Purpose: This investigation quantified the effect of changes in plasma osmolality on the measurement of hematocrit (Hct) and the implications for the subsequent use of these data to calculate changes in plasma volume and application to the World Anti-Doping Agency Athlete Biological Passport.
Methods: Two groups of eight male volunteers visited the laboratory after an overnight fast. In study 1, a 20-mL blood sample was collected and aliquoted into collection tubes containing saline of varying concentrations to alter the sample osmolality. In study 2, plasma osmolality was manipulated in vivo through prolonged exercise. Samples were analyzed for hemoglobin concentration and Hct using manual methods and using an automated hematology analyzer (AHA).
Results: Changes in blood, plasma, and red cell volumes were calculated. Although AHA Hct values did not change (P = 0.652), spun packed cell volume fell progressively as the osmolality of the sample increased (P < 0.001, study 1). Consequently, there was a significant increase in apparent plasma volume as osmolality increased (P < 0.001): regression analysis revealed that a 10 mOsm·kg−1 change in plasma osmolality produced a difference of 0.8 Hct units and a 1.6% change in plasma volume. In study 2, exercise produced a 12 ± 3 mOsm·kg−1 increase in plasma osmolality. No difference in Hct was apparent at rest (P = 0.659), but spun packed cell volume was 1.0 ± 0.9 Hct units lower during exercise compared with AHA data (P < 0.001). There was a difference in the degree of plasma volume change calculated, with a reduction of 8.7% ± 3.4% and 11.3% ± 3.5% reported with the manual and AHA methods, respectively (P = 0.002).
Conclusions: Conditions or interventions that result in a marked change in plasma osmolality produce a discrepancy in Hct measured using an AHA, consequently introducing errors into any calculation of changes in plasma volume using these data. These findings may also have implications for the measurement of Hct by World Anti-Doping Agency-accredited laboratories.
School of Sport, Exercise and Health Sciences, Loughborough University, Leicestershire, England, UNITED KINGDOM
Address for correspondence: Ronald J. Maughan, Ph.D., School of Sport, Exercise and Health Sciences, Loughborough University, Leicestershire LE11 3TU, England, United Kingdom; E-mail: firstname.lastname@example.org.
Submitted for publication February 2013.
Accepted for publication June 2013.
Direct measurement of blood and plasma volumes can be made using a variety of dilution techniques, typically requiring the serial measurement of a tracer substance after allowing sufficient time for uniform distribution throughout the compartment. Such techniques often require access to labeling facilities and are not amenable to repeated measurements over a relatively short time frame. For these reasons, it has become common to measure changes in hemoglobin (Hb) concentration as an index of blood volume changes. Hb is present within the erythrocyte and does not enter or leave the circulation in appreciable quantities over the time scale of most human physiology experiments, so a change in Hb concentration can be used to calculate changes in blood volume. Dye dilution methods for plasma volume measurement are also limited because of the impossibility of making repeated measurements on a short time scale. Changes in plasma volume can be calculated, however, if the Hb concentration and hematocrit (Hct) are known. The Hct reflects the volume of red blood cells as a fraction of the total volume of a sample, usually expressed as a percentage. The possibility of using these measures has long been recognized, but Dill and Costill (4) published an improved method for the calculation of changes in blood, red cell, and plasma volumes from serial measurements of Hb concentration and Hct. Because a wide variety of exercise, environmental, nutritional, and pharmacological interventions produces marked changes in blood and plasma volumes, this method remains in common use today.
