Frequency of the C34T mutation of the AMPD1 gene inworld-class endurance athletes: does this mutation impair performance?

Juan C. Rubio,1 Miguel A. Martın,1 Manuel Rabadan,2 Felix Gomez-Gallego,3 Alejandro F. San Juan,3

Juan M. Alonso,4 Jose L. Chicharro,5 Margarita Perez,3 Joaquın Arenas,1 and Alejandro Lucia3

1Research Centre, University Hospital 12 de Octubre; 2Department of Physiology, Sport MedicineCenter, Higher Sports Council; 3European University of Madrid; 4Medical Department,Spanish Track and Field Federation; and 5Complutense University, Madrid, Spain

Submitted 13 December 2004; accepted in final form 20 January 2005

Rubio, Juan C., Miguel A. Martın, Manuel Rabadan, FelixGomez-Gallego, Alejandro F. San Juan, Juan M. Alonso, Jose L.Chicharro, Margarita Perez, Joaquın Arenas, and AlejandroLucia. Frequency of the C34T mutation of the AMPD1 gene inworld-class endurance athletes: does this mutation impair perfor-mance? J Appl Physiol 98: 2108–2112, 2005. First published January27, 2005; doi:10.1152/japplphysiol.01371.2004.—The C34T muta-tion in the gene encoding for the skeletal muscle-specific isoform ofAMP deaminase (AMPD1) is a common mutation among Caucasians(i.e., one of five individuals) that can impair exercise capacity. Thepurpose of this study was twofold. First, we determined the frequencydistribution of the C34T mutation in a group of top-level Caucasian(Spanish) male endurance athletes (cyclists and runners, n � 104).This group was compared with randomly selected Caucasian (Span-ish) healthy (asymptomatic) nonathletes (n � 100). The second aim ofthis study was to compare common laboratory indexes of enduranceperformance (maximal oxygen uptake or ventilatory thresholds)within the group of athletes depending on their C34T AMPD1 geno-type. The frequency of the mutant T allele was lower (P � 0.05) in thegroup of athletes (4.3%) compared with controls (8.5%). On the otherhand, indexes of endurance performance did not differ (P � 0.05)between athlete carriers or noncarriers of the C34T mutation (e.g.,maximal oxygen uptake 72.3 � 4.6 vs. 73.5 � 5.9 ml �kg�1 �min�1,respectively). In conclusion, although the frequency distribution of themutant T allele of the AMPD1 genotype is lower in Caucasian eliteendurance athletes than in controls, the C34T mutation does notsignificantly impair endurance performance once the elite-level statushas been reached in sports.

AMP deaminase; cycling; running

DURING INTENSE EXERCISE CAUSING adenosine monophosphate(AMP) accumulation, the enzyme AMP deaminase (AMPD;EC 3.5.4.6) is activated in skeletal muscle. This enzyme is avery important regulator of muscle energy metabolism duringexercise, and its expression is highly regulated (11). By con-verting AMP to inosine monophosphate (IMP) with liberationof ammonia, AMPD displaces the equilibrium of the myoki-nase reaction toward ATP production (27). Also, the AMPDreaction is the initial reaction of the purine nucleotide cycle,which plays a central role in the salvage of adenine nucleotidesand in determining energy charge (15). Other possible impor-tant functions of the purine nucleotide cycle are the deamina-tion of amino acids (aspartate) and the regulation of the glycolyticpathway by the formation of ammonia and IMP (11, 15).

The skeletal muscle-specific isoform (M) of AMPD is en-coded by the AMPD1 gene, located on the chromosome 1p13-

p21 (30). A nonsense mutation [C to T transition in nucleotide34 (C34T)] in exon 2 of AMPD1 converting the codon CAAinto the premature stop-codon TAA, and thus resulting inpremature stop of protein synthesis, appears to be the maincause of AMPD deficiency (21). Approximately 2% of thegeneral Caucasian population is homozygous (TT) and nearly20% heterozygous (CT), for the aforementioned mutation (21,22, 27, 28, 37). Although some sedentary individuals with thisdefect (mostly TT) present with easy fatigability, cramps, ormyalgia after exercise, the mutation is also present in asymp-tomatic individuals (31). On the other hand, both heterozygousand homozygous statuses for the C34T mutation have beenassociated with increased severity of coexisting disorders (29).

