. In view of the biological and environmental risks,

we expected that singletons would perform better compared to twins in the

intelligence test (unpublished data)

4. In the light of Finnish culture and education ideals of independence, we

expected that Type A attached children would show no more mental

symptoms at preschool age than securely attached children (III).

36

37

4 Materials and methods

4.1 Participants

This is a prospective, longitudinal follow-up study of twins and singletons and

their parents from Oulu, Finland, focusing on early attachment and later

development and mental health.

The twin sample was drawn from consecutive twin deliveries in 1992–1994

at the University Central Hospital (UCH) of Oulu. A group of singletons of the

same sex as the firstborn twin were selected for twin-singleton comparisons. In

all, 30 singleton families (SF) and 30 twin families (TF) took part in the study

(Table 1). In order to be eligible for the study the infants had to be the firstborn in

their families and both infants of the twin pair had to survive. All the infants were

free of major malformations. The singletons were born during gestational weeks

33–42 (mean 40, SD 2.34) and the twins during gestational weeks 31–40 (mean

37, SD 2.47). The difference between the duration of pregnancy of singletons and

twins was significant (p < 0.001). The age of mothers of the singletons ranged

from 22 to 35 years (mean 27.8, SD 2.8), and the age of the fathers from 25 to 41

years (mean 30.3, SD 3.7) The age of the mothers of the twins ranged from 22 to

37 years (mean 30.0, SD 3.9), while the fathers fell in the age range 24–45 years

(mean 32.8, SD 5.59). The difference between the age of mothers of singletons

and twins was significant (p = 0.007), but not significant among fathers.

Eight of the twin pregnancies were initiated using assisted reproductive

techniques (ARTs), as reported by the parents. At 18 months of age zygosity was

determined by a questionnaire using questions on physical similarity, a method

which has shown high reliability in Finnish twin data (Sarna et al. 1978). On the

basis of this estimation only one pair was MZ.

The families were categorized by their socio-economic status into social

classes I (professional, managerial), II (intermediate occupations), III (skilled

occupations), and IV (unskilled occupations, housewives, retired) according to

the parent who had the higher socio-economic status (Alestalo 1978). Students

were classified into the class they would belong to after graduation. The majority

(71%) of the families were assigned to the social classes II-III (Table 1). All the

infants were of Caucasian ethnic background, except for one twin pair which had

an African father

38

Table 1. Characteristics of the sample (N = 30 singletons and 60 twins).

Characteristics Singletons Twins

n 27 % Mean SD n 60 % Mean SD

Boys 16 59.3 26 43.3

Girls 11 40.7 34 56.7

Gestational weeks 40.0 2.3 37.0 2.5

Mother's age1 27.8 2.8 30.0 3.9

Father's age1 30.3 3.7 32.8 5.6

SES of the family1

I 5 18.5 8 27.6

II 12 44.4 11 37.9

III 9 33.3 10 34.5

IV 1 3.7 0 0

Preterm2 3 11.1 24 40.0

Neonatal hospital treatment 4 14.8 24 37.9

Primary caretaker

Mother 11 45.8 11 21.2

Both 12 50.0 34 65.4

Father 1 4.2 7 13.5

1 according to the parent with higher socioeconomic status, 2 < 37 gestational weeks

4.2 Procedure

The study was approved by the Ethics Committee, Medical Faculty, University of

Oulu, Finland.

The subjects were identified by the Obstetric Department of the Central

University Hospital of Oulu. 30 consecutive twin deliveries were selected (no

parents refused to participate), and a group of singletons of the same sex as the

firstborn twin were invited for twin-singleton comparisons (two couples of the

invited parents refused to participate and were replaced by other parents of

singletons). In order to be eligible for the study the infants had to be the firstborn

in their families and both infants of the twin pair had to survive. (Table 2)

The hospital records were examined and the following indicators were noted:

ages of mother and father at delivery, length of gestation, pregnancy and delivery

complications, neonatal hospital treatment and social status of the family.

Altogether, 40 boys (24 twins, 16 singletons) and 44 girls (33 twins, 11

singletons) were assessed in the Strange Situation Test (SST) at the age of 18

months (corrected for degree of prematurity). There were 56 mothers and 54

fathers, owing to the fact that one of the twin pairs and one of the singletons came

39

from a single-parent family, and thus these three children were assessed only with

the mother. Two fathers (one father of singletons and one of twins) refused to

participate in the study, and the parents of one of the twin pairs refused to

participate with one of their twins after considering the situation to be too

distressing for their child. Taking these cases into account, a total of 84 mother-

infant (57 mother-twin and 27 mother-singleton) and 78 father-infant assessments

(53 father-twin and 25 father-singleton) were conducted. Assessment of

attachment took place in a hospital laboratory playroom equipped with a one-way

mirror, microphones and a video recorder. All attachment classifications were

assessed from videotapes by the author according to the Preschool Assessment of

Attachment (PAA) developed by Crittenden (1992a, 1992b).

The sample of the first original publication (I) consisted of 47 mother–twin

and 18 mother-singleton SSTs, altogether 36 girls (8 singletons, 28 twins) and 29

boys (10 singletons, 19 twins). In this analysis the ages of mothers of twins

ranged from 22 to 37 years (M = 29.9) and mothers of singletons from 22 to 35

(M = 27.8). Forty-five per cent of twins (21/47) were preterm (< 37 gestational

weeks) and of singletons 17% (3/18). Thirty-six per cent of twins (17/47) and

17% (3/18) of singletons were treated in the neonatal ward.

The intelligence of the children (52 twins and 27 singletons) was measured

by a clinical psychologist using the Stanford-Binet test (Thorndike, Hagen &

Sattler 1986) at an average age of 36 months corrected by the degree of maturity.

This test is often used to evaluate the intelligence of preschool-aged children. The

abilities measured by the test items include general comprehension, visual-motor

ability, arithmetic reasoning, memory and concentration, vocabulary and verbal

fluency, judgement and reasoning. Behaviour in the testing session was assessed

on a five-point scale. The observed traits were concentration on the task, interest

in the task, endurance in difficult tasks, self-confidence, working speed and

verbal expression. (Unpublished data).

Later, when the children were four years old, the Child Behavior Checklist

(CBCL) questionnaires were posted to both parents with the instruction to fill in

the CBCL independently. After two notifications the mothers of 22 singletons and

54 twins and the fathers of 17 singletons and 60 twins completed the

questionnaires. (III)

40

Table 2. Study population (N = 30 singletons, 60 twins).

Age Months Measure Participating parent Participants Attrition

60 twins 30 singletons Twins Singletons

18 SST1 Mother 57 27 3 3

Father 53 25 7 5

Mother or father 60 27 0 3

18 Parental Questionnaire Both parents 58 27 2 3

36 IQ2 52 27 8 3

48 CBCL3 Mother 54 22 6 8

Father 60 17 0 10

1 Strange Situation Test, 2 Intelligence Quotient, 3 Child Behavior Checklist

4.3 Methods

4.3.1 Parental Questionnaire at 18 months

At 18 months of age, corrected by the degree of prematurity, a questionnaire was

sent to both parents about their children’s health, development and temperament,

asking about their own education and employment, as well as some facts of

family life – for example family pattern, the primary caretaker of the child, or

whether the parents had shared caretaking responsibility for their infants.

4.3.2 Strange Situation Procedure and Preschool Assessment of

Attachment (PAA)

The most common procedure for identifying attachment patterns (secure, avoidant,

ambivalent/coercive and disorganized) in infancy, the Strange Situation Procedure

(Ainsworth & Bell 1970), was used in this study. The procedure consists of a series of

separation and reunion episodes with a caregiver in laboratory conditions. In the 20 -

minute procedure, the caregiver and the infant are placed in an unfamiliar playroom

equipped with toys while a researcher observes/records the procedure through a one-

way mirror. The procedure consists of eight sequential episodes in which the child

experiences both separation from and reunion with the mother as well as the presence

of an unfamiliar stranger.

41

The protocol is conducted in the following format according to Ainsworth and

Bell 1978:

– Episode 1: Caregiver, Child, Experimenter (30 seconds)

– Episode 2: Caregiver, Child (3 minutes)

– Episode 3: Caregiver, Child, Stranger (3 minutes)

– Episode 4: Stranger, Child (3 minutes or less)

– Episode 5: Caregiver, Child (3 minutes)

– Episode 6: Child Alone (3 minutes or less)

– Episode 7: Stranger, Child (3 minutes or less)

– Episode 8: Caregiver, Child (3 minutes)

All attachment classifications were assessed from videotapes by the author

according to the PAA method developed by Crittenden (1992a, 1999b). The

author, who served as the rater, found 91% inter-rater reliability in the

standardized reliability test and 94% agreement with Crittenden in assessing 16

tapes randomly selected from the data set (I, II).

4.3.3 Stanford-Binet Intelligence Scale (SB)

The intelligence level of the children was measured at an average age of 36 moths

using the Stanford-Binet Intelligence Scale (SB), which was developed in 1916 and

was revised in 1937, 1960, 1986 and 2003 (Bain & Allin 2005). In the present study

the intelligence of the children was measured by the highly experienced leading

clinical psychologist at the Department of Pediatrics, University Hospital of Oulu

using the 4th version of SB (SB- IV) at an average age of 36 months corrected by

the degree of maturity. SB-IV (SB-IV; Thorndike, Hagen & Sattler 1986) has been

widely accepted and included among tests endorsed for assessment of cognitive

ability in preschool children. Concurrent validity studies suggest that SB-IV is likely

to yield composite scores similar to those provided by other acceptable measures of

cognitive functioning, such as the WISC-R, WAIS-R, and Form L-M (Hollinger &

Baldwin 1990).

4.3.4 Child Behavior Checklist (CBCL)

The Child Behavior Checklist is a parent-report measure that has been widely

used internationally.