The introduction of automated hematology analyzers (AHA) has allowed the rapid analysis of both Hb and Hct, along with several other hematological measures of clinical interest. These instruments typically require only a small volume of blood and offer results within a short time frame, making them attractive in a variety of clinical and research settings. Typically, AHA uses a spectrophotometric technique to determine Hb concentration similar to that used in manual procedures (i.e., cyanmethemoglobin assay), but the analytical method used by these instruments to determine Hct differs markedly from that used in conventional laboratory settings. Before the introduction of AHA, Hct was determined by centrifugation of anticoagulated blood in a capillary tube, with the height of the column of erythrocytes compared with the total height of the column of blood in the tube. This is called the Hct or spun packed cell volume (PCV), with an appropriate correction typically applied to account for plasma trapped between the red cells (3). Most AHA determines Hct by measuring the electrical impedance of a sample of cells suspended in a conductive reagent. Because red blood cells have a lipid membrane, they account for the majority of the electrical resistivity of the whole blood. The impedance variation generated by the passage of cells through a small aperture is used to determine the number and the size of the red blood cells. A value for the Hct is calculated from these two measures.
The volume of each individual red blood cell is highly responsive to changes in extracellular osmolality (13). In individuals who are acutely hyperosmotic, red cell volume will rapidly and reversibly decrease because of free water movement out of the cells to balance intracellular and extracellular osmotic force. Stott et al. (16) previously demonstrated this expected inverse relationship between the osmolality of a sample and the spun PCV, but this response was not apparent in Hct values determined by an AHA method. These differences are explained by the sample handling procedure used by most commercially available AHA (e.g., COULTER® AC·T™ 5diff (Beckman Coulter, High Wycombe, UK) or Sysmex XT-2000i (Sysmex Corp., Kobe, Japan)); samples of whole blood are subjected to a series of dilution steps in relatively large volumes of “isotonic” reagent, which effectively normalizes the sample osmolality and results in the measurement of an “isotonic Hct” (16), which may be very different from the in vivo osmolality. These Hct measurements will not reflect the true circulating red cell volume present in vivo if the plasma osmolality is markedly different from that of the “isotonic” diluent. Calculated changes in the plasma volume made using values derived from AHA will therefore also contain an unknown degree of error if plasma osmolality has changed during an experiment.
The Athlete Biological Passport (ABP) was introduced by the World Anti-Doping Agency (WADA) in 2009. The fundamental principle of the ABP is based on the monitoring of selected biological parameters over time that will indirectly reveal the effects of doping rather than detect a particular doping substance or method. Unexplained or abnormal changes in passport data can be used to identify and target athletes for specific analytical testing, as well as can be used directly to pursue antidoping rule violations. The hematological module of the ABP, which is used to detect the use of prohibited methods to enhance oxygen transfer, relies on the measurement of the following parameters: Hct, Hb, red blood cells count, percentage of reticulocytes, reticulocyte count, mean corpuscular volume, mean corpuscular Hb, and the mean corpuscular Hb concentration. Hct is used along with these other measures to determine an “Abnormal Blood Profile Score” (15). In an effort to standardize analytical results, WADA-accredited laboratories are required to use analyzers with comparable technical characteristics, and these instruments are subject to regular internal and external quality assessment tests (WADA Technical Document TD2010BAR 2011). Antidoping laboratories use a commer cially available AHA (typically, Sysmex XT-2000i) to produce these data, and consequently, ABP samples are subject to the analytical artifacts described previously. Because presample conditions are not standardized in doping control, distinct variation in sample osmolality is likely, and this may limit the reliability of the ABP to detect the use of prohibited methods to enhance oxygen transfer.
Many laboratories routinely measure Hb and Hct using either traditional assay methods (Hb by cyanmethemoglobin method and Hct by spun PCV) or an AHA. The aim of the present study, therefore, was to determine the influence of changes in plasma osmolality on the measurement of Hb and Hct using these different analytical approaches. Given the importance of Hct to the equations described by Dill and Costill (4), any discrepancy between these measures is likely to introduce error into the calculation of changes in red cell and plasma volumes. In addition, Hct is monitored by antidoping authorities using an AHA-based approach as part of the ABP. Because many exercise, nutritional, and pharmacological interventions result in marked changes in plasma osmolality, this will likely invalidate the use of AHA data to determine changes in plasma volume and may have significant implications for reliability of the ABP. If Hct values obtained from hematology analyzers are influenced by osmolality in a predictable manner, then it should be possible to apply a correction factor to these data to account for the use of an “isotonic Hct” produced by AHA devices. These objectives were addressed through the completion of two separate, but interrelated, studies. First, the effect of changes in sample osmolality on measured Hb and Hct was directly determined through the addition of known concentrations of saline to blood samples drawn at rest (study 1). To illustrate this response in vivo, a separate study examined Hb and Hct measured during prolonged exercise in the heat, an intervention where a significant change in sample osmolality was expected (study 2).
