Introduction
Psychosocial problems in children
Psychosocial problems among children cause serious limitations in daily functioning (Brugman et al. Reference Brugman, Reijneveld, Verhulst and Verloove-Vanhorick2001) and have large social and economic consequences for family and society (World Health Organization, 2001). The term ‘psychosocial problems’ is used for an extensive group of phenomena and are frequently divided into two areas: externalizing problems (behavioural problems) and internalizing problems (emotional problems). In particular, behavioural problems such as oppositional-defiant disorder, conduct disorder and attention-deficit and hyperactivity disorder (ADHD) and emotional problems such as anxiety disorders and depression are frequently diagnosed among youths (Verhulst et al. Reference Verhulst, van der Ende, Ferdinand and Kasius1997b). Both externalizing and internalizing problems are related to various negative outcomes on adult age. For example, conduct problems during childhood are associated with crime and drugs use (Fergusson et al. Reference Fergusson, Horwood and Ridder2005). In addition, internalizing problems in adolescents are related to increased anxiety and depression in early adulthood (Pine et al. Reference Pine, Cohen, Gurley, Brook and Ma1998; Weissman et al. Reference Weissman, Wolk, Goldstein, Moreau, Adams, Greenwald, Klier, Ryan, Dahl and Wickramaratne1999). In general, children with high levels of parent-reported emotional and behavioural problems are more likely to meet the criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) psychiatric diagnoses in adulthood than children without these problems (Hofstra et al. Reference Hofstra, van der Ende and Verhulst2002).
Prevalence rates of psychosocial problems vary with age, methods used to assess these problems and the sample. The prevalence of psychosocial problems in children and pre-adolescents has been estimated in several European countries (i.e. Great Britain, Germany, Italy and The Netherlands) and appears to be approximately 10–15% (Goodman et al. Reference Goodman, Ford, Simmons, Gatward and Meltzer2000; Meltzer et al. Reference Meltzer, Gatward, Goodman and Ford2000; Crone et al. Reference Crone, Vogels, Hoekstra, Treffers and Reijneveld2008; Ravens-Sieberer et al. Reference Ravens-Sieberer, Wille, Erhart, Bettge, Wittchen, Rothenberger, Herpertz-Dahlmann, Resch, Holling, Bullinger, Barkmann, Schulte-Markwort and Dopfner2008; Frigerio et al. Reference Frigerio, Rucci, Goodman, Ammaniti, Carlet, Cavolina, De Girolamo, Lenti, Lucarelli, Mani, Martinuzzi, Micali, Milone, Morosini, Muratori, Nardocci, Pastore, Polidori, Tullini, Vanzin, Villa, Walder, Zuddas and Molteni2009).
Sex differences
Boys more often have psychosocial problems than girls (Zeijl et al. Reference Zeijl, Crone, Wiefferink, Keuzenkamp and Reijneveld2005) and are three to four times more diagnosed as having conduct disorders than girls (Loeber et al. Reference Loeber, Burke, Lahey, Winters and Zera2000). The difference in prevalence of psychosocial problems is partly attributable to the fact that boys have more externalizing disorders, which are more visible to others (Costello et al. Reference Costello, Mustillo, Erkanli, Keeler and Angold2003). During adolescence, the prevalence of mood disorders increases in girls whereas it remains stable in boys. These differences persist during adulthood (Hankin et al. Reference Hankin, Abramson, Moffitt, Silva, McGee and Angell1998).
Risk factors
Several factors have been associated with a higher risk for the development of psychosocial problems. Psychosocial problems are more often found in children with parents with a low socio-economic status, in children whose mother was younger than 27 years old at time of birth of the child, and in children living in a single parent family (Kroes et al. Reference Kroes, Kalff, Steyaert, Kessels, Feron, Hendriksen, van Zeben, Troost, Jolles and Vles2002; Zeijl et al. Reference Zeijl, Crone, Wiefferink, Keuzenkamp and Reijneveld2005; Amone-P'Olak et al. Reference Amone-P'Olak, Burger, Ormel, Huisman, Verhulst and Oldehinkel2009). Children living with single mothers are twice as likely to have definite or severe difficulties with psychosocial problems that interfere in their everyday life than children living with two parents (Simpson et al. Reference Simpson, Bloom, Cohen, Blumberg and Bourdon2005). Furthermore, children who experience stressful life events in their early years, for example migration and loss of a family member, are more at risk for developing externalizing problems in the later childhood (Mesman & Koot, Reference Mesman and Koot2001; Kroes et al. Reference Kroes, Kalff, Steyaert, Kessels, Feron, Hendriksen, van Zeben, Troost, Jolles and Vles2002). Also, in children with chronic physical illness, the risk of psychosocial problems is doubled compared to healthy children (Bilfield et al. Reference Bilfield, Wildman and Karazsia2006). Finally, family conflicts, parental strain and growing up with a step-parent are related to mental health problems in adolescents (Wille et al. Reference Wille, Bettge and Ravens-Sieberer2008).
