Area deprivation and child psychosocial problems

Abstract

Background
We examined the association of area deprivation with the occurrence of psychosocial problems among children aged 4–16 in a representative national sample of children based on standardised measures of parent-reported problems and diagnoses made by doctors and nurses working in child healthcare (child health professionals, CHPs).
Methods
The study comprised 4480 children aged 4–16 years, eligible for a routine health assessment (response: 90.1 %), in 19 Child Healthcare Services across the Netherlands that routinely provided preventive child healthcare to nearly all school-aged children. Parents completed the Child Behaviour Checklist (CBCL). CHPs examined the child and interviewed parents and child during their routine health assessments. Main outcome measures concerned psychosocial problems as reported by parents (i. e. a clinical score on the CBCL) and as identified by CHPs.
Results
Prevalence rates of psychosocial problems were 8.6% for parent-reported problems and 10.1 % for CHP-identified problems. They were much higher in the most deprived third of the areas. Odds ratios (95 % confidence intervals) compared with the least deprived third were 1.93 (1.41–2.64) regarding parent-reported problems and 1.76 (1.30–2.38) regarding CHP-identified problems. Regarding parent reports, associations were slightly stronger for behavioural problems than for emotional problems. Less than a quarter of the area differences could be explained by individual and family characteristics.
Conclusions
Child psychosocial problems occur more frequently in deprived areas. Both preventive and curative health services should be better equipped for this concentration of child and adolescent morbidity in deprived areas.

Full text

Page 1

Soc Psychiatry Psychiatr Epidemiol (2005) 40:18–23DOI 10.1007/s00127-005-0850-0
■ Abstract Background We examined the association
of area deprivation with the occurrence of psychosocial
problems among children aged 4–16 in a representative
national sample of children based on standardised mea-
sures of parent-reported problems and diagnoses made
by doctors and nurses working in child healthcare (child
health professionals, CHPs). Methods The study com-
prised 4480 children aged 4–16 years,eligible for a rou-
tine health assessment (response: 90.1%), in 19 Child
Healthcare Services across the Netherlands that rou-
tinely provided preventive child healthcare to nearly all
school-aged children. Parents completed the Child Be-
haviour Checklist (CBCL).CHPs examined the child and
interviewed parents and child during their routine
health assessments.Main outcome measures concerned
psychosocial problems as reported by parents (i.e. a
clinical score on the CBCL) and as identified by CHPs.
Results Prevalence rates of psychosocial problems were
