Sensations and reaction times evoked by electrical sinusoidal stimulation

Sensations and reaction times evoked by electrical sinusoidal stimulation

Neurophysiologie Clinique/Clinical Neurophysiology (2009) 39, 283—290 ORIGINAL ARTICLE/ARTICLE ORIGINAL Sensations and reaction times evoked by elec...

405KB Sizes 0 Downloads 106 Views

Neurophysiologie Clinique/Clinical Neurophysiology (2009) 39, 283—290

ORIGINAL ARTICLE/ARTICLE ORIGINAL

Sensations and reaction times evoked by electrical sinusoidal stimulation Types de sensations et temps de réaction à des stimuli sinusoïdaux E.P.V. Félix , L.M.P. Giuliano , C.J. Tierra-Criollo , G. Gronich , N.I.O. Braga , C.A. Peres , J.A.M. Nóbrega , G.M. Manzano ∗ Federal University of Sao Paulo, Neurology and Neurosurgery, Rua Dr. Tirso Martins, 264 ap.52, 04120-050 São Paulo, Brazil Received 5 March 2009; accepted 25 October 2009 Available online 6 November 2009

KEYWORDS Perception; Reaction times; Electric current; Sensations; Sensory fibers; Pain



Summary Objective. — To determine whether 5 Hz and 2000 Hz sinusoidal electric currents evoke different sensations and to indirectly evaluate which peripheral nerve fibers are stimulated by these different frequencies. Methods. — One hundred and fifty subjects chose three among eight descriptors of sensations evoked by 5 Hz and 2000 Hz currents and the results were submitted to factor analysis. In 20 subjects, reaction times to 5, 250 and 2000 Hz currents were determined at 1.1x ST and reaction times to 5 Hz currents were also determined at 2x ST. Results. — Responses were grouped in four factors: Factor 1, which loaded mainly in descriptors related to tweezers stimulation, was higher than the other factors during 2000 Hz stimulation at 1.5x ST. Factor 2, which loaded mainly in descriptors related to needle stimulation, was higher than the other factors during 5 Hz stimulation. Factor 1 increased and Factor 2 decreased with an increase in 5 Hz intensity from 1.5 to 4x ST. Reaction times measured from the fastest responses were significantly different: 0.57 s (0.16 to 1.60), 0.34 s (0.12 to 0.71) and 0.22 s (0.08 to 0.35) for 5, 250 and 2000 Hz, respectively, and 0.22 s (0.11 to 0.34) for 5 Hz at 2x ST. Conclusions. — Sinusoidal electrical stimulation of 5 Hz and 2000 Hz evoke different sensations. At juxta-threshold intensities, RT measurements suggest that 2000 Hz stimulates A␤-fibers, 250 Hz A␤- or A∂-fibers, 5 Hz A␤-, A∂- or C-fibers. The fiber type, which was initially stimulated by the lower frequencies, depended on inter-individual differences. © 2009 Elsevier Masson SAS. All rights reserved.

Corresponding author. E-mail address: [email protected] (G.M. Manzano).

0987-7053/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.neucli.2009.10.001

284

MOTS CLÉS Perceptions ; Temps de réaction ; Courant électrique ; Sensations ; Fibres nerveuses sensitives

E.P.V. Félix et al. Résumé Objectifs. — Déterminer si des stimulations au moyen de courants électriques sinusoïdaux à 5hz et 200 Hz évoquent différentes sensations et identifier quels types de fibres nerveuses périphériques sont stimulées par ces différentes fréquences. Méthodes. — Cent cinquante sujets étaient invités à choisir trois parmi huit descripteurs des sensations évoquées par des courants de 5 Hz et 2000 Hz. Les résultats ont été soumis à une analyse factorielle. Chez 20 sujets, nous avons mesuré les temps de réaction à des courants de 5, 250 et 2000 hertz appliqués à 1,1 fois le seuil sensitif ainsi qu’aux courants de 5 Hz à deux fois le seuil sensitif. Résultats. — Les réponses ont été regroupées en quatre facteurs : Le Facteur1, correspondant principalement à des sensations de pincement, était plus représenté que les autres facteurs pendant les stimulations à 200 Hz à 1,5 fois le seuil ; Le Facteur2, correspondant principalement à des sensations de piqûre, dominait les autres facteurs pour des stimulations à 5 Hz. L’importance du Facteur 1 augmentait et celle du Facteur2 diminuait lors d’une augmentation d’intensité des stimuli à 5 Hz de 1,5 fois à quatre fois le seuil. Les temps de réaction mesurés à partir des réponses les plus rapides étaient significativement différents : 0,57 s (0,16 à 1,60), 0,34 s (0,12 à 0,71) et 0,22 s (0,08 à 0,35) pour les réponses aux stimulations à 5, 250 et 2000 Hz respectivement et 0,22 s (0,11 à 0,34) lorsque des stimuli à 5 Hz étaient appliqués à deux fois le seuil. Conclusions. — Des stimulations électriques sinusoïdales à 5 Hz et 2000 Hz évoquent différentes sensations. Pour des intensités juxtaliminaires, les temps de réaction suggèrent que les stimulations à 2000 Hz stimulent des fibres A␤, à 250 Hz des fibres A␤ ou A∂, à 5 Hz des fibres A␤, A∂ ou C. Le type de fibres initialement stimulées par les fréquences les plus faibles était variable d’un individu à l’autre. © 2009 Elsevier Masson SAS. Tous droits réservés.

