Low Response to Aerobic Training in Metabolic Disease: Role of Skeletal Muscle

Abstract

Aerobic exercise is established to increase cardiorespiratory fitness (CRF), which is linked to reduced morbidity and mortality. However, people with metabolic diseases such as type 1 and type 2 diabetes may be more likely to display blunted improvements in CRF with training. Here, we present evidence supporting the hypothesis that altered skeletal muscle signaling and remodeling may contribute to low CRF with metabolic disease.

Summary for table of contents:

Metabolic disease impairs the ability of skeletal muscle to adapt to exercise training, contributing to low aerobic capacity.

Introduction

Cardiorespiratory fitness (CRF) is one of the single best predictors of longevity in humans (1). In addition, having high CRF may reverse cardiovascular disease risk and mortality imparted by other important risk factors such as smoking, hypertension, diabetes, and obesity (2, 3). In contrast, having low CRF is associated with a 5-fold increase in risk for developing metabolic syndrome and type 2 diabetes (4), and up to 8-fold higher risk for cardiovascular disease (5). Thus, improving cardiorespiratory fitness is a key strategy to reduce the risk of developing cardio-metabolic disease to prolong health- and life-span.

The gold standard measurement of CRF is maximal or peak oxygen uptake (VO2_peak), which is defined as the highest value of oxygen consumption attained upon an strenuous exercise test, designed to bring the subject to the limit of exercise tolerance. VO2_peak is a measure that encompasses cardiovascular, skeletal muscle, and respiratory function and reflects the integrated ability to transport oxygen from atmospheric air to the working tissues (6). Striated muscles represent the largest consumers of oxygen during aerobic exercise (6).

Regular aerobic exercise is the only clinically validated treatment to improve VO2_peak, and consequently, CRF. However, improvements in CRF with training display a large degree of heterogeneity (7). In response to standardized, laboratory-supervised exercise training, some individuals have large improvements in exercise capacity (high-responders) while others demonstrate little or no change in aerobic exercise capacity (low-responders) (8, 9). The fact that some individuals are less responsive to improvements in CRF with training, infers “exercise resistance” (10), or a phenotype of Low Response to Training (LRT) (11). Exercise resistance can refer to a blunting of many different metabolic and health benefits of exercise, as reviewed previously (12, 13). For the purposes of this review, the abbreviation LRT will be used to refer specifically to impaired improvement in VO2_peak in response to aerobic training in humans and animal models.

Findings from the HERITAGE family study indicate that approximately 50% of the variance in CRF exercise-response is due to inherited (genetic/epigenetic) factors, with the remainder attributed to environmental factors (7). There has been significant scientific interest in determining the mechanisms that contribute to the LRT phenotype. This review focuses on the potential contributions of metabolic disease and diabetes on LRT, and the molecular mechanisms that link impaired glucose metabolism to blunted improvements in CRF. We present evidence to support a novel hypothesis that systemic conditions that occur with metabolic disease such as chronic hyperglycemia and insulin resistance lead to abnormal skeletal muscle molecular signaling and morphophysiological adaptations in response to aerobic exercise training. In turn, we propose this aberrant adaptive response of muscle to training contributes, in part, to persistently low CRF in individuals with metabolic disease.

PART 1: METABOLIC DISEASE AND CARDIORESPIRATORY RESPONSE TO TRAINING

As observed by the HERITAGE family study (7) and subsequently several other investigations (8, 9, 14), LRT can be observed in the general population, independent of chronic disease status, at an incidence of ~25–30% (15). The LOOK AHEAD study demonstrated that individuals with metabolic disease (obesity and type 2 diabetes) can improve CRF through intensive lifestyle intervention including exercise training and weight loss (16). However, whether increases in CRF with training are both clinically and statistically significant requires careful consideration of several factors (8, 17). Inclusion of a sedentary control group to assess the error of CRF measurement over time, or a control group without metabolic disease, can better allow conclusions regarding the effect of metabolic disease on CRF in response to training. In that regard, some clinical studies indicate that individuals with metabolic disease have lower CRF compared to people without metabolic impairments. Impaired glucose tolerance, insulin resistance, Type 1 diabetes and Type 2 diabetes have all been associated with low CRF compared to age-, weight-, and activity-matched controls without metabolic disease (18–21). This indicates that individuals with metabolic diseases may experience an even higher level of heterogeneity in exercise response than populations without metabolic impairments

Individuals with impaired glucose tolerance and diabetes represent an increasing proportion of the population, and persistently low CRF can compound increased risk of cardiovascular disease and mortality already present in this group. Therefore, determining the mechanisms that underlie LRT is an essential first step toward improving the response to exercise and reversing the extensive health risks associated with low CRF and metabolic disease.