42

The CBCL/4–18 was normed on a sample of 2 368 non-handicapped 4 to 18

year old American children (Achenbach 1991). The CBCL has strong content and

construct validity and high test–retest reliability (r = 0.91–0.95) (Achenbach &

Rescorla 2001). In the present study mothers and fathers evaluated their

children’s behavioural and emotional problems by using CBCL questionnaires for

ages 4–6, modified and translated for Finnish pre-schoolers by Almqvist 1996. In

this modified version school-words are changed to kindergarten-words, because

in Finland children start school at age seven. Parents rate their children’s

problematic behaviours and competencies by completing 20 items assessing

competencies and 113 items assessing behavioural/emotional problems during the

previous 6 months. Items are rated on a 3-point Likert-type scale (i.e. 0 = not true;

1 = somewhat or sometimes true; 2 = very often true or often true). Normal,

borderline, and clinical ranges are defined in relation to the scores. There are

subscales for the problem items that are used to evaluate internalizing (i.e.

withdrawal, somatic complaints, and anxiety/depression) and externalizing

symptoms (i.e. aggressive and delinquent behaviour) as well as social and thought

problems. (Achenbach 1991).

4.3.5 Statistical methods

Analyses were performed using The Statistical Package for Social Sciences

(SPSS) version 7.0 (I), 8.0.1 (II) and 22.0.0.0 (III). Pearson’s Chi-Square tests and

Fisher’s Exact tests were used to compare children’s attachment type to mothers

and fathers between twins and singletons, boys and girls as well as mothers and

fathers as the primary caretaker of the child. The Mann-Whitney U -test was used

to study the gender- and attachment-type differences in CBCL because of non-

normally distributed data and the small size of the sample. Differences between

IQ scores among attachment types to mothers and fathers were analysed using the

Kruskal-Wallis test. The influences of socioeconomic status (SES) and

prematurity on IQ were analysed by using two-way ANOVA, with the IQ score as

response variable and SES and prematurity as fixed explanatory variables.

Statistical significance was evaluated using two-tailed 0.05-level tests. No

multiple testing corrections were made, based on a suggestion by Rothman

(1990).

43

5 Results

5.1 Attachment of singletons and twins to both parents (I, II)

Within the whole sample 37% of the children were scored Type A, 35% as Type

B, 20% as Type C, and 8% outside the normal range (IO), when assessed

attachment strategy with the mother. When the children were assessed with their

fathers, 36% Type A, 33% Type B, 28% Type C and 3% outside the normal range.

There was no significant difference between the distributions of attachment to the

father and the mother. Attachment to mother and father in the whole Finnish

sample is presented in Figure 2, along with the results of an international meta-

analysis from eight countries (van Ijzendoorn & Kroonenberg 1988).

Fig. 2. Attachment (%) to mother and father in the whole Finnish sample (N = 84) and

in the international sample (N = 1 990) from eight countries (van Ijzendoorn &

Kroonenberg 1988).

The attachment of singletons and twins to the mother and father was analysed

separately, showing twins to be more often securely attached to the mother than

singletons (p = 0.013). By contrast, attachment to the father did not differ

significantly (p = 0.680) between singletons and twins (Figures 3–6).

3735

20

8

3633

28

3

20

65

14

0

10

20

30

40

50

60

70

80

A B C IO

Finnish sample,attachment to mother

Finnish sample,attachment to father

International sample,attachment to mother

44

Fig. 3. Sub-classifications of attachment type with the mother in singletons at 18

months.

Fig. 4. Sub-classifications of attachment type with the mother in twins at 18 months.

15 %

18 %

7 %

19 %

4 %

4 %

33 %

B1‐2

C1‐2

C4

IO

A4

A3

A1‐2

28 %

3 %

12 %

16 %

2 %4 %

35 %

B1‐2

B3

B4

C1‐2

C4

IO

A1‐2

45

Fig. 5. Sub-classifications of attachment type with the father in singletons at 18

months.

Fig. 6. Sub-classifications of attachment type with the father in twins at 18 months.

16 %

12 %

24 %

4 %

44 %

B1‐2

B4

C1‐2

C4

A1‐2

19 %

17 %

21 %2 %

5 %

4 %

32 %B1‐2

B4

C1‐2

C3

C4

IO

A1‐2

46

In Tables 3 and 4 the sub-classifications of attachment of both girls and boys and

singletons and twins to the mother and to the father are shown separately. Gender

did not statistically significantly affect the pattern of attachment in the whole

group, although 43% (19/44) of the girls and only 25% (10/40) of the boys were

Type B attached to their mothers. The main difference was found between

singleton and twin girls, as only one of the 11 singleton girls (9%) was Type B

attached to the mother, whereas 18 of the 33 twin girls (55%) were Type B

attached (p = 0.02). No such difference was observed between singleton and twin

boys. Although the comparison of the attachment in twin boys and girls to the

mother revealed 55% (18/33) of the girls, and 29% (7/24) of the boys to be Type

B, this difference did not reach statistical significance. Gender did not affect the

twins’ attachment to their fathers, or the singletons’ attachment to either their

mothers or fathers.

The relationship between the three biological risk factors of newborns were

analyzed in relation to attachment with the mothers, in the combined group of both

singletons and twins. These risk factors were twinship, prematurity, and neonatal

morbidity in such an extent that hospital treatment was needed. The more risk factors

the infant had encountered, the higher was the probability that the toddler was

classified as Type B attached. Attachment type in relation to cumulative risk, as

number of risks, is presented in publication I Table 8.8.

47

Ta

ble

3. M

ain

an

d s

ub

-cla

ss

ific

ati

on

s o

f a

tta

ch

me

nt

typ

e t

o t

he

mo

the

r in

sin

gle

ton

s a

nd

tw

ins

.

Cla

ssifi

catio

n A

A

A

4 A

3 A

1-2

B

B

B

1-2

B3

B4

C

C

C

1-2

C3

C4

IO

IO

D

/A/C

A/C

Tot

al

Tot

al

N

%

N

N

N

N

%

N

N

N

N

%

N

N

N

N

%

N

N

N

%

Sin

glet

ons

11

40%

4

15%

7

26%

5

19%

- G

irls

0

0 5

1

0 0

2

0 1

1

1

11

41%

- B

oys

1

1 4

3

0 0

3

0 1

3

0

16

59%

- T

ota

l

1

1 9

4

0 0

5

0 2

4

1

27

100%

Tw

ins

20

35%

25

43

%

10

18%

2

4%

- G

irls

0

0 10

10

1 7

3

0 1

1

0

33

58%

- B

oys

0

0 10

6 1

?

6 0

0

1 0

24

42

%

- T

ota

l

0

0 20

16

2 7

9

0 1

2

0

57

100%

Tot

al

31

37%

29

35

%

17

20%

7

8%

with

mot

her

1 1

29

20

2

7

14

0 3

6

1

84

100%

Ta

ble

4. M

ain

an

d s

ub

-cla

ss

ific

ati

on

s o

f a

tta

ch

me

nt

typ

e t

o t

he

fath

er

in s

ing

leto

ns

an

d t

win

s.

Cla

ssifi

catio

n A

A

A

4 A

3 A

1-2

B

B

B

1-2

B3

B4

C

C

C

1-2

C3

C4

IO

IO

D

/A/C

A/C

Tot

al

Tot

al

N

%

N

N

N

N

%

N

N

N

N

%

N

N

N

N

%

N

N

N

%

Sin

glet

ons

11

44%

7

28%

7

28%

0

0%

- G

irls

0

0 4

0

1 2

4

0 0

0

0

11

44%

- B

oys

0

0 7

2

0 2

2

0 1

0

0

14

56%

- T

ota

l

0

0 11

2 1

4

6 0

1

0 0

25

10

0%

Tw

ins

17

32%

19

36

%

15

28%

2

4%

- G

irls

0 0

9

7 0

5

5 0

2

0 2

30

57

%

- B

oys

0 0

8

3 0

4

6 1

1

0 0

23

43

%

- T

otal

0

0 17

10

0 9

11

1

3

0 2

53

10

0%

Tot

al

28

36%

26

33

%

22

28%

2

3%

with

fat

her

0 0

28

14

0

12

17

1

4

0 2

78

10

0%

48

Among singletons the mother was reported to be the primary caretaker in 46%,

both parents equally in 50% and the father in 4% of cases and the corresponding

figures among twins were 21%, 65% and 14%. (p = 0.058). If the child-mother

attachment pattern was Type B, the primary caretaker among singletons was more

often the mother (67%, 4/6) and among twins both parents equally (73%, 16/22).

Correspondingly, if the attachment strategy with the father was Type B, the

primary caretaker was both equally among singletons (67%, 4/6) and twins (53%,

9/17) (Table 5).

Table 5. Parental preference and attachment to mother and to father.

Primary Caretaker A B C IO Total p

N % N % N % N % N %

Attachment to mother

Singletons

Mother 2 19% 4 36% 3 27% 2 19% 11 100%

Both 7 58% 2 17% 2 17% 1 8% 12 100%

Father 1 100% 0 0% 0 0% 0 0% 1 100% Ns

Twins

Mother 4 40% 4 40% 2 20% 0 0% 10 100%

Both 13 38% 16 47% 3 9% 2 6% 34 100%

Father 2 33% 2 33% 2 33% 0 0% 6 100% Ns

Attachment to father

Singletons

Mother 6 60% 2 20% 2 20% 0 0% 10 100%

Both 3 25% 4 33% 5 42% 0 0% 12 100%

Father 1 100% 0 0% 0 0% 0 0% 1 100% Ns

Twins

Mother 3 37% 4 50% 1 13% 0 0% 8 100%

Both 12 40% 9 30% 9 30% 0 0% 30 100%

Father 0 0% 4 50% 2 25% 2 25% 8 100% Ns

In the whole group, 51% of the children had the same main attachment pattern

with the mother and the father (35% of singletons and 62% of twins).

When the attachment patterns were dichotomized to either secure or insecure,

up to 70% (54/77) of the children had the same attachment quality to both parents

and in 52% (40/77) it was insecure (Table 6). Within the whole twin sample,

inter-twin concordance of attachment to mothers was found to be 52%, whereas

the concordance of attachment to fathers was 46%.

49

Table 6. Attachment to father versus attachment to mother in both twins and

singletons.

