NMES is excito-motor which means that it directly activates muscle fibres by bypassing motor neurons (even if they are simultaneously activated as well as sensory neurons). To this end, it must be clearly above the motor threshold, that is, the minimal intensity of stimulation that produces a direct motor response and generates involuntary non-controlled segmental movements. It turns out that the higher the intensity, the greater the extent of the electrical field and the greater the number of recruited muscle fibres – that is, both I and II fibre types (Collins, 2007). In the context of regular or chronic application, it appears that the higher the current intensity, the greater the physiological effects/benefits induced (Paillard, 2018). NMES produces strong muscle contractions and provokes pain through surface electrodes placed over the bellies or motor points of muscle(s) targeted (Lake, 1992; Vanderthommen and Duchateau, 2007). In order to recondition the basal functional capabilities in older (frail) subjects, NMES should be applied specifically to the muscles of their lower limbs with high intensities (maximal tolerable by subject), high frequencies (>30 Hz and rather 50–80 Hz), optimal width pulses (matching to the chronaxy of the stimulated muscle that is, for instance 300–450 μs for quadriceps femoris) short contractions interspersed with long recovery times – for example, 3–10 s/10–30 s on/off – for 10–15 min, 20 min maximum (Paillard, 2018).

Evidence suggests that NMES regularly applied on quadriceps femoris or dorsi/plantarflexor muscles (e.g., tibialis anterior, and soleus) in older subjects (>60 years old) improves lower-limb muscle strength (Caggiano et al., 1994; Paillard et al., 2004; Bezerra et al., 2011; Caulfield et al., 2013; Kern et al., 2014; Mignardot et al., 2015; Von Stengel et al., 2015; Mani et al., 2018; Acaröz Candan et al., 2019; Langeard et al., 2021) and postural balance (Amiridis et al., 2005; Paillard et al., 2005a,b; Nejc et al., 2013; Mignardot et al., 2015; Alptekin et al., 2016; Bondi et al., 2021).

In this context, the functional improvements of motor and postural functions in older subjects are mainly linked to enhancements of motor output through muscle structural (mass) and functional (neural networks) adaptations – Figure 2 (Paillard, 2018). The naturally irreversible atrophy of lower-limb muscle in inactive or sedentary individuals can be reversed with the chronic application of NMES (Boncompagni et al., 2007; Carraro et al., 2015). NMES stimulates not only anabolic pathways (e.g., secretion of insulin-like growth factor-1), but also negatively modulates muscle catabolism, which increases protein synthesis and reduces protein degradation and activates satellite cells in aged individuals (Barberi et al., 2015; Mancinelli et al., 2019). Hence, NMES induces an increase in the size of muscle fibres especially in the type II muscle fibres which are particularly affected by the effects of advancing age (Zampieri et al., 2015; Mancinelli et al., 2019). From a functional viewpoint, NMES results in greater contribution of muscles regularly stimulated (i.e., electromyographic activity of tibialis anterior and medial gastrocnemius) in the postural regulation (Amiridis et al., 2005). It also entails increased musculotendinous stiffness of the muscles regularly stimulated (Mignardot et al., 2015).

However, even if one cannot ignore that NMES applied on selected muscle groups engenders sensory information likely to be used as part of the command and control of voluntary movement – for example, during walking – Paillard (2020) inferred that it generates little sensory information likely to be exploited by the central nervous system to optimise and refine perceptual and motor skills as part of the postural balance regulation (Figure 2). In fact, NMES simultaneously activates both sensory and motor neurons and would provoke conflict between the antidromic motor action potentials and the orthodromic sensory action potentials at the spinal level (Bergquist et al., 2011), thus causing interference in the central integration of induced afferents (although the presence of central adaptations cannot be totally excluded).

In order to improve the functional capabilities, NMES would be particularly interesting since it is likely to enhance motor output especially muscle strength/power in older frail subjects with a high risk of falling precisely due to a lack of muscle strength/power.

SES is not excito-motor (it cannot directly activate the muscle fibres) and only (or almost only) activates the sensory pathways (this is the principle of this stimulation). To this end, it must be well below the motor threshold and close to the sensory threshold, that is, the minimum intensity of stimulation that can be perceived by individual. Indeed, the current intensity must be around the sensory threshold without exceeding it or only slightly, since a non-excito-motor current (low intensity) is already likely to lead to recruitment of type I muscle fibres from reflex pathways – mediated at spinal level, that is, homolateral monosynaptic connections of the Ia fibres with the α-motoneurons generate depolarisation of the latter and induce the contraction of the muscle fibres they innervate (Zeronian et al., 2021). SES is painless (although there are possible sensations) and applied to a peripheral nerve, belly muscles (motor points) and/or joints at current intensities below, at or slightly above the sensory threshold (Schröder et al., 2018; Paillard, 2021). In order to recondition the motor and postural functions in older subjects, SES should be applied with large pulses (e.g., 1 ms) and high frequencies (80–100 Hz) to facilitate potentiation and central integration of emitted signals (Collins, 2007; Bergquist et al., 2011) for several tens of minutes in a uninterrupted way.

Evidence suggests that SES regularly applied to lower-limb muscle (motor points), peripheral nerve or joints in older subjects (>60 years old) improves sensorimotor function (Park et al., 2014; Ng et al., 2016; Paillard, 2021). However, since SES generates no muscle deformation or contraction, it thus cannot stimulate synthesis of contractile proteins responsible for muscle structural adaptations (Coffey and Hawley, 2007). The adaptations induced cannot occur at the level of the motor output of muscle. In fact, SES generates sensory cues that are detected by sensory sensors which transmit signals throughout sensory pathways to cortical areas (Figure 1). With chronic application, SES induces adaptations at different stages (sites) throughout sensory pathways – sensory sensors and spinal and supraspinal structures (Figure 2).

At the peripheral level (i.e., sensory sensors), SES would modify the sensitivity threshold of mechanoreceptors that result from altering the ion permeability of their membrane (Gravelle et al., 2002; Ross et al., 2007). This sensory adaptation would optimise the potentiation and transmission of proprioceptive afferents and thus enhance the sensory contribution in the motor and postural regulation (Ross et al., 2007; Severini and Delahunt, 2018; Paillard, 2021). At the spinal level, SES is likely to engender a reduction or increase in the amplitude of the induced H-reflex (Goulet et al., 1997; Hardy et al., 2002, respectively) that, in both cases, can improve motor and postural abilities (Paillard, 2017a). The reduction of the Ia-afferent excitation to the α-motoneuron pool can attenuate destabilising joint movements while the enhancement of the activation of the α-motoneuron pool can facilitate instantaneous segmental reactions to the demands of the motor or postural task (Paillard, 2017a). At the cortical level, SES is likely to induce durable changes in motor cortex excitability that can be assimilated to enhance sensorimotor activity and connectivity (Kaelin-Lang et al., 2002; Schröder et al., 2018; Insausti-Delgado et al., 2021). Enhanced sensorimotor activity could result from favourable impact of SES over postcentral and parietal cortices (primary somatosensory cortex and posterior parietal cortex) to precentral cortices (primary motor cortex) in the stimulated hemisphere (Paillard, 2021).

SES would be of particular interest as it is likely to improve sensorimotor output in older subjects who are not at high risk of falling due to lack of muscle strength/power or very weak motor output.