Two separate studies were performed. First, the effect of manipulating sample osmolality on measured Hb and Hct was determined. The influence of these data on calculated changes in blood, red cell, and plasma volumes was then assessed using manual and AHA methods (study 1). A separate study examined the effect of a change in in vivo osmolality, induced by exercise in the heat, on measurements of Hct and calculated changes in plasma volume using both manual and AHA methods (study 2). Before volunteering, all participants received written information regarding the nature and purpose of the study. After an opportunity to ask any questions, a written statement of consent was signed. The experimental protocol was approved by the local Ethical Advisory Committee (ref. R11-P138).
Eight healthy men (age, 27 ± 4 yr; height, 1.75 ± 0.06 m; body mass, 73.9 ± 5.4 kg) volunteered to participate in this study. Each volunteer was asked to visit the laboratory on two separate occasions separated by at least 7 d. The experimental trials were undertaken in the morning after an overnight fast. On both occasions, volunteers were seated in a comfortable environment for 15 min before a 20-mL blood sample was collected into a dry syringe from a superficial antecubital vein by venipuncture.
An aliquot (10 mL) of the whole blood was immediately added to a tube containing K2EDTA (1.5 mg·mL−1) (6,10) and mixed thoroughly by gently inverting at least 10 times by hand. A series of samples, each of 1 mL volume, were then dispensed into a series of collection tubes containing 0.100 mL of saline of known concentrations in increments between 150 and 600 mOsm·kg−1 (Osmomat auto; Gonotec; Farnborough, UK). These tubes were mixed gently before being analyzed for Hb concentration and Hct using the methods described in the succeeding part of this article. After completion of the hematology analysis, samples were centrifuged to yield plasma that was used to determine osmolality. The remaining 10 mL of whole blood was added to the same range of collection tubes containing saline, omitting the anticoagulant step. These samples were left to clot at room temperature for 60 min before being centrifuged to yield serum, which was used to determine serum osmolality.
Eight physically active men (age, 26 ± 4 yr; height, 1.79 ± 0.07 m; body mass, 76.3 ± 10.2 kg; V˙O2max, 55.7 ± 4.0 mL·kg−1·min−1) were recruited to take part in the second part of this investigation. Trials took place in the morning after an overnight fast. Volunteers sat in a comfortable environment (22°C–24°C) for 20 min. One hand was immersed in warm water (42°C) for 10 min before a 21-gauge cannula was introduced into a superficial forearm vein to allow the collection of serial blood samples. To ensure the cannula remained patent, it was flushed with a small volume of heparinized saline immediately after each collection. A baseline sample (10 mL) was drawn at the end of the rest period. Volunteers then entered a climatic chamber maintained at a temperature of 30.0°C ± 0.2°C and a relative humidity of 50% ± 1%, and commenced cycle exercise at a power output corresponding to 60% V˙O2max until volitional exhaustion. Blood samples (10 mL) were drawn after 15 and 30 min as well as at the point of exhaustion.
Blood samples collected throughout the experimental protocol were drawn into dry syringes, with 1 mL dispensed into a tube containing K2EDTA (1.5 mg·mL−1). These tubes were mixed thoroughly by gently inverting at least 10 times by hand before being analyzed for Hb concentration and Hct using the methods described in the succeeding part of this article. On completion of the hematology analysis, samples were centrifuged to yield plasma that was used to determine osmolality.