Introduction Peripheral nerves are composed of nervous fibers of different diameters, which can be stimulated by electrical currents applied to the skin. It is well-known that, especially for short pulse durations, electrical rectangular pulses of increasing intensities first stimulate thick, and then thin fibers [3,12,37]. This recruitment of sensory fibers has both sensory and perceptual consequences [7,33]. It was suggested that current perception threshold (CPT) to sinusoidal electrical currents of 5, 250 and 2000 Hz relates to stimulation of C, A∂- and A␤-fibers, respectively [4]. However, the basis for this statement is not clear. Both the evaluation of drug effects in normal subjects [17,18,30,36] and the effect of ischemia [2] are suggestive of the selective character of sinusoidal stimuli. Recently, a correlation was reported in normal subjects between thermal thresholds and 5 Hz stimulations, as well as between vibratory thresholds and 2000 Hz stimulations [19]. These findings are in keeping with some previous observations [20,22,27] but at odds [29,35], or just in partial agreement with others [6,25]. The same measurements in patients with peripheral neuropathies also gave variable results. While some authors found that 5 Hz stimuli could be useful in the evaluation of C-fibers [20,22,27], others found that it might not be the case [29,35]. In some of these reports, comments can be found on the sensations that were evoked by the stimuli, suggesting a correlation between 5 Hz stimuli and thin fiber system activation [6]. On the other hand, there seems to be some agreement that 2000 Hz stimuli activate the thick fiber system, although there is still debate on whether only the thick fiber system is stimulated [5,6,35].

The aim of our study is to determine what the relationships are between the frequency of sinusoidal stimulation and the types of sensory-nerve fibers that are stimulated. We first evaluated the relationship between percepts and stimulus frequency and tried to determine whether different sensations were evoked by 5 Hz and 2000 Hz sinusoidal currents. Second, we examined whether these percepts were coherent with the expectations on what fiber systems are stimulated. Third, we examined the effect of increasing the intensity of 5 Hz stimuli. Fourth, we measured the reaction times to the different frequencies as an indirect way to evaluate what types of fibers are stimulated. Preliminary observations based on part of this material were previously published [23,24,31].

Methods General procedures This project was approved by the Federal University of São Paulo’s Ethic Committee and all volunteers signed an informed consent. Volunteers were recruited mainly among people from the University (students and health professionals) through personal contact with the involved researchers. All experiments were conducted in a sound attenuated and at ambient temperature room (23.7 ± 1.9◦ C). Volunteers sat in a comfortable armchair with hands resting on the chair’s arm. For stimulation, the distal phalanx of the left index finger was first cleaned with alcohol embedded gauze and then wiped with dry gauze. Stimulation was provided through gold electrodes that were applied to the medial and lateral surfaces of the phalanx with a thin amount of conductive gel. For Experiment II, two recording electrodes

Sensations and reaction times evoked by electrical sinusoidal stimulation Table 1

285

Factor loadings from each variable.

Picada (sting) Pontada (prick) Agulhada (pins and needles) Queimac ¸ão (burning) Aperto (squeeze) Pressão (pressure) Vibrac ¸ão (vibration) Movimento (movement) Expl. Var. Prop. total

Factor

Factor

Factor

Factor

1

2

3

4

−0.004951 0.184753 0.133355 0.115064 0.888705 0.765181 0.528996 0.112723 1.733023 0.216628

0.771624 0.753013 0.760224 0.163382 0.104701 0.105978 0.313393 0.093376 1.896195 0.237024

0.175217 0.058117 0.033912 0.052981 0.009686 0.238621 0.559506 0.948792 1.308323 0.163540

0.142729 -0.073298 0.230675 0.950812 -0.052262 0.223037 0.245229 0.002122 1.095617 0.136952

(0.9 mm Ag/AgCl discs) covered by a thin amount of conductive gel were applied to the skin overlaying the right extensor indicis proprius muscle, another pair of similar electrodes were applied to the dorsum of the left hand, and an accelerometer was attached to the distal phalanx of the right index finger. Sensory thresholds (ST) for currents at 5 Hz, 250 Hz, and 2000 Hz were then determined by the method of limits, followed by double forced choice, in protocols of 3-second ON and 2-second OFF for the 5 Hz stimulus, and 2-second ON and 2-second OFF for the 250 Hz and 2000 Hz stimulus. This choice was made according to the protocol suggested by the manufacturer of the NeurometerTM CPT (USA), which was used as the stimulating device. Statistical analyses were performed using the software STATISTICA 7.1. Detected differences were considered significant when associated probabilities were lower than 5%.

