The best way to counteract this assertion would be with a solid and scientifically sound study.
First, you would need to define physical and mental abnormality before you go looking at criminals. You cannot look at each criminal and try to find an abnormality. This would bias you in favor of finding such abnormalities and you might see abnormalities where none exist.
Second, you would need to look at the criminal population. You would need to check them for the characteristics that you have defined as abnormalities.
Third, you would need to look at a sample of people who are not criminals. You would need to determine if those people had the abnormal characteristics you had identified.
Overall, you would need to compare the levels of abnormality in the criminal population with that of the non-criminal population. If there was a difference, you would have to run statistical tests to ensure that it was a statistically significant difference.
If you did find a statistically significant difference, you would need to be sure that the abnormality caused the crime, rather than the crime causing the abnormality.
In other words, it would take a lot of resources and study to prove or disprove this assertion.