Sample analysis and calculations
All hematological analyses were completed within 3 h of sample collection as recommended by Kennedy et al. (7). Before analysis, samples were placed on a rolling mixer (Denley Spiramix 5; Denley, Sussex, UK). Blood samples were first analyzed in duplicate for Hb concentration using the cyanmethemoglobin method (Sigma-Aldrich, St. Louis, MO). Spun PCV was determined in triplicate by drawing blood into plain borosilicate glass capillary tubes, one end of which was then sealed with putty. Tubes were placed into numbered slots of an Hct centrifuge and centrifuged for 5 min at 14,000g (HaematoSpin 1400; Hawksley, Sussex, UK). Spun PCV values were adjusted downward by 1% to account for plasma trapped between the red cells (3). Hb and Hct were also determined in duplicate using an AHA (COULTER® AC·T™ 5diff, Beckman Coulter). The cell counter was calibrated before use as described by the manufacturer’s guidelines, and calibration was regularly checked using quality control samples. The osmolality of plasma and serum samples was determined in duplicate by freezing point depression (Gonotec Osmomat auto).
Changes in blood, plasma, and red cell volumes were calculated as described by Dill and Costill (4). In all cases, these data were generated using Hb and Hct data either from the AHA (labeled “AHA”) or from values obtained from the manual methods described previously (labeled “Manual”). In study 1, the values were calculated relative to the midpoint of the data range (i.e., 300 mOsm·kg−1). In study 2, all changes were calculated relative to the preexercise sample.
The intraassay coefficients of variation (n = 20) for the methods used were cell counter Hb (1.0%), cell counter Hct (1.0%), spun PCV (0.5%), and cyanmethemoglobin assay Hb (0.6%).
Data are presented as mean ± SD unless otherwise stated. The range of values is presented where this information was deemed to be relevant. To identify differences in normally distributed results, two-way (analytical method-by-osmolality) repeated-measures ANOVA was used. Where a significant interaction was apparent, pairwise differences were evaluated using the Tukey post hoc procedure. Relations between variables were examined using the Pearson correlation. The effects of sample osmolality on measured Hct and the calculated change in blood and plasma volumes were assessed by fitting linear models and examining the significance of fitted terms in these models via regression. Statistical significance was accepted at P < 0.05.
The serum osmolality of the samples when drawn was 286 ± 2 mOsm·kg−1 (range, 280–289 mOsm·kg−1). The addition of 1.5 mg·mL−1 K2EDTA resulted in a 10 ± 1 mOsm·kg−1 increase in sample osmolality, and the addition of the samples to saline resulted in an extracellular osmolality range between 263 ± 4 and 335 ± 3 mOsm·kg−1.
Hb concentration of the undiluted samples was 15.7 ± 0.9 g·dL−1 (range, 14.8–17.7 g·dL−1) and 15.5 ± 1.0 g·dL−1 (range, 14.1–17.1 g·dL−1) measured using the cyanmethemoglobin method and the AHA, respectively. There was a strong relation between the Hb concentrations obtained using the manual assay and those determined by the cell counter (r = 0.861, P < 0.001). Manipulation of plasma osmolality did not influence Hb measurements made using the cyanmethemoglobin method (P = 0.339) or the AHA (P = 0.406).
Hct of the undiluted samples was 45.9% ± 2.9% (range, 41.5%–49.5%) and 45.6% ± 2.3% (range, 41.5%–49.5%) measured using the spun PCV and AHA methods, respectively. Hct values from the AHA were not influenced by a change in plasma osmolality (P = 0.652, Fig. 1A), but spun PCV was progressively lower as the osmolality of the sample increased from 150 to 600 mOsm·kg−1 (P < 0.001, Fig. 1B). Linear regression analysis was then applied to predict the change in Hct that would result from a known change in plasma osmolality. The change in plasma osmolality significantly predicted the change in sample Hct (B = −0.078, t(112) = −6.2, P < 0.001), with the results indicating that this explained 26% of the variance (R2 = 0.258, F(1,112) = 38.2, P < 0.001). Each 10 mOsm·kg−1 change in plasma osmolality will result in a 0.8-unit difference in Hct measured by spun PCV; when the plasma is hypertonic, this difference will be in a negative direction, and it will be positive when the plasma is hypotonic (Fig. 1A–B).
