![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Department of Anthropology, University of Washington, Seattle, WA 98195-3100
Department of Anthropology, University of Washington, Seattle, WA 98195-3100 Laboratory for Human Biology Research, Department of Anthropology, Northwestern University, Evanston, IL 60208
Dear Editor:
In their commentary, Zeng et al. raise a number of concerns regarding our method for measuring transferrin receptor (TfR) in dried blood spots. While their commentary is directed at a recent application of this method (1), most of their concerns are addressed in our earlier publication validating the blood spot TfR method (2). We refer readers to this publication, but we also respond directly to their concerns here.
First, Zeng et al. do not believe that our method qualifies as "commercially available" because the assay calibrators and calibrator diluent are not included with the Ramco kit. This is an important point because we claim that the availability of commercially available reagents is a major advantage of our method. However, as stated in our methods publication, both the calibrator and calibrator diluent are provided by Ramco, and would thus be consistent across laboratories. We therefore believe that they can reasonably be called "commercially available," even though they are not part of the off-the-shelf serum kit.
Second, the commentary questions our attempt to minimize matrix differences by diluting calibrators in washed erythrocytes, and proposes instead that we use a monoclonal antibody to minimize any matrix effect. While we agree that this may work in principle, we do not believe it is advisable in this case because it defeats a major advantage of our method: all reagents are available from a single supplier, and these reagents have been validated by the supplier to work together for optimal performance in the measurement of TfR. Furthermore, while we recognize that the use of washed erythrocytes is an imperfect solution to the problem of matrix differences, we believe that it is a favorable solution. Other blood spot protocols developed for a wide range of analytes have taken a similar approach (3–6).
Third, Zeng et al. are concerned with the uneven distribution of TfR across the whole blood spot, an issue we discuss in detail (2). To overcome this potential source of bias we recommend that discs of blood be punched out from the periphery of dried blood spots exclusively, and that the punch from the center of the disc be avoided. Zeng et al. propose premeasuring capillary blood prior to its application to the filter paper—an approach that has been used previously (7). This is a reasonable alternative, although as we noted in our methods paper, this procedure adds processing steps at the sampling site, increases the cost of blood collection, and generates additional biohazardous waste.
Fourth, our method is criticized for its relatively high lower detection limit of 0.55 mg/L. This seems particularly high given that serum protocols typically report lower detection limits of <0.1 mg/L. However, it should be emphasized that these protocols typically require 10 to 20 uL of serum. In contrast, our method uses two 3.2 mm discs of dried whole blood, corresponding to 
3 uL of serum. Lower sensitivity is an obvious trade-off here, but we believe the benefits of the blood spot method outweigh the costs to sensitivity, particularly because iron deficiency is associated with elevated TfR, at concentrations that are an order of magnitude greater than our lower detection limit.
Because our method uses blood spots and not serum, the lower detection limit and the reference values will be significantly different from the other serum protocols summarized by Zeng et al. Their Table 1 ignores this fact, and our method therefore looks like an outlier. A more meaningful comparison would be between our method and the previously developed blood spot method of Cook et al. (7).
Lastly, Zeng et al. claim that we did not use the Bland and Altman (8) method to evaluate agreement between TfR results for matched blood spot and plasma samples. In fact, although we did not present these results graphically, we reported that "On average, plasma TfR concentrations were 0.741 mg/L higher than blood spot TfR (SD = 0.991)" [p. 3761 of (2)]. We interpreted these results as indicating an acceptable level of agreement. We also compared the results from the two methods with a scatterplot and best-fit regression line. We recognized the limits of this approach (and for that reason also pursued the Bland and Altman approach), but we presented the scatterplot and regression line to facilitate comparison with previous blood spot TfR methods (7). Lastly, Zeng et al. question the validity of our cutoff value for dried blood spot samples, but we explicitly acknowledged the limitations of this value, and noted that "a better approach would be to develop TfR reference values specific to the whole blood spot assay, as has been done for current plasma assays" [p. 3763 of (2)]. Additional analyses from a range of populations will be required to achieve this objective.
Any useful analytic method represents a reasonable compromise between the desire to maximize assay accuracy, validity, and precision, while minimizing its cost in terms of time, money, quantity of sample, and participant burden. Field-based, community-level research on iron deficiency—particularly with infants and children—places a premium on the ease of sample collection, storage, and transport, and we believe that blood spots provide a minimally-invasive tool that will facilitate such research. In our methods paper (2) we highlight both the advantages and disadvantages of measuring TfR in dried blood spots so investigators can make an informed decision regarding the appropriateness of our method for their own research goals and settings.
LITERATURE CITED
1. Shell-Duncan, B. & McDade, T. W. (2004) Use of combined measures from capillary blood to assess iron deficiency in rural Kenyan children. J. Nutr. 134:384-387.
2. McDade, T. W. & Shell-Duncan, B. (2002) Whole blood collected on filter paper provides a minimally-invasive method for assessing transferrin receptor level. J. Nutr. 132:3760-3763.
3. McDade, T. W., Burhop, J. & Dohnal, J. (2004) High sensitivity enzyme immunoassay for C reactive protein in dried blood spots. Clin. Chem. 50:652-654.
4. Shirtcliff, E. A., Granger, D. A., Schwartz, E., Curran, M. J., Booth, A. & Overman, W. H. () Assessing estradiol in biobehavioral studies using saliva and blood spots: Simple radioimmunoassay protocols, reliability, and comparative validity. Horm. Behav. 38:137-147.
5. Worthman, C. M. & Stallings, J. F. (1997) Hormone measures in finger-prick blood spot samples: New field methods for reproductive endocrinology. Am. J. Phys. Anthropol. 104:1-22.[Medline]
6. Worthman, C. M. & Stallings, J. F. (1994) Measurement of gonadotropins in dried blood spots. Clin. Chem. 40:448-453.
7. Cook, J. D., Flowers, C. H. & Skikne, B. S. () An assessment of dried blood-spot technology for identifying iron deficiency. Blood 92:1807-1813.
8. Bland, J. M. & Altman, D. G. (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307-310.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
