Spinal cord injury as a trigger to develop
periodic leg movements during sleep:
an evolutionary perspective
Lesão medular como gatilho para desenvolver movimentos periódicos das pernas durante
o sono: uma perspectiva evolutiva
Susana Cristina Lerosa Telles, Rosana S. Cardoso Alves, Gerson Chadi
The periodic limb movement (PLM) disorder is character-
ized by periodic episodes of repetitive and highly stereotyped
limb movements, which occur during sleep, and by clinical
sleep disturbance that cannot be accounted for by another
primary sleep disorder1. They consist of the big toe extension
in combination with partial flexion of the ankle and, some-
times, hip (Fig 1). Similar movements can occur in the upper
limbs. This occurrence in spinal cord injury (SCI) patients2
(Fig 2) suggests the spinal origin of those movements, which
could be due to the disruption of REM-related inhibitory spi-
nal pathways, producing the disconnection or disinhibition of
a spinal generator, which is a kind of local pacemaker3.
Restless legs syndrome (RLS) is a sensorimotor disorder
frequently related to PLM and was recently reported in SCI pa-
tients2. It is characterized by a complaint of a strong, nearly irre-
sistible, urge to move the legs, which is made worse by rest (ly-
ing or sitting) and is at least partially and temporarily relieved by
walking or moving the legs. The relief is usually immediate. The
urge to move the legs worsens in the evening and at night1.
Scientific reviews have addressed the probable central pat-
tern generator (CPG) in SCI patients and its outcome to reha-
bilitation. We based this review on an evolutionary perspective
of CPG development from invertebrates to mammals, in order
to point to the probable CPG action in humans during PLM.
Neuroregeneration Center, Experimental Neurology, Department of Neurology, School of Medicine, Universidade de São Paulo, São Paulo SP, Brazil.
Correspondence: Susana Cristina Lerosa Telles; Avenida Doutor Arnaldo 455 / 2.115; 01246-903 São Paulo SP - Brasil; E-mail: email@example.com
Support: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).
Conflict of interest: There is no conflict of interest to declare.
Received 25 January 2012; Received in final form 05 April 2012; Accepted 13 April 2012
The primary trigger to periodic limb movement (PLM) during sleep is still unknown. Its association with the restless legs syndrome (RLS) is
established in humans and was reported in spinal cord injury (SCI) patients classified by the American Spinal Injury Association (ASIA) as A.
Its pathogenesis has not been completely unraveled, though recent advances might enhance our knowledge about those malfunctions. PLM
association with central pattern generator (CPG) is one of the possible pathologic mechanisms involved. This article reviewed the advances
in PLM and RLS genetics, the evolution of CPG functioning, and the neurotransmitters involved in CPG, PLM and RLS. We have proposed that
SCI might be a trigger to develop PLM.
Key words: nocturnal myoclonus syndrome, restless legs syndrome, spinal cord regeneration.
O gatilho principal para o desenvolvimento de movimentos periódicos dos membros (MPM) durante o sono ainda é desconhecido. A associa-
ção entre o MPM e a síndrome das pernas inquietas (SPI) em seres humanos já foi previamente estabelecida e relatada em pacientes com
lesão medular (LM), classificados pela American Spinal Injury Association (ASIA) como A. A patogênese do MPM não foi completamente des-
vendada, apesar de avanços recentes poderem ampliar o conhecimento sobre essas disfunções. Um dos possíveis mecanismos patológicos
envolvidos é o gerador de padrão central (GPC). Este artigo revisou os avanços na genética do MPM e da SPI, a evolução do funcionamento
do GPC e os neurotransmissores relacionados ao GPC, ao MPM e à SPI. Foi proposta a hipótese de que a LM poderia ser um gatilho para
deflagrar os MPM.
Palavras-Chave: síndrome da mioclonia noturna, síndrome das pernas inquietas, regeneração da medula espinal.
