Vious report, in this study ferritin, transferrin and TIBC had the lowest sensitivities to diagnose ID [22]. The low sensitivity of ferritin is explained for being an acute phase reactant [19], and thus, its plasma concentration may not reflect the actual iron status in the presence of inflammation, which was very prevalent in the study 64849-39-4 population (88 ) [19,48]. To solve this limitation, it is usually recommended to measure another acute phase protein [such as CRP or a-1-acid glycoprotein], and to adjust the ferritin level by the presence of order Dimethylenastron inflammation [49]. However, in this study the 25033180 sensitivity of ferritin did not improve after adjustment by the level of CRP, which could be explained by the stabilization of ferritin levels once iron stores are exhausted [48]. The observed low sensitivities of both transferrin and TIBC may also be due to their alteration during an inflammatory process [19,50]. Transferrin is an acute negative protein, i.e., it decreases during an inflammatory process, while TIBC values derive from the measurement of transferrin and therefore are also affected by inflammation. The TfR-F index has been suggested as a useful parameter for the identification of iron depletion even in settings with high infection pressure [18], and it was shown to be the best predictor of bone marrow iron stores deficiency in a previous report [22]. In contrast, in this study the TfR-F index showed a low sensitivity (42 ), and only its adjustment by the level of CRP [44] increased the sensitivity to 75 , while reducing the specificity from 91 to 56 . We found that sTfR, TfR-F index (adjusted by the level of CRP), and transferrin saturation showed the highest sensitivities. Moreover, sTfR and TfR-F index showed the highest AUCROC ( 0.75). The sTfR ROC curve indicated that there was no alternative cut-off with higher sensitivity than that of the current one (1.76 mg/l) without lowering the specificity below 50 . For the TfR-F index, the ROC curve showed that the sensitivity of this marker could be improved from 42 to 78 by changing the current cut-off from 1.5 to 0.86. It can be noticed that the performance of TfR-F index with the cut-off of 0.86 is similar to the performance of TfR-F index corrected by the CRP level (1.5 if 
CRP,1 mg/dl; 0.8 if CRP 1 mg/dl). However, this similarity is not coincidental, since 88 of the study participants had a CRP 1 mg/dl. This observation is 1326631 in contrast with that of a previous study, whereby in spite of a similar prevalence of inflammation (89 ) it was found that the TfR-F index unadjusted by the CRP level was a good marker of ID [22]. The findings of the current study show that the TfR-F index should be adjusted by the CRP level for maximal prediction of bone marrow iron stores deficiency in our setting, and indicate a lack of consistency of the diagnostic efficiency of current iron markers across different populations. In this study, the MCHC, which could be a potentially feasible iron marker for resource poor settings, had an AUCROC of only0.59 (p = 0.3382). This finding is also in contrast with the performance of this marker observed in the Malawian study where the AUCROC of MCHC was 0.68 (p = 0.001) [22]. The poor performance of MCHC in our study could be due to the high prevalence of a-thalassaemia in this population (64 among the 121 anaemic children in the case-control study; 78 among the 41 study participants included in this analysis). It has been reported that a-thalassaemia carriers have.Vious report, in this study ferritin, transferrin and TIBC had the lowest sensitivities to diagnose ID [22]. The low sensitivity of ferritin is explained for being an acute phase reactant [19], and thus, its plasma concentration may not reflect the actual iron status in the presence of inflammation, which was very prevalent in the study population (88 ) [19,48]. To solve this limitation, it is usually recommended to measure another acute phase protein [such as CRP or a-1-acid glycoprotein], and to adjust the ferritin level by the presence of inflammation [49]. However, in this study the 25033180 sensitivity of ferritin did not improve after adjustment by the level of CRP, which could be explained by the stabilization of ferritin levels once iron stores are exhausted [48]. The observed low sensitivities of both transferrin and TIBC may also be due to their alteration during an inflammatory process [19,50]. Transferrin is an acute negative protein, i.e., it decreases during an inflammatory process, while TIBC values derive from the measurement of transferrin and therefore are also affected by inflammation. The TfR-F index has been suggested as a useful parameter for the identification of iron depletion even in settings with high infection pressure [18], and it was shown to be the best predictor of bone marrow iron stores deficiency in a previous report [22]. In contrast, in this study the TfR-F index showed a low sensitivity (42 ), and only its adjustment by the level of CRP [44] increased the sensitivity to 75 , while reducing the specificity from 91 to 56 . We found that sTfR, TfR-F index (adjusted by the level of CRP), and transferrin saturation showed the highest sensitivities. Moreover, sTfR and TfR-F index showed the highest AUCROC ( 0.75). The sTfR ROC curve indicated that there was no alternative cut-off with higher sensitivity than that of the current one (1.76 mg/l) without lowering the specificity below 50 . For the TfR-F index, the ROC curve 
showed that the sensitivity of this marker could be improved from 42 to 78 by changing the current cut-off from 1.5 to 0.86. It can be noticed that the performance of TfR-F index with the cut-off of 0.86 is similar to the performance of TfR-F index corrected by the CRP level (1.5 if CRP,1 mg/dl; 0.8 if CRP 1 mg/dl). However, this similarity is not coincidental, since 88 of the study participants had a CRP 1 mg/dl. This observation is 1326631 in contrast with that of a previous study, whereby in spite of a similar prevalence of inflammation (89 ) it was found that the TfR-F index unadjusted by the CRP level was a good marker of ID [22]. The findings of the current study show that the TfR-F index should be adjusted by the CRP level for maximal prediction of bone marrow iron stores deficiency in our setting, and indicate a lack of consistency of the diagnostic efficiency of current iron markers across different populations. In this study, the MCHC, which could be a potentially feasible iron marker for resource poor settings, had an AUCROC of only0.59 (p = 0.3382). This finding is also in contrast with the performance of this marker observed in the Malawian study where the AUCROC of MCHC was 0.68 (p = 0.001) [22]. The poor performance of MCHC in our study could be due to the high prevalence of a-thalassaemia in this population (64 among the 121 anaemic children in the case-control study; 78 among the 41 study participants included in this analysis). It has been reported that a-thalassaemia carriers have.