Dear Editors,

We read with great interest the recent systematic review on dietary interventions and behaviour of prisoners by Poulter and colleagues^{(Reference Poulter, Coe and Graham1)}. The authors are to be commended for conducting a deep analysis of an important topic, a subject matter that been too often overlooked in mainstream academic discourse, and carceral policy. We are hopeful that their efforts, and their suggestions for further research, will be widely disseminated and used as a roadmap for investigators to follow.

The authors set out clear inclusion/exclusion criteria for the type of studies that would be included in the analysis, most notably the exclusion of those that did not include a separate comparator group. Our concern is that the complete dismissal of studies wherein subjects acted as their own control may dilute the findings, leaving the reader with an impression that there has been little historical investigation of dietary alterations in prison systems. The inclusion of studies wherein subjects acted as their own control significantly expands the overall data pool. At least a half-dozen intervention studies, all conducted in correctional facilities in the 1980s and held in the PsycINFO database, were not considered^{(Reference Schoenthaler2–Reference Schoenthaler and Doraz7)}. Collectively, these studies involved several thousand inmates in a variety of juvenile correctional facilities, wherein dietary interventions to reduce processed foods and sugar content were employed. One additional study by the same group introduced a flavonoid-rich orange juice at mealtimes^{(Reference Schoenthaler8)}. The outcome measures included pre-intervention facility records (typically months before) and records during and after the interventions. While there was some turnover during the dietary intervention periods, most of the subjects were in the facilities before, during and after the interventions, acting as their own controls. The generalised findings included significant reductions in violence, antisocial behaviour and correction officer-noted infractions^{(Reference Schoenthaler and Bier9,Reference Schoenthaler10)} .

Of course, the research designs of these quasi-experimental time-series studies are open to critique, as were the studies chosen for qualitative synthesis by the authors. In order to advance the science of diet, behaviour and mental health in carceral systems, we agree that controlled clinical trials are critical. However, compared with supplement trials, these are notoriously difficult to blind and deliver, measure adherence, and analysis of the composition of the consumed dietary pattern is a challenge^{(Reference Staudacher, Rocks and Jacka11)}.

It is our contention that the studies omitted from the Poulter and colleagues systematic review can offer vital learnings, including ways to develop international transdisciplinary research. Awareness of these studies might encourage investigations that are not necessarily comparator-controlled, yet still informative. For example, in Maine, USA, the entire carceral system is undergoing a transformation in dietary policy, including limitations on highly processed foods, and expansion of nutrient-rich whole foods^{(Reference Cebula12)}. This represents enormous research opportunity, including investigations of unexplored mechanisms such as the microbiome. What we know from the studies omitted from the systematic review is that changes to official diet policies are associated with significant changes in official records of undesirable behaviour^{(Reference Schoenthaler13)}. Notwithstanding advances in mechanistic science, what remains unknown is why.