Prevention
Screening and treatment of psychosocial problems may improve the prognosis of children with psychosocial problems, especially when the screening occurs with good validated instruments (Durlak & Wells, Reference Durlak and Wells1998). Still, a minority of the children with psychosocial problems is in contact with the mental health care (Verhulst & van der Ende, Reference Verhulst and van der Ende1997a; Brugman et al. Reference Brugman, Reijneveld, Verhulst and Verloove-Vanhorick2001). The prevention of depression is one of the specific areas of attention in the Dutch collective prevention. It is known that having psychosocial problems in childhood is a risk factor for negative outcomes (i.e., crime, drug use and psychiatric diagnosis) at adult age (Reinherz et al. Reference Reinherz, Giaconia, Hauf, Wasserman and Silverman1999; Fergusson et al. Reference Fergusson, Horwood and Ridder2005). Therefore, it is important to know the prevalence of psychosocial problems among children. So far, empirical data about psychosocial problems in Dutch children aged 8–12 years and the relationship with risk factors and quality of life are sparse. Information about prevalence of psychosocial problems can be used by the Municipal Health Service and municipalities to set priorities for their policy in order to prevent or reduce the prevalence of psychosocial problems among young children.
Aim of the study
Therefore, the aim of this current study was (a) to determine the prevalence of psychosocial problems among Dutch children aged 8–12 years and (b) to study its association with risk factors and quality of life. A large population-based sample of 2703 children (1392 boys and 1311 girls) was used. The parents or caregivers completed the relatively new, validated and reliable Strengths and Difficulties Questionnaire (SDQ) and the KIDSCREEN, measuring psychosocial problems and quality of life, respectively. Data about several important mediating risk factors were collected. All data were analysed separately for boys and girls because of hypothesized sex differences.
Methods
Recruitment and sample selection
This study was conducted within the framework of a community-based health study carried out in two regions in the north west of The Netherlands in 2007. The presented data were part of a larger study in children aged 0–12 years conducted by the Municipal Health Service and induced by the law on public health. The overall aim of the total study was to describe the health situation of the children.
In the regions Kop van Noord-Holland and West-Friesland, a random sample of children aged 0–12 years was drawn per municipality from the Municipal Population Registration. Before starting with the sample selection, the population of one of the two regions was additionally stratified for age groups (0–4, 4–8 and 8–12 years).
The questionnaires about psychosocial problems, quality of life and several risk factors were completed by the parents and/or caregivers (proxies), because most children in this sample were too young to formulate a response. Parents are commonly used as proxies who provide information about the behaviour of children (Najman et al. Reference Najman, Williams, Nikles, Spence, Bor, O'Callaghan, Le Brocque, Andersen and Shuttlewood2001). In this study a population-based study sample of children aged 8–12 years were included.
Procedure
A postal questionnaire was sent to the parents/caregivers of the randomly selected children. On the first page of the questionnaire a letter was printed in which the study was introduced and explained. The questionnaires had to be returned by mail. In order to raise the response, a reminder was sent to the parents after 3 weeks. The parents/caregivers in West-Friesland also received a second reminder: a new questionnaire was sent 3 weeks after the first reminder. The data were collected from February to April 2007.
Response
In the Kop van Noord-Holland, the questionnaire was sent to the parents of 2360 children aged 8–12 years. A total of 1497 questionnaires (response: 63%) were appropriate for analyses. For West-Friesland, the questionnaire was sent to the parents of 1654 children aged 8–12 years. A total of 1206 questionnaires were suitable for analysis (response: 73%). The overall response rate was 67%.