Susana Cristina Lerosa Telles et al. Periodic leg movements
CENTRAL PATTERN GENERATOR
Graham Brown proposed the following: “when the geo-
graphical relations between animal and environment re-
mained adverse, there would occur a local asphyxia of the
efferent neuron — the primitive centre — and a motor activa-
tion of the limbs, even if uncoordinated, would serve to drive
the animal to a new region in which the deficiency of nutritive
materials would be compensated”. This statement can be in-
terpreted as a defense mechanism in case of any movement
malfunction causing spasticity, which would cause a poor mo-
tor coordination, but would elicit movements so that the an-
imal could search for food4. As evolution proceeded and the
primitive nervous system became enclosed with the CPG in-
side, the environment local asphyxia would no longer directly
affect the nerves centers and its place would be taken by the
blood stimulus (hunger) that acts even in higher forms4.
Fig 2. Periodic leg movements cluster and arousals. Polysomnographic recording of a 21 years-old spinal cord injury male patient
with periodic leg movements, severe restless legs syndrome, and diurnal somnolence. In this case, the movements took place
during stage 2 (S2) of nonrapid eye movement sleep and were associated with arousals. (I) Electroencephalogram in central
derivation C3-A2; (II) electroencephalogram in occipital derivation O2-A1; (III) electromyogram showing leg movements, which are
characteristic of PLM; (IV) airway derivation showing normal breathing during the events.
Fig 1. Periodic leg movements cluster. Five-minute polysomnographic recording of a 28 years-old female volunteer with restless
leg syndrome, without diurnal somnolence or any other health problems in order to demonstrate periodic leg movements cluster.
The movements took place during stages 2, 3, and 4 of nonrapid eye movement sleep. (I) Electroencephalogram in central
derivation showing stage 2 (S2), 3 (S3), 4 (S4), and wake; (II) electroencephalogram in occipital derivation; (III) tibialis anterior
electromyogram showing leg movements, which are characteristic of PLM that last four minutes and 24 seconds; (IV) airway
derivation showing regular airway breathing during the periodic leg movement event.
Arq Neuropsiquiatr 2012;70(11):880-884
The highest forms have similar CPG, but it is difficult to
explain how more complex patterns made by many differ-
ent joints with multiple movements would be controlled and
executed by an isolated system. Most of all, it would hardly
come from Brown’s CPG model alone. Many researches point
out more complex systems in invertebrates and vertebrates.
One of the first CPG models can be found among the in-
vertebrates. The stomach of the crab is composed of striat-
ed muscles that are controlled by a CPG, located inside the
stomatogastric ganglion. It is controlled by extrinsic neuro-
modulatory substances, which can change or activate specif-
ic motor patterns. The most important neurotransmitters in
this case are serotonin for neuromodulation and acetylcho-
line for a faster neurotransmission. On the other hand, the
Tritonia diomedia mollusk uses only serotonin to activate its
escape swim CPG, when attacked by a predatory starfish5.
Also in the vertebrates, the neuromodulators are respon-
sible for behavioral flexibility in the CPGs. There are two ma-
jor mechanisms by which those neuromodulatory influences
alter the function of the CPG: the combination of intrinsic
membrane properties, such as the ability to generate plateau
potentials or post-inhibitory rebound, or interactions be-
tween members of the pattern generator circuits5.
In vertebrates, the CPG is activated by interactions be-
tween glycine and glutamate. The external modulations are
made by monoamines, glutamines, and peptides. Also, mul-
tiple ion channels and signal transduction mechanisms in
pre- and post-synaptic cellular compartments play a role
in rhythm generation. They are controlled by co-activation
of α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid
(AMPA) and N-methyl d-aspartate (NMDA) receptors by glu-
tamate release from descending fibers, such as the reticulo-
spinal system6. For instance, in tadpoles, swimming requires
continuous excitatory output that comes from hindbrain
neurons, which excite one another through the co-release of
glutamate and acetylcholine, causing persistent activity in
the entire group of neurons7.
The inhibition that occurs in Brown’s model is elicit-
ed by glycine in all preparations of vertebrates. Gamma-
aminobutyric acid (GABA) inhibits the CPG, and nitric acid
acts as modulator of the swimming pattern and of the re-
lease of other modulators, such as noradrenalin. This neuro-
modulator is released by the ventral horns and it is attached
to glycinergic receptors7. GABA receptors play a similar
role in some cases6. The mechanism for hyperpolarization
is also shared among mammalian species, turtles, and lam-
preys. During the plateau depolarization, Ca2+ enters the
cytosol. During activity-dependent-release of glutamate,
NMDA and metabotropic glutamate receptors are activated.