Experiment I We studied a sample of 150 unpaid healthy subjects (111 females, mean age 33.8 years, range from 15 to 61 years). Descriptors of sensations were selected in two pilot experiments. Briefly, in the first pilot experiment (nine men and 11 women, mean age 36.2 years), manual stimulations of the distal phalanx of the left index finger by cotton (slightly brushed on the ventral surface), needle (applied perpendicular to the ventral surface with enough force to elicit mild pain) and tweezers (flat stimulating surfaces

Table 2 Descriptive statistics (mean ± S.D.) for the different factor scores at 5 Hz and 2 kHz stimulation. 5 Hz

2 kHz

Intensities ( × ST)

1.5

1.5

Factor Factor Factor Factor

0.24 0.51 0.27 0.17

1 2 3 4

(pressure) (prick) (vibration) (movement)

× ST: times sensory threshold.

± ± ± ±

0.45 0.50 0.49 0.46

0.52 0.21 0.34 0.19

± ± ± ±

0.48 0.41 0.51 0.47

Figure 1 Mean (+1 se) of the scores for each factor after 5 and 2000 Hz stimulation at 1.5 × ST. A) Within factors (* P < 0.00). B) Between factors at 2 kHz (*1 P < 0.00;*2 P < 0.02). C) Between factors at 5 Hz (*P < 0.00).

286 applied on the ventral and dorsal surfaces with enough force to be clearly felt) were performed and the subjects were asked to describe the evoked sensations. From the elicited sensations by the needle and the tweezers, eight words were chosen, based on their higher frequency of occurrence. In order to reduce the number of words on the final lists, we performed a second pilot experiment (seven men and 13 women, mean age 44.3 years), in which the same manual stimulations of the same body region by the needle and the tweezers were applied, but in which the subjects had to choose three from a list of the eight words that were derived from the first pilot experiment. Based again on their higher frequency of occurrence, three descriptors related to needle stimulation [‘‘picada’’ (‘‘prick’’), ‘‘pontada’’ (‘‘pins and needles’’) and ‘‘agulhada’’ (‘‘sting’’)], and two related to tweezers stimulation [‘‘aperto’’ (‘‘squeeze’’) and ‘‘pressão’’ (‘‘pressure’’)] were chosen to elaborate the final list of descriptors to be used on the main experiment. Although these stimuli are commonly used to test thick and thin fiber function in neurological examination, tweezers may stimulate only these receptors that are related to thick fibers (provided that they should have flat stimulating surfaces and that applied pressures should not be strong enough to produce pain), while the needle that is used to elicit pain sensation does stimulates thick as well as thin fibers (e.g., [14]). Therefore, to complete the final list, three descriptors were added based on the supposition that they describe sensations related to activation of thin and thick fibers and the inability of the natural stimuli used in eliciting those sensations, i.e. ‘‘queimac ¸ão’’ (‘‘burning’’), ‘‘movimento’’ (‘‘movement’’) and ‘‘vibrac ¸ão’’ (‘‘vibration’’). After these procedures, these eight descriptors were used to elaborate lists, with these eight words appearing in random order. Data collection started following application of electrodes and ST determinations. We first presented a 10-second burst of 5 Hz stimulation at threshold intensity and the subject was asked to choose the three words from the presented list that were closest to the sensation felt during the stimulation. If he either couldn’t find appropriate words in the list or did not feel the stimulus, he was asked to choose any of the words to complete the set of three words. The same procedure was repeated with bursts changed to 5 Hz at 1.5 × ST, 2000 Hz at ST, 2000 Hz at 1.5 × ST, and 5 Hz at 2 and 4 × ST, in this order. Each series of stimuli was followed by a brief rest of at least 60 s. Data derived from the threshold stimulation were not analyzed, given the random nature of this intensity and the frequent occurrence of absent sensation at this intensity. With other intensities, all volunteers mentioned stimulus perceptions. Data were analyzed in the following way: the frequency of each pair of words was initially calculated relative to the first word of the pair, next then to the second word of the pair; thereafter, these two relative frequencies were averaged and a symmetrical 8 × 8 matrix was built being the diagonal obtained by the frequency of each word divided by itself. This matrix was submitted to a factor analysis followed by a VARIMAX normalized procedure; the factors extracted were then compared according to the different experimental situations. Newman-Keuls test was used to compare the mean of the different factors at the 1.5 × ST level for the two frequencies; the same test

E.P.V. Félix et al. was used to compare the mean of the different factors at the three levels of stimulation for the 5 Hz stimulation.

