Walking and running are the most common human gaits. Humans walk at low speed and change gait to increase their locomotion speed, minimizing the energy expenditure. Thus, at speeds below walk-run transition speed, the oxygen consumption of walking is lower than that of running, while at higher speeds the relationship is reversed.¹ The energetic cost to travel a given distance is called the cost of transport (C), and has long been known to strongly depend on speed in human walking.^{2, 3} The cost is minimized at intermediate walking speeds of 4.5–5.4 km·h⁻¹ (or 1.25–1.5 m·s⁻¹) and grows as speed increases above or decreases below this optimum value. This trend is related to the pendular transfer of kinetic and potential energy, which is greater at intermediate speeds, ⁴ reducing the total mechanical work that must be performed by muscles at these speeds. Equally, the described energetically optimal walking speeds are in accordance with the contractile physiology of skeletal muscle and the biomechanical models of terrestrial locomotion.⁵

Humans are exceptionally economical walkers ^{2, 6} with their long legs and capacity to store and recover elastic strain energy.⁷ If the activity of individual muscles is minimized at intermediate walking speeds, as suggested above, we would expect the various muscles of the human locomotor system to be tuned to the same optimal speed to reduce the energetic cost of locomotion.⁵

Patients with chronic heart failure (CHF) and heart transplant (HT) experience fatigue and dyspnea during daily activities as well as during maximal exercise testing.^{8, 9} HT is the preferred therapeutic strategy for the treatment of end-stage CHF. HT patients have different responses to physical exercise because of sustained pre-transplant physical conditions. Wilson el al.,⁸ suggested that nutritive flow to skeletal muscles was defective in patients with CHF, and that their diminished maximal exercise capacity was correlated to the degree of impairment of nutrient flow. The exertional fatigue suffered by these patients was attributed to skeletal muscle underperfusion.^{8, 10} However, Wilson et al.,¹¹ demonstrated that patients with CHF and exertional fatigue were limited by skeletal muscle abnormalities rather than by skeletal muscle underperfusion. Further studies indicated that early skeletal muscle anaerobic metabolism, with lactate production, is the primary factor limiting exercise performance in patients with this disorder.^{9, 11} The skeletal muscle metabolic response to exercise is altered in patients with CHF.

Studies with EMG activity in CHF patients demonstrated that surface EMG can be used to detect skeletal muscle fatigue, altered muscle activity and atrophy.^{12, 13} Information of muscle activity in groups of CHF and HT showed high coactivation and high “global” EMG (called electromyographic cost) when compared to healthy controls.¹⁴ The correlation between metabolic and electromyographic variables can help to understand the changes in patients walking parameters and propose adequate goals during physical rehabilitation.

It is known that patients with cardiac diseases undergo musculoskeletal peripheral disorders, as well as anomalies in oxygen consumption, cost of transport, electromiographic activity and muscular coactivation.^{11, 13} In a previous study, it has been shown that ventilatory efficiency disorders in CHF patients could explain some changes during physical activity.¹⁵ However, others factors should be taken in account, like the connection between metabolic and electromyographic variables proposed in the present study.

The aim of this study was to verify correlations between metabolic (oxygen consumption and cost of transport) and electromyographic (coactivation and electromyographic cost) variables, as well as to obtain the regression equation over a range of walking speeds in CHF and HT patients and healthy controls. A strong positive correlation was hypothesized between cost of transport and electromyographic cost, and between oxygen consumption and the main muscle coactivations in all of the experimental and control groups.

Better understanding of complex interactions involved in measuring mechanical work and efficiency might best be obtained using a multidisciplinary approach.²⁸ The present study proposed an evaluation of the relationship between electromyographic and physiological variables, using specific (coactivation percentage) and global (C and CEMG) measures. The global measures reflect the sum of more localized contributions throughout the body.

The linear regression analyses showed a strong relationship between CEMG and C in the three analyzed populations. However, the metabolic cost appears slightly underestimated in the controls, but the CEMG seems to be a bit overestimated, with HT set in between, with the slope at nearly 1 (Fig. 1). As shown, it is possible to estimate the C towards the CEMG in the three analyzed populations. CEMG could replace C in some clinical situations: when a gas analyzer is not available, when using a mask is not recommended (like in cerebral palsy), or when the population present some alteration in respiratory parameters (like in a respiratory disease), which could to change the real values.

A strong and significant correlation was found in all the analyzed groups, between the CEMG and the metabolic C, and between oxygen consumption and the main muscle coactivation. According to Bona et al.¹⁷ the mechanical work was higher, while the mechanical efficiency was lower in CHF and HT, when compared with the controls. Those differences were partially due to differences of SSWS. However, muscle activation and CEMG also play a secondary role.¹⁴ According to Williams (1985),²⁸ muscular efficiencies are typically between 20-35% for walking. Mechanical efficiency during walking was near 25% at SSWS. For CHF and HT, the mechanical efficiency during walking achieve values near 20% at the same speed of the control group.¹⁷

The EMG technique presents some limitations, much like other physiological evaluations. EMG recordings provide a local quantitative measure of total muscle fiber activity, and changes in total rate of oxygen uptake result primarily from changes in the oxygen uptake of the general working of muscles.²³ A strong relationship was reported between EMG data and mechanical work.²⁹ Decrease in the recovery index and increase in the external mechanical work in CHF and HT patients is determined by a reduction in the pendulum-like energy interchange due to an altered trajectory of the COMB.¹⁷ The authors suggest that the reported differences in external mechanical work and recovery could determine an increased muscle activity to sustain locomotion. All these data confirm the relationship between EMG data and biomechanical parameters during walking.