When using data obtained from manual methods, manipulation of plasma osmolality did not alter the calculated change in blood volume (P = 0.299, Fig. 2B), but a greater change in red cell (P < 0.001) and plasma volume (P < 0.001, Fig. 3B) was observed as sample osmolality increased. Again, a linear regression analysis was undertaken to predict the change in plasma volume resulting from a known change in plasma osmolality. The change in plasma osmolality significantly predicted the change sample in plasma volume (B = −0.160, t(112) = 44.7, P < 0.001), with the results indicating that this explained 95% of the variance (R2 = 0.948, F(1,112) = 1995.8, P < 0.001). These data suggest that a 10 mOsm·kg−1 change in plasma osmolality produces a 1.6% change in the calculated plasma volume, with similar effects seen for both hypertonicity and hypotonicity; when the plasma is hypertonic, this difference will be in a positive direction, and it will be negative when the plasma is hypotonic (Fig. 3A–B). In contrast to these observations, when Hb and Hct data from the AHA were used to calculate changes in blood (P = 0.248, Fig. 2A), red cell (P = 0.525), or plasma (P = 0.291, Fig. 3A) volumes, manipulation of plasma osmolality produced no significant effects on the values obtained.
Study 2 data
Preexercise plasma osmolality was 296 ± 1 mOsm·kg−1. These values are approximately 10 mOsm·kg−1 higher than that of the in vivo plasma osmolality because of the addition of EDTA (see previous part of this article). A significant increase in osmolality was observed during exercise, with values of 309 ± 3 mOsm·kg−1 measured at exhaustion. Hb concentration was measured as 15.7 ± 0.9 and 15.5 ± 0.8 g·dL−1 before exercise using the cyanmethemoglobin method and the AHA, respectively (Fig. 4A). A significant increase in Hb concentration was observed during exercise (P < 0.001), with no significant difference in the values obtained between analytical methods (P = 0.355). Although no difference in Hct was apparent at rest (P = 0.659), mean spun PCV was 1.0 ± 0.9 Hct units lower during exercise when compared with data from the AHA (P < 0.001, Fig. 4B).
Percentage changes in blood, red cell, and plasma volumes are presented in Table 1. Exercise resulted in a reduction in blood volume (P < 0.001), but there was no difference in the magnitude of change determined by the two analytical methods (P = 0.109). Although red cell volume did not significantly change during exercise (P = 0.657), there was a difference in the response reported by the manual and AHA methods (P = 0.047). When compared with the preexercise sample, plasma volume fell during exercise (P < 0.001). There was a marked difference in the degree of change in plasma volume calculated by the two different methods, with a reduction of 8.7% ± 3.4% and 11.3% ± 3.5% observed at exhaustion with the manual and AHA methods, respectively (P = 0.002).
Serial measurements of Hb concentration and Hct are widely used to calculate changes in blood, red cell, and plasma volumes, most often using the method described by Dill and Costill (4). Many laboratories routinely measure Hb and Hct using either traditional assay methods (Hb by cyanmethemoglobin method and Hct by spun PCV) or by the use of an AHA. The results of the present study demonstrate that changes in plasma osmolality within the range expected in normal physiology can significantly alter the measurement of Hct by spun PCV. This response is eliminated, however, when Hct is determined using a commercially available AHA. Because red cell volume is highly responsive to changes in plasma osmolality, conditions or interventions that result in a marked change in plasma osmolality may result in an inappropriate value for Hct if this is measured using an AHA. Despite these observations, a large number of published studies, as well as antidoping laboratories undertaking hematological analysis as part of the ABP, still rely on Hct data obtained from an AHA. Where substantial alterations in plasma osmolality have occurred, it is clear that these data are invalid unless an appropriate correction can be applied.