Since the original report by Fishbein et al. (9), who firstproposed that a deficiency of AMPD causes exercise limita-tion, several studies using different exercise models haveanalyzed the extent to which this condition alters functionalcapacity and the mechanisms behind this potential functionallimitation (6, 23, 27, 32, 34, 36). Some controversy arises fromthese excellent reports. On the other hand, these studies wereconducted in nonathletes. No study has yet analyzed thefrequency distribution of C34T AMPD1 genotypes amongtop-level endurance athletes, e.g., Olympic-class runners orprofessional cyclists able to successfully complete 3-wk tourraces, nor whether the C34T mutation in the AMPD1 genemight affect their performance. Although one would expectthat AMPD deficiency might affect performance mainly duringshort-term, supramaximal [�100% maximal oxygen uptake(VO

2 max)] exercise inducing depletion of phosphocreatine (PCr)

and fall in the total adenine nucleotide pool (e.g., 400-m trackraces or short velodrome events), several studies have notedthe accumulation of IMP to occur at fatigue during prolonged,submaximal exercise (e.g., �1 h at 70–75% VO2 max), partic-ularly in the presence of low intramuscular glycogen stores bythe end of exercise (2, 24, 25, 33). The subsequent decrease inATP provision from carbohydrate sources may lead to atransient increase in ADP concentration, stimulating themyokinase reaction. This reaction results in the formation ofAMP, which must be rapidly deaminated to IMP and ammoniavia the activity of AMPD (35). The aforementioned phenom-ena at the muscle level are likely to occur at the end oftop-level endurance competitions, e.g., marathon races (13), oreach daily stage throughout 3-wk cycling races, each of whichlasts 5 or more hours but can include some bouts (�20–30

Address for reprint requests and other correspondence: A. Lucıa, EuropeanUniv. of Madrid, 28670 Madrid, Spain (E-mail: alejandro.lucia@uem.es).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

J Appl Physiol 98: 2108–2112, 2005.First published January 27, 2005; doi:10.1152/japplphysiol.01371.2004.

8750-7587/05 $8.00 Copyright © 2005 the American Physiological Society http://www. jap.org2108

min) of very intense exercise (�90% VO2 max) (16). ThusAMPD could also play a very important role in regulatingmuscle metabolism during exhausting endurance events.

Therefore, the purpose of this study was twofold. First, wedetermined the frequency distribution of C34T AMPD1 geno-types in a group of top-level Caucasian (Spanish) enduranceathletes (cyclists and runners). This group was compared withrandomly selected Caucasian (Spanish) healthy (asymptomatic)nonathletes. The second aim of this study was to compare com-mon laboratory indexes of endurance performance (VO2 max orventilatory thresholds) within the group of athletes dependingon their C34T AMPD1 genotype. Our hypothesis was that, inelite endurance athletes, the frequency distribution of themutant T allele is lower than in the general population, andparticularly the TT genotype makes achievement of elite statusunlikely. Nevertheless, on the basis of the results of previousresearch (23, 27, 36), common indexes of endurance perfor-mance would be overall unaffected in heterozygous athletes(CT) who have reached the status of “elite athlete.”

METHODS

Subjects. Written consent was obtained from each subject and thestudy was approved by the institutional ethics committee (UniversidadEuropea de Madrid, Spain).

Our sample comprised 104 elite endurance athletes, i.e., 50 unre-lated top-level male Spanish riders from the four best professionalcycling teams who ranked among the top 65–70 Spanish cyclists interms of a 3-wk stage race performance in this country and 54 runnerswho were the best Olympic-class Spanish male runners [specialists inmiddle-distance events (1,500 m), 5,000 and 10,000 m track races,3,000-m steeplechase events, or marathon] as determined by actualperformance during international competitions for the 1999–2004 period.

We specifically chose professional cyclists for this study who metthe following criteria: 1) being enrolled in a professional cycling team,2) having at least 2 yr experience in the professional category of theInternational Cycling Union, and 3) having participated in (andfinished) one or more classic 3-wk stage races between 1999 to 2004.During this period, 11 cyclists won at least one mass-start stage ortime trial of the Tour, Giro, or Vuelta, and 7 and 11 were among top-3and top-10 finishers, respectively, in these main 3-wk races.

All of the runners participate in cross-country races (usual distanceof 10,000–12,000 m) during winter months (including world cross-country championships for some of them). Irrespective of their spe-cialty, the training loads of these 54 subjects typically include �150km/wk (150–200 km/wk) and approach 250 km/wk in marathonersduring some periods of the year. Among the most important compe-tition awards of our runners during the last years are the following:European champion (1,500 m, 3,000-m steeplechase, 5,000 m, ormarathon), world champion, and top 3 in world championships (e.g.,marathon, 1,500 m, or 3,000-m steeplechase), or Olympic medalistand finalist in Olympic Games (1,500 m, 3,000-m steeplechase, 5,000m, 10,000 m, or marathon).

