Environmental risk factors associated with the persistence of conduct difficulties in children with intellectual disabilities and autistic spectrum disorders

Research in Developmental Disabilities 35 (2014) 3508–3517

Contents lists available at ScienceDirect

Research in Developmental Disabilities

Environmental risk factors associated with the persistence of conduct difficulties in children with intellectual disabilities and autistic spectrum disorders Eric Emerson a,c,*, Jan Blacher b, Stewart Einfeld a, Chris Hatton c, Janet Robertson c, Roger J. Stancliffe a a b c

University of Sydney, Australia University of California at Riverside, USA Lancaster University, UK

A R T I C L E I N F O

A B S T R A C T

Article history: Received 12 June 2014 Received in revised form 28 August 2014 Accepted 28 August 2014 Available online 18 September 2014

We investigated the association between exposure to environmental risks in early childhood and the prevalence and persistence of conduct difficulties (CD) in children with intellectual disability (ID) who did not have autistic spectrum disorder (ASD), children with ASD and typically developing (TD) children. Results indicated that: (1) exposure to risk was associated with elevated prevalence of CD at age three and, for TD children and children with ID, increased risk of CD persisting to ages five and seven; (2) at all levels of risk, children with ASD were more likely to show persistent CD than other children; (3) children with ID were no more likely to show persistent CD than TD children at low levels of exposure to environmental risk. ß 2014 Elsevier Ltd. All rights reserved.

Keywords: Conduct difficulties Problem behavior Socio-economic position Environmental factors Autism Intellectual disability

1. Introduction Children with intellectual or developmental disabilities show markedly higher rates of behavioral difficulties (‘challenging’ or externalizing problematic behaviors) than their non-disabled peers (Baker et al., 2003; Einfeld, Ellis, & Emerson, 2011; Totsika, Hastings, Emerson, Lancaster, & Berridge, 2011). The presence of such behaviors, especially if persistent over time, can have a detrimental impact on child wellbeing and the wellbeing of siblings and parents (Emerson & Einfeld, 2011). Behavioral difficulties such as aggression are relatively common in typically developing children, with prevalence reaching a peak at age two to three years after which it reduces markedly (Broidy et al., 2003; Cairns, Cairns, Neckerman, Ferguson, & Garie´py, 1989; Nagin & Tremblay, 1999; Tremblay, 2000, 2006; Tremblay et al., 1999, 2004). While early onset of behavioral difficulties predicts aggression and a range of adverse personal outcomes in later life (Campbell, Shaw, & Gilliom, 2000; Moffitt & Scott, 2008; Tremblay, 1999, 2000, 2006; Tremblay et al., 1999, 2004), outcomes for children who show persistent behavioral difficulties across early and middle childhood are particularly poor (Petitclerc & Tremblay, 2009; Tremblay, 2006; Tremblay et al., 2004). As such, there has been a growing interest in identifying personal and environmental

* Corresponding author at: Lancaster University, UK. E-mail address: [email protected] (E. Emerson). http://dx.doi.org/10.1016/j.ridd.2014.08.039 0891-4222/ß 2014 Elsevier Ltd. All rights reserved.

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3509

factors that are associated with different trajectories of behavioral difficulties over time. This body of research has highlighted the importance of the additive or interactive effects of multiple risk factors including: genetic factors; executive dysfunction; temperament; social cognition; and exposure to environmental risk factors such as family poverty, less than optimal parenting practices and neighborhood deprivation (Moffitt & Scott, 2008; Tremblay, 2006, 2012; Tremblay et al., 2004; Vitaro & Tremblay, 2008). It has also led to the development of a range of targeted preventative interventions for children at risk of developing conduct difficulties (Boisjoli, Vitaro, Lacourse, Barker, & Tremblay, 2007; Conduct Problems Prevention Research Group, 2011; Lochman, Wells, Qu, & Chen, 2012; Prinz, Sanders, Shapiro, Whitaker, & Lutzker, 2009; Sanders, 2008; Tremblay, Pagani-Kurtz, Masse, Vitaro, & Pihl, 1995; Vitaro, Barker, Brendgen, & Tremblay, 2012; WebsterStratton & Taylor, 2001; Webster-Stratton, Reid, & Stoolmiller, 2008; Zubrick et al., 2005). In contrast, relatively little longitudinal research has been undertaken on the persistence of behavioral difficulties over time in children with intellectual or developmental disabilities. The results of this body of research have drawn attention to the potential importance of a range of factors, with increased persistence being related to: increased severity of intellectual disability (Einfeld et al., 2006; Gray et al., 2012); male gender (Gray et al., 2012); child regulatory strategies (Gerstein, Crnic, Ryu, Baker, & Blacher, 2011); poorer maternal health and wellbeing (Baker, Neece, Fenning, Crnic, & Blacher, 2010; Totsika et al., 2013), although for conflicting results see (Eisenhower, Blacher, & Baker, 2013); family adaptability (Baker, Seltzer, & Greenberg, 2011); parenting (Baker et al., 2010); and, for children with borderline intellectual disability, exposure to environmental disadvantage (Emerson, Einfeld, & Stancliffe, 2011). Interestingly, the latter association was not found in a study of children with autism (Gray et al., 2012). The aims of the present study are to add to this emerging body of research by examining the relationship between exposure to a range of environmental risk factors at ages nine months and three years and the persistence of conduct difficulties from age three years to ages five and seven in population-based cohorts of British children with intellectual disability or autism spectrum disorder (ASD). We hypothesize that exposure to environmental risk factors at ages nine months and three years will be associated with increased prevalence of behavioral difficulties at age three years and increased persistence of behavioral difficulties from age three to ages five and seven years. We used the Family Stress Model (FSM) (Conger et al., 1992; Conger & Donnellan, 2007) to guide our selection and categorization of indicators of environmental risk factors. The FSM posits that the association between exposure to low socio-economic position (SEP) and the developmental health of children is mediated by the impact of stressors associated with low SEP on the mental health of parents and the relationship quality between parents which in turn has a deleterious impact on parenting behaviors and practices. 2. Method The study is based on secondary analysis of the first four waves of data collected by the UK’s Millennium Cohort Study (MCS). MCS data are managed by the Centre for Longitudinal Studies at the University of London and are available to researchers registered with the Economic and Social Data Service (www.esds.ac.uk) through its data archive (www.dataarchive.ac.uk). Full details of the design of MCS are available in a series of reports and technical papers (Hansen, Jones, Joshi, & Budge, 2010; Hansen, 2012; Johnson, 2009, 2012; Jones & Ketende, 2010; Plewis, 2007; Plewis & Ketende, 2006), key aspects of which are summarized below. 2.1. Sampling Participant families were randomly selected from Child Benefit Records, a non means-tested welfare benefit available to all UK children at the time the cohort was established. Sampling was geographically clustered to include all four countries of the UK (England, Wales, Scotland, Northern Ireland), and disproportionately stratified to over-sample children from ethnic minority groups and disadvantaged communities (Plewis, 2007). Children and families were drawn from 398 randomly selected electoral wards in the UK. The first survey (MCS1) took place when children were nine months old and included a total of 18,552 families. Children were followed up at ages three (MCS2; 15,590 families, 84% retention rate from MCS1), five (MCS3; 15,246 families, 82% retention rate from MCS1) and seven (MCS4; 13,857 families, 75% retention rate from MCS1). For each family, information was collected on the target child falling within the designated birth date window. For multiple births (e.g., twins, triplets) information was collected on all children. To avoid the statistical problems associated with the clustering of multiple births within households, the present analyses are restricted to one randomly selected target child in multiple birth households. 2.2. Identification of children with intellectual disability and autistic spectrum disorder 2.2.1. Intellectual disability Child cognitive ability was assessed at age three using the Bracken School Readiness Assessment (Bracken, 2002) and Naming Subscale of the British Ability Scales (BAS) (Elliott, Smith, & McCulloch, 1997), selected subscales of the BAS at ages five and seven, and the NFER Progress in Maths test at age seven (Hansen, 2012). For ages five and seven we extracted the first component (‘g’) from a principle component analysis of all age-standardised subscale/test scores. The first component accounted for 63% of score variance at age seven and 55% of score variance at age five. We identified children as having

