Illegal performance enhancement through artificially induced erythrocythemia (blood doping) has been a major issue in all endurance sports in the last three decades. It had an important impact on the development of performances in disciplines, where oxygen uptake is the limiting factor. Oxygen transport is a major determinant of oxygen uptake, especially in exercise conditions that involve large muscle groups, such as most competitive endurance sports. Oxygen transport is determined through cardiac output and the red cell volume of the organism.
From a historic point of view, the first methods that have been applied for the purpose of increasing red cell volume in sports were heterologous and autologous blood transfusions. Informal opinion suggests that these techniques were used since the early 1970s. The 1984 United States Olympic cycling team confessed the infusion of red blood cells before their Olympic competition (8). However, it is conceivable that the method of blood doping through red cell transfusion was only available to a small group of elite athletes, as the logistic requirements of this technique are high.
These premises changed with the commercial introduction of recombinant human erythropoietin (rhEPO) in 1987: The advantages of increased red cell volume and improved oxygen transport became accessible to a broad range of athletes without the logistical constraints of blood transfusions. It is unknown to what extent the illicit use of rhEPO has influenced performance in endurance events as well as in the setting of current world records. There is widespread concern that rhEPO abuse may be considerable. Physiologically, rhEPO administration provides the same benefit on performance as red cell transfusion as it results in increased Hb mass.
RhEPO became detectable in urine tests in 2000 (16) and with an increasing number of out-of-competition tests, some abusers have apparently shifted back to transfusion blood doping: The recent doping scandal in Spain of 2006, where "Operacion Puerto" led to the exclusion of several Tour de France riders because of possible involvement in doping with autologous blood transfusions, shows that the latter are again a problem in endurance sports. Although transfusions have been added to the list of forbidden substances in 1984, homologous blood transfusions can only be detected since 2002 (17), and autologous transfusions are still virtually undetectable.
Many sporting federations use concentration-based variables as [Hb] and hematocrit (Hct) as well as indirect detection models based on blood variables related to the erythroid system (3,13) as screening tools for blood manipulations but their sensitivity in uncovering blood transfusions is unknown. A recent study by Damsgaard (9) did not show any relevant changes in the indirect models described by Gore (13) but suggested that individual variations in [Hb] exceeding 15% between samples obtained shortly before any major competition would be indicative of autologous blood manipulation. As these variables are mainly based on concentrations and thus, are highly affected by fluctuations in plasma volume, they may not adequately reflect the absolute changes of Hb mass induced by blood transfusions. In this context, and intuitively, the most sensitive parameter would be determination of Hb mass itself. The recent development of an optimized CO-rebreathing method by Schmidt and Prommer (24), who modified the concept initially introduced by Burge and Skinner (6), provides a method potentially useful to screen athletes in an anti-doping setting with sufficient reliability and accuracy. CO rebreathing is much less cumbersome and time-consuming than other common blood volume techniques (e.g., 51-Cr, Evans blue) and has a benign toxicity profile in the doses used. Importantly, CO rebreathing has the lowest measurement error for Hb mass determination (90% CI, 1.4-3.5%) as evaluated in a recent meta-analysis by Gore et al (12). So far, Burge and Skinner (6) as well as Schmidt and Prommer (24) investigated the amount of Hb loss through blood donation using CO rebreathing and measured a difference in Hb mass after donation of approximately 9% (6), or of 95 g, respectively (24). However, until present, it is still unknown if Hb mass measurements are suitable to detect the absolute changes induced by blood transfusions.
The aim of this investigation was to evaluate the hypothesis that the withdrawal and reinfusion of autologous red blood cells can be detected through Hb mass measurements using the optimized CO-rebreathing technique. The results could help to improve strategies for the detection of blood doping as red cell volume is the target for all blood-doping manipulations (transfusions, rhEPO, etc.), irrespective of their technique.
This study included 10 male volunteers (mean age 28 ± 6 yr, mean height 182 ± 6 cm, mean weight 73.4 ± 7 kg). None of the subjects was a license holder in any sports discipline, so no antidoping regulations were violated. All subjects were nonsmokers and in a complete state of health as assessed by the responsible physician. The institutional ethics committee at the University of Freiburg approved the study procedures. Written informed consent was obtained from all subjects before participation in the study.