The mean (�SD) age, height, and mass of all the athletes was:27 � 4 yr, 176.6 � 6.2 cm, and 64.3 � 6.8 kg, respectively.

A group of 100 sedentary unrelated, healthy Spanish nonathletes(60 men, 40 women, aged 18–50 yr) without symptoms of metabolicmuscle diseases served as controls.

Ethnic and geographic origin of subjects. All the control subjectsare Spanish Caucasians of the same ethnic and geographic origin [i.e.,from Castile (the main, central area of Spain)]. Concerning athletes,58.7% are also from Castile (and 3.8% are from areas geographicallylocated next to Castile as Aragon or Extremadura), 12.5% are from theMediterranean area, 10.6% are from the Southern part of Spain(Andalucıa), 5.8% are non-Basques from the Atlantic, Northern part

of Spain (e.g., Asturias), and 8.6% are of Basque origin. Althoughprevious research has evidenced some peculiar genetic characteristicsin the Basque population compared with the rest of Spaniards andother European Caucasians, e.g., in human leukocyte antigen (HLA)gene frequencies (4), to the best of our knowledge no study hasreported differences in the genes regulating muscle metabolism be-tween Basque and non-Basque Spaniards, nor between differentSpanish, non-Basque subpopulations. Thus both groups of controlsand athletes were comparable in terms of ethnic origin, especiallywhen considering that the percent of Basque subjects was 0% and�9% in controls and athletes, respectively.

Determination of indexes of endurance performance. During the2001 period, we measured the gas exchange of each athlete duringexercise tests until exhaustion (see below) using a face mask appara-tus attached to a continuous, breath-by-breath monitoring system(Oxycon Champion System; Jaeger, Wuerzburg, Germany). The cy-clists and runners were evaluated while pedaling on a cycle ergometer(Ergometrics 900; Ergo-line; Bitz, Germany) or running on a treadmill(Technogym Run Race 1400 HC, Gambettola, Italy), respectively,after an incremental protocol until exhaustion, i.e., workload increasesof 25 W/min starting at 25 W for the cyclists and of 1 km/h startingat 8 km/h (with constant 1% upgrade) for the runners.

The following variables were measured during each test: oxygenuptake (VO2), pulmonary ventilation (VE), ventilatory equivalents foroxygen (VE/VO2) and carbon dioxide (VE/CO2), and end-tidal partialpressure of oxygen (PETO2) and carbon dioxide (PETCO2).

VO2 max was recorded as the highest VO2 value obtained for anycontinuous 1-min period during the tests. At least two of the followingcriteria were also required for the attainment of VO2 max: a plateau inVO2 values despite increasing workload, a respiratory exchange ratio�1.15 or the attainment of a maximal heart rate value above 95% ofthe age-predicted maximum. The ventilatory threshold (VT) wasdetermined by using the criteria of an increase in both VE/VO2 andPETO2 with no increase in VE/VCO2, whereas the respiratory compen-sation threshold (RCT) was determined by using the criteria of anincrease in both VE/VO2 and VE/VCO2 and a decrease in PETCO2 (19).Two independent observers detected VT and RCT. If there wasdisagreement, the opinion of a third investigator was obtained (19).

Genotype determinations. Genomic DNA was extracted from pe-ripheral blood anticoagulated with EDTA according to standard phe-nol-chloroform procedures followed by alcohol precipitation. To de-tect the C34T (C to T transition in nucleotide 34) in exon 2 ofAMPD1, a PCR fragment containing the mutation was amplified byusing the primers and PCR conditions indicated by Tsujino et al. (37).The fragment was digested with Mae II and electrophoresed through2% agarose gel. The wild-type PCR product is cleaved by the enzyme,whereas the mutant is not.

In those athletes showing the C34T mutation we also studied theG468T mutation of AMPD1 as previously described by Gross et al.(12). [This mutation has also been associated with deficiency ofmuscle AMPD and exercise-induced myalgia in Caucasians (12)].

Statistical analysis. The frequency of both the mutant T allele andheterozygous subjects with the C34T mutation was compared betweenthe two groups (athletes and nonathletes) by use of a z-test forbinomial population proportions. A Mann-Whitney’s test was used tocompare VO2 max, VT, and RCT levels in carriers and noncarriers ofthe C34T mutation, both within the total group of athletes and withineach of both subgroups of athletes (cyclists, on one hand, and runners,on the other). The level of significance was set at 0.05.

