stroke - brain waves - electrophysiology
acidente vascular cerebral - ondas encefálicas - eletrofisiologia
Stroke often leads to a significant decrease in patient quality of life, yet it has
been the subject of few electrophysiological studies[1],[2]. Following the pioneering work of McComas et al.[3], electromyography studies showed that changes in muscle fiber density occur during
the chronic phase of stroke[4],[5], and that positive sharp waves and fibrillation are observed two to three weeks
after stroke[6],[7]. However, few studies have used electrophysiological examinations to evaluate F-wave
alterations in the acute phase of stroke.
F-waves are late muscle responses elicited by the firing of antidromically stimulated
motor neurons[8]. F-wave measurements are helpful in evaluating conduction along the entire flength
of peripheral motor neuron axons, including the most proximal segment[9]. Mesrati and Vecchierini[10] reported that F-wave changes occur in central nervous system (CNS) diseases, and
concluded that F-waves are absent during the acute phase of CNS lesions but persist
in the chronic phase in association with spasticity and hyperreflexia. Alterations
in F-waves are associated with major severity of the CNS diseases and a poor long-term
motor prognosis[10].
In the only study that investigated the relationship between level of consciousness
and F-waves, it was concluded that F-waves may be useful as an objective measure of
the severity of consciousness impairment, but an association between the neurological
condition and laboratory exam results was not reported[8]. Given that stroke is a main cause of chronic incapacity in adults, F-wave alterations
in the acute phase of stroke could be important for future prognostic studies. Therefore,
we hypothesized that the persistence of F-waves correlates with the severity of stroke
in the acute phase. The aim of this study was to describe F-wave alterations in the
acute phase of stroke and to evaluate their association with clinical and laboratory
exam results.
METHODS
Study design, setting and participants
A single-center cross-sectional study was conducted in accordance with the principles
of the Declaration of Helsinki[11]. Patients were selected following approval of the study protocol by the Institutional
Review Board of the Botucatu Medical School (UNESP). All participants or their legal
representatives were aware of the study objectives and provided written informed consent.
The study included 20 individuals with a stroke diagnosis that was confirmed by CT-scan
or MRI. The individuals were admitted to the stroke unit of Universidade Estadual
Paulista (UNESP) between March 2012 and November 2013, and were included only if they
met the following eligibility criteria: first stroke event, presence of hemiparesis
with scores ≥ 1 in items 5 or 6 of the National Institutes of Health Stroke Scale/Score
(NIHSS), absence of previous cranial trauma, absence of previous myopathy, absence
of diabetes (glycated hemoglobin < 6.5%) as specified by the expert committee on the
diagnosis and classification of diabetes mellitus[12], absence of alcoholism and other known causes of peripheral neuropathy, and normal
sensory and motor conduction in all four limbs as measured by conventional electrophysiological
techniques[9].
Variables
Exposures
The independent variables of clinical condition were: age (years), race (non-Caucasian
vs. Caucasian), glycemia (mg/dL), glycated hemoglobin (HbA1c, %), creatine phosphokinase
(CPK, U/L), NIHSS score upon admission (total score), and type of stroke (ischemic
vs. hemorrhagic). Glycemia, HbA1c, and CPK were measured by routine laboratory examination.
The NIHSS was administered prior to the electrophysiological examination and aimed
to quantify the severity of the neurological condition and was subdivided into 11
items: consciousness level, speech, language, visual field, unilateral spatial neglect,
ocular movement, strength, coordination, and sensitivity. Each item was scored on
a scale from 0 to 4, where a higher score was associated with a poorer neurological
clinical condition[3]. Stroke was classified as either ischemic or hemorrhagic by means of a CT-scan or
MRI evaluation according to international guidelines.
Outcomes
The outcomes of the electrophysiological examination after admission were: F-wave
persistence and latency. The temperature of the lower limb was maintained above 32°C
and F-waves were measured from the deep peroneal nerve using symmetrical techniques
on the hemiparetic side. Analysis time was set to 10 ms/cm with a sensitivity of 200 μV and the band-pass
filter was set to 20–3,000 Hz. A minimum of two series of 16 supramaximal stimuli
were applied to the deep peroneal nerve at the head of the fibula and it was captured
in the extensor digi
torum brevis muscle in order to determine the persistence and latency of F-waves. The minimum
amplitude to evoke a positive response (i.e., F-waves) was 10% of the amplitude of
the M-wave. The examination was performed by a single examiner, who was certified
by the Brazilian Society of Clinical Neurophysiology. All independent variables were
measured upon admission to the stroke unit, and the outcomes were measured within
24–48 h.
