Summary
Slow oxidative muscle, most notably the soleus, is inherently well equipped with the molecular machinery for regulating blood-borne substrates. However, the entire human musculature accounts for only ∼15% of the body’s oxidative metabolism of glucose at the resting energy expenditure, despite being the body’s largest lean tissue mass. We found the human soleus muscle could raise local oxidative metabolism to high levels for hours without fatigue, during a type of soleus-dominant activity while sitting, even in unfit volunteers. Muscle biopsies revealed there was minimal glycogen use. Magnifying the otherwise negligible local energy expenditure with isolated contractions improved systemic VLDL-triglyceride and glucose homeostasis by a large magnitude, e.g., 52% less postprandial glucose excursion (∼50 mg/dL less between ∼1 and 2 h) with 60% less hyperinsulinemia. Targeting a small oxidative muscle mass (∼1% body mass) with local contractile activity is a potent method for improving systemic metabolic regulation while prolonging the benefits of oxidative metabolism.
Introduction
By ∼2010 over half of American adults and 80% of those >65 years old had either prediabetes or diabetes (Menke et al., 2015; Xia et al., 2022). There is also currently a high prevalence of prolonged sitting between 9 and 11 h/day (Craft et al., 2012; Healy et al., 2015; Matthews et al., 2018; van der Berg et al., 2016) at a low metabolic rate during seated behaviors (Newton et al., 2013), especially in people who are at high risk for age-associated metabolic diseases such as metabolic syndrome and type 2 diabetes (van der Berg et al., 2016). Even in nondiabetics, postprandial glucose concentration in the 60–120 min range of an oral glucose tolerance test (OGTT) has often been described as one of the strongest independent metabolic risk factors for chronic disease because of linkages to Alzheimer disease (Kakehi et al., 2018; Ohara et al., 2011), neuropathies (Buysschaert et al., 2015; Papanas et al., 2011), dyslipidemia (DeFronzo and Abdul-Ghani, 2011; Festa et al., 2004), and cardiovascular conditions (DeFronzo and Abdul-Ghani, 2011; Succurro et al., 2009). Of concern, glucose tolerance is relatively difficult to improve by a meaningful amount during most therapies, including after substantial amounts of weight loss or exercise (Jansen et al., 2022; King et al., 1995; Knudsen et al., 2014; Magkos et al., 2016; Rose et al., 2001; Ross et al., 2000, 2015; Slentz et al., 2016).
There is no doubt that inactive muscle fibers require little energy (Dela et al., 2019; Kelley et al., 1994; Rolfe and Brown, 1997) and that the whole-body oxidative metabolism is low throughout many hours of the day when sitting with inactive muscles (Newton et al., 2013); this may be one of the most fundamental yet overlooked issues guiding the way toward discovering metabolic solutions to assist in preventing some age-associated chronic diseases. During periods of inactivity, skeletal muscle accounts for only ∼15% of the whole-body postprandial glucose oxidation in nondiabetic controls of similar age and BMI as in the present studies (Kelley et al., 1994), despite being the body’s largest lean tissue mass (∼21–31 kg in women and men) (Heymsfield et al., 2022). Consistent with this, multiple studies using the arteriovenous balance method of the lower limb have calculated that the oxygen consumption (VO2) of inactive muscle is ∼1–2 mL/min/kg Dela et al. (2019); Kelley et al. (1994); (Rolfe and Brown, 1997). Therefore, during acute inactivity, the muscle-mass-specific VO2 (in units of mL/min/kg muscle) is even less than the modest value of ∼3.0–3.5 mL/min/kg body weight for the basal metabolic rate lying down or during prolonged sitting throughout the day in a whole room calorimeter (Newton et al., 2013). Thus, contrary to a common notion, even though skeletal muscle is the body’s largest lean tissue mass, it is unlikely the dominant contributor to the oxidative metabolism of either glucose or lipids when sitting at resting energy expenditure. The prevailing perspective (mostly from epidemiology) has been that there is a whole-body metabolic rate threshold that must be exceeded to induce a robust gain in metabolic health responses. Furthermore, the specific muscles recruited and types of contractile activity have largely been disregarded in human research. Taking a step back, we took a more physiological perspective in the current experiments. Herein, we tested the direct and immediate effects of sustaining a high duration of elevated oxidative muscle metabolism when sitting. We used 2 guiding principles in our approach.