RESULTS

C34T AMPD1 genotype distributions. The T allelic fre-quency in our control population (8.5%) was overall similar tothat previously reported in large samples of white Caucasians,e.g., 10.8% of 503 subjects in the study by Rico-Sanz et al.(27), but significantly higher (�50%) than in the group of

2109AMPD AND SPORTS PERFORMANCE

J Appl Physiol • VOL 98 • JUNE 2005 • www.jap.org

athletes (4.3%) (P � 0.05) (Fig. 1). No homozygosity (TT)individual for the C34T mutation was found among the twogroups. [Previous research from our laboratory, however, hasrevealed a 1.5% prevalence of homozygosity for this mutationin a larger population sample (n � 400) of individuals withclinical suspicion of metabolic myopathy (28)]. The distribu-tion frequency of heterozygous (CT) genotypes for the C34Tmutation in the control group (17%) was similar to the distributionreported in the literature for Caucasian populations, i.e., �1 ofevery 5 individuals (21, 22, 27, 38), but was �50% higher (P �0.05) than in athletes (8.7%), i.e., �1 of every 12 athletes.

Clinical and performance history in genotyped athletes.Nine athletes (4 cyclists and 5 runners) were heterozygous forthe C34T mutation. None carried the G468T mutation ofAMPD1 previously described by Gross et al. (12). All of themhad successfully pursued their sporting careers, e.g., winner ofthe Young Riders’ classification in the Tour de France, top 9(1st non-African) in 5,000 m (World Championships) and 2ndin European cross-country championships, or Europe cham-pion in 3,000-m steeplechase. During the last years, theirresting blood levels of total creatine kinase levels measuredseveral times during the season (in noncompetition days) haveconsistently ranged within normal limits (�167 IU/l), i.e.,reflecting no excessive muscle damage (7).

One of the runners had suffered an episode of acute liver andrenal failure [due to a combination of both excessive training(overtraining) and severe viral infection] that required hospi-talization. He did recover adequately (i.e., to reach an Olympicfinal) after a resting period of a few months.

Six subjects had reported no previous cramps, abnormalmyalgia, and/or delayed muscle soreness or limited exercisecapacity during their career except two cyclists. One of themcomplained of powerless feeling and mild leg pain (with nocramps) during the last season of his career. These symptomswere due to endofibrosis of the external iliac artery, a diseasethat has been described in highly trained athletes, resulting insignificantly reduced blood flow to working muscles (1). An-other cyclist reported both increased myalgia and musclespasms in several competitions and frequent episodes of pain-ful contractures in the dorsum of the feet. His levels of thyroidhormones (total triiodothyronine, total thyroxin) and thyroid-stimulating hormone, growth hormone, cortisol, and testoster-one were consistently within normal limits during the lastseasons. Thus the possibility that his clinical symptoms hadbeen induced by partial AMP deficiency is not to be ruled out.Further physiological evaluations were performed in this cy-clist, including a surface EMG record of his vastus lateralis andrectus femoris muscles during the aforementioned type of

incremental cycle-ergometer test and a 20-min constant-loadtest at 400 W. Both tests, however, showed excellent adapta-tion to endurance exercise, i.e., a significant upward shift inEMG activity after a power output of 400 W was surpassed[indicating neuromuscular fatigue (17) to clearly occur onlyafter 400 W was surpassed during this test], and a high valueof gross mechanical efficiency (24%) (18).

VO2 max, VT, and RCT. The main characteristics of laboratoryindexes of endurance performance are shown in Table 1. Nosignificant differences in age, height, body mass, VO2 max, VT,or RCT were found between homozygous or heterozygousathletes (P � 0.05). Similarly, no significant differences werefound in these variables between homozygous and heterozy-gous athletes within each group of cyclists and runners, respec-tively (P � 0.05).

DISCUSSION

In this study, we have studied for the first time the frequencydistribution of C34T AMPD1 genotypes in top-level enduranceathletes. The main finding of our investigation was twofold: 1)in Caucasian elite endurance athletes the frequency distributionof the mutant T allele is lower (�50%) than in nonathletes, but,2) in general, the T allele does not seem to be associated tolimited endurance performance once the elite-level status hasbeen reached in sports.

