Original Contributions

The Effects of Infusion on 10-km Race Time Anthony J. Brien, PhD. Toby L. Simon, MD

>urpose of this study was to investigate the effect of infusion of 400 mL of red • enhanced performance (Sports Illus- blood cells (RBCs) on 10-km track race time, submaximal heart rate, , trated, Jan 21,1985, p 12). 2,3-diphosphoglycerate, and partial pressure of oxygen at 50% hemoglobin The efficacy of induced erythrocy- saturation. Six highly trained, male, distance runners twice donated a unit of themia as an aid to competitive en- RBCs, which was frozen for subsequent reinfusion. Eleven weeks after the durance performance is associated with skepticism.* This study determined the second donation, they undertook a series of three competitive 10-km races on a effects of induced erythrocythemia on standard 400-m track: before infusion, after 100 mL of saline solution, and after highly trained, male, distance runners 400 mL of autologous, previously frozen deglycerolized RBCs. All subjects took in a 10-km track race. all trials in this double-blind, placebo, crossover, experimental design. Running time was recorded at each 400-m split, and blood was collected prior to each trial. METHODS The data were analyzed by analysis of variance. Results following the RBC Six male distance runners were se- infusion showed a significantly higher hematocrit concentration, a significantly lected to participate in a double-blind faster 10-km run, a nonsignificant decrease in submaximal heart rate (10 beats study of induced erythrocythemia. Both per minute), and no significant changes in either 2,3-diphosphoglycerate or subjects and personnel involved in the measurements were kept unaware of partial pressure of oxygen at 50% hemoglobin saturation. Erythrocythemia treatment used. All subjects were resi- induced by the infusion of 400 mL of autologous packed RBCs effectively dents of Albuquerque 18 months prior to increased performance capacity in a 10-km track race, probably due to an the beginning of the study and were increase in oxygen delivery to the working muscles. acclimatized to the altitude of the city (JAMA 1987:257:2761-2765) (1500 to 1800 mX Their previous 10-km times prior to the study are shown in Table 1 These times and resting heart SUBJECT of red blood cell (RBC) stances where this practice has ap- rates showed them to be relatively well sions as an ergogenic aid has been of peared to benefit athletes. According to trained for amateur runners. The sub- special interest to coaches, athletes, the Sunday Times of London (Feb 5, jects were instructed to maintain their and scientists since Pace et af first 1984, p 30X when Italian Francesco normal daily activities and training conducted a study on homologous blood Moser twice shattered the outdoor one- schedule throughout the course of the infusion in 1947. The underlying theory hour cycling record in Mexico City in study. Relative anthropometric and of induced erythrocythemia is that in- 1984, he was assisted by an entourage of other data are shown in Table 1 as creased oxygen-carrying capacity in- two cardiologists and eight men 18 to 20 means ±1 SD. Informed consent was creases performance of working muscle years of age who were chosen several obtained from all subjects after ap- by increasing oxygen supply. Studies in months previously because of their proval by the University of New Mexico laboratory settings have produced con- blood type compatibility with Moser. Human Research Review Committee. flicting results.1 Finnish distance runner Kaarlo A double-blind, crossover, experi- There have been several reported in- Maaininka freely admitted receiving 2 mental design was utilized. The sub- units of blood from team doctors shortly jects were prevented from observing From the Departments of Health, Physical Education, before he won silver and bronze medals experimental treatments by wearing and Recreation (Or Brien}. Pathology {Dr Simon), and in the 10- and 5-km events, respectively, blacked-out goggles and having loud Medicine (Dr Stmon\ University of New Mexico, and United Blood Services. Albuquerque. Dr Brien is now at the 1980 Olympic games in Moscow.* music played through headphones dur- with the Bahrain Sports Institute. Manama. State ol The latest disclosure of blood boosting ing all infusion processes. Those record- Bahrain. among several members of the success- ing study results were blinded as to Reprint requests to Room 337 BRF, Department of Pathology. School ol Medicine. University ol New Mex- ful US cycling team in the 1984 Olym- treatment order. ico. Albuquerque. NM B7I31 (Dr Simon). pics has rekindled the question of The subjects underwent two phle-