First and foremost, as described earlier, the energy demand is minimal in resting muscle fibers. Therefore, mitochondrial oxidative phosphorylation is capped at a relatively low ceiling during inactivity (Dela et al., 2019; Kelley et al., 1994; Rolfe and Brown, 1997). Related to this, the elevated energy demands and fuel requirements for carbohydrate oxidation quickly come to an end when an exercise bout ends (Horton et al., 1998; Wasserman et al., 1991). For these reasons, there is a need for understanding the biochemical effects of sustaining an elevated rate of oxidative metabolism by skeletal muscle, but with a subtle rate of whole-body energy expenditure.
Second, slow oxidative muscle has multiple intrinsic molecular and phenotypic features favoring specialization in prolonged contractile activity, in part because of the capacity for using more blood-borne fuels and hypothetically less glycogen in some physiological conditions; this is supported by animal (Bey and Hamilton, 2003; Cartee et al., 2016; Deshmukh et al., 2021; Halseth et al., 1998; James et al., 1985; Mackie et al., 1980; McDonough et al., 2005; Terry et al., 2018) and human (Deshmukh et al., 2021; Gollnick et al., 1974a, 1974b; Jensen et al., 2012; Johnson et al., 1973; Murgia et al., 2021) studies that have long described the heterogeneous qualities between different fiber types within a muscle and between different muscles. The soleus has a greater predominance of slow-oxidative fibers (∼88% of the soleus mass is type I slow-twitch fibers) than 36 other human muscles that have also been fiber typed (Johnson et al., 1973). The soleus is a slow-twitch postural muscle that has motor neurons and other features favoring a lower threshold of effort needed to recruit it for more time and intensity than other limb muscles (Hodgson et al., 2005; Monster et al., 1978). Compared with other leg muscles, highly controlled studies in rodents have found the soleus has a phenotype favoring more uptake of both plasma TG (Bey and Hamilton, 2003; Mackie et al., 1980) and blood glucose (Halseth et al., 1998; James et al., 1985). It has distinctive vascular features enhancing delivery of blood-borne fuels and oxygen (McDonough et al., 2005), relatively high levels of hexokinase II and GLUT4 (Jensen et al., 2012), and a relatively low concentration of glycolytic enzymes and glycogen phosphorylase (Gollnick et al., 1974a, 1974b). However, walking can cause rapid rates of glycogen depletion in the soleus as in other muscles (Jensen et al., 2012). Therefore, it is far from certain whether there is an effective physiological approach to capitalize on the phenotype of this slow oxidative muscle to improve systemic lipid and glucose metabolism.
The scientific challenges and potential impact of developing a method for raising oxidative metabolism locally by a small tissue is perhaps best understood in light of the already much more established scientific interest (Chen et al., 2020) in activating another small tissue with an oxidative phenotype, brown adipose tissue (BAT). The soleus (Bey and Hamilton, 2003; Halseth et al., 1998; James et al., 1985; Jensen et al., 2012; Mackie et al., 1980; Petersen et al., 2003; Song et al., 1999) and BAT (Chondronikola et al., 2014; McNeill et al., 2020) are both tissues making up too small a percentage of total body mass to alter energy expenditure unless methods are developed to cause an intense local metabolic activation. Yet under some conditions, both might possibly be equipped with a phenotype favoring exceptional metabolic rates over prolonged periods of time. The specific questions we posed are analogous to the hurdles already faced in BAT research; how can people consistently activate tissue specific oxidative metabolism at a meaningful rate to increase whole-body oxygen consumption and then sustain it for hours at a time? Even if methods were developed to make that possible, would raising the local metabolic rate by a small mass of tissue be sufficient to impact systemic metabolic parameters as complex as very-low-density lipoprotein (VLDL)-TG concentration and postprandial glucose tolerance?