The subjects were randomly assigned to two groups. Group I (N = 6) was to donate and receive one autologous blood transfusion of packed red cells (PRC), whereas group II (N = 4) was to donate and receive two PRC. The uneven distribution of the subjects was attributable to limited availability of the apheresis donation (see below). The exact preparation method of the PRC for each group is described below. All measurements were conducted on consecutive days, as depicted in Figure 1.
FIGURE 1- Study desi...Image Tools
Each study day included the determination of Hb mass by the optimized CO-rebreathing method as described by Schmidt and Prommer (24). Venous blood samples were obtained in addition on days 2-4 and on day 6 by standard venipuncture. On day 3, the blood donation took place (see below for different PRC-extraction techniques of each group); the reinfusion followed on day 5. With only 1 d between the donation and reinfusion, it could be assumed that the changes observed can only be attributed to the donation and reinfusion. Furthermore, with this methodical approach, it is also very unlikely to induce any supraphysiologically elevated Hb mass.
Determination of hemoglobin mass and blood volume.
Briefly, the method used in our study (24) consists of a 2-min rebreathing procedure of a known CO volume (in this study, 1.20-1.40 mL·kg−1 body weight were used) with a spirometer (Spico-CO-Respirations-Applikator, Universität Bayreuth, Germany). A portable CO gas analyzer (Fluke C-220 Carbon Monoxide meter, Fluke Corp., Everett, WA) was used during the rebreathing maneuver to check the experimental setting at the mouthpiece, nose clip, and spirometer system for CO leakage. Before and at defined time points after the CO application, CO Hb concentration ([COHb], (%)) was measured in capillary samples from the hyperemized earlobe (by applying Finalgon cream, Boehringer Ingelheim, Ingelheim, Germany) using an AVL Omni Blood Gas analyzer (AVL OMNI 9, AVL Medizintechnik, Roche Diagnostics, Bad Homburg, Germany). For this study, two samples were obtained before the rebreathing process and two samples at 6 and 8 min after the start of CO rebreathing. The 6- and 8-min values were averaged for determination of the [COHb] after CO administration. Every sample was measured once. The analyzer was subject to regular internal and external quality controls. [COHb] and Hb concentration (g·dL−1) [Hb] of the blood are determined using an optical method. The accuracy of the analyzing equipment has been investigated elsewhere (29), where a precision of 0.9% and variance of 0.5% have been found. During and after the CO-rebreathing process, the test subjects were monitored for any signs of early CO toxicity such as headache, dizziness, weakness, nausea, confusion, disorientation, and visual disturbances (19). Total Hb mass was calculated from the difference in COHb before and after CO application, as outlined by Schmidt and Prommer (24). [Hb] was determined by averaging the [Hb] of all capillary measurements. At the beginning of sampling capillary blood, two samples were used for the determination of Hct (%) using microcentrifugation technique (Hettich Mikro 20, A. Hettich, Tuttlingen, Germany). The results of both measurements were averaged. All procedures were performed by the same, experienced investigators at the same time of day for any given subject and within a limited time frame (within 3 h) in relation to the blood donation and reinfusion on days 3 and 5. Although all subjects were advised to increase their fluid intake before and after the blood donation, food and fluid intake were not standardized during the study.
Blood donation and reinfusion.
One PRC was obtained in group I by standard blood bank procedures from a whole-blood sample. During the extraction procedures and after centrifugation a certain amount of Hb mass was lost in the buffy coat and leukocyte filter, which was quantified as follows:
The specific weight used for whole blood was 1.058 g·mL−1, for PRC 1.1 g·mL−1 (7). The extracted PRC were weighed and evaluated for Hb content (g); the formula used for determination of Hb content (g) in the PRC was
Equation (Uncited)Image Tools
To obtain two PRC for the subjects of group II, a special extraction method (apheresis donation) (27) was used (MCS+, Haemonetics Corp, Braintree, MA) where, in brief, whole blood is directly centrifuged and plasma instantly reinfused through the same venous needle, ensuring that subjects would lose the amount of Hb necessary for two PRC while the amount of plasma lost almost would equal that of the standard PRC preparation. In contrast to group I, the loss of Hb mass to the system is minimal and can be calculated from
Equation (Uncited)Image Tools
The determination of Hb mass in each PRC bag was analogous to those of group I.
