SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

Compelling evidence exists that pain may affect the motor system, but it is unclear if different sources of peripheral limb pain exert selective effects on motor control. This systematic review evaluates the effects of experimental (sub)cutaneous pain, joint pain, muscle pain and tendon pain on the motor system in healthy humans. The results show that pain affects many components of motor processing at various levels of the nervous system, but that the effects of pain are largely irrespective of its source. Pain is associated with inhibition of muscle activity in the (painful) agonist and its non-painful antagonists and synergists, especially at higher intensities of muscle contraction. Despite the influence of pain on muscle activation, only subtle alterations were found in movement kinetics and kinematics. The performance of various motor tasks mostly remained unimpaired, presumably as a result of a redistribution of muscle activity, both within the (painful) agonist and among muscles involved in the task. At the most basic level of motor control, cutaneous pain caused amplification of the nociceptive withdrawal reflex, whereas insufficient evidence was found for systematic modulation of other spinal reflexes. At higher levels of motor control, pain was associated with decreased corticospinal excitability. Collectively, the findings show that short-lasting experimentally induced limb pain may induce immediate changes at all levels of motor control, irrespective of the source of pain. These changes facilitate protective and compensatory motor behaviour, and are discussed with regard to pertinent models on the effects of pain on motor control.


Databases

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information
  • Literature search in Medline, EMBASE and Web of Knowledge.
  • 112 papers satisfied the inclusion criteria: (1) effects of pain on motor parameters; (2) experimentally induced pain without structural damage; (3) pain localized in the upper or lower extremity; and (4) healthy human subjects.

What does this study add?

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information
  • Comprehensive overview of pain-induced changes at all levels of motor control.
  • Experimental limb pain may lead to protective and compensatory motor behaviour.

1. Introduction

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

Pain may have profound effects on motor behaviour that are mediated at various levels of the nervous system, ranging from spinal reflex circuits to (pre)motor cortices. The effects of pain thus may become manifest in a variety of motor parameters such as reflex amplitude, muscle activity, force production, kinematics, movement strategy and activation of cortical areas involved in motor control. Empirical studies typically focus on a single or a limited number of parameters, whereas pertinent descriptive models generally comprise only a selection of the complex interactions between pain and the motor system. Consequently, a coherent view on the consequences of pain on motor behaviour has been lacking. Only recently, Hodges and Tucker (2011) proposed a new theory on motor adaptation to pain involving changes at multiple levels of the motor system.

Acute intense pain has been thought to elicit motor responses that serve to protect the painful limb from further damage. Although such behaviour serves a clear short-term benefit for the injured part, it may have long-term negative consequences if it does not dwindle with healing of the initial injury. The presence of chronic pain may lead to abnormalities in motor control, either as a direct effect of pain or as a consequence of adopting a movement strategy that compensates for such direct effects (Hodges and Tucker, 2011). Conversely, abnormalities in motor control may lead to the development of (chronic) pain, e.g., when tissues are overloaded (Sterling et al., 2001; Arendt-Nielsen and Graven-Nielsen, 2008). In clinical pain conditions, multiple factors may play a role in the effects of pain on motor behaviour, rendering it difficult to disentangle the separate effects of nociceptive input on the motor system. Experimental procedures to induce pain in healthy volunteers allow for establishing clear-cut, reproducible cause-and-effect relations, and thus provide invaluable means to isolate motor consequences of acute pain.

The various procedures to induce pain in a controlled manner have specific advantages and methodological limitations (Staahl and Drewes, 2004; Graven-Nielsen, 2006; Arendt-Nielsen and Graven-Nielsen, 2008). Ideally, the effects of pain are studied by selective activation of nociceptive afferents without causing structural tissue damage. In reality, however, experimental procedures cannot exclusively target nociceptive afferents but also activate non-nociceptive afferents (Mense, 1993; Graven-Nielsen, 2006). Appropriate control conditions that take into account stimulation of these non-nociceptive afferents are required to draw inferences on the effects of pain on motor control.

This systematic review evaluates the effects of experimentally induced pain on the motor system in healthy humans in order to obtain more insight into the empirical evidence for interactions between pain and the motor system. The source of pain (i.e., skin, joint, muscle or tendon) is expected to have differential effects on motor control, considering that these tissues have different roles in the motor system and projections of nociceptive afferents may vary among various tissue types (Millan, 1999; Almeida et al., 2004). This review is limited to pain localized in the extremities and focuses on the effects of controlled external stimuli that lead to localized pain without causing structural damage.