This work was part of an effort to develop a method of muscular contractile activity specifically geared for sustaining the possible distinct benefits of oxidative metabolism for prolonged periods, instead of sitting with inactive muscle at a low metabolic rate. The present experiments were designed to test the potential physiological influence of the human soleus muscle during hours of prolonged contractile activity.
Discussion
These results indicate that the human soleus muscle is capable of raising, and sustaining for hours, the local rate of oxidative metabolism to high levels. From a physiological perspective, this kind of contractile activity was effective at improving systemic metabolic regulation quickly and by a biochemically meaningful amount to improve glucose regulation (Figure 3), even at the lowest SPU intensity studied (SPU1; Tables 2, 3, and 4 and Figure 3). Furthermore, the high level of endurance of the soleus during SPU contractions provides a tool for reversing the otherwise slow rate of muscle metabolism during long periods of inactivity (Dela et al., 2019; Kelley et al., 1994; Rolfe and Brown, 1997). We are unaware of any existing or promising pharmaceuticals that come close to raising and sustaining whole-body oxidative metabolism at the magnitude in the current study (Tables 1 and 2), including drugs that may activate BAT (Loh et al., 2019). A beta-3 agonist given at double the Food and Drug Administration (FDA)-approved dose increased energy expenditure in 3 h, but by only an average of ∼8.5 kcal/h (Loh et al., 2019). By comparison, SPU contractions were capable of raising the energy demand 10-fold greater (91 kcal/h above control during the 270 min of activity in our first experiment; Table 1), without any evidence of progressive fatigue or other physiological limitations in the unfit individuals studied (Table 1).
We learned that through this method of contractile activity specifically geared for sustaining oxidative metabolism, only an extra 100–200 mg/min of local carbohydrate oxidation (Table 2) by a small muscle mass was potent enough to improve glucose regulation by a large amount after ingesting a large glucose load (Table 3). As described more later, carbohydrate oxidation increases profoundly during the acute minutes of relatively heavy whole-body exercise. However, there is not compelling evidence to believe that either in humans or animal models, carbohydrate oxidation remains elevated by even a small amount in the hours after ending an acute exercise bout (Horton et al., 1998; Wasserman et al., 1991). There was not an increase in carbohydrate oxidation after large muscle mass exercise within subgroups of people who had a significant excess post-exercise oxygen consumption (EPOC) (Horton et al., 1998). This finding of no increase in post-exercise carbohydrate oxidation above sedentary control levels was also demonstrated when experimentally changing the insulin concentration over a wide range (Wasserman et al., 1991).
There is a competition at the cellular level between various substrates to supply the necessary energy for contractions. During exercise, muscle glycogen is generally the primary carbohydrate to fuel contractions and often accounts for about 72%–95% of the carbohydrate oxidation when it is available in normal concentrations (Bergman et al., 1999; Helge et al., 2007; Horowitz et al., 1999); this is observed even with lower intensity exercises such as cycle ergometry at 20%–30% of VO2 max (Gollnick et al., 1974a, 1974b) or quadricep extensions at 25% of the local VO2 peak (Helge et al., 2007). During isolated knee extensions at a local VO2 of 190 mL/min/kg muscle (when roughly comparable to SPU contractions in Figure 2C), the VL muscle depleted twice as much glycogen in 35 min (Helge et al., 2007) as the soleus did during 270 min of SPU contractions (Table 1); this is consistent with the possibility that with SPU contractions, the soleus may in some conditions deplete glycogen at a rate 10–15 times slower than the VL muscle of the thigh during knee extensions (Helge et al., 2007) or cycling (Bergman et al., 1999; Gollnick et al., 1974a, 1974b). A thorough review by Sylow and Richter concluded that blood glucose accounts for only ∼10%–18% of the energy during the time frame of a typical exercise session (Sylow et al., 2017). By comparison, these types of large muscle mass exercises (Bergman et al., 1999) potentially increase the total carbohydrate oxidation more than 10-fold greater than SPU contractions (Table 2). With cycle ergometry at ∼5 METs for 1 h, the rate of total carbohydrate oxidation was ∼1600 mg/min (Bergman et al., 1999), which was many times greater than SPU contractions could induce; this is also consistent with the carbohydrate oxidation in another study during 60 min of moderate intensity cycling combined with glucose ingestion (Horowitz et al., 1999). However, once glycogen is accounted for, the remaining carbohydrate oxidation due to blood glucose by the entirety of both legs during cycling was ∼180 mg/min (Bergman et al., 1999), which was relatively similar to the estimated rate of carbohydrate oxidation by the working soleus muscle during SPU contractions (Table 2). Prior work has suggested that when recruiting a smaller instead of a larger mass of muscle, there may be a shift in the relative reliance on fuels from glycogen to utilization of more lipids (Helge et al., 2007) and/or more blood glucose (Richter et al., 1988) to fuel contractions. A plausible hypothesis is the intrinsic metabolic phenotype of the soleus (Bey and Hamilton, 2003; Gollnick et al., 1974a, 1974b; Halseth et al., 1998; Hodgson et al., 2005; James et al., 1985; Jensen et al., 2012; Mackie et al., 1980; McDonough et al., 2005; Monster et al., 1978) requires less glycogen to fuel contractions, especially when there is not epinephrine stimulation (Richter et al., 1982). As described below, in addition to local muscle glycogen use, there are also systemic processes that tend to offset the ability of contractions to attenuate postprandial hyperglycemia during and after large muscle mass exercise (Devlin et al., 1989; Hamilton et al., 1996; Knudsen et al., 2014; Maehlum et al., 1978; Steenberg et al., 2020).
Results from many studies support the concept that it is often difficult to improve postprandial glucose tolerance. Studies have identified specific mechanistic explanations for why oral glucose tolerance is generally not improved (Devlin et al., 1989; Hamilton et al., 1996; Knudsen et al., 2014; Maehlum et al., 1978; Rose et al., 2001) or even made worse (Flockhart et al., 2021; Knudsen et al., 2014; Rose et al., 2001) in the hours following traditional types of exercise that would rely predominantly on muscle glycogen as the source of carbohydrate. Those observations include people with normal glucose tolerance (Flockhart et al., 2021; Knudsen et al., 2014; Rose et al., 2001) to people with diabetes (Knudsen et al., 2014). The mechanistic explanations to date include compelling evidence of increased insulin resistance within the unrecruited muscle fibers after exercise sessions (Devlin et al., 1989; Steenberg et al., 2020), elevated rates of glucose appearance into the bloodstream (Hamilton et al., 1996; Knudsen et al., 2014; Maehlum et al., 1978), and also possible impairment of intrinsic mitochondrial function from intense exercise training (Flockhart et al., 2021). In studying the chronic effects of exercise training, a landmark study reported that there was a modest 13 mg/dL decrease in the 2-h glucose without a significant improvement in the entire AUC; this only occurred when doubling the recommended weekly volume of exercise training combined with a vigorous intensity (7 METs) and nutritional conditions, allowing for a 4.6 cm reduction in waist circumference (Ross et al., 2015). Other carefully controlled studies showed unequivocally no reduction in the 2 h glucose (or AUC) with progressive weight loss of 5%, 11%, or even 16% (Magkos et al., 2016); this was confirmed in another large experimental weight loss study after inducing 15% weight loss (Jansen et al., 2022). Nevertheless, given that carbohydrate oxidation is not elevated after an acute exercise session ends, the post-exercise glucose utilization by muscle is apparently restricted to the nonoxidative pathways. Therefore, as described next, there has been the need to better understand the immediate and direct effects of contractile activity.