The amount of Hb lost in the transfusion system itself was estimated to be small because it was mechanically evacuated from blood after donation and reinfusion procedures using a specific evacuation tool. The international standards of transfusion medicine were respected during donation, preparation and reinfusion of the blood samples. The PRC were stored at 4°C in temperature-regulated fridges as per German blood-banking standards.
Venous blood samples.
Measurements of [Hb], Hct, serum erythropoietin (EPO), and reticulocytes from venous blood samples have been used to establish indirect detection models as described by Gore et al. (13) for EPO abuse or by Damsgaard et al. for autologous blood transfusion (9). Use of the different models (ON-he, OFF-hre, OFF-hr) leads to scores that allow identification of conspicuous constellations during different phases of possible rhEPO administration based on cutoff scores corresponding to selected false-positive rates (see Table 2 in Gore et al. (13)).
Calculations and statistical analysis.
In addition to the formulas for Hb loss by PRC extraction, blood loss through venous blood sampling was quantified according to the mean volume lost per day (mL) and the individual [Hb] on the same day, separately for each group:
Equation (Uncited)Image Tools
Blood volume (BV), plasma volume (PV) and red cell volume (RCV) were calculated from Hct (%), Hb (g/dl), and Hb mass (g) as follows:
To evaluate the reliability of the method separately at the time points baseline, after donation and after reinfusion, the typical error of measurement (TEM) for two consecutive measurements using the optimized CO method (capillary blood sampling, 95% confidence intervals (CI)) was calculated as proposed by Hopkins (15). This method has been used recently in a meta-analysis (12) to determine reliability of various approaches to measure Hb mass (meta-analysis: 90% CI) and evaluates the reliability of the measurements independent of the mean. The standard error of measurement (SEM) was calculated as
Equation (Uncited)Image Tools
for the measurements of each athlete to determine the within-subject variation for the consecutive measurements at each time point. Differences between the three time points were evaluated using matched-pairs analysis (paired t-tests), separately for each group. A Bland-Altman analysis (4) was used to compare the calculated with the measured Hb mass for significant differences at two occasions (after donation vs baseline and after reinfusion vs after donation).
Data are presented as means ± standard deviations (SD) unless otherwise specified. All error calculations are displayed as absolute values (g) and percentages of mean Hb mass (%) at the given time points. P values are indicated with the data; P < 0.05 was considered to represent statistical significance. Statistical analysis was performed using JMP version 5.1 (SAS Institute Inc. Cary, NC).
The CO-rebreathing procedures were tolerated well, none of the subjects showed any signs of CO toxicity. [COHb] increased from 0.9 ± 0.3% before CO administration to 8.1 ± 0.9% after CO administration. No CO leaks were detected at the spirometer, connected mouthpiece, or nose clip during any of the investigations. In addition, the donation of whole-blood samples and subsequent reinfusion of PRC did not cause any side effects. None of the subjects or measurements had to be excluded from further analysis.
Hb mass loss.
During the preparation of whole blood, Hb mass was lost in the buffy coat after centrifugation (mean 12.3 ± 1.2 g) and in the leukocyte filter (mean 5 ± 0 g) in group I. On the other hand, the Hb loss by PRC preparation in group II was minimal and was calculated to be 1 ± 0 g for every subject. The mean Hb content of the PRC in group I was 59.0 ± 3.9 g (one PRC) and 108.3 ± 1.3 g in group II (two PRC).
The volume lost through venous blood samples was the same in both groups and was estimated to be 86 mL per subject, thus resulting in a total mean Hb mass loss of 14.2 ± 1.0 g Hb for each subject (11.4 ± 1.0 g Hb after donation; additional 2.8 ± 0.1 g after reinfusion).