Experiment II We studied a sample of 20 unpaid healthy subjects (13 females). The mean age of the subjects was 40.2 years, range from 23 to 59 years. Upper-limb length, from now on referred as ‘‘arm length’’, was measured from the tip of the third finger to the spinous process of the C7 vertebra with the arm 90◦ abducted. Following application of electrodes and ST determinations, the protocol of data collection started, which consisted, initially, to instruct the subjects to extend the right index finger as soon and as fast as possible after the perception of a stimulus on the left index finger. After the instruction, 10 series of stimulus (2-second ON and 8-second OFF) were presented for each tested frequency and intensity. Stimuli consisted of currents at 5 Hz (1.1 × ST), 250Hz (1.1 × ST), 2000Hz (1.1 × ST), and 5Hz (2.0 × ST), each series of stimulation was followed by a brief rest. Reaction times were measured from the onset of the stimulus artifact, recorded from the dorsum of the stimulated hand, to the onset of the EMG signal. For each series of stimuli, the fasted responses were chosen for further analysis. For comparison of the means for the different stimuli an ANOVA for repeated measures was used, followed by Fisher least significant difference test when necessary; possible association of gender or age was evaluated through t test or calculation of Pearson correlation coefficient.

Results Experiment I From the factor analysis, we extracted four factors that together explained 75.4% of the data variability. As can be seen from Table 1, Factor 1 loaded on ‘‘aperto’’ (squeeze) and ‘‘pressão’’ (pressure), Factor 2 on ‘‘picada’’ (prick), ‘‘pontada’’ (pin and needles) and ‘‘agulhada’’ (sting), Factor 3 on ‘‘movimento’’ (movement) and ‘‘vibrac ¸ão’’ (vibration), and Factor 4 on ‘‘queimac ¸ão’’ (burning). Therefore Factor 1 was named ‘‘pressure’’, Factor 2 ‘‘prick’’, Factor 3 ‘‘vibration’’, and Factor 4 ‘‘burning’’. At 1.5 × ST (Table 2), during 5 Hz and 2000 Hz stimulation, the following was found: • within factors (Fig. 1A): Factor ‘‘pressure’’ was significantly higher at 2000 Hz than at 5 Hz (P < 0.0000) and Factor ‘‘prick’’ was significantly higher at 5 Hz than at 2000 Hz (P < 0.0000); • between factors: at 2000 Hz (Fig. 1B), Factor ‘‘pressure’’ was higher than Factor ‘‘prick’’ (P < 0.0000), Factor ‘‘vibration’’ (P < 0.0023), and Factor ‘‘burning’’ (P < 0.0000); Factor ‘‘vibration’’ was higher than Factor ‘‘burning’’ (P < 0.0402); at 5 Hz (Fig. 1C) Factor ‘‘prick’’ was higher than Factor ‘‘pressure’’ (P < 0.0000),

Sensations and reaction times evoked by electrical sinusoidal stimulation Table 3

287

Descriptive statistics (mean ± S.D.) for the different factor scores at 5 Hz stimulation with different intensities. 5 Hz

Intensities ( × ST)

1.5

Factor Factor Factor Factor

0.24 0.51 0.27 0.17

1 2 3 4

(pressure) (prick) (vibration) (movement)

2 ± ± ± ±

0.45 0.50 0.49 0.46

0.33 0.45 0.25 0.15

4 ± ± ± ±

0.47 0.47 0.50 0.48

0.45 0.37 0.21 0.14

± ± ± ±

0.50 0.45 0.48 0.47

× ST: times sensory threshold.

Factor ‘‘vibration’’ (P < 0.0000), and Factor ‘‘burning’’ (P < 0.0000). During 5 Hz stimulation at different intensities it was found that (Table 3, Fig. 2): Between factors: • At 1.5 × ST, Factor ‘‘prick’’ was significantly higher than Factor ‘‘pressure’’ (P < 0.0000), Factor ‘‘vibration’’ (P < 0.0001), and Factor ‘‘burning’’ (P < 0.0000); • at 2 × ST Factor, ‘‘prick’’ was significantly higher than Factor ‘‘pressure’’ (P < 0.0435), Factor ‘‘vibration’’ (P < 0.0006), and Factor ‘‘burning’’ (P < 0.0000), Factor ‘‘pressure’’ was significantly higher than Factor ‘‘burning’’ (P < 0.0055) • at 4 × ST, Factor ‘‘pressure’’ and Factor ‘‘prick’’ were significantly higher than Factor ‘‘vibration’’ (P < 0.0241) and Factor ‘‘burning’’ (P < 0.0005). Within factors: • Factor ‘‘pressure’’ showed a steady increase from 1.5 to 4 × ST, being significantly higher at 4 × ST than at 1.5 × ST (P < 0.0005); • Factor ‘‘prick’’ showed a steady decrease with increasing intensities, being lower at 4 × ST than at 1.5 × ST (P < 0.0368); • Factor ‘‘vibration’’ and Factor ‘‘burning’’ showed a steady decrease with increased intensities although no significant differences were detected.

Figure 2 Mean (± 1 se) of the scores for each factor after 5 Hz stimulation at different multiples of the ST. Within factors, factor1 (pressure) increased and factor 2 (prick) decreased with stimulus increases (* P < 0.0005; ** P < 0.0368).

