heart weight has been reported in greyhounds (r = 0.73, ... the ventricular myocardium and should therefore be ac- .... and body weight in greyhound dogs. Aust.
Athletic heart syndrome in dogs competing in a long-distance sled race PETER D. CONSTABLE, KENNETH W. HINCHCLIFF, JEANNE OLSON, AND ROBERT L. HAMLIN College of Veterinary Medicine, The Ohio State University, Columbus, Ohio 43210 Constable, Olson, and
Peter Robert
D., Kenneth W. Hinchcliff, Jeanne L. Hamlin. Athletic heart syndrome in in a long-distance sled race. J. Appl. Physiol.
dogs competing 76(l): 433-438,1994.-The cardiac effects of endurance training were evaluated by cardiac auscultation and electrocardiographic examination of 48 heavily trained sled dogs (3,0005,000 km of training), 18 lightly trained sled dogs (300-800 km of training), 19 untrained sled dogs, and 14 mongrel dogs. A grade I-II/VI early- to midsystolic cardiac murmur was auscultated with increasing frequency as training level increased. The QRS duration (66.1 t 7.4 ms) and QT interval (236 ~fr:20 ms) were significantly (P < 0.05) longer in heavily trained sled dogs than in mongrel dogs (QRS, 60.6 t 4.6; QT, 219 t 11 ms). A long QT interval (>250 ms) was observed in 8 (16.7%) heavily trained dogs but not in the other groups. A significant rightward shift in the mean electrical axis of ventricular depolarization in the frontal plane was observed in heavily trained sled dogs. The auscultatory and electrocardiographic findings in heavily trained sled dogs were remarkably similar to those reported in elite human endurance athletes, suggesting that endurance-trained sled dogs provide a naturally occurring model for the athletic heart syndrome. auscultation; cardiac murmur; electrocardiography; pertrophy; QT interval
cardiac hy-
involvedin endurance events have auscultable cardiac murmurs, bradyarrhythmias, cardiac hypertrophy, and electrocardiographic evidence of vagotonia and cardiac chamber enlargement (17, 38). This constellation of “abnormalities” has been termed the athletic heart syndrome, and the changes are thought to reflect functional and morphological adaptations of the athlete’s heart in response to sustained endurance exercise rather than the presence of cardiovascular disease (17). In many respects, the athletic heart syndrome represents an intraspecies example of physiological adaptation in that mammalian species that have depended on endurance for their evolutionary survival have a relatively large heart and slow resting heart rate (8, 1% Although experimental intraspecies studies in dogs, rats, and horses demonstrated that endurance training is accompanied by cardiac hypertrophy and bradycardia (2, 21, 29,30, 34, 35, 37), a naturally occurring intraspecies phenomenon comparable to the athletic heart syndrome in humans has not been well documented in animals. We therefore studied selected cardiac factors in Alaskan sled dogs, a breed that has been associated with legendary feats of stamina and endurance (36). The main purpose of this study was to determine whether cardiac changes indicative of the athletic heart syndrome exist in dogs trained for endurance events. MANY ELITE HUMAN ATHLETES
0161-7567/94
$3.00 Copyright
MATERIALS
AND METHODS
Dogs. Four groups of dogs of similar body weight, age, and conformation, but different levels of exercise training, were examined by thoracic auscultation and electrocardiography. The first group of dogs (n = 48) represented elite canine endurance athletes (13, 36) consisting of heavily trained (HT) sled dogs participating in the 1992 Yukon Quest International sled dog race. This race is 1,600 km long and is held in February each year between Fairbanks, AK and Whitehorse, Yukon, Canada. The dogs were in a team of 8-12 animals, pulled a sled with a musher, and were reported to have completed 3,000~5,000 km of training and racing in the winter before examination. The ambient temperature during the race ranged from -43 to -9”C, and the dogs averaged 5.5 km/h for the duration of the race. The dogs were examined after they had been fed, watered, and rested for 12-48 h after running in the Yukon Quest, when serum electrolyte, protein, and creatinine concentrations are within the normal range (16). The mushers were asked to categorize each dog in their team as 1) above average for the team or 2) average or below average for the team. The second group of dogs (n = 18) contained lightly trained (LT) sled dogs located in Fairbanks and Eagle, AK. They had completed 300-800 km of training in the winter before examination, were in good health, and had not been exercised for 224 h before examination. These dogs were used for sled dog races of ~100 km and for recreational sledding. The third group of dogs (n = 19) consisted of heartworm-free untrained (UT) sled dogs located in Yellow Springs, OH. They were in good health, were kept in a dog run, and had not been exercised for 212 mo before examination. The fourth group of dogs (n = 14) consisted of heartwormfree mongrel dogs located in Columbus, OH that were in good health, had not been exercised for 248 h before examination, and had a normal left ventricular end-diastolic pressure, mean arterial pressure, right atria1 pressure, pulmonary arterial pressure, heart rate, and cardiac output during subsequent unrelated studies performed under cu-chloralose anesthesia. The exercise history of these dogs was unknown, but none was a sled dog. Auscuhtion. The thorax of each dog was auscultated by one of the authors (PDC) to detect and characterize cardiac murmurs. Cardiac auscultation was performed in a quiet environment with the dog standing and not panting. When a cardiac murmur was detected, its intensity was graded on a scale of l-6 (1 l), the point of maximal intensity determined, the timing in relationship to the cardiac cycle estimated, and the nature of the murmur characterized. Ekctrocardiogmphy. Electrocardiograms were obtained with all dogs in a standing position, the forelimbs and hindlimbs perpendicular to the ground, all legs separate, and the head and neck parallel to the ground and in the same plane as the long axis of the body. The electrocardiographic leads were attached to the dog’s skin by alligator clips at the palmar aspect of the left and right foreleg over the olecranon; at the cranial aspect of the left and right hindlimb over the patella ligament; over the dorsal spinous process of the seventh thoracic vertebra (lead
0 1994 the American
Physiological
Society
433
434
ATHLETIC
HEART
SYNDROME
IN DOGS
tude >0.4 mV in lead II (right atria1 enlargement); PR interval VlO); and over the apex beat at the left lateral thorax, level >13O ms (1st degree atrioventricular block); R wave amplitude with the olecranon (lead V3). The points of electrode attach>3.O mV in lead II and ~2.5 mV in lead aVF (left ventricular ment were saturated with 70% alcohol, and the electrocardioenlargement); notched QRS complex (notching in the upstroke gram was monitored for 230 s for rhythm disturbances before being recorded by a canine electrocardiograph (model EL-l, or downstroke of the R or S wave in lead V3); ST segment depression or >0.2-mV elevation in lead II, III, or aVF (abnorMortara Instrument, Milwaukee, WI). This battery-operated portable machine provides a real-time image of the electrocarmal); and QT interval ~250 ms (abnormal). Right ventricular diogram on a liquid crystal screen. Heating packs were applied enlargement, right bundle branch block, left bundle branch to the batteries and the liquid crystal screen when the unit was block, and incomplete right bundle branch block were diagused in subfreezing temperatures. When directed, the electronosed using established criteria (22, 23). cardiographic unit will record and store in digitized form sigBecause heart rate variability has been used as a noninvasive nals from leads I, II, III, aVR, aVL, aVF, V3, and VlO obtained index of cardiac autonomic tone (10,18), the natural logarithm over a 10-s period. The resultant electrocardiogram was of heart period variance was determined from the 10-s lead II printed at 25 and 50 mm/s by an X-Y plotter (Hewlett Packard recording. This variable was examined to test the hypothesis 7470A), and data were extracted from the digital record for that endurance training was associated with changes in autosubsequent analysis. nomic tone. Data analysis. The electrocardiographic unit uses a threshStutisticuZ analysis. Categorical data