2. Literature search methods

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

A literature search was performed in three electronic databases (Medline, EMBASE and Web of Knowledge) to identify potentially relevant studies. The search strategy was developed in collaboration with a specialist in information retrieval of the Leiden University Medical Center Library and comprised a combination of MeSH terms and free text terms related to motor function (including proprioception) combined with terms related to experimental pain (see Supporting Information Methods S1). The results were limited to articles in English, German and Dutch. The most recent search was performed on 7 March 2011.

The identified studies were first screened by title and abstract, after which the full text of potentially relevant articles was studied (see Supporting Information Methods S2). Papers were included if all of the following four criteria were met: (1) effects of experimentally induced pain on one or more of the following parameters were studied: spinal reflexes, muscle activity, movement characteristics, proprioception and activation in motor-related brain areas; (2) experimental pain was induced by a controlled stimulus leading to localized pain without causing structural damage (i.e., thermal, mechanical, electrical or chemical stimulation); (3) pain was localized in the upper extremity or lower extremity; and (4) data were obtained from healthy human subjects. Note that papers were excluded if pain was induced by ischaemia or eccentric exercise because effects of ischaemia are non-specific and eccentric exercise may lead to inflammatory reactions and structural damage to muscle tissue (Friden and Lieber, 1992). In addition, reference lists of all included publications as well as reviews on this topic were tracked following the procedure described above. In case there was any uncertainty about inclusion or exclusion, a second, independent reviewer was consulted. Discrepancies between the reviewers were to be resolved by consensus agreement. However, no such discrepancies arose during the process.

Data regarding the type, intensity and location of pain, the test protocol, outcome parameters, and performed motor tasks were extracted using a standard form. Outcome parameters were categorized into one or more of the following aspects of motor function: (1) spinal reflexes; (2) muscle activity; (3) task performance, movement kinetics and kinematics; (4) proprioception; and (5) brain activation. The systematic review of studies in each category is preceded by a short introduction providing a framework for the topic in question. All studies fulfilling the inclusion criteria are presented in Supporting Information Tables S1–8, which summarize the results on the effects of (sub)cutaneous pain, joint pain, muscle pain and tendon pain. If possible, pain intensity is categorized into mild [visual analogue score (VAS) <30 mm], moderate (VAS 30–54 mm) and severe pain (VAS >54 mm; Collins et al., 1997). Note that some authors did not explicitly report measures of pain intensity, whereas others expressed pain intensity as peak VAS, mean VAS or the area under the VAS curve (on the basis of which the average pain level was calculated). Results for each of the aforementioned aspects of motor function are summarized in the text, with n indicating the number of studies on which the description of results is based. A detailed description of the results is provided online (see Supporting Information Results S1).

3. Results

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

The distribution of studies over the various aspects of motor function (Fig. 1) reveals that research has concentrated on muscle activity and movement characteristics. These aspects of motor function were examined predominantly by means of experimentally induced muscle pain, whereas spinal reflexes and brain activation patterns were mainly examined by means of experimentally induced cutaneous pain. A direct comparison between different pain sources was made in only 9 out of 112 studies. Overall, the effects of (sub)cutaneous pain and muscle pain have received considerable attention, unlike the effects of tendon pain and joint pain.

figure

Figure 1. Overview of the studies included in this review, grouped by aspects of motor function and sources of pain under study. Several studies addressed more than one aspect of motor function and are thus presented in more than one column. Within a given column, each study can only appear once; the source(s) of pain under study (e.g., ‘muscle’ or ‘muscle vs. skin’) are indicated by various patterns.

Download figure to PowerPoint

3.1 Spinal reflexes

Spinal reflexes belong to the most basic elements of motor behaviour. On this level, the effects of experimentally induced pain have been studied for the nociceptive withdrawal reflex (NWR; n = 13), the phasic stretch reflex (n = 2), the H-reflex (n = 17) and inhibitory spinal circuits (n = 7; see Supporting Information Table S1).