3510

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

intellectual disability if they scored two or more standard deviations below the mean on the first principle component at age seven (n = 419 [3.3%] of 12,820 children for whom test results were available). Interviewers did not administer the assessments if the child ‘has a learning disability/serious behavioral problem (e.g., severe ADHD, autism) which prevents them from carrying out the assessments’, ‘is unable to respond in the required manner for each assessment, e.g., reading, writing, manipulating objects’, ‘is not able to speak or understand English (or Welsh if applicable)’ or if consent and co-operation were not forthcoming. If cognitive test scores were missing at age seven, we identified children as having intellectual disability if they scored two or more standard deviations below the mean on the first principal component at age five (n = 146 [6.5%] of 2250 children). If cognitive test scores were missing at age five and at age seven, we identified children as having intellectual disability if they scored two or more standard deviations below the mean on the Bracken School Readiness Assessment at age three (n = 49 [4.4%] of 1105 children). If Bracken scores were not available, we identified children as having intellectual disability if they scored two or more standard deviations below the mean on the BAS Naming Subscale at age three (n = 54 [7.6%] of 711 children). This process allowed us to classify intellectual disability on the basis of cognitive test scores for 99.1% of children participating at age seven (MCS4). For 125 children no cognitive test results were available at any age. Cognitive testing was not administered for a variety of reasons including lack of parental consent, failure to co-operate with testing and severity of child disability. For these children we identified intellectual disability on the basis of parental report at age seven. A child was identified as having intellectual disability if both of the following two criteria were met: (1) the child was reported to be receiving special education due to their ‘learning difficulty’ (the term used in educational services in the UK to refer to intellectual disability); (2) the child was reported to have ‘great difficulty’ in all three areas of reading, writing and maths. This led to the identification of another 11 children as having intellectual disability. This procedure led to the identification of 522 of the 13,857 (3.8%) children participating at age seven as having intellectual disability. Boys were significantly more likely than girls to be identified as having intellectual disability (4.6% vs 2.7%; OR = 1.75, 95%CI 1.45–2.11). Given the distribution of test scores (and of intelligence itself) very few of the children identified as having intellectual disability will have had severe intellectual disability. For example, using estimates of the prevalence of severe intellectual disability suggests that MCS is likely to contain less than 40 children with severe or profound intellectual disability (Maulik, Mascarenhas, Mathers, Dua, & Saxena, 2011). 2.2.2. ASD ASD was identified on the basis of key informant report at age seven (in 96.7% of cases the child’s biological mother) to two questions: (1) ‘Has [name] ever been diagnosed by a doctor as having autism or Asperger’s syndrome?’; (2) ‘Has [name] been identified as having special educational needs? If so, was it for ASD?’ An affirmative response to either question led to the child being identified as having ASD (n = 224, 1.6%). The majority of children with ASD (62%) were identified by positive answers to both questions. Boys were significantly more likely than girls to be identified as having ASD (2.9% vs 0.6%; OR = 5.26, 95%CI 3.68–7.52). Children with ASD were significantly more likely to be identified as having intellectual disability (18.2% vs 3.4%; OR = 6.22, 95%CI 4.40–8.81). 2.3. Measures 2.3.1. Child behavior problems The Conduct Difficulties subscale of the Strengths and Difficulties Questionnaire (SDQ: www.sdqinfo.com) (Goodman, 1999) was used to measure children’s conduct difficulties at ages three, five and seven. The SDQ was developed to measure difficulties and predict likely clinical levels of behavior problems in the community (Goodman, 2001). It is now extensively used in research and clinical practice. It has been shown to maintain its good psychometric properties with children with intellectual disabilities and children with autism (Emerson, 2005; Iizuka et al., 2010). In our sample, internal consistency (Cronbach’s alpha) of the five-item conduct difficulties subscale for the full sample was 0.68 at age three, 0.56 at age five and 0.60 at age seven. Subscale scores were converted to a binary measure for ‘borderline or abnormal vs no’ conduct difficulties using the recommended scale cut points (see www.sdqinfo.com). 2.3.2. Environmental risk factors The MCS includes a range of indicators of potential environmental risks. We used the Family Stress Model (Conger et al., 1992; Conger & Donnellan, 2007) to guide our selection and categorization of indicators of environmental risk factors into four general classes: (1) household poverty and neighborhood deprivation; (2) maternal resources; (3) maternal health and wellbeing; (4) parenting behaviors. We restricted the selection of indicators to those collected at ages nine months (MCS1) and three years (MCS2) (i.e., preceding the period of persistence of conduct difficulties we were investigating). 2.3.2.1. Household poverty and neighborhood deprivation. Five indicators of household poverty and neighborhood deprivation were available: (1) income poverty defined as living in a household whose equivalised income was 60% less than the national median (Emerson, Graham, & Hatton, 2006); (2) experiencing material hardship (at MCS1 not owning three or more material assets from a list of eight, e.g., refrigerator, microwave; at MCS2 not being able to afford two or more of a list of nine

