Early Prevention of Adult Antisocial Behaviour - Assets - Cambridge

26 Pages·2003·0.13 MB·English

Early Prevention of Adult
Antisocial Behaviour
Editedby
David P. Farrington
UniversityofCambridge
Jeremy W. Coid
StBartholomew’sandtheRoyalLondonSchoolofMedicineandDentistry
PUBLISHEDBYTHEPRESSSYNDICATEOFTHEUNIVERSITYOFCAMBRIDGE
ThePittBuilding,TrumpingtonStreet,CambridgeCB21RP,UnitedKingdom
CAMBRIDGEUNIVERSITYPRESS
TheEdinburghBuilding,CambridgeCB22RU,UK
40West20thStreet,NewYork,NY10011-4211,USA
477WilliamstownRoad,PortMelbourne,VIC3207,Australia
RuizdeAlarco´n13,28014Madrid,Spain
DockHouse,TheWaterfront,CapeTown8001,SouthAfrica
http://www.cambridge.org
(cid:1)C CambridgeUniversityPress2003
Thisbookisincopyright.Subjecttostatutoryexception
andtotheprovisionsofrelevantcollectivelicensingagreements,
noreproductionofanypartmaytakeplacewithout
thewrittenpermissionofCambridgeUniversityPress.
Firstpublished2003
PrintedintheUnitedKingdomattheUniversityPress,Cambridge
TypefacePlantin10/12pt SystemLATEX2ε []
AcataloguerecordforthisbookisavailablefromtheBritishLibrary
ISBN0521651948hardback
Contents
Listoffigures pageix
Listoftables x
Listofcontributors xi
Preface xiii
1 Advancingknowledgeabouttheearlypreventionofadult 1
antisocialbehaviour
 . 
2 Formulatingstrategiesfortheprimarypreventionof 32
adultantisocialbehaviour:“Highrisk”or“population”
strategies?
 . 
3 Riskfactorsforadultantisocialpersonality 79
 ,  .  
 . 
4 Preventingtheintergenerationalcontinuityofantisocial 109
behaviour:Implicationsofpartnerviolence
 .    
5 Protectivefactorsandresilience 130
    
6 Preventionduringpregnancy,infancyandthepreschool 205
years
 .    
7 Preventionthroughfamilyandparentingprogrammes 243
 
8 Preventionintheschoolyears 265
.     . 
vii
viii Contents
9 Preventionofantisocialbehaviourinfemales 292
 -
10 Economiccostsandbenefitsofprimarypreventionof 318
delinquencyandlateroffending:Areviewoftheresearch
 . 
11 Conclusionsandthewayforward 356
 .    . 
Index 369
Figures
2.1 Personsfoundguiltyof,orcautionedfor,indictable page34
offencesper100,000populationbyagegroup,1999
2.2 Fourstagesofcriminalcareers 37
2.3 DiagnosticcriteriaforDSM-IV,AxisII.301.7Antisocial 39
PersonalityDisorder
2.4 DiagnosticcriteriaforDSM-IV312.8ConductDisorder 40
componentofAntisocialPersonalityDisorder
2.5 Lifetimerates/100ofAntisocialPersonalityDisorder 43
(DSM-III,DSM-III-R)basedoncommunitysurveysor
surveysofrelatives
2.6 Lifetimerates/100ofAntisocialPersonalityDisorderin 47
prisonsurveys
2.7 Schematicrepresentationoftherelationbetweenriskof 56
crimeandthedistributionofdifferentlevelsofexposure
toacause
2.8 Vulnerabilityandstressfactorsrelatingtoadverseadult 59
outcome.KauaiLongitudinalStudy
2.9 Protectivefactorsrelatingtopositiveoutcome.Kauai 60
LongitudinalStudy
3.1 AhierarchicaldevelopmentalmodelofADHD,ODD,CD, 84
andantisocialpersonality
5.1 Hypotheticalmodelandempiricalresultsontherelationship 140
betweenaccumulatedriskfactors,accumulatedprotective
factors,andtheintensityofproblembehaviour
5.2 Asimplemodelofcumulatingrisksinthedevelopmentof 144
persistentantisocialbehaviour
6.1 Frequencyofhitting,biting,andkickingforboysandgirls 206
aged2to11years
6.2 Possiblelinksbetweenmanipulatedfactorandoutcomefor 208
developmentalpreventionexperiments
ix
Tables
3.1 DefinitionsofAntisocialPersonalityDisorder, page82
PsychopathicPersonalityDisorder,andDissocial
PersonalityDisorder
5.1 Multilevelexamplesofpotentialprotectivefactors 179
againstantisocialbehaviour
6.1 Studieswithjuveniledelinquencyasanoutcome 210
6.2 Studieswithsociallydisruptivebehaviourasan 212
outcome
6.3 Studieswithcognitiveskillsasanoutcome 214
6.4 Studieswithfamilycharacteristicsasanoutcome 216
10.1 Summaryofprimaryandsecondaryprevention 326
studies
10.2 EconomicbenefitsoftheQuantumOpportunities 344
Programme
x
1 Advancing knowledge about the early
prevention of adult antisocial behaviour
David P. Farrington
Themainaimsofthisbookaretoreviewwhatisknownaboutthecauses
and prevention of adult antisocial behaviour. The book aims to specify
whatweknow,whatwedonotknow,andwhatweneedtoknow,recom-
mending priority research that would address key questions and fill key
gaps in knowledge. The main aim of this introductory chapter is to set
the scene for the more detailed chapters that follow by outlining some
of the key topics, issues and questions arising in the early prevention of
adult antisocial behaviour. This chapter defines the territory by briefly
reviewing epidemiology, development, risk and protective factors, and
preventionprogrammes.