The NWR (for reviews see Clarke and Harris, 2004; Sandrini et al., 2005) is a spinal reflex elicited by noxious stimulation of cutaneous afferents and has a ‘modular organization’ in animals (Clarke and Harris, 2004) and humans (Andersen et al., 1999, 2001, 2003; Sonnenborg et al., 2001; Schmit et al., 2003). The evoked motor response represents the most appropriate movement to withdraw the stimulated area from the offending source. Pain stimuli applied to the skin (Grönroos and Pertovaara, 1993; Andersen et al., 1994; Ellrich and Treede, 1998; Ellrich et al., 2000) or muscle (Andersen et al., 2000) consistently caused a modulation of the NWR, which probably served to protect the painful tissue, a finding in line with those obtained from animal research (Clarke and Harris, 2004). However, effects of phasic muscle pain may depend on the exact timing of the painful intramuscular electrical stimulation (Andersen et al., 2006; Ge et al., 2007). Suppression of the NWR was observed during painful heterotopic stimulation, which was attributed to a ‘diffuse noxious inhibitory control’ mechanism (Willer et al., 1984, 1989; Roby-Brami et al., 1987; Terkelsen et al., 2001; Serrao et al., 2004).

At the spinal level, motor output is modulated by various excitatory and inhibitory circuits (Pierrot-Deseilligny and Burke, 2005). The most well-known excitatory spinal circuit consists of Ia afferents originating from muscle spindles and projecting to α-motoneurons of the homonymous muscle and synergists, which leads to a reflex contraction of a previously stretched muscle (i.e., the phasic stretch reflex). The H-reflex is evoked by direct stimulation of Ia afferents, thereby bypassing the muscle spindles and fusimotor activity that are involved in the stretch reflex, and is generally assumed to reflect excitability of the motor neuron pool (Knikou, 2008). Inhibitory spinal circuits mediated by Ib interneurons play a role in coordinating the activity of muscles operating at several joints (Jankowska, 1992; Rossi and Decchi, 1997) and protective negative feedback circuits mediated by Renshaw cells inhibit contracting muscles (recurrent inhibition; for a review see Katz and Pierrot-Deseilligny, 1999). We found insufficient evidence for systematic modulation of the phasic stretch reflex, the H-reflex or inhibitory circuits by pain stimuli applied to the skin or muscle (see Supporting Information Results S1). The failure to detect systematic modulation of those reflexes does not necessarily imply that pain has no influence on the associated spinal circuits. Findings regarding these spinal reflexes – which are typically prone to methodological issues – were often highly variable and obtained from small samples. Nevertheless, there were some indications that the effects of pain may to some extent be mediated by inhibitory spinal circuits, e.g., through reinforcement of recurrent inhibition during contraction of a painful muscle (Rossi et al., 2003a) or through modulation of the Ib inhibitory pathway (Rossi and Decchi, 1995, 1997; Rossi et al., 1999a,b). As it stands, it is not clear if this modulation depends on the origin of pain.

3.2 Muscle activity

The effects of experimentally induced pain on muscle activity have been examined at rest (Supporting Information Table S2: n = 10), during isometric and dynamic contractions (Supporting Information Table S3: n = 23 and Supporting Information Table S4: n = 14), and during low-load repetitive work (Supporting Information Table S5: n = 5). For isometric and dynamic contractions, the effects of pain on the activity of the (painful) agonist muscle, the non-painful synergists and the non-painful antagonists are presented in separate columns.

