Abstract
Childhood head injuries and conduct problems increase the risk of aggression and criminality and are well-known correlates. However, the direction and timing of their association and the role of their demographic risk factors remain unclear. This study investigates the bidirectional links between both from 3 to 17 years while revealing common and unique demographic risks. A total of 8,603 participants (50.2% female; 83% White ethnicity) from the Millennium Cohort Study were analysed at 6 timepoints from age 3 to 17. Conduct problems were parent-reported for ages 3 to 17 using the Strengths and Difficulties Questionnaire (SDQ) and head injuries at ages 3 to 14. A cross-lagged path model estimated the longitudinal bidirectional effects between the two whilst salient demographic risks were modelled cumulatively at three ecological levels (child, mother, and household). Conduct problems at age 5 promoted head injuries between 5 and 7 (Z = 0.07; SE = 0.03; 95% CI, 0.02–0.13), and head injuries at ages 7 to 11 promoted conduct problems at age 14 (ß = .0.06; SE = .0.03; 95% CI, 0.01–0.12). Head injuries were associated with direct child-level risk at age 3, whereas conduct problems were associated with direct risks from all ecological levels until 17 years. The findings suggest a sensitive period at 5–11 years for the bidirectional relationship shared between head injuries and conduct problems. They suggest that demographic risks for increased head injuries play an earlier role than they do for conduct problems. Both findings have implications for intervention timing.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
Childhood conduct problems and head injuries are both significant risk factors for lifelong aggression and criminality [1, 2] and are known correlates [3]. However, how and when conduct problems and head injuries increase the other during childhood, particularly when controlling for demographic risk factors, remains unknown. This poses a serious problem for professionals in health, social care, and education. Without knowing when and to what extent head injuries pose a risk for conduct problems (and vice versa) it is difficult to design and deploy interventions with the greatest potential for impact.
Conduct problems can be defined as repeated violations to age-appropriate societal norms [4], such as fighting, threatening, and bullying. One of the potential causes of conduct problems is head injury [5]. Head injury is the main cause of death and disability in the UK, with approximately 1.4 million admissions of head injury every year, of which 33–55% are children [6].
Clinical studies have shown increased conduct problems following traumatic brain injuries (TBI) [5, 7]. Mild head injuries (those that do not disrupt normal brain functioning) are similarly associated with increased odds of delinquent behaviours at ages 11 and 14 [8], and with greater levels of conduct problems in adolescence and early adulthood [9]. Mechanisms explaining how head injuries pose a risk for increased conduct problems include changes to brain areas linked with executive functioning. In particular, over-activation of attention networks [10] and changes in neural connectivity resulting in task switching difficulty [11].
Research, however, investigating conduct problems influence on the risk of head injuries is limited. Studies typically investigate this relationship alongside ADHD [12] or from adolescent to adulthood [13]. However, a recent study suggests that childhood conduct problems at age 5 can similarly predict an increased risk of sustaining head injuries from ages 7 to 11 [3]. Mechanisms to explain this association similarly include changes to brain areas. For example, the ventral striatum (associated with reward processing) has been shown to be impaired in those with conduct problems and is related to greater real-life risk-taking [14]. Such risk-taking may provide a greater opportunity to sustain a head injury including through rough and tumble play, which has been shown to be more common in those with conduct problems [15].
Although the current literature suggests a potential bidirectional association between childhood conduct problems and head injury, no published study explicitly investigated this association, nor identified a sensitive age in which these associations take place. This information is critical to inform effective interventions. Limitations of many previous studies is their focus on TBIs, while 95% of head injuries are mild or never reported [6], the inclusion of clinical samples, self-reported head injuries, long delays in reporting of head injuries, and failure to control for common factors influencing both conduct problems and head injuries. We sought to account for such limitations by investigating whether there is a bidirectional association between head injuries and conduct problems during child development from 3 to 17 years in a large, longitudinal UK cohort. Importantly, the current study controls for salient demographic risk factors concerning the child, their mother, and their household, leading to two research questions:
-
1.