Attachment to mother Attachment to father

Insecure (A, C, IO) Secure (B) Total N (%)

N % N % N %

Insecure (A, C, IO) 40 52% 11 14% 51 66%

Secure (B) 12 16% 14 18% 26 34%

Total 52 68% 25 32% 77 100%

5.2 Attachment and cognitive development (Previously

unpublished data)

According to SB-IV the mean IQ at 36 months in this data was very high (M

129.4, SD = 19.0). However, the socioeconomic status of the sample was also

above the average. No statistically significant difference (p = 0.58) was observed

between the mean IQs of singletons and twins (singletons M 132.1, SD = 18.8, A-

twins M 129.3, SD = 18.1, B-twins M 126.8, SD = 20.0) In the testing session

singletons had slightly better mean scores in verbal expression (singletons mean

3.0, SD = 1.0, twins M 2.6, SD = 1.1, p = 0.05).

High intelligence levels of pre-schoolers appear to be clearly related to the

high social status of their families (F (3,72) = 13.8, p < 0.001) as well as to

maturity (F (1,72) = 23.1, p < 0.001). The interaction is however non-significant.

The difference is evident even in this sample, in which the socioeconomic status

is above the average (Figure 7).

50

Fig. 7. The influences of socioeconomic status (SES) and prematurity on IQ in the

whole sample.

Table 7 shows the means and standard deviations of intelligence level in relation

to the main attachment classification with the mother and with the father. Among

Type A and Type B attached children the mean IQs were almost equal, being

clearly higher compared to the Type C children. There were only two subjects

whose attachment was out of the normal range (IO) to the father and they

appeared to have the highest intelligence levels. However, no statistically

significant differences were observed in the mean IQ distribution between the

patterns of attachment to the mother or to the father.

Estimated marginal means of IQs

Estimatedmarginalmean

s160 ‐

150 ‐

140 ‐

130 ‐

120 ‐

110 ‐

100 ‐

IProfessional and

managerial

IIIntermediateoccupation

IIISkilled

occupation

IVUnskilled occupation,housewives, retired

Sosioeconomic status (according to the parent with higher SES)

Non‐estimable means are not plotted

Duration ofpregnancy

<37 weeks>37 weeks

51

Table 7. Means (M) and standard deviations (SD) of IQ per pattern of attachment with

the mother and the father in the whole sample.

Pattern of attachment Mother Father

N M SD N M SD

Type A 29 131.3 21.5 25 131.7 21.3

Type B 23 130.5 16.8 24 130.9 15.4

Type C 16 125.3 20.3 22 125.7 21.7

IO (D/A/C, A/C) 6 126.5 13.5 2 141 24

Total 74 129.4 19 73 129.9 19.5

The pre-term children, 3 singletons (M IQ = 116.3, SD = 7.1) and 24 twins

(M = 122.1, SD = 19.2), had lower intelligence levels at 36 months of age

compared with full-term children (IQ of singletons M = 134.1, SD = 18.9 and of

twins M = 133.1, SD = 17.8), even though the ages were corrected for the degree

of maturity (p = 0.01).

5.3 Attachment and mental well-being (III)

The mothers of singletons reported more anxious/depressed symptoms

(p = 0.049) and aggressive behaviour (p = 0.012) in their children (Figure 8).

Externalizing problems (p = 0.026) compared to twins and the total problem raw

scores were also significantly higher (p = 0.032) among singletons (Table 8).

52

Fig. 8. CBCL syndrome scale raw score medians of singletons and twins reported by

mother.

Table 8. Broad-band scale scores according to the CBCL questionnaire completed by

mothers of singletons and twins.

Singletons

N = 22

Twins

N = 54

p-value for

singletons vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total problems 23 16 33 29.0 16 11 24 18.5 0.032

Internalizing 3 1 8 5.9 3 1 5 3.2 0.068

Externalizing 11 5 15 11.2 6 4 11 7.6 0.026

Medians (md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test

The fathers of singletons also reported more externalizing (p = 0.002) problems in

their children’s behaviour compared to twins. In addition, the fathers reported

more internalizing problems (p = 0.004) and Total problem scores were also

significantly higher (p = 0.001) among singletons (Table 9), although there were

no statistically significant differences in syndrome scale scores (Figure 9).

Symptoms according to the CBCL questionnaire completed by mothers and

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

53

fathers of singletons and twins are also presented in publication III Tables 1a and

1b.

Fig. 9. CBCL syndrome scale raw score medians of singletons and twins reported by

father.

Table 9. Symptoms according to the CBCL questionnaire completed by fathers of

singletons and twins.

Singletons

N = 17

Twins

N = 60

p-value for

singletons vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total Problems 23 20 36 27.6 13 6 24 16.4 0.001

Internalizing 4 2 8 6.2 2 1 4 2.8 0.004

Externalizing 10 8 10 10.6 5 3 10 6.9 0.002

Medians (Md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test.

The medians of CBCL syndrome raw scores of singleton and twin girls as

reported by the mothers are presented in Figure 10 and of singleton and twin boys

in Figure 11. The mothers of singleton girls reported more attention problems

(p = 0.04, Figure 10) in their daughters, and the total problem raw scores were

also higher compared to those of twin girls (p = 0.032, Table 10). Among boys

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

54

there were no significant differences between singletons and twins according to

mothers’ reports (Figure 11).

Fig. 10. CBCL syndrome scale raw score medians of singleton and twin girls reported

by mother.

Table 10. Broad-band scale scores according to the CBCL questionnaire completed by

mothers of singleton and twin girls.

Singletons

N = 9

Twins

N = 30

p-value for

singletons vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total problems 23 16 43 30.4 13 8 24 16.7 0.032

Internalizing 5 2 12 6.9 2 1 4 3.1 0.064

Externalizing 6 5 17 10.9 5 3 11 7.0 0.089

Medians (Md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

55

Fig. 11. CBCL syndrome scale raw score medians of singleton and twin boys reported

by mother.

Table 11. Broad-band scale scores according to the CBCL questionnaire completed by

mothers of singleton and twin boys.

Singletons

N = 13

Twins

N = 24

p-value for

singletons vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total problems 25 16 36 28.1 21 13 26 20.7 0.408

Internalizing 3 1 7 5.2 4 1 5 3.3 0.442

Externalizing 11 6 15 11.4 8 4 11 8.3 0.196

Medians (Md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test

The CBCL syndrome scale raw score medians reported by fathers of singleton

and twin girls are presented in Figure 12 and those of singleton and twin boys in

Figure 13. The fathers of singleton girls reported more aggressive behaviour

(p = 0.016, (Figure 12), internalizing problems (p = 0.035) and externalizing

problems (p = 0.029) compared to twin girls (Table 12).

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

56

Fig. 12. CBCL syndrome scale raw score medians of singleton and twin girls reported

by father.

Table 12. Broad-band scale scores according to the CBCL questionnaire completed by

fathers of singleton and twin girls.

Singletons

N = 7

Twins

N = 34

p-value for

singletons vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total problems 27 18 46 29.9 12 6 26 16.9 0.052

Internalizing 4 2 8 6.3 2 1 3 2.6 0.035

Externalizing 11 10 19 13.1 5 3 10 7.3 0.029

Medians (Md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test

Among boys (Figure 13) the symptoms in the CBCL questionnaire reported by

fathers were stronger in singletons compared to twins in several subcategories:

thought problems (p = 0.007), attention problems (p = 0.020) and aggressive

behaviour (p = 0.022) and also Total problem raw scores and externalizing raw

scores were significantly higher compared those of twin boys (p = 0.005 and

0.023, Table 13).

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

57

Fig. 13. CBCL syndrome scale raw score medians of singletons and twin boys

reported by father.

Table 13. Broad-band scale scores according to the CBCL questionnaire completed by

fathers of singleton and twin boys.

Singletons

N = 10

Twins

N = 26

p-value for

singleton vs

twins

CBCL

Md q1 q3 M Md q1 q3 M

Total problems 22 21 31 26.1 13 7 22 15.8 0.005

Internalizing 4 2 9 6.2 2 1 4 3.1 0.078

Externalizing 10 8 10 8.9 5 2 9 6.4 0.023

Medians (Md), first quartiles (q1), third quartiles (q3) and means (M) are given, and the significances of

differences in broad-band scale scores between singletons and twins were calculated using the

nonparametric Mann-Whitney U-test

The syndrome scale scores reported by mothers according to different attachment

types with the mother are presented in Tables 14 and 15. Among singletons, Type-

A-attached toddlers had the highest withdrawn sub-score. Significantly higher

scores were also found with regard to somatic complaints and total internalizing

problems. However, no such associations were found among twins. With regard

to CBCL data from the fathers, no significant differences were found between

different attachment types among singletons or among twins. The relation of the

0

2

4

6

8

10

12

14Singletons

Twins

p<0.05

58

attachment type with the father and symptoms according to the CBCL

questionnaire completed by fathers of singletons and twins is presented in

Publication III Table 2b.

59

Ta

ble

14

. A

tta

ch

men

t ty

pe

w

ith

th

e m

oth

er

an

d

sym

pto

ms

ac

co

rdin

g

to

the

C

BC

L q

ue

sti

on

na

ire

c

om

ple

ted

b

y

mo

the

rs

of

sin

gle

ton

s (

A =

, B

=,

C =

, IO

=)

.