JAMA, May 22/29,1987—Vol 257, No. 20 Red Blood Cell Infusion—Brien & Simon 2761 RESPONDENTS 70 SCA001870 botomies (460 mL each) spaced eight weeks apart. Red blood cells were sepa- rated from whole blood and stored, us- ing the high-glycerol freezing tech- nique.' The subjects continued training (112 to 176 km per week) for approx- imately 11 weeks after the second phle- botomy before receiving the infusions. Six days before the first infusion, collection of baseline data was initiated. 15 33 Treadmill studies were done six days before infusion, blood specimens five days before, and the 10-km race three 4 35 177.80 61.82 39 41 (0.41) 177.0 17 32 15 days before. Two days before the first 5 41 167.64 58.18 45 43 (0.43) 123.9 19 ' 33 51 infusion, treadmill was repeated; at 6 33 182.88 65.00 40 42 (0.42) 112.7 16 33 41 three days, blood studies; and at five 5? 33.6 173.99 6121 41.8 41.6 (0.416) 134.64 14.3 33 25 days, the second race. Three days after SD 5.2 5.96 2.57 2.3 1.4 (0.014) 22.59 4.3 44 the second race, the second infusion was given, and the procedures were re- peated at the same intervals. Infusion treatments consisted of an autologous infusion of 400 mL of previously frozen deglycerolyzed RBCs or a placebo infu- sion of 100 mL of normal saline (the Group 1 approximate amount of saline used to resuspend the RBCs). The subjects were randomly divided into two groups (1 and 2) and were randomly assigned treatments, with the order of receipt of the infusion treatments reversed using a crossover double-blind design. The subjects ran a series of three 10-km open-competition track races at nine-day intervals on a 40-m rubber- ized, asphalt, outdoor track located at altitude (1803 m). The first race was used as a baseline control measure, while the second and third races were run five days after the first and second infusions, respectively. (0.40] Three submaximal treadmill tests Baseline Postplacebo Poslrelnfusiori were taken three days prior to each race to determine any change in heart rate (HRX measured by recording electro- cardiogram. Prior to each test, subjects performed a standard six-minute warm- Group 2 up. Testing began at a velocity of 8 kra/h (5 mph) and increased 1.6 km/h (1 mph) each minute until 14.4 km/h (9 mph) was reached. At that time, the treadmill speed was held constant and the grade elevated to 5%. Subjects ran at this load for six minutes, with the submaximal rate being recorded between the fifth and" sixth minutes. Blood samples (10 mL) were collected in heparinized syringes from the ante- cubital vein two days prior to each race. Samples were analyzed for hematocrit (Hct), red-cell 2,3-diphoaphoglycerate (2,3-DPG), and partial pressure of oxy- 40T gen at 50% hemoglobin saturation (PJi (0.40)? The Hct was measured on a micro- ' BaseS ne Postreihfusfon Postplacebo hematocrit centrifuge (coefficient of variation, 1.2%). The 2,3-DPG was Fig 1.—Relationship between hematocrit and infusions. HemalocrH increase occurred after infusion of red assayed enzymatically using a method Hood eels and was sustained tor week between 10-km race after relnfusion and 10-km race after placebo In that has a coefficient of variation of group 2. 2.0%. A device that generates an oxy-

2762 JAMA, May 22/29,1987—\tol 257. No. 20 Red Blood Cell Infusiorv-Bnen & Simon SCA001871 hemoglobin dissociation curve by mea- suring arterial oxygen pressure Groupl through a membrane and saturation 35 -i spectrophotometricaUy on deoxyge- nated blood, which is reoxygenated in stages, was used to determine PM values 34- (reproducible to 1 mm Hg). Three subjects received the saline placebo first and were designated as 33- group L The other three subjects re- ceived the R6C infusion initially and were designated as group 2. . Analysis of variance was performed 32- for the major end points. Between-sub- jects effects and within-subjects effects were analyzed. Because of multiple 31- comparisons, P<.01 was chosen as the determinant of significance.

