6.1.6 - Diet as offender treatment

Diet and the proposed mechanism

Dietary treatment aims to correct nutritional inadequacies that may affect the processes supporting behaviour regulation.

The brain needs a continuing supply of nutrients. Vitamins and minerals contribute to metabolic processes, including processes involved in neural functioning; essential fatty acids contribute to cell membranes. If intake is inadequate, impaired functioning could contribute to difficulties with attention, emotional regulation or impulse control. Correcting inadequacy may therefore make regulation easier. This is a proposed pathway rather than a claim that one nutrient directly switches criminal behaviour on or off.

Improved diet can mean providing nutritionally adequate meals and helping people choose them. Supplementation adds measured nutrients to the existing diet. These are related approaches, but evidence for capsules does not automatically demonstrate the effect of changing a whole menu. Capsules permit a known dose and a matched placebo, making a controlled trial more feasible.

People differ in their nutritional needs, intake and absorption. A person with adequate intake may have little to gain from extra supplementation, while an individual with poor intake may benefit more. Social circumstances also affect access to food and choices. Nutritional support may complement psychological and social interventions; it should not replace an individual assessment of why an offence occurs. This lesson explains research, not a prescription for self-treatment.

Gesch: a controlled nutritional trial

Gesch et al. (2002) tested whether adding vitamins, minerals and essential fatty acids reduced prisoners’ antisocial behaviour.

The researchers recruited 231 volunteer young adult prisoners, aged at least 18, who gave informed consent. This was an experimental, randomised, double-blind, placebo-controlled trial conducted within ordinary prison operations. Volunteering concerns entry into the sample; random allocation concerns assignment to treatment. Random allocation does not turn volunteers from a prison into a representative sample of everyone who offends.

Participants received either active vitamin/mineral and essential-fatty-acid capsules or matched placebos, while their normal diet continued. Random allocation was organised within prison wings, helping balance conditions associated with location and the prison regime. The capsules were coded so participants and research staff did not know the allocation during the trial. Similar appearance and lack of an obvious distinctive taste helped maintain this double blind.

The outcome was proven disciplinary incidents, recorded in Governor reports for more serious events and minor reports for less serious breaches. These records were compared before and during supplementation. Researchers considered time at risk because participation varied as prisoners entered or left; the average supplementation period was about 142 days in each group. Counts without an exposure period would be misleading if one person was observed much longer than another.

A subset completed food diaries, and psychological assessments helped check baseline equivalence. The main test concerned the rate of disciplinary incidents, not whether participants claimed to feel better. No adverse reactions to supplementation were reported. Participants received a written debrief, and those given placebo were offered active supplements after the trial. The complete original paper, pp22–28 provides the method and findings.

Gesch: findings and comparisons

Gesch’s different percentage figures describe different comparisons and should not be exchanged.

In the analysis of all 231 recruited participants, those allocated active capsules committed an average of 26.3% fewer disciplinary offences than the placebo group during the trial. The reported 95% confidence interval was approximately 8.3% to 44.3% and the difference was statistically significant at about $p=0.03$, two-tailed. This is a between-group comparison, not the percentage reduction from the active group’s own baseline.

A further analysis considered 172 participants treated for at least two weeks after exclusions: 82 active and 90 placebo. Within this analysis the active group showed a reported 35.1% reduction from its own baseline, whereas the placebo reduction was 6.7%, which was not statistically significant. The active group’s baseline comparison was significant at $p<0.001$, two-tailed. The 172 is the total for both groups, not 172 people receiving active capsules.