Are there bidirectional associations between head injuries and conduct problems from ages 3 to 17 years?
-
2.
Is combined risk at the child, mother and household levels associated with conduct problems and/or head injuries from ages 3 to 17 years?
Methods
Study design and participants
Participants were part of the Millennium Cohort study (MCS), a longitudinal birth cohort study of 18,786 individuals born in the UK between 2000 and 2002 [16]. They were studied at seven time points, at 9 months (T1), 3 (T2), 5 (T3), 7 (T4), 11 (T5), 14 (T6), and 17 years (T7). Analyses were limited to those with complete conduct problem data at the last wave (T7) [17, 18, 19]. Further exclusions were made to those who were not first-born children to allow independence of observation[20] and due to different levels of aggression related schemas and head injury risk in siblings [21, 22]. Final exclusions were made to those whose main respondent was not their biological mother as the focus of mother-related risk (see below) such as mother to child attachment were measured only for the biological mother. This resulted in an analytic sample of 8,603 individuals (4,322 female [50.2%]; 83% White ethnicity; see flow chart in Supplementary Fig. 1).
All procedures and analyses were approved by the University of Southampton Ethics Committee (ID = 62,100). Families provided written informed consent to take part and consented for their data to be shared for secondary analysis. Data were downloaded from the UK Data Archive [beta.ukdataservice.ac.uk/datacatalogue/series/series?id = 2000031].
Measures
Conduct problems
These were assessed from age 3 (T2) using the five items from the Conduct Problem Subscale of the parent-report version of the Strengths and Difficulties Questionnaire (SDQ) [23]. Items are scored on a 3-point scale (0—2) with a higher total score indicating a higher level of conduct problems (possible range: 0—10). Cronbach’s alpha values within this study ranged from 0.52 to 0.66 across the MCS waves. Previous research has shown the SDQ to have over 75% sensitivity in identifying clinically relevant conduct problems [24], the parent version specifically has strong validity in identifying conduct disorder [25], and has been shown to be invariant across timepoints [26].
Head injuries
Parents were asked if their child had ever, or since the last wave, sustained a head injury that resulted in them being taken to the doctor, health centre, or hospital. Head injuries (coded 1) included responses categorised as a ‘bang on the head’ or ‘loss of consciousness’. The ‘loss of consciousness’ group was extremely small meaning that there would not have been the statistical power to warrant analysing the groups separately. The overall ‘head injury’ variables also capture everyday head injuries sustained in the general population as opposed to the moderate-severe head injuries that are often the focus of the literature. Head injury data was analysed from T2 onwards to achieve temporal ordering with the studies risk factors.
Demographic risks
Demographic risks were divided by ecological level (child, mother, and household) and combined risk from each level was measured via a cumulative risk index (CRI). Each CRI consisted of five items dichotomised into 0’s (low risk) and 1’s (high risk) based on the literature and summated. A higher score indicated the presence of more risks in a child’s development. Further details of each CRI can be seen in Supplementary Table 1.
Child level risk
Child level risk factors were taken from the parent interview at T1 and included male sex [2, 27], low birth weight (< 2.5 kg) and premature birth (< = 252 days gestation) [28, 29], and whether the child’s biological mother smoked or drank alcohol during pregnancy [30].
Mother level risk
Mother level risk factors were from the parent interview at T1 and included pregnancy before 18 years [27, 31], no high-school qualification [30, 31, 32], current unemployment [30], low attachment with child (< = 22 on Condon Maternal Attachment Scale) [33, 34], and psychological distress (> 4 on Rutter Malaise Inventory) [27, 35].
Attachment with child was measured using a subset of six items from the Condon Maternal Attachment Questionnaire [33]. The items were scored on a scale from 1 (almost all the time) to 5 (never; possible range: 0—30). A lower score indicates greater difficulties in mother–child attachment. Maternal psychological distress was measured using the MCS’s 9-item composite variable of the Rutter Malaise Inventory’s original 24-item scale [35]. The items were coded as 0 (no) and 1 (yes) and summed (possible range: 0 – 9) with a higher score indicating higher psychological distress.
