Cognitions do play an important part in mediating the relationship between challenging behaviour and stress. Further research is needed to examine the similarities and differences between the mediation models for aggression and self-injurious behaviour.

Fragile X Syndrome (FXS) is characterised by features including anxiety and autistic-like behaviour, which led to early hypotheses that aberrant physiological arousal may underlie the behavioural phenotype. In line with this, several lines of evidence suggest that the HypothalamicPituitary-Adrenal (HPA) axis may be altered in the syndrome. This review collates evidence to determine the nature of HPA axis baseline activity and reactivity (as measured by glucocorticoid levels) differences in FXS, and its relationship to behaviour. Through a search electronic databases, 15 papers were identified which provided data on humans with FXS or the FMR1 knockout mouse model. The findings across studies are mixed, though trends in the findings can be seen, including elevations in cortisol levels, particularly in males. Preliminary findings also highlight associations between cortisol levels and key behaviours associated with the syndrome, such as gaze avoidance. Areas for future research are discussed. Many thanks to the editor and to the reviewers for taking the time to kindly re-review the manuscript and provide useful feedback. I have addressed the issues raised and made amendments, as detailed individually below, using tracked changes:As requested, I have re-structured the article to align to the review questions. I have, however, addressed gender differences according to each aspect of the cortisol release (baseline activity, magnitude/ duration of response, for humans and animals) following the discussion on that specific topic, and clearly marked this with sub-headings. However, I hope that I have now better highlighted the themes in gender differences observed, and have included a separate table on this issue, as requested.When referring to each study in the text, I have briefly stated how many individuals were involved and what the ages of the individuals. I have done this for the studies with human participants, as from drafting this for the animal studies (which are addressed often is less detail in the text) this became very cumbersome to read. However, the animal results are presented directly alongside the sample sizes in Table 1, which I hope will make this interpretation more simple.On p15 we mention the findings of Roberts et al (2009) in relation to the magnitude of cortisol responses. The authors of this study split their analyses according to two groups: those individuals with FXS who met the criteria for autism, and those who did not. The reviewer requested that this discussion be moved to the section on cortisol and behaviour. Given the paucity of studies discussing the magnitude of cortisol responses in this group, I have proposed that we leave this study here (as there is no option to present the results without referring to the grouping according to autism symptomatology) but to note that a further discussion on the topic is included later in the manuscript. Please do let me know if this is acceptable.The first paragraph on page 17 was noted to be out of place. I have now clarified the int...

Fragile X Syndrome (FXS) appears to be associated with an increased risk for engaging in challenging behaviour, particularly self-injury, relative to those with mixed aetiology learning disabilities. Such behavioural issues are reported to be of high concern for those providing support. As such, this systematic review aimed to gain further epidemiological data regarding challenging behaviours in individuals with FXS, including: self-injurious behaviour (SIB), hand-biting as a specific topography of SIB, aggression and property destruction. Twenty eight manuscripts were identified which reported the prevalence of a relevant topography of behaviour, with widely varying prevalence estimates. Weighted averages of the prevalence of behaviours were calculated across studies. Comparison of proportions revealed significant gender differences and differences in the prevalence of types of behaviour. It is hoped that this comprehensive overview of data on this clinically significant topic will help to inform and drive future investigation to understand and provide effective intervention for the benefit of those with FXS.

No abstract