CB

CL

qu

est

ion

na

ire

A

B

C

IO

N

= 1

0

N

= 4

N =

4

N

= 4

p-v

alu

e

M

d

q1

q

3

M

M

d

q1

q

3

M

M

d

q1

q

3

M

M

d

q1

q

3

M

A

vs

BC

IO

CB

CL s

yndro

me s

cale

s

With

dra

wn

2

1

4

3.1

0

0

7

2.2

0

0

2

1.0

0

0

2

0.8

0.0

17

Som

atic

Com

pla

ints

2

1

2

1.9

0

0

2

0.2

0

0

1

0.5

1

0

2

1.0

0.0

16

Anxi

ous/

De

pre

sse

d

2

0

3

3.0

2

0

7

3.2

2

1

4

2.0

1

0

3

1.5

0.6

87

So

cia

l Pro

ble

ms

2

1

3

2.6

1

0

6

2.2

0

0

3

1.2

2

0

3

1.5

0.1

48

Th

ou

gh

t P

rob

lem

s 0

0

2

0.6

0

0

3

1.0

0

0

1

0.2

0

0

1

0.2

0.4

71

Att

en

tion

Pro

ble

ms

2

1

4

2.6

4

2

8

4.2

2

1

4

2.2

3

1

4

2.8

0.7

38

Delin

qu

ent B

ehavi

our

2

1

3

1.7

2

0

5

2.8

0

0

4

1.2

1

1

2

1.2

0.8

13

Ag

gre

ssiv

e B

eh

avi

ou

r 9

5

13

9.0

10

6

19

11.8

6

5

18

9.5

8

5

12

8.2

0.6

39

CB

CL b

road-b

and s

cale

s

To

tal P

rob

lem

s 28

18

55

31.8

26

12

66

34.2

16

11

40

22.2

24

17

29

23.0

0.5

31

Inte

rnaliz

ing

6

3

10

7.7

2

0

15

5.8

3

2

6

3.5

2

1

7

3.2

0.0

40

Ext

ern

aliz

ing

11

5

15

11.7

13

6

24

14.5

6

5

22

10.8

9

6

13

9.5

0.7

65

Media

ns

(Md),

first

quart

iles

(q1),

thir

d q

uart

iles

(q3)

an

d m

eans

(M)

are

giv

en, and the s

ignifi

cance

s of

diff

ere

nce

s betw

een a

ttach

ment ty

pes

were

calc

ula

ted

usi

ng the n

onpara

metr

ic M

ann-W

hitn

ey

U-t

est

60

Ta

ble

15

. A

tta

ch

men

t ty

pe

wit

h t

he

mo

ther

an

d s

ym

pto

ms a

cco

rdin

g t

o t

he

CB

CL

qu

esti

on

na

ire

co

mp

lete

d b

y m

oth

ers

of

twin

s

(A =

, B

=,

C =

, IO

=).

CB

CL

qu

est

ion

na

ire

A

B

C

IO

N

= 1

8

N

= 2

2

N

= 9

N =

2

p-v

alu

e

M

d

q1

q

3

M

M

d

q1

q

3

M

M

d

q1

q

3

M

M

d

q1

q

3

M

A

vs

BC

IO

CB

CL s

yndro

me s

cale

s

With

dra

wn

1

0

2

1.1

1

0

1

1.1

0

0

2

0.6

0

0

- 0.5

0.3

67

Som

atic

Com

pla

ints

1

0

2

1.2

1

0

2

1.0

0

0

2

0.8

0

0

- 0.5

0.4

47

Anxi

ous/

De

pre

sse

d

1

0

2

1.4

0

0

3

1.4

1

0

2

0.9

0

0

- 0.5

0.4

21

So

cia

l Pro

ble

ms

2

0

3

1.8

1

0

2

1.2

1

0

2

1.2

2

1

- 1.5

0.4

29

Th

ou

gh

t P

rob

lem

s 0

0

1

0.7

0

0

1

0.6

0

0

2

0.6

0

0

0

0.0

0.5

95

Att

en

tion

Pro

ble

ms

3

2

4

2.7

2

0

4

2.0

0

0

3

1.6

2

1

- 1.5

0.0

92

Delin

qu

ent B

ehavi

our

2

1

2

1.5

0

0

2

1.0

1

0

2

1.3

1

1

1

1.0

0.1

45

Ag

gre

ssiv

e B

eh

avi

ou

r 4

3

7

5.1

6

3

12

7.6

4

0

10

5.6

2

1

- 2.5

0.5

79

CB

CL b

road-b

and s

cale

s

To

tal P

rob

lem

s 18

12

23

18.8

14

10

31

19.3

18

2

22

15.4

10

6

- 10.0

0.3

49

Inte

rnaliz

ing

3

2

4

3.7

3

1

6

3.4

1

0

4

2.2

2

0

- 1.5

0.2

10

Ext

ern

aliz

ing

6

4

9

6.6

7

4

12

8.6

4

0

12

6.9

4

2

- 3.5

1.0

00

Media

ns

(Md),

first

quart

iles

(q1)

and t

hird q

uart

iles

(q3)

and m

eans

(M)

are

giv

en, and t

he s

ignifi

cance

s of diff

ere

nce

s betw

een a

ttach

ment ty

pe

s w

ere

calc

ula

ted u

sing t

he n

onpara

metr

ic M

ann-W

hitn

ey

U-t

est

61

6 Discussion

The objectives of this study were:

1. to examine the attachment of 18 months old Finnish twins and singletons to

both parents.

2. to examine cognitive skills in the same cohort at 36 months of age and to

compare the results with the recorded attachment type at 18 months.

3. at age 48 months, to relate mental well-being of the children as reported by

their parents to the 18-month attachment results.

6.1 Attachment of singletons and twins to both parents

In our study Type A attachment strategy is strongly represented in comparison to

data of a meta-analysis of attachment patterns elicited in almost 2000 Strange

Situations in eight different countries (van Ijzendoorn & Kroonenberg 1988), in

which 65% of the children were classified as Type B, 20% as Type A, and 14% as

Type C. Their data suggested a pattern of cross-cultural differences, in which

Type A classifications emerged as relatively more prevalent in the Western

European countries and Type C classifications as relatively more frequent in

Israel and Japan.

We assumed that because of Finland’s Northern European cultural heritage

and the sparse population, normative Finnish children would display a

distributional bias towards Type A attachment strategy, similar to other Western

cultures. We also expected that because twins are obliged to share parental

attention with their co-twin there would be fewer Type B attached twins than

singletons. Because of the egalitarian values of Finnish culture, we expected no

gender differences.

In this Finnish sample, only one third of the children were classified to be Type B

attached, and Type A (avoidant/defended) was predominating, especially its

mildest form A1-2. There were no significant gender differences found in this

data, although a slight, non-significant trend was seen for the boys’ attachment to

be more frequently insecure in quality, and the girls’ more often secure. In order

to enable further considerations of a statistical significance of gender differences,

a larger sample size would be needed. Comparing the differences within the

whole sample of singletons and twins, our data revealed a trend showing twins to

be more often Type B attached to the mother; among girls the difference was

62

significant. This result was contrary to our hypothesis of less securely attached

twins, due to their obligation to share the attention of the mother with their co-

twin. Instead, this may support the secondary hypothesis of special parental

attention and support, which increases the probability of secure attachment.

Additionally, the two months longer maternal leave in twin families may have a

positive effect on attachment development in the last two months of the first year

of life. For one child, parental allowance is paid for 159 working days; for twins,

218 working days (KELA 2015). Furthermore, twins may together enjoy more

maternal presence and response than singletons. Even though the mother’s

nurture is not always directed to one particular child, the mother is in any case

often near.

There was no significant difference between the distributions of attachment to

the mother and to the father. Within the whole sample the concordance of

attachment quality (secure vs. insecure) between mothers and fathers was as high

as 70%. This concordance was comparable with a meta-analysis by Fox,

Kimmerly & Schafer (1991), in which about 50 % of the children were assessed

as Type B attached to both parents and 19 % were assessed as insecurely attached

to both parents. Fathers in twin families are needed to take part in childcare,

although Finnish men (especially within the younger generation) are active in

singleton families as well. Many Finnish fathers spend some weeks with their

new-born babies (paternal leave). Nowadays fathers of twins can stay at home at

the same time as the mother for a total of about 13 weeks (Kela 2015). Fathers also

share the responsibility of childcare with mothers later on as the children grow

up. The participation in childcare enables the fathers to be available at times of

family stress, and thus to be able to become important attachment figures for the

infants. In the present study, when the parents shared the caretaking of the child,

this appeared more often to predict Type B attachment. When both parents take

care of the child equally, this might increase their time and opportunity to respond

to the child’s needs for closeness and security.

6.2 Attachment and cognitive development

This study supports the hypothesis that children with Type A and Type B

attachment patterns at 18 months attain the highest intelligence levels at 36

months of age. Attachment figures of Type A attached children may actively

support distal types of play and, particularly, “academic” play (Crittenden 1995).

They often also strongly support successful solutions in the play or game. Thus, it

63

is not surprising that Type A attached children succeed well in intelligence tests,

which measure analytical skills rather than creative intelligence. Type B children

have a positive relationship with their parent, who sensitively provides a child with

help when needed and in this way fosters a child's cognitive development. Only in

two cases was the attachment type between the child and the father outside the

normal set of patterns (A/C), and in both of these cases the children had the

highest intelligence levels. It could also be considered that the parents of these

children place particular emphasis on cognitive skills, whereas the emotional side

is left with less attention. The result might also be a coincidence due to the small

sample size. More data will be needed to confirm this finding. On the other hand,

if the attachment to the mother was outside the normal range (IO, N = 6) the

child’s IQ was clearly lower compared to the IQs of Type A and Type B attached

pre-schoolers.

The SES of the family and gestational maturity were significantly associated

with better success in the intelligence test. The SES of this sample was high and

the highest classes were over-represented, and thus the comparison between IQ

and SES was not valid in this case.

It was surprising that the toddlers’ attachment classifications were not linked

to concentration, endurance, self-confidence or working speed in the intelligence

tests. Previous research has shown that the cognitive development of twins is

slower than that of singletons in early childhood (Alin Åkerman & Fischbein

1991, Hay et al. 1984, Myrianthopoulos 1976). This is understandable in view of

the risks, both biological and environmental, associated with twins. However, the

main developmental risk of twins appears to be associated with prematurity. At

least in this study the full-term twins performed almost as well in the intelligence

test as the full-term singletons.

6.3 Attachment and mental well-being in singletons and twins

Parents of singletons reported more emotional and behavioural problems

compared to twins. The mean raw score for total problems was among singleton

girls 30.4 and among twin girls 16.7 according to the mothers. According to the

fathers the corresponding figures were 29.9 and 16.9. Parallel, but non-significant

means of total problem raw scores were reported by mothers of singleton and

twin boys (28.1 vs. 20.7) and there was a statistically significant difference in

total problem scores reported by the fathers (26.1 vs. 15.8, p = 0.005).

