Exercise-induced changes in flow shear stress, vascular circumferential stress, and stretch stress can modulate endothelial function, and the effect of this modulation is related to the intensity of exercise. The selection of the appropriate exercise intensity is therefore essential for the maintenance and improvement of endothelial function. Exercise intensity is usually defined in clinical and experimental studies in terms of the percentage of maximal oxygen uptake (%VO${}_{\text{2max}}$, maximal oxygen uptake), percentage of maximal heart rate (%HR${}_{\text{max}}$), percentage of heart rate in reserve (%HRR, heart rate in reserve), and one-repetition maxmimum (1RM) (for resistance exercise). In resistance exercise, the 1RM that the body can lift in one repetition is used to measure its exercise intensity. In human exercise, low intensity is defined as 40% VO${}_{\text{2max}}$ or HRR 64% HR${}_{\text{max}}$, and 30–50% 1RM; moderate intensity is defined as 40–60% VO${}_{\text{2max}}$ or HRR at 64–76% HR${}_{\text{max}}$, and 50–70% 1RM; and high intensity is defined as $>$60% VO${}_{\text{2max}}$ or HRR $>$76% HR${}_{\text{max}}$, and 70–85% 1RM [23]. The grading of exercise intensity often refers to the Bedford exercise protocol for the approximate setting of exercise load in animal experiments, with a running treadmill speed of 8–10 m/min and an incline of 0° for low-intensity exercise; a running treadmill speed of 15–20 m/min and an incline of 5° for moderate-intensity exercise; and a running treadmill speed of $>$20 m/min and an incline of 10° for high-intensity exercise.

In addition, according to different exercise frequencies, exercise with different intensity is further divided into acute and long-term exercise. Acute exercise refers to a single bout of exercise, and long-term exercise is repetitive bouts of regular exercise. It is worth noting that acute high-intensity exercise includes acute high-intensity continuous exercise and acute high-intensity interval training (HIIT) or high-intensity interval exercise (HIIE). HIIT/HIIE is a form of intermittent exercise that has emerged in recent years, and encompasses the completion of numerous high-intensity exercise in a short period, and is interspersed with low-intensity exercise or rest between two high-intensity exercises [11]. Accordingly, long-term high-intensity exercise contains long-term high-intensity continuous exercise and long-term HIIT/HIIE.

To analyze the impact of different intensity exercise on persons with normal or impaired endothelial function to provide appropriate exercise intensity for them, the different populations were included in this review. These populations mainly include healthy young people, older adults, postmenopausal women, obese persons and patients with diabetes, hypertension or other cardiovascular diseases. Healthy young people tend to possess normal endothelial function, while impaired endothelial function is commonly associated with the other populations mentioned above. Some healthy elderly men also possess normal endothelial function. Accordingly, animals with normal and impaired endothelial function were included in the review [24, 25, 26]. It should be noted that human and animal experimental data on different intensity exercise regulating endothelial function were searched on PubMed, Web of Science and Google Scholar from 2003 to 2023. 70 pieces of literature were selected to review.

Extant studies have shown differential results in the modulation of arterial endothelial function through low-intensity exercise (Table 1, Ref. [9, 11, 13, 22, 27, 28]). Goto et al. [9] and Birk et al. [27] did not uncover any significant changes in FMD levels or forearm blood flow in response to ACh in healthy young adults for either acute exercise session or a long-term low-intensity exercise intervention lasting 12 weeks; this indicated that endothelial function did not undergo a significant improvement. However, Shimizu et al. [11] found that RHI levels after 4 weeks of a low-intensity resistance-exercise intervention in healthy elderly people were elevated from 1.8 $\pm$ 0.2 vs. 2.1 $\pm$ 0.3 and that their vWF levels were reduced, leading to enhanced NO-mediated vasodilation and reduced endothelial cell injury in blood vessels—suggesting that short-term low-intensity exercise improved vascular endothelial function in the elderly. In addition, Merino et al. [22] demonstrated that low-intensity walking exercise for 4 consecutive months improved antioxidant capacity and vascular endothelial function in overweight and obese postmenopausal women by augmenting the activity of the antioxidant enzymes SOD (9506 $\pm$ 3408 vs. 12,628 $\pm$ 2472 μmol/min/grHb) and GPX (1.97 $\pm$ 0.51 vs. 2.26 $\pm$ 0.77), increasing their small-artery reactive congestion index.