One potential limitation from our investigation stems fromthe fact that we did not study athletes engaging in short-term,supramaximal (�110% VO2 max) exercise (e.g., 400-m trackrunning races or short cycling events in a velodrome). Indeed,the activity of AMPD is especially important during short,supramaximal efforts (e.g., 110% VO2 max), inducing a markeddepletion of muscle PCr and a fall in the total adenine nucle-otide pool (ATP � ADP � AMP) (8). Nevertheless, severalstudies on healthy humans have noted the accumulation of IMPto occur at fatigue during prolonged, submaximal exercise(e.g., �1 h at 70–75%VO2 max), particularly in the presence oflow intramuscular glycogen stores by the end of exercise (2,24, 25 33). For instance, Norman et al. (25) studied changes inmuscle energy state during prolonged exercise (�70%VO2 max)until exhaustion in healthy men. Muscle biopsies were ob-tained at rest, after 15 and 45 min of exercise, and at exhaustionand analyzed for ATP, ADP, AMP, IMP, hypoxanthine, andPCr content. Glycogen content at exhaustion was �30% of thepreexercise level. The PCr content decreased steeply during the

Table 1. Main characteristics of athletes homozygous withno mutation (CC) or heterozygous (CT) for theC34T mutation of AMPD1

CC (n � 95) CT (n � 9)

Age, yr 27�4 26�4Height, cm 176.7�6.1 175.9�7.2Mass, kg 64.5�6.6 62.5�8.9VO2 max, ml�kg�1�min�1 73.5�5.9 72.3�4.6%VO2 max at VT 68.4�6.4 67.1�4.9VO2 at VT, ml�kg�1�min�1 50.3�5.3 48.5�5.0%VO2 max at RCT 86.9�5.1 85.9�4.9VO2 at RCT, ml�kg�1�min�1 63.9�6.6 62.1�4.4

Values are means � SD. VO2, oxygen uptake; VO2 max, maximal VO2; VT,ventilatory threshold; RCT, respiratory compensation threshold. No significantdifferences between means were found in VO2 max, VT, or RCT (P � 0.05).

Fig. 1. Frequency distributions of the C (no mutation) and T (C34T mutation)alleles of among athletes (n � 104) and nonathletes (n � 100). *P � 0.05 forathletes vs. nonathletes.

2110 AMPD AND SPORTS PERFORMANCE

J Appl Physiol • VOL 98 • JUNE 2005 • www.jap.org

first 15 min of exercise and continued to decrease during therest of the exercise period (P � 0.05). Pronounced increases(64%) in contents of IMP (P � 0.001) were found whenexhaustion was approaching. Furthermore, energy charge[(ATP � 0.5 ADP)/(ATP � ADP � AMP)] was decreased atexhaustion (P � 0.05). The increases in IMP and hypoxanthinethat occurred when exhaustion was approaching during pro-longed submaximal exercise together with the decrease inenergy charge during this phase of exercise suggest a failure ofthe exercising skeletal muscle to regenerate ATP at exhaustion.Glycogen depletion and the aforementioned phenomena at themuscle level could also occur at the end of top-level endurancecompetitions, e.g., daily stages throughout 3-wk cycling races,each of which lasts 5 or more hours but might include somebouts (�20–30 min) of very intense (�90% VO2 max) exercise,particularly high mountain stages (16). In our athlete sample,there were also world-class specialists in middle-distance trackevents (1,500 m). In this competition, elite runners mustperform well above 100% VO2 max (i.e., with a total anaerobiccontribution as high as 20%) (14). Among our subjects wereOlympic-level specialists in 3,000-m steeplechase, 5,000 and10,000 m, which are events requiring a continuous near-maximal effort (�90% VO2 max) (5, 39). Even elite maratho-ners such as the ones studied here must run consistently at�80–90% of VO2 max to achieve successful competitive per-formances (13).

On the other hand, the AMPD reaction is also the majorcontributor to the production of ammonia, a biochemical indi-cator of the intensity of exercise (10, 15). In this regard,previous research using different workloads has reported sig-nificant increases in ammonia levels to occur at high, yetsubmaximal intensities (i.e., �VT) (26). In fact, the so-called“ammonia threshold” (i.e., nonlinear increases in ammonialevels once the VT or lactate threshold has been reached) hasbeen described during gradual testing (40). Dudley et al. (8)reported significant increases in ammonia levels after exerciseabove 85% VO2 max. Thus AMPD could also play a veryimportant role in regulating muscle metabolism during theendurance, yet intense events in which our elite athletes se-lected as subjects participate.

