Differences in Electrical Stimulation Thresholds between Men and Women Nicola A. Maffiuletti, PhD,1 Azael J. Herrero, PhD,2 Marc Jubeau, PhD,3 Franco M. Impellizzeri, MSc,1 and Mario Bizzini, MSc, PT1 Objective: Surface electrical stimulation (ES) of skeletal muscle is used in a variety of clinical settings in healthy and unhealthy subjects of both sexes. Although women generally present larger amounts of subcutaneous adipose tissue than men, which could limit current flow to the stimulated muscle, sex-related differences in ES current levels have not been clearly demonstrated to date. We report data from healthy men and wom Methods: Sensory (current perception), motor (minimal knee extension torque production), and supramotor thresholds (10% of the maximal voluntary knee extension torque) and perceived pain during surface ES of the quadriceps femoris muscle were investigated in 40 healthy volunteers (20 men, 20 women). Results: Sensory threshold was lower in women than in men (⫺43%; p ⬍ 0.001). Similarly, female muscles required lower current amplitudes to attain the supramotor threshold (⫺17%; p ⬍ 0.01). The Visual Analogue Scale pain score was significantly greater in women than in men at motor threshold (⫹112%; p ⬍ 0.01) but not at supramotor threshold (⫹36%; p ⬎ 0.05). Interpretation: Collectively, our data demonstrate higher sensory and supramotor excitability to surface ES in female subjects and provide further evidence for a neurophysiological explanation for more pronounced pain perception in women. These observations may help clinicians to better understand the sex-specific response to ES and to design more rational stimulation treatments with the ultimate goal of optimizing patient care and safety. Ann Neurol 2008;63:507–512 Surface electrical stimulation (ES) of skeletal muscle has several clinical applications, including restoration/ improvement of muscle function (eg, in neurorehabilitation)1 and pain management.2,3 Depending on treatment objectives, ES can be applied with a variety of protocols and parameters. Low-amplitude currents are adopted for stimulation of sensory nerves. At these current levels, ES is perceived by the subject through somatic sensory receptors mainly located in cutaneous and subcutaneous tissues.4 When current amplitude is increased above sensory levels, an increasing number of efferent terminal axon branches,5 in addition to myelinated afferents, are excited. This results in contractile protein interaction. At these high current levels, fat cells, located between the skin and the sarcolemma, inevitably limit current diffusion to the targeted muscle.6 Individual current levels are determined on an empirical basis by clinicians and practitioners.7 Men are generally considered to be more electrically excitable than women,8 because the former have proportionately more muscle mass and less adipose mass.9 However, to our knowledge, few or no data exist regarding sex differences in ES current levels. Because men and women can respond to ES differently and would therefore require different current doses to optimize treatment effectiveness, it is important to determine the influence of sex on ES thresholds and perceived pain. The main objective of this study was to examine sex differences in electrical sensory, motor, and supramotor thresholds with associated pain scores. We selected the quadriceps femoris because it is the muscle most often stimulated.10 From the 1Neuromuscular Research Laboratory, Schulthess Clinic, Zurich, Switzerland; 2Faculty of Health Sciences, European University Miguel de Cervantes, Valladolid, Spain; and 3Institut National de la Sante et de la Recherche Médicale U887, University of Burgundy, Dijon, France. Published online Feb 25, 2008, in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.21346 Received Aug 28, 2007, and in revised form Dec 12. Accepted for publication Jan 2, 2008. Subjects and Methods Subjects Forty healthy volunteers (20 men, 20 women) without any known orthopedic, neuromuscular, or cardiovascular problems participated in this study (Table 1). Subjects were recruited from the