Household level risk
Household level risk factors were taken from the parent interview at T1 and T2. These included single parent household [31, 36], low household income (< 60% of median household income) [31, 36], household overcrowding (fewer rooms than people excluding bathrooms and hallways) [31, 36], low household occupational status (highest occupational status in the household being semi-skilled or lower) [32], and a low-quality home learning environment (bottom quartile of early home learning environment scale) [37]. The home learning environment was measured at T2 using six items used in the home learning environment scale available in the MCS dataset (excluding ‘playing with numbers’) [37]. These measured the frequency at which the child engaged in learning activities. These items were scored on a rating-scale from 0 (not at all) to 7 (everyday) and summed (possible range 0—42). A higher score indicates a higher quality home learning environment.
Covariates
ADHD
ADHD was controlled for due to its high comorbidity with conduct problems [38] and its high association with sustaining a head injury [12]. ADHD was measured from age 5 (T3) to age 14 (T6) by asking the parent if their child had a diagnosis of ADHD. A binary variable was generated (0 = no diagnosis, 1 = diagnosis of ADHD).
Epilepsy
Epilepsy was controlled for due to its increased association with sustaining a head injury [39]. Epilepsy was measures from age 3 (T2) to age 17 (T7) by asking the parent if their child had a diagnosis of epilepsy. A binary variable was generated (0 = no diagnosis, 1 = diagnosis of epilepsy).
Statistical analysis
Mplus (v7.4) was used to run a cross-lagged path model (see Fig. 1) to test the relationships between head injury and conduct problems over time while controlling for salient demographic risks, ADHD, and epilepsy. MCS sample weights from T7 were applied to account for stratification, attrition, and nonresponse bias. The internal validity of the statistical estimates concerning the binary head injury variables were improved through use of the weighted least square estimation procedure. Missing data were accounted for through the use of the Full Information Maximum Likelihood procedure.
Stylized Illustration of the Structural Equation Model Implemented in This Study. This figure shows the cross-lagged path model conducted on conduct problem variables from age 3 (T2) to 17 (T7) and head injury variables from age 3 (T2) to 14 (T6). These are connected by contemporaneous correlations as well as lagged paths to T + 1 within and across variables. The three cumulative risk indices (CRI) at the child, mother, and household-levels are connected to each head injury and conduct problem variable (dotted lines). Solid lines represent pathways between conduct problem and head injury variables. Dashed lines represent correlations within timepoints
Contemporaneous correlations were included to account for the relationship within-timepoints [40]. As the correlations were between a binary and continuous variable, Mplus calculated point-biserial (rpbis) correlations.
Total, direct, and indirect effects were modelled and reported (see Primer for further information on these effects [41]). Indirect effects (e.g., the indirect effect of T2 head injuries on T4 conduct problems via T3 head injuries) were reported as total indirect effects (sum of all indirect effects). However, where a total indirect was not significant but an individual indirect effect was, the individual indirect effect was reported.
Model fit was evaluated based on the following criteria: Tucker-Lewis Index (TLI; acceptable fit ≥ 0.90, good fit ≥ 0.95), the Comparative Fit Index (CFI; acceptable fit ≥ 0.90, good fit ≥ 0.95), and Root Mean Square Error of Approximation (RMSEA; acceptable fit < 0.08, good fit < 0.05) [42, 43]. Where conduct problems (continuous) were the dependent variable, standardised beta values (ß) were reported. Where head injury (binary) was the dependent variable, the standardised Z-value (index of probit regression) was reported. Results were considered significant with α = 0.05.
Data availability
The MCS dataset used in this study is available via the UK Data Service. The Mplus output for the direct and indirect effects as well as the code needed to create the CRI variables can be accessed via Pure.
Results
Participants and demographics
Table 1 provides a summary and comparison of sample characteristics between the excluded and analytical samples. The samples differed significantly on all variables, though these effects were weak (Cramér’s V < 0.20, Cohen’s d < 0.20). A breakdown of the head injury variable can be seen in Supplementary Table 2.