Statistical methods
The univariate association between each clinical exposure and electrophysiological
outcome was analyzed using non-parametric tests (Spearman and Mann-Whitney). The exposures
most strongly associated (p < 0.0083 by Bonferroni correction, α = 0.05/6) with the
outcomes were included in a model of multiple linear regression. The number of independent
variables associated with each outcome was 6 in the univariate regression model. After
adjustment of the model, the Shapiro-Wilk diagnostic did not reveal any violation
of the presuppositions of normality and homoscedasticity. Significance was set at
p < 0.05. Statistical analyses were performed using SPSS version 21.0 (IBM®, Chicago, Illinois, USA).
RESULTS
Over the course of this study, only 20 of the 168 individuals recruited met the study
eligibility criteria. The principal reasons for exclusion of individuals were: absence
of hemiparesis (18 patients), presentation of cranial trauma (8 patients), previous
myopathy (2 patients), previous diagnosis of diabetes (57 patients), history of alcoholism
(24 patients), previous presentation of peripheral neuropathy (36 patients), and presentation
of abnormal sensory and motor conduction before F-wave examination (13 patients).
The characteristics of participants in this study are displayed in [Table 1]. The majority of patients were Caucasian (85%) and presented with hyperglycemia
(78%) and ischemic stroke (75%) at admission.
Table 1
Characteristic of participants (n = 20).
|
Variable
|
Participants
|
|
Age (years)
|
66 (26-88)1
|
|
Gender
|
|
Male
|
10 (50.0%)
|
|
Female
|
10 (50.0%)
|
|
Race
|
|
Non-Cau
casian
|
3 (15.0%)
|
|
Caucasian
|
17 (85.0%)
|
|
NIHSS
|
|
1-7
|
15 (75,0%)
|
|
9-23
|
5 (25.0%)
|
|
Glycemia (mg/dL)
|
96.5 (61.0–172.0)1
|
|
HbA1c (%)
|
5.6 (5.2–6.0)1
|
|
CPK (xx)
|
86.5 (24.0–225.0)1
|
|
Type of stroke
|
|
Ischemic
|
15 (75.0%)
|
|
Hemorrhagic
|
5 (25.0%)
|
|
F-waves
|
|
Persistence (%)
|
81.2 (8.0–100.0)1
|
|
Latency (ms)
|
43.7 (30.7–50.0)1
|
1Values in median; HbA1c: glycated hemoglobin; NIHSS: National Institutes of Health
Stroke Scale.
The latency of F-waves was not significantly associated with any clinical or laboratory
exam result ([Table 2]). However, the persistence of F-waves was significantly correlated with glycemia
(r = 0.71; p < 0.001; [Figure A]) and NIHSS scores (NIHSS 1–7 = 65.0 × NIHSS 9-23 = 100; p = 0.004; [Figure B]).
Table 2
Association between F-waves and clinical and laboratory exams.
|
cPK (U/L)
|
Persistence
|
p
|
Latency
|
p
|
|
Age (years)
|
r = 0.31
|
0.181
|
r = -0.008
|
0.975
|
|
Race
|
|
Non-caucasian (n = 3)
|
87.5 (81.2–84.0)
|
0.241
|
42.5 (41.1–48.4)
|
0.791
|
|
Caucasian (n = 17)
|
80.0 (8.0–100.0)
|
44.3 (30.7–50.0)
|
|
NIHSS
|
|
1-7 (n = 15)
|
65.0 (8.0–84.0)
|
0.004
|
44.7 (30.7–50.0)
|
0.407
|
|
9-23 (n = 5)
|
100 (81.3–100.0)
|
43.2 (37.0–44.8)
|
|
Glycemia (mg/dL)
|
r = 0.71
|
< 0.001
|
r = 0.10
|
0.672
|
|
CPK (U/L)
|
r = -0.32
|
0.158
|
r = 0.30
|
0.188
|
|
Type of stroke
|
|
Ischemic (n = 15)
|
80.0 (8.0–100.0)
|
0.292
|
43.2 (30.7–48.4)
|
0.965
|
|
Hemorrhagic (n = 5)
|
81.2 (40–100.0)
|
44.3 (37.0–50.0)
|
NIHSS: National Institutes of Health Stroke Scale; CPK: Creatine phosphokinase.