Measures
Psychosocial problems
Psychosocial problems in children were measured with the extended version of the SDQ, completed by their parents/caregivers. The SDQ is a relatively short questionnaire where strengths as well as difficulties of a child in the previous month are being asked (Goodman et al. Reference Goodman, Ford, Simmons, Gatward and Meltzer2000). It consists of 25 items that can be grouped in the following five 5-item subscales: hyperactivity/inattention, emotional symptoms, conduct problems, peer relationship problems and prosocial behaviour. Each item has three answer possibilities, to which can be assigned a value (0 = ‘not true’, 1 = ‘somewhat true’, 2 = ‘certainly true’). For each subscale, the scores can be summed, by which a sub score for each scale is calculated. This score varies from 0 to 10. High scores on the prosocial behaviour subscale indicate strengths, whereas high scores on other subscales indicate difficulties. A total problem score can be calculated by adding the scores of the subscales, with the exception of the prosocial behaviour scale. The total SDQ score varies between 0 and 40. A total SDQ score of ≥14 indicates psychosocial problems (Goedhart et al. Reference Goedhart, Treffers and Van Widenfelt2003). The cut-offs for problems on the subscales are as follows: emotional problems ≥5, conduct problems ≥4, hyperactivity problems ≥7 and peer relationship problems ≥4. We defined prosocial behaviour problems as scores of <5 on the prosocial behaviour subscale.
An extended version of the SDQ includes an impact supplement of eight questions, asking if the respondent thinks the child has a problem, and explores the chronicity, distress, social impairment and burden for others. This provides useful additional information about the degree of impact of problems on the child's daily life. From these questions, an impact score can be calculated (range 0–10). An impact score of ≥2 indicates raised impact of the problems on the child's daily life. Because of its brevity, the SDQ is suitable and commonly used in epidemiological studies. The SDQ is a validated, reliable and responsive measure with good psychometric properties (Goodman, Reference Goodman2001; Muris et al. Reference Muris, Meesters and van den Berg2003; van Widenfelt et al. Reference van Widenfelt, Goedhart, Treffers and Goodman2003). Concurrent validity with several psychopathology measures (e.g. Child Behaviour Checklist (CBCL), Youth Health Questionnaire) was good (Muris et al. Reference Muris, Meesters and van den Berg2003; van Widenfelt et al. Reference van Widenfelt, Goedhart, Treffers and Goodman2003) and high SDQ scores were associated with DSM-IV diagnosis of several psychiatric disorders in children (Goodman, Reference Goodman2001).
Risk factors
Information about risk factors, i.e. education level of the parents, ethnicity, family structure, level of income of the household, chronic diseases and life events was obtained by the questionnaire. Parental education level was included as indicator of the social economic status and concerned the highest degree obtained by a parent. It was divided into three categories: low (lower secondary education or below), medium (upper secondary education) and high (tertiary education). Ethnicity was categorized as Dutch nationality, non-Dutch western nationality and non-western nationality, based upon the country of birth of the child and of the parents. The family structure is categorized into whether the child lives in a single parent family or not (all other family structures). Income of the household was divided into below or equal to a modal level (net income ≤1750 euro/month), and above the modal level, according to the classification of income by Statistics Netherlands (CBS) in 2006. A list of 15 chronic diseases commonly seen in children and a list of 22 life events were included in the questionnaire. Parents could mark the diseases that the child had and life events that the family had experienced. Chronic diseases and life events were summed and categorized into ≥1 v. none chronic diseases and life events.
Quality of life
Quality of life is determined with the KIDSCREEN-10 proxy version for parents or primary caregivers (Ravens-Sieberer et al. Reference Ravens-Sieberer, Gosch, Abel, Auquier, Bellach, Bruil, Dur, Power and Rajmil2001). Quality of life can be described as a multidimensional construct covering physical, emotional, mental, social and behavioural components of well-being and function as perceived by patients or other individuals (Ravens-Sieberer et al. Reference Ravens-Sieberer, Gosch, Abel, Auquier, Bellach, Bruil, Dur, Power and Rajmil2001). The KIDSCREEN is a quality of life measurement that is applicable to healthy and chronically ill children and adolescents aged 8–18 years. The KIDSCREEN-10 is the short version of the KIDSCREEN instrument (The KIDSCREEN Group Europe, 2006) and consists of ten questions about quality of life in the previous week. The reliability and validity was tested in 13 participating European countries. For the KIDSCREEN-10 proxy version, a satisfactory item internal consistency has been shown. The overall reliability was good (Cronbach's alpha: 0.78) and the agreement between youth and proxy report was good (intra class correlation coefficient: 0.56). The correlation between the self-report and the proxy version of the KIDSCREEN-10 index was 0.57. The Dutch population norms were available for the parent version, for the total group of children as well as for males and females separately. For each item of the KIDSCREEN-10, a score of 1–5 can be obtained, resulting in a total quality of life score with a range of 10–50. Higher scores indicate a better quality of life.