Association between head injury and conduct problems across development
The cross-lagged path model showed acceptable fit (χ2(32) = 468.34; p < 0.001; RMSEA = 0.02 [0.018, 0.022]; CFI = 0.93; TLI = 0.84) with all except the TLI meeting the predefined acceptable threshold [42, 43].
The contemporaneous correlations between head injury and conduct problems were small, positive, (rpbis < 0.10) and significant (p < 0.05) at age 3 (T2) and 17 (T7).
Head injury at each time point had significant direct (Fig. 2, Supplementary Table 3) and indirect effects (Fig. 2, Supplementary Table 4) for an increased likelihood of subsequent head injury, as did conduct problems for increased subsequent conduct problems (Fig. 2, Supplementary Tables 3 and 4).
The Direct and Indirect Effects Within and Between Conduct Problems and Head Injury From Ages 3 to 17. This figure shows the significant direct effects (solid lines) within and between the head injury and conduct problem variables, the significant total indirect (dashed lines), and the individual indirect (dotted lines) effects. All indirect effects from head injury to later head injury variables (T + 1 onwards) and from conduct problems to later conduct problem variables were significant but omitted for clarity. Only significant pathways are shown to p < 0.05 (*) and p < 0.001 (**)
Head injury at ages 7 to 11 had a direct effect for increased conduct problems at age 14 (ß = 0.06; SE = 0.03; 95% CI 0.01–0.12). Head injuries at ages 3 to 5 and 5 to 7 had significant individual indirect effects linked to greater conduct problems at age 14 (ß = 0.01; SE = 0.004; 95% CI 0.001–0.02; ß = 0.02; SE = 0.01; 95% CI 0.002–0.05 respectively). Head injuries at ages 5 to 7 and 7 to 11 had significant individual indirect effects linked to greater conduct problems at age 17 (ß = 0.02; SE = 0.01; 95% CI 0.002–0.03; ß = 0.04; SE = 0.02; 95% CI 0.01–0.08, respectively). See Fig. 2 for visualisation.
Conduct problems at age 5 had a direct effect for an increased likelihood of head injury between ages 5 and 7 (Z = 0.07; SE = 0.03; 95% CI 0.02–0.13). There were significant total indirect effects from conduct problems at age 3 for an increased likelihood of head injuries at ages 5 to 7 (Z = 0.05; SE = 0.02; 95% CI 0.01–0.08). Significant individual indirect effects were identified from conduct problems at ages 3 and 5 for an increased likelihood of head injuries at ages 7 to 11 (Z = 0.02; SE = 0.01; 95% CI 0.003–0.03; Z = 0.03; SE = 0.01; 95% CI 0.01–0.05, respectively) and of head injuries at ages 11 to 14 (Z = 0.01; SE = 0.002; 95% CI 0.001–0.01; Z = 0.01; SE = .0.004; 95% CI 0.001–0.02, respectively)). See Fig. 2 for visualisation.
The influence of child, mother and household-level demographic risk factors
Child-level cumulative risk had a significant direct effect for increased conduct problems at ages 3, 5, 11, and 17 (Table 2, Supplementary Fig. 2). Mother-level cumulative risk had a significant direct effect for increased conduct problems at ages 3, 5, 7, and 17 (Table 2, Supplementary Fig. 3). Household-level cumulative risk had a significant direct effect for increased conduct problems at ages 3, 11, and 17 (Table 2, Supplementary Fig. 4). All three CRIs had significant total indirect effects for increased conduct problems from ages 5 to 14 (Table 2, Supplementary Figs. 2, 3, and 4) and significant individual indirect effects for age 17 (Child: ß = 0.07, SE = 0.02, 95% CI 0.04 –0.10; Mother: ß = 0.02, SE = 0.004, 95% CI 0.02–0.03; Household: ß = 0.05, SE = 0.01, 95% CI 0.03–0.08).
