Till startsida
University of Gothenburg
To content Read more about how we use cookies on gu.se

Child maltreatment and neurodevelopmental disorders (ESSENCE)


This edition of the Researcher’s Corner presents the result of a newly published article on associations between child maltreatment and neurodevelopmental disorders. We caught up with the article’s first author Lisa Dinkler for a conversation about the study behind the article. How and why did it come about and what were the results of the study? Lisa Dinkler is a psychologist and a full-time PhD student at the GNC. In her PhD project Lisa investigates the development and course of eating disorders and their association with ADHD and Autism Spectrum Disorder (ASD) with the help of three longitudinal cohorts.

You and your colleagues recently published an article focusing on associations between maltreatment and neurodevelopmental disorders. Could you provide us with some background into this study? How did the study come about and what were you interested in? What previous research is available within this field?

  • To start off my colleagues got interested in this because in their clinical practice and their research they were struck by the number of neurodevelopmental or psychiatric problems some maltreated children have. My colleagues have met children who had six or seven co-existing diagnoses, which just did not seem possible, and we wanted to know if this is real or the exception to the rule. Childhood maltreatment has been shown to be strongly associated with psychiatric disorders in childhood and adulthood, but whether maltreated children also are more likely to have lots of neurodevelopmental disorders compared to non-maltreated children had not been investigated. So we decided to have a look at this in a population context.


  •  It is a common notion among clinicians that maltreatment causes the developmental or psychiatric problems that the individuals who have been maltreated suffer from. We, however, knew that neurodevelopmental disorders like ADHD and ASD are strongly heritable, so we wondered if perhaps genetic factors might be underlying any association. The question as to whether maltreatment could cause neurodevelopmental disorders had not really been investigated in a population-representative sample. Our objective was to determine whether maltreated children have an elevated number of neurodevelopmental disorders and, if so, whether child maltreatment might cause this increased ‘load’ of neurodevelopmental disorders.


How did you go about studying this?

  • To investigate these questions we used a sample from the Child and Adolescent Twin Study in Sweden (CATSS), which is a nationwide, ongoing birth cohort study aiming to include all twins born in Sweden since 1st July 1992. The parents of all 9-year-old twins are contacted and invited to participate in a telephone interview. In this interview, the parents are asked if their children have ever experienced emotional abuse/neglect, physical neglect, physical abuse, and/or sexual abuse. The interview also contains the Autism-Tics, ADHD, and other comorbidities inventory (A-TAC), a well-validated instrument with very good psychometric properties for the screening of neurodevelopmental disorders. In our study we looked at the association between child maltreatment and four neurodevelopmental disorders – ADHD, ASD, learning disorders, and tic disorders – in about 8,200 nine-year-old twins.


What are the advantages of twin studies?

  • We were mainly interested in the ‘load’ of neurodevelopmental disorders, that is, the number of neurodevelopmental disorders out of the four possible ones we included in the study. To investigate if child maltreatment could cause an increased load of neurodevelopmental disorders, we used the co-twin control method. This method is designed to investigate causal effects in observational research, so it can tell us, if an association is consistent with a causal effect, or if it is better explained by familial factors, such as genetic factors and common environmental factors (e.g. parental socioeconomic status, parental education level). We know that identical twins share 100% of their genetic makeup and they are also matched for common environmental factors, since they are brought up in the same family. This means that by looking at identical twin pairs we can investigate the causal effect of a random factor that the twins do not share on any outcome while we are controlling for the familial factors. So, if we take identical twin pairs who are discordant for child maltreatment, that is, one twin has been maltreated while the other one has not, we can see if child maltreatment has an elevating effect on the number of neurodevelopmental disorders in the maltreated twin, while we control for the genetic and environmental factors shared within the twin pairs. If, for example, the maltreated and the non-maltreated twin in a pair share similar numbers of neurodevelopmental disorders, then it is very likely that the maltreatment has not caused the load of neurodevelopmental disorders.


How would you sum up the results of your study?

  • Regarding our first question we found what we expected: maltreated children had an elevated number of neurodevelopmental disorders compared to children who have not been maltreated. In fact, maltreated children were nearly six times as likely, compared to their non-maltreated peers, of having three or more neurodevelopmental problems. In other words, maltreated children have a bigger load to carry in terms of both – the maltreatment and the neurodevelopmental disorders.


  •  In the co-twin control design we found that identical twins discordant for child maltreatment – one twin maltreated and the other one not – did not differ significantly in their number of neurodevelopmental disorders. So the association between child maltreatment and increased load of neurodevelopmental disorders disappeared when we took genetic and shared environmental factors into account. This in turn is not consistent with the hypothesis that the child maltreatment caused the increased load of neurodevelopmental disorders in maltreated children. Instead, most of the association between child maltreatment and load of neurodevelopmental disorders seems to be explained by common genetic effects, a so-called genetic confounding of the association.