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3511

goods/activities that have been identified as basic necessities for UK families, e.g., a warm weatherproof coat for each child (Pantazis, Gordon, & Levitas, 2006); (3) living in a ‘workless’ household (a household in which no adult was in employment for more than 16 h a week); (4) living in a household that key informants reported to suffer from damp or condensation; (5) living in a neighborhood within the lowest quintile of scores on the Index of Multiple Deprivation for their country (Emerson et al., 2006; Johnson, 2009). These five variables were combined into a cumulative risk exposure scale assigning one point for each exposure across the first two interview waves and, given that exposure to persistent poverty has a particularly damaging impact on the developmental health of children (Hobcraft & Kiernan, 2010), an additional point if the child was exposed to the same risk factor (e.g., income poverty) on both waves, giving a potential range of 0–15. The scale had excellent internal consistency (Alpha = 0.88). Given the non-normal distribution of scores on the scale and the relatively small sample sizes for the ASD group in particular, it was converted for descriptive analyses by taking a median split (0–2 risks vs 3–15 risks) to create a simple binary measure. 2.3.2.2. Maternal resources. Two indicators of maternal resources were selected: (1) low educational attainment; and (2) being a single parent when the child was age three. Low educational attainment was defined as having a highest educational qualification less than a General Certificate of Secondary Education at Grade ‘C’ (Department for Education, 2013). 2.3.2.3. Maternal health and wellbeing. Three indicators of poorer maternal health and wellbeing were collected at both interview waves: (1) low life satisfaction (scoring in the bottom quartile [below 7] on a 0–10 scale of overall life satisfaction); (2) possible mental health problem (MCS1 scoring positive on four or more of an abbreviated nine item version of the Malaise Scale developed by the MCS research team [within sample internal consistency = 0.73]; MCS2 scoring above the recommended cut-off for possible common mental health disorder on the K6 screen [within sample internal consistency = 0.85]) (Kessler et al., 2002, 2003; Rutter, Tizard, & Whitmore, 1970); (3) reporting having ‘fair’ or ‘poor’ health to an item about overall general health status or reporting having a limiting long-standing illness. An additional item (low self-efficacy) was only available at MCS1. Low self-efficacy was defined as scoring in the lowest quartile in response to the following three items: ‘I usually have a free choice and control over my life’; ‘I never really seem to get what I want out of life’ (reverse scored); ‘Usually I can run my life more or less as I want to’. As for household poverty and neighborhood deprivation, these items were combined into a cumulative risk exposure scale (potential range 0–9) that had acceptable internal consistency (Alpha = 0.77). Again, given the non-normal distribution of scores on the scale and small sample size of the ASD group it was converted for descriptive analyses by taking a median split to create a simple binary measure. 2.3.2.4. Parenting behaviors. Two indicators of parenting behavior were selected: (1) inconsistent parenting (parental report that the child often did not have both regular mealtimes and a regular bedtime); (2) harsh parenting (parental report that they shouted at or smacked their child when they misbehaved more than rarely). These measures were only collected when the child was three years of age. 2.3.3. Total risk exposure A total risk exposure scale was derived by summing scores from the Household Poverty and Neighborhood Deprivation Scale, the Maternal Health and Wellbeing Scale and the indicators for Parental Resources and Parenting Behaviors. The scale has good internal consistency (Alpha = 0.81). Given the non-normal distribution of scores on the scale it was converted for descriptive analyses by into a three point ordinal scale (low, medium and high risk) using tercile cut points based on the smallest analytic group (children with ASD). 2.4. Approach to analysis To address missing data (item non-response) for the risk factors, multiple imputation was employed in IBM SPSS 19. This involved the imputation of five parallel data sets. The results presented are pooled estimates from separate analyses run on the five data sets. All analyses were undertaken using sample weights to correct for the oversampling of specific populations and the effects of unit non-response bias in recruitment and retention. In order to create mutually exclusive groupings the ‘intellectual disability’ group was restricted to children who did not have ASD (n = 449, 3.4%), whereas the ASD group included children with and children without intellectual disability (n = 224). Due to the small numbers involved it was not possible to include a separate group of children with both intellectual disability and ASD. 2.5. Ethical approval Ethical approval for the MCS1 was granted by the South-West Multi-Centre Research Ethics Committee (England), and by the London Multi-Centre Research Ethics Committee for MCS2-4. The current study is a secondary analysis of MCS data, and the ethical responsibilities of the present authors included the protection of participants’ anonymity and confidentiality.

3512

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3. Results 3.1. Child disability status and risk exposure Prevalence of exposure to environmental risk factors at ages nine months and three years is presented in Table 1. As can be seen, there were significant between group differences in exposure to all categories of risk. On all indicators children with intellectual disabilities (but not ASD) were significantly more likely to be exposed to environmental risk than typically developing children. Children with ASD were significantly more likely to be exposed to environmental risk than typically developing children for five of the six indicators, the exception being level of parental education. Relative to children with ASD, children with intellectual disability were significantly more likely to be exposed to household poverty/neighborhood deprivation, low maternal education and inconsistent parenting. 3.2. The prevalence of conduct difficulties at age three The prevalence of conduct difficulties at age three was 49% for typically developing children (the reference group), 62% for children with intellectual disability (OR = 1.69 (1.37–2.08), p < 0.001) and 66% for children with ASD (OR = 2.00 (1.48–2.70), p < 0.001). Unadjusted associations between environmental risk factors and the prevalence of conduct difficulties at age three are presented in Table 2 separately for typically developing children, children with intellectual disability and children with ASD. As can be seen, the strength of association between indicators of environmental risk and conduct difficulties was very similar for typically developing children and children with intellectual disability. For children with ASD the associations were weaker for all indicators, with none reaching the conventional level of statistical significance. However, none of the differences in strength of association between children with ASD and children with either intellectual disability or typically developing children were statistically significant with the point estimate of risk for the latter two groups all falling within the 95% confidence intervals of risk for the ASD group. For all groups of children there were weak to moderate associations between the overall number of risks to which children were exposed at ages nine months and three years and the prevalence of conduct difficulties (Spearman’s r typically developing children = 0.243, p < 0.001; children with intellectual disability r = 0.274, p < 0.001; children with ASD r = 0.220, p < 0.01). Given the similar strength of association between total risk exposure and prevalence across the three groups multivariate logistic regression was used to estimate the extent to which between-group differences in prevalence of conduct difficulties Table 1 Prevalence of exposure to environmental risk at ages nine months and three years by child disability status. Risk factor