The present study tested the effects of sustaining continuous contractile activity, at two levels of energy demand, throughout the entire 3-h postprandial period. The contractions were already attenuating glucose concentration quickly within the first 30 min when glucose was still rising (Table 3). Thereafter, the glucose improvements continued to increase and remained significantly lower than the sedentary control level throughout the entire 3-h postprandial time course. Earlier studies also found that activity involving a large muscle mass (cycle ergometry at 59%–67% HRmax) completed after 45 min of the postprandial period transiently blunted the rise in blood glucose and insulin during the contractions (Aadland and Hostmark, 2008). However, this effect was short-lived and evident only between 30 and 45 min of the postprandial period, followed by significantly greater hyperglycemia than the inactive sitting trial after exercise stopped. Similarly, Kanaley’s group (Holmstrup et al., 2014) found that subjects with low glucose tolerance were able to decrease the hyperglycemia during 60 min of exercise (60%–65% VO2max). After stopping the exercise, the blood glucose increased above the level when they never exercised, and thus the total postprandial iAUC was not improved (Holmstrup et al., 2014). Although an intermittent exercise pattern slightly improved the glucose iAUC compared with the 60 min bolus of exercise, spreading out the contractile activity with brief breaks did not reduce the glucose iAUC compared with sitting inactive in their volunteers who had low glucose tolerance as the present participants (Holmstrup et al., 2014). That said, it is important to understand that as previously emphasized, there are distinct molecular and physiological processes impacting metabolic regulation with specific inactivity/activity approaches (Bey and Hamilton, 2003; Hamilton et al., 1998, 2004, 2007, 2014; Zderic and Hamilton, 2012). Here we have focused on a method of raising slow oxidative muscle metabolism to complement (not replace) existing approaches.
Prolonged sitting has become ubiquitous across the lifespan (Craft et al., 2012; Healy et al., 2015; Matthews et al., 2018; van der Berg et al., 2016) and is not significantly less in regular exercisers (Craft et al., 2012). But it is important to remember that regardless of whether ∼45 min of activity is administered as many separate brief breaks or as a single daily bolus, it mathematically increases the energy demand for muscular work for only ∼5% of the waking day. Thus, currently recommended activity approaches are a not a direct solution for a high amount of time when skeletal muscle has a low muscle metabolism. A large number of studies have focused on “brief breaks” from sitting inactive, e.g., 2–5 min activity breaks each half hour or 6 min once each hour (Henson et al., 2020; Larsen et al., 2015; Loh et al., 2020; Thorsen et al., 2019). What may be missed is that brief breaks are, by definition, brief amounts of contractile activity and thus only brief periods to potentially benefit from oxidative metabolism. A systematic review has reported that lifestyle interventions replacing sitting time with standing and/or brief walking breaks have decreased sitting time on average only 30.4 min/day (Peachey et al., 2020). That said, a promising early study reported that taking 3 brief interruptions each hour to walk at 2 mph (∼3 METs) may reduce the blood glucose by approximately 5, 15, and 6 mg/dL at 1, 2, and 3 h respectively, with the average insulin concentration (based on AUC) reduced by about 12% (Larsen et al., 2015). In a thorough recent study, no differences were observed for plasma glucose or insulin with any of the 3 patterns of walking breaks to interrupt inactive sitting in men with abdominal obesity (Thorsen et al., 2019). Henson et al. (2020) summarized results when combining 4 large postprandial lab studies and concluded that the glucose and insulin were not decreased by standing versus sitting in the postprandial period. Also, the brief walking breaks generally reduced glucose <10 mg/dL and thus were physiologically small and/or nonsignificant in some large categories of people (e.g., no statistical influence in males) (Henson et al., 2020). Physiological studies relevant to this found that muscle glucose uptake appears to have a relatively slow time course at the onset of some types of contractile activity (Bergman et al., 1999; Horowitz et al., 1999; Mossberg et al., 1993). In studying the time course in the human lower limb musculature during cycling, there was no increase in glucose uptake in the leg after either 5 or 15 min of continuous contractions, and significant responses were not evident until after 30 min (Bergman et al., 1999). Taken together, there is a biochemical basis for the need to develop methods to understand and benefit biochemically from prolonged contractile activity.