Only the child-level cumulative risk had a significant direct effect for an increased likelihood of head injuries from 9 months to 3 years (see Table 2, Supplementary Fig. 2). Total indirect effects were significant only at the child-level for head injuries sustained at ages 3 to 5, 5 to 7, and 11 to 14 (Table 2, Supplementary Fig. 2). However, significant individual indirect effects were present from the mother and household-levels to head injuries sustained at ages 5 to 7 (Z = 0.01, SE = 0.002; 95% CI = 0.001–0.01; Z = 0.01, SE = 0.004; 95% CI = 0.002–0.02, respectively)), and the household-level for ages 7 to 11 (Z = 0.004, SE = 0.002; 95% CI = 0.001–0.01).
Discussion
The aim of this study was to identify if there were bidirectional associations between conduct problems and head injuries in a UK population between the ages of 3 and 17 years, while controlling for salient demographic risk factors. The results showed that higher levels of conduct problems at age 5 promoted an increased likelihood of head injury between the ages of 5 to 7 whilst a head injury sustained between the ages of 7 and 11 promoted increased conduct problems at age 14. Thus, this study shows a longitudinal, bidirectional relationship between head injuries and conduct problems during a sensitive period between the ages of 5 and 11 years. Further, the bidirectional relationship between head injury and conduct problems exists over and above the effects of salient demographic risk factors at the child, mother, and household-level as well as ADHD and epilepsy.
These results provide further evidence that childhood head injuries are associated with increased levels of conduct problems [2]. However, it elaborates on the previous literature by suggesting that this relationship is bidirectional and that conduct problems also promote head injuries during the sensitive period of 5 to 11 years. This was only previously identified when there was a co-morbid diagnosis of ADHD [12] or in a young adult population [13]. This clarifies results shown by Brandt and colleagues [3] whilst controlling for salient demographic risk. Thus, the current study provides novel insight into a potential bidirectional association between head injury and conduct problems that warrants further investigation.
In line with existing literature, child, mother, and household demographic risks all had direct and indirect effects for increased conduct problems over the course of development (from 3 to 17 years) [27, 30, 31]. However and surprisingly, the mother and household risks were found to play no direct role in promoting head injuries during childhood (from age 9 months). Direct risk instead lied solely at the level of the child with all but one of these risk factors (male sex) being themselves socially stratified.
Strengths and limitations
The key strength of the current study is its use of a large birth cohort dataset which enabled the statistical unpacking of the complex relationships linking conduct problems to head injuries and vice versa over time. Another strength is the comprehensive inclusion of all head injuries, which increases the ecological validity to the findings.
A limitation of this paper is the stringent inclusion criteria for participants, which limited the generalizability of the findings to the general UK population. The analytical sample differed from the total sample on demographics including ethnicity. Therefore, the results may not reflect the ethnic diversity within the UK population meaning that these results must be read with caution. The analytical sample also had significantly lower mean conduct problems than the total sample, suggesting that the sample may not be representative of conduct problems presented in the general UK population.
Parent-report for both head injuries and conduct problems might be considered a limitation. Though this addresses the limitations of previous head injury research whereby self-report is likely to inhibit accuracy (i.e. due to infantile amnesia) [44], it could introduce a social desirability bias. Therefore, this research (as with all research using parent measures) requires smaller-scale follow-up using more objective measures, such as clinical records.
Implications
Parents and teachers may work together to identify those children with high levels of conduct problems when they enter primary school as these children are at an increased risk for sustaining a head injury. This is particularly important as this is a critical developmental period where children enter school and begin to have reduced parental supervision and increased peer interaction. All of which could result in greater opportunities to sustain a head injury. Additional safety precautions may be administered in schools to try to counteract the increased risk for head injuries, which poses a subsequent risk for an increase in conduct problems until age 14.
Examples include limiting or prohibiting contact sports where there is the potential to sustain a head injury [45] and interventions to encourage helmet usage when riding a bike [46].
Conclusions
The results of this study suggest a sensitive period between the ages of 5 and 11 where conduct problems and head injuries are risk factors for one another with consequences for interventions that run both before and during this period.