Experiment II The arm length varied from 74 to 95 cm with a mean of 83 cm. Means (standard deviation) of the ST were 0.48 (0.16) mA for 5 Hz, 0.54 (0.17) mA for 250 Hz, and 1.74 (0.35) mA for 2000 Hz currents. The fastest reaction times (RT), at intensity of 1.1 × ST for the three studied frequencies and at 2 × ST for 5 Hz currents were significantly different (P < 0.0000, Fig. 3): RT to 5 Hz (1.1 × ST) was significantly higher than to 250 Hz (1.1 × ST) (P < 0,0036), 2000 Hz (1.1 × ST) (P < 0,0000), and 5 Hz (2.0 × ST) (P < 0,0000); RT to 250 Hz (1.1 × ST) was significantly higher than to 2000 Hz (1.1 × ST) (P < 0,0265) and 5 Hz (2.0 × ST) (P < 0,01912); no significant differences were detected from RT to 2000 Hz (1.1 × ST) and 5 Hz (2.0 × ST) (P < 0,8937). The range of RT between stimuli was also markedly different: for 2000 Hz, it varied from 0.08 s to 0.35 s, for 250 Hz from 0.12 s to 0.71 s, for 5 Hz (1.1 × ST) from 0.16 sec to 1.60 s and for 5 Hz (2.0 × ST) from 0.11 s to 0.34 s. Considering a mean distance of 83 cm from the stimulus site to the spinous process of C7 and a conduction velocity range range from 4 to 30 m/s for A∂-fibers [34], one may assume that the arrival of the peripheral volleys from the stimulus site to C7 would take from 27.7 to 207.5 ms. It has been estimated that the distance from C7 to SI cortex is about 210 mm [9] and that the conduction velocities for A∂fibers trough the spinal thalamic tract (STT) varies from 5.8 to 22.1 m/s [26]. Therefore, it can be estimated that the

Figure 3 Mean (± 1 se) of the fastest reaction time, from the different subjects, for the different stimuli at different intensities. ST: sensory threshold.

288

Figure 4 Fastest reaction time, from the different subjects, for the different stimuli at different intensities (open symbols) and their respective means (filled symbols). Dashed line — theoretical inferior limit of the reaction times related to C-fibers. Dotted line — theoretical inferior limit of the reaction times related to A∂-fibers. ST: sensory threshold.

time spent from C7 to cortex in this pathway varies from 9.5 ms to 36.2 ms. Overall, for A∂-fibers, our calculations lead to an estimated time from the stimulus site to cortex from 37,2 to 243,7 ms. For C-fibers, considering a conduction velocity range from 0.4 to 1.8 m/s [34] and a STT conduction velocity range from 1.4 to 4.0 m/s [26], the same calculation lead to an estimate of 513.6 ms to 2225.0 ms. Adding 200 ms to these values as an estimate to central processing time plus efferent time to response [38], we arrive to an expected RT from 237 to 444 ms for A∂fibers-mediated pathways and from 714 to 2425 ms for the C-fibers-mediated pathways. These estimates are in agreement with previously measured RT for warm and cold stimuli [38]. Fig. 4 shows the scatter of individual RT values for the tested stimuli and the theoretical lower limit of the expected RT mediated by different diameter fibers based on these rough estimates above.

Discussion Experiment I Our observations demonstrate that sinusoidal electrical currents of different frequencies relate to different descriptors of sensations, and that increasing the intensity of 5 Hz currents changed the described sensations. At 1.5 × ST, 5 Hz frequencies related more frequently to descriptors usually associated with Factor ‘‘prick’’ whereas 2000 Hz currents to those associated with Factor ‘‘pressure’’. Factor ‘‘pressure’’ loaded mainly on ‘‘aperto’’ (squeeze) and ‘‘pressão’’ (pressure), that is, descriptors associated with tweezers stimulation, whereas Factor ‘‘prick’’ loaded mainly on ‘‘picada’’ (prick), ‘‘pontada’’ (pins and needles), and ‘‘agulhada’’ (sting), that is, descriptors associated with needle stimulation. This implies that our observations are in keeping with the proposed association of 2000 Hz with thick fibers stimulation and 5 Hz with thin fibers stimulation, in agreement with our previous findings [23,24,31]. Following this rationale, the change in perceptions with the increase in intensity of the 5 Hz stimulus suggests an inverse