RESULTS Baseline Postplacebo Postreinfuston In the three xraces, the temperature 08.3°C, 16.1°C, and 17.2°C), humidity (29%, 28%, and 24%X and wind velocity (8, 3.2, and 6.4 km/h [5, 2, and 4 mph]) were similar. Group 2 Figure 1 shows that Hcts were in- 35-n creased more than 5% by the RBC infusions. The treatment relation- ship with Hct was highly significant 34H (P = .004). The comparisons between baseline Hct and Hct after reinfusion and between placebo and reinfusion 33-i were similarly significant (P<.001 and P=.003, respectively). In group 2, there was a carryover effect in that the Hct increase from baseline caused by 32 —I the reinfusion persisted after the placebo (P=.004 for the interaction with order). Hematocrit after placebo 31 — was higher than baseline in group 2 (when it followed the reinfusion with RBCs) but not in group 1 (P = .001 for this difference), and Hct after placebo Baseline PostreWusion Postplacebo was much lower than after reinfusion in group 1 but not in group 2 (P=.008 for Fig 2.—Relationship between time to run 10-km race and reWusioci or placebo. In group 1, time fen after this interaction with order). reinfusion but not after placebo, In group 2, time fefl after reinfusion, and Improved time was sustained for 13 The treatment effect was significant days from reinfusion of red blood certs to 10-km race after placebo infusion. on the 10-km race time. Results are shown in Fig 2 and Table 2. Reinfusion of RBCs caused a reduction of approx- infusion of RBCs. These changes were Blood volume increases with increased imately one minute in time to run the not statistically significant. There were cardiac output and HR, working mus- 10-km race. The comparisons of baseline no differences for PM and 2,3-DPG. cles experience vasodUation, sympa- to saline were not significant, but base- Mean Pn at baseline was 26.8±0.5mm thetic nervous system stimulation oc- line to reinfusion and saline to reinfu- Hg. After RBC infusion, it was curs, a linear increase in oxygen sion were both significant at P<.001. 27.4 ±1.0 mm Hg. Mean 2,3-DPG level consumption occurs, and lactate pro- Order did affect the results, because of was 2.38 ±0.13 ujnol/mL at baseline and duction increases and causes increased the carryover of the reduced time in 2.36 ±0.36 u.molAnL after infusion. ventilation. Endurance exercise, such group 2 from the reinfusion through to as running, results in increases in max- placebo. Running time after placebo COMMENT imum oxygen consumption but not in was much lower than baseline in group 2 The purpose of this investigation was muscle hypertrophy or strength in- (where it followed the reinfusion with to determine the effects of a reinfusion creases.' Thus, in the context of these RBCs) but not in group 1 (P=.005 for of 400 mL of autologous RBCs on the complex physiological changes, this this difference), and running time after 10-km track race time in highly trained, study asks the question if increased placebo was much slower than after male, distance runners. Supporting RBC mass can provide a sufficient in- reinfusion in group 1 but not in group 2 data were measures of Hct, 2,3-DPG, crease in oxygen utilization to improve (P<.001 for this interaction with order). P^.andHR. performance. Figure 3 shows the decrease in sub- A number of important physiological The RBC infusion resulted in a faster maximal HR of varying amounts after changes occur in running a 10-km race. overall 10-km race time in comparison

