Counterpoint: Exercise training does not induce vascular adaptations beyond the active muscle beds

Dick H. J. Thijssen, Maria T. E. Hopman

Exercise training is a well known and powerful strategy to improve cardiovascular structure and function in healthy individuals as well as in patients with various diseases (10). It has been suggested that exercise-induced vascular adaptations, at least partly, explain the decrease in cardiovascular risk (11). Therefore, insight into localization and magnitude of exercise-induced vascular adaptations is essential to optimally benefit from exercise training to improve vascular health.

Previous studies that examined the effects of exercise training in animal studies primarily evaluated the exercised hindlimbs (5, 29), whereas molecular responses to cycling exercise in humans were analyzed from biopsies taken from the highly active quadriceps muscle (7, 12). In contrast, studies examining vascular adaptations to lower limb training (e.g., cycling, walking, and running) mainly focus on the nonactive forearm vascular bed. This is even more surprising considering the stimuli for arteriogenesis and angiogenesis during exercise training, i.e., increments in blood flow and shear stress on the endothelium and reduced oxygen tension and related expression of vascular endothelial growth factor. During cycle exercise, blood flow and oxygen consumption only minimally increase in the nonactive upper limbs (1, 9, 24). Accordingly, vascular adaptations are unlikely to be expected in inactive regions.

While many previous studies examined lower limb exercise-induced vascular adaptations in one region (e.g., forearm; Refs. 4, 14, 17, 23, 30), only a few studies assessed adaptations in the active leg as well as in the nonactive upper extremity. Interestingly, these latter studies reported vascular adaptations in the exercised region, but not in the nontrained vascular beds. For example, in 40 patients with coronary artery disease, 10-wk (predominantly cycling and walking) exercise training resulted in an improved posterior tibial artery endothelial function, with no change in brachial vascular function (8). In addition, adaptations in lower limb conduit and resistance artery function and structure were found after 4–6 wk electrically stimulated cycle training in spinal cord-injured individuals (27, 28), 3 mo cycling training in heart failure patients (3, 15), 12 wk walking exercise in heart transplant recipients (21), and 3 wk (19) to 1 yr (6) lower limb training in postmyocardial infarct patients, while forearm vascular function and structure did not change. This indicates that large muscle mass exercise in these studies induced insufficient stimuli in the nonactive forearm to result in vascular adaptations.

Despite the above findings, several studies, which examined the forearm only, reported functional and structural vascular adaptations to lower limb exercise training (e.g., walking, running, and cycling) in the nonactive brachial (30) and radial artery (17) or forearm vasculature (4, 14, 23). Especially during walking and running exercise, which involves upper body movements, the nonactive forearm vascular bed is at least moderately active. Apparently, the angiogenic stimuli shear stress and hypoxia exceeded the threshold to induce vascular adaptations in the nonactive forearm vascular bed in these studies (4, 14, 17, 23, 30). On the basis of this limitation in many in vivo studies, vascular adaptations in nonactive regions do not necessarily result from the exercise training stimulus in the lower limbs per se.

Accurately addressing the question whether exercise training induces vascular changes in nonactive areas is challenging, especially when examining large muscle mass exercise. Studying individuals with a spinal cord injury (SCI) offers a unique opportunity to examine active and nonactive areas during exercise training. Below the lesion level, SCI individuals are subject to a complete loss of motor and sensory control, excluding the possibility for “polluting” muscle activity in the paralyzed muscles during exercise. Accordingly, arm-crank exercise in SCI individuals offers the opportunity to study whether leg vascular function adapts in the paralyzed legs. Although involving a smaller muscle mass than cycling, arm-crank exercise can result in ∼80% of maximal oxygen uptake and ∼90% of maximal heart rate (18, 22). A previous cross-sectional study examined the effects of upper extremity exercise training on artery size above and below the lesion level in paraplegic endurance athletes (n = 29) and inactive paraplegic subjects (n = 20). The ∼50% higher physical fitness level and doubling in cross-sectional area of the subclavian artery in the paraplegic endurance athletes compared with sedentary paraplegics indicates the difference in upper extremity activity level between both paraplegic groups. Nonetheless, both groups demonstrated similar femoral artery dimensions (13), which reinforces our hypothesis.

Functional electrical stimulation (FES) in SCI individuals provides the opportunity to stimulate individual paralyzed muscle groups. FES cycling substantially increases heart rate and oxygen uptake (20) and is demonstrated to change leg vascular function after at least 2 wk training (27, 28). Adjacent, nonstimulated paralyzed regions are subject to passive movement, while no polluting muscle activity will be present. Since passive movement does not induce acute or chronic blood flow changes (25), these regions provide a unique opportunity to study the effects of exercise training in active and adjacent, nonactive muscles. Recently, we studied vascular adaptations before and after 4 wk of FES cycling in the stimulated thigh muscles and the adjacent inactive calf in SCI individuals (28). While functional and structural vascular adaptations were reported in the thigh, no exercise-induced vascular changes were reported in the calf vascular bed. In another study, 4-wk unilateral limb stimulation in SCI individuals significantly changed superficial femoral artery structure and function of the trained leg, whereas vascular characteristics in the untrained leg were not altered (2). These findings demonstrate the presence of local exercise-induced vascular adaptations, while contralateral or adjacent regions from the exercised area do not benefit from the exercise training.

Taken together, we conclude that exercise does not induce vascular adaptations beyond the active muscles. Most likely, if vascular adaptations in the nonactive regions were present, these regions were moderately active during exercise training, resulting in changes in shear stress acting on the endothelium and small increases in oxygen consumption and hypoxia. Although these changes in shear stress and oxygen consumption in the nonactive regions are likely to be much lower than in the exercised region, both stimuli may exceed the minimum level necessary to result in vascular adaptations. Accordingly, active regions may demonstrate functional and structural vascular changes at an earlier stage than the nonactive muscles, leading to a possible underestimation of the effects of exercise training when examining the nonactive muscles only. Therefore, future studies that examine the effects of physical (in)activity on the vasculature should examine active as well as nonactive regions. Many human in vivo studies automatically choose to examine forearm vascular adaptations to assess the effects of exercise training. However, calf and thigh baseline and peak blood flow using plethysmography reported a good reproducibility (26), while the superficial femoral artery FMD response represents a largely NO-mediated endothelium-dependent vasodilation (16). This indicates the robustness of these tools to examine lower limb vascular function and structure of conduit and resistance vessels in humans in vivo.