were compared by the x2 old value for rate of change in the voltage signal to determine test of association. Fisher’s exact test (PROC FREQUENCY, the onset and end of the P, QRS, and T waves in each lead. The SAS Institute) was used whenever the expected cell count in duration of the QRS wave and PQ or QT intervals was deter>20% of the cells was ~5 (27). mined by the electrocardiographic unit from the time that the P Continuous data were analyzed by one-way analysis of varior QRS wave starts in any lead to the time that the QRS or T ance. The variance, kurtosis, and skewness of continuous variwave ends in any lead. The QRS duration, PQ interval, and QT ables were evaluated using the PROC UNIVARIATE proceinterval values recorded by the electrocardiograph can theredure (SAS Institute), and variables with unequal variances or fore exceed those measured in any single lead but are suspected nonnormal distributions were log-transformed before analysis to provide a more accurate value. The electrocardiographic unit of variance was performed (27). Whenever the F ratio was also produced a printout of the P, Q, R, S, R’, S’, T, and net QRS found to exceed the critical value (P ~0.05) for each variable, wave amplitudes (in mV), the ST segment displacement at the comparisons were made between HT and UT sled dogs, LT and J point (in mV), and the duration of the Q, R, S, R’, and S’waves UT sled dogs, HT sled dogs and mongrel dogs, and UT sled (in ms) for each of the eight recorded leads. The mean electrical dogs and mongrel dogs with use of the Bonferroni inequality to axis of ventricular depolarization in the frontal plane was estimated by plotting the net value for QRS complexes from leads I maintain the experimentwise error rate at ~0.05. Pearson’s product-moment correlation coefficients (r) were calculated for and III and determining the direction of the resultant vector. continuous variables to explore the relationship between elecPopulation-based studies examining the association betrocardiographic variables and body weight, age, and heart tween the duration of electrical systole (represented by the QT rate. All continuous data are presented as means rt SD. interval) and heart rate demonstrated that the QT interval decreases in an approximately linear fashion with heart rate (1, 19). Accordingly, numerous disparate formulas have been proRESULTS posed as heart rate-independent measures for the duration of electrical systole. The most widely used method for correcting The ambient temperature during thoracic auscultation the QT interval for heart rate (QT,) is Bazett’s formula, where and electrocardiographic recording ranged from -30 to QT, = QT/@R, with QT and R-R intervals being measured in 22°C. The body weight of mongrel dogs was 24 t 2 kg. seconds (3). Unfortunately, QT, increases in an approximately The body weight of sled dogs ranged from 15 to 33 kg, linear fashion with heart rate, indicating that QT, is not a heart rate-independent measure (1). Because the magnitude of the with mean values for UT, LT, and HT sled dogs of 27 t 4, change in QT and QT, with changes in heart rate appears to be 23 t 5, and 26 t 3 kg, respectively. The age of mongrel equal and opposite, we speculated that the average of the QT dogs was unknown, but all animals were ~1 yr old. The and QT, values for each dog would provide a heart rate-indeage of sled dogs ranged from 18 mo to 8 yr, with mean pendent measure for the duration of electrical systole. Accordvalues for UT, LT, and HT sled dogs of 5 t 2,5 t 2, and ingly, a new QT index was developed such that QThde, = 4 t 2 yr, respectively. Body weight and age were poorly (QT + QTJZ. Although this index does not appear to have a correlated (r < 0.3) with other study variables and were sound physiological basis (19), it is derived from the two most therefore not considered to be confounding factors. commonly used indexes for the duration of electrical systole Auscultable cardiac murmurs were detected with in(QT and QT,). The relationship between electrical systole and creasing frequency as training level increased (Table 1). heart rate was also evaluated by the electrical diastole-systole In 17 HT sled dogs and all LT and UT sled dogs with (D/S) quotient, calculated from the (TQ interval)/(QT interval) quotient, where the QT interval represents the duration of auscultable cardiac murmurs, the murmur was a grade electrical systole and the TQ interval represents the duration I-to-II early- to midsystolic crescendo-decrescendo murof electrical diastole, and QT + TQ = R-R (19). mur, with the point of maximal intensity being the left The following criteria were used to categorize the electrocarheart base. Three HT sled dogs also had a similar murdiographic findings (21): heart rate ~70 beats/min (bradycarmur of lower intensity detected at the right heart base. dia); heart rate ~160 beats/min (tachycardia); variation of One HT sled dog had a grade II early-systolic murmur >lO% in R-R interval in lead II, such that (R-R,, - R-Rmti)/ heard best over the left heart apex, and another HT sled R-R,, > 0.1, unaccompanied by a change in P wave configuradog had a grade II early-systolic murmur that was louder tion (sinus arrhythmia); variation of >10% in R-R interval in over the right heart base than the left heart base. The lead II, accompanied by a change in P wave configuration and minor changes in the PR interval (wandering pacemaker); P performance level of HT sled dogs with cardiac murmurs (9 above average, 10 average or below average), as aswave duration >85 ms (left atria1 enlargement); P wave ampli-
ATHLETIC TABLE
1. Number
of mongrel
435
HEART SYNDROME IN DOGS
dogs and UT, LT, and HT sled dogs with abnormal
auscultatory
and ECG findings
Sled Dogs Mongrels (n = 14) Diagnosis
n
Systolic cardiac murmur Left atria1 enlargement Right atria1 enlargement AV block 1” 2” (Mobitz type I) Left ventricular enlargement Notched QRS in V3 ST segment abnormalities Prolonged QT interval
0 0 0
UT, untrained;
LT, lightly
UT (n = 19) %
0 0 0
%
n
%
n
%
P
3 2 0
15.8 10.5 0
5 3 1
27.8 16.7 5.6
19 8 5
39.6 16.7 10.4
0.012 0.426 0.440
0 0
6 1 17 3 1 8
12.5 2.1 35.4 6.3 2.1 16.7
0.383 0.767 0.003 0.773 0.579 0.041
15.8 5.3 0 5.3 0 0
HT, heavily trained;
in mongrel
QT
index
D/S quotient Mean electrical axis, O In (heart period variance), ms2
ST segment was depressed >0.2 mV in lead II in one HT and one LT sled dog. In both cases, the ST segment was also elevated >0.2 mV in lead V3. A prolonged QT interval (QT >250 ms) was present in eight (16.7%) HT sled dogs but no other groups (Table 1). The QT interval was significantly longer in HT sled dogs than in UT sled dogs and mongrel dogs, whereas QT, was significantly longer in HT sled dogs than in mongrel dogs (Table 2). The QT interval and QTindex were weakly correlated with QRS duration (r = 0.40, P QS, syndrome: a new mortality risk indicator in coronary artery disease. Am. J. Cardiol. 50: 1229-1235,1982. 7. Carew, T. E., and J. W. Covell. Left ventricular function in exercise-induced hypertrophy in dogs. Am. J. Cardiol. 42: 82-88, 1978. 8. Clark, A. J. Comparative Physiology of the Heart. New York: Macmillan, 1927, p. 71-86.
438
ATHLETIC
HEART
The QT interval during reflex adaptation. Circulation 69: 22-25, 1984. Eckberg, D. L. Human sinus arrhythmia as an index of vagal cardiac outflow. J. Appl. Physiol. 54: 961-966,1983. Freeman, A. R., and S. A. Levine. The clinical significance of the systolic murmur. A study of 1,000 consecutive “noncardiac” cases. Ann. Intern. Med. 6: 1371-1382,1933. Grande, F., and H. L. Taylor. Adaptive changes in the heart, vessels, and patterns of control under chronically high loads. In: Handbook of Physiology. Circulation. Washington, DC: Am. Physiol. Sot., 1965, sect. 2, vol. III, chapt. 74, p. 2615-2678. Grandjean, D., and B. M. Paragon. Nutrition of racing and working dogs. Part II. Determination of energy requirements and the nutritional impact of stress. Compr. Contin. Educ. Pratt. Vet.
9. Davidowski,
SYNDROME
right bundle-branch possible misnomer.
T. A., and S. Wolf.
cardiovascular
10. 11.
12.
13.
15: 45--T&1993. 14. Hamlin, R. L., and C. R. Smith.
basis for interpretation
Anatomical of the electrocardiogram.
and physiologic Am. J. Vet. Res.
21: 701-708,196O. 15. Herrmann, G. R. Experimental
heart disease. I. Method of dividing hearts with sectional and proportional weights and ratios for two hundred normal dogs’ hearts. Am. Heart J. 1: 213-231,1925.