Hyperglycemia and CRF

Chronic hyperglycemia, or elevated blood glucose, is a common feature of impaired glucose tolerance, insulin resistance, type 1 diabetes, and type 2 diabetes. Nadeau et al. identified a negative association between HbA1c, which is a marker of blood glucose concentrations in the preceding 2–3 months, and low CRF in male and female adolescents with and without type 1 diabetes (18) and type 2 diabetes (19). Likewise, Solomon and colleagues demonstrated that indices of hyperglycemia, including high HbA1c and elevated fasting glucose, are predictive of low CRF in males and females of heterogeneous age with glucose tolerances ranging from normal to highly impaired with type 2 diabetes (21). The same laboratory demonstrated an inverse relationship between baseline HbA1c and training-induced changes in CRF (ΔVO_2peak) in older men and women with type 2 diabetes, suggesting that low CRF associated with chronic hyperglycemia may be due to LRT (20).

In contrast, a smaller study of pre-menopausal women (age range: 30–50 yrs) found that those with T2D and higher starting HbA1c had greater improvements in VO₂peak after exercise training compared to pre-menopausal women without T2D (22). Thus, although several key clinical investigations have identified chronic hyperglycemia as a potential physiological contributor to LRT, sex and age may also have an impact on improved CRF with training in the presence of metabolic dysfunction. Another study examining the effect of low or high glycemic diets on exercise-induced changes in CRF, found that older (>60 yrs) men and women with pre-diabetes had similar increases in VO₂max with both dietary strategies (23). As a low glycemic index diet would be expected to mitigate rises in blood glucose levels after meals, these data may argue against an important role for hyperglycemia in CRF response to exercise. However, it should be noted that participants began the study with HbA1c in the normal range, and fasting glucose was decreased similarly by both diets in this population (23).

To determine whether chronic hyperglycemia may be causally related to LRT, we embarked on a series of studies using pre-clinical models of chronic hyperglycemia (24–26). First, we studied novel rat models generated by selective breeding for low and high CRF response to aerobic exercise training (LRT and HRT, respectively). After 16 generations of breeding, HRT had a 50% increase in CRF with training, which was completely absent in LRT (24). Importantly, sedentary LRT showed lower insulin sensitivity and increased blood glucose levels in response to a glucose tolerance test compared to sedentary HRT, indicating impairment in glucose metabolism as a potential mechanism driving poor cardiorespiratory response to aerobic exercise training (24).

Subsequent studies demonstrated that induction of moderate (>200 mg/dL) chronic hyperglycemia 8-wks before voluntary wheel running in CD-1 mice blunted training-induced improvements in CRF (26). Notably, improvements in CRF were similarly blunted if hyperglycemia was induced by 1) consumption of a Western-style diet high in sucrose and saturated fat, or 2) by reducing insulin levels with streptozotocin treatment (26). As both models differ with respect to common confounding factors such as serum insulin and fat mass, these results suggest that hyperglycemia itself may be causative in LRT. Additional time-course experiments in these models indicate that chronic hyperglycemia in the range of weeks to months is necessary to blunt aerobic adaptation to training. In contrast, glucose levels during acute exercise, which may be impacted by carbohydrate intake during exercise, do not appear to negatively impact the adaptive response.

In support of a role for chronic hyperglycemia in LRT, we demonstrated that prevention of hyperglycemia in streptozotocin-treated mice using the sodium-glucose cotransporter-2 (SGLT2) inhibitor, canagliflozin, can restore improvements in CRF with training (25). Importantly, exercise-induced improvements in CRF in that study were found to be independent of fat mass loss, which is a common side-effect of SGLT2 inhibition (25, 27). Taken together, studies in humans and rodent models point to hyperglycemia as a cause of LRT associated with metabolic disease and identify pharmacological anti-hyperglycemics as a potential therapy to improve CRF.