  •  Lastly, we also looked at individual diagnoses and wanted to know, whether child maltreatment could still cause a maltreated child having more symptoms of a specific neurodevelopmental disorder when controlling for familial effects. Again, we looked at identical twins discordant for child maltreatment. We found that the maltreated twins of those pairs had slightly higher symptoms of ADHD and ASD than the non-maltreated twins in the pairs. With this result we cannot exclude that child maltreatment might cause an increase of ADHD and ASD symptoms. However, since we only have a cross-sectional sample here and no reliable information on the times of onset of child maltreatment and neurodevelopmental symptoms, we cannot determine the direction of causality in the association of child maltreatment and number of neurodevelopmental symptoms; thus nothing can be said with certainty regarding causal effects. So the results could either suggest that child maltreatment caused the increase in symptoms of ADHD and ASD, or that children with more symptoms of ADHD and ASD are more likely to be maltreated.


  • Summarising, we can conclude that child maltreatment does not cause an increase in prevalence of neurodevelopmental disorders, while mainly genetic factors are underlying this association. However, we cannot exclude that child maltreatment might be responsible for an increase of symptoms of these neurodevelopmental disorders. It is likely that this increase of symptoms happens above or below the threshold for being labelled as having a neurodevelopmental disorder, and it does therefore not affect the prevalence of neurodevelopmental disorders (i.e. number of neurodevelopmental disorders).

What are the clinical implications of your results?

  • The clinical implications of these results are quite profound. In our clinical experience, maltreated children are less likely to be assessed for neurodevelopmental disorders because we tend to assume that their behavioural problems can be explained by the fact they have been maltreated. What this study is telling us is that ALL children who have been maltreated must have a detailed assessment for neurodevelopmental disorders because they are at much higher risk of having any one of these disorders, and are at much higher risk of having a quite complex presentation (i.e. multiple neurodevelopmental disorders).


  • As the association between child maltreatment and neurodevelopmental disorders goes both ways, we also need to realise that if children present with multiple neurodevelopmental disorders, there is a possibility that they are at high risk of being maltreated – so we need to watch for that as well.


  • The third important implication is that the study should also make us think again about what we call “trauma-related disorders” – even in adults – because many of these adults could actually have neurodevelopmental disorders like ASD and ADHD. We need, as clinicians, to get away from the idea that maltreatment can explain all of a child’s or an adult’s problems.


What are the research implications of your results?

  • As mentioned before, we did not have sufficient information in our data to investigate the temporal sequence of onset of child maltreatment and symptoms of neurodevelopmental disorders. Therefore, we could not determine the direction of causality in the association between child maltreatment and increased neurodevelopmental symptoms which we found in identical twins. Prospective twin studies starting prenatally or in very early childhood would be needed to clarify the temporal sequence of the onset of child maltreatment and symptoms of neurodevelopmental disorders to disentangle causal effects. Further studies could also estimate the magnitude of common genetic effects on the association of child maltreatment and neurodevelopmental disorders applying traditional quantitative twin modelling. It would furthermore be interesting to have a closer look at sex differences with higher-powered datasets; for example, we found that child maltreatment did increase ADHD symptoms in boys, but not in girls.


Last but not least, you wrote your master thesis on the association of childhood maltreatment with antisocial personality disorder and psychopathy among young adult offenders. Can you tell us a little bit about that study and its results?

  • The main objective of the study was to map experiences of physical and emotional abuse in childhood among violent and sexual offenders. Furthermore we wanted to investigate differences in the prevalence and severity of childhood abuse between inmate groups that were classified based on the presence or absence of antisocial personality disorder (ASPD) and psychopathy.


  • We used a nationally representative cohort of 270 male prisoners between 18 and 25 years who had been convicted for violent or sexual crimes. The inmates were asked about their experiences of physical and emotional abuse in childhood in a semi-structured clinical interview and the Childhood Trauma Questionnaire (CTQ). Psychopathy was measured with the Psychopathy Checklist-Revised (PCL-R) and ASPD was assessed by the interviewer.


  • We found that childhood abuse experiences were very common in this population of violent and sexual offenders: 73% indicated the occurrence of abuse in at least one measurement (interview or CTQ). In the face-to-face interview, inmates with ASPD reported childhood abuse significantly more often (54%) than inmates without ASPD (37%). However, inmates with ASPD and psychopathy did not report more childhood abuse than inmates with ASPD without psychopathy. The CTQ scores were not significantly associated with ASPD diagnosis or psychopathy (PCL-R score). There were also no differences in the severity of experienced emotional and physical abuse between the inmate groups as measured with the dimensional CTQ scales.


  • We know that offenders with ASPD exhibit more persistent and violent criminal behaviour than offenders without this personality disorder. Our results indicate that childhood abuse might lead to a more established pattern of antisocial behaviours. Addressing abuse experiences and preventing (further) childhood abuse might improve the effectiveness of early prevention programs for antisocial children and inhibit the development of criminal careers. When we compared the answers given by an inmate in the clinical interview and in the CTQ we found that the agreement between the two measures was poor and false-negative reports occurred in both measures (i.e. inmates reported abuse in one measure, but not in the other). This tells us that multiple approaches should be applied to reliably measure childhood maltreatment experiences in violent and sexual offenders. Further research should investigate reliable measurements of childhood maltreatment experiences in offenders.




Lisa Dinkler; Sebastian Lundström; Ruchika Gajwani; Paul Lichtenstein; Christopher Gillberg; Helen Minnis (2017)
. Maltreatment-associated neurodevelopmental disorders: a co-twin control analysis. ISSN: 1469-7610, GUP 250147

Photo: Josefin Bergenholtz

By Nanna Gillberg

Page Manager: Anna Spyrou|Last update: 11/7/2017

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?