The literature is understandably filled with language about designing programs to “engage as much skeletal muscle mass as possible” and the necessity of “a sufficient amount of muscle mass” (Ivy et al., 1999; Laughlin, 2016). The findings presented here do not in any way interfere with traditional exercise programs for their own distinct benefits. There are hundreds of skeletal muscles and many microvascular and metabolic exercise training adaptations that are restricted to the recruited muscles (Laughlin, 2016). Thus, it is understandable that the paradigm for glucose management and other cardiometabolic outcomes has focused less on the quality of contractile activity than on the quantity of recruited muscle and raising the total metabolic rate to higher levels for short exercise bouts (e.g., 150 min/week). The Diabetes Prevention Program (DPP) and other recommendations either implied or explicitly stated that health-enhancing metabolic effects may require an energy expenditure in the >3 MET range (Knowler et al., 2002). The current findings obviously are inconsistent with the notion that all contractile activity near ∼1.3 or ∼1.7 METs is a “stepping stone” to the goal of exercising at >3 METs (Dunstan et al., 2021). All of this points to the “specificity principle” that is a key axiom in exercise physiology (Hamilton et al., 2007, 2014). Applied here, it means that the phenotype of the muscles recruited, and the duration of the elevated metabolic rate, will determine the distinct biochemical processes that regulate the effectiveness of physical activity.
There have more recently been national guidelines proposing that people sit less and/or move more in addition to traditional methods of exercise (Dunstan et al., 2021). Unfortunately, this advice is still lacking in specifics about how to reduce sedentary time enough for the most meaningful health gains. The rapid and large decreases in skeletal muscle TG uptake observed in rodent inactivity physiology studies (Bey and Hamilton, 2003; Hamilton et al., 2007) have been heavily cited by epidemiologists and clinical trials specialists to provide biological plausibility to understand the observational associations of diseases related to sedentary time. What has been overlooked is the same publications cited to provide the biological plausibility for why muscular inactivity is unhealthy also alluded to a potential solution; those studies were primarily based upon the large local molecular and biochemical responses in the soleus muscle (in rodents) that are dependent on prolonged contractile activity (Bey and Hamilton, 2003; Bey et al., 2003). The present findings provide evidence that the human soleus muscle has the potential to contribute to systemic metabolic regulation.
Finally, it might prove to be that the most interesting hypothesis raised by these results is that the human soleus, although only ∼1% of body weight, can sustain a sufficient metabolic rate for an impressive duration and improve glucose and lipid metabolism. Others have demonstrated significant beneficial correlations between the slow-red oxidative fiber type and chronic disease states (Gaster et al., 2001; Hickey et al., 1995). There has been interest in applying molecular biology techniques to therapeutically enhance the quality of skeletal muscle by increasing the amount of slow oxidative muscle fibers (Gan et al., 2013). It is important to note that the results were obtained from adults across the lifespan (22–82 years of age) and with a wide BMI range and habitual physical activity levels (Table 4). Findings reveal that the human soleus of these ordinary people was already physiologically capable of producing these responses. Looking back, studies in the 1870–1880s by Ranvier described the soleus remarkably well as a red muscle with curvy capillaries that is relatively slowly contracting and fatigue resistant, even when the muscle was obtained from highly sedentary animals such as the domesticated rabbits and cats (Ranvier, 1873, 1880). The SPU method is specifically geared for sustaining positive effects of prolonging an elevated muscle metabolism for hours (not minutes), but with a very subtle increase in whole-body energy expenditure while sitting (Figure S1). This low effort method by a muscle that is naturally geared toward prolonged contractile activity may avoid the potentially serious deleterious cardiac effects of prolonged endurance training in some people or the impairment of mitochondrial function after training with excessive exercise intensity (Eijsvogels et al., 2016; Flockhart et al., 2021). Many other questions will need to be addressed in order to understand the full translational potential. One understandable viewpoint is that prolonged periods of elevated muscle metabolism is an unrealistic expectation. Another perspective is that it is an opportunity for gaining the distinct health effects of elevating muscle metabolism by a biochemically meaningful amount.
Acknowledgments
We appreciate the financial support of the American Diabetes Association