Insulin resistance and CRF

Insulin resistance is a condition that frequently accompanies hyperglycemia and has been linked to low CRF. Solomon et al. demonstrated in a group of men and women with normal glucose tolerance, impaired glucose tolerance, or type 2 diabetes, that insulin sensitivity as assessed during an oral glucose tolerance test (siOGTT), displayed a strong positive association (R=0.734, P<0.0001) with CRF (21). Consistent with this, our laboratory demonstrated a similar relationship (R=0.635, P<0.001) between the siOGTT insulin sensitivity index and CRF in men and women without diabetes (26). In addition, there is evidence that insulin resistance in healthy weight adolescents with type 1 diabetes can independently predict low VO_2peak (18). Thus, insulin resistance is associated with low CRF in people with type 1 and type 2 diabetes, as well as individuals without diagnosed metabolic disease.

It is difficult to separate the contributions of chronic hyperglycemia and insulin resistance to low CRF, as these two conditions are inter-dependent. Which defect, if either, represents the primary causative factor in low CRF with metabolic disease is unknown. However, 4 months of the insulin sensitizing treatment, rosiglitazone, increased CRF by ~7% in people with uncomplicated type 2 diabetes (28). In addition to improving insulin sensitivity, Rosiglitazone significantly reduced fasting glucose in that trial, illustrating the close relationship between hyperglycemia and insulin resistance. In another trial of older men and women with pre-diabetes, a low glycemic index diet combined with exercise was found to produce greater improvements in insulin sensitivity compared to a high glycemic index diet (23). Yet, CRF was improved similarly, along with decreases in body weight and fasting glucose in both diet groups, suggesting no connection between insulin sensitivity and CRF in that population (23). Nevertheless, the results of some clinical studies, considered in conjunction with robust associations between low CRF and insulin resistance, suggest insulin sensitizers as a potential therapeutic avenue to improve CRF in metabolic disease (28).

PART 2: ROLE OF SKELETAL MUSCLE REMODELING IN CRF WITH DIABETES

Exercise capacity can be improved by several physiological adaptations that occur with endurance training, including increased cardiac output, neuronal remodeling, and higher blood hemoglobin concentration (6). In addition, aerobic training induces adaptations to skeletal muscle that improve oxygen delivery and metabolism and contribute to improved exercise capacity (29). These muscle adaptations include a shift toward a more oxidative fiber type, and exercise-induced angiogenesis to increase capillary density. Here, we present evidence that exercise-induced remodeling processes are impaired in skeletal muscle under conditions of metabolic disease.

Blunted exercise-induced angiogenesis and muscle blood flow

Increased skeletal muscle capillary density is a key exercise-induced morpho-physiological adaptation contributing to improved cardiorespiratory fitness (30). Capillary expansion (i.e. angiogenesis) in response to exercise training is a vital chronic adaptation that allows the active tissues to receive the necessary supply to withstand continuous long-term increase in work capacity and metabolic demand (30).

Some clinical studies demonstrate that muscle capillary density is lower in people with impaired glucose tolerance and diabetes compared to those without metabolic disease (31–33). In addition, impaired muscle blood flow is associated with low CRF in people with type 1 and type 2 diabetes (18, 34). What contribution lower capillary density makes to impaired muscle blood flow in diabetes is unknown, as blood flow is regulated by several factors, including endothelial function and sensitivity to hyperemic stimuli (35, 36). However, there is some evidence from clinical studies that lower muscle capillary density with metabolic disease is due to impaired exercise-induced angiogenesis. Wallberg-Henrikson et al., demonstrated that young men with type 1 diabetes had no increase in muscle capillary density following 8-wks of endurance training, while control participants had a 14% increase in capillary density in response to the same training protocol (37). It should be noted that participants with type 1 diabetes had normal increases in CRF (VO₂peak) in that study, thus observing a disassociation between capillary density and CRF (37). In contrast, another study found that VO₂peak was significantly associated with muscle capillary density in participants with normal glucose tolerance and type 2 diabetes (38).