8.6% for parent-reported problems and 10.1% for CHP-
identified problems.They were much higher in the most
deprived third ofthe areas.Odds ratios (95% confidence
intervals) compared with the least deprived third were
1.93 (1.41–2.64) regarding parent-reported problems
and 1.76 (1.30–2.38) regarding CHP-identified prob-
lems. Regarding parent reports, associations were
slightly stronger for behavioural problems than for
emotional problems.Less than a quarter of the area dif-
ferences could be explained by individual and family
characteristics. Conclusions Child psychosocial prob-
lems occur more frequently in deprived areas.Both pre-
ventive and curative health services should be better
equipped for this concentration of child and adolescent
morbidity in deprived areas.
■ Key words mental health – socio-economic factors –
child health – area deprivation – multilevel
Introduction
Psychosocial problems,such as behavioural,emotional,
and educational problems, are highly prevalent among
children and adolescents and may severely interfere
with their everyday functioning [1–3].Child and family
factors such as age [4,5],gender [5],socio-economic sta-
tus [4–7],ethnic background [8–10] and family compo-
sition [4,7,11] have frequently been shown to be associ-
ated with their occurrence.
Recent reviews indicate that the socio-economic po-
sition of the areas in which children and adolescents live
is associated with behavioural and emotional outcomes
[12,13].A UK national study on twins shows that neigh-
bourhood deprivation accounted for about 5% of the ef-
fects of a shared family environment on the occurrence
of behavioural and emotional problems among 2-year-
old twins [14]. As the effect of the area of residence
seems to be largest in late childhood and early adoles-
cence [12,13],this is likely to be a lowest estimate.
Nationwide studies on the occurrence of psychoso-
cial problems in areas of varying deprivation are rare,
however,as most studies concern local or regional sam-
ples with a limited range of the socio-economic (SE) po-
ORIGINAL PAPER
Sijmen A.Reijneveld · Emily Brugman · Frank C.Verhulst · S.Pauline Verloove-Vanhorick
Area deprivation and child psychosocial problems
A national cross-sectional study among school-aged children
Accepted: 4 July 2004
SPPE 850
S.A.Reijneveld,MD,PhD · E.Brugman,MSc ·
S.P.Verloove-Vanhorick,MD,PhD
TNO (Netherlands Organisation of Applied Scientific Research)
Prevention and Health
Leiden,The Netherlands
Prof.S.A.Reijneveld,MD,PhD (?)
Groningen University
Dept.of Health Sciences
P.O.Box 196
9700 AD Groningen,The Netherlands
Tel.: +31-50/363-2860
Fax: +31-50/363-6251
E-Mail: s.a.reijneveld@med.rug.nl
F.C.Verhulst,MD,PhD
Erasmus University Rotterdam
Academic Hospital Rotterdam-Sophia
Rotterdam,The Netherlands
S.P.Verloove-Vanhorick,MD,PhD
Leiden University Medical Centre
Dept.of Paediatrics
Leiden,The Netherlands