The results of experimental animal studies revealed that after 6 weeks of 10 m/min treadmill training in diabetic rats, serum NO levels and eNOS expression levels were elevated and vWF was decreased in the exercise group, proving beneficial to endothelial function; and appeared to prevent and improve diabetic cardiomyopathy [13]. Similarly, blood pressure was significantly reduced in rats suffering from severe hypertension after long-term low-intensity exercise training, and their impaired endothelium-dependent vasodilatory function and insulin sensitivity were also improved [29].

The reasons subserving the production of differential modulatory effects on arterial endothelial function with low-intensity exercise described in the aforementioned studies may have related to whether the human subjects or rats initially possessed healthy arterial endothelial function. As previously mentioned, healthy young people usually possess normal endothelial function, while impaired endothelial function tends to occur in healthy young people, older adults, postmenopausal women, obese persons and patients with cardiovascular diseases. However, whether the initial healthy or impaired endothelial function is a critical factor in different intensity exercise regulating endothelial function requires further study.

A series of studies have shown that acute and long-term moderate-intensity exercise promote arterial endothelial function in healthy individuals as well as in the elderly, hypertensives, diabetics, and patients after myocardial infarction (Table 2, Ref. [9, 21, 28, 30, 31, 32, 33, 34, 35, 36, 37]). Boeno et al. [30] ascertained that acute moderate-intensity resistance exercise accelerated vasodilation by increasing nitrites and nitrates (NO${}_{\text{X}}$) level from 6.8 $\pm$ 3.3 to 12.6 $\pm$ 4.2 µM and FMD from 12.5 $\pm$ 4.10 to 17.2 $\pm$ 3.9% in sedentary middle-aged men. In addition, patients undergoing percutaneous coronary intervention (PCI) after acute myocardial infarction showed significant progress in endothelial function after an acute moderate-intensity (50–60% HRR) exercise intervention [31]. Landers-Ramos et al. [32] additionally found a significant increase in FMD levels from 10 $\pm$ 1.3 to 16 $\pm$ 1.4% after moderate-intensity exercise training for 10 days in sedentary older adults following an aerobic exercise intervention at 70% VO${}_{\text{2max}}$ for ten consecutive days. Another study comprising hypertensive patients as subjects undergoing long-term exercise at 60–80% HRR exercise intensity for 12 weeks revealed a 1.7 $\pm$ 2.8% increase in FMD, and a decrease in ET-1 levels and the related inflammatory factors CRP, MCP-1, and VACM-1 levels; with improvements in blood pressure, inflammation, and endothelial function associated with cardiovascular health [33]. Meta-analysis also showed that long-term exercise training for more than 8 weeks significantly increased overall FMD levels in patients with type II diabetes, and low-to-moderate-intensity exercise augmented FMD levels compared with high-intensity continuous exercise [38, 39]. Most studies [9, 21, 28, 30, 31, 32, 33, 34, 35, 36, 37] have demonstrated that moderate-intensity exercise can ameliorate endothelial dysfunction. However, Shah et al. [24] found that prior acute moderate-intensity exercise compared with no exercise did not affect FMD, ET-1 and NO concentration following a high sugar meal in postmenopausal women. A previous study confirmed that impaired endothelial function caused by high sugar intake was restored with acute moderate intensity in young men [40]. Therefore, we speculate the dual disadvantage factors acting on endothelial function, including high sugar and menopause, inhibit the effectiveness of moderate intensity exercise on improving endothelial function in Shah et al.’s study [24].

The effect of an exercise intervention on arterial endothelial function also exerts a significant temporal effect, with the effect on endothelial function diminishing or even disappearing after the end of the exercise. Naylor et al. [34] found that 12 weeks of combined aerobic and resistance exercise in adolescent type II diabetic patients produced a significant elevation in brachial artery FMD levels from 7.62 $\pm$ 1.2 to 9.82 $\pm$ 1.0% while effectively controlling patient blood glucose levels. However, after 12 weeks of discontinuation of training, the data showed that brachial FMD gradually returned to resting levels and those endothelial functional changes gradually disappeared.