Although there is evidence of impaired muscle capillary density with metabolic disease, there are also some contradictory data. Monaco et al. demonstrated that young, physically active adults with type 1 diabetes had similar muscle capillary density than age- and activity-matched controls without diabetes (39). Prior et al. demonstrated that older individuals with impaired glucose tolerance had a significant improvement in muscle capillary density in response to endurance training (40). However, no control group with normal glucose tolerance with which to compare the angiogenic response to exercise was included (40). Another investigation found muscle CD31 content, which can be a marker of capillary density, was increased following single-leg exercise training in the skeletal muscle of obese men and women with or without type 2 diabetes (41). However, training was more akin to resistance, rather than aerobic training in that study, and participants with diabetes had relatively mild hyperglycemia (HbA1c<6.5). Thus, whether capillary density and exercise-induced angiogenesis is impaired in humans with metabolic disease likely depends on participant characteristics such as age, sex, and habitual activity levels. Severity and duration of disease, exercise protocol, and other aspects of study design may also play a role.

Rat models selectively bred for low (LRT) or high (HRT) response to training have provided valuable insight into the potential role of exercise-induced angiogenesis on CRF response to aerobic training. Blunted CRF improvements following 8-wks of treadmill running in female LRT were mirrored by a failure to increase muscle capillary density (24). As a result, HRT had ~ 50% greater CRF and muscle capillary density compared to LRT after training, suggesting impaired exercise-induced angiogenesis may contribute to the LRT phenotype. Consistent with a role for impaired exercise-induced skeletal muscle angiogenesis in LRT associated with metabolic disease, blunted CRF with training was associated with lower post-training (voluntary wheel running) capillary density in mice chronically exposed to hyperglycemia (26). Notably, we demonstrated blunted increases in muscle capillary density whether hyperglycemia was caused by Western diet consumption or streptozotocin treatment (26). In addition, if hyperglycemia was prevented using canagliflozin in streptozotocin treated mice, exercise-induced increases in capillary density were restored (25), suggesting chronic hyperglycemia as a potential mechanism for impaired exercise-induced angiogenesis.

More glycolytic muscle fiber-type

Along with increased vascularization, endurance exercise training promotes a shift towards a more oxidative skeletal muscle fiber phenotype. Type I and type IIA fibers are more oxidative and fatigue resistant while type IIX fibers are more glycolytic and less resistant to fatigue. Having a high proportion of oxidative muscle fibers is characteristic of elite endurance athletes (29), and this adaptive remodeling is thought to be one mediator of improved endurance exercise capacity.

Several studies demonstrate that individuals with obesity, insulin resistance, or diabetes have a lower proportion of oxidative muscle fibers compared to controls without metabolic disease (38, 42–45). Segerstrom et al. demonstrated that type I and type IIX fiber content were positively and negatively associated, respectively, with CRF in age- and BMI-matched older male participants with normal glucose tolerance or type 2 diabetes (38). In addition, there are reports that the proportion of type 1 fibers can predict metabolic health markers such as insulin sensitivity in humans (43, 45). In women with polycystic ovary syndrome (PCOS), the failure to improve insulin sensitivity with endurance training was associated with a blunted fiber-type switch from IIX to IIA, compared to controls without PCOS (46). Whether a reduction in oxidative muscle fiber proportion plays a causative role in low CRF with metabolic disease is unknown. Few studies have compared fiber-type switching with endurance exercise in people with metabolic disease compared to a control group. Furthermore, there is some contention over whether fiber-type switching plays an important role in exercise adaptation in humans, although a growing consensus suggests that it does (47).