Plasma osmolality is dictated largely by the circulating concentrations of electrolytes and, to a lesser extent, glucose (except in diabetic hyperglycemia), and it is typically maintained within a relatively narrow range (275–295 mOsm·kg−1). Although this is true under normal conditions, values in excess of 300 mOsm·kg−1 have been reported after prolonged exercise, representing a change of 12–15 mOsm·kg−1 from baseline (17). Similar responses are seen after short-term very intense exercise, where blood lactate concentration may exceed 20 mmol·L−1 with corresponding hypertonicity. In severe dehydration, serum sodium concentration may exceed 175 mmol·L−1, with accompanying osmolality in excess of 360 mOsm·kg−1 (9,14). Likewise, it is possible for low values to be reported after ingestion of large volumes of dilute beverages during a short period (11) or in some cases of hyponatremia (approximately 245 mOsm·kg−1 (12)).
The use of Hb and Hct to determine changes in blood and plasma volumes is widespread in physiology research. Dill and Costill (4) improved on previous methods used to determine changes in plasma volumes by using the Hb concentration to account for changes in the total blood volume that may occur between sampling periods. Given the established relation between red cell volume and plasma osmolality (13), it would be expected that any change in plasma osmolality would significantly alter the volume of the red cells. This response has been characterized previously, with a linear reduction in spun PCV reported as the osmolality of the sample was increased from 260 to 420 mOsm·kg−1 through the addition of NaCl or sorbitol (16). The present data support these previous findings and furthermore demonstrate that the use of spun PCV preserves in vivo effects of changes in plasma osmolality on calculated changes in plasma volume. These responses were abolished when these samples were analyzed using an AHA. The measurement of Hct used by many AHA is made by measuring the electrical impedance of cells suspended in a conductive reagent when passed through a small aperture. The intensity of the electronic pulse generated from each red blood cell is proportional to the cell volume, and the Hct is directly determined on the basis of the count and volume detection of each individual red blood cell.
The instrument used in the present study reports a final dilution of 1/10,000 (sample to reagent) before the volume measurement is undertaken; the osmolality of the commercially available AHA diluent (product ref. 8547169, COULTER® AC·T™ 5diff Diluent, Beckman Coulter) was measured as 342 mOsm·kg−1. Given this level of dilution, any variation in red cell volume due to the osmolality of the sample will be completely lost. Baseline values will also be distorted because of the high osmolality of the “isotonic” diluent. This wholly explains the discrepancy in the results obtained between the manual and AHA methods in the present study. Although there are likely to be subtle differences in the sensors, reagents, and algorithms used between AHA manufacturers (Coulter, Sysmex, etc.), the basic electronic particle counting principle remains very similar, leaving all the analyzers using this method open to this source of error. The methods used by some small handheld devices (iStat, Nova, etc.) do not significantly dilute the blood and are not directly influenced by changes in serum osmolality of the sample. These analyzers estimate an expected conductivity value by measuring the plasma cation concentrations; this in itself is likely to introduce a significant source of measurement error.
To demonstrate the presence of this response in vivo, measurements of Hb and Hct were made in a group of volunteers before and during a bout of exhausting exercise in the heat (exercise time, 67.5 ± 12.4 min). Although the two methods gave the same values for Hct before exercise when the mean plasma osmolality was 296 ± 1 mOsm·kg−1, mean spun PCV was significantly lower (46.3% ± 2.3%) during exercise than the values obtained from the AHA (47.4% ± 2.7%). An increase of 12 ± 3 mOsm·kg−1 in plasma osmolality was observed during the exercise bout, and the difference in Hct data between the two methods is consistent with the change expected from study 1 that involved passive manipulation of the sample osmolality after collection. Clearly, the difference in Hct values, with no effect on Hb concentration, resulted in an effect on the calculated changes in red cell and plasma volumes. This again supports the notion that it is not appropriate to use AHA-derived values in these equations, because an error arises when the osmolality of the sample changes. Given that plasma osmolality almost invariability changes when plasma volume changes, repeat measures in this case have little or no validity. The data from study 1 suggested that a 10 mOsm·kg−1 increase in plasma osmolality would result in a 0.8-unit reduction in Hct, corresponding to a difference of 1.6% in the calculated change in plasma volume. Given that the mean spun PCV during exercise was 1.0 ± 0.9 units lower than the Hct measured using the AHA, it appears that there is indeed scope to use these regression data as a retrospective correction factor to account for the isotonic Hct data produced by the hematology analyzers. It is worth noting that the magnitude of blood and plasma volume change observed was likely influenced by the change in posture experienced when the volunteers moved from a seated position to the bicycle saddle (5). Although this may lead to an overestimation of the change in plasma volume induced by the exercise itself, it does not influence the outcome of the present study.