Experimentally induced pain generally did not affect resting muscle activity (Cobb et al., 1975; Graven-Nielsen et al., 1997a,b; Svensson et al., 1998; Madeleine and Arendt-Nielsen, 2005; Serrao et al., 2007; Birznieks et al., 2008; Ge et al., 2008; Fernández-Carnero et al., 2010; Xu et al., 2010), but facilitated muscle cramps when latent myofascial trigger points were stimulated (Serrao et al., 2007; Ge et al., 2008; Xu et al., 2010). During both isometric and dynamic contractions, activity of the (painful) agonist (Graven-Nielsen et al., 1997a; Madeleine et al., 1999a,b, 2006; Birch et al., 2000b; Ciubotariu et al., 2004, 2007; Ervilha et al., 2004a,b, 2005; Ge et al., 2005; Qerama et al., 2005; Farina et al., 2005a; Del Santo et al., 2007; Falla et al., 2007, 2008, 2009, 2010; Henriksen et al., 2007, 2009a,b, 2011; Martin et al., 2008) and its non-painful antagonist (Madeleine et al., 1999b; Ervilha et al., 2004a,b; Henriksen et al., 2009b) generally was reduced by pain arising from the skin, joint, muscle or tendon. However, parameters derived from the surface electromyography (sEMG) signal of these muscles (i.e., amplitude and median power frequency) often remained unaffected during contractions at a relatively low intensity, i.e., <25% of maximum voluntary contraction (MVC) for muscles located in the upper or lower limb (Birch et al., 2000a; Farina et al., 2004, 2005b, 2008; Madeleine and Arendt-Nielsen, 2005; Hodges et al., 2008; Tucker and Hodges, 2009; Hirata et al., 2010) and <15% MVC for muscles located in the shoulder-neck region (Diederichsen et al., 2009; Samani et al., 2010). During dynamic tasks, several findings indicated that inhibition of a muscle was most pronounced when activity was highest (Graven-Nielsen et al., 1997a; Ervilha et al., 2004a,b; Henriksen et al., 2007, 2009a; Hodges et al., 2009). These findings suggest that the likelihood of a muscle being inhibited by pain may depend on its activation level: the stronger the activity, the more likely that its activity will be reduced. However, activity of the painful muscle remained unaffected despite a relatively high intensity of isometric contraction in four out of eight studies (Graven-Nielsen et al., 1997a; Schulte et al., 2004; Ge et al., 2005; Madeleine and Arendt-Nielsen, 2005; Bandholm et al., 2008).

Notably, findings at low-intensity muscle contractions may depend on the applied EMG technique since sEMG recordings failed to detect any pain-induced alterations (Birch et al., 2000a; Schulte et al., 2004; Farina et al., 2004, 2005b; Hodges et al., 2008; Tucker et al., 2009), whereas intramuscular EMG recordings showed adaptations in motor unit firing and recruitment (Birch et al., 2000a; Farina et al., 2004, 2005b, 2008; Hodges et al., 2008; Tucker et al., 2009; Tucker and Hodges, 2009, 2010). Several findings, including unaltered sEMG parameters following electrical stimulation of motor axons (Qerama et al., 2005; Farina et al., 2005a) and unaffected muscle fiber conduction velocity in three out of four studies (Farina et al., 2004, 2005a,b, 2008 vs. Schulte et al., 2004), indicated that alterations in muscle activation were due to central rather than peripheral effects of pain. Furthermore, several studies revealed signs of redistribution of activity, both within the (painful) agonist (Falla et al., 2008, 2009, 2010; Tucker et al., 2009; Tucker and Hodges, 2009, 2010) and among muscles involved in the task (Graven-Nielsen et al., 1997a; Madeleine et al., 1999a,b, 2008; Ciubotariu et al., 2004; Schulte et al., 2004; Ervilha et al., 2004b, 2005; Falla et al., 2007; Bandholm et al., 2008; Diederichsen et al., 2009; Hodges et al., 2009; Samani et al., 2009, 2010). Reduced activation of painful muscles in some cases was compensated by activation of non-painful synergists (Madeleine et al., 1999b; Ciubotariu et al., 2004; Schulte et al., 2004; Ervilha et al., 2004b, 2005; Bandholm et al., 2008; Diederichsen et al., 2009). However, non-painful synergists often remained unaffected (Birch et al., 2000b; Schulte et al., 2004; Hodges et al., 2008; Henriksen et al., 2009a) or were inhibited just like the (painful) agonist (Ciubotariu et al., 2004, 2007; Ervilha et al., 2004a,b, 2005; Henriksen et al., 2007, 2009b, 2011).

3.3 Task performance, movement kinetics and kinematics

The effects of experimentally induced pain on characteristics of motor control have been examined in terms of kinetics, kinematics and other indices of task performance during isometric contractions (Supporting Information Table S3: n = 22), dynamic contractions (Supporting Information Table S6: n = 23) and low-load repetitive work (Supporting Information Table S5: n = 7).