Effect of blood donation and reinfusion on Hb mass.
Hb mass changed significantly after blood donation in both groups and returned almost to baseline levels after reinfusion of the PRC. All Hb mass results (including individual values and errors of measurement) are displayed in Table 1, and the individual alterations also are depicted in Figure 2.
The measured mean decrease of Hb mass after donation was 89 ± 16 g in group I and 120 ± 14 g in group II. After the reinfusion, mean Hb mass increased significantly compared with after donation (group I: 70 ± 16 g; group II: 90 ± 9 g) but was slightly lower than at baseline (group I: −19 ± 17 g; group II: −30 ± 14 g). The TEM calculated from two consecutive measurements for each time point is displayed in Table 1 and ranged between 0.8% (6.4 g of Hb) and 3.1% (22.1 g of Hb) in group II for the after donation measurements. The individual differences between measured and calculated (expected) Hb mass changes are presented in Table 2. It was observed that Hb mass determination lead to under- as well as overestimation of the calculated (expected) Hb mass change. A Bland-Altman test did not reveal significant differences between measured and calculated Hb mass when comparing after donation versus baseline and after reinfusion versus after donation in both groups.
Changes in vascular compartments.
The changes of the vascular compartments are shown in Table 3. When summarized to the three time points, the greatest changes were observed in RCV, which was 13% lower after donation than at baseline in group I, and 16% lower in group II. RCV was lower after reinfusion compared with baseline in both groups (group I, −4%; group II, −6%), although the changes were only significant in group I. In contrast to RCV, PV was not significantly decreased after donation in neither of the groups. In group I, PV was significantly higher after reinfusion than after donation (+11%), and at baseline (+9%), no significant changes of PV were observed in group II. In summary, in both groups the alterations in BV mainly resulted from the changes in RCV.
Venous parameters and indirect detection models.
The results are summarized in Table 3. [Hb] and Hct decreased significantly in both groups after blood donation. The serum level of EPO had a significant mean 2.4-fold increase in group I and a 2.5-fold increase in group II. At the same time, reticulocytes did not change after donation in group I and increased in group II.
After reinfusion, [Hb] increased by 8.5% in group I and 9.6% in group II compared with after donation, and the indirect parameters almost returned to baseline levels in group I, whereas group II showed significantly lower levels of [Hb] and Hct after reinfusion compared with baseline. Additionally, the reticulocyte count was significantly higher after reinfusion than at baseline in group II. EPO serum levels did not reveal significant differences after reinfusion compared with baseline levels. The highest scores of the indirect models observed were those at baseline; for instance, OFF-hr at baseline was 100.2 in group I and 101.6 in group II. In summary, none of the scores exceeded the thresholds suggested for the indirect detection of blood manipulations.
The major finding of this study is that the withdrawal of whole blood and reinfusion of autologous PRC are detectable using the optimized CO-rebreathing method, because Hb mass changed significantly after these interventions.
Although detection methods exist for doping with rhEPO (16) and homologous blood transfusions (17), autologous blood transfusions remain virtually undetectable. The aforementioned practices target Hb mass as the key variable to improve performance. Measurement of this parameter, as performed in this study with the optimized CO-rebreathing method, would therefore allow monitoring of absolute changes triggered by doping techniques. In this context, two conditions have to be met. On the one hand, Hb mass has to be relatively stable under physiological circumstances so that conspicuous alterations can be ascribed to doping practices. On the other hand, the method for determination of Hb mass must be reliable, with a small error of measurements, and adequately reflect the absolute changes induced by blood manipulations.
Hb mass is altered by endurance training and physical performance, as can be concluded from the fact that endurance-trained subjects usually have an Hb mass of 12 g·kg−1 or above compared with untrained controls, who have approximately 8 g·kg−1 (14,21,23). In contrast, it has been shown that Hb mass is not subject to longitudinal variations in elite athletes who have a history of long-term training, especially those in a professional set-up. In this context, Gore and coworkers (11) have reported that Hb mass did not change after 12 wk of intense rowing training or 4 wk of heat training (32°C). Thus, it can be assumed that, if at all, changes in Hb mass develop over several months of specific training rather than weeks. Other influences such as altitude exposure could potentially lead to variations in Hb mass (2,5,28). Furthermore, it has been shown that athletes originating from altitude have a higher Hb mass than athletes from sea level (23), though we speculate that Hb mass individually is not subject to acute physiological changes in those athletes, analogous to sea level-born athletes.