The lack of significant difference in indicators of enduranceperformance between CT and CC genotypes found here is inoverall agreement with previous research with nonathletes.Norman et al. (23) showed short-term, high-intensity exerciseperformance (30-s Wingate test) to be unaffected by AMPD1genotypes. At least in their exercise model, AMPD deficiencywould cause higher oxidative metabolism during exertion, as aresult of the increased adenosine levels enhancing blood flowand increased ADP levels stimulating oxidative phosphoryla-tion, respectively, which in turn would compensate for thedecreased purine nucleotide cycling associated with the Tallele (23). Tarnopolsky et al. (36) found no significant differ-ences among AMPD1 genotypes in time to exhaustion duringincremental cycle ergometry tests; e.g., the mean value ofheterozygous subjects was 94% of CC homozygous individu-als. It was concluded that complete (TT) or partial AMPDdeficiency (CT) does not affect tricarboxylic acid cycle anaple-rosis, PCr hydrolysis, or cellular energy charge during exhaus-tive exercise. In contrast, the capacity for repetitive submaxi-mal isometric muscle contractions (determined during a 20-min test of repetitive voluntary isometric contractions at 40%

of maximal force-generating capacity) has been shown to bereduced in AMPD-deficient subjects (TT genotype) (6). In linewith our findings, Rico-Sanz et al. (27) found no significantdifferences between CC and CT genotypes in the VO2 max

values attained by previously sedentary individuals duringincremental testing. Their pioneer study was the only one todescribe the effects of endurance exercise training (20 wk)among the different C34T AMPD1 genotypes. Interestingly,the training-induced increase in VO2 max was significantlyhigher (P � 0.006) in heterozygous than in homozygous withno C34T mutation, although no explanation for this findingwas mentioned in their report. In this regard, one could hy-pothesize that, during the natural selection process to reach thestatus of elite athletic competition, the partial metabolic defi-ciency in heterozygous elite athletes (i.e., decreased purinenucleotide cycling) is compensated for by other training adap-tations such as increased blood flow and oxidative phosphor-ylation in working muscles because of higher levels of aden-osine and ADP, respectively. Another explanation for the lackof differences in VO2 max among homozygous and heterozy-gous athletes might derive from the fact that, in endurance-trained humans, maximal endurance performance and VO2 max

are largely constrained by the oxygen delivery to workingmuscles (particularly, cardiac pump capacity), more so than byperipheral metabolic factors (i.e., at the muscle level) (3). Thusany partial metabolic limitation at the peripheral muscle level(e.g., decreased purine nucleotide cycling) might not signifi-cantly affect the VO2 max values of top-level athletes. On theother hand, we did not find any evidence of alteration in twoother powerful indicators of endurance capacity, VT andRCT (20). Both thresholds are good indicators of the adaptationsthat occur after endurance training, mostly at the peripheralmuscle level, i.e., improved buffer capacity, increased muscleoxidative capacity, or higher fatigue tolerance of type I (oxidative)fibers before type II (glycolytic) fibers are recruited. Extensiveresearch has shown that both variables are related with competi-tive performance in endurance sports as cycling or running [seeMeyer et al. (20) for a review]. Although our design is limited bythe fact that we did not perform muscle biopsies, the high VO2 max

values (mean of �73 ml �kg�1 �min�1 and consistently �65ml �kg�1 �min�1) and the very high workloads at which both VTand RCT occurred in heterozygous athletes (�67%VO2 max and86% VO2 max, respectively) seem to suggest that partial AMPDdeficiency does not impair aerobic energy production. No suchhigh values of VT, RCT, or VO2 max have been reported to date incarriers of the T allele. Finally, some hypotheses that explain thefact that AMPD deficiency is not necessarily associated withclinical symptoms or reduced exercise capacity could also helpexplain the lack of detrimental effect of this mutation in theperformance of our heterozygous athletes. The most likely hy-pothesis is alternative splicing of exon 2 harboring the mutation(11). [Because exon 2 is only 12 nucleotides in length (4 times 3),skipping of exon 2 would not affect the reading frame (11)].

Although no significant differences were found in commonindicators of endurance performance between genotypes, cau-tion must be taken when extrapolating laboratory data to actualcompetition. Indeed, small physiological differences mightresult in significant changes in competition performance, assmall differences (�2–3% in terms of performance time)usually make the difference between winning and losing anOlympic final or a cycling main race. In addition, the distri-

2111AMPD AND SPORTS PERFORMANCE

J Appl Physiol • VOL 98 • JUNE 2005 • www.jap.org

bution frequency of the T allele was lower in athletes than innonathletes. In this regard, further research is necessary tocorroborate that the C34T is harmless in terms of top-levelperformance. For instance, future studies might discernwhether humans who are homozygous for the C34T mutationcan also reach the status top-level endurance performance.Further research is also needed with elite athletes participatingin shorter, supramaximal (�100%VO2 max) events.

In conclusion, although the frequency distribution of themutant T allele of the C34T AMPD1 genotype seems to belower in Caucasian elite endurance athletes than in controls,this mutation does not appear to affect endurance performanceonce the elite-level status has been reached in sports.

ACKNOWLEDGMENTS

We express gratitude to the athletes who participated in this study and to thephysicians who collected blood from the athletes (Pedro Celaya, AlfredoCordova, Jesus Hoyos, and Jose G. Villa).