NMDA receptor activation coupled with depolarization al-
lows a direct entry of Ca2+ at the synapse8,9.
The complexity increases in the control of mamma-
lian respiratory system. The site of pattern generator is
called Bötzinger complex. In such place, the CPGs are con-
trolled by two kinds of pacemaker neurons, with bursting sus-
tained by different electrical currents: sodium (cadmium in-
sensitive pacemaker) and a nonspecific Ca2+ dependent cation
current (cadmium sensitive pacemakers), in addition to ex-
tensive synaptic interactions. This system is subjected to neu-
romodulation by several different amines and peptides, such
as serotonin, norepinephrine, substance P, and acetylcholine
functioning through muscarinic acetylcholine receptors9.
Mammalian spinal cord functioning is still not under-
stood. By means of genetic techniques, many neurons have
been identified as being part of an alleged CPG in mice. Five
classes of progenitor cells give rise to motoneurons, and four
others of ventrally located interneurons form reflex or loco-
motor circuits in the adult mouse10.
Vilensky and O’Connor11 reviewed the literature on step-
ping in nonhuman and human primates and discussed their
experiment of trying to elicit steps in a squirrel monkey with
complete transected spinal cord (level T8), based on tread-
mill training. They pointed out the high probability that loco-
motion in monkeys is based on spinal cord CPG, but differ-
ent species might require other types of stimulation to turn
the CPG on. It is a matter of understanding that the primate
must have full body support, differently from spinalized cats
or dogs, which are able to support their body after a spinal
cord lesion. It can be assumed that the human CPG would be
even harder to stimulate.
There are important neuroanatomical changes in primate
evolution. Heffner and Masterton12 found a correlation of two
factors and increased digital dexterity: a reduction of the number
of synapses between the neocortex and the spinal motoneurons
innervating the digits, and an extension of direct neocortical
control beyond the cervical enlargement. Those neuroanatomi-
cal characteristics can be translated into the theory of increasing
encephalization across the evolution. The increase in finger-mo-
tor coordination has led to higher survival chance for primates.
The consequence was the spinal cord submission to neocortex
in a larger scale than what is seen in other species.
PERIODIC LIMB MOVEMENT AND RESTLESS
LEGS SYNDROME GENETICS
There is strong evidence about a presumed hereditary
component of PLM. Five loci (12, 14, 9, 2, and 20 p) have already
been described13. A recent article showed the association be-
tween a sequence variant in chromosome 6 p with PLM in dis-
tinct Icelandic and American cohorts of subjects with RLS and
their families14. A simultaneous report by Winkelman15 showed
an association between the same sequence variant and two
additional single nucleotide polymorphisms in German and
Canadian cohorts with RLS. Thus, PLM could serve as a heri-
table biologic marker or endophenotype for the RLS.
Susana Cristina Lerosa Telles et al. Periodic leg movements
PERIODIC LIMB MOVEMENT IN SPINAL CORD
Since the first neural CPG theory proposed by Graham
Brown through his experiments on cats4, the most compel-
ling evidence of human CPG is the occurrence of PLM in
SCI patients. Telles et al.2 reported the presence of RLS and
PLM in the American Spinal Injury Association (ASIA) A
patients. In this study, 24 patients were submitted to a full
night polysomnography and were assessed with Epworth
Sleepiness Scale and an adapted form of International RLS
Scale Rating Scale (IRLS Rating Scale). The Control Group
(CG) was composed of 16 subjects, 50% of each sex, and
ages: 24.38±4 years-old. The SCI Group (SCIG) was com-
posed of eight subjects (29±5 years-old) with a complete
SCI (ASIA A) of about three and a half years of duration,
100% males. The result was 100% of the SCIG had RLS com-
pared to 17% in the CG (p<0.0001). SCIG had 18.11±20.07
of PLM index, while the CG had 5.96±11.93 (p=0.01).