To evaluate whether the optimized CO-rebreathing method adequately reflects the absolute changes induced by blood transfusion, quantitative assessment has to be performed for validation. Burge and Skinner (6) and Schmidt and Prommer (24) have validated their experiments comparing calculated and measured Hb mass loss induced by phlebotomy. Whereas Burge and Skinner state that measured Hb mass 1 wk after venesection was not different from the calculated estimate (although they do not provide absolute values of the difference of measured and calculated Hb mass), Schmidt and Prommer have calculated the amount of Hb loss according to (550 mL (donation volume) × [Hb]) and have described a mean difference between measured and calculated Hb values of +9 g. The present study has confirmed these results and also has used a reinfusion setting as an additional means of assessing the optimized CO method accuracy and precision when using capillary blood samples. Furthermore, a very thorough approach was chosen to quantify the amount of blood lost by donations, additional venous sampling and the PRC preparation process. In this context, the net and gross sums of Hb mass after donation and after reinfusion were determined. The measurements of Hb mass after donation showed the greatest agreement between calculated (expected) and measured values (see Table 2). From an individual perspective, underestimation as well as overestimation of "true" Hb mass changes by the optimized CO-rebreathing method were observed. This variation was found to be within the limits of the errors of measurement (see Tables 1 and 2). Interestingly, Hb mass was found to be significantly lower after reinfusion than at baseline in both groups. This shows that the optimized CO-rebreathing method has sufficient validity, allowing conclusive results with respect to the Hb mass additionally lost by venous sampling and PRC preparation. Another potential factor in this context might be that some of the Hb did not survive or was not viable after reinfusion.
The reliability expressed by the typical error (15) of 0.8-3.1% for each pair of measurements (capillary sampling, 95% CI) is in line with the previously published reliability measures of Schmidt and Prommer (24) of 1.9-2.0% (venous sampling, 95% CI) and the values described in the meta-analysis by Gore et al. (12) between 1.4 and 3.5% (CO rebreathing as described by Burge and Skinner (6), 90% CI).
The administered units of PRC contained, on average, either 59 g (one unit) or 108 g (two units) of Hb. On the basis of the TEM of 0.8-3.1% (12-43 g of Hb mass (95% CI)), the infusion of one unit of blood might be problematic to detect in an individual if only one baseline measurement is available. If the TEM is small (e.g., 0.8% (smallest TEM in our study)), the likelihood of reliable detection is high. On the other hand, the amount of Hb in two PRC distinctly exceeds the above mentioned range and therefore, in an antidoping setting, the infusion of two units would trigger changes in Hb mass that would allow identification by the current method. In this context, previous literature suggests that an increase in aerobic capacity (e.g., an improvement of V˙O2max by 3-5%) is only achieved if the equivalent of two units of blood are infused (see (10) for a concise review of the subject). Additionally, lay press articles and current informal opinion report that the amount of blood misused for doping practices is believed to be about 800 mL. Nevertheless, it seems feasible that abusers infuse only one PRC, which might also enhance physical performance with a smaller, but nevertheless profitable, effect that would be difficult, if not impossible, to assess using the optimized CO-rebreathing method.
A possible strategy to reveal autologous blood doping could be the long-term screening of Hb mass and calculations of the 95% confidence and 95% tolerance intervals, especially because a single baseline measurement might not be suited to reveal autologous transfusion with the next Hb mass determination. Suspicious alterations could be defined as values that exceed those intervals. In any case, Parisotto et al. (18) have shown that rhEPO administration significantly increased Hb mass by 6.9-12%, corresponding to a mean Hb mass of about 58-107 g. Thus, the same approach of long-term Hb mass screening may unmask other blood-doping procedures such as homologous transfusion and administration of rhEPO for which further detection methods exist (16,17). Because it is possible to maintain elevated [Hb] with microdose rhEPO administration after high-dose EPO injections, which reduces the window of urine detection to as little as 12-18 h (1), irregularities might be revealed by repeated Hb mass measurements.