Page 2

19
sition of these areas [12, 13, 15]. If national studies are
available [12, 13, 15], they mostly concern unstandard-
ised measures of behavioural and emotional problems
that are based on parent reports and not on assessments
by professionals [15].
The aim of this study is to examine the association of
area deprivation with the occurrence of psychosocial
problems in a representative national sample ofchildren
aged 4–16 including standardised measures of parent-
reported problems and diagnoses made by doctors and
nurses working in child healthcare.
Subjects and methods
We collected data within the framework of the routine preventive
health assessments that are provided regularly to all Dutch children,
from October 1997 to June 1998.The study was approved by the local
Medical Ethical Committee.
■ Participants
We obtained a national sample using a two-stage selection procedure.
In the first stage, a random sample of 19 out of the 63 Dutch Child
Healthcare Services was drawn,after stratification by region and de-
gree of urbanisation of their district.In the second stage,each Service
provided a random sample of 75 children (three classes) for each of
four grades (second, fourth, and seventh of primary school and sec-
ond of secondary school),insofar as they provided preventive health
assessments to these grades.Of the 4970 eligible children aged 4–16,
4480 (90.1%) participated.Differences between responding and non-
responding children by sex,age,ethnic background,and degree of ur-
banisation were small,according to Cohen effect size index w (range
of w, 0.01–0.08) [16]. Further details have been presented elsewhere
[3].
■ Measures and procedures
Data on the occurrence of psychosocial problems were based on par-
ent report and on an assessment by a doctor or nurse working in pre-
ventive child healthcare (child health professional CHP). The Child
Behaviour Checklist (CBCL) was used to measure parent-reported
psychosocial problems.It was mailed to parents along with an invita-
tion for the routine health assessment.Dutch [17],English [18],Turk-
ish [8,10] and Arabic versions were available,depending on the reg-
istered country of birth of the parents.Parents returned the CBCL at
the assessment in a sealed envelope; it was not inspected by the CHP.
The CBCL is a highly reliable and valid measure of behavioural and
emotional problems over the preceding 6 months, used in various
countries and cultures [17–21].It contains 120 problem items on the
basis of which a Total Problems Score can be computed (the higher
the score,the more problematic the child),as well as separate scores
regarding behavioural and emotional problems designated as the Ex-
ternalising and Internalising problems broad-bands,respectively.In-
ternalising problems comprises problems that are mainly within the
individual; it consists of the Withdrawn, Somatic Complaints, and
Anxious/Depressed CBCL syndrome scales. Externalising comprises
problems that mainly involve conflicts with other people and with
their expectations for the child; it consists of the Delinquent and Ag-
gressive Behaviour CBCL syndrome scales.The CBCL Total Problems
score further comprises three syndrome scales that are not allocated
to either Internalising or Externalising problems (Social Problems,
Thought Problems, and Attention Problems). Children are allocated
to a normal range or a clinical (elevated) range, using the 90thper-
centiles of the Dutch normative sample for the Internalising, Exter-
nalising and Total Problems scores as cut-off [17].
CHP-identified psychosocial problems concerned the response of
the CHP to the following question:“Does the child have a psychoso-
cial problem at this moment?”(yes or no).The CHP had to answer this
question after each assessment,based on routine history-taking and
physical examination. Children that only had risk indicators for the