Data from preclinical studies support a role for impaired fiber-type switching in low CRF associated with metabolic disease. Many investigations have demonstrated genetic manipulations that lead to a more oxidative muscle fiber-type also result in increased CRF (48, 49). In addition, our data demonstrate that rats selectively bred as low responders for improved CRF with training (LRT) have lower type 1 fiber content in plantaris muscle than high-responders (HRT) (24). Interestingly, the fiber-type defect in LRT was present in both sedentary and trained rats, suggesting it is not the sole contributor to post-training differences in CRF in those models (24). In mouse models of chronic hyperglycemia, we demonstrate no difference in gastrocnemius fiber composition in sedentary mice, but a blunted increase in type I/IIA fibers with training compared to controls with normoglycemia (26). Moreover, glucose lowering with canagliflozin restored training-induced fiber-type switching in mice exposed to chronic hyperglycemia (25). Thus, studies from animals and humans demonstrate that metabolic disease is associated with a less oxidative fiber type, which may be due to blunted training-mediated fiber type switching.

Muscle mitochondrial function

Aerobic exercise induces a well-characterized and reproducible increase in mitochondrial number and muscle oxidative capacity. Whether muscle mitochondrial capacity contributes to improved CRF with training has been a point of contention, with some studies demonstrating that mitochondrial oxidative capacity exceeds demands during maximal exercise (6). However, these conclusions were largely based on studies of metabolically healthy male humans and rodents. In contrast, many metabolic disease states are associated with impairments to muscle mitochondrial density and function, which introduces the possibility that mitochondria may be limiting to CRF under these conditions.

Monaco et al. (2018) demonstrated impaired mitochondrial oxidative capacity and altered mitochondrial ultrastructure in muscle from physically active young adults with type 1 diabetes compared to matched controls without diabetes, although no training intervention was performed (39). Notably, overall mitochondrial content was similar between people with and without diabetes in that study, suggesting a defect in mitochondrial respiration rather than biogenesis (39). Using ³¹P-magnetic resonance spectroscopy, Cree-Green et al. (2015) demonstrated that youth with type 1 diabetes had slower rate of oxidative phosphorylation and conversion of ADP back to ATP after cessation of isometric exercise, compared to age- and activity-matched controls without diabetes (50). Consistent with this, an independent study found that defects in mitochondrial complex IV – a major regulatory site for oxidative phosphorylation-dependent ATP restoration (51) – accompanied low CRF and blunted exercise-induced increase in skeletal muscle oxygen uptake in young adults with type 1 diabetes compared to BMI- and activity-matched controls (52).

Whether mitochondrial defects in the muscle of individuals with type 1 diabetes are due to impaired improvements in mitochondrial capacity with aerobic exercise is unknown, as few training intervention studies have examined this. However, one study demonstrated that baseline alterations in mitochondrial function were partially restored by a combination of endurance and resistance training in people with type 1 diabetes, although overall adaptation to training was lower compared to control participants without diabetes (53).

The potential contribution of impaired muscle mitochondrial capacity to low CRF in people with insulin resistance and type 2 diabetes is less clear than with type 1 diabetes. Some investigations demonstrate normal mitochondrial function in type 2 diabetes and insulin resistance (54), while others have observed impairments (55). This apparent discordance may be due to the use of populations with different disease severity and duration, or utilization of different methodologies to assess mitochondrial content and function. Hey-Mogensen et al. demonstrated improvements in mitochondrial respiration after 10-wks of aerobic training in in men with type 2 diabetes, despite displaying blunted exercise-induced increases in mitochondrial gene transcripts compared to controls without diabetes (56). It should be noted that basal mitochondrial respiration was not impaired in those participants (56).

When considered collectively, data from humans with type 1 diabetes, type 2 diabetes and insulin resistance suggests that baseline impairments in muscle mitochondrial capacity and function can occur. More studies are needed to determine whether mitochondrial impairments in these populations can be normalized by aerobic exercise training. In addition, it is not fully understood whether increases in mitochondrial capacity with exercise are similar in people with metabolic disease compared to matched controls without metabolic dysfunction.

Abnormal extracellular matrix expansion and remodeling

The causes for impaired skeletal muscle remodeling with exercise, including blunted angiogenesis and fiber-type switching, in rodent models of diabetes and human participants are unknown. However, the extracellular matrix (ECM) of skeletal muscle plays a key role in muscle remodeling, signaling, health, and disease. Moreover, the ECM compartment is frequently dysregulated in diabetes, contributing to many of its complications including nephropathy and cardiovascular disease. ECM remodeling also occurs in the skeletal muscle of animals and humans with metabolic disease, with an accumulation of collagens being associated with insulin resistance and obesity (57, 58).