64

Thus the scores for the singletons were clearly higher than the corresponding

value (M = 23.1) of a normative American sample (N = 619) in the age group 4–

11 (Achenbach 1991), although the scores for the twins were lower.

Correspondingly in the American sample the raw score mean among boys in the

age group 4–11 was 24.2 (Achenbach 1991). The raw score means between the

present sample and the American sample are not fully comparable because in the

American sample only 17% of the age group (4–11) were 4–5 years old.

The tendency of twins to have fewer emotional and behavioural problems than

singletons might be explained by the fact that twins always have someone close for

support – their co-twin. In order to have comparable study groups we selected

firstborn singletons and twins, which however made it impossible to compare the

effects of sisters and brothers for the well-being of singletons.

Finnish society is considered to respect an individuals’ independence; the

parents tend to encourage their children’s active and autonomous behaviour,

while dependence-related actions are often discouraged. In Finland most of the

fathers participate in childcare actively and fathers are known to be less binding

and to encourage more independence than do mothers (Minde et al. 1990).

Independence can also be encouraged in practice, because the Finnish society is

safe. Thus we assumed that type A attachment reflects and predicts independence,

not a risk for a child´s mental health. Nonetheless in this study the association

between Type A attachment and emotional symptoms was clear among singletons.

Type A attached singletons had significantly more internalizing problems, somatic

complaints and withdrawal, as reported by mothers, compared to twins. This

finding is in line with attachment studies (Ericksson et al. 1985, Greenberg &

Speltz 1988, Lyons-Ruth et al. 1993, 1997, Pierrehumbert et al. 2000) which have

indicated that Type A attachment is associated with toddlerhood or preschool

symptoms of a more internalizing nature, such as anxiety and withdrawal. The

avoidant individual attempts to self-regulate negative affect rather than to

acknowledge vulnerability and ask for help from others (Main and Cassidy 1988).

It is easy to understand that Type A attachment may lead via feelings of alienation

and aloneness to withdrawal, somatization and even depression.

Why was this association between avoidant attachment and internalizing

problems not found among twins? Even if the inter-twin relationship may also

include some stressful aspects such as sibling rivalry, the twin toddlers in fact do

not suffer from loneliness and the twins might use one another as attachment

figures. Bowlby (1969/1982) discussed the notion that infants could have multiple

attachment objects and that “responsiveness to crying and readiness to interact

65

socially are among the most relevant variables” (p. 315) in determining who will be

chosen as an attachment figure. The mother is the most likely candidate in most cases,

but fathers, siblings, aunts, uncles, and grandparents may also serve as attachment

figures (Ainsworth et al. 1991). When the mother is perceived as inaccessible, for

example, siblings may turn to each other for comfort and care. However, the presence

of the co-twin does not diminish the need for the parent, nor does the twin prefer the

company of the co-twin to that of the parent.

To ensure favourable development and mental well-being, every child should

have an attachment figure, with which to secure the bad moments and share

moments of joy. And lucky he or she, who has in addition one or more secure

secondary attachment objects - like a co-twin.

6.4 Limitations

The main limitation of this study is the small sample size and especially the over-

representation of twins (30 singletons and 60 twins), which prevents generalization

of the results. However, this was the first research of attachment in Finland

concerning healthy children and ordinary families and can thus be seen as a

pioneering work, which could be replicated with bigger study groups.

Unfortunately there is no reported data of Finnish attachment classification of

a normative sample of CBCL/4–18 reports of 4-year-old Finnish children, which

would allow national comparison of results.

Another limitation in this sample is that the data was collected from one city,

Oulu, which is the capital of northern Finland with several research institutes, the

University Hospital and its well-known technology. The Oulu Region has over

200 000 inhabitants and it is the fastest growing region in Finland. On the other

hand no differences were found in another Finnish study of three-year-old

children’s emotional and behavioural problems from two very different cities,

Kemijärvi and Uusikaupunki; one from the north, another from further south and

from different sociocultural areas of Finland (Sourander 2001). Thus, the Finnish

society appears to be rather homogenous, and the practical decision to collect data

from only one population, the City of Oulu, will have had little effect on the

validity of the results.

The third limitation of this study is the high socioeconomic status of the

families, which also limits the generalizability of the sample results. Thus, we

must be very careful in drawing conclusions concerning correlations of SES with

attachment pattern, cognitive development and emotional and behavioral

66

symptoms of children because of the very low numbers in some SES classes. This

decreases the reliability of the results.

67

7 Conclusions: Implications and future research

The present study examined the attachment in 18 months old Finnish singletons

and twins to both parents and then compared attachment pattern with the

cognitive development at 36 months of age and mental well-being reported by

both parents at 48 months.

The twin birth rate has increased because of the growing numbers of older

mothers and more successful fertility treatments. About half of the twins are born

before 37 weeks, which increases risks for development and well-being. All in all

twin pregnancy and delivery, awareness of all risks, and finally the simultaneous

care of two infants represent both mentally and physically a great challenge for

the parents of twins. Therefore these parents are often particularly worried about

the well-being of their twins. However, this study shows that twinship can also

serve as a protective factor in the development of mental health; twins performed

well compared to singletons.

Our results indicated that in infancy, twins more often had Type B, secure

attachment strategy to their parents and the accumulation of the two other risk

factors in addition to twinship – premature birth and hospital treatment – further

increased the probability that the attachment strategy would be classified as Type

B.

Parents of twins reported significantly less emotional and behavioural

problems in their children compared to parents of singletons. Twins also

performed well in the intelligence test, which revealed that the children in this

study had above-average intelligence, although pre-term twins and singletons

appeared to have distinctly lower intelligence levels at 36 months of age

compared with full-term children.

Finnish society is a safe environment in which to live and develop, and

independence might thus be encouraged. Type A attachment strategy is over-

represented especially among singletons and for them it appears to lead to an

increasing trend towards internalizing symptoms at preschool age. The co-twin

appears to protect a twin with Type A attachment strategy from the pathway of

developing internalizing symptoms such as withdrawal and somatic complaints.

However, these findings, although very interesting, must be considered as

preliminary and further investigation with a larger sample size is needed.

As we suggested that the co-twin attachment may help to prevent Type A

parental attachment from developing into child’s internalizing symptoms, it might

68

be worthwhile to assess inter-twin attachment in further follow-up research on the

development and mental well-being of twins and singletons. Furthermore,

attachment between siblings might be worth studying when searching for

protective factors in children’s well-being in difficult situations.

69

References

Achenbach TM (1991a) Integrative guide for the 1991 CBCL/4–18, YSR, and TRF profiles, Burlington, University of Vermont Department of Psychiatry.

Achenbach TM (1991b) Manual for the Youth Self-Report and profile. Burlington: Department of Psychiatry, University of Vermont Department of Psychiatry.

Achenbach, TM & Edelbrock CS (1978) The classification of child psychopathology: A review and analysis of empirical efforts. Psychol Bull 85: 1275–1301.

Achenbach TM & Rescorla LA (2001) Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families.

Ainsworth MDS, Bell SM & Stayton DJ (1971) Individual differences in Strange- Situation behavior of one-year-old. In The origins of human social relations, Schauffer HR, ed. New York, Academic Press, pp. 17–52.

Ainsworth MDS & Bell SM (1970) Attachment, exploration, and separation: Illustrated by the behavior of one-year-olds in a strange situation. Child Dev 41: 49–67.

Ainsworth MDS, Blehar MC, Waters E & Wall S (1978) Patterns of attachment. New Jersey: Erlbaum.

Ainsworth MDS & Bowlby J (1991) An ethological approach to personality development. Am Psychol 46: 331–341.

Ainsworth MDS, Parkes, CM, Stevenson-Hinde J & Marris P, eds (1991) Attachment across the lifecycle. New York: Routledge.

Alestalo M & Uusitalo H (1978) Occupational prestige and its determinants: the case of Finland (No.20). University of Helsinki.

Alin Åkerman B & Fischbein S (1991) Twins: are they at risk? A longitudinal study of twins and non-twins from birth to 18 years of age. Acta Genet Med Gemellol 40: 29−40.

Amato PR (2000) The consequences of divorce for adults and children. J Marriage Fam 62(4): 1269–1287.

Ananth CV, Joseph KS, Demissie K & Vintzileos AM (2005) Trends in twin preterm birth subtypes in the United States, 1989 through 2000: Impact on perinatal mortality. Am J Obstet Gynecol 193: 1076–1082.

Bain SK & Allin JD (2005) Book review: Stanford–Binet intelligence scales, fifth edition. J. Psychoeduc Assess 23: 87–95.

Bakker P (1987) Autonomous language of twins. Acta Genet Med Gemellol 36(2): 233–238.

Bates JE & Bayles K (1988) Attachment and the development of behavior problems. In Clinical implications of attachment, Belsky J & Nezworski T, Eds. Hillsdale, NJ, Erlbaum: pp. 253–299.

Belsky J & Fearon RM (2002) Infant-mother attachment security, contextual risk, and early development: a moderational analysis. Dev Psychopathol Spring 14: 293–310.

Bleker OP, Oosting J & Hemrika DJ (1988) On the retardation of fetal growth in multiple gestations. Acta Genet Med Gemellol 37: 41–46.

70

Bohlin G & Hagekull B (2000) Behavior Problems in Swedish four-year-olds. In Organization of Attachment Relationships, McKinsey Crittenden P & Hartl Claussen A, eds. New York: Cambridge University Press, pp. 75–96.

Bongers IL, Koot, HM, Van der Ende J & Verhulst FC (2003) The normative development of child and adolescent problem behavior. J. Abnorm. Psychol 112: 179–192.

Bonellie SR, Curriw D & Chalmers J (2005) Comparison of risk factors for cerebral palsy in twins and singletons. Dev Med Child Neurol 47: 587–591.

