H567/03 Section B Option 2 - Raine et al. (1997) and biological crime prevention

Aim, sample and matched comparison

The aim was to investigate whether severely violent offenders referred in connection with a not-guilty-by-reason-of-insanity defence had differences in cortical and subcortical functioning. The predicted areas included the prefrontal cortex, angular gyrus, corpus callosum, amygdala, hippocampus and thalamus.

The original paper describes 41 people in California charged with murder or manslaughter, labelled 'murderers' for convenience. Referrals concerned an insanity defence, competence to stand trial or diminished capacity at sentencing. This is a selected forensic sample, rather than a random sample of all people who kill. There were 39 men and 2 women; mean age was 34.3 years. A control group of 41 had the same sex distribution, mean age 31.7, and was matched on age and sex. Six offenders diagnosed with schizophrenia were matched with six controls with schizophrenia.

This is a quasi-experimental, matched-group comparison: offender status existed before research and could not be randomly allocated. Group was the naturally occurring independent variable; relative glucose metabolism in specified brain regions was the main dependent measure. Matching reduces particular alternative explanations, but cannot match every aspect of participants' lives.

PET during an attention task

Participants were instructed to be free of medication for the two weeks before scanning. The paper reports negative urine screens for referred offenders. A degraded-stimulus continuous performance task required detecting target signals. Practice began ten minutes before the tracer injection; the actual task began thirty seconds before injection to reduce the initial novelty effect. A radioactive glucose tracer, fluorodeoxyglucose, was taken up during 32 minutes of task performance. Participants then moved to the PET scanner, with an individually moulded holder keeping the head still. Ten slices at 10 mm intervals were recorded. PET measured tracer uptake as an indicator of regional metabolic activity. Both groups underwent the same task and imaging procedures.

The task challenged sustained attention. Researchers compared relative metabolism, expressing regional uptake in relation to other regions in the same slice. PET here measures functioning during this task; it does not directly show a structural lesion, a criminal intention or a permanent level of self-control. The groups did not differ significantly in behavioural performance on the attention task, making a simple explanation based on poorer task performance less plausible.

Standardised scanning and quantification support replicability and objective measurement. However, anatomical localisation had limited resolution and slice placement could vary between people. The study used protocols and consent forms approved by the University of California, Irvine Human Subjects Committee. Consent procedures are a strength, while radioactive exposure and voluntariness in a forensic/custodial setting still require ethical consideration; committee approval does not remove those questions.

Findings and proposed pathways

Compared with controls, the offender group showed lower metabolism in bilateral prefrontal regions, the corpus callosum and superior parietal regions. Lower left angular-gyrus metabolism was a two-tailed trend, rather than a clear result at the conventional 5% level. Activity patterns in the amygdala, medial temporal region including the hippocampus, and thalamus showed abnormal left-right asymmetries. These are patterns relative to controls, not a claim that every offender had every abnormality.

Occipital metabolism was higher. The cerebellar increase was only a nonsignificant trend; the study did not find significant group differences in the caudate, putamen, globus pallidus or midbrain. The distinction prevents describing every region as impaired.

Reduced prefrontal regulation could contribute to impulsivity and difficulty considering consequences. Altered emotional processing, memory and communication between hemispheres could interact with this difficulty. These proposed pathways connect the brain findings to the topic 'What makes a criminal?' through possible vulnerability to violence. They were not directly demonstrated as causes of the offences.

For a small quantitative illustration, original Table 1 reports left lateral prefrontal relative metabolism means of 1.12 for controls and 1.09 for offenders. The difference is $1.12-1.09=0.03$. Relative to the control mean, this is $(0.03/1.12)\times100\approx2.68\%$ lower. These are relative metabolic values, not percentages of damaged brain or an individual's chance of offending.