References
Erskine HE, Norman RE, Ferrari AJ, Chan GC, Copeland WE, Whiteford HA et al (2016) Long-term outcomes of attention-deficit/hyperactivity disorder and conduct disorder: a systematic review and meta-analysis. J Am Acad Child Adolesc Psychiatry 55(10):841–850
Fullerton AF, Jackson NJ, Tuvblad C, Raine A, Baker LA (2019) Early childhood head injury attenuates declines in impulsivity and aggression across adolescent development in twins. Neuropsychology 33(8):1035–1044. https://doi.org/10.1037/neu0000570
Brandt V, Hall C, Eisenbarth H, Hall J (2022) Mutual relationships between head injury and conduct problems in children aged 9 months to 14 years in the UK Millennium Cohort Study. Medrxiv. 41:22268970. https://doi.org/10.1101/2022.01.09.22268970
Girard L-C, Tremblay RE, Nagin D, Côté SM (2019) Development of aggression subtypes from childhood to adolescence: a group-based multi-trajectory modelling perspective. J Abnorm Child Psychol 47(5):825–838
Pastore V, Galbiati S, Recla M, Colombo K, Beretta E, Strazzer S (2018) Psychological and behavioural difficulties following severe TBI in adolescence: a comparison with a sample of peers with brain lesions of other origin and with a control group. Brain Inj 32(8):1011–1020. https://doi.org/10.1080/02699052.2018.1469041
Excellence NIfHaC. 2014. Head injury: assessment and early management. https://www.nice.org.uk/guidance/cg176. Accessed 28 Feb 2022.
Bellesi G, Barker ED, Brown L, Valmaggia L (2019) Pediatric traumatic brain injury and antisocial behavior: are they linked? A systematic review Brain Injury 33(10):1272–1292. https://doi.org/10.1080/02699052.2019.1641621
Mongilio J (2022) Childhood head injury as an acquired neuropsychological risk factor for adolescent delinquency. J Res Crime Delinq 59(6):756–790. https://doi.org/10.1177/00224278221081140
Buckley L, Chapman RL (2017) Associations between self-reported concussion with later violence injury among Australian early adolescents. J Public Health (Oxf) 39(1):52–57. https://doi.org/10.1093/pubmed/fdw009
Kramer ME, Chiu CY, Walz NC, Holland SK, Yuan W, Karunanayaka P et al (2008) Long-term neural processing of attention following early childhood traumatic brain injury: fMRI and neurobehavioral outcomes. J Int Neuropsychol Soc 14(3):424–435. https://doi.org/10.1017/S1355617708080545
Caeyenberghs K, Leemans A, Heitger MH, Leunissen I, Dhollander T, Sunaert S et al (2012) Graph analysis of functional brain networks for cognitive control of action in traumatic brain injury. Brain 135(4):1293–1307. https://doi.org/10.1093/brain/aws048
Ramos Olazagasti MA, Klein RG, Mannuzza S, Belsky ER, Hutchison JA, Lashua-Shriftman EC et al (2013) Does childhood attention-deficit/hyperactivity disorder predict risk-taking and medical illnesses in adulthood? J Am Acad Child Adolesc Psychiatry 52(2):153. https://doi.org/10.1016/j.jaac.2012.11.012
Vassallo JL, Proctor-Weber Z, Lebowitz BK, Curtiss G, Vanderploeg RD (2007) Psychiatric risk factors for traumatic brain injury. Brain Inj 21(6):567–573. https://doi.org/10.1080/02699050701426832
van Hoorn J, McCormick EM, Perino MT, Rogers CR, Telzer EH (2020) Differential behavioral and neural profiles in youth with conduct problems during risky decision-making. J Res Adolesc 30(3):599–615. https://doi.org/10.1111/jora.12546
Garcia M, Aubron V, Salla J, Hanne-Poujade S, Teymoori A, Michel G et al (2020) Association between adolescent rough-and-tumble play and conduct problems, substance use, and risk-taking behaviors: Findings from a school-based sample. Aggressive Behav 46(1):37–48. https://doi.org/10.1002/ab.21866
Fitzsimons E, Haselden L, Smith K, Gilbert E, Calderwood L, Agalioti- Sgompou V, et al. (UCL Centre for Longitudinal Studies, London, 2020)
