SIRS, We are most grateful to Grønbaek and Eivindson for the interest in our article, and we are pleased to reply to their comments.First of all, based on the criticism raised, we have now performed correlation analyses between the GH/insulin-like growth factor (IGF)-1 axis components and markers of inflammation at each individual time point, but no significant results were obtained. We had not, in fact, previously performed these analyses for one simple reason: as documented by the studies cited, 1, 2 it is not reasonable to expect significant correlations between the GH/IGF-1 axis components and markers of inflammation in inflammatory bowel disease (IBD) patients when the sample is limited in size, as too many factors other than inflammation are known to influence the endocrine axis, i.e. age, gender, nutritional status, other hypothalamic peptides, sex steroid availability, etc.
3As clearly emerges in Figure 1(a/b) and, as has been correctly pointed out, correlation analyses have been carried out pooling data from different time points. We agree with Eivindson and Grønbaek that, statistically, this is not the most appropriate method when seeking correlations, because not all the observations are independent, and results can be affected by time trends. However, we would explain why we were persuaded to conduct this kind of analysis. Also in previous studies by Eivindson et al.,4 and by Street et al., 5 with a con-