The results of experimental animal studies showed that long-term moderate-intensity wheel exercise was observed to normalize diabetes-related endothelial dysfunction and improve insulin sensitivity in a diabetic mouse model, and the results suggested a reversal of type II diabetic endothelial dysfunction by enhancing NO bioavailability through elevated production of mitochondrial manganese superoxide dismutase (Mn-SOD), total eNOS protein, and phospho-eNOS (Ser1177) [21]. Furthermore, Ye et al. [35] trained hypertensive rats to run at 18–20 m/min (55–65% VO${}_{\text{2max}}$) for 60 min per day, 5 days per week for 8 weeks, and discerned that long-term moderate-intensity exercise prevented hypertension-related endothelial ultrastructural remodeling and endothelial dysfunction by alleviating oxidative stress and enhancing NO-mediated diastolic response in the mesenteric arteries of hypertensive rats. In another study, the authors established a mouse model of hypoxia-induced endothelial cell injury and found that moderate-intensity exercise for 4 weeks augmented the release of circulating exosomes derived from endothelial progenitor cells (EPCs) and elevated the expression of exosome miR-126. This exercise reduced the apoptotic rate of endothelial cells, thereby protecting and enhancing endothelial cell function [28].

In conclusion, most acute and long-term moderate-intensity exercise training effectively improve endothelial function in different populations, but their effects on endothelial function have certain time limitations. Therefore, both healthy individuals with normal endothelial function and those with endothelial dysfunction need to maintain an effective exercise regimen to improve endothelial function through long-term exercise. In addition, double or multiple unfavorable factors acting on endothelial function may be able to weaken or inhibit the improvement of moderate intensity exercise on endothelial function. Thus, people with endothelial dysfunction who expect to obtain a beneficial effect of exercise need to minimize the influence of adverse factors, such as reducing high sugar intake.

Some studies suggest that one bout or repetitive bouts of sustained high-intensity exercises cause oxidative stress and the development of cellular inflammatory responses, leading to impaired endothelial function, while others suggest that repetitive bouts of HIIT/HIIE exerts a positive effect on the regulation of arterial endothelial function (Table 3, Ref. [9, 12, 27, 30, 35, 41, 42, 43, 44, 45, 46]). Birk et al. [27] described a significant diminution in FMD levels from 6.6 $\pm$ 1.6 to 3.6 $\pm$ 2.2% after acute high-intensity continuous exercise, and Nyborg et al. [41] discerned that athletes participating in the high-intensity continuous Norwegian triathlon (an all-around sport) exhibited a transient 5.6% decrease in FMD response immediately after the race, with a corresponding drop in L-arginine (NO precursor) levels and a rise in the levels of the endothelial inflammatory markers E-selectin, vascular cell adhesion molecule-1 (VCAM-1), and intercellular cell adhesion molecule-1 (ICAM-1). This implied that the inflammatory response induced by overly intense exercise led to endothelial dysfunction. Other studies have shown that long-term sustained high-intensity exercise also impaired endothelium-dependent vasodilation and reduced NO synthesis and secretion by reducing antioxidant levels and increasing oxidative stress. For example, Goto et al. [9] subjected healthy young men to intense cycling training at 75% VO${}_{\text{2max}}$ for 12 weeks, and found that the subjects’ blood reflected an increase in 8-hydroxy-2${}^{\prime }$-deoxyguanosine (8-OHdG) from 6.7 $\pm$ 1.1 to 9.2 $\pm$ 2.3 ng/mL and malondialdehyde-low density lipoprotein (MDA-LDL) from 69.0 $\pm$ 19.5 vs. 82.4 $\pm$ 21.5 U/L, neither of which was conducive to endothelium-dependent vasodilation. However, HIIT/HIIE with alternating high- and low-exercise intensities was shown to generate a significantly beneficial effect on endothelial function. HIIT was found to improve vascular endothelial function in healthy older adults in the short term, and to reduce the risk of cardiovascular disease to a greater extent than with moderate-intensity continuous exercise, positively affecting the vascular system [42]. Jo et al. [43] conducted a comparative exercise intervention between long-term HIIT and moderate-intensity continuous exercise for 8 weeks in 34 patients with hypertension syndrome, and their results suggested that FMD was significantly improved by 6.1% after exercise in both groups, while NO and EPC expression in patients in the HIIT group rose after the intervention. However, there was no change after moderate-intensity continuous exercise.