Typically developing children (n = 11,103)

Children with ID but not ASD (n = 449)

Children with ASD (n = 224)

Test statistics x2 = (df)

High household poverty/neighborhood deprivation

36%

71%

54%

Overall 258.5(2)*** ID/ASD 19.3(1)*** TD/ID 230.9(1)*** TD/ASD 32.9(1)***

Low maternal education

27%

58%

29%

Overall 197.9(2)*** ID/ASD 49.0(1)*** TD/ID 197.8(1)*** TD/ASD 0.5(1)

Single parent status

16%

32%

25%

Overall 76.0(2)*** ID/ASD 3.1(1) TD/ID 66.6(1)*** TD/ASD 11.4(1)**

Low maternal wellbeing

51%

58%

59%

Overall 15.5(2)*** ID/ASD 0.2(1) TD/ID 24.9(1)*** TD/ASD 7.1(1)**

Inconsistent parenting

13%

26%

18%

Overall 65.2(2)*** ID/ASD 6.1(1)* TD/ID 62.2(1)*** TD/ASD 4.1(1)*

Harsh parenting

26%

33%

33%

Overall 11.9(2)** ID/ASD 0.0(1) TD/ID 7.6(1)** TD/ASD 4.6(1)*

* p < 0.05. ** p < 0.01. *** p < 0.001.

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3513

Table 2 Unadjusted associations (odds ratios with 95% confidence intervals) between environmental risk factors and the prevalence of conduct difficulties at age three. Risk factor

Typically developing children (n = 11,103)

Children with ID but not ASD (n = 449)

Children with ASD (n = 224)

High household poverty/neighborhood deprivation Low maternal education Single parent status Low maternal wellbeing Inconsistent parenting Harsh parenting

2.03*** 1.89*** 1.94*** 2.08*** 1.81*** 3.34***

2.05** (1.34–3.12) 1.81** (1.19–2.74) 1.58* (1.00–2.50) 2.36*** (1.51–3.69) 1.78 (0.96–3.29) 3.86*** (2.26–6.60)

1.35 1.41 0.94 1.83 1.03 1.89

(1.87–2.21) (1.73–2.06) (1.73–2.17) (1.90–2.27) (1.50–2.19) (3.04–3.67)

(0.73–2.49) (0.71–2.82) 0.45–1.95) (0.95–3.54) (0.30–3.54) (0.96–3.71)

* p < 0.05. ** p < 0.01. *** p < 0.001.

at age three may be attributable to between-group differences in levels of exposure to environmental risk. Compared to typically developing children, unadjusted risk (odds ratio) of showing conduct difficulties at age three was 1.71 (1.38–2.13), p < 0.001 for children with intellectual disabilities and 2.02 (1.49–2.73), p < 0.001 for children with ASD compared to typically developing children. Adjustment for between-group differences in levels of exposure to environmental risk eliminated the elevated risk of conduct difficulties for children with intellectual disabilities (OR = 1.06 (0.84–1.34), p = n.s.) and reduced the elevated risk of conduct difficulties for children with ASD (OR = 1.65 (1.20–2.25), p < 0.01). 3.3. The persistence of conduct difficulties from age three to age five and age seven Of those children who exhibited conduct difficulties at age three 24% of typically developing children (the reference group), 54% of children with intellectual disability (but not ASD) (OR = 4.33 (3.07–6.12), p < 0.001) and 70% of children with ASD (OR = 6.83 (4.35–10.71), p < 0.001) also exhibited conduct difficulties at both ages five and seven. In the analyses presented below we compared children who showed ‘persistent’ conduct difficulties (i.e., at age three, five and seven) with those who showed conduct difficulties at age three but not at any subsequent age. Unadjusted associations between environmental risk factors and the persistence of conduct difficulties to ages five and seven are presented in Table 3 separately for typically developing children, children with intellectual disability (but not ASD) and children with ASD. Data presented in Table 3 are restricted to participants who: (1) exhibited conduct difficulties at age three; and (2) for whom age 5 and age 7 conduct difficulties data are available. As can be seen, while these associations were broadly similar for typically developing children and children with intellectual disability, children with ASD showed a different pattern of relationships between environmental risk and persistence of conduct difficulties. However, only one of the twelve differences in strength of association between children with ASD and children with either intellectual disability or typically developing children was statistically significant (stronger association between low maternal wellbeing and persistence for children with intellectual disability) with the point estimate of risk for the children with intellectual disability falling outside the 95% confidence intervals of risk for the ASD group. The relatively large 95% confidence intervals of estimated risk reflect the relatively small sample sizes for children with intellectual disability and children with ASD. Increasing rates of exposure to multiple risks was associated with marked increases in persistence of conduct difficulties for both typically developing children (Spearman’s r with total risk scale = 0.283, p < 0.001) and children with intellectual disability (Spearman’s r = 0.378, p < 0.001). A much weaker association was apparent for children with ASD (Spearman’s r = 0.076, p > 0.05). The strength of association for children with ASD was significantly weaker than for children with intellectual disability (z = 2.19, p < 0.05) and showed a trend to be weaker than for typically developing children (z = 1.94, p = 0.052).

Table 3 Unadjusted associations (odds ratios with 95% confidence intervals) between environmental risk factors and the persistence of conduct difficulties to ages five and seven. Risk factor

Typically developing children (n = 3825)

Children with ID but not ASD (n = 122)

Children with ASD (n = 92)

High household poverty/neighborhood deprivation Low maternal education Single parent status Low maternal wellbeing Inconsistent parenting Harsh parenting

2.32*** 2.60*** 2.07*** 2.47*** 2.51*** 2.38***

2.45* (1.10–5.43) 2.33* (1.13–4.84) 1.46 (0.69–3.08) 5.56*** (2.51–12.29) 1.28 (0.50–3.27) 2.37* (1.08–5.18)

1.54 (0.61–3.85) 2.76 (0.93–8.22) 4.70* (1.01–21.97) 1.04 (0.42–2.54) 3.12 (0.22–44.13) 0.98 (0.37–2.60)

* p < 0.05. ** p < 0.01. *** p < 0.001.