E.P.V. Félix et al. order of recruitment of fibers in relation to rectangular pulse stimulation. A limitation of our approach is that we cannot rule out the possibility that even low-intensity stimulation gives rise to percepts that would depend on some interaction of different fibers and give the impression of isolated stimulation of the different systems; in fact, recent observations in animals suggest that while the 2000 Hz stimulus stimulates only thick fibers, the 5 Hz stimulus stimulates both thin and thick fibers [16,28]. Another possibility to be considered refers to the existence of different subsystems within the different fiber systems, the thick fiber system is composed of at least four subsystems [13], and central interactions of stimulation of different subsystems may lead to different perceptions, as illustrated by the proposed explanation for the ‘‘grill illusion’’ within the thin fiber system [8]. It should also be remembered that fibers related to wide — dynamic range — neurons may also question the conclusions of these experiments, however, the distinction among the factors, their relation with the different stimuli and the results of the reaction times measurement suggest that this is unlikely to have great impact on the results. As a possible explaining hypothesis for the different fiber recruitment with different frequencies, some authors [4,15,20] suggested that, while thick fibers respond to the rapid 2000 Hz stimulation, thin fibers would require several milliseconds of continuous stimulation to do so. Moreover, because the large fibers repolarize more rapidly than the 5 Hz stimulus would depolarize them, their threshold would not be reached. According to our findings, these explanations might hold true only for certain intensities of stimulation (Fig. 2). Although perceptions to these stimuli have been previously mentioned, on one occasion, it was concordant with the proposed selectivity of the stimulation [20] while, on another, it was somewhat discordant [6]. However, both of these experiments apparently were based on an open description of sensations. Our experiment used a forced choice method with a restricted choice of options derived from observations with natural forms of stimulation. This seemed a better way to try quantifying the subjective experiences, even though this implied to lose extra, perhaps important, information. Our descriptors were extracted after tweezers and needle stimulation, which do not cover all the sensations that are expected from stimulation of the thin and thick fiber systems. In particular, we didn’t use variations of temperature that, depending on the heat source, could stimulate C-fibers related receptors in isolation. It is well-known that stimulation of thick fibers seems to obscure the sensations of the thin fiber system, a premise for the utilization of transcutaneous electrical stimulation of nerves (TENS) in some painful syndromes. Our observations are in keeping with this, in the sense that increasing the intensity of the 5 Hz stimulus was followed by an increase in Factor ‘‘pressure’’ paralleled by a decrease in Factor ‘‘prick’’ percepts. It was particularly interesting to note that the 5 Hz stimulation at twice ST was already increasing Factor ‘‘pressure’’ percepts. This could be an explanation for some of the inconsistencies found in the literature, as for the capacity of the ST at 5 Hz to detect involvement of thin fibers in neuropathies. Some authors found that the threshold

Sensations and reaction times evoked by electrical sinusoidal stimulation determination for perception of 5 Hz currents was unable to predict the presence of thin fiber involvement in certain peripheral neuropathies [29,35], while others found it useful [20,27]. Looking at the distribution of normal values for the current perception threshold at 5 Hz, it was shown that about 49% of subjects would display twice the ST and still be within the normal range [11]. This suggests that a similar proportion of patients with isolated involvement of thin fibers would detect the 5 Hz stimulus through the thick fibers, while perception thresholds would be within the normal range. On the other hand, abnormalities would be detected in pathologies involving both systems. An apparent exception [27] can be explained when we look at the abnormal data reported, they were lower than the reference range, suggesting a positive instead of negative consequences of the studied neuropathy, which was, in fact, mainly characterized by positive symptoms. It is also important to consider that our observations probably depend on the duration of the used stimulus trains. Since our protocol mainly favored spatial summation, increasing the intensity of the 5 Hz stimulus in a continuous way might well lead directly to painful sensations, which would evidence thin fiber involvement with higher intensities, in this case due to temporal summation. Another possible interpretation would be that only thick fibers were stimulated after the 5 Hz currents, particularly considering that, for pain, a case has been raised for thick fiber involvement [10]. However, at twice ST, the 5 Hz stimulus evoked ‘‘thick fiber percepts’’ and the RT also suggested thick fiber stimulation (to be discussed below), which makes this interpretation less likely. Another interesting observation is that, while the measurement in clinical situations sometimes performed poorly (discussed above), the reviewed studies with drugs and tourniquet paralysis showed evidence towards selective fiber involvement; it should be noted that, in these last studies, differences were detected between sample means, as in the present work.

289

lation of A∂-fibers, however the stimulation with the 5 Hz stimulus seems to give results faster than what would be expected from stimulation of C-fibers. Considering the distribution of the responses obtained in the present work (Fig. 4), the data show that the response variability is most marked for the 5 Hz, followed by the 250 Hz and the 2000 Hz stimulus. Based on these variations, the higher frequency stimulus seems to be related to A␤fibers stimulation (although stimulation of fast A∂-fibers cannot be ruled out for some subjects), the 250 Hz stimulus seemed to be mainly related to A∂-fibers (or A␤-fibers in some individuals), and the 5 Hz currents to be related with either type of fibers, depending on individuals. In fact, these observations are in keeping with recent simulations and animal studies reporting the spectrum of fibers stimulated by these stimuli [16,28]. These last observations may explain why we and others found correlations between thin fiber functions and 5 Hz stimulation, thick fiber functions and 2000 Hz stimulation, and the poor performance of this test in some clinical studies. Some limitations of our approach should be remembered. First, we assumed in our calculations that the central processing and efferent time of the responses were similar (about 200 ms) for the different fibers types. Although, this is not proven by our data, laser studies with measurements of RT gave similar results for A∂- and C-fiber stimulation [21]. Second, the RT might be mediated by fibers that would be different as a function of which percept is evoked. However, as the 5 Hz stimulus, at low intensities, gave results suggesting stimulation of thin fibers with compatible RT and the same stimulus, at higher intensities, gave percepts and RT suggestive of stimulation of thick fibers, makes this possibility unlikely. Third, this last argument, also makes another possibility less probable, namely that the 5 Hz stimulus might only stimulate thick fibers and the differences observed in RT would, in this case, be secondary to the need for a few current cycles to be delivered (for the 5 Hz stimulation that would mean up to 200 ms per cycle).