Burgundy University (Dijon, France) community. The majority of them were graduate students (85% of women, 80% of men), French mother tongue (85% of both women and men), and residents in urban or semiurban areas of Dijon at the time of the study (100% of women, Address correspondence to Dr Maffiuletti; Neuromuscular Research Laboratory, Schulthess Clinic, Lengghalde 2, 8008 Zurich, Switzerland. E-mail: email@example.com © 2008 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services 507 Table 1. General Characteristics (means ⴞ Standard Deviation) of the Experimental Subjects by Sex Characteristics Men (n ⴝ 20) Women (n ⴝ 20) Age (yr) 27.3 ⫾ 4.7 25.6 ⫾ 4.5 Height (m) 1.79 ⫾ 0.07 1.68 ⫾ 0.06a Body mass (kg) 75.8 ⫾ 12.7 60.9 ⫾ 8.1a Femur length (cm) 45.4 ⫾ 2.6 43.1 ⫾ 1.9a Thigh circumference (cm) 54.2 ⫾ 3.7 51.2 ⫾ 4.1 Skinfold thickness (mm) 6.9 ⫾ 2.9 22.0 ⫾ 5.4a Quadriceps crosssectional area (cm2) 77.1 ⫾ 6.3 53.1 ⫾ 8.4a 301.7 ⫾ 52.8 176.0 ⫾ 44.9a MVC torque (Nm) a p ⱕ 0.006 different from men. MVC ⫽ maximal voluntary contraction. 95% of men). None of them had previously engaged in systematic ES training. The local ethical committee approved the study, and written informed consent was always obtained. The study was conducted according to the Declaration of Helsinki. Protocol Subjects were asked not to take part in vigorous physical activity for 2 days before their test date. On arrival at the laboratory, the dominant lower extremity was determined by asking subjects which limb they preferred to use when kicking a ball. All the assessments were subsequently performed on this limb. Measurements First, femur length, thigh circumference, and skinfold thickness were measured while the subject was seated with an angle of 90 degrees at the hip and knee joints. Femur length was measured from the greater trochanter to the lateral epicondyle of the knee. Thigh circumference was obtained at half of femur length while the thigh muscles were relaxed. Femur length and thigh circumference were measured to the nearest 1mm with a tape measure. Skinfold thickness was measured using a commercially available caliper (Harpenden; British Indicators, West Sussex, United Kingdom) at two different sites: half of femur length on the anterior aspect of the thigh, and distal one-third of femur length on the lateral aspect of the thigh (ie, under the stimulating electrode; see later). All circumferences and skinfolds were the average of two measurements (same observer). Quadriceps crosssectional area was estimated according to the following formula: (2.52 ⫻ thigh circumference [in cm]) ⫺ (1.25 ⫻ skinfold thickness at half of femur length [in mm]) ⫺ 45.13.11 Subsequently, subjects were positioned into the chair of an isokinetic device with the hip at 90 degrees and the knee 508 Annals of Neurology Vol 63 No 4 April 2008 joint at 60 degrees of flexion. In these conditions, electrical sensory, motor, and supramotor thresholds with the associated pain scores (see later) were quantified at two experimental frequencies. Isometric knee extension torque was continuously recorded throughout the testing session by means of an isokinetic dynamometer (Biodex, Shirley, NY). Torque signal was fed directly from the dynamometer into a 16-bit A/D converter, then into a computer sampling at 1,000Hz using Tida software (Tida, Heka Elektronik, Lambrecht/ Pfalz, Germany). Weight of the tested limb was assessed by the dynamometer to allow torque to be corrected for that resistance. The dynamometer lever arm was attached 2 to 3cm above the lateral malleolus by using a strap. Extraneous movements of the upper body were limited by two crossover shoulder harnesses and a belt across the abdomen. The quadriceps femoris muscle was stimulated using a battery-powered ES unit (Compex Sport-P; Medicompex SA, Ecublens, Switzerland) and two surface electrodes. One electrode (5 ⫻ 10cm) was placed on the proximal (anterior) aspect of the quadriceps femoris muscle (5cm below the femoral triangle), and the other electrode (5 ⫻ 5cm) was positioned on the distal third of femur length (lateral aspect of the thigh), over the