It may seem to be a limitation that the time span between donation and reinfusion of 2 d does not reflect possible doping practice, where that range is most likely 4-6 wk. However, the aim of this study was to determine whether the absolute changes in Hb mass caused by blood donation and reinfusion might be identified using the optimized CO-rebreathing method regardless of the pre-reinfusion values.
As a final point, practical approaches to detect autologous transfusion by indirect indicators are discussed. First models were already published in 1987 by Berglund et al. (3), and further strategies were recently described by Damsgaard et al. (9). Berglund et al. (3) conclude that, after reinfusion of 1350 mL of blood 4 wk after withdrawal, a combination of measurements of Hb, bilirubin, iron, and erythropoietin in serum could detect 50% of the blood-doped athletes by a single test sample during the first week after reinfusion. In the study of Damsgaard et al. (9), withdrawal and reinfusion included a partly larger quantity of blood compared with that in our study, which triggered [Hb] changes above normal variation (26), suggesting autologous blood manipulations if variation in [Hb] exceeds 15% between samples obtained shortly before any major competition. The results of our study are in accordance with those investigations: [Hb] increased after reinfusion versus after donation by 8.5% in group I and 9.6% in group II compared with a 14% increase after reinfusion in Berglund et al.'s work. Although rapid alterations in EPO serum levels could be demonstrated after donation and after reinfusion (2.4- to 2.5-fold change) in our investigation, the study period was probably too short to induce changes in the reticulocyte count. Additionally, our results were in line with Damsgaard et al. in reference to the OFF-hr model by Gore et al. (13), because none of the subjects demonstrated positive OFF-hr scores. However, a critical appraisal of these presently used variables is necessary. The main weakness is the fact that they are measures based on concentrations and, thus, are highly affected by fluctuations in plasma volume, such as those induced by altitude, posture, hydration status, and exercise (20,22,25). For most erythropoietic manipulations, [Hb] and Hct return to baseline after the vascular compartments (plasma volume) have adapted to newly increased red cell volume. Therefore, this approach based on concentration related variables probably yields a high number of false-negative blood tests, and determination of RCV or Hb mass themselves would be the most sensitive parameters in this context.
In summary, this investigation supports the potential of Hb mass measurement as a screening tool for blood doping, with certain limitations. Changes in Hb mass induced by artificial blood loss and subsequent reinfusion can be detected using the optimized CO-rebreathing method. In this context, assessment of method precision and knowledge of the physiological variation are of utmost importance. Especially in longitudinal settings, Hb mass measurements could unveil suspicious constellations. Further studies are needed to assess the individual variability of Hb mass and the impact of other biological factors, such as age and training status, to improve the differentiation between artificially induced variation and physiological variability, as well as detection of supraphysiological elevation in Hb mass.
The authors are indebted to the subjects who participated in the study. The authors do not have a professional relationship with companies or manufacturers that may benefit from the results of this study.
1. Ashenden, M., E. Varlet-Marie, F. Lasne, and M. Audran. The effects of microdose recombinant human erythropoietin regimens in athletes. Haematologica
2. Ashenden, M. J., C. J. Gore, G. P. Dobson, and A. G. Hahn. "Live high, train low" does not change the total haemoglobin mass of male endurance athletes sleeping at a simulated altitude of 3000 m for 23 nights. Eur. J. Appl. Physiol. Occup. Physiol.
3. Berglund, B., P. Hemmingsson, and G. Birgegard. Detection of autologous blood transfusions in cross-country skiers. Int. J. Sports Med.
4. Bland, J. M., and D. G. Altman. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet
5. Brugniaux, J. V., L. Schmitt, P. Robach, et al. Eighteen days of "living high, training low" stimulate erythropoiesis and enhance aerobic performance in elite middle-distance runners. J. Appl. Physiol.