Siddique J, Brown CH, Hedeker D, Duan N, Gibbons RD, Miranda J et al (2008) Missing data in longitudinal trials–part B, analytic issues. Psychiatr Ann 38(12):201
Speyer LG, Hang Y, Hall HA, Murray AL (2022) The role of harsh parenting practices in early- to middle-childhood socioemotional development: An examination in the Millennium Cohort Study. Child Dev. https://doi.org/10.1111/cdev.13761
Cheng H, Furnham A (2020) Correlates of maternal Emotional Stability: Findings from the Millennium Cohort Study. Person Ind Differ. 164:110119. https://doi.org/10.1016/j.paid.2020.110119
Grawitch MJ, Munz DC (2004) Are your data nonindependent? A practical guide to evaluating nonindependence and within-group agreement. Underst Stat 3(4):231–257
Ardebili EF, Golshani F (2016) Early maladaptive schemas and aggression based on the birth order of children. Mod Appl Sci 10(9):14–21
Honda C, Yamana H, Matsui H, Nagata S, Yasunaga H, Naruse T (2020) Age in months and birth order in infant nonfatal injuries: A retrospective cohort study. Public Health Pract. 1:100005. https://doi.org/10.1016/j.puhip.2020.100005
Goodman R (2001) Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry 40(11):1337–1345. https://doi.org/10.1097/00004583-200111000-00015
Goodman R, Ford T, Simmons H, Gatward R, Meltzer H (2000) Using the Strengths and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a community sample. Br J Psychiatry 177(6):534–539. https://doi.org/10.1192/bjp.177.6.534
Becker A, Woerner W, Hasselhorn M, Banaschewski T, Rothenberger A (2004) Validation of the parent and teacher SDQ in a clinical sample. Euro Child Adol Psych 13(2):11–16. https://doi.org/10.1007/s00787-004-2003-5
Sosu EM, Schmidt P (2017) Tracking emotional and behavioral changes in childhood: does the strength and difficulties questionnaire measure the same constructs across time? J Psychoeduc Assess 35(7):643–656. https://doi.org/10.1177/0734282916655503
McKinlay A, Kyonka EG, Grace RC, Horwood LJ, Fergusson DM, MacFarlane MR (2010) An investigation of the pre-injury risk factors associated with children who experience traumatic brain injury. Inj Prev 16(1):31–35. https://doi.org/10.1136/ip.2009.022483
Reijneveld SA, de Kleine MJ, van Baar AL, Kollée LA, Verhaak CM, Verhulst FC et al (2006) Behavioural and emotional problems in very preterm and very low birthweight infants at age 5 years. Arch Dis Child Fetal Neonatal Ed 91(6):F423–F428. https://doi.org/10.1136/adc.2006.093674
Whiteside-Mansell L, Bradley RH, Casey PH, Fussell JJ, Conners-Burrow NA (2009) Triple risk: do difficult temperament and family conflict increase the likelihood of behavioral maladjustment in children born low birth weight and preterm? J Pediatr Psychol 34(4):396–405. https://doi.org/10.1093/jpepsy/jsn089
Van Adrichem DS, Huijbregts SCJ, Van Der Heijden KB, Van Goozen SHM, Swaab H (2020) Aggressive behavior during toddlerhood: Interrelated effects of prenatal risk factors, negative affect, and cognition. Child Neuropsychol 26(7):982–1004. https://doi.org/10.1080/09297049.2020.1769582
Trentacosta CJ, Hyde LW, Shaw DS, Dishion TJ, Gardner F, Wilson M (2008) The relations among cumulative risk, parenting, and behavior problems during early childhood. J Child Psychol Psychiatry 49(11):1211–1219. https://doi.org/10.1111/j.1469-7610.2008.01941.x
Greitemeyer T, Sagioglou C (2016) Subjective socioeconomic status causes aggression: A test of the theory of social deprivation. J Pers Soc Psychol 111(2):178–194. https://doi.org/10.1037/pspi0000058
Condon JT, Corkindale CJ (1998) The assessment of parent-to-infant attachment: development of a self-report questionnaire instrument. J Reprod Infant Psychol 16(1):57–76