(1.99–2.71) (2.22–3.03) (1.72–2.49) (2.12–2.89) (1.87–3.37) (2.04–2.78)

3514

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

In the previous section we used multivariate logistic regression to estimate the extent to which between-group differences in prevalence of conduct difficulties may have been attributable to between-group differences in levels of exposure to environmental risk. However, the presence of marked between-group differences in the strength of association between total risk exposure and persistence (see above) indicates it would be inappropriate to use regression methods which assume a constant level of association across subgroups. Instead, the data were stratified and between-group comparisons were made for different levels of exposure to environmental risk. At all levels of risk, children with ASD were significantly more likely to show persistent conduct difficulties than typically developing children (low risk OR = 9.13 (6.82– 12.21), p < 0.001; medium risk OR = 5.69 (4.25–7.61), p < 0.001; high risk OR = 4.25 (2.73–6.61), p < 0.001) and children with intellectual disability without ASD (low risk OR = 6.31 (3.76–10.58), p < 0.001; medium risk OR = 2.20 (1.51–3.20), p < 0.001; high risk OR = 1.87 (1.14–3.06), p < 0.05). Children with intellectual disability were significantly more likely to show persistent conduct difficulties than typically developing children only at levels of medium or high exposure to environmental risk (low risk OR = 1.45 (0.94–2.23), p = n.s.; medium risk OR = 2.59 (2.02–3.32), p < 0.001; high risk OR = 2.27 (1.79–2.89), p < 0.001). 4. Discussion The results of our study indicate that: (1) exposure to environmental risk in the early years (age nine months and three years) is associated with increased risk of showing conduct difficulties at age three and, for typically developing children and children with intellectual disability, increased risk of conduct difficulties persisting from age three to ages five and seven; (2) adjusting for the likely effects of increased exposure to risk among children with intellectual disability and ASD eliminated the elevated prevalence of conduct difficulties among children with intellectual disability and reduced the elevated prevalence of conduct difficulty among children with ASD; (3) at all levels of risk, children with ASD were significantly more likely to show persistent conduct difficulties than typically developing children and children with intellectual disability; (4) children with intellectual disability were significantly more likely to show persistent conduct difficulties than typically developing children only at levels of medium or high exposure to environmental risk. These results add to the existing literature in four important ways. First, they add to the sparse literature on the extent of persistence of conduct difficulties in young children with intellectual disability or ASD by indicating that these behaviors can be highly persistent from age three across middle childhood and (especially for children with ASD) and are more persistent than similar behaviors shown by typically developing children. Second, they suggest that presence of ASD moderates the influence of exposure to environmental adversities in early life on the subsequent persistence of conduct difficulties. That is, while there is a moderately strong relationship between exposure to environmental adversities and persistence among typically developing children and children with primarily mild intellectual disability who do not have ASD, no such association is apparent for children with ASD. Gray and colleagues have also recently reported a lack of association between exposure to environmental adversity and the persistence of emotional and behavioral disorders (Gray et al., 2012). Our study adds significantly to this literature by: (1) including comparison groups of typically developing children and children with intellectual disability; and (2) using a much broader range of indictors of exposure to adversity. The reasons for the relatively weaker association between exposure to environmental adversity and the persistence of conduct difficulties among children with ASD are unknown. One possibility may be related to the neurobiology of ASD compared to children with predominantly mild intellectual disability. It is now recognized that ASD is associated with a range of genetic abnormalities which give rise to disturbances of neuronal migration, growth, neurotransmission and cell regulation (Parellada et al., 2014); abnormalities which give rise to the characteristic behavioral features of ASD. This contrasts significantly with the (predominantly mild) intellectual disability and typically developing groups. Although there are some disorders where brain pathology manifests as mild intellectual disability (e.g., fetal alcohol spectrum disorders, Prader Willi syndrome) the majority of children with mild intellectual disability have brain function similar to that of typically developing children (Einfeld & Emerson, 2008). Thus, autism and accompanying behavior disturbances may be less affected by environmental factors because some of the variance is accounted for by the effects of congenital brain pathology. It should be noted, however, that children with ASD exhibited a very similar strength of overall association between exposure to environmental adversity and the initial prevalence of conduct difficulties at age three. There were, however, some potentially important between-group differences in associations with particular indicators. For example, the main risk factors for persistence in children with ASD were single parenting status and inconsistent parenting. These are also potential indicators of an interpersonally unpredictable environment, a form of environmental adversity that may be particularly important for children with ASD. Third, our results add to the limited evidence on factors associated with the prevalence of conduct difficulties among young children with intellectual disability or ASD (Einfeld et al., 2011), drawing attention to the importance of exposure to a wide range of environmental adversities in early childhood. While there is robust evidence of elevated prevalence of emotional and behavioral disorder in children with intellectual disability (Einfeld et al., 2011), our results suggest that this may reflect the increased risk of exposure to environmental adversities among this group (rather than their intellectual disability per se) (Emerson & Hatton, 2007a, 2007b). Finally, our results suggest that the poorer emotional/behavioral health of children with intellectual disability who do not have ASD may also be accounted for by their decreased resilience when exposed to risk. That is, differential risk of persistence