The performance of various motor tasks mostly remained unimpaired by pain arising from skin, joint, muscle or tendon. Subjects were able to produce a given submaximal force level (Schulte et al., 2004; Farina et al., 2004, 2005a,b, 2008; Madeleine and Arendt-Nielsen, 2005; Del Santo et al., 2007; Bandholm et al., 2008; Hodges et al., 2008; Martin et al., 2008; Tucker et al., 2009; Tucker and Hodges, 2009, 2010), which was often associated with reduced sEMG activity level of the (painful) agonist muscle (section 3.2). Maximum voluntary force production was reduced in four out of five studies (Graven-Nielsen et al., 1997a, 2002; Slater et al., 2003; Henriksen et al., 2010b; vs. Slater et al., 2005). In dynamic motor tasks, effects of pain were mostly reflected in the movement kinetics, e.g., in a reduction of peak moments around the joint upon which a painful muscle acted (Bonifazi et al., 2004; Henriksen et al., 2007, 2009a,b, 2010a,b), although impact force just after heel strike was unaffected by pain in a knee extensor muscle (Henriksen et al., 2008). Despite observed alterations in kinetics, changes in movement kinematics were absent (Henriksen et al., 2007, 2008, 2009b; Maihöfner et al., 2007; Diederichsen et al., 2009) or rather subtle (Madeleine et al., 1998, 1999a,b, 2008; Jaberzadeh et al., 2003; Bonifazi et al., 2004; Ervilha et al., 2004a,b, 2005; Henriksen et al., 2009a, 2011). The performance of computer work (Birch et al., 2000b, 2001; Samani et al., 2009, 2010) or manual dexterity tasks (Smith et al., 2006) was not deteriorated by experimentally induced muscle pain. During quiet standing, pain applied to the lower leg muscles led to weight shifting to the non-painful leg (Hirata et al., 2010). Postural stability was slightly reduced by severe pain induced in the bilateral upper trapezius muscles (Vuillerme and Pinsault, 2009) and by pain applied to cutaneous tissue or muscular tissue of the lower leg (Madeleine et al., 1998, 1999b; Blouin et al., 2003; Corbeil et al., 2004; Hirata et al., 2010), but it remained unaffected by pain arising from other sources (Madeleine et al., 1999a, 2004; Corbeil et al., 2004; Bennell and Hinman, 2005).

3.4 Proprioception

The sense of positions and movements of one's body parts, as well as the perception of forces produced by muscles, are basic requirements for adequate motor control. Proprioception not only provides information about the internal state of a limb to facilitate movement planning; it also allows for more flexible movement control (Gentilucci et al., 1994; Park et al., 1999) and assists the voluntary control of goal-directed movements, e.g., by triggering muscle activation sequences (Park et al., 1999) or by timing and coordinating movement sequences (Cordo et al., 1994).

Indications were found of a slight deterioration of proprioception (Supporting Information Table S7: n = 8). Pain stimuli applied to skin or muscle caused a deterioration of movement sense – albeit not in all conditions (Matre et al., 2002; Weerakkody et al., 2008) – and the perception of produced force (Weerakkody et al., 2003). Although it has been reported that pain caused a distortion or loss of position sense (Rossi et al., 1998, 2003b), no quantitative evidence has been presented for impaired joint position sense (Matre et al., 2002; Bennell et al., 2005) or alterations in firing of muscle spindle afferents (Birznieks et al., 2008). Indications were found that muscle pain interfered with processing of non-nociceptive afferent signals from the muscle (Rossi et al., 1998, 2003b; Niddam and Hsieh, 2008). Although most studies did not include a direct comparison between the effects of cutaneous pain, joint pain and muscle pain, it appears that pain arising from muscle tissue may have more pronounced effects on proprioception than pain arising from (sub)cutaneous tissue (Weerakkody et al., 2003).

3.5 Brain activation

Since the emergence of functional brain imaging techniques, considerable efforts have been made to identify the cortical and subcortical structures that are activated by pain. Research has concentrated on examining brain activation in response to acute painful stimulation in healthy subjects lying quietly in a scanner. It is well known that pain activates cortical areas involved with perception of intensity and location of the painful stimulus, the regulation of emotional responses accompanying pain, and the distribution of attention (for a review, see Peyron et al., 2000). Although activation of motor-related areas (i.e., primary motor cortex, supplementary motor cortex, premotor cortex, cerebellum and/or basal ganglia) has occasionally been reported, this topic has mainly been regarded a side issue. Unfortunately, research has not yet focused on the effects of pain on brain activation patterns during movement planning or execution.