There are many other theories regarding PLM in non-
injured patients. Summarizing, it is not known whether
PLM would be a consequence of the absence of superior
efferences to the spinal cord, which would cause an in-
crease in the excitability of the motoneurons, or if it is a
primary phenomena coming from a generator located in-
side the human spinal cord. If its origin is in the superior
central nervous system, it is not known if the abnormal
movements would be correlated with the sleep stage3, rel-
ative concentration of iron in the red nucleus16, and lack of
iron in the blood caused by anemia or central neurotrans-
mitters disorders17. If its origin is intrinsic to the spinal
cord, the action of the neurotransmitters in the spinal
cord might be one of the causes of PLM.
NEUROTRANSMITTERS INVOLVED IN
PERIODIC LIMB MOVEMENT AND IN THE
RESTLESS LEGS SYNDROME
Considering the previous explanations of spinal cord
physiology in the vertebrates, animal models of PLM pro-
vide a promising way to study neuronal connections that lie
within the spinal cord. Ondo et al.18 reviewed animal models
for RLS and PLM caused by iron deprivation and destruc-
tion of the A11 diencephalic spinal tract. They specifical-
ly discussed RLS 6-hydroxydopamine rat model with and
without iron deficiency and dopamine D3 receptor knock-
out mice. For PLM, they discussed the effect of haloperidol,
a dopamine antagonist, in rats. They concluded that even
though the RLS animal models might help in the evaluation
of the involvement of dopaminergic action or iron deple-
tion in RLS, this is not a perfect model, since it comes with-
out the clinical sensory components that characterize the
syndrome in humans18. Since PLMs are seen as behavioral
markers for RLS in animal models, those models might be
closer to human PLM itself than to human RLS.
Esteves et al.19 described a mouse animal model in ten
rats subjected to T9 lesions. There were four different types
of lesion: Group 0, with no histopathological alterations
(SHAM Group); Group I, predominantly affecting central
part of dorsal column; Group II, predominantly the entire
dorsal column; Group III, affecting mainly the half spinal
cord, including the whole dorsal column; and Group IV, al-
most all the spinal cord. Their results showed that 10 out of
11 rats developed limb movements during sleep, while the
SHAM Group presented no limb movements during the ex-
periment. Their conclusions were that these movements are
probably generated by a spinal CPG, without the involve-
ment of cortical inputs. This model might help to found out
spinal mechanisms responsible for PLM in animals.
In humans, there are major controversies with regard
to the localization of the neural structures involved in the
pathologic process of RLS, however there is most likely a
major contribution of the spinal cord. Bara-Jimenez et al.20
studied the flexion reflex of patients with primary RLS by
electrically stimulating the plantar nerve. They found a
facilitation of the late component of flexor reflex afferent
(FRA), indicating hyperexcitability of motoneurons in this
condition. Included in the FRA are high-threshold muscle,
joint, and cutaneous afferents. They also noted that the late
components shared several features with PLM. Those stud-
ies suggest the presence of a CPG for the periodic motor
manifestations of RLS. According to the authors, this CPG
might be facilitated by the suppression of or decrease in
supraspinal inhibitory inputs. The FRA would be the tool
through which the spinal CPG would act and would be re-
sponsible for the relief caused by movement in RLS21.
A question can be raised: is PLM an evolutional trace
of ancient CPGs instead of pathology in itself? An answer
should take into consideration the high prevalence of PLM
in SCI. In this condition, the encephalization that distin-
guishes the human species is compromised, and the spi-
nal cord and all its structures are no longer submitted to
the neocortex. This leads to many neurological features
that characterize SCI injuries, such as alterations in tonus,
muscular strength, sensation, and others. Those are neu-
roplasticity-related events that like in brain injuries22 lead
to neurological function deterioration. Maybe in those
cases, PLM is a manifestation of the release of the spinal
generator resulting in leg movements.
The fact that there is a genetic component in PLM
does not exclude the possibility of this phenomenon being
884 Download full-text
Arq Neuropsiquiatr 2012;70(11):880-884
1. American Academy of Sleep Medicine. International classification
of sleep disorders, revised: diagnostic and coding manual. Chicago
(Illinois): The Academy; 2001.