Analysis of the energy recovery showed that, according to literature, the exchange between kinetic energy and potential energy was limited, especially at slow speeds, while it reached maximum values at the higher walking speeds.^30-32 By optimizing the system conversion, less metabolic energy is needed to keep walking. The SSWS adopted by CHF and HT patients directly determines the efficiency of the pendulum mechanism (recovery). Reducing their speed, patients increase the stability of walking, but increase the variability of movements as well,^{33, 34} a factor that could affect the muscular activity and, therefore, the C. Improvements in walking speed, leading to the optimal walking speed, would increase the vertical displacement of the COMB, providing larger oscillations of kinetic energy and potential energy of the COMB and, generally, an improvement of the efficiency of the pendulum mechanism of walking.

The rate of oxygen consumption was significantly correlated to leg, trunk, and arm muscle activation in all groups, suggesting that these are key factors affecting exercise performance. This finding suggests that dynamic muscle coactivation was an important factor specific to CHF and HT as a quality of movement index. The authors estimated coactivations in both stance and balance phases of walking, and at specific phases and requests of movement. These data support the idea that peripheral muscle limitations play an important role in people with CHF and HT, confirming and extending prior studies.^{14, 17} The same behavior has been shown for a number of other muscles and for a variety of different types of dynamic contractions.²⁹

There is compelling evidence that skeletal muscle dysfunctions play an important role in exercise intolerance in CHF. The problem involves the large muscles of locomotion, small muscles of the arms, and even respiratory muscles. These changes lead to increased muscle fatiguability, decreased oxidative metabolism, increased oxidative stress, and ineffective high energy phosphate, most likely resulting in an early accumulation of lactate during walking.

The beneficial effect of exercise training on reducing skeletal muscle alterations in these patients demonstrates that this process is reversible.³⁵ The changes lead to an improvement in peak VO₂ and lactate threshold, and delayed onset of anaerobic metabolism. Exercise training also has anti-inflammatory effects, and, as described above, can increase local expression of the anabolic peptide insulin-like growth factor 1. Additionally, reduced sympathetic hyperactivation and improved endothelial dysfunction with exercise training contribute to improved muscle blood flow and clinical performance.³⁵

One recent study³⁶ found increased coactivation in people with multiple sclerosis when compared to healthy subjects walking at similar speeds. This fact appeared not to be due to different gait velocities for this population; by contrast, it was likely an adopted strategy, suggesting that this mechanism may be part of a characteristic implemented pattern that may explain differences in gait ability. For CHF and HT patients, this increase in percentage of coactivation may well contribute to higher C and CEMG,¹⁴ lower recovery, and mechanical efficiency at SSWS.¹⁷

Another study³⁷ analyzed the activity of agonist and antagonist muscles during gait in patients with multiple sclerosis, demonstrating that abnormal lower limb muscle coactivation occurs during the stance phase of gait. This increased lower limb muscle coactivation during gait is an adaptive strategy that tends to compensate for muscle weakness. The relationship between muscle coactivation during gait and disability, postural stability, joint stiffness, and gait performance are regulated by variations in the forces produced. This is important, since an inappropriate coactivation (excessive and/ or prolonged) reduces gait performance by reducing gait speed and increasing metabolic cost. The results of the present study increase understanding of the function of coactivation in cardiac patients.

Research in medicine sports for decades has demonstrated that the skeletal muscle function improved through exercise training, which reduces the debilitating symptoms of chronic heart failure through its effects on the cardiovascular and musculoskeletal systems. Analysis of muscle activation can provide important clinical information and the ability of patients to produce a desired functional outcome.³⁶ In these way, mortality and hospital admission are significantly reduced after exercise training in patients with CHF, and this benefit was not restricted to any particular subgroup of patients.

The present study proposes that for CHF and HT patients who have SSWS lower than OWS, rehabilitation programs should focus on improving the speed of exercise, such as dynamic muscle contraction (and coactivation) for a better efficiency and economy of movement. Thus, further studies are necessary to evaluate other populations, like chronic obstructive pulmonary disease, multiple sclerosis, or cerebral palsy with these methods.

It is important to consider some limitations of the present study. The SSWS on the ground, for all groups, was greater than SSWS on the treadmill. This, as described by Figueiredo et al. (2013),¹⁵ may have affected the results (see Table 1). SSWS are expected to be lower on the treadmill than on the ground, hence the control group's lower SSWS. The assessment of the SSWS is an important task in an attempt to define the characteristics of walking and the physical tolerance of each group. Our CHF patients showed only a slight decrease in function capacities and, for this reason, our outcomes cannot be extended to patients with more severe CHF. However, with the current clinical therapies, more and more patients of CHF clinics have a profile equivalent to that of our group.

The aim of this study was to verify the correlation between oxygen consumption and coactivation, as well as the correlation between C and CEMG in CHF and HT patients and healthy controls. Accordingly, the results highlighted complementary strategies to reduce the C caused by muscle economy denominated CEMG, in addition to the inverted pendulum mechanism. More specifcially, muscle coactivation seemed to play a role in the increase of oxygen consumption throughout postural stability, stance, and balance phases for the entire gait cycle to determine total body energy consumption. Identification of inappropriate activation and how this may lead to specific locomotor deficits could lead to more effective training interventions.