Relevant experimental studies with animal models have shown that acute and long-term sustained high-intensity exercise leads to impaired endothelial function. Przyborowski et al. [44] ascertained that reduced NO production and elevated superoxide anion levels in mice after acute progressive-to-exhaustive exercise approached basal levels after 4 hours of recovery. Other investigators reported that long-term high-intensity continuous exercise increased oxidative stress levels in spontaneously hypertensive rats, leading to eNOS uncoupling, excess ROS production, and attenuated NO bioavailability - which, in turn, adversely affected endothelial function [12, 35]. In the study by Groussard et al. [45], SOD and GPx activities in Zucker obese rats rose after 10 weeks of HIIT exercise intervention, improving endothelial function by enhancing antioxidant-defense capabilities.

In summary, high-intensity exercise provokes specific changes to the regulation of arterial endothelial function. Acute or long-term high-intensity sustained exercises can cause inflammatory responses and oxidative stress, reduce NO bioavailability, and adversely affect endothelial function. Acute or long-term HIIT/HIIE, however, enhances arterial endothelial function; and, therefore, those individuals with insufficient exercise time can select HIIT/HIIE. However, because HIIT/HIIE requires several high-intensity exercises in a short period, strict control is needed in exercise intensity, exercise interval time, and exercise frequency to prevent the occurrence of adverse cardiovascular events.

As a vasodilator derived from endothelial cells, NO is catalyzed by activated eNOS to metabolize L-arginine. NO regulates vascular tone and also inhibits platelet aggregation and leukocyte adhesion, and is an important indicator of arterial endothelial function [47]. ET-1 is an active peptide secreted by endothelial cells and exerts a robust vasoconstrictive effect. Vascular endothelial dysfunction, then, is associated with a decrease in NO secretion as well as an increase in ET-1 secretion. Studies have shown that mechanical forces (including blood shear stress, circumferential stress, and stretch stress) are important physiologic regulators of the production of both NO and ET-1 in endothelial cells, with blood shear stress being the most important mechanical force stimulus regulating vascular endothelial function [3]. Acute and long-term moderate-intensity exercise and acute HIIE intervention augment perfusion, alter hemodynamic signaling, and induce the upregulation of the phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) pathway by increasing the frequency and amplitude of shear stress acting on the eNOS Ser1177 phosphorylation at the vascular wall. Furthermore, this activity activates and increases endogenous NO bioavailability and reduces ET-1 production, thereby improving vascular tone and arterial endothelial function [48, 49, 50, 51]. In contrast, acute high-intensity sustained exercise can uncouple eNOS, leading to a further drop in NO production as well as an elevation in ET-1 concentrations [30, 33]. In addition, acute HIIT exercise mediates increased levels of adipocytokine C1q/tumor necrosis factor-related protein 9 (CTRP9), which may benefit endothelial function in obese individuals by promoting eNOS phosphorylation [52].

Oxidative stress is a pathophysiologic state that results from an imbalance between the oxidative and antioxidant systems and arises when redox homeostasis within an organism is damaged. Dysregulation of redox homeostasis occurs when the production of oxidants such as ROS (including superoxide anion, hydrogen peroxide, and hydroxyl radical) exceeds the production of antioxidants such as superoxide dismutase and glutathione, leading to impaired vascular endothelial function [53]. Studies have shown that low levels of ROS are conducive to maintaining normal cellular homeostasis and blood vessel function, while overproduction of ROS leads to oxidative stress reactions that then lead to the development and progression of cardiovascular diseases. Long-term low-to-moderate-intensity exercise increases the expression of the body’s antioxidant enzymes SOD and GPX, improves antioxidant enzyme defense systems, maintains vascular homeostasis, and reduces the risk of cardiovascular disease by augmenting antioxidant capacity as well as by reducing the overexpression of oxidative enzymes (e.g., reduced nicotinamide adenine dinucleotide phosphate-reduced oxidase and xanthine oxidase [35, 54]). In addition, researchers have identified a mitochondrial inner membrane uncoupling protein 2 (UCP2) that is an important negative regulator of ROS production. Furthermore, they have found that the long-term moderate-intensity exercise intervention upregulated UCP2 expression via the peroxisome proliferator-activated receptor gamma coactivator-1 alpha (PGC1$\alpha$)/peroxisome proliferator-activated receptor-delta (PPAR-$\delta$) pathway and thereby down-regulated ROS and increased eNOS expression; this, in turn, mitigated metabolic disorders concerning NO bioavailability to alleviate endothelial dysfunction [35, 55, 56]. However, excessive and sustained acute and long-term high-intensity sustained exercise can cause substantial ROS production, induce vascular remodeling, and alter normal physiologic processes, thereby reducing intracellular NO bioavailability and further exacerbating endothelial dysfunction and injury [48, 57, 58].