GRANTS

This study was partially financed by a grant from the Consejo Superior deDeportes (02/UPR10/04).

REFERENCES

1. Abraham P, Bouye P, Quere I, Chevalier JM, and Saumet JL. Past,present and future of arterial endofibrosis in athletes: a point of view.Sports Med 34: 419–425, 2004.

2. Broberg S and Sahlin K. Adenine nucleotide degradation in human skeletalmuscle during prolonged exercise. J Appl Physiol 67: 116–122, 1989.

3. Calbet JAL, Jensen-Urstad M, van Hall G, Holmberg H-C, RosdahlH, and Saltin B. Maximal muscular vascular conductance during wholebody upright exercise in humans. J Physiol 558: 319–331, 2004.

4. Calderon R, Wentzel J, and Robertz DF. HLA frequencies in Basques inSpain and in neighbouring populations. Ann Hum Biol 20: 109–120, 1993.

5. Costill DL. Inside Running: Basics of Sports Phyiology. Indianapolis, IN:Benchmark, 1986, p. 33.

6. De Ruiter CJ, May AM, van Engelen BG, Wevers RA, Steenbergen-Spanjers GC, and de Haan A. Muscle function during repetitive mod-erate-intensity muscle contractions in myoadenylate deaminase-deficientDutch subjects. Clin Sci (Lond) 102: 531–539, 2002.

7. Donnelly AE, Clarkson PM, and Maughan RJ. Exercise-induced mus-cle damage: effects of light exercise on damaged muscle. Eur J ApplPhysiol 64: 350–353, 1992.

8. Dudley GA, Staron RS, Murray TF, Hagerman FC, and Luginbuhl A.Muscle fiber composition and blood ammonia levels after intense exercisein humans. J Appl Physiol 54: 582–586, 1983.

9. Fishbein WN, Armbrustmacher VW, and Griffin JL. Myoadenylatedeaminase deficiency: a new disease of muscle. Science 200: 545–548, 1978.

10. Fishbein WN, Foellmer JW, and Davis JI. Medical implications of thelactate and ammonia relationship in anaerobic exercise. Int J Sports Med11, Suppl 2: S91–S100, 1990.

11. Gross M. Clinical heterogeneity and molecular mechanisms in inbornmuscle AMP deaminase deficiency. J Inherit Metab Dis 20: 186–192, 1997.

12. Gross M, Rotzer E, Kolle P, Mortier W, Reichmann H, Goebel HH,Lochmuller H, Pongratz D, Mahnke-Zizelman DK, and Sabina RL.G468-T AMPD1 mutant allele contributes to the high incidence ofmyoadenylate deaminase deficiency in the Caucasian population. Neuro-muscul Disord 12: 558–565, 2002.

13. Hagerman FC. Energy metabolism and fuel utilization. Med Sci SportsExerc 24, Suppl 9: S309–S314, 1992.

14. Hill DW. Energy system contributions in middle-distance running events.J Sports Sci 17: 477–483, 1999.

15. Lowenstein JM. Ammonia production in muscle and other tissues: thepurine nucleotide cycle. Physiol Rev 52: 382–414, 1972.

16. Lucia A, Earnest C, and Arribas C. The Tour de France: a physiologicalreview. Scand J Med Sci Sports 13: 275–283, 2003.

17. Lucia A, Pardo J, Durantez A, Hoyos J, and Chicharro JL. Physio-logical differences between professional and elite road cyclists. IntJ Sports Med 19: 342–348, 1998.

18. Lucia A, Hoyos J, Perez M, Santalla A, and Chicharro JL. Inverserelationship between VO2 max and economy/efficiency in world-class cy-clists. Med Sci Sports Exerc 34: 2079–2084, 2002.

19. Lucia A, Hoyos J, Santalla A, Earnest C, and Chicharro JL. Tour deFrance versus Vuelta a Espana: which is harder? Med Sci Sports Exerc 35:872–878, 2003.

20. Meyer T, Lucia A, Earnest CP, and Kindermann W. A conceptualframework for performance diagnosis and training prescription fromsubmaximal parameters—theory and application. Int J Sports Med, Suppl1: S38–S48, 2004.

21. Morisaki T, Gross M, Morisaki H, Pongratz D, Zollner N, and HolmesEW. Molecular basis of AMP deaminase deficiency in skeletal muscle.Proc Natl Acad Sci USA 89: 6457–6461, 1992.

22. Norman B, Glenmark B, and Jansson E. Muscle AMP deaminasedeficiency in 2% of healthy population. Muscle Nerve 18: 239–241, 1995.

