Key points are not available for this paper at this time.
We have concerns over a study, published recently in the British Journal of Pharmacology (BJP) (Michael et al., 2025). Our points, discussed below, could be addressed by strengthening the procedure for peer review and editorial scrutiny of submitted manuscripts. We acknowledge that authors are required to confirm the local ethical approval of their preclinical research. Unfortunately, ethics assessments are not consistent internationally, and so the advice of the BJP Consulting Editor (ARRIVE Guidelines and Animal Welfare) should be sought as a matter of routine in respect of articles with key words such as stress, pain, trauma, and so on. The research in question used a ‘Chronic Social Defeat Stress’ (SDS) model, a procedure that first involves the confinement of a male test mouse within the home cage of a resident male mouse that had been pre-screened to confirm its innate aggression. The initial hostile encounter lasted for 10 min, and was followed by continuous psychological stress for 10 days, where the test mouse was separated by a perforated Plexiglas divider from a different aggressive mouse each day. Citation of an earlier publication was used to provide key details of this procedure, in which it was disclosed that ‘In cases in which repeated defeats lead to the development of open wounds exceeding 1 cm, removal of the mouse from the study and immediate euthanasia is indicated’ and that the procedure ‘… rarely, if ever, leads directly to death in the mice undergoing defeat. However, death may occur in the hours following cessation of daily defeats’ (Golden et al., 2011). After this prolonged bout of severe stress, the behaviour of the SDS mice was assessed in a battery of tests; the outcome measures were used as indices of animals’ subjective state and cognitive performance. One aim of these screens was to assess the stress resilience (or lack of it) of animals as a risk factor for opiate misuse. Another aim was to test the efficacy of the ketamine metabolite, (2R,6R) 6-hydroxynorketamine, as a candidate treatment. Our main concern is that there was no justification of the need to use such a severe stress in order to meet the experimental objectives. Consideration of the 3Rs (Replacement, Reduction and Refinement) when planning animal experiments is now regarded as an ethical imperative, internationally, but there is no indication that Michael et al. (2025) had considered the opportunities for Refinement. Another is the lack of transparency about the number of physical attacks experienced by the animals or steps taken to monitor their welfare and ameliorate ensuing harms. Both those points would need to be addressed for compliance with ARRIVE 2.0 guidelines (therein, see Item 9 Experimental Procedures and Item 16 Animal Care and Monitoring). Given that unambiguous interpretation of research findings is predicated on a CLEAR description of the experimental procedures, this information is essential both ethically and scientifically (George et al., 2025). A third concern focuses on scientific validity (see ARRIVE 2.0: Item 12 Justification for the validity of animal species or model selection). The conclusions from this study assumed that the subjective state of an animal can be ‘diagnosed’ based on its motor, gustatory or vegetative behaviours. Examples include comparing the behaviour of stressed and unstressed animals in the ‘sucrose preference test’ (SPT) and the ‘female urine sniff test’, which were used to evaluate ‘anhedonia’. The severe stressor used in this study deviated markedly from the chronic unpredictable mild stress protocol that was validated for use in combination with the SPT (Monleon et al., 1995). Also, there are many reasons why severe stress might reduce sniffing of female urine by males. Consequently, it cannot be certain that changes in animal behaviour in either of these tests indicate anhedonia, which is an anthropomorphic interpretation that defies objective confirmation. Finally, it is hard to imagine how the severe stress imposed on the animals is relevant to humans in ways that would inform any translational research objectives. This lack of ‘face validity’ undermines one of the criteria used to endorse the use of animal models of human disorders, generally. In short, we would expect the journal (BJP) to require authors to confirm that experimental procedures that impose pain, suffering, distress or lasting harm on animals were necessary to meet the research objectives: that is, opportunities for Refinement were considered and ruled out. A description of the welfare monitoring of the animals also should be included, together with steps taken to ameliorate any harms and, when appropriate, the humane end-point(s) that would prompt euthanasia. In fact, these requirements are consistent with the BJP's existing editorial policy (Lilley et al., 2020; table 2, Points 2, 3 and 6), and so we assume that the publication of Michael et al. (2025), in its current form, was a procedural oversight. For that reason, we urge the journal to take steps to ensure that, in future, articles that do not comply with ARRIVE 2.0 and Lilley et al. (2020) do not slip through the editorial net. S. Clare Stanford: Conceptualization; writing—original draft; writing—review and editing. Amrita Ahluwalia: Writing—review and editing. Michael J. Curtis: Writing—review and editing. Farhad Dehkhoda: Writing—review and editing. Christopher H. George: Writing—review and editing. John Kolassa: Writing—review and editing. Vanessa Minervini: Writing—review and editing. Vandana Nikam: Writing—review and editing. None of the authors has any conflicts to declare.
Stanford et al. (Thu,) studied this question.