Several studies have addressed the interference between pain and cortical correlates of motor function by examining motor evoked potentials (MEPs) evoked by transcranial magnetic stimulation (TMS) or transcranial electric current stimulation (TECS) over the primary motor cortex (Supporting Information Table S8: n = 18). In general, MEP amplitude was reduced as a consequence of pain signals originating from skin (Uncini et al., 1991; Kaneko et al., 1998; Kofler et al., 1998, 2001; Valeriani et al., 1999; Farina et al., 2001; Tamburin et al., 2001; Urban et al., 2004; Fierro et al., 2010) or muscle (Le Pera et al., 2001; Svensson et al., 2003; Martin et al., 2008), but in some studies, it was found to be unaffected (Le Pera et al., 2001; Cheong et al., 2003; Fadiga et al., 2004; Martin et al., 2008) or even elevated (Cheong et al., 2003; Fadiga et al., 2004; Del Santo et al., 2007). In the case of distally localized pain, inhibition of MEPs evoked in the biceps brachii may be followed by excitation, probably reflecting preparations for hand withdrawal (Kofler et al., 1998, 2001; Urban et al., 2004). Painful stimulation at the hand sometimes caused modulation of corticospinal excitability of a muscle in the contralateral hand (Valeriani et al., 1999; Kofler et al., 2001) or arm (Hoeger Bement et al., 2009).

MEPs provide a measure of corticospinal excitability, which encompasses both cortical and spinal processes. In order to disentangle the effects of pain on cortical processes, assessment of MEPs should therefore be complemented by assessment of cervicomedullary MEPs or H-reflexes. Unfortunately, such measurements were present in only 5 out of 18 studies, and the results were inconclusive. The reduction of MEP amplitude induced by pain applied at the skin appears not attributable to decreased excitability of spinal motor neurons (Farina et al., 2001; Urban et al., 2004), whereas the reduction of MEP amplitude as a consequence of muscle pain may (partly) reflect decreased excitability of spinal rather than cortical motor neurons (Le Pera et al., 2001; Svensson et al., 2003). In contrast, the findings of Martin et al. (2008) suggest that muscle pain may lead to decreased cortical excitability accompanied by opposing alterations at the spinal level.

Furthermore, studies analysing electroencephalogram (EEG) signals in terms of oscillation frequencies (Babiloni et al., 2008) or the exact timing of evoked potentials (Tarkka et al., 1992) provided indications of interference between pain and sensorimotor processes related to the planning and execution of movement (Supporting Information Table S8: n = 2).

4. Discussion

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

This systematic review was conducted to obtain a better understanding of how pain affects motor behaviour. Since it is unclear if pain from different tissues differentially affects the motor system, we evaluated the various sources of pain separately.

Notably, some motor components (spinal reflexes) were mainly examined by means of experimentally induced cutaneous pain, while others (muscle activity and movement characteristics) were predominantly examined by means of experimentally induced muscle pain (Fig. 1). Although these differences hamper comparisons across studies, the observed effects on various components of the motor system were largely similar irrespective of the source of pain. Studies on spinal reflexes indicated differential influences of cutaneous pain and muscle pain, but due to the limited amount of available data and the heterogeneity of results, it is not possible to draw firm conclusions in this regard.

Although some findings of this systematic review are congruent with existing models, others are not. In line with earlier reports (Knutson, 2000; Arendt-Nielsen and Graven-Nielsen, 2008; Hodges and Tucker, 2011), regardless of the pain source, we found no evidence of muscle hyperactivity as predicted by the ‘vicious cycle’ model (Travell et al., 1942; Johansson and Sojka, 1991). This model is based on the assumption that pain leads to muscle spasms, whereas evidence pointed at inhibition of painful muscles. The observed inhibition of the (painful) agonist muscle is consistent with the ‘pain-adaptation’ model (Lund et al., 1991). However, this model predicts excitation of antagonist muscles, for which no evidence was found in this review. On the contrary, a muscle's susceptibility to inhibition by pain seemed to depend on its activity state, rather than its function within a particular movement. Also the ‘neuromuscular adaptation’ model, which predicts alterations in synergies, does not provide a conclusive explanation for the interaction between pain and motor control (Sterling et al., 2001). In particular, this model cannot account for the finding that the synergist's behaviour often paralleled that of the (painful) agonist muscle. However, the theory on motor adaptation to pain recently proposed by Hodges and Tucker (2011) was largely in accord with our findings showing that pain affected many components of motor behaviour mediated at multiple levels of the motor system. Under circumstances of acute pain, the observed redistribution of activity within and among muscles as well as the (subtle) changes in mechanical behaviour indeed seem to reflect adaptations leading to protection from further pain or injury. Hodges and Tucker (2011) intended to offer an explanation for the substantial variance observed between individuals and tasks, which resulted in a widely applicable theory. Given the relatively homogenous picture that emerges from the present review, it might be suggested that effects of short-lasting, experimentally induced limb pain can be described with higher specificity (Fig. 2).