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3515

between children with intellectual disability and typically developing children was related to risk exposure, with no difference in risk of persistence when children were not exposed to adversity. These results are consistent with trends in a similar direction reported in cross-sectional studies of prevalence (Emerson & Hatton, 2007a, 2007b). 4.1. Strengths and limitations The three main strengths of the present study are: (1) the use of a population-based sample of children with and without intellectual disability and ASD; (2) the use of a longitudinal design; (3) the measurement of multiple indicators of environmental adversity (Emerson, 2013). However, as in all studies, there were limitations that impact the interpretation of these findings. First, while having access to a large, longitudinal dataset is an asset, datasets (such as the MCS) that are designed for multiple purposes commonly utilize abbreviated forms of measures or constructs. For example, using maternal responses to two questions about ASD is clearly not equivalent to an empirically validated diagnosis. However, similar approaches to assessment have been used in other large-scale studies (Liptak et al., 2008). Liptak and colleagues, for example, used data from the National Survey of Children’s Health in their large-scale study of access to health services for children with ASD in the U.S. In this study, ASD diagnosis was determined from parental response to one question, ‘Has a doctor or health professional every told you that your child has autism?’ Interestingly, Liptak et al. (2008) found that Latinos and poor families rated their children’s autism as more severe than Whites or parents who were higher socioeconomically. Not surprisingly, being poor was associated with decreased access to services. Similarly, the MCS used abbreviated scales of cognitive functioning rather than complete IQ tests. Second, while the overall sample was relatively large, it was of insufficient size to examine the extent to which our results generalized to children with severe intellectual disability or children with intellectual disability and co-occurring ASD. We estimate, for example, that the available sample of 11,776 children at age three would have contained less than 40 children with severe or profound intellectual disability (Maulik et al., 2011). Given our focus was on the persistence understanding factors influencing the experience of a subset of children (those who exhibited conduct difficulties at age three), the sample was far too underpowered to allow for such sub-group analyses. It is important, therefore, to keep in mind that our results regarding intellectual disability primarily relate to less severe intellectual disability. A further sample-size issue arose for the results reported in Table 3, where the relatively wide confidence intervals mirrored the relatively small sample sizes for children with intellectual disability (n = 122) and children with ASD (n = 92). Third, there are clear limitations associated with sole reliance on self-report data, especially when applied to factors that may be subject to significant social desirability biases (e.g., parenting practices). Unfortunately, the MCS does not contain observational measures of parenting that could be used to evaluate the validity of self-report data. Finally, the non-normal distribution of many measures when combined with the relatively small sample size of the ASD group led to us collapsing some measures of risk exposure into simple binary indicators using a median split. While this is relatively common practice, other approaches to abbreviating the scales may have resulted in differences in results. 4.2. Future research Future research needs to build on the strengths and address the limitations of the present study. Particularly important is the need for future research to: (1) employ robust indicators of exposure to a variety of environmental adversities (Emerson, 2013; Emerson & Brigham, in press; Emerson et al., 2006); (2) test the generalization of these results to key sub-groups including children with severe intellectual disability and children with co-occurring intellectual disability and ASD; (3) examine aspects of child and family resilience and vulnerability in relation to exposure to adversity; (4) use observational as well as self-report measures of family functioning and parenting. It is important to note that our study did not address factors that can mitigate or off-set environmental risks and their effects on child outcome. For example, recent research has drawn attention to the importance of such factors as parental education, wealth, optimism, social support and health as protective factors among parents of children with intellectual disability when faced with adversity in early childhood (Azad, Blacher, & Marcoulides, in press; Ellingsen, Baker, Blacher, & Crnic, 2014; McConnell, Savage, & Breitkreuz, 2014). Finally, while our analyses cannot attribute causal relationships between exposure to adversity and the prevalence and persistence of conduct difficulties, they do provide some circumstantial evidence to support the case (especially for children with intellectual disability) for increased investment in the development, evaluation and scaling up of environmentally focused early intervention strategies to reduce the incidence and persistence of conduct difficulties (Allen et al., 2013; Einfeld, Tonge, & Clarke, 2013; Emerson & Einfeld, 2011; Tellegen & Sanders, 2013).

References Allen, D., Langthorne, P., Tonge, B. J., Emerson, E., McGill, P., & Fletcher, R. (2013). Toward the prevention of behavioural and emotional difficulties in people with intellectual disabilities. Journal of Applied Research in Intellectual Disabilities, 26, 501–514. Azad, G., Blacher, J., & Marcoulides, G. M. (2014). Longitudinal models of socio-economic status: Impact on positive parenting behaviors. International Journal of Behavioral Development (in press).