23. Norman B, Sabina RL, and Jansson E. Regulation of skeletal muscleATP catabolism by AMPD1 genotype during sprint exercise in asymp-tomatic subjects. J Appl Physiol 91: 258–264, 2001.

24. Norman B, Sollevi A, Kaijser L, and Jansson E. ATP breakdownproducts in human muscle during prolonged exercise to exhaustion. ClinPhysiol 7: 503–509, 1987.

25. Norman B, Sollevi A, and Jansson E. Increased IMP content in glyco-gen-depleted muscle fibres during submaximal exercise in man. ActaPhysiol Scand 133: 97–100, 1988.

26. Ogino K, Kinugawa T, Osaki SM, Kato M, Endoh A, Furuse Y,Uchida K, Shimoyama M, Igawa O, Hisatome I, and Shigemasa C.Ammonia response to constant exercise: differences to the lactate re-sponse. Clin Exp Pharmacol Physiol 27: 612–617, 2000.

27. Rico-Sanz J, Rankinen T, Joanisse DR, Leon AS, Skinner JS, Wil-more JH, Rao DC, and Bouchard C. Associations between cardiorespi-ratory responses to exercise and the C34T AMPD1 gene polymorphism inthe HERITAGE Family study. Physiol Genomics 14: 161–166, 2003.

28. Rubio JC, Martın MA, Del Hoyo P, Bautista J, Campos Y, Segura D,Navarro C, Ricoy JR, Cabello A, and Arenas J. Molecular analysis ofSpanish patients with AMP deaminase deficiency. Muscle Nerve 23:1175–1178, 2000.

29. Sabina R. Myoadenylate deaminase deficiency. Neurol Clin 18: 185–194,2000.

30. Sabina RL, Morisaki T, Clarke P, Eddy R, Shows TB, Morton CC,and Holmes EW. Characterization of the human and rat myoadenylatedeaminase genes. J Biol Chem 265: 9423–9433, 1990.

31. Sabina RL, Swain JL, and Colmes EW. Myoadenylate deaminasedeficiency. In: The Metabolic and Molecular Bases of Inherited Disease(7th ed.), edited by Scriver CR, Beaudet AL, Sly WS, and Walle D. NewYork: McGraw-Hill, 1995, p. 1769–1780.

32. Sabina RL, Swain JL, Olanow CW, Bradley WG, Fishbein WN,DiMauro S, and Holmes EW. Myoadenylate deaminase deficiency.Functional and metabolic abnormalities associated with disruption of thepurine nucleotide cycle. J Clin Invest 73: 720–730, 1984.

33. Sahlin K, Katz A, and Broberg S. Tricarboxylic acid cycle intermediatesin human muscle during prolonged exercise. Am J Physiol Cell Physiol259: C834–C841, 1990.

34. Sinkeler SP, Binkhorst RA, Joosten EM, Wevers RA, CoerwinkeiMM, and Oei TL. AMP deaminase deficiency: study of the humanskeletal muscle purine metabolism during ischaemic isometric exercise.Clin Sci (Lond) 72: 475–482, 1987.

35. Spencer MK, Yan Z, and Katz A. Effect of low glycogen on carbohy-drate and energy metabolism in human muscle during exercise. Am JPhysiol Cell Physiol 262: C975–C979, 1992.

36. Tarnopolsky MA, Parise G, Gibala MJ, Graham TE, and Rush JW.Myoadenylate deaminase deficiency does not affect muscle anaplerosisduring exhaustive exercise in humans. J Physiol 533: 881–889, 2001.

37. Tsujino S, Shanske S, Carroll JE, Sabina RL, and DiMauro S. Doubletrouble: combined myophosphorylase and AMP deaminase deficiency in achild homozygous for nonsense mutations at both loci. NeuromusculDisord 5: 263–266, 1995.

38. Verzijl HT, van Engelen BG, Luyten JA, Steenbergen GC, van denHeuvel LP, ter Laak HJ, Padberg GW, and Webers RA. Geneticcharacteristics of myoadenylate deaminase deficiency. Ann Neurol 44:140–143, 1998.

39. Weston AR, Mbambo Z, and Myburgh KH. Running economy of Africanand Caucasian distance runners. Med Sci Sports Exerc 32: 1130–1134, 2000.

40. Yuan Y and Chan KM. A longitudinal study on the ammonia thresholdin junior cyclists. Br J Sports Med 38: 115–119, 2004.

2112 AMPD AND SPORTS PERFORMANCE

J Appl Physiol • VOL 98 • JUNE 2005 • www.jap.org