figure

Figure 2. Motor consequences of experimental limb pain induced in skin, muscle or tendon. The main findings regarding the effects of nociceptive afferent signals (red) on motor control, which involves interaction between non-nociceptive afferent signals (blue) and motor efferent signals (green), are presented in text boxes. Dotted arrows between the lower three text boxes suggest a causal relationship between the effects of pain on muscle activity, kinetics and kinematics, but the relation between these parameters has not been directly addressed as such.

Download figure to PowerPoint

The majority of studies focused on the influence of pain on muscle activation during various types of contraction (section 3.2). Pain generally caused a reduction of activity of the (painful) agonist as well as non-painful synergists and antagonists, especially at higher intensity of contraction. Despite the influence of pain on muscle activation, the performance of various motor tasks mostly remained unimpaired (section 3.3), presumably as a result of redistribution of activity, both within the (painful) agonist and among muscles involved in the task. The finding that a given force level was associated with less sEMG activity in the painful muscle also pointed at compensation by other (not recorded) motor units or muscles. Effects of pain were mainly reflected in movement kinetics as a reduction of maximum force or peak moment, resulting in subtle alterations in movement kinematics. Although there were indications of a slight deterioration of proprioception (section 3.4), no evidence was found of detrimental effects on motor control. Given the observed activation-dependent inhibition of muscle activity, this sensory impairment is unlikely to play a major role in mediating the effects of pain on motor control because, if so, more complex alterations in timing and coordination of movement sequences would have been expected (Cordo et al., 1994).

Because muscle activation and movement characteristics are the result of many processes mediated at various levels of the motor system, their responses to pain do not allow identification of the exact mechanisms that underpin the interaction between pain and the motor system. Given that small-diameter afferents have projections both at the spinal and the supraspinal level (Millan, 1999; Almeida et al., 2004), it is not surprising that motor control was found to be affected by pain at its most basic level, i.e., spinal reflexes, as well as its highest level, i.e., cortical processes related to movement planning and execution. Findings regarding spinal reflexes were often highly variable and obtained from small samples. This may partly explain why we found limited evidence for systematic modulation of the H-reflex, the stretch reflex or inhibitory circuits (section 3.1). In contrast, pain caused a consistent modulation of the NWR, as has also been observed in animal research (Clarke and Harris, 2004).

At the highest level of motor control, EEG studies provided indications of pain interfering with cortical sensorimotor processes related to movement planning and execution. Studies using TMS or TECS over the primary motor cortex showed that pain arising from skin or muscle leads to reduced excitability of the corticospinal motor pathway (section 3.5). The question remains, however, if and to what extent these changes are attributable to altered spinal excitability. Attempts to disentangle the cortical and spinal contributions to alterations in corticospinal excitability have been made in a limited number of studies, which provided inconclusive results. Unfortunately, studies using functional brain imaging techniques have focused on brain activation at rest (Peyron et al., 2000), leaving the issue of how pain affects brain activation during movement planning or execution unaddressed.

The research on experimentally induced pain covered in this review delineates the effects of acute and transient pain stimuli on motor control and allows to disentangle the intricate cause-and-effect relation between pain and movement. However, the findings cannot be translated to clinical pain conditions (Edens and Gil, 1995), which likely are associated with long-term adaptations to pain, a key aspect of the theory proposed by Hodges and Tucker (2011). Additionally, clinical pain conditions are commonly associated with structural damage that may induce additional effects on movement. Moreover, emotional and cognitive responses to (chronic) pain may greatly affect motor control, e.g., movement strategies may be altered by fear of pain (Vlaeyen and Linton, 2000). Such responses, if present, are probably different in experimental conditions as participants are aware that the pain will quickly resolve. Also, in some studies, participants were made familiar with the nociceptive stimulus prior to the experimental session in order to minimize a potential emotional component of the pain.