JAMA. May 22/29.1987—Vol 257, No. 20 Red Blood Cell Infusion—Brten 8. Simon 2763 SCA001872 Table 2.—Individual Performance Tests Results: Race Time in Incremente of Four Laps (Minutes and s) with the preinfusion race or the race after saline when saline was first. In 10000m Subject Race 1600m 3200m 4*OO m 6400m 8000 m 9600m (to km) group 1, there was no change after 1 5.00 1021 15.49 21.24 27.02 32.42 34.02 saline infusion but an improvement 1 2 S.03 10,26 1B.01 21.43 27.24 33.08 34.32 after the RBC infusion. In group Z, the 3» 5.03 10.16 15.32 2O54 26.18 31.45 33.02 change was seen after infusion and re- 1 5.00 10.13 15.32 20.53 28.21 31.40 32.51 mained throughout the rest of the study 2 2 5.01 10.12 15.31 20.51 28.13 31.30 32.42 (ie, persisted after the second infusion 3» 5.00 10.00 14.59 20.05 25.16 3022 31.33 of salineX In either case, it was the RBC 1 5.00 10.20 15.45 21.11 2B.40 32.09 33.30 infusion that caused the decrease, and it 3 2 5.00 10.20 15.49 21.19 26.49 32.18 33.37 lasted at least 13 days. The improved 3* 5.00 10.07 1520 20.35 25.50 3104 3230 performance was related to increased 1 5.01 10.08 15.21 20.39 26.00 31.19 3251 Hct, facilitating aerobic metabolism in 4 2« 5.00 10.04 15.00 19.59 25.00 29.59 31.12 the working muscles. The identical pat- 3 4.59 9.59 14.58 20.00 25.02 30.01 31.14 tern of interaction between RBC reinfu- 1 5.14 10.41 16.09 21.41 27.16 32.48 34.09 sion and order seen for Hct and 10-kra 5 2' 5.03 10.12 15.31 2021 26.13 31.31 32.48 race time effectively excludes a psycho- 3 S.03 10.16 15.32 20.53 28.13 3155 32.51 logical effect. Thus, only with elevation 1 5.07 1028 15.56 21.30 27.05 32.31 33.48 in Hct was the improvement in the 6 V 5.00 10.12 15:31 2065 26.28 31.51 33.04 10-km running time Been, 3 5.03 10.16 15.33 20.58 26.26 31.50 33.03 Astrand and Rodahl' note that max- imal performance in events of approx- " 'Red Mood cell Infusion. Race 1 a baseline. Tne other race vras alter placebo saltne Infusion (2 or 3 without an imately 30 minutes' duration represents asterisk). an energy output of 95% from aerobic and 5% from anaerobic sources. Fbllow- ing induced erythrocythemia, the mean 190- Group 1 performance time for the 10-km track race was 69 s faster, ie, an average of almost 7 s/km. Similar results have been reported by Goforth et al," who found 180- that runners improved an average of 9 s per mile in a 3-mile time trial. The environmental conditions for the track 170- races showed little variation and are unlikely to have affected performance mi time. Questioning prior to the two 160- postinfusion races revealed that the subjects did not know exactly what they had received, ie, saline or RBCs. 150- Erythrocythemia may have influ- enced submaximal performance in this study, as evidenced by lower HRs, but the finding was not statistically signifi- Baseline Postplacebo Postrelnfuslon cant. During a constant level of submax- imal steady-state exercise, HRs were 10 beats per minute lower at 48 hours after RBC reinfusion. Brack et al' re- ported comparable results and claimed Group 2 that ervthrocythemia, through an ele- 190-1 vation in arterial oxygenation, permits submaximal aerobic metabolism to be attained at a lower cardiac output g. 180- Infusion of 400 mL of RBCs did not cause a compensatory shift in Pn, and this agrees with findings by Williams et 170- al" and Goforth et aL' The concentration of 2,3-DPG in the RBCs is important relative to the af- finity of oxygen to hemoglobin. An in- 160- creased level of 2,3-DPG is associated with a reduced affinity of oxygen to hemoglobin, thereby facilitating there- 150- lease of oxygen from hemoglobin when

Fig 3.—Submaximal heart rate decreased alter botti placebo and reinfusion In group 1 and after Baseline Postreinluston Poslftecebo reinfusion in group Z. Thais was greater fall after reinfusion than placebo.