development ofpsychosocial problems,for instance parents with psy-
chiatric problems or other family problems,had to be coded as no be-
cause our aim was to assess existing child problems and not risk fac-
tors for their development (similar to the CBCL).Next,the CHP had
to rate the severity of the problem as mild,moderate,or severe.CHPs
could only participate in the study if they had attended a half-day’s
training session in the research procedures,including answering this
question.
Area deprivation concerned the Dutch general practitioner (GP)
deprivation score,based on data from Statistics Netherlands regard-
ing 1995 [22,23].It is the sum of the standardised scores of each area
(postcode sector) regarding degree of urbanisation (i.e. mean num-
ber of addresses per km2in the area),proportion of ethnic minorities,
mean income per earner (further:‘mean income’) and proportion of
residents aged 16–64 who are dependent on social benefits (further:
‘social benefits’).Regarding the 5% of the population living in the ar-
eas with the most unfavourable scores, Dutch GPs receive an addi-
tional fee for each patient [24],comparable to the UK Jarman-system
[25].For this study,we classified each child based on its home address
and next divided the sample in tertiles of deprivation.
Relevant child background characteristics concerned ethnic back-
ground,parental educational level and employment,and family com-
position.Ethnic background was measured by country of birth of the
child’s parents. On the basis of the migration histories of various
groups living in the Netherlands,it was coded as Dutch-born; from a
(former) Dutch colony (at least one parent born in Surinam or Dutch
Antilles); from countries in which Dutch employers recruited un-
skilled labourers in the 1960s and 1970s (at least one parent born in
Turkey or Morocco);and others.Parental educational level concerned
the highest degree obtained by a parent.Family structure focused on
the number ofparents in the family (biological father and mother,two
parents but at least one not-biological,only one,and other situations
such as two parents of the same gender).All characteristics were reg-
istered by the CHP at the assessment. Categories of all variables are
presented in Table1.
■ Analysis
Firstly, we examined the prevalence of psychosocial problems ac-
cording to the CBCL and to the CHP, by background characteristics
and by area deprivation. In order to focus on the more severe mor-
bidity,we limited our analyses regarding CHP-identification to mod-
erate and severe problems (46.4% of all identified problems). Sec-
ondly,we examined whether differences in these prevalence rates by
area deprivation remained after adjustment for differences in age and
gender by category of area deprivation.Finally,we examined whether
remaining differences could be explained by other background char-
acteristics, parental educational level and employment, ethnic back-
ground and family composition, using multilevel logistic regression
[26].Levels always concerned child and area,and CHP in the models
regarding CHP-identification.In the latter case,we used cross-classi-
fied models because a CHP examines children from a number of ar-
eas, and children from one area can be examined by a number of
CHPs [26,27].We limited the analyses to children with complete data
on the CBCL and on area deprivation (n=4080). We repeated these
analyses for two separate indicators of the SE position of an area,i.e.
‘mean income’and ‘social benefits’,to examine how much these con-
tributed to differences by area deprivation.
Results
Data concerned 4080 children, 483 areas and 78 CHPs.
Table 1 shows the distribution of various child charac-
teristics by levels of area deprivation. Table 2 provides
similar data regarding areas.They show that virtually all