Bowlby J (1970) Reasonable fear and natural fear. Int J of Psychiatry 9: 79–88. Bowlby J (1969/1982) Attachment and loss: Vol I. Attachment. 2 nd ed. New York: Basic

Books 1969/1982. Bowlby J (1973) Attachment and loss: vol II. Separation, anxiety and anger. London:

Hogarth Press. Bowlby J (1979) The making and breaking of affectional bonds. London: Routledge. Bowlby J (1980) Attachment. Attachment and Loss. Vol.III. Loss. New York: Basic Books. Braungart-Rieker JM, Garwood MM, Powers BP & Wang X (2001) Parental sensitivity,

infant affect, and affect regulation: Predictors of later attachment. Child Dev 72: 252–270.

Bretherton I (1980) Young children in stressful situations: The supporting role of attachment figures and unfamiliar caregivers. In Uprooting and development, Coelho GV & Ahmed P, eds. New York: Plenum, pp. 179–210.

Brumariu L & Kerns K (2010) Parent–child attachment and internalizing symptoms in childhood and adolescence: A review of empirical findings & future directions. Dev Psychopathol 22: 177–203.

Burton A & Ananth CV (2010) Contributions of ischaemic placental disease to preterm birth in twin gestations. J Matern Fetal Neonatal Med 23: 1183–1186.

Bus AG & van Ijzendoorn MH (1988). Mother-child interactions, attachment and emergent literacy: A cross-sectional study. Child Dev 59: 1262–1272.

Calvin C, Fernandes C, Smith P, Peter M, Visscher PM & Deary IJ (2009) Is there still a cognitive cost of being a twin in the UK? Intelligence 37: 243–248.

Carlson EA (1998) A prospective longitudinal study of disorganized/disoriented attachment. Child Dev 69: 1970–1979.

Cassidy J (1994) Emotion regulation: Influences of attachment relationship In: Fox NA (Ed), The development of emotion regulation: Biological and behavioural consideration. Monographs of the Society for Research in Child Development, 59 (2–3): 228–249.

Cassidy J (1999) The nature of the child’s ties. In Handbook of Attachment: Theory, Research, and Clinical Implications, Cassidy J & Shaver PR, eds. New York: Guilford, pp. 3–20.

Chaudhari S, Bhalerao MR, Vaidya U, Pandit A & Nene U (1997) Growth and development of twins compared with singletons at ages one and four years. Indian Pediatr 34: 1081–1086.

Chauhan SP, Scardo JA, Hayes E, Abuhamad AZ & Berghella V (2010) Twins: Prevalence, problems, and preterm births, Am J Obstet Gynecol, 203: 305–315.

71

Cicchetti D & Toth SLA (1991) Developmental perspective on internalizing and externalizing disorders. Rochester Symposium on Dev Psychopathol 2: 1–20.

Cincotta RB, Gray PH, Phythian G, Rogers, YM, Chan FY (2000) Long term outcome of twin-twin transfusion syndrome. Arch Dis Child Fetal Neonatal Ed 83: 171–176.

Collins AL & Sullivan PF (2013) Genome-wide association studies in psychiatry: what have we learned? Br J Psychiatry 202: 1–4.

Contreas JM, Kerns K A, Weimer B L, Gentzler A L & Tomich P L (2000) Emotion regulation as a mediator of associations between mother-child attachment and peer relationships in middle childhood. J Fam Psychol 14: 111–124.

Crittenden P (1992a) Quality of attachment in the preschool years. Dev Psychopathol 4: 209–241.

Crittenden P (1992b) Coding manual: Classification of quality of attachment for preschool-aged children. University of New Hampshire: Family Research Laboratory.

Crittenden PM (1995) Attachment and psychopathology. In John Bowlby's attachment theory: Historical, clinical, and social significance, Goldberg S, Muir R, Kerr J, eds. New York: Analytic Press, pp. 367–406.

Dales RJ (1969) Motor and language development in twins during the first three years. J Genet Psychol 114: 263–271.

De Moutzon J, Goossens V, Bhattacharya S, Ferraretti AP, Korsak V, Kupka M, Nygren KG & Nyboe Andersen A (2012) Assisted reproductive technology in Europe, 2007: results generated from European registers by ESHRE. Hum Reprod 27: 954–966.

De Rosnay M & Harris PL (2002) Individual differences in childrenʹs understanding of emotion: the roles of attachment and language. Attach Hum Dev 4: 39–54.

de Ruiter C & van IJzendoorn MH (1993) Attachment and Cognition: A review of the literature. Int J Educ Res 18: 525–540.

Dhont M, De Sutter P, Ruyssinck G, Martens G & Bekaert A (1999) Perinatal outcome of pregnancies after assisted reproduction: A case-control stud. Amer J Obstet and Gynecol 181: 688–695.

Erickson MF, Sroufe LA & Egeland B (1985) The relationship between quality of attachment and behavior problems in preschool in a high-risk sample. In Growing points of attachment theory and research, Bretherton I & Waters, eds. Monographs of the Society of Research in Child Development (1–2 serial no 209): 147–156.

Eriksson AW & Fellman J (2007) Temporal trends in the rates of multiple maternities in England and Wales. Twin Res Hum Genet 10: 626–632.

Estourgie-Van Burk GF, Bartels M, Van Beijsterveldt TC, Delemarre-Van de Waal HA & Boomsma DI (2006) Body size in five-year-old twins: heritability and comparison to singleton standards. Twin Res Hum Genet 2006 9: 646–655.

Fearon RP, Bakermans-Kranenburg MJ, van Ijzendoorn MH, Lapsley AM & Roisman GI (2010) The significance of insecure attachment and disorganization in the development of children´s externalizing behaviour, a meta-analytic study. Child Dev 81: 435–56.

Feng X, Shaw DS & Silk JS (2008) Developmental trajectories of anxiety symptoms among boys across early and middle childhood. J Abnorm Psychol 117: 32–47.

72

FIGO (2001) Committee for the Ethical Aspects of Human Reproduction and Women's Health. Ethical guidelines in the prevention of iatrogenic multiple pregnancy. Euro J Ob Gyn Repro Biol 96?: 209–210.

Fox N, Kimmerly N & Schafer W (1991) Attachment to mother/attachment to father: a meta-analysis. Child Dev 62: 210–25.

Fraley RC & Tancredy CM (2012). Twin and Sibling Attachment in a Nationally Representative Sample. Pers Soc Psychol Bull 38: 308–16.

Gau JS, Silberg JL, Erickson MT & Hewitt JK (1992) Childhood behavior problems: a comparison of twin and non-twin samples. Acta Genet Med Gemellol 41: 53–63.

Gilliom M & Shaw DS (2004) Codevelopment of externalizing and internalizing problems in early childhood. Dev Psychopathol 6: 313–333.

Gjone H & Novik S (1995) Parental ratings of behaviour problems: a twin and general population comparison. J. Child Psychol. Psychiatry 36: 1213–1224.

Grant KE, Compas BE, Stuhlmacher AF, Thurman AE, McMahon SD & Halpert JA (2003) Stressors and child and adolescent psychopathology: moving from markers to mechanisms of risk. Psychol Bull 129: 447–466.

Greenberg MT & Speltz ML (1988) Attachment and the ontogeny of condact problems in: Belsky J & Nezworski T (eds) Clinical implications of attachment Hillsdale: NJ Erbaum, pp. 177–218.

Grether JK, Nelson KB & Cummins SK (1993) Twinning and cerebral palsy: Experience in four northern California counties, births 1983 through 1985, Pediatr 92: 854–858.

Groh AM, Roisman GI, van Ijzendoorn MH, Bakermans-Kranenburg MJ & Fearon RP (2012) The significance of insecure and disorganized attachment for children’s internalizing symptoms: a meta-analytic study. Child Dev 83, 591–610.

Groh AM, Fearon RP, Bakermans-Kranenburg MJ, van Ijzendoorn MH, Steele RD & Roisman GI (2014) The significance of attachment security for children’s social competence with peers: a meta-analytic study. Attach Hum Dev 16: 103–136.

Hay DA, O'Brien PJ, Johnston CJ & Prior M (1984) The high incidence of reading disability in twin boys and its implications for genetic analyses Acta Genet Med Gemellol 33: 223−236.

Hay DA, Prior M, Collet S & Williams M (1987) Speech and language development on twins. Acta Genet Med Gemellol 36: 213–223.

Hazen NL & Durrett ME (1982) Relationship of security of attachment to exploration and cognitive mapping aabilities in two-year olds. Dev Psychol 18:751–759.

Hetherington EM & Stanley-Hagan M (1999) The adjustment of children with divorced parents: a risk and resiliency perspective. J Child Psychol Psychiatry 40(1), 129–140.

Hirshfeld-Becker DR, Micco J, Henin A, Bloomfield A, Biederman J & Rosenbaum J (2008) Behavioral inhibition. Depress Anxiety 25: 357–367.

Hollinger CJ & Baldwin C (1990) The Stanford-Binet, fourth edition: A small study of concurrent validity. Psychol Rep 66: 1131–1336.

Husen T (1961) Abilities of Twins. Acta Psychol 19: 736–738. Imaizumi Y (1997) Trends of twinning rates in ten countries, 1972–1996. Acta Genet Med

Gemellol (Roma) 46: 209–218.

73

Jacobsen T & Hoffman V (1997) Children’s attachment representations: Longitudinal relations to school behavior and academic competency in middle childhood and adolescence. Dev Psychol 33: 703–710.

Jacobvitz D & Hazen N (1999). Developmental pathways from infant disorganization to childhood peer relationships. In Attachment disorganization. Solomon J & George C, eds. New York: Guilford Press, pp. 127–159.

Kalland M & Sinkkonen J 2013 Children exposed to parental substance abuse who are placed in foster care: An attachment perspective. In Parenting and substance Abuse: Developmental approaches to Intervention. Suchman NE, Pajulo M & Mayes LC, eds. New York: Oxford University Press USA, pp. 344–364.

Kayla N, Anderson KN, Koh BD, Connor JJ, Ascan F, Koerner AF, Damario M & Rueter MA (2014) Twins conceived using assisted reproduction: Parent mental health, family relationships, and child adjustment at middle childhood. Hum Rep 29: 2247–2255.