Fourtypesofpreventioncanbedistinguished(TonryandFarrington,
1995).Criminaljusticepreventionreferstotraditionaldeterrence,inca-
pacitationandrehabilitationstrategiesoperatedbylawenforcementand
criminal justice agencies. Situational prevention refers to interventions
designed to reduce the opportunities for antisocial behaviour and to in-
crease the risk and difficulty of committing antisocial acts. Community
prevention refers to interventions designed to change the social condi-
tions and social institutions (e.g. community norms and organisations)
thatinfluenceantisocialbehaviourincommunities.Developmentalpre-
vention refers to interventions designed to inhibit the development of
antisocial behaviour in individuals, by targeting risk and protective fac-
torsthatinfluencehumandevelopment(seeFarrington,2000a).
This book concentrates on early developmental prevention pro-
grammes, including those implemented in pregnancy and infancy, par-
enting programmes, preschool programmes, individual skills training,
andschoolprogrammes.Manyoftheseinvolveprimaryprevention,tar-
geting unselected individuals in the whole community, but secondary
preventionprogrammestargetingchildrenatriskarealsoreviewed.The
focusofthebookisonriskfactorsandearlypreventioninchildhoodand
adolescence; for reviews of risk factors and early interventions for con-
ductdisorderanddelinquency,seeFarrington(1999)andRutter,Giller
andHagell(1998).
1
2 D.Farrington
Definitionandmeasurement
Definitionofantisocialbehaviour
There is clearly a syndrome of adult antisocial behaviour defined by a
cluster of antisocial symptoms. This syndrome is given different names
indifferentcountriesanddifferentclassificationsystems:antisocialper-
sonalitydisorderinDSM-IV(AmericanPsychiatricAssociation,1994),
dissocial personality disorder in ICD-10 (World Health Organisation,
1992)andpsychopathicdisorderintheEnglishMentalHealthAct1983,
forexample.
Both types of behaviour and features of personality are included in
the antisocial behaviour syndrome. Types of behaviour include prop-
erty crimes such as burglary, violent crimes, drug use, heavy drinking,
drunk or reckless driving, sexual promiscuity or risky sex behaviour,
divorce/separation or unstable sexual relationships, spouse or partner
abuse, child abuse or neglect, unemployment or an unstable employ-
ment history, debts, dependence on welfare benefits, heavy gambling,
heavysmoking,andrepeatedlyingandconning.Personalityfeaturesin-
clude impulsiveness and lack of planning, selfishness and egocentricity,
callousness and lack of empathy, lack of remorse or guilt feelings, low
frustrationtoleranceandhighaggressiveness.
An important question is the relative importance of behavioural and
personality symptoms in defining antisocial personality disorder. Hare
andhiscolleagues(e.g.Hare,HartandHarpur,1991)haveconsistently
criticised the DSM criteria for antisocial personality as too behavioural
andinsufficientlyconcernedwithpersonalityfeatures.Hare’sPsychopa-
thyChecklist(PCL-R)distinguishestwofactors.Factor1consistsofper-
sonalityfeaturessuchasegocentricity,lackofremorse,andcallousness,
while factor 2 describes an impulsive, antisocial, and unstable lifestyle.
Theproblemisthatsomefeaturesofanantisociallifestyle(e.g.unemploy-
mentanddependenceonwelfarebenefits)mayeitherreflectanantisocial
personalityormaybecausedbycircumstancesoutsidetheperson’sown
control. Because of this, it is desirable to include both behavioural and
personalityfeaturesinthedefinitionofantisocialpersonality.