Page 3

20
characteristics of both individual children and of areas
are less favourable in deprived areas.
Table 3 shows prevalence rates of psychosocial prob-
lems by background characteristics and by area depri-
vation. Rates were higher in families without both bio-
logical parents, and with parents of low educational
level, without substantial paid jobs; and of non-Dutch
origin. Furthermore, rates of psychosocial problems
were higher in deprived areas.
Table 4 shows the results of multilevel logistic regres-
sion analyses using clinical CBCL Total Problems,Exter-
nalising and Internalising scores, and moderate or se-
vere CHP-identified problems as outcomes. They show
that prevalence rates ofpsychosocial problems generally
increased by area deprivation. Characteristics of the
child and its family explained only a small proportion of
these increased rates,on average 20.5%.Remaining dif-
ferences are relatively large for clinical CBCL Total Prob-
lems and Externalising problems and for CHP-identi-
fied problems.
Analyses for the separate measures of the area SE po-
sition ‘social benefits’ yielded similar, though slightly
smaller, area differences. For instance, age/gender-ad-
justed odds ratios (95% confidence intervals) for the in-
Least deprivedModerately deprivedMost deprivedP-value
Age (school type)
4–12 years (primary school)
12+ –15 years (secondary school)
<0.0001b
1057 (72.0)
411 (28.0)
684 (53.6)
593 (46.4)
1052 (78.8)
283 (21.2)
Mean age in years (SD)9.0 (3.36) 10.1 (3.56)8.8 (3.13) <0.0001c
0.29b
Gender
Male
Female
723 (49.3)
745 (50.7)
654 (51.2)
623 (48.8)
643 (48.2)
692 (51.8)
Family composition
With both biological parents
With two parents, maximum one biological
One-parent family
Other
Parental educational levela
Low
Medium
High
Parental employmenta
At least one parent >16 h/week
No parent employed >16 h/week
<0.0001b
1340 (91.3)
44 (3.0)
58 (4.0)
26 (1.8)
1134 (88.8)
40 (3.1)
84 (6.6)
19 (1.5)
1053 (78.9)
70 (5.2)
174 (13.0)
38 (2.8)
<0.0001b
429 (30.2)
492 (34.6)
500 (35.2)
403 (33.6)
393 (32.7)
405 (33.7)
521 (40.8)
408 (32.0)
347 (27.2)
<0.0001b
1413 (96.8)
46 (3.2)
1218 (95.8)
53 (4.2)
1154 (87.2)
169 (12.8)
Ethnic background
Dutch
Turkish/Moroccan
Surinamese/Antillean
Other
<0.0001b
1444 (98.4)
0 (0.0)
2 (0.1)
22 (1.5)
1237 (96.9)
4 (0.3)
3 (0.2)
33 (2.6)
1157 (86.7)
67 (5.0)
52 (3.9)
59 (4.4)
All children146812771335
aData were missing for 182 children regarding parental educational level and for 27 children regarding parental employment;
bChi-square tests regarding differences by area deprivation; cF-tests in ANOVA regarding differences by area deprivation
Table1 Distribution of various child
background characteristics by levels of
area deprivation. Values are numbers
(column percentages) of each back-
ground characteristic. Due to rounding,
not all percentages add up to 100%
Least deprived Moderately deprivedMost deprived P-valuea
Mean number of inhabitants (SD)
Mean number of children in the study (SD)
Percentage of families with children (SD)
Mean family size in area (SD)
Mean net income per earner in Euro (SD)b
Percentage social benefits (SD)b
Percentage non-Dutchb
Mean number of addresses per km2b
Number of areas (percentage of total)
2860 (2463)
8.1 (14.9)
63.1 (7.9)
3.15 (0.23)
14678 (2038)
13.0 (5.7)
1.1 (0.6)
348 (420)
182 (37.7)
6209 (3579)
8.0 (10.8)
62.5 (7.3)
3.07 (0.19)
13903 (1017)
19.0 (4.8)
1.8 (1.5)
1092 (800)
160 (33.1)
7747 (3082)
9.5 (15.5)
61.7 (7.6)
2.93 (0.17)
12443 (1013)
28.6 (8.2)
10.5 (11.2)
2540 (1543)
141 (29.2)
<0.0001
0.58
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
aValues of F-tests in ANOVA regarding differences by level of area deprivation; units of analysis in all tests were areas; bArea
deprivation has been defined on the basis of these four area characteristics
Table2 Distribution of various area
characteristics by levels of area depriva-
tion. Values are statistical measures
(numbers or percentages) of each char-
acteristic using areas as units of analysis