Our work in rodent models has demonstrated that chronic hyperglycemia is associated with glucose-induced modifications to the extracellular matrix (ECM) of skeletal muscle, including glycation and abnormal ECM expansion/remodeling (25, 26). This chronic hyperglycemia-associated ECM expansion was associated with blunted improvements in CRF in response to aerobic training, along with impaired aerobic remodeling of skeletal muscle including oxidative fiber-type switching and angiogenesis (25, 26). Treatment of hyperglycemic mice with canagliflozin prevented excess muscle ECM deposition and maintained aerobic adaptation to training (25).

Studies from cultured muscle lines also demonstrate that high-glucose conditions lead to increased ECM gene expression and a less oxidative muscle phenotype (26). Plating endothelial cells on ECM modified by glycation can blunt tube formation in vitro, indicating that exposure to high glucose conditions may inhibit angiogenic remodeling by altering the ECM (26). Thus, preclinical and in vitro studies point to chronic hyperglycemia-induced ECM accumulation as a potential mechanism for failed muscle remodeling with exercise. This hypothesis is supported by data demonstrating that altered muscle ECM gene expression is associated with the magnitude of improvement in CRF with aerobic training in humans (59).

PART 3: MOLECULAR MEDIATORS OF CRF WITH METABOLIC DISEASE

Exercise training-induced morphological adaptations in skeletal muscle such as angiogenesis occur as a cumulative result of molecular signals initiated during each acute bout of exercise (60). Signaling events that are specific to the type of exercise or muscle contraction, such as protein phosphorylation, lead to altered gene expression, which over time, can result in altered protein expression and changes in muscle phenotype (60). Thus, it is likely that phenotypic changes that are associated with low CRF in people with metabolic disease occur, at least in part, due to changes in signal transduction and gene transcription with acute exercise.

Altered exercise-induced signal transduction in skeletal muscle

Using a bioinformatics approach, we determined that the acute response to aerobic exercise in rat models of low response to training was characterized by hyperactivation of the c-Jun N-terminal kinase (JNK), while high-responders had little JNK activation with exercise (24). JNK has been well-characterized as a stress-activated kinase, but its role in muscle remodeling was previously unknown. Our recent work demonstrates that JNK activation with exercise in LRT leads to increased phosphorylation of a downstream target, SMAD2, in its linker region (pSMAD2-L) (24, 61). Furthermore, we show that JNK-mediated SMAD2 phosphorylation leads to inhibition of the Transforming Growth Factor β/Myostatin pathway (61), which is an important regulator of muscle phenotype. Activation of other prominent mediators of muscle remodeling with exercise, including AMPK and Akt, were not divergently regulated in low and high responders to training (24).

To determine whether this novel JNK/SMAD signaling axis mediates the response to aerobic exercise, we performed exercise training studies in muscle-specific JNK knockout mice. Ablation of JNK/SMAD2 signaling resulted in increased muscle capillary density, a higher proportion of oxidative muscle fibers, and a greater increase in exercise capacity with aerobic training compared to controls; thus identifying JNK activation in muscle as a negative regulator of aerobic training response. Our experiments also determined that JNK/SMAD2 signaling in muscle is activated by mechanical stress such as muscle stretch, in contrast to other exercise signaling molecules such as AMPK, which are activated primarily by metabolic stress, but not stretch (61). In addition, we found that JNK/SMAD is typically activated in healthy humans during resistance (loaded knee extension), but not endurance (cycling) exercise (61).

