Discussion
The present study was conducted to investigate the association between V̇O2peak and MBF in ET and UT subjects and whether the degree of LV hypertrophy associates to these variables. We found that ET males with significantly higher V̇O2peak had decreased hyperemic MBF compared with that of UT male subjects. Furthermore, we observed that this decreased hyperemic MBF was inversely related to LV mass and the difference between the groups disappeared when the LV mass differences were controlled for.
The ET subjects had an increased SV but similar ejection fraction and cardiac output at rest compared with those of UT subjects. These are normal and comparable findings with those of several earlier reports. In addition, the ET subjects demonstrated significantly increased wall thicknesses and, thereby, increased LV mass compared with those of the UT subjects. The average LV mass (330 g) and LV mass index (169 g·m) in trained subjects are among the highest values reported earlier. We also found a strong association between LV variables and V̇O2peak, as observed in earlier studies when both groups were pooled together. In the subgroup analysis of the ET group, the trained subjects with the highest LV mass (or wall thickness) did not have superior V̇O2peak compared with that of the trained subjects with much lower mass. This indicates that pronounced myocardial hypertrophy is not a necessary adaptation to achieve a high V̇O2peak in trained subjects, as also suggested earlier, but has significant effect on myocardial vascular function, as suggested in the present study.
Basal MBF was similar in both groups, which is also in line with some but not all previous investigations. We also observed that the total LV work was similar between the ET and UT subjects, but when LV work was normalized with LV mass (LV work index), the relative LV work was almost 40% lower in the ET group. However, although cardiovascular resistance is increased in athletes aiming to limit myocardial overperfusion (oxygen supply vs demand), an athlete's heart seems to be at least slightly luxuriously perfused at rest, as basal MBF normalized for LV work was significantly higher in athletes. This reasoning is based on the fact that LV work is normally closely coupled with oxygen consumption and myocardial perfusion.
During adenosine infusion, the ET group had, on average, 23% lower absolute hyperemic myocardial perfusion than that in the UT group. This is in accordance with the studies by Heiss et al. and Heinonen et al., whereas some other investigations have observed both unchanged or increased hyperemic perfusion in endurance athletes. We have previously shown an inverse relation between hyperemic MBF and exercise performance (maximal power output during cycling) in highly trained cross-country skiers with pronounced cardiac hypertrophy. In the present study, V̇O2peak was inversely related to hyperemic perfusion not only in the entire study population but also in the ET group alone. Moreover, when the group comparisons were also adjusted for age (Table 3), the results remained essentially similar. First, this observation suggests that endurance training diminishes the pharmacologically induced hyperemic blood flow. Second, it also highlights the importance of controlling for aerobic power when studying myocardial reactivity in healthy subjects. Indeed, when the association between V̇O2peak and hyperemic blood flow was controlled for LV mass, both the group difference and correlation (both groups combined) were not significant anymore. On the other hand, within the ET group, correlation remained significant despite this adjustment.
The major finding in the present study was that the ET group had significantly lower hyperemic MBF compared with that in the UT group. This finding is similar with that observed in many pathophysiological states, e.g., in hypertension and in hypertrophic cardiomyopathy (HCM), in which reduced hyperemic MBF or perfusion reserve has been observed. In addition, our observation of higher perfusion resistance in ET subjects during hyperemic conditions is in line with the results from Knaapen et al., as performed with HCM patients. This suggests that the reduced hyperemic MBF is due to higher extravascular compressive forces. Because many anatomical changes in HCM are similar with those observed in athlete's heart. one can hypothesize that these changes would also be related to decreased perfusion reserve in ET subjects. In this regard, we found a trend toward a lower reserve in ET men, as measured in absolute terms. When the blood flow reserve was calculated as a ratio between MAP-normalized hyperemic and work-normalized basal blood flow, ET subjects demonstrated almost 50% lower reserve. Consequently, the alteration in hyperemic MBF in ET subjects with very pronounced LV hypertrophy may also have physiological or even clinical significance, but further research is warranted to investigate this possibility.
Mechanistically, it has been suggested that the reduced capillary density, which accompanies the increased LV mass, induces microvascular dysfunction in HCM. Myocardial perfusion measures, especially during hyperemia as also determined in the present study, serve as a noninvasive functional surrogate for capillary density. Studies in healthy animals have proposed that endurance training leads to increased density of coronary arterioles, whereas capillary density remains largely unchanged when cardiac hypertrophy remains minor. However, capillary density can also be decreased even in healthy animals when training has induced marked cardiac hypertrophy. Furthermore, although the size of coronary arteries is usually closely associated to the LV mass in healthy humans, evidence that the growth or enlargement of coronary vasculature does not necessarily always follow the rate of increase in myocardial mass exists, which may explain the finding of decreased hyperemic blood flow with increasing LV mass in the present study.
Importantly, rather than associated with the so-called concentric hypertrophy that is typical for many pathophysiological cardiac diseases such as hypertension, we found here that decreased hyperemic blood flow correlated negatively with variables that depict eccentric hypertrophy (larger cavity dimension) and not, for example, with RWT that is a good marker of concentric hypertrophy. Eccentric hypertrophy is a very typical feature for athletes exposed to high-volume loads during their training, like our trained subjects. Eccentric hypertrophy may also be associated with decrease in capillary length density, which is also known to occur as a result of training, leading to cardiac hypertrophy. However, although increasing exercise hyperemia pharmacologically has been shown to improve cardiac function and oxygen consumption at near-maximal exercise, our findings do not necessarily mean that athletes with very pronounced hypertrophy are compromised in their cardiac pumping function, let alone experience ischemia even at maximal exercise. The ET subjects we studied reported having never experienced any chest pain or related sensations. In addition, many of them had succeeded well in international competitions such as in different world championships. Accordingly, it seems that despite reduced cardiac hyperemic vasodilator reserve, LV hypertrophy does not lead to greater susceptibility to myocardial ischemia during external stress, but fine-tuned regulatory pathways are able to control bodily function so that capacities of any organ are not exceeded even at maximal exertion.
Finally, in line with our finding of impaired pharmacologically induced hyperemic response, it has also been previously reported that coronary artery blood flow response to reactive hyperemia is decreased in vigorously trained dogs. Reactive hyperemia is a widely used procedure also in human peripheral arteries, correlating positively with fitness in healthy but UT subjects. However, exercise training has been shown to cause dissociation between peripheral flow capacity and maximal aerobic power. Thus, these earlier findings and our results strongly support the proposition of "athlete's paradox" in which highly trained and performing athletes can have even decreased hyperemic responses when assessed by standard determination procedures at resting state.
Taken together, the present study supports the findings from the peripheral vasculature of ET subjects that the cardiac vasodilatory capacity is also not improved but can even be reduced, compared with that in UT controls. In particular, MBF during adenosine-induced hyperemia is diminished in ET subjects and this decrease is associated to V̇O2peak, and especially the degree of LV hypertrophy.