The ABP was introduced by WADA to identify and target athletes for specific analytical testing and to directly pursue antidoping rule violations. The hematological module of the ABP relies on serial measurements of Hct, along with several other hematological parameters, to determine an “Abnormal Blood Profile Score” (15) that is monitored longitudinally. WADA-accredited laboratories are required to use a commercially available AHA to produce these data (typically Sysmex XT-2000i), and consequently, ABP samples are subject to the analytical artifacts illustrated in the present study. Presample conditions are not standardized as part of the doping control process, meaning that distinct variation in osmolality is possible from sample to sample because of exposure to a wide variety of factors (exercise, environment, and nutrition). This may significantly limit the reliability of the ABP to detect the use of prohibited methods to enhance oxygen transfer, particularly because plasma osmolality is not routinely determined.
Although spun PCV is widely used in the calibration of many AHA, it is not without some potential sources of error. Spun PCV can be influenced by several variables, including plasma trapped between the red cells, contamination with white blood cells, an indistinct boundary between the red and white cell layers, nonflat tube seals, red cell dehydration by excessive anticoagulant, and red cell oxygenation (1). It appears, however, that these potential errors occur in both positive and negative directions and ultimately tend to counterbalance each other, leading to a small net effect on the value obtained (2). Original corrections for plasma trapped between the red cells were in the region of 2%–4%, but the value of 1% suggested by Dacie and Lewis (3) appears more appropriate in the light of recent recommendations (1,2). The International Council for Standardization in Hematology suggests that K2EDTA is the anticoagulant of choice for blood cell counting and sizing (6). Clearly, K2EDTA at high concentrations can produce a significant effect on spun PCV (8). When present in a concentration of 1.5–2.0 mg·mL−1 of blood, previous evidence suggests that it does not appear to have any significant effect on the blood count parameters (6,10). Addition of 1.5 mg·mL−1 K2EDTA in the present study produced a 10 ± 1 mOsm·kg−1 change in the measured osmolality of the samples. As highlighted previously, this change in plasma osmolality can alter the spun Hct and any subsequent calculation of changes in plasma volume, but this response will be consistent across all samples, assuming that the quantity of K2EDTA is uniform.
In conclusion, the results of the present study suggest that changes in plasma osmolality within the normal physiological range have a marked effect on spun PCV. This effect is eliminated when Hct is measured using an AHA that bathes the sample in an “isotonic” medium before analysis. Consequently, the use of AHA-derived data to determine percentage changes in plasma volume is inappropriate if a significant change in plasma osmolality occurs during an experimental protocol. In addition, WADA-accredited laboratories monitor Hct using AHA methodologies as part of the ABP. In this case, varying sample osmolality is likely, and this may limit the reliability of the ABP to detect the use of prohibited methods to enhance oxygen transfer. The present data suggest that it is possible for a correction factor to be retrospectively applied to the Hct value obtained using an AHA on the basis of the change in osmolality observed during a trial (see recommendations in the succeeding part of this article). It is worth noting that many studies using AHA to determine changes in plasma volume do not routinely measure plasma osmolality. Perhaps this needs to be addressed if this analytical method continues to be used in this manner.
Practical Application of These Findings
1. A change in plasma osmolality occurring during a study will result in a consistent change in Hct in vivo. This response is preserved when Hct is determined using spun PCV methods (i.e., microcentrifugation) but is lost when Hct is measured using AHA that mixes samples with a reagent before analysis (Coulter, Sysmex, etc.).
