Novel object recognition task for mice: Is it a test for memory, object neophobia or innate preference?

Novel object recognition task (NOR) is one of the most widely used memory tests for mice. Over the years of intensive use, the test protocol has become largely standardized except for one detail, the objects themselves. In three experiments including in total 41 C57BL/6 J mice, we tested the impact of the object novelty vs. its general appeal to an individual mouse using 18 object pairs. This was possible by having a cross-over design where the same object was familiar to half of the mice and novel for the other half. Some objects were consistently approached or avoided across mice. However, an object pair that showed a strong novelty preference in one group of mice, could fail to do so in another group. Therefore, a counterbalance design of the object assignments is important. Pre-test habituation of mice to the presence of objects in their home cage increased object exploration in the test for the first three object pairs. We also assessed how well the object-related details in NOR are disclosed in current literature. As a representative sample, we searched for NOR reporting in mice published in top-tier Nature journals during 2024 and found 69 articles. Notably, only 24.6 % of the articles described the objects and only 14.5 % employed a cross-over design. Most alarming was that 20.3 % of the reports mixed up object neophobia and memory as the task outcome. Obviously, more detailed reporting is needed to ensure reproducibility of results obtained with the NOR test in mice.