Page 4

21
Table3 Prevalence rates of psychosocial problems by child background characteristics and area deprivation
CBCL Problems
Total
CHP-identified
psychosocial problems AllExternalisingInternalising
n n (%)P1
n (%)P1
n (%)P1
n (%)P1
Age (school type)
4–11 years (primary school)
12–16 years (secondary school)
0.870.043 0.0340.17
2793
1287
241 (8.6)
109 (8.5)
254 (9.1)
143 (11.1)
265 (9.5)
96 (7.5)
295 (10.6)
118 (9.2)
Gender
Male
Female
0.480.630.26 0.086
2020
2060
167 (8.3)
183 (8.9)
192 (9.5)
205 (10.0)
189 (9.4)
172 (8.3)
221 (10.9)
192 (9.3)
Family structure
Two biological parents
Two parents, max. one biological
One-parent family
Other
<0.0001 <0.0001<0.0001<0.0001
3527
154
316
83
260 (7.4)
23 (14.9)
51 (16.1)
16 (19.3)
302 (8.6)
26 (16.9)
54 (17.1)
15 (18.1)
280 (7.9)
18 (11.7)
52 (16.5)
11 (13.3)
321 (9.1)
22 (14.3)
56 (17.7)
14 (16.9)
Parental educational level*
Low
Medium
High
0.002 <0.00010.770.049
1353
1293
1252
137 (10.1)
111 (8.6)
80 (6.4)
163 (12.0)
128 (9.9)
79 (6.3)
125 (9.2)
117 (9.0)
102 (8.1)
162 (12.0)
121 (9.4)
112 (8.9)
Parental employment*
At least one parent >16 h/wk
No parent >16 h/week
0.0220.160.006 0.001
3785
268
312 (8.2)
35 (13.1)
359 (9.5)
35 (13.1)
320 (8.5)
37 (13.8)
366 (9.7)
45 (16.8)
Ethnic background
Dutch
Turkish/Moroccan
Surinamese/Antillean
Other
0.26 0.700.008 0.021
3838
71
322 (8.4)
9 (12.7)
8 (14.0)
11 (9.6)
370 (9.6)
8 (11.3)
8 (14.0)
11 (9.6)
332 (8.7)
14 (19.7)
7 (12.3)
8 (7.0)
375 (9.8)
12 (16.9)
7 (12.3)
19 (16.7)
57
114
Area deprivation
Least deprived
Moderately deprived
Most deprived
1468
1277
1335
92 (6.3)
103 (8.1)
155 (11.6)
<0.0001110 (7.5)
114 (8.9)
173 (13.0)
<0.0001106 (7.2)
103 (8.1)
152 (11.4)
<0.0001108 (7.4)
135 (10.6)
170 (12.7)
<0.0001
All children 4080350 (8.6)397 (9.7) 361 (8.8)413 (10.1)
1Values of chi-square tests regarding differences in the proportion of children with the outcome concerned by characteristic
* Data were missing for 182 children regarding parental educational level and for 27 children regarding parental employment
Age/gender-adjustedAdjusted for all background variables
OR (95% CI)OR(95% CI)Decrease
CBCL Total Problems
Least deprived
Moderately deprived
Most deprived
1
1.28
1.93
1
1.26
1.67
(0.92 to 1.79)
(1.41 to 2.64)
(0.90 to 1.77)
(1.20 to 2.32)
7.1%
28.0%
CBCL Externalising Problems
Least deprived
Moderately deprived
Most deprived
1
1.14
1.88
1
1.12
1.69
(0.86 to 1.51)
(1.45 to 2.43)
(0.85 to 1.49)
(1.30 to 2.21)
14.3%
21.6%
CBCL Internalising Problems
Least deprived
Moderately deprived
Most deprived
1
1.09
1.58
1
1.07
1.35
(0.78 to 1.53)
(1.15 to 2.17)
(0.76 to 1.50)
(0.97 to 1.88)
22.2%
39.7%
CHP-identified problems
Least deprived
Moderately deprived
Most deprived
1
1.39
1.76
1
1.37
1.56
(1.03 to 1.87)
(1.30 to 2.38)
(1.01 to 1.85)
(1.14 to 2.13)
5.1%
26.3%
1OR odds ratio; CIconfidence interval; background characteristics concern family structure, parental educational
level and employment; and ethnic background
Table4 Results of multilevel logistic regression
analyses using clinical Child Behaviour Checklist
(CBCL) Total Problems, Externalising and Internalis-
ing scores, and moderate or severe CHP-identified
problems as outcomes1