3516

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

Baker, B. L., McIntyre, L. L., Blacher, J., Crnic, K., Edelbrock, C., & Low, C. (2003). Preschool children with and without developmental delay: Behavior problems and parenting stress over time. Journal of Intellectual Disability Research, 47, 217–230. Baker, B. L., Neece, C. L., Fenning, R. M., Crnic, K. A., & Blacher, J. (2010). Mental disorders in five-year-old children with or without developmental delay: Focus on ADHD. Journal of Clinical Child & Adolescent Psychology, 39, 492–505. Baker, J. K., Seltzer, M. M., & Greenberg, J. S. (2011). Longitudinal effects of adaptability on behavior problems and maternal depression in families of adolescents with autism. Journal of Family Psychology, 25, 601–609. Boisjoli, R., Vitaro, F., Lacourse, E., Barker, E. D., & Tremblay, R. E. (2007). Impact and clinical significance of a preventative intervention for disruptive boys: 15 year follow up. British Journal of Psychiatry, 191, 415–419. Bracken, B. A. (2002). Bracken school readiness assessment. London: Harcourt Assessment. Broidy, L. M., Nagin, D. S., Tremblay, R. E., Bates, J. E., Brame, B., & Dodge, K. (2003). Developmental trajectories of childhood disruptive behaviours and adolescent delinquency: A six site, cross national study. Developmental Psychology, 39, 222–245. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Ferguson, L. L., & Garie´py, J. L. (1989). Growth and aggression, 1: Childhood to early adolescence. Developmental Psychology, 25, 320–330. Campbell, S. B., Shaw, D. S., & Gilliom, M. (2000). Early externalizing behavior problems: Toddlers and preschoolers at risk for later maladjustment. Development and Psychopathology, 12, 467–488. Conduct Problems Prevention Research Group (2011). The effects of the Fast Track preventive intervention on the development of conduct disorder across childhood. Child Development, 82, 331–345. Conger, R. D., & Donnellan, M. B. (2007). An interactionist perspective on the socioeconomic context of human development. Annual Review of Psychology, 58, 175– 199. Conger, R. D., Conger, K. J., Elder, G. H., Lorenz, F. C., Simons, R. L., & Whitbeck, L. B. (1992). A family process model of economic hardship and adjustment of early adolescent boys. Child Development, 3, 526–554. Department for Education (2013). GCSE and equivalent attainment by pupil characteristics in England, 2012/13. London: Department for Education. Einfeld, S., & Emerson, E. (2008). Intellectual disability. In M. Rutter, D. Bishop, D. Pine, S. Scott, J. Stevenson, E. Taylor, & A. Thapar (Eds.), Rutter’s child and adolescent psychiatry (5th ed.). Oxford: Blackwell. Einfeld, S., Piccinin, A., Mackinnon, A., Hofer, S., Taffe, J., & Gray, K. (2006). Psychopathology in young people with intellectual disability. Journal of the American Medical Association, 296(16), 1981–1989. Einfeld, S., Ellis, L., & Emerson, E. (2011). Comorbidity of intellectual disability and mental disorder: A systematic review. Journal of Intellectual & Developmental Disability, 36, 137–143. Einfeld, S. L., Tonge, B. J., & Clarke, K. S. (2013). Prevention and early intervention for behaviour problems in children with developmental disabilities. Current Opinion in Psychiatry, 26, 263–269. Eisenhower, A., Blacher, J., & Baker, B. L. (2013). Mothers’ perceived physical health during early and middle childhood: Relations with child developmental delay and behavior problems. Research in Developmental Disabilities, 34, 1059–1068. Ellingsen, R., Baker, B. L., Blacher, J., & Crnic, K. (2014). Resilient parenting of preschool children at developmental risk. Journal of Intellectual Disability Research, 58(7), 664–678. http://dx.doi.org/10.1111/jir.12063 Elliott, C., Smith, P., & McCulloch, K. (1997). British ability scales: Second edition. London: nferNelson. Emerson, E. (2005). Use of the strengths and difficulties questionnaire to assess the mental health needs of children and adolescents with intellectual disabilities. Journal of Intellectual & Developmental Disability, 30, 14–23. Emerson, E. (2013). Commentary: Childhood exposure to environmental adversity and the well-being of people with intellectual disabilities. Journal of Intellectual Disabilities Research, 57, 589–600. http://dx.doi.org/10.1111/j.1365-2788.2012.01577.x Emerson, E., & Brigham, P. (in press). Exposure of children with developmental delay to social determinants of poor health: Cross sectional study. Child: Care Health Development. http://dx.doi.org/10.1111/cch.12144 Emerson, E., & Einfeld, S. (2011). Challenging behaviour (3rd ed.). Cambridge: Cambridge University Press. Emerson, E., & Hatton, C. (2007a). The contribution of socio-economic position to the health inequalities faced by children and adolescents with intellectual disabilities in Britain. American Journal on Mental Retardation, 112, 140–150. Emerson, E., & Hatton, C. (2007b). The mental health of children and adolescents with intellectual disabilities in Britain. British Journal of Psychiatry, 191, 493–499. Emerson, E., Graham, H., & Hatton, C. (2006). The measurement of poverty and socio-economic position in research involving people with intellectual disability. In L. M. Glidden (Ed.), International review of research in mental retardation (pp. 77–108). New York: Academic Press. Emerson, E., Einfeld, S., & Stancliffe, R. J. (2011). Predictors of the persistence of conduct difficulties in children with cognitive delay. Journal of Child Psychology and Psychiatry and Allied Disciplines, 52, 1184–1194. Gerstein, E. D., Crnic, K. A., Ryu, E., Baker, B. L., & Blacher, J. (2011). Developmental risk and young children’s regulatory strategies: Predicting behavior problems at age five. Journal of Abnormal Child Psychology, 39, 351–364. Goodman, R. (1999). The extended version of the Strengths and Difficulties Questionnaire as a guide to child psychiatric caseness and consequent burden. Journal of Child Psychology and Psychiatry, 40, 791–801. Goodman, R. (2001). Psychometric properties of the Strengths and Difficulties Questionnaire (SDQ). Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345. Gray, K., Keating, C., Taffe, J., Brereton, A., Einfeld, S., & Tonge, B. J. (2012). Trajectory of behavior and emotional problems in autism. American Journal on Intellectual and Developmental Disabilities, 117, 121–133. Hansen, K. (Ed.). (2012). Millennium cohort study: First, second, third and fourth surveys. A guide to the datasets (7th ed.). London: Centre for Longitudinal Studies, Institute of Education, University of London. Hansen, K., Jones, E., Joshi, H., & Budge, D. (Eds.). (2010). Millennium cohort study fourth survey: A user’s guide to initial findings (2nd ed.). London: Centre for Longitudinal Studies, Institute of Education, University of London. Hobcraft, J. N., & Kiernan, K. E. (2010). Predictive factors from age 3 and infancy for poor child outcomes at age 5 relating to children’s development, behaviour and health: Evidence from the Millennium Cohort Study. York: University of York. Iizuka, C., Yamashita, Y., Nagamitsu, S., Yamashita, T., Araki, Y., & Ohya, T. (2010). Comparison of the strengths and difficulties questionnaire (SDQ) scores between children with high functioning autism spectrum disorder (HFASD) and attention-deficit/hyperactivity disorder (AD/HD). Brain and Development, 32, 609–612. Johnson, J. (Ed.). (2009). Millennium cohort study: Geographic identifiers in MCS. London: Centre for Longitudinal Studies, Institute of Education, University of London. Johnson, J. (Ed.). (2012). Millennium cohort study: Psychological, developmental and health inventories. London: Centre for Longitudinal Studies, Institute of Education, University of London. Jones, E., & Ketende, S. (Eds.). (2010). Millennium cohort study: User guide to analysing MCS data using SPSS. London: Centre for Longitudinal Studies, Institute of Education, University of London. Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., & Normand, S.-L.T. (2002). Short screening scales to monitor population prevalences and trends in nonspecific psychological distress. Psychological Medicine, 32, 959–976. Kessler, R. C., Barker, P. R., Colpe, L. J., Epstein, J. F., Gfroerer, J. C., & Hiripi, E. (2003). Screening for serious mental illness in the general population. Archives of General Psychiatry, 60, 184–189. Liptak, G. S., Benzoni, L. B., Mruzek, D. W., Nolan, K. W., Thingvoll, M. A., & Wade, C. (2008). Disparities in diagnosis and access to health services for children with autism: Data from the National Survey of Children’s Health. Journal of Developmental and Behavioral Pediatrics, 29, 152–160. Lochman, J. E., Wells, K. C., Qu, L., & Chen, L. (2012). Three year follow-up of coping power intervention effects: Evidence of neighborhood moderation? Prevention Science. http://dx.doi.org/10.1007/s11121-012-0295-0