Experiment II Our findings fully support our starting hypothesis that, if the different stimuli activated different set of nerve fibers then different reaction times would be obtained. Moreover, these are in agreement with other observations [12,17,18,23,24,31,36]. However, the observed differences raised some doubts about the specificity of the stimulation, at least for some of the tested stimuli. Measurements of RT related to the fastest peripheral fibers had been previously reported in the literature and mean values around 150 to 250 ms were described when responses were based on index finger movements. Variations were possibly related to the type of signal that was used to measure the response, stimulus, and age [1,32]. The responses recorded after stimulation at 2000 Hz (1.1 × ST) and 5 Hz (2 × ST) are similar to the reported values and therefore suggest that these responses are related to stimulation of A␤-fibers. Considering the rough estimates we used to calculate the expected RT after different fibers stimulation, the mean RT obtained from 250 Hz stimulation are suggestive of stimu-

Conclusion Our findings show that electrical currents of sinusoidal shape at frequencies of 5 Hz and 2000 Hz do evoke different sensations, which may be explained by the stimulation of different subsets of nerve fibers. Furthermore, we showed that changing the intensity of the 5 Hz stimulus changes the evoked sensations. At juxta-threshold intensities, RT measurements suggest that 2000 Hz stimulates A␤-fibers, 250 Hz A␤- or A∂-fibers and 5 Hz A␤-, A∂- or C-fibers. The fiber type, which was initially stimulated by the lower frequencies, depended on inter-individual differences.

Acknowledgements This work financially supported by FAPESP grant 05337-6 and CNPQ grant 478476/2004-3. No authors have any conflict of interest.

290

References [1] Anstey KJ, Mack HA, Christensen H, Li SC, Reglade-Meslin C, Maller J, et al. Corpus callosum size, reaction time speed and variability in mild cognitive disorders and in a normative sample. Neuropsychologia 2007;45:1911—20. [2] Baron CB, Irving GA. Effects of tourniquet ischemia on current perception thresholds in healthy volunteers. Pain Pract 2002;2:129—33. [3] Blair EA, Erlanger J. A comparison of the characteristics of axons through their individual electrical responses. Am J Physiol 1933;106:524—64. [4] Chado HN. Current perception threshold evaluation of sensory nerve function in pain management. Pain Dig 1995;5:127—34. [5] Cheng W, Jiang Y, Chuang L, Huang C, Heng L, Wu H, et al. Quantitative sensory testing and risk factors of diabetic sensory neuropathy. J Neurol 1999;246:394—8. [6] Chu N. Current perception thresholds in toe-to-digit transplantation and digit-to-digit replantation. Muscle Nerve 1996;19:183—6. [7] Collins WF, Nulsen FE, Randt CT. Relation of peripheral nerve fibre size and sensation in man. Arch Neurol 1960;3:381—5. [8] Craig AD, Bushnell MC. The thermal grill illusion: unmasking the burn of cold pain. Science 1994;265:252—5. [9] Desmedt JE, Cheron G. Central somatosensory conduction in man: neural generators and interpeak latencies of the far-field components recorded from neck and right or left scalp and earlobes. Electroenceph clin Neurophysiol 1980;50:382—403. [10] Djouhri L, Lawson SL. A␤-fiber nociceptive primary afferent neurons: a review of incidence and properties in relation to other afferent A-fibre neurons in mammals. Brain Res Rev 2004;46:131—45. [11] Galvão MLS, Manzano GM, Braga NIO, Nóbrega JAM. Determinac ¸ão do limiar de percepc ¸ão de corrente elétrica em uma amostra de voluntários normais. Arq Neuropsiq 2005;63:289—93. [12] Hallin RG, Torebjörk HE. Electrically induced A and C fibre responses in intact human skin nerves. Exp Brain Res 1973;16:309—20. [13] Johnson KO, Yoshioka T, Vega-Bermudez F. Tactile functions of mechanoreceptive afferents innervating the hand. J Clin Neurophysiol 2000;17:539—58. [14] Kakigi R, Shibasaki H. Scalp topography of mechanically and electrically evoked somatosensory potentials in man. Electroenceph clin Neurophysiol 1984;59:44—56. [15] Katims JJ. Electrodiagnostic functional sensory evaluation of the patient with pain: a review of the neuroselective current perception threshold (CPT) and pain tolerance threshold (PTT). Pain Dig 1998;8:219—30. [16] Koga K, Furue H, Rashid H, Takak A, Katafuchi T, Yoshimura M. Selective activation of primary afferent fibers evaluated by sine-wave electrical stimulation. Mol Pain 2005;1:13—21. [17] Liu SS, Gerancher JC, Bainton BG, Kopacz DJ, Carpenter RL. The effects of electrical stimulation at different frequencies on perception and pain in human volunteers: epidural versus intravenous administration of fentanyl. Anesth Analg 1996;82:98—102. [18] Liu S, Kopacz DJ, Carpenter RL. Quantitative Assessment of differential sensory nerve block after lidocaine spinal anesthesia. Anesthesiology 1995;82:60—3. [19] Lowenstein L, Jesse K, Kenton K. Comparison of perception threshold testing and thermal-vibratory testing. Muscle Nerve 2008;37:514—7.