Curran EA, Cryan JF, Kenny LC, Dinan TG, Kearney PM, Khashan AS (2016) Obstetrical mode of delivery and childhood behavior and psychological development in a British cohort. J Autism Dev Disord 46(2):603–614
Rutter M, Tizard J, Whitmore K (1970) Education, health and behaviour: psychological and medical study of childhood development. Longman Group Limited, Newyork
Northerner LM, Trentacosta CJ, McLear CM (2016) Negative affectivity moderates associations between cumulative risk and at-risk toddlers’ behavior problems. J Child Fam Stud 25(2):691–699
Sylva K, Melhuish E, Sammons P, Siraj-Blatchford I (2004) Taggart B. University of London/Department for Education and Skills, Institute of Education
Gnanavel S, Sharma P, Kaushal P, Hussain S (2019) Attention deficit hyperactivity disorder and comorbidity: A review of literature. World J Clin Cases. 7(17):2420–2426. https://doi.org/10.12998/wjcc.v7.i17.2420
Annegers JF, Coan SP (2000) The risks of epilepsy after traumatic brain injury. Seizure 9(7):453–457
Chen J, Liu L, Chen J, Ng MS, Lou VW, Wu B et al (2021) The cross-lagged association between depressive symptoms and health-related quality of life in patients receiving maintenance hemodialysis: a three-wave longitudinal study. Qual Life Res 30:1–11
Buhi ER, Goodson P, Neilands TB (2007) Structural Equation Modeling: A Primer for Health Behavior Researchers. Am J Health Behav 31(1):74–85. https://doi.org/10.5993/AJHB.31.1.8
Lt Hu, Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Mod 6(1):1–55
Marsh HW, Hau K-T, Wen Z (2004) In Search of Golden Rules: Comment on Hypothesis-Testing Approaches to Setting Cutoff Values for Fit Indexes and Dangers in Overgeneralizing Hu and Bentler’s (1999) Findings. Struct Equ Mod 11(3):320–341. https://doi.org/10.1207/s15328007sem1103_2
Alberini CM, Travaglia A (2017) Infantile Amnesia: A Critical Period of Learning to Learn and Remember. J Neurosci 37(24):5783–5795. https://doi.org/10.1523/jneurosci.0324-17.2017
Pollock AM, White AJ, Kirkwood G (2017) Evidence in support of the call to ban the tackle and harmful contact in school rugby: a response to World Rugby. Br J Sports Med 51(15):1113–1117. https://doi.org/10.1136/bjsports-2016-096996
Jullien S (2021) Prevention of unintentional injuries in children under five years. BMC Pediatr 21(1):1–11. https://doi.org/10.1186/s12887-021-02517-2
Acknowledgements
HC was funded by the Economic and Social Research Council South Coast Doctoral Training Partnership (Grant Number ES/P000673/1). We are grateful to the Centre for Longitudinal Studies (CLS), UCL Social Research Institute, for the use of the MCS dataset and to the UK Data Service for making them available. Neither CLS nor the UK Data Service bear any responsibility for the analysis or interpretation of these data.
Funding
No funding was received towards this work.
Author information
Authors and Affiliations
Contributions
HC had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors meet all four ICMJE criteria for authorship. HC, JH, HE, and VB conceived and designed the study. HC developed the models, analysed and interpreted the data, and wrote the first draft. All authors discussed the results, critically edited the paper, and approved the final version submitted here for publication. All authors agree to be accountable for this work.
Corresponding author
Ethics declarations
Conflict of interest
The authors report no competing interests.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Carr, H.R., Hall, J.E., Eisenbarth, H. et al. The bidirectional relationship between head injuries and conduct problems: longitudinal modelling of a population-based birth cohort study. Eur Child Adolesc Psychiatry 33, 411–420 (2024). https://doi.org/10.1007/s00787-023-02175-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00787-023-02175-y