Keiley MK, Bates JE, Dodge K A. and Pettit, G. S. (2000). A cross-domain growth analysis: externalizing and internalizing behaviors during 8 years of childhood. JACP 28: 161–179.

Keith LG, Oleszczuk JJ & Keith DM (2000) Multiple gestation: Reflections on epidemiology, causes and consequences. Int J Fertil Womens Med 45:206–214.

KELA (2015) http://www.kela.fi/web/en/multiple-birth-families. Keller T, Spieker S & Gilchrist L (2005) Patterns of risk and trajectories of preschool

problem behaviors: A person-oriented analysis of attachment in context. Dev Psychopathol 17: 349–84.

Kinzler WL, Ananth CV, Vintzileos AM (2000) Medical and economic effects of twin gestations. J Soc Gynecol Invest 7: 321–327.

Klann-Delius G & Hofmeister C (1997) The development of communicative competence of securely and insecurely attached children in interactions with their mothers J Psycholinguistic Res 26: 69–88.

Lahti J, Räikkönen, K, Kajantie E, Heinonen K, Pesonen AK, Järvenpää AL & Strandberg T (2006) Small body size at birth and behavioral symptoms of ADHD in children aged five to six years. J of Child Psychol Psychiatry 47, 1167–1174.

Lambalk CB, Boosma DI, De Boer L, De Konig CH, Schoute, E, Popp-Snijders C & Schoemaker J. (1998). Increased levels and pulsatility of follicle-stimulating hormone in mothers of hereditary dizygotic twins. JCEM 83: 481–486..

Lavigne JV, Arend R, Rosenbaum D, Binns HJ, Christoffel KK & Gibbons RD (1998) Psychiatric disorders with onset in the preschool years: II: correlates and predictors of stable case status. J Am Acad Child Adolesc Psychiatry 37: 1255–1261.

Leonard MR (1961) Problems in identification and ego development in twins. Psychoanal Study Child 16: 300–320.

Leve LD, Kim H K & Pears K C (2005) Childhood temperament and family environment as predictors of internalizing and externalizing trajectories from ages 5 to 17. J Abnorm Child Psychol 33: 505–520.

Lorenz JM (2012) Neurodevelopmental outcomes of twins. Semin Perinatol 36: 201–212.

74

Luyckx K, Tildesley, E A, Soenens, B, Andrews JA, Hampson, SE, Peterson M, Duriez B

(2011) Parenting and trajectories of children’s maladaptive behaviors: a 12-year prospective community study. J Clin Child Adolesc Psychol 40: 468–478.

Lyons-Ruth K, Alpern L & Repacholi B (1993) Disorganized infant attachment classification and maternal psychosocial problems as predictors of hostile-aggressive behavior in the preschool classroom. Child Dev 64: 572–585.

Lyons-Ruth K, Eaterbrooks MA, Cibelli CD (1997) Infant attachment strategies, infant mental lag, and maternal depressive syndroms: Predictors of internalizing and externalizing problems at age 7. Dev Psychol 33: 681–692.

Lyons-Ruth K & Jakobvitz D (1999) Attachment disorganization: Unresolved loss, relational violence, and lapses in behavioural and attentional strategies. In Cassidy J, Shaver PR, eds. Handbook of attachment: Theory, research, and clinical applications. New York: Guilford, pp. 520–554.

Lytton H (1977) Parent–child interaction: The socialization process observed in twin and singleton families. New York: Plenum.

Main M & Cassidy J (1988) Categories of response to reunion with the parent at age 6: predictable from infant attachment classifications and stable over a 1-month period. Dev Psychol 24: 415–426.

Main M & Hesse E (1990) Parent- insolved traumatic experiences are related to infant disorganized status: Is frightened and/or frightening parental behavior the linking mechanism? In Attachment in the preschool years: Theory, research, and intervetions, Greenberg MT & Cicchetti D, eds. Chicago: University of Chicago Press, pp. 161–182.

Main M & Kaplan N & Cassidy J (1985) Security in infancy, childhood and adulthood: A move to the level of representation In Growing points of attachment theory and research, Bretherton I & E Waters, eds. Monographs of the Society of Research in Child Development 50 (1–2, Serial No209): 66–104.

Main M &Solomon J (1986) Discovery of a new insecure-disorganized/disorientated attachment pattern. In Affective development in infancy Norwood, Brazelton TB & Yogman MW, eds. NJ: Ablex, pp. 95–124.

Main M (1990) Cross-cultural studies of attachment organization: Recent studies, changing methodologies, and the concept of conditional strategies. Hum Dev 33: 48–61.

Martin JA, Hamilton BE & Osterman MJK (2012) Three decades of twin births in the United States, 1980–2009. NCHS data brief, no 80. Hyattsville, MD: National Center for Health Statistics.

Matas L, Arend R A and Sroufe L A. (1978) Continuity of adaptation in the second year: The relationship between quality of attachment and later competence. Child Dev 49: 547–556.

McCulloch A, Wiggins RD, Joshi, HE & Sachdev D (2000) Internalizing and externalizing chil- dren’s behaviour problems in Britain and the U.S.: Relationships to family resources. Child Soc 14: 368–383.

75

Mesman J, Bongers I, Koot H (2001) Preschool developmental pathways to preadolescent internalizing and externalizing problems. J Child Psychol Psychiatry 42: 679–689.

Minde K, Corter C, Goldberg S & Jeffers D (1990) Maternal preference between premature twins up to age four. J Am Acad Child Adolelesc Psychiatry 29: 369–374.

Miner JL & Clarke-Stewart KA (2008) Trajectories of externalizing behavior from age 2 to age 9: relations with gender, temperament, ethnicity, parenting, and rater. Dev Psychol 44: 771–786.

Mizuta I, Zahn-Waxler C, Cole P, Hiruma N (1996) A cross-cultural study of preschoolers’ attachment: security and sensitivity in Japanese and US dyads. Int J Beh Dev 19: 141–59.

Moilanen I (1987) Dominance and submissiveness between twins. II Consequences for mental health. Acta Genet Med Gemellol 36: 257–65.

Moilanen I & Ebeling H (1998) To be born as a twin – risks and sequelae. Int J Circumpolar Health 57: 138–47.

Moilanen I, Linna S-L, Ebeling H, Piha J, Kumpulainen, K, Tamminen T & Almqvist F (1999) Are twins’ behavioural/emotional problems different from singletons’? Eur Child Adolesc Psychiatry 8 Suppl 4: IV 62- IV/67.

Moilanen I & Pennanen P (1997) ”Mother´s child” and ”father´s child” among twins. Acta Genet Med Gemellol 46: 219–230.

Moise J, Laor A, Armon Y, Gur I & Gale R (1998) The outcome of twin pregnancies after IVF. Hum Rep 13: 1702–1705.

Moss, E & St-Laurent D (2001). Attachment at school age and academic performance. Dev Psychol 37: 863–874.

Muris P, Van Brakel AML, Arntz A & Schouten E (2011) Behavioral inhibition as a risk factor for the development of childhood anxiety disorders: a longitudinal study. J Child Fam Stud 20, 157–170.

Myrianthopoulos NC (1976) Congenital malformations in twins. Acta Genet Med Gemellol (Roma) 25: 331–335.

National Vital Statistics Report (2013) Births: Final Data for 2011, 62 (1). O´Connor TG & Croft, CM (2001) A twin study of attachment in preschool children. Child

Dev, 72: 1501–1511. Official Statistics of Finland (OSF): Official Statistics of Finland [e-publication]. Helsinki:

Advisory Board of OSF. Ozcakar ZB, Sahin F, Beyazova U & Soysal S (2003) Physical and mental development of

Turkish twins. Pediatr Int 45: 712–718. Pierrehumbert B, Miljkovitch R, Plancherel B, Halfon O & Ansermet F (2000) Attachment

and temperament in early childhood: Implications for later behavior problems. Infant Child Dev 9: 17–32.

Phillips DI, Davies MJ & Robinson JS Fetal growth and the fetal origins hypothesis in twins- problems and perspectives. Twin Research 4: 327–331.

Pihlakoski L, Sourander A, Aromaa M, Rautava P, Helenius H & Sillanpää M (2006) The continuity of psychopathology from early childhood to preadolescence: a prospective cohort study of 3–12-year-old children. Eur Child Adolesc Psychiatry 15: 409–417.

76

Pulkkinen L, Vaalamo I, Hietala R, Kaprio J & Rose RJ (2003) Peer reports of Adaptive Behavior in Twins and Singletons: Is Twinship a Risk or an Advantage? Twin Research 6: 106–118.

Record RG, McKeown T & Edwards JH (1970) An investigation of the difference in measured intelligence between twins and single births. Ann Hum Genet 34: 11–20.

Robbers Sc, Bartels M, van Oort FV, van Beijsterveldt CE, van der Ende J, Verhulst FC,Boomsma DI & Huizink AC (2010) A twin-singleton comparison of developmental trajectories of externalizing and internalizing problems in 6- to 12-year-old children. Twin Res Hum Genet 13: 79–87.

Ronalds GA, De Stavola BL & Leon DA (2005). The cognitive cost of being a twin: Evidence from comparisons within families in the Aberdeen children of the 1050s cohort study. BMJ 331: 1306.

Rothman KJ No Adjustments Are Needed for Multiple Comparisons. Epidemiology 1990 1: 43–46.

Rutter M, Thorpe KJ, Greenwood R, North K & Golding J (2003) Twins as a natural experiment to study the causes of mild language delay: I Twin–singleton differences in language, and obstetric risk. J Child Psychol Psychiatry 44: 326–341.

Rutter M, Tizard J & Whitmore K (1970) Education, Health and Behavior. London: Longman. Rydhstroem H & Heraib F (2001) Gestational duration, and fetal and infant mortality for

twins vs singletons. Twin Research 4: 227–231. Räikkönen K, Pesonen AK, Heinonen K, Kajantie E, Hovi P, Järvenpää AL, Eriksson JG

& Andersson S (2008) Depression in young adults with very low birth weight: the Helsinki study of very low-birth-weight adults. Arc Gen Psychiat 65: 290–306.