2764 JAMA, May 22/29.1987—Vbl 257, No. 20 Red Btood Cell Infusion—Brien & Simon SCA 001873 the blood reaches the muscles. On the span to allow the Hct and blood volume with autologous reinfusion under care- other hand, hemoglobin affinity for oxy- to stabilize between phlebotomy and ful medical supervision, as in this study, gen is increased when the level of 10-km race. the procedure is safe. Homologous 2,3-DPG is decreased."'12 Frozen RBCs Within the limitations of this study, it blood and autologous blood without maintain 2,3-DPGB; thus, there were no appears that the infusion of 400 mL of these careful controls are not safe. significant differences in 2,3-DPG levels RBCs into a distance runner with stable Overtransfusion to could in this study. Oxygen affinity of the Hct will significantly increase the total result in ischemic episodes due to de- RBCs was not adversely affected. Com- RBC concentration, which, in turn, will creased flow in vessels. parable findings have been reported by contribute to a significant decrease in Autologous transfusion has raised Buick et al* and Goforth et al* the time required to run a 10-km track ethical questions. It is regarded by the Increases in blood viscosity could af- race. The effect of the infusion and the International Olympics Committee as fect oxygen delivery following an eleva- subsequent increase in RBC mass may dishonest.' Our purpose was not to en- tion in Hct. Research indicates that the also relate to the observation that run- courage dishonesty in sports but rather ideal Hct for the delivery of oxygen to ners usually experience some element to explore the effects of increase in RBC contracting muscle is between 50% and of hemolysis and iron depletion, which mass on runners' performance. Well- 60% volume (0.50 and 0.60). This finding keeps their Hct at a lower level than trained individuals appear to utilize the differs from the generally accepted op- they would have if not running.* The additional oxygen carried by the in- timal Hct of 45% volume (0.45) at rest" RBC infusion thus increases the run- creased RBC mass advantageously in In this study, the mean Hct after RBC ners' Hct to a level that might be their athletic performance. Increased RBC infusion was only 47.4% volume (0.474), normal value if not running. In this mass can improve performance even in one not likely to compromise blood flow study, the Hcts were within accepted runners at a stable, presumably normal due to increased viscosity. normal range but below mean levels Hct level. The results of this study are in agree- seen in men acclimated to altitude in ment with those of a number of previous Albuquerque.24 The subjects in this This study was supported by grants from Gen- M< eral Clinical Research Center Program DRB, Na- »tjgations* -* that reported sigmf- study, although good-caliber runners, tional Institutes of Health M01-RR00997-09, and increases in endurance perform- were not elite distance runners, and the Public Health Service Transfusion Medicine Aca- demic Award 5 K07 HL01232-03 from the National ance. On the other hand, the results results obtained in this study may or Heart, , and Blood Institute, National Insti- were at odds with those of a number of may not be applicable to the latter. 0 0 tutes of Health, and Blood Systems Research former studies '* that reported no This procedure of "blood doping," or Foundation. significant effect of erythrocythemia on induced erythrocythemia, is controver- Statistical consultation was provided by Richard Harris, PhD. The authors also thank Lee Gates, endurance performance. The decisive sial in sports medicine (Sunday Times MD, of the Lovelace Medical Center Laboratory elements seem to be the quantity of [London], Feb 5, 1984, p 30; Sports and the staffs of the Conical Research Center and RBCs infused and an appropriate time Illustrated, Jan 21,1985, p 12X" Done United Blood Services. References 1. Face N, Lozner EL, Consolazio WV, et al: The J Appl Physiol 1980:48:636-642. circulation and work capacity after red blood cell increase in hypoxia tolerance of normal men accom- 10. Williams MH, Ltodhejm H, Schuster R: The reinfusion under normoxia and hypoxia in women, panying the polycythemia induced by transfusion effect of blood infusion upon endurance capacity abstracted. Med Sn Sportt 1979:11:98. of erythrocyteg. Am } Pkysiol 1947;148:152-163. and ratings of perceived exertion. Med Sri Sportt 18. Von Rost R, HoUman W, Llesen H, et at Uber 2. Williams MB: Blood doping: A review. J Drug 1978:10:113-118. den Einfluss einer Erthrozyten-Retransfmion auf Itntet 1980:10:331-339. 11. Edington BT, Edgerton R: The oxygenation of die Kardio-puhnonale Leistungsfahigkeit. Sport- 3. Higdon H: Blood-doping among endurance ath- hemoglobin in the presence of diphosphoglycerate: ant Sportrradian 1975^6:137-144. letes. American Medical Newt, Sept 27, 1985, Effect of temperature, pH, ionic strength and 19. Robinson B, Epstein S, Kahler R, et al: Circu- pp 37, 39-41. hemoglobin concentration. Biochemistry 1969& latory effects of acute expansion of blood volume. 4. Klein HG: and athletics. 2567-2571. Ctrc Rts 1966:29:26-32. N Engl J lied 1985;312:864-S56. 12. Klein J, Laver Di The effect! of deoxygenation 20. Frye A, Ruhling R: RBC infusion, exercise, 5. Widnunn FK: Technical Manual. Arlington, of aduR and fetal hemogtobinon the synthesis of red hemoconcentration and VOj, abstracted. Med Set Va, American Association of Blood Banka, 1985. cell 2,3-diphosphoglycerate and its in vivo con- Sport* 1977:9:69. 8. Marino N, De Pasquak E, Bruno H, et at sequences. J din Invest 1970:49:400-407. 21. VkJemanT,HytomaaT:EffectDfbloodremoval Cardiovascular system, in Scott WN, Nisonson B, 13. GujianA-.Textbook if Medical Pkytiology,*& and autotransfusion on heart rate response to a jjjAo\

JAMA. May 2Z/29.1987—Vbl 257, No 20 Red Blood Cell Infusion—Brien & Simon 2765 SCA001874