Anotherimportantquestioniswhetherindividualsdifferqualitatively
(in kind) or quantitatively (in degree) in antisocial personality (Clark,
LivesleyandMorey,1997).Peoplecanbescoredaccordingtotheirnum-
ber of symptoms. For example, Robins and her colleagues (e.g. Robins
and Price, 1991) have consistently argued that the number of child-
hood conduct disorder symptoms predicts the number of adult antiso-
cial behaviour symptoms, rather than any specific childhood behaviour
Advancingknowledgeofearlyprevention 3
predicting a specific adult behaviour. The key problem is where to set
the boundary between normal and pathological, or between health and
illness. Existing boundaries depend largely on clinical judgement. For
example, according to DSM-IV, ‘only when antisocial personality traits
areinflexible,maladaptiveandpersistentandcausesignificantfunctional
impairmentorsubjectivedistressdotheyconstituteAntisocialPersonal-
ityDisorder’(AmericanPsychiatricAssociation,1994,p.649).Farmore
is known about the early prevention of particular types of antisocial be-
haviourthanabouttheearlypreventionofantisocialpersonalitydisorder
orpsychopathy.
Measurementofantisocialbehaviour
Antisocial behaviour can be measured in a variety of ways. Interviews
by psychiatrists are necessary to yield psychiatric diagnoses in Great
Britain, where explicit diagnostic criteria are not as widely used as in
North America. However, psychiatrist interviews are not very practical
for large-scale epidemiological studies. One possible strategy is to use a
two-stage procedure in which the population is initially screened using
briefsymptomquestionnaires(e.g.Bebbingtonetal.,1981).Then,more
intensiveclinicalinterviewscanbegiventoallthosewithhighsymptom
scoresandtoarepresentativesampleofthosewithlowscores.
Another possible method is to use an interview designed for non-
clinicians, such as the NIMH Diagnostic Interview Schedule used in
theEpidemiologicalCatchmentAreaproject(RobinsandRegier,1991).
Ratingsorchecklistscompletedbyinformantssuchasinstitutionalstaff
can also be used, based on interviews and records, as in the case of
thePCL-R(Hare,1991).Alternatively,semi-structuredinterviewswith
informants such as relatives or close friends can be used, as with the
Standardized Assessment of Personality (Pilgrim and Mann, 1990), or
psychologicaltestsandself-completionquestionnairescanbeused(e.g.
Blackburn,1975).
Itisimportantwithallmeasurementtechniquestoassessvalidityand
reliability. However, one problem in assessing validity is that the exter-
nal criterion for antisocial personality disorder or psychopathy is often
basedonpsychiatricdiagnoses,whichmayhavelowreliability(Malgady,
Rogler and Tryon, 1992). It is especially important to measure the pre-
dictive validity of instruments given at a relatively early age or stage of
development.
Inthischapter,IwillrefertoresultsobtainedintheCambridgeStudy
in Delinquent Development, which is a prospective longitudinal survey
of 411 South London males from age 8 to age 46 (Farrington, 1995,
4 D.Farrington
2002c).Atage32,ameasureofantisocialpersonalitywasdevised,based
onthefollowingtwelveitems:convictedinthelastfiveyears,self-reported
offender,involvedinfights,drug-taker,heavydrinker,poorrelationship
with parents, poor relationship with wife/cohabitee, divorced/child liv-
ingelsewhere,frequentunemployment,anti-establishmentattitude,tat-
tooed, and impulsive (Farrington, 1991). These were measured in a
structured social interview. The reliability of this scale was 0.71, and
the worst quarter of the males had four or more adverse features out of
twelve.
Inter-relationshipsbetweenbehaviours
In general, all the behavioural and personality symptoms listed above
tend to be intercorrelated, since people who show one of them have an
increased risk of also showing any other. For example, the two factor
scores of the PCL-R are highly intercorrelated (over 0.5: Hare et al.,
1991),andthetotalPCL-Rscoreishighlycorrelatedwiththediagnosis
ofantisocialpersonalitydisorder(0.67inHare,1985).Comorbidityisa
common finding, and it is assumed that all of the symptoms reflect the
sameunderlyingtheoreticalconstruct.However,itisimportanttoquan-
tifythedegreeofversatilityinantisocialbehaviour,andtoassesswhether
itismorereasonabletoassumetwoormoreunderlyingconstructsrather
than only one. Another important question is whether conclusions are
differentwithcontinuousasopposedtodichotomousmeasuresofsymp-
toms.
To the extent that intercorrelated clusters of symptoms are identified
within the general category of antisocial behaviour, it may be useful to
distinguish typologies of individuals. For example, Moffitt (1993) dis-
tinguished between ‘life-course-persistent’ individuals, who began their
antisocial behaviour at an early age and persisted for a long time, and
‘adolescence-limited’ ones who began later and desisted earlier. How-
ever,itisunclearhowfarthesecategoriesdifferindegreeratherthanin
kind.
Epidemiologyanddevelopment
Epidemiology
It is important to establish the prevalence of antisocial symptoms, and
of antisocial personality disorder, at different ages. It is useful to deter-
minethepeakagesofdifferenttypesofantisocialbehaviour,andthepeak
agesforaccelerationanddecelerationinprevalence.Informationisalso