Many other factors affect the impact of pain on motor function as well, e.g., the severity, duration and location of pain, additional activation of non-nociceptive afferents, and the state of the motor system (i.e., at rest or during planning or execution of movement). Unravelling the impact of pain on motor function thus requires diligent experimental control of many factors, which represents a major methodological challenge. Moreover, similar to gender differences in perception and tolerance of pain (Fillingim et al., 2009; Racine et al., 2011), the results of several studies suggest that gender-specific factors may influence the motor responses to pain (Ge et al., 2005; Madeleine et al., 2006; Falla et al., 2008, 2010). However, only two studies explicitly assessed potential gender-specific differences in motor consequences of pain. Surprisingly, the potential influence of this factor on the results was not addressed in the majority of studies (e.g., gender distribution was not reported in 23% of studies).

As regards the overall picture emerging from the studies included in this review, several limitations have to be acknowledged. Firstly, sample size was typically small (ranging from 1 to 36 subjects, with only 48% of the studies including more than 10 subjects). Secondly, 63% of the studies lacked an appropriate condition to control for possible non-nociceptive effects of pain stimuli. Thirdly, research has concentrated on relatively easily accessible aspects of motor function (i.e., muscle activity and movement characteristics; Fig. 1), culminating in a limited number of coherent parameters. Several findings indicated that the observed changes result from central rather than peripheral effects of pain. Due to methodological issues and heterogeneity regarding outcome parameters, however, findings remained largely inconclusive for parameters that may provide insight into processes mediating the effects of pain at different levels of the central nervous system (i.e., spinal reflexes and cortical correlates of motor function). This motivates future examination of the impact of pain on spinal reflexes and cortical correlates of motor function, taking special care of methodological considerations. In this context, it should be noted that research on the effects of pain on spinal reflexes and brain activation has mainly focused on the motor system at rest. For a full appreciation of the interaction between pain and motor control, it is essential to examine the effects of pain on spinal reflexes and brain activation not only at rest, but also during movement planning and execution.

Author contributions

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

Conception and design: P.J.M. Bank, C.E. Peper, J. Marinus, P.J. Beek, J.J. van Hilten

Collection and assembly of data: P.J.M. Bank

Data analysis and interpretation: P.J.M. Bank, C.E. Peper, J. Marinus

Manuscript writing: P.J.M. Bank, C.E. Peper, J. Marinus, P.J. Beek, J.J. van Hilten

Final approval of manuscript: P.J.M. Bank, C.E. Peper, J. Marinus, P.J. Beek, J.J. van Hilten

References

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Databases
  4. What does this study add?
  5. 1. Introduction
  6. 2. Literature search methods
  7. 3. Results
  8. 4. Discussion
  9. Acknowledgement
  10. Author contributions
  11. References
  12. Supporting Information
FilenameFormatSizeDescription
ejp186-sup-0001-metS1.doc39K Methods S1. Search strategy.
ejp186-sup-0002-metS2.doc35K Methods S2. Flow diagram of included studies.
ejp186-sup-0003-resS1.doc116K Result S1. Detailed description of results.
ejp186-sup-0004-tblS1.doc183K Table S1. Effects of experimentally induced pain on spinal reflexes.
ejp186-sup-0005-tblS2.doc55K Table S2. Effects of experimentally induced pain on muscle activity at rest.
ejp186-sup-0006-tblS3.doc208K Table S3. Effects of experimentally induced pain on muscle activity and performance during isometric contractions.
ejp186-sup-0007-tblS4.doc107K Table S4. Effects of experimentally induced pain on muscle activity during dynamic contractions.
ejp186-sup-0008-tblS5.doc58K Table S5. Effects of experimentally induced pain on muscle activity and movement characteristics during low load repetitive work.
ejp186-sup-0009-tblS6.doc110K Table S6. Effects of experimentally induced pain on movement characteristics during dynamic contractions.
ejp186-sup-0010-tblS7.doc68K Table S7. Effects of experimentally induced pain on measures of proprioception.
ejp186-sup-0011-tblS8.doc127K Table S8. Effects of experimentally induced pain on cortical correlates of motor function.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.