Here is the typical story. A researcher shows a colorful picture of the brain, with a certain part being too small or too large, or having too little or too much blood flow. Then the researcher claims that these brain imaging changes support the validity of a certain diagnosis. In children, the common claim is that brain imaging changes support the validity of the diagnosis of ADD.
The problem, as all know, is that correlation cannot be assumed to be causation. The brain is always changing. How do we know we are dealing with an illness? There are differences in the brain PET scans of Republicans and Democrats. Are those diseases?
Further, which way does the arrow of causality go? Are the brain changes causing the illness? This is what is assumed usually. But there is another possibility: It could be that the illness is causing the brain changes.
This study is able to answer this question in relation to the construct of childhood ADD because it is prospective: it examines clinical diagnosis and brain imaging at different times moving forward in childhood. Its conclusion: brain changes are effects, not causes. They do not prove validity of ADD.
In this study 845 children received brain MRI testing at ages 8 and 10. They also had clinical assessments at ages 6, 8, and 10. The researchers divided clinical symptoms into two basic categories of “externalizing” (agitation, excitation, impulsivity) and “internalizing” (sadness, anxiety, withdrawal) symptoms.
Higher externalizing symptoms at age 6 were associated with later smaller total brain volume, with reductions in both gray and white matter, and smaller subcortical brain volume. Over four years, the association of externalizing symptoms remained strongest for smaller subcortical brain volume.
Higher internalizing symptoms at age 6 had no associations with brain changes initially, but after four years, there was some association again with smaller subcortical brain volume.
Neuroimaging at age 8 did not predict changes in externalizing or internalizing symptoms at age 10.
So, the prospective design allows for some claims of causality. If one event happens before a change in another event, the first event is more likely to be a cause. If one event does not precede another event, then the first event is not likely to be a cause. Here, MRI brain findings did not correlate with any later changes in psychiatric symptoms. In other words, MRI findings did not “cause” anything. In contrast, psychopathological states were associated with later changes in brain MRI. Or, psychopathology seemed to cause brain MRI changes. Specifically, externalizing symptoms of impulsivity and excitation caused atrophy of subcortical brain structures.
What are we to conclude? Rather, this study tells us what not to conclude: Brain MRI abnormalities occurring along with psychiatric symptoms does not mean that the former cause the latter; it could be the other way around.
Lastly, how can symptoms change the brain? We know that psychomotor excitation (e.g., mania, stress reactions, psychosis) is associated with overactivity of the hypothalamic-pituitary axis (HPA), with overproduction of steroid hormones, which are excitotoxic in the brain, causing cortical atrophy. This link is rather well established. It could be a mechanistic explanation for the clinical/biological finding of this study: clinical psychopathology leads to brain abnormality, not vice versa.
These results could support the view that controlling psychopathological symptoms is beneficial for the brain, and that short-term symptom improvement could have long-term neurological benefits.