Figure (A) Scatterplot of F-waves versus glycemia; (B): Boxplot of the association between
F-waves and stroke severity.
The exposures most strongly associated with our outcomes were included in the multivariate
regression model ([Table 3]). Glycemia was the most important predictor for an increase in the persistence of
F-waves (p < 0.001), with each 1 mg/dl of glycemia augmenting the persistence of F-waves
by 0.59% [β = 0.59 (0.44–0.74); p < 0.001].
Table 3
Linear regression adjusted to account for the F-wave persistence as a function of
glycemia and NIHSS.
|
Variable
|
β
|
SE
|
p
|
CI 95%
|
|
Glycemia
|
0.594
|
0.070
|
< 0.001
|
0.448
|
0.741
|
|
NIHSS (9-23)
|
17.921
|
14.645
|
0.237
|
-12.848
|
48.689
|
R2aj = 82.1; pSW = 0.261; NIHSS: National Institutes of Health Stroke Scale; β: estimated
coefficient; SE: standard error; CI: confidence interval.
DISCUSSION
The generation of F-waves is dependent upon the excitability of motor neurons. An
increase in the persistence of F-waves indicates alpha motor neuron hyperexcitability,
whereas a reduction in the persistence of F-waves indicates hypoexcitability[13]. In the acute phase of stroke, F-wave persistence is significantly reduced on the
paretic side. In our study, the mean F-wave persistence was 81.2%. Other studies have
reported that F-wave amplitudes and persistence are decreased in clinically involved
limbs, and this finding is compatible with observations of decreased central excitability
(e.g., decreased tone and reflexes) in patients who are studied early after stroke[14],[15].
In the acute phase of stroke, F-waves may be absent owing to the hyperpolarization
of spinal motor neurons; however, in the late phase, many authors report that F-waves
are increased (100%) owing to the augmentation of central nervous system excitability,
the disinhibition of supraspinal descending pathways, and a slow conduction period[10]. A higher persistence of F-waves in the acute phase of stroke may be related to
a poorer neurological outcome[15]. Several trials have reported an association between the NIHSS score and functional
outcome[16],[17], but an association between the persistence of F-waves and NIHSS scores was not
significant in the final model of regression in this study. Some studies have reported
differences in F-wave latencies observed on the normal side versus the paretic side[18], whereas other studies did not observe any differences between the normal and paretic
sides[19],[20],[21]. In this study, we did not find any differences in F-waves latencies between the
normal and paretic sides.
Regarding the biochemical variables in our study, the majority of patients presented
with hyperglycemia in the acute phase of stroke. This effect may be related to the
elevation of catecholamines during ischemia or alternatively the activation of the
hypothalamic-pituitary axis and suprarenal glands, which can lead to activation of
the sympathetic nervous system, elevated glucocorticoid production, and stimulates
the production of glucose by glycogenolysis, glyconeogenesis, proteolysis and/or lipolysis[22],[23]. Hyperglycemia in the acute phase of stroke may play a role in ischemic neuronal
damage[24] and, additionally, hyperglycemia is associated with a poorer long-term prognosis[25],[26],[27].
Hyperglycemia has been reported to alter nerve conduction in some preclinical studies[28],[29]. In clinical studies, acute hyperglycemia did not alter nerve conduction velocities
and amplitudes[30], but chronic hyperglycemia has been hypothesized to be more detrimental to nerves[31]. The relationship between hyperglycemia and nerve conduction is therefore controversial
in the literature, and no study to date has reported a relationship between the persistence
of F-waves and an acute glucose disorder. Therefore, we postulate the hypothesis that
hyperglycemia leads to neuronal hyperexcitability and exacerbated brain damage after
cerebral ischemic injury.
The suggestion that F-waves are a useful prognostic indicator in the acute phase of
stroke requires further study in a larger patient sample and with follow-up examination
of patients in the chronic phase of stroke. The principal limitation of our study
is the small sample size. The methodological rigor of the adopted inclusion criteria
was a barrier to the recruitment of patients. Additionally, electrophysiological evaluations
were operator-dependent and therefore limited our ability to screen patients. The
long-term prognosis was not evaluated in this manuscript. Despite the limitations
of our study, we report a novel finding in a Brazilian population of the principal
alterations of F-waves in the acute phase of stroke, and these results provide useful
parameters for future research studies.
We conclude that increases in the persistence of F-waves are associated with hyperglycemia
in the acute phase of stroke.