Data analysis
All analyses were performed separately for boys and girls because of the statistically significant effect modification by sex of the relations under condition. Descriptive data (means, standard deviation and percentage) were presented for the total study sample and for boys and girls separately. Differences in study sample characteristics by sex were examined using Student t test for continuous variables and χ2-tests for categorical variables. As no differences in study sample characteristics were found between the two regions in the prevalence of psychosocial problems (SDQ score ≥14) and on any of the five SDQ subscales, the data of the two regions were merged and analysed together. χ 2-tests were used to study the differences in psychosocial problems and the five SDQ subscales scores by sex, using the predefined cut-off scores for the SDQ and its subscales. Next, univariable and multivariable stepwise logistic regression analysis were performed to determine the influence of the risk factors on psychosocial problems. To investigate multicollinearity, the variance-inflation factor (VIF) was determined in a regression model including all risk factors. All VIF values were between 1.00 and 1.81, indicating no multicollinearity. In sensitivity analysis, we assessed the correlation between the risk factors and each SDQ subscale, except for prosocial behaviour problems because its prevalence was too low to reliably assess its association with the risk factors. Because of the skewed distribution of the total SDQ score and KIDSCREEN score, we used the non-parametric Spearman's rho to assess the association between psychosocial problems (both SDQ total and SDQ subscale scores) and quality of life (the total KIDSCREEN-10 score). For all statistical testing, we used two-sided hypothesis testing with an alpha level of 0.05. Statistical analyses were performed using the SPSS 15 software package for Windows.
Results
Study sample
The characteristics of the study sample are presented in Table 1. A total of 2703 children were included in the analysis (51.5% (n = 1392) boys and 48.5% (n = 1311) girls). Boys had significantly more often a chronic disease, experienced a life event and had a non-Dutch nationality. No difference between boys and girls was found for parental education level, family structure and income level.
Table 1. Characteristics of the study sample by sex
*Standard deviation.
Significant differences (p < 0.05) between sex are printed bold and cursive.
Prevalence of psychosocial problems
The prevalence of psychosocial problems (SDQ score ≥14) in the total sample was 10.4%. Boys had more psychosocial problems than girls (13.9% v. 6.6%, OR = 2.28; 95% CI = 1.75–2.97; Fig. 1). Boys have significantly more hyperactivity/inattention, conduct, peer relationship and prosocial behaviour problems compared with girls, but not regarding emotional problems. Hyperactivity and attention deficit problems were the most common problems (14.8%) in boys. Emotional problems (7.5%) were the problems most frequently found in girls. In 9.6% of all children, the impact score was ≥2, indicating that their problems interfere definitely or severely with the child's everyday life in home life, friendships, classroom learning or leisure activities. The raised impact of psychosocial problems in children's daily life was more often found in boys than in girls (11.6% v. 7.6%, OR = 1.59; 95% CI = 1.22–2.08; data not shown).
Fig. 1. Prevalence rates (%) for psychosocial problems (SDQ-score ≥14) and the five SDQ subscales scores by sex (N = 2703). *Significant (p < 0.05) difference in prevalence rates for psychosocial problems between boys and girls.
Risk factors
The odds ratio (OR) and the 95% CI of the univariable logistic regression are presented in Table 2. Univariable analyses revealed that having one or more chronic diseases or having experienced one or more life event was associated with psychosocial problems. Low parental educational level was associated with psychosocial problems compared to high educational level, but not to medium educational level. Living in a single parent family and having a below modal or similar to modal household income were also associated with psychosocial problems. Region, ethnicity and age were not related to psychosocial problems.
Table 2. Univariable OR for high levels of psychosocial problems (SDQ-score≥14) by sex with risk factors
The reference categories are: Kop van Noord-Holland, aged 8–9 years, low educational level, Dutch nationality, non-single parent family, above modal income, no chronic diseases diagnosed, no life events experienced.
Significant relationships (p < 0.05) are printed bold and cursive.
When the stepwise multivariable logistic regression was performed (Table 3), only four of the eight variables contributed significantly to the model. One or more chronic diseases, having experienced life event(s) and an income below the modal level are associated with psychosocial problems. Parental educational level was associated with psychosocial problems in boys but not in girls.