We next performed acute exercise experiments in mice and humans to determine if JNK/SMAD2 signaling is associated with low improvements in CRF that occur with metabolic disease. Our data demonstrate that chronic hyperglycemia in mice and impaired glucose tolerance in humans is associated with exacerbated JNK activation during acute endurance exercise and poor aerobic capacity (26). Treatment of hyperglycemic mice with canagliflozin was able to prevent JNK hyper-activation with acute exercise in conjunction with improved aerobic remodeling of skeletal muscle and improvements in aerobic capacity with training (25). Importantly, we demonstrated that increased JNK activation with hyperglycemia is a progressive signaling defect that worsens with longer durations of hyperglycemia, but is not caused by acute exposure to hyperglycemia during exercise (26). In humans and mice, JNK activation with acute exercise is associated with muscle extracellular matrix accumulation (25, 26), indicating that changes to muscle phenotype that occur with metabolic disease may impact muscle signaling with acute exercise.

In aggregate, these findings identify the JNK/SMAD2 axis as one mechanism underlying the blunting effects of chronic hyperglycemia on aerobic remodeling in individuals with impaired glucose metabolism. Whether interventions can reverse this signaling defect in humans with metabolic disease to improve CRF with training remains to be determined. Given the complex signaling response of skeletal muscle to exercise, it is likely other, yet unidentified, signaling mechanisms can contribute to impaired remodeling in low responders.

Altered exercise-induced gene expression in skeletal muscle

In addition to aberrant exercise-induced signal transduction, there is evidence that an altered transcriptional response to acute exercise may occur in metabolic disease. Pillon et al. demonstrated that the transcriptional response to 30-min of cycling exercise was significantly different in men with type 2 diabetes compared to those with normal glucose tolerance (62). Specifically, men with diabetes demonstrated enrichment of genes associated with inflammation and extracellular matrix remodeling, and reduced expression of genes associated with metabolic control (62). These data are consistent with data from our laboratory and others demonstrating an altered inflammatory and ECM remodeling signatures in muscle from rodent models of low exercise response and impaired glucose tolerance in response to exercise (24, 63). In contrast, some exercise-induced genes, such as GLUT4 and PGC1α are normally regulated in men and women with type 2 diabetes (64).

Available literature regarding the mechanisms of LRT have pooled data from males and females, or studied a single sex. However, increasing evidence demonstrates that sex-differences in signaling and gene regulatory response of skeletal muscle to exercise training likely exist (65). A recent large-scale meta-analysis on this topic pooled the findings of multiple studies encompassing different exercise modalities (66). Together, these studies reported 247 differentially expressed genes in response to exercise between males and females (66). However, whether or not metabolic disease may differently affect gene expression in response to exercise in males and females remains unclear, and should be a topic for future study.

SUMMARY

In summary, several factors have been associated with blunted improvements in CRF with metabolic disease. At the systemic level, both chronic hyperglycemia and insulin resistance appear to be linked to attenuated CRF responses to aerobic training (Figure). In skeletal muscle, blunted exercise-induced angiogenesis, a less oxidative fiber-type, impaired mitochondrial function, and extracellular matrix expansion may contribute to impaired adaptive response to exercise. At the molecular level, changes in muscle signal transduction and gene transcription with acute exercise likely contribute to the low-responder phenotype. Given the importance of improving CRF to reducing morbidity and mortality, a better understanding of how metabolic disease impacts exercise response will be essential for developing therapies to improve health span and lifespan.

Figure. Mechanisms associated with blunted improvements in CRF in people with metabolic disease.

Systemic factors that accompany metabolic disease, such as chronic hyperglycemia and insulin resistance are associated with blunted improvements in CRF with aerobic training. In skeletal muscle, ECM accumulation is associated with impaired aerobic remodeling, including lower muscle blood flow and capillary density, a less oxidative muscle fiber type, and impaired mitochondrial function. At the molecular level, altered signal transduction with exercise and altered gene expression likely contribute to the impaired adaptive response to training. Figure was created with Biorender.com.

Key points:

• Having low cardiorespiratory fitness (CRF) is associated with increased risk for cardiovascular disease and mortality

• Individuals with metabolic diseases, such as insulin resistance and diabetes, have lower CRF than those without metabolic disease

• Metabolic disease is linked to blunted improvements in CRF with aerobic training, or Low Response to Training (LRT)

• Impaired skeletal muscle signaling and aerobic remodeling may be key mechanisms underlying LRT in individuals with metabolic dysfunction

• Glucose-lowering and insulin-sensitizing treatments have been identified as potential therapies to improve the response to training