E. Emerson et al. / Research in Developmental Disabilities 35 (2014) 3508–3517

3517

Maulik, P. K., Mascarenhas, M. N., Mathers, C. D., Dua, T., & Saxena, S. (2011). Prevalence of intellectual disability: A meta-analysis of population-based studies. Research in Developmental Disabilities, 32, 419–436. McConnell, D., Savage, A., & Breitkreuz, R. (2014). Resilience in families raising children with disabilities and behavior problems. Research in Developmental Disabilities, 35, 833–848. Moffitt, T. E., & Scott, S. (2008). Conduct disorders of childhood and adolescence. In M. Rutter, D. Bishop, D. Pine, S. Scott, J. Stevenson, E. Taylor, & A. Thapar (Eds.), Rutter’s child and adolescent psychiatry (pp. 543–564). Oxford: Blackwell. Nagin, D., & Tremblay, R. E. (1999). Trajectories of boys’ physical aggression, opposition, and hyperactivity on the path to physically violent and non violent juvenile delinquency. Child Development, 70, 1181–1196. Pantazis, C., Gordon, D., & Levitas, R. (Eds.). (2006). Poverty and social exclusion in Britain: The millennium survey. Bristol: Policy Press. Parellada, M., Penzol, M. J., Pina, L., Moreno, C., Gonza´lez-Vioque, E., & Zalsman, G. (2014). The neurobiology of autism spectrum disorders. European Psychiatry, 29, 11–19. Petitclerc, A., & Tremblay, R. E. (2009). Childhood disruptive behaviour disorders: Review of their origin, development, and prevention. Canadian Journal of Psychiatry, 54, 222–231. Plewis, I. (2007). The millennium cohort study: Technical report on sampling (4th ed.). London: Centre for Longitudinal Studies. Plewis, I., & Ketende, S. (2006). Millennium cohort study: Technical report on response. London: Centre for Longitudinal Studies, Institute of Education, University of London. Prinz, R. J., Sanders, M. R., Shapiro, C. J., Whitaker, D. J., & Lutzker, J. R. (2009). Population-based prevention of child maltreatment. Prevention Science, 10, 1–12. Rutter, M., Tizard, J., & Whitmore, K. (1970). Education, health and behaviour. London: Longmans. Sanders, M. R. (2008). Triple P-positive parenting program as a public health approach to strengthening parenting. Journal of Family Psychology, 22, 506–517. Tellegen, C. L., & Sanders, M. R. (2013). Stepping stones Triple P-positive parenting program for children with disability: A systematic review and meta-analysis. Research in Developmental Disabilities, 34, 1556–1571. Totsika, V., Hastings, R. P., Emerson, E., Lancaster, G. A., & Berridge, D. M. (2011). A population-based investigation of behavioural and emotional problems and maternal mental health: Associations with autism spectrum disorder and intellectual disability. Journal of Child Psychology and Psychiatry and Allied Disciplines, 52, 91–99. Totsika, V., Hastings, R. P., Emerson, E., Berridge, D. M., Lancaster, G. A., & Vagenas, D. (2013). Is there a bidirectional relationship between maternal well-being and child problem behaviors? Longitudinal evidence in autism spectrum disorders. Autism Research, 6, 201–211. Tremblay, R. E. (1999). When children’s development fails. In D. P. Keating & C. Hertzman (Eds.), Developmental health and the wealth of nations: Social, biological and educational dynamics (pp. 55–71). New York: Guilford Press. Tremblay, R. E. (2000). The development of aggressive behavior during childhood: What have we learned in the past century? International Journal of Behavioral Development, 24, 129–141. Tremblay, R. E. (2006). Prevention of youth violence: Why not start at the beginning? Journal of Abnormal Child Psychology, 34(4), 481–487. Tremblay, R. E. (2012). Environmental, genetic and epigenetic influences on the developmental origins of aggression and other disruptive behaviors. In T. Bliesener, A. Beelmann, & M. Stemmler (Eds.), Antisocial behavior and crime: Contributions of developmental and evaluation research to prevention and intervention (pp. 3–16). Cambridge, MA: Hogrefe Publishing. Tremblay, R. E., Pagani-Kurtz, L., Masse, L. C., Vitaro, F., & Pihl, R. O. (1995). A bimodal preventive intervention for disruptive kindergarten boys: Its impact through mid-adolescence. Journal of Consulting and Clinical Psychology, 63, 560–568. Tremblay, R. E., Japel, C., Perusse, D., McDuff, P., Boivin, M., & Zoccolillo, M. (1999). The search for the age of ‘onset’ of physical aggression: Rousseau and Bandura revisited. Criminal Behaviour and Mental Health, 9, 8–23. Tremblay, R. E., Nagin, D. S., Seguin, J. R., Zoccolillo, M., Zelazo, P. D., & Boivin, M. (2004). Physical aggression during early childhood: Trajectories and predictors. Pediatrics, 114(1), e43–e50. Vitaro, F., & Tremblay, R. E. (2008). Clarifying and maximising the usefulness of targeted preventative interventions. In M. Rutter, D. Bishop, D. Pine, S. Scott, J. Stevenson, E. Taylor, & A. Thapar (Eds.), Rutter’s child and adolescent psychiatry (5th ed., pp. 989–1008). Oxford: Blackwell. Vitaro, F., Barker, E. D., Brendgen, M., & Tremblay, R. E. (2012). Pathways explaining the reduction of adult criminal behaviour by a randomized preventive intervention for disruptive kindergarten children. Journal of Child Psychology and Psychiatry, 53, 749–756. Webster-Stratton, C., & Taylor, T. (2001). Nipping early risk factors in the bud: Preventing substance abuse, delinquency, and violence in adolescence through interventions targeted at young children (0 to 8 years). Prevention Science, 2(3), 165–192. Webster-Stratton, C., Reid, J., & Stoolmiller, M. (2008). Preventing conduct problems and improving school readiness: Evaluation of the Incredible Years Teacher and Child Training Programs in high-risk schools. Journal of Child Psychology and Psychiatry, 49(5), 471–488. Zubrick, S. R., Northey, K., Silburn, S. R., Lawrence, D., Williams, A. A., & Blair, E. (2005). Prevention of child behavior problems through universal implementation of a group behavioral family intervention. Prevention Science, 6, 287–304.