2. Telles SC, Alves RC, Chadi G. Periodic limb movements during sleep
and restless legs syndrome in patients with ASIA A spinal cord injury.
J Neurol Sci 2011;303:119-123.
3. Dickel MJ, Renfrow SD, Moore PT, Berry RB. Rapid eye movement
sleep periodic leg movements in patients with spinal cord injury.
4. Brown TG. On the nature of the fundamental activity of the nervous
centres; together with an analysis of the conditioning of rhythmic
activity in progression, and a theory of the evolution of function in the
nervous system. J Physiol 1914;48:18-46.
5. Katz PS. Comparison of extrinsic and intrinsic neuromodulation in
two central pattern generator circuits in invertebrates. Exp Physiol
6. Alford S, Schwartz E, Viana di Prisco G. The pharmacology of vertebrate
spinal central pattern generators. Neuroscientist 2003;9:217-228.
7. Dickinson PS. Neuromodulation of central pattern generators in
invertebrates and vertebrates. Curr Opin Neurobiol 2006;16:604-614.
8. Kettunen P, Krieger P, Hess D, El Manira A. Signaling mechanisms of
metabotropic glutamate receptor 5 subtype and its endogenous role
in a locomotor network. J Neurosci 2002;22:1868-1873.
9. Krieger P, Grillner S, El Manira A. Endogenous activation of
metabotropic glutamate receptors contributes to burst frequency
regulation in the lamprey locomotor network. Eur J Neurosci
10. Gordon IT, Whelan PJ. Deciphering the organization and modulation
of spinal locomotor central pattern generators. J Exp Biol
11. Vilensky JA, O’Connor BL. Stepping in humans with complete spinal cord
transections: a phylogenetic evaluation. Motor Control 1997;1:284-292.
12. Heffner RS, Masterton RB. The role of the corticospinal tract
in the evolution of human digital dexterity. Brain Behav Evol
13. Pichler I, Hicks AA, Pramstaller PP. Restless legs syndrome: an update
on genetics and future perspectives. Clin Genetics 2008;73:297-305.
14. Winkelmann J, Muller-Myhsok B, Wittchen HU, et al. Complex
segregation analysis of restless legs syndrome provides evidence
for an autosomal dominant mode of inheritance in early age at onset
families. Ann Neurol 2002;52:297-302.
15. Winkelman JW. Periodic limb movements in sleep – endophenotype
for restless legs syndrome? N Engl J Med 2007;357:703-705.
16. Allen RP, Barker PB, Wehrl F, Song HK, Earley CJ. MRI measurement
of brain iron in patients with restless legs syndrome. Neurology
17. Barriere G, Cazalets JR, Bioulac B, Tison F, Ghorayeb I. The restless
legs syndrome. Prog Neurobiol 2005;77:139-165.
18. Ondo WG, Zhao HR, Le WD. Animal models of restless legs syndrome.
Sleep Med 2007;8:344-348.
19. Esteves AM, de Mello MT, Lancellotti CL, Natal CL, Tufik S. Occurrence
of limb movement during sleep in rats with spinal cord injury. Brain
20. Bara-Jimenez W, Aksu M, Graham B, Sato S, Hallett M. Periodic limb
movements in sleep: state-dependent excitability of the spinal flexor
reflex. Neurology 2000;54:1609-1616.
21. Paulus W, Schomburg ED. Dopamine and the spinal cord in restless
legs syndrome: does spinal cord physiology reveal a basis for
augmentation? Sleep Med Rev 2006;10:185-196.
22. Cecatto RB, Chadi G. The importance of neuronal stimulation in
central nervous system plasticity and neurorehabilitation strategies.
Funct Neurol 2007;22:137-143.
generated by CPG. From an evolutionary perspective, one
could suppose that the SCI and the consequent release of the
human CPG would be the necessary triggers to develop PLM.
The important point is that one theory does not exclude the
other. There might be simultaneous or multiple events that
lead to PLMs, not only in SCI patients, but also in all others
that present this disturbance. This knowledge may in the fu-
ture be another tool to deal with SCI rehabilitation.