Inflammatory responses can induce arterial endothelial dysfunction that constitutes an important trigger for atherosclerosis and structural changes in arteries. The exercise intervention effectively suppresses the inflammatory response, and long-term low-to-moderate-intensity exercise reduces the expression of pro-inflammatory proteins, such as CRP, MCP-1, tumor necrosis factor-alpha (TNF-$\alpha$) , ICAM-1, and VCAM-1, while increasing the expression levels of anti-inflammatory factors such as IL-4 and IL-10, thereby rectifying arterial endothelial function [33, 59]. Hong et al. [56] found that the long-term moderate-intensity exercise mitigated coronary endoplasmic reticulum stress-related endothelial dysfunction by reducing endoplasmic reticulum stress and thioredoxin-interacting protein (TXNIP)/nucleotide-binding and oligomerization (NACHT), leucine-rich repeat (LRR), and N-terminal pyrin domain (PYD) domains-containing protein 3 (NLRP3) inflammatory vesicle expression. In contrast, long-term high-intensity sustained exercise increased the levels of inflammatory factors such as IL-6, creating an inflammatory response [60].

EPCs can differentiate into endothelial cells that are involved in mediating the repair of endogenous vascular endothelial injury and that play a key role in maintaining the structural and functional integrity of the vascular endothelium [61]. Long-term Exercise can increase the number of EPCs, increase the activity of eNOS, enhance the bioavailability of NO, regulate endothelial repair in angiogenesis, and prevent endothelial dysfunction by targeting EPCs [62]. Acute and long-term moderate-intensity aerobic or resistance exercise affects the mobilization of EPCs by augmenting related pro-angiogenic factors such as vascular endothelial growth factor (VEGF),stromal cell-derived factor-1, hypoxia-inducible factor-1, and matrix metalloproteinase-9 to enhance endogenous endothelial repair, which, in turn, repairs and maintains the vascular cytoarchitecture [63, 64, 65].

The release of a range of bioactive molecules in exercise via extracellular vesicles has been identified as a novel phenomenon in mediating intercellular communication that promotes beneficial effects in many systems in vivo, and the exercise intervention can induce the release of exosomes from a variety of tissues and thereby improve endothelial function. Exosomes are nano-sized tiny extracellular vesicles that contain proteins, lipids, nucleotides, and other biologically active substances; and these target endothelial cells through direct lipid membrane fusion, receptor-ligand interactions, macropinocytosis, endocytosis, and other pathways to regulate cellular behavior and mediate biological effects. These actions then promote vascular neogenesis, the regulation of vasoconstriction and diastole, and the inhibition of apoptosis and other regulatory endothelial functions [66]. Exercise-induced exosome secretion provides a novel and direct endogenous cardioprotective effect that is closely related to the advancement of intercellular information exchange [67] by exercise, the activation of the sprouty related EVH1 domain containing 1 (SPRED1)/VEGF signaling pathway [28], and increased Ca${}^{2+}$ release [68]. Exercise-induced circulating exosomes are thought to manifest a positive regulatory role in mediating signaling processes associated with adaptive responses to exercise, and both acute exercise and long-term chronic exercise exert beneficial actions on endothelial function and protect cardiovascular health by promoting the expression of exosomes and their contents (miRNAs and proteins)-including miRNA-126 and miRNA-342-5P, which enhance intercellular communication [28, 66, 69]. Ma et al. [28] demonstrated that long-term low- and moderate-intensity exercise reduced endothelial cell apoptosis and improved endothelial function by increasing miR-126 expression in circulating exosomes, which in turn activated the downstream sprouty related EVH1 domain containing 1 (SPRED1)/VEGF-signal-transduction pathway. Another study revealed that plasma exosomes secreted by mice after 2 weeks of moderate-intensity wheel exercise significantly enhanced endothelial cell migration and angiogenesis compared with sedentary mice and that exosomal SOD3 was important in angiogenesis [70].