16. Hinchcliff, W. Royle,
H. W., J. Olson, C. Crusberg, J. Kenyon, W. Weber, and J. Burr. Serum biochemical
dogs competing
202: 401-405,1993. 17. Huston, T. P., J. C. Puffer,
and R. W. M. MacMillan. The athletic heart syndrome. N. Engl. J. Med. 313: 24-32, 1985. 18. Katona, P. G., and F. Jih. Respiratory sinus arrhythmia: nonin-
vasive measure of parasympathetic id. 39: 801-805,1975. 19. Kovacs, S. J. The duration 20. 21. 22. 23.
24. Palatini, bardoni,
of ventricular
arrhythmias
enigma
A. Calzavera,
and
M. Li-
Prevalence and possible in athletes. Am. Heart J.
110: 560-567,1985. 25. Pantano, J. A. Reply to the letter “Arrhythmias and QT interval in athletes.” Am. Heart J. 107: 608, 1984. 26. Pocchiari, R. J., R. L. Hamlin, and S. A. McCune. Electrocar-
diographic
findings
in rats with cardiomyopathy.
Am. J. Vet. Res.
54: 607-611,1993. 27. SAS Institute. Users Guide. Cary, NC: SAS Inst., 1986, version 5. 28. Schneider, H. P., R. C. Truex, and J. K. Knowles. Compara-
tive observations of the hearts of mongrel Anat. Rec. 149: 173-180, 1968. 29. Stahl, W. R. Organ weights of primates Science Wash. DC 150: 1039~1042,1965.
and greyhound
dogs.
and other mammals.
30. Steel, J. D., R. I. Taylor, P. E. Davis, G. A. Stewart, and P. W. Salmon. Relationships between heart score, heart weight, and body weight in greyhound dogs. Aust. Vet. J. 52: 561-564,1976. 31. Steinhaus, A. H., R. W. Boyle, and T. A. Jenkins. The chronic
effects of running and swimming exercise on the hearts of growing dogs as determined electrocardiographically. Am. J. Physiol. 99: 503-511,1932. 32. Steinhaus, A. H., L. A. Hoyt,
33.
cardiac control. J. Appl. Phys-
of the QT interval as a function of heart rate: a derivation based on physical principles and a comparison to measured values. Am. Heart J. 110: 872-878,1985. Lepeschkin, E. The configuration of the T wave and the ventricular action potential in different species of mammals. Ann. NY Acad. Sci. 127: 170-177,1965. Marsland, W. P. Heart rate response to submaximal exercise in the Standardbred horse. J. Appl. Physiol. 24: 98-101, 1968. Miller, M. S., and L. P. Tilley. Electrocardiography. In: Canine and Feline Cardiology, edited by P. R. Fox. New York: Churchill Livingstone, 1988, p. 43-89. Moore, E. N., J. P. Boineau, and D. F. Patterson. Incomplete
block. An electrocardiographic Circulation 44: 678-687, 1971.
P., G. Maraglino, G. Sperti, A. C. Pessina, and C. D. Palu.
mechanisms
R. Long,
change in in a long distance sled race. J. Am. Vet. Med. Assoc.
IN DOGS
34. 35. 36.
and H. A. Rice. The effects of running and swimming on the organ weights of growing dogs. Am. J. Physiol. 99: 512-520, 1932. Stewart, G. A. The heart score theory in the racehorse. Au&. Vet. J. 57: 422-428, 1981. Stone, H. L. Cardiac function and exercise training in conscious dogs. J. Appl. Physiol. 42: 824-832, 1977. Tipton, C. M., R. J. Barnard, and T. K. Tcheng. Resting heart rate investigations with trained and nontrained hypophysectomized rats. J. Appt. Physiol. 26: 585-588, 1969. Van Citters, R. L., and D. L. Franklin. Cardiovascular performance of Alaska sled dogs during exercise. Circ. Res. 24: 33-42,
1969. 37. Wyatt,
Influences of physical training 35: 883-889,1974. 38. Zehender, M., T. Meinertz, J. Keul, and H. Just. ECG variants and cardiac arrhythmias in athletes: clinical relevance and prognostic importance. Am. Heart J. 119: 1378-1391,199O. H. L., and J. H. Mitchell.
on the heart of dogs. Circulation