Page 5

22
termediate and unfavourable levels of social benefits
were 1.46 (1.05 to 2.02) and 1.87 (1.36 to 2.57) regarding
a clinical CBCL Total Problems score, and 1.35 (1.01 to
1.82) and 1.43 (1.05 to 1.94) regarding moderate and se-
vere CHP-identified problems.
Discussion
Our study shows that the prevalence rates of psychoso-
cial problems in a Dutch national sample of school-aged
children were about 80% higher in the most deprived
third of the areas than in the least deprived one. Look-
ing at the type of parent-reported problems, associa-
tions were slightly stronger for behavioural problems
than for emotional problems,but they were rather sim-
ilar for parent-reported problems compared with mod-
erate and severe CHP-identified problems. A relatively
small part,less than a quarter,of the area differences can
be explained by individual and family characteristics as
studied.
■ Explanations for our findings
The concentration of health problems such as child psy-
chosocial problems in deprived urban areas can be ex-
plained in two ways. Firstly, low socio-economic status
(SES) children concentrate in these deprived areas.
Among them, psychosocial problems occur more fre-
quently [4–7].Secondly,factors at area level,i.e.the en-
vironment in which children live,may contribute to the
occurrence of psychosocial problems,in addition to the
aforementioned contribution of individual SES.Our re-
sults show that the latter is the case to a large degree as
adjustment for characteristics ofthe individual child ex-
plains only a small part of the area differences.
Our finding of a relatively strong association of the
area context with child psychosocial problems calls for
an explanation. Leventhal and Brooks-Gunn identified
three factors [12]: lack of institutional resources in de-
prived areas such as health- and day care; child-parent
relationships in which the parents transfer their own
economic, social and health difficulties and resulting
psychological problems to the child;and a lack of norms
and collective efficacy in these areas that shape child be-
haviour. Regarding the latter, it is also interesting that
area differences regarding CBCL Externalising prob-
lems are somewhat larger than those regarding CBCL
Internalising problems,which may be a result of the en-
vironment shaping behaviour more than emotion. Our
measure of area deprivation has been designed to mea-
sure lack of primary care and other institutional re-
sources,and has been proven to be highly valid as such
[24]. Additional research is needed on the explanatory
power of the other pathways.
Methodological aspects may have influenced our
findings,but do not explain them.Firstly,we used post-
code sectors as units of analyses,which may not be per-
ceived by people as their neighbourhood [12,15].How-
ever, a previous study has shown that differences be-
tween postcode sectors and neighbourhoods as units of
analyses are small, with postcode sectors performing
slightly better [28].Secondly,more elaborate psychiatric
assessments might yield more valid estimates of child
psychosocial problems. However, parent-report and
CHP-identification showed very similar associations,
which supports their validity.We previously also showed
that in all school-aged children,CHPs identify problems
in 57% of the children with a clinical CBCL Total Prob-
lems score,and that this percentage was higher for prob-
lems rated as moderate or severe by the CHP [3].Finally,
the response in our study was high and we used mea-
sures of proven validity.
■ Implications
Our study shows large differences in child psychosocial
problems by area deprivation. Firstly, prevention in
these areas should be reinforced. This concerns both
preventive services and policies aiming at fields outside
healthcare such as educational services and urban plan-
ning [29,30].
Secondly,primary care and other child healthcare fa-
cilities should be made available to a larger degree in de-
prived areas to cope with the much larger number of
child psychosocial problems in these areas.This implies
that systems for additional financing,similar to those in
primary care,may be needed [24,25].The results of the
present study show a large child morbidity to be treated
and prevented in these areas.
■ Acknowledgements This study was financially supported by the
Dutch Ministry of Health, Welfare and Sports, the Praeventiefonds
(grant No. 28–2628–2), and Zorg Onderzoek Nederland (grant
No.101004–12).It was carried out by TNO Prevention and Health in
co-operation with the Dutch Association of Municipal Public Health
Services. The authors wish to thank the 19 Dutch Municipal Health
Services that participated in this study for their co-operation.
References
1. Leaf PJ,Alegria M, Cohen P, Goodman SH, Horwitz SM, Hoven
CW, Narrow WE, Vaden-Kiernan M, Regier DA (1996) Mental
health service use in the community and schools: results from
the four-community MECA Study. Methods for the Epidemiol-
ogy of Child and Adolescent Mental Disorders Study.J Am Acad
Child Adolesc Psychiatry 35:889–897
2. Verhulst FC,van der Ende J (1997) Factors associated with child
mental health service use in the community. J Am Acad Child
Adolesc Psychiatry 36:901–909
3. Brugman E,Reijneveld SA,Verhulst FC,SP (2001) Identification
and management of psychosocial problems by preventive child
health care.Arch Pediatr Adolesc Med 155:462–469
4. 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 sta-
tus.J Am Acad Child Adolesc Psychiatry 37:1255–1261

Page 6

23
5. Moffitt TE, Caspi A (2001) Childhood predictors differentiate
life-course persistent and adolescence-limited antisocial path-
ways among males and females.Dev Psychopathol 13:355–375
6. McLeod JD,Nonnemaker JM (2000) Poverty and child emotional
and behavioral problems: racial/ethnic differences in processes
and effects.J Health Soc Behav 41:137–161
7. Harland P, Reijneveld SA, Brugman E, SP, Verhulst FC (2002)
Family factors and life events as risk factors for behavioural and
emotional problems in children. Eur Child Adolesc Psychiatry
11:176–184
8. Bengi-Arslan L,Verhulst FC, van der Ende J, Erol N (1997) Un-
derstanding childhood (problem) behaviors from a cultural per-
spective:comparison of problem behaviors and competencies in
Turkish immigrant,Turkish and Dutch children.Soc Psychiatry
Psychiatr Epidemiol 32:477–484
9. Crijnen AA, Bengi-Arslan L, Verhulst FC (2000) Teacher-re-
ported problem behaviour in Turkish immigrant and Dutch chil-
dren: a cross-cultural comparison. Acta Psychiatr Scand 102:
439–444
10. Sowa H,Crijnen AA,Bengi-Arslan L,Verhulst FC (2000) Factors
associated with problem behaviors in Turkish immigrant chil-
dren in The Netherlands.Soc Psychiatry Psychiatr Epidemiol 35:
177–184
11. Agerbo E,Nordentoft M,Mortensen PB (2002) Familial,psychi-
atric, and socioeconomic risk factors for suicide in young peo-
ple: nested case-control study.BMJ 325:74
12. Leventhal T,Brooks-Gunn J (2000) The neighborhoods they live
in: the effects of neighborhood residence on child and adoles-
cent outcomes.Psychol Bull 126:309–337
13. Ingoldsby EM,Shaw DS (2002) Neighborhood contextual factors
and early-starting antisocial pathways. Clin Child Fam Psychol
Rev 5:21–55
14. Caspi A,Taylor A,Moffitt TE,Plomin R (2000) Neighborhood de-
privation affects children’s mental health: environmental risks
identified in a genetic design.Psychol Sci 11:338–342
15. Boyle MH,Lipman EL (2002) Do places matter? Socioeconomic
disadvantage and behavioral problems of children in Canada. J
Consult Clin Psychol 70:378–389
16. Cohen J (1988) Statistical power analysis for the behavioral sci-
ences.Lawrence Erlbaum,Hillsdale
17. Verhulst FC, van der Ende J, Koot HM (1996) Manual for the
CBCL/4–18 (in Dutch). Erasmus University/Dept of Child and
Adolescent Psychiatry,Sophia Childrens’Hospital,Rotterdam
18. Achenbach TM (1991) Manual for the Child Behavior Check-
list/4–18 and 1991 Profile. University of Vermont, Dept of Psy-
chiatry,Burlington (VT)
19. Verhulst FC, van der Ende J (1995) The eight-year stability of
problem behavior in an epidemiologic sample. Pediatr Res 38:
612–617
20. Crijnen AA,Achenbach TM,Verhulst FC (1997) Comparisons of
problems reported by parents of children in 12 cultures: total
problems,externalizing,and internalizing.J Am Acad Child Ado-
lesc Psychiatry 36:1269–1277
21. Weisz JR,McCarty CA (1999) Can we trust parent reports in re-
search on cultural and ethnic differences in child psychopathol-
ogy? Using the bicultural family design to test parental culture
effects.J Abnorm Psychol 108:598–605
22. Statistics Netherlands (1997) Kerncijfers vierpositie postcodege-
bieden 1995 (Keyfigures on four digit postcode-areas 1995).
Voorburg/Heerlen,Statistics Netherlands
23. Reijneveld SA (2002) Neighbourhood socioeconomic context
and self reported health and smoking: a secondary analysis of
data on seven cities.J Epidemiol Community Health 56:935–942
24. Verheij RA,de Bakker DH,Reijneveld SA (2001) GP income in re-
lation to workload in deprived urban areas in The Netherlands.
Before and after the 1996 pay review. Eur J Public Health 11:
264–266
25. Jarman B (1983) Identification of underprivileged areas.Br Med
J (Clin Res Ed) 286:1705–1709
26. Goldstein H, Rasbash J, Plewis I, Draper D, Browne W,Yang M,
Woodhouse G,Healy M (1998) A user’s guide to MLwiN.Univer-
sity of London,Multilevel Models Project,London
27. Goldstein H (1995) Multilevel statistical models.Edward Arnold,
London
28. Reijneveld SA,Verheij RA, de Bakker DH (2000) The impact of
area deprivation on differences in health: does the choice of the
geographical classification matter? J Epidemiol Community
Health 54:306–313
29. Lock K (2000) Health impact assessment.BMJ 320:1395–1398
30. van Herten LM,Reijneveld SA,Gunning-Schepers LJ (2001) Ra-
tionalising chances ofsuccess in intersectoral health policy mak-
ing.J Epidemiol Community Health 55:342–347