E.P.V. Félix et al. [20] Masson EA, Veves A, Fernando D, Boulton AJM. Current perception thresholds: a new, quick, and reproducible method for the assessment of peripheral neuropathy in diabetes mellitus. Diabetol 1989;32:724—8. [21] Mouraux A, Guérit JM, Plaghki L. Non-phase locked electroencephalogram (EEG) responses to CO2 laser skin stimulations may reflect central interactions between A[delta]- and C-fiber afferent volleys. Clin Neurophysiol 2003;114:710—22. [22] Oishi M, Mochizuki Y, Katsuhiko O, Naganuma T, Nishijo Y, Mizutani T. Current perception threshold and sympathetic skin response in diabetic and alcoholic polyneuropathies. Intern Med 2002;41:819—22. [23] Pereira MTS, Giuliano LMP, Tierra-Criollo CJ, Paula Jr AR, Manzano GM. Sensac ¸ões a estimulac ¸ão somato-sensitiva senoidal. IFMBE Proceed 2004;5:1119—21. [24] Pimentel JM, Petrillo R, Vieira MMF, Giuliano LMP, Tierra-Criollo CJ, Braga NIO, et al. Perceptions and electric senoidal current stimulation. Arq Neuropsiq 2006;46:10—3. [25] Pitei DL, Watkins PJ, Stevens MJ, Edmonds ME. The value of neurometer in assessing diabetic neuropathy by measurement of the current perception threshold. Diab Med 1994;11: 872—6. [26] Qiu Y, Inui K, Wang X, Tran TD, Kakigi R. Conduction velocity of the spinothalamic tract in humans as assessed by CO2 laser stimulation of C-fibers in men. Neurosci Lett 2001;311: 181—4. [27] Ro L, Chen S, Tang L, Hsu W, Chang H, Huang C. Current perception threshold testing in Fabry’s disease. Muscle Nerve 1999;22:1531—7. [28] Sundar S, González-Cueto JA. On the activation threshold of nerve fibers using sinusoidal electrical stimulation. Conf Proc IEEE Eng Med Biol Soc 2006;1:2908—11. [29] Tack CJJ, Netten PM, Scheepers MH, Meijer JWG, Smits P, Lutterman J. Comparison of clinical examination, current and vibratory perception threshold in diabetic polyneuropathy. Netherl J Med 1994;44:41—9. [30] Tay B, Wallace MS, Irving G. Quantitative assessment of differential sensory blockade after lumbar epidural lidocaine. Anesth Analg 1997;84:1071—5. [31] Tierra-Criollo CJ, Pereira MTS, Camêlo PM, Giuliano LMP, Manzano GM, Paula Jr AR. Agrupamento de sensac ¸ões somatossensoriais sem e durante estimulac ¸ão de corrente senoidal. Rev Bras Eng Biomed 2006;22:105—13. [32] Tomberg C, Levarlet-Joye H, Desmedt JE. Reaction times recording methods: reliability and EMG analysis of patterns of motor commands. Electroenceph clin Neurophysiol 1991;81:269—78. [33] Torebjörk HE, Hallin RG. Identification of afferent C units in intact human skin nerves. Brain Res 1974;67:387—403. [34] Vallbo AB, Hagbarth KE, Torebjork HE, Wallin BG. Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves. Physiol Rev 1979;59:919—57. [35] Vinik AI, Suwanwalaikorn S, Stansberry KB, Holland MT, McNitt PM, Colen LE. Quantitation measurement of cutaneous perception in diabetic neuropathy. Muscle Nerve 1995;18:574—84. [36] Wallace MS, Dyck JB, Rossi SS, Yaksh TL. Computer-controlled lidocaine infusion for the evaluation of neuropathic pain after peripheral nerve injury. Pain 1996;66:69—77. [37] Wiederholt WC. Threshold and conduction velocity in isolated mixed mammalian nerves. Neurology 1970;20:347—52. [38] Yarnitsky D, Ochoa JL. Warm and cold specific somatosensory systems: Psychophysical thresholds, reaction times and peripheral conduction velocities. Brain 1991;114:1819—26.