Sajaniemi N, Mäkelä J, Salokorpi T, von Wendt L, Hämäläinen T & Häkämies-Blomqvist L 2001 Cognitive performance and attachment patterns at four years of age in extremely low birth weight infants after early intervention. Eur Child Adolesc Psychiatry 10: 122–129.

Sarna S, Kaprio J, Sistonen P & Koskenvuo M (1978) Diagnosis of twin zygosity by mailed questionnaire. Hum Hered 28: 241–54.

Shaw DS, Owens EB, Vondra JI, Keenan K & Winslow EB (1996) Early risk factors and pathways in the development of early disruptive behavior problems. Dev Psychopathol 8: 679–699.

Shaw DS, Gilliom M, Ingoldsby, EM & Nagin DS (2003) Trajectories leading to school-age conduct problems. Dev Psychol 39: 189–200.

Schore AN (2005) Attachment, affect regulation, and the developing right brain: Linking developmental neuroscience to pediatrics. Pediatr Rev 26: 204–217.

Shaw DS, Lacourse E & Nagin DS (2005) Developmental trajectories of conduct problems and hyperactivity from ages 2 to 10. J Child Psychol and Psych 46: 931–942.

Shaw DS & Vondra JI (1995) Infant attachment security and maternal predictors of early behaviour problems: A longitudinal study of low income families. JACP 23: 335–357

Solomon J & George C (1999) The measurement of attachment security in infancy and childhood. In Handbook of Attachment: Theory. Research, and Clinical Implications.J CassidyJ & Shaver PR, eds. New York: Guilford, pp. 287–318.

77

Sourander A (2001) Emotional and Behavioural problems in a sample of Finnish 3-year-old. Eur Child Adolesc Psychiatry 10: 98–104.

Spellacy WN, Handler A, Ferre CD (1990) A case-control study of 1253 twin pregnancies from a 1982–1987 perinatal data base. Obstet and Gynecol 75: 168–171.

Stanger C, Achenbach TM & Verhulst FC (1997) Accelerated longitudinal comparisons of aggressive versus delinquent syndromes. Dev Psychopathol 9: 43–58.

Steele H, Steele M & Croft C (2008) Early attachment predicts emotion recognition at 6 and 11 years. Attach Hum Dev 10: 379–393.

Sterba SK, Prinstein MJ & Cox MJ (2007) Trajectories of internalizing problems across childhood: heterogeneity, external validity, and gender differences. Dev Psychopathol 19: 345–366.

THL National Institute for Health and Welfare. Medical Birth Register, unpublished statistics.

Thompson R (1999) Early attachment and later development In Handbook of Attachment: Theory, Research, and Clinical Implications, Cassidy J & Shaver PR eds. New York: Guilford, pp. 265–286.

Thorndike RL, Hagen EP & Sattler JM (1986) The Stanford-Binet Intelligence Scale: Fourth edition. Chicago: Riverside.

Thorpe K, Golding J, Mac Gillivray I & Greenwood R, Comparison of prevalence of depression in mothers of twins and mothers of singletons. Brit Med J 302: 875–878.

Tirkkonen T, Kunelius A, Ebeling H, Pylkkönen M, Tuomikoski H & Moilanen I (2008) Attachment in Finnish singletons and twins to both parents. Psychiatria Fennica 39: 89–98.

Trias T, Ebeling H, Penninkilampi-Kerola V, Kunelius A, Tirkkonen T & Moilanen I (2006) How Long Do the Consequences of Parental Preference Last: A Study of Twins From Pregnancy to Young Adulthood. Twin Res Hum Genet 9: 240–249.

van den Oord EJCG, Koot HM, Boomsma DI, Verhulst FC & Orlebeke JF (1995) A twin-singleton comparison of problem behaviour in 2–3-years-olds. J Child Psychol Psyc 36: 449–458.

van Ijzendoorn M, Kroonenberg P (1988) Cross-cultural patterns of attachment: a meta-analysis of the strange situation. Child Dev 59: 147–56.

van Ijzendoorn MH, Moran G, Belsky J, Pederson D, Bakermans-Kranenburg MJ & Kneppers K (2000) The similarity of siblings’ attachment to their mothers. Child Dev 4: 1086–1098.

van Ijzendoorn MH, Dijkstra J & Bus AG (1995) Attachment, intelligence and language: A meta-analysis. Soc Dev 4: 115–128.

van Ijzendoorn M & Kroonenberg P (1988) Cross-cultural patterns of attachment: a meta-analysis of the strange situation. Child Dev 59: 147–56.

van Oort FVA, Verhulst FC, Ormel J & Huizink AC (2010) Prospective community study of family stress and anxiety in (pre)adolescents: the Trails study. Eur Child Adolesc Psychiatry 19: 483–491.

Wang S (1997) Traumatic stress and attachment. Acta Physiol Scand Supplement 640: 164–169.

78

Waters E, Wippman J & Sroufe LA (1979) Attachment, positive affect, and competence in peer group: two studies in construct validation. Child Dev 50: 821–829.

West KK, Mathews BL & Kerns KA (2013) Mother-child attachment and cognitive performance in middle childhood: An examination of mediating mechanisms. Early Child Res Q 28: 259–270.

Webbink D, Posthuma D, Boomsma DI, de Geus, EJC & Visscher PM (2008). Do twins have lower cognitive ability than singletons? Intelligence 36: 539–547.

Weinfeld NS, Sroufe LA, Egeland B & Carlson EA (1999) The nature of individual differences in infant-caregiver attachment. In Handbook of Attachment: Theory, research and Clinical Implications J Cassidy & P R Shaver, eds. NewYork: Guilford, pp. 68–89.

Wetter E & El-Sheikh M (2012) Trajectories of children’s internalizing symptoms: The role of maternal internalizing symptoms, respiratory sinus arrhythmia, and child sex. J Child Psychol Psychiatry 53: 168–177.

Williams MC & O’Brien WF (1998) Low weight/length ratio to assess risk of cerebral palsy and perinatal mortality in twins. J Perinatol 15: 225–228.

Wilson RS (1974) Twins: mental development in the preschool years. Dev Psyc 10: 580–588.

Zazzo R (1979) The twin condition and the couple effects on personality development. Acta Genet Med Gemellol 33: 75–80.

Wetter E & El-Sheikh M (2012) Trajectories of children’s internalizing symptoms: The role of maternal internalizing symptoms, respiratory sinus arrhythmia, and child sex. J Child Psychol Psyc 53: 168–177.

Williams MC & O’Brien WF (1998) Low weight/length ratio to assess risk of cerebral palsy and perinatal mortality in twins. J Perinatol 15:225–228.

Wilson RS (1974) Twins: mental development in the preschool years. Dev Psyc 10: 580–588.

Zazzo R (1979) The twin condition and the couple effects on personality development. Acta Genet Med Gemellol 33: 75–80.

79

Original publications

I Moilanen I, Kunelius A, Tirkkonen T & Crittenden P (2000) Attachment in Finnish twins. In Crittenden P & Claussen A (eds) The Organization of Attachment Relationships. Cambridge University Press: 125–40.

II Tirkkonen T, Kunelius A, Ebeling H, Pylkkönen M, Tuomikoski H & Moilanen I (2008) Attachment in Finnish singletons and twins to both parents. Psychiatria Fennica 39: 89–98.

III Tirkkonen T, Joskitt L, Kunelius A, Huhtaniska M & Moilanen I (2015) Twinship as a protective factor against behavioural and emotional problems at preschool age. Early Child Development and Care. DOI: 10.1080/03004430.2015.1066783.

The original publications have been reproduced with the kind permission of the

copyright holders.

Original publications are not included in the electronic version of the dissertation.

80

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

1307. Kämppi, Antti (2015) Identifying dental restorative treatment need in healthyyoung adults at individual and population level

1308. Myllymäki, Satu-Marja (2015) Specific roles of epithelial integrins in chemical andphysical sensing of the extracellular matrix to regulate cell shape and polarity

1309. Antonoglou, Georgios (2015) Vitamin D and periodontal infection

1310. Valtokari, Maria (2015) Hoitoon pääsyn moniulotteisuus erikoissairaanhoidossa

1311. Toljamo, Päivi (2015) Dual-energy digital radiography in the assessment of bonecharacteristics

1312. Kallio-Pulkkinen, Soili (2015) Effect of display type and room illuminance inviewing digital dental radiography : display performance in panoramic andintraoral radiography

1313. Roivainen, Eka (2015) Validity in psychological measurement : an investigation oftest norms

1314. Puhto, Ari-Pekka (2015) Prosthetic joint infections of the hip and knee :treatment and predictors of treatment outcomes

1315. Pentikäinen, Ilkka (2015) Distal chevron osteotomy for hallux valgus surgery :role of fixation and postoperative regimens in the long-term outcomes of distalchevron osteotomy – a randomised, controlled, two-by-two factorial trial of 100patients

1316. Mikkonen, Paula (2015) Low back pain and associated factors in adolescence : acohort study

1317. Hakalahti, Antti (2015) Efficacy and safety of radiofrequency catheter ablation inthe treatment of atrial fibrillation

1318. Hannula, Virva (2015) The prevalence of diabetic retinopathy and its effect onsocial well-being and health related quality of life in children and young adultswith type 1 diabetes

1319. Leppänen, Virpi (2015) Epävakaan persoonallisuuden hoitomallitutkimus Oulunmielenterveyspalveluissa

1320. Piira, Olli-Pekka (2015) Effects of emotional excitement on cardiovascularregulation

1321. Sonkajärvi, Eila (2015) The brain's electrical activity in deep anaesthesia : withspecial reference to EEG burst-suppression

UNIVERSITY OF OULU P .O. Box 8000 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

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

University Lecturer Anu Soikkeli

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0983-8 (Paperback)ISBN 978-952-62-0984-5 (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 1322

ACTA

Tiina T

irkkonen

OULU 2015

D 1322

Tiina Tirkkonen

EARLY ATTACHMENT, MENTAL WELL-BEING AND DEVELOPMENT OF FINNISH CHILDREN AT PRESCHOOL AGETWINSHIP – RISK OR OPPORTUNITY?

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF MEDICINE