Engage: Child abuse and neglect – new evidence they can cause mental health problems
Our research team wanted to know if childhood maltreatment directly causes poor mental and physical health. One way to investigate this is using genetics. To begin, we combined data on childhood maltreatment with genetic data in over 180,000. This allowed us to check whether the differences in people’s experience of childhood maltreatment was linked to differences in their genetics.
We found that only around 9% of these experiences of childhood maltreatment are linked to genetics. But this does not mean that genetics is causal for experiencing childhood maltreatment, as we shall see later. This also does not mean that a child is to blame, or that childhood maltreatment cannot be reduced regardless of the child’s genetics. The genetic signal reflects the environment a child grows in and it shows how important the environment is (such as the conditions a child is raised in) as a known and modifiable risk factor for addressing childhood maltreatment.
Children are absolutely not responsible for being maltreated – so how is it possible that our study has identified a genetic component associated with experiencing childhood maltreatment?
The genetic signal we found is a reflection of the environment that the participants grew up in through a mechanism called gene-environment correlation. Gene-environment correlation suggests that a person’s genetics impact how they engage with their environment – and subsequently how their environment (including friends and family) interacts with them.
For example, a child who is short may be bullied because of their height. Because height is substantially genetic, genes for height may be correlated with being bullied. So height genes don’t actually cause bullying, but rather the environment a child grows up in (where being short is viewed negatively) does.
To understand gene-environment correlation in childhood maltreatment, we investigated differences in experiences of maltreatment between siblings and within families. We found that parental behaviour, a child’s behaviour, and a parent’s response to that behaviour were all important factors in childhood maltreatment. For example, children who are overactive, which is partly genetic, might be at increased risk for punitive maltreatment because of the mismatch between the parent’s expectations and the child’s actual behaviour. This suggests that we need better interventions addressing the complex family dynamics to prevent maltreatment.
After identifying a genetic component associated with experiencing childhood maltreatment, the next step was proving whether childhood maltreatment really does cause poorer health. To do this, we used a recently developed method called Mendelian randomisation.
Typically, when researchers want to prove causality we would use randomised controlled trials. These randomly assign people to groups to prove if A (in this case, childhood maltreatment) causes B (poorer health).
But to investigate if childhood maltreatment causes health conditions using this method, researchers would have to randomly split children into two groups, where one group would be exposed to maltreatment. The researchers would then need to wait for several years to see if this group is more likely to develop health conditions. However, this is extremely unethical and cannot be done.
Instead, Mendelian randomisation splits people randomly into groups based on their genetics – acting similar to a randomised controlled trial. In other words, we used the genetic variants linked with the experience of childhood maltreatment as a stand-in for maltreatment.
Using this method, we found that people who had higher scores on genetic indicators of childhood maltreatment also had higher risk of mental health problems, including depression, ADHD and schizophrenia. These findings not only confirm previous results from epidemiological studies which have shown links between the two, but also provide evidence that childhood maltreatment, at least in part, causes poorer mental health.
We did not find evidence to suggest that experiencing childhood maltreatment is causal for later physical health conditions, like type 2 diabetes or coronary artery disease.
Mendelian randomisation is not without limitations. For one, both childhood maltreatment and mental health conditions are genetically complex. Experiencing different types of maltreatment and at different points in childhood may also have different effects on mental health. However, it’s one of the best methods we currently have to establish causality without experimental manipulation.
Ultimately, our research shows genetics can be used as a tool in answering questions about the role of environment in many health conditions, including those that are difficult – or even impossible – to address solely using other types of study. Our findings also highlight the urgent need for better protective measures to safeguard children at risk for maltreatment – which would also reduce the risk of mental health conditions later in life.
About the Authors
Varun Warrier receives funding from the Wellcome Trust, the Templeton World Charity Foundation, St. Catharine’s College, Cambridge, and the Autism Research Trust.
Charlotte Cecil receives funding from from the European Union’s Horizon 2020 Research and Innovation Programme under the Marie Skłodowska-Curie grant agreement No 707404 and grant agreement No 848158 (EarlyCause Project).
Marinus van IJzendoorn receives funding from the Dutch Ministry of Education, Culture, and Science and the Netherlands Organization for Scientific Research (grant number 024.001.003, Consortium on Individual Development) and a Spinoza Prize of the Netherlands Organization for Scientific Research. Some of his work on child maltreatment has been supported by the Lumos Foundation, a charity that financed some activities of The Lancet Commission on Institutionalisation and Deinstitutionalisation of Children, of which he was a member. He is a founder and member of the non-profit Beagle Advice, Research and Development company, with no conflict of interests with the current work.
This article was originally published on the ConversationPicture (c) Jon Challicom / NSPCC / PA.