Published online November 1, 2005
PEDIATRICS
Vol. 116
No. 5
November 2005, pp.
1258-1259
(doi:10.1542/peds.2005-1689)
Anemia Is an Important Finding and Is Likely to Respond to Iron Therapy
Charlotte M. Wright, MD
PEACH Unit
Community Child Health
University of Glasgow
Glasgow G3 8SJ, Scotland
Geoffrey P. Summerfield, DM
Department of Haematology
Queen Elizabeth Hospital
Gateshead NE9 6SX, United Kingdom
Robert Moy, MD
Institute of Child Health
Community Child Health
University of Birmingham
Birmingham B15 2TT, United Kingdom
To the Editor.—
We were interested to see the recent article1 that concluded that anemia was a poor predictor of iron deficiency (ID), defined as 
2 abnormal values out of 3 known markers of ID (ferritin, transferrin saturation, and free erythrocyte protoporphyrin). This seems to be a brave conclusion, because all available markers of ID are known to be highly nonspecific, and none can be said to provide a gold standard for diagnosis.
Our recent study, published after this article was accepted for publication, allows for a very different conclusion.2 We had been aware for some time of great uncertainty about the value of different markers of ID and that a therapeutic response to iron would theoretically be the best indicator. We thus examined the predictive value of a number of potential markers of ID (mean cell volume, hemoglobin [Hb], mean cell hemoglobin (MCH), ferritin, and zinc protoporphyrin) not simply on the basis of their overlap with other measures but also the extent to which they predicted a response to treatment with orally administered iron. We took venous blood from 462 children aged 13 months, and all of them with 1 abnormal value (147 [32%]) were treated with iron for 6 weeks; 125 returned to be retested after 3 months. Because individual abnormal test results will always tend to improve on remeasurement, we calculated a therapeutic-response score (TRS) that standardized and summed changes in each of the measures after treatment, thus giving a global indicator of changes in iron sufficiency.
We also found only a modest overlap between anemia and ID (defined as 2 abnormal values of ferritin, zinc protoporphyrin, or mean cell volume), but to our surprise both the Hb and MCH level proved to be powerful predictors of a therapeutic response, and even in isolation this was of an equivalent order to the response seen for ID (see Tables 1 and 2). The intercorrelation between markers was not generally high, with ferritin in particular essentially unrelated to the other markers.
Ours was not a very large study, and the combined prevalence of anemia and ID was only 9%. Thus, there is scope for reexamining this question in a larger population with a higher prevalence of ID. However, unless markers are tested against either a therapeutic response or bone marrow aspirate iron staining, all that is really being demonstrated is how poorly most of them actually function as indicators of ID. For the time being, we suggest that anemia is still an important finding that is likely to respond iron therapy.
REFERENCES
- White KC. Anemia is a poor predictor of iron deficiency among toddlers in the United States: for heme the bell tolls.
Pediatrics. 2005;115
:315
–320[Abstract/Free Full Text]
- Wright CM, Kelly J, Trail A, Parkinson KN, Summerfield G. The diagnosis of borderline iron deficiency: results of a therapeutic trial.
Arch Dis Child. 2004;89
:1028
–1031[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2005 by the American Academy of Pediatrics