2. This response is consistent in both hypertonic and hypotonic environments, meaning that a correction factor can be applied to AHA-derived Hct values to account for this measurement artifact.
3. A 10 mOsm·kg−1 change in plasma osmolality results in a 0.8-unit difference in Hct measured by microcentrifugation. If plasma osmolality increases during a trial, then this correction should be subtracted from AHA-derived Hct values. Likewise, if plasma osmolality falls, then this correction should be added to AHA-derived Hct values.
4. These corrections should be applied before using these data to calculate percentage changes in blood, plasma, and red cell volumes using the methods described by Dill and Costill (4) or in any application related to the WADA ABP.
No external funding was obtained to undertake this work. No conflict of interest is present. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Bull BS, Fujimoto K, Houwen B, et al. International Council for Standardization in Haematology (ICSH) recommendations for “surrogate reference” method for the packed cell volume. Lab Hematol. 2003; 9: 1–9.
2. Bull BS, Hay KL. Is the packed cell volume (PCV) reliable? Lab Hematol. 2001; 7: 191–6.
3. Dacie JV, Lewis SM. Practical Haematology. 4th ed. London: Churchill; 1968. pp. 45–9.
4. Dill DB, Costill DL. Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. J Appl Physiol. 1974; 37: 247–8.
5. Hagan RD, Diaz FJ, Horvath SM. Plasma volume changes with movement to supine and standing positions. J Appl Physiol. 1978; 45: 414–7.
6. International Council for Standardization in Haematology: Expert Panel on Cytometry. Recommendations of the International Council for Standardization in Haematology for ethylenediaminetetraacetic acid anticoagulation of blood for blood cell counting and sizing. Am J Clin Pathol. 1993; 100: 371–2.
7. Kennedy JB, Maehara KT, Baker AM. Cell and platelet stability in disodium and tripotassium EDTA. Am J Med Technol. 1981; 47: 89–93.
8. Lampasso JA. Error in hematocrit value produced by excessive ethylenediaminetetraacetate. Am J Clin Pathol. 1965; 44: 109–10.
9. Latcha S, Lubetzky M, Weinstein AM. Severe hyperosmolarity and hypernatremia in an adipsic young woman. Clin Nephrol. 2011; 76: 407–11.
10. Lewis SM, Stoddart CT. Effects of anticoagulants and containers (glass and plastic) on the blood count. Lab Pract. 1971; 20: 787–92.
11. Maughan RJ, Leiper JB. Sodium intake and post-exercise rehydration in man. Eur J Appl Physiol Occup Physiol. 1995; 71: 311–9.
12. Oster JR, Singer I. Hyponatremia, hyposmolality, and hypotonicity: tables and fables. Arch Intern Med. 1999; 159: 333–6.
13. Ponder E, Robinson EJ. The measurement of red cell volume: V. The behaviour of cells from oxalated and from defibrinated blood in hypotonic plasma and saline. J Physiol. 1934; 83: 34–48.
14. Riedesel ML, Allen DY, Peake GT, Al-Qattan K. Hyperhydration with glycerol solutions. J Appl Physiol. 1987; 63: 2262–8.
15. Sottas PE, Robinson N, Giraud S, et al. Statistical classification of abnormal blood profiles in athletes. Int J Biostat. 2006; 2: 3.
16. Stott RA, Hortin GL, Wilhite TR, Miller SB, Smith CH, Landt M. Analytical artifacts in hematocrit measurements by whole-blood chemistry analyzers. Clin Chem. 1995; 41 (2): 306–11.
17. Watson P, Black KE, Clark SC, Maughan RJ. Exercise in the heat: effect of fluid ingestion on blood-brain barrier permeability. Med Sci Sports Exerc. 2006; 38 (12): 2118–24.
Keywords:© 2014 American College of Sports Medicine
PLASMA OSMOLALITY; PACKED CELL VOLUME; ERYTHROCYTE VOLUME; BLOOD VOLUME; DILL AND COSTILL