| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Columbia University Department of Epidemiology Mailman School of Public Health New York, NY
aAddress correspondence to this author at: Cancer Epidemiology, Columbia University, Mailman School of Public Health, 722 W. 168th St., 7th Floor, New York, NY 10032. Fax 212-305-9413; e-mail LBL10{at}columbia.edu.
To the Editor:
The concentration of C-reactive protein (CRP) in blood is considered a marker of inflammation; it is also predictive of cardiovascular disease and possibly cancer (1)(2). Although CRP is frequently measured in large studies, the effects of storage conditions and delays in processing on those measurements are largely unknown. In whole blood samples stored at ambient temperature, CRP has been reported to be stable for 6 h (3) and for 3 days(4), and in samples stored on ice it has been reported as stable for 36 h (5). However, in multicenter studies for which analyses are conducted at a single site, specimens are often shipped overnight. It is possible to ship specimens under refrigerated or frozen conditions, but doing so adds to the cost of the study. We therefore assessed the stability of high-sensitivity CRP (hsCRP) in unprocessed whole blood samples shipped overnight at ambient temperature.
This study was part of a larger study of changes in immune and inflammatory markers and quality of life among people staying at an institute that teaches about a raw vegan diet (Hippocrates Health Institute, W. Palm Beach, FL). The protocol of our study was approved by the Institutional Review Board at Columbia University, and informed consent was obtained from all participants before blood sample collection. We obtained 2 fasting blood specimens from each of 21 volunteers. Blood from each participant was collected into two 4-mL serum separator tubes (Becton Dickinson). One specimen from each participant underwent immediate processing. Within 3 h of collection, these specimens were centrifuged at 1200g for 10 min, and the serum was then poured into a microcentrifuge tube. The other specimen from each participant underwent delayed processing. These specimens were shipped overnight at ambient temperature and were centrifuged 
28 h after collection at 1510g for 10 min. The serum was then poured into microcentrifuge tubes. All serum specimens were then stored at –80 °C. The hsCRP concentrations in the immediate and delayed processing samples were measured simultaneously, in duplicate, with a standardized ELISA, according to the manufacturer’s directions (Life Diagnostics, Inc.; detection limit, 0.1 mg/L). Absorbance was read at 450 nm with a microtiter well reader. hsCRP values <0.1 mg/L were recorded as 0.09 mg/L.
The results did not follow a gaussian distribution; we therefore used nonparametric statistical methods (SPSS, Ver. 10.0.5). We log-transformed the data to determine the intraclass correlation coefficient. A two-sided P value <0.05 was considered significant.
As is evident in Fig. 1
, hsCRP concentrations were increased with delayed processing in 67% of the participants. The greatest variability occurred with values <2.5 mg/L: in a comparison of immediate vs delayed processing for hsCRP values >2.5 mg/L, 6 of 8 differed by <18%. The median (interquartile range) hsCRP concentration was 1.63 (0.55–4.33) mg/L in the group of specimens processed immediately and 1.49 (0.69–4.82) mg/L in the delayed processing group, a 9% decrease in the latter group. A Wilcoxon signed-ranks test showed a significant difference between immediate and delayed processing (P = 0.037); however, the Spearman correlation coefficient was 0.97 (P <0.01) and the intraclass correlation coefficient was 0.94 (P <0.01).
|
In summary, the correlation between hsCRP for immediate and delayed processing was strong; however, most values were higher after delayed processing. The lower median hsCRP with delayed processing likely reflects the relatively wide variation in hsCRP results between processing procedures when the values are low. The imprecision associated with hsCRP processing, 3.5%–7.0% (4), is smaller than the 9% difference found in this study. In one of the previous studies, mean serum CRP concentrations in 15 specimens analyzed after a 6-h delay in processing were 6% lower than those in the samples processed immediately (3). In another study with 5 participants, the concentrations in EDTA–whole blood specimens processed after a 3-day delay differed by <10%, with no mention of direction (4). In a third study of 17 participants, mean concentrations in heparinized specimens stored on ice were 2% higher after a 36-h delay in processing (5).
In conclusion, when clotted whole blood samples are shipped at ambient temperature and serum processing is delayed for 28 h, there may be small changes in hsCRP concentrations measured with ELISA, particularly with lower values. These results should be confirmed in a larger sample and for hsCRP measured with automated methods.
Acknowledgments
Support for this work was provided by a National Cancer Institute Grant (5 R25 CA094061). The Biomarkers Core Facility, supported by P30 ES09089 and P30 CA013696, assisted in blood processing and the ELISA.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. Hartweg, M. Gunter, R. Perera, A. Farmer, C. Cull, C. Schalkwijk, A. Kok, H. Twaalfhoven, R. Holman, and A. Neil Stability of Soluble Adhesion Molecules, Selectins, and C-Reactive Protein at Various Temperatures: Implications for Epidemiological and Large-Scale Clinical Studies Clin. Chem., October 1, 2007; 53(10): 1858 - 1860. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