6. Burge, C. M., and S. L. Skinner. Determination of hemoglobin mass and blood volume with CO: evaluation and application of a method. J. Appl. Physiol.
7. Council of Europe Publishing. Guide to the Preparation, Use and Quality Assurance of Blood Components
. 11th ed. Strasbourg, France: Council of Europe Publishing, 2005.
8. Cramer, R. B. Olympic cheating: the inside story of illicit doping and the U. S. cycling team. Rolling Stone
9. Damsgaard, R., T. Munch, J. Morkeberg, S. P. Mortensen, and J. Gonzalez-Alonso. Effects of blood withdrawal and reinfusion on biomarkers of erythropoiesis in humans: implications for anti-doping strategies. Haematologica
10. Gledhill, N. Blood doping and related issues: a brief review. Med. Sci. Sports Exerc.
11. Gore, C. J., A. G. Hahn, C. M. Burge, and R. D. Telford. VO2max and haemoglobin mass of trained athletes during high intensity training. Int. J. Sports Med.
12. Gore, C. J., W. G. Hopkins, and C. M. Burge. Errors of measurement for blood volume parameters: a meta-analysis. J. Appl. Physiol.
13. Gore, C. J., R. Parisotto, M. J. Ashenden, et al. Second-generation blood tests to detect erythropoietin abuse by athletes. Haematologica
14. Heinicke, K., B. Wolfarth, P. Winchenbach, et al. Blood volume and hemoglobin mass in elite athletes of different disciplines. Int. J. Sports Med.
15. Hopkins, W. G. Measures of reliability in sports medicine and science. Sports Med.
16. Lasne, F., and J. De Ceaurriz. Recombinant erythropoietin in urine. Nature
17. Nelson, M., H. Popp, K. Sharpe, and M. Ashenden. Proof of homologous blood transfusion through quantification of blood group antigens. Haematologica
18. Parisotto, R., C. J. Gore, K. R. Emslie, et al. A novel method utilising markers of altered erythropoiesis for the detection of recombinant human erythropoietin abuse in athletes. Haematologica
19. Raub, J. A., M. Mathieu-Nolf, N. B. Hampson, and S. R. Thom. Carbon monoxide poisoning-a public health perspective. Toxicology
20. Sawka, M. N., V. A. Convertino, E. R. Eichner, S. M. Schnieder, and A. J. Young. Blood volume: importance and adaptations to exercise training, environmental stresses, and trauma/sickness. Med. Sci. Sports Exerc.
21. Sawka, M. N., A. J. Young, K. B. Pandolf, R. C. Dennis, and C. R. Valeri. Erythrocyte, plasma, and blood volume of healthy young men. Med. Sci. Sports Exerc.
22. Schmidt, W., B. Biermann, P. Winchenbach, et al. How valid is the determination of hematocrit values to detect blood manipulations? Int. J. Sports Med.
23. Schmidt, W., K. Heinicke, J. Rojas, et al. Blood volume and hemoglobin mass in endurance athletes from moderate altitude. Med. Sci. Sports Exerc.
24. Schmidt, W., and N. Prommer. The optimised CO-rebreathing method: a new tool to determine total haemoglobin mass routinely. Eur. J. Appl. Physiol.
25. Schumacher, Y. O., J. Temme, D. Bueltermann, A. Schmid, and A. Berg. The influence of exercise on serum markers of altered erythropoiesis and the indirect detection models of recombinant human erythropoietin abuse in athletes. Haematologica
88: 712-714, 2003.
26. Sharpe, K., W. Hopkins, K. R. Emslie, et al. Development of reference ranges in elite athletes for markers of altered erythropoiesis. Haematologica
27. Shi, P. A., and P. M. Ness. Two-unit red cell apheresis and its potential advantages over traditional whole-blood donation. Transfusion
28. Wehrlin, J. P., P. Zuest, J. Hallen, and B. Marti. Live high-train low for 24 days increases hemoglobin mass and red cell volume in elite endurance athletes. J. Appl. Physiol.
29. Zander, R., and W. Schaffartzik. Haem oxymeter. Qualitest