Table 3. Multivariable OR for high levels of psychosocial problems (SDQ-score≥14) by sex with risk factors
The reference categories are: low educational level, above modal income, no chronic diseases diagnosed, no life events experienced. Significant relationships (p < 0.05) are printed bold and cursive.
In sensitivity analysis, we observed that chronic diseases and life events were significantly related to elevated scores on each SDQ subscale in the total sample (range OR: 1.70–2.69 and OR: 1.76–3.68, respectively). In addition, low income was related to the presence of emotional problems and peer problems. Furthermore, single parent family was associated with conduct problems and hyperactivity problems. A younger age was related to conduct problems, and low education level was related to hyperactivity problems (data not shown).
Quality of life
Psychosocial problems are significantly inversely related to children's quality of life in the total sample (Spearman's rho = −0.47, p < 0.001). Because virtually similar associations were found for boys and girls, we only report associations for the total group. From the correlations between the five subscales of the SDQ and children's quality of life, emotional problems showed the largest negative correlation with quality of life (rho = −0.42), followed by peer relationship problems (rho = −0.33, p < 0.001), hyperactivity problems (rho = −0.30, p < 0.001) and conduct problems (rho = −0.28, p < 0.001). As expected, a positive correlation was found for the subscale pro-social behaviour (rho 0.21, p < 0.001). Then 40.6% of the children with psychosocial problems had a low quality of life compared with 4.0% of the children with no psychosocial problems (p < 0.001) (data not shown). A high quality of life was found in 9.8% of the children with psychosocial problems, while 53.1% of the children without psychosocial problems had a high quality of life (data not shown).
Discussion
In this Dutch population-based study, we observed that psychosocial problems are prevalent in children aged 8–12 years and are related to a reduced quality of life. In this present study, we found that 10.4% of children aged 8–12 years had psychosocial problems. Our prevalence of psychosocial problems in children is in line with other studies that used the SDQ (Goodman et al. Reference Goodman, Ford, Simmons, Gatward and Meltzer2000; Crone et al. Reference Crone, Vogels, Hoekstra, Treffers and Reijneveld2008) or other methods to assess psychosocial problems (Meltzer et al. Reference Meltzer, Gatward, Goodman and Ford2000; Ravens-Sieberer et al. Reference Ravens-Sieberer, Wille, Erhart, Bettge, Wittchen, Rothenberger, Herpertz-Dahlmann, Resch, Holling, Bullinger, Barkmann, Schulte-Markwort and Dopfner2008; Frigerio et al. Reference Frigerio, Rucci, Goodman, Ammaniti, Carlet, Cavolina, De Girolamo, Lenti, Lucarelli, Mani, Martinuzzi, Micali, Milone, Morosini, Muratori, Nardocci, Pastore, Polidori, Tullini, Vanzin, Villa, Walder, Zuddas and Molteni2009).
Zeijl et al. (Reference Zeijl, Crone, Wiefferink, Keuzenkamp and Reijneveld2005) found that 6% of the Dutch children aged 8–12 have psychosocial problems. This proportion is lower than the prevalence found in our study. However, this prevalence was based on CBCL that was completed by the parents, which uses American reference scores for psychosocial problems (Zeijl et al. Reference Zeijl, Crone, Wiefferink, Keuzenkamp and Reijneveld2005). The use of a different outcome measure is most likely explaining the higher prevalence of psychosocial problems in our sample.
Furthermore, we found that psychosocial problems, in particular externalizing symptoms such as hyperactivity and conduct problems, are more common in boys than in girls. This is consistent with other studies (Brugman et al. Reference Brugman, Reijneveld, Verhulst and Verloove-Vanhorick2001; Zeijl et al. Reference Zeijl, Crone, Wiefferink, Keuzenkamp and Reijneveld2005) but is in contrast with Crone et al. (Reference Crone, Vogels, Hoekstra, Treffers and Reijneveld2008), where no significant difference between boys and girls was found. However, the study sample of Crone et al. (Reference Crone, Vogels, Hoekstra, Treffers and Reijneveld2008) was considerably smaller (n = 707) than our study and therefore possibly no difference could be detected. Achenbach et al. (Reference Achenbach, McConaughy and Howell1987) stated that parents are well able to assess the externalizing problems of a child but that they are less accurate in the appraisal of internalizing problems of the child. Because boys more often have externalizing problems that are more visible to others and girls more often have internalizing problems that are more difficult to observe (Costello et al. Reference Costello, Mustillo, Erkanli, Keeler and Angold2003), it is possible that the psychosocial problems in girls are generally underestimated. This might also explain the similar levels of emotional problems for boys and girls reported by the parent.
This study confirmed that having a chronic disease is a risk factor for psychosocial problems, as found elsewhere (Bilfield et al. Reference Bilfield, Wildman and Karazsia2006). Furthermore, life events were related to psychosocial problems. Harland et al. (Reference Harland, Reijneveld, Brugman, Verloove-Vanhorick and Verhulst2002) measured the association of several life events with psychosocial problems. They found that parental divorce and having an ill or hospitalized family member were significantly related to psychosocial problems in children.
In our study, age was not related to psychosocial problems. This could be due to the small age range. The prevalence of psychosocial problems in this study was not related to ethnicity. This is in accordance with another study conducted in the Netherlands (Zwirs et al. Reference Zwirs, Burger, Schulpen, Wiznitzer, Fedder and Buitelaar2007) but in contradiction with the results of Harland et al. (Reference Harland, Reijneveld, Brugman, Verloove-Vanhorick and Verhulst2002).
Psychosocial problems are related to a lower quality of life. The correlations between psychosocial problems and quality of life (Spearman's rho = −0.47) can be seen as a moderately negative association. In a European validation study of the KIDSCREEN instrument, an effect size of 0.67 was found, indicating a moderate to large association between the SDQ and the KIDSCREEN-10 (The KIDSCREEN Group Europe, 2006). This association is higher than the association we found. However, the European study was not limited to children aged 8–12 years and was conducted in a different study population.
Several strengths of our study need to be emphasized. First, we used data from a large population-based sample. Second, we used validated and reliable instruments for measuring both psychosocial problems and quality of life. Finally, the outcomes are presented separately for boys and girls because of the effect modification by sex.
Our study also has several limitations. First, the present study has a cross-sectional design, thus we cannot infer conclusions regarding causality. Second, the response was about 67% in the two regions. In this analysis, a non-response analysis could not be performed. In Muris et al. (Reference Muris, Meesters and van den Berg2003) and Van Widenfelt et al. (Reference van Widenfelt, Goedhart, Treffers and Goodman2003), it was found that children whose parents did not respond, had higher self-rated psychosocial problems. It is possible that the response here may underestimate the real prevalence of problems. Third, another concern is the high item non-response for the question about level of income. Yet, additional analyses showed that the prevalence of psychosocial problems did not differ between responders and non-responders on this question. The same risk factors for psychosocial problems were found when the level of income was excluded from the model, with the addition of living in a single parent family. However, it is likely that single parent families generally have lower incomes than two parent families and therefore a significant contribution of family structure is found in the model without the level of income. Fourth, the prevalence is based on single informants, i.e. parents or caregivers. Measurements of psychosocial health based on several informants (e.g. parents, children and teachers) are found to be more accurate (Goodman et al. Reference Goodman, Ford, Simmons, Gatward and Meltzer2000). However, SDQs completed by parents are better predictors of psychiatric disorders than SDQs completed by adolescents themselves (Goodman et al. Reference Goodman, Ford, Simmons, Gatward and Meltzer2000). When the children's self-report of the KIDSCREEN is available, this should be used instead of the proxy report. The proxy report can be regarded as complementary information (The KIDSCREEN Group Europe, 2006). Finally, the present study was limited to a population living in a rural area in the northwest of The Netherlands. This population largely consisted of Dutch Caucasians. A previous study in the Netherlands showed that the prevalence of psychosocial problems is higher in immigrant children (Reijneveld et al. Reference Reijneveld, Harland, Brugman, Verhulst and Verloove-Vanhorick2005). It is unclear whether the relationship between psychosocial problems, risk factors and quality of life is similar in other racial and ethnic groups from more urban areas.
Conclusion
Our results give important insight into the prevalence of psychosocial problems and its relationship with risk factors and quality of life for both Dutch boys and girls separately. The identified risk factors in this study can be used for developing targeted prevention strategies aimed at children at high risk for psychosocial problems. Moreover, further well-designed prospective research is needed, to study the differences in psychosocial problems across sex and racial/ethnic groups over time ideally using parents, children and teachers reports.
Declaration of Interest
None declared. Financial support was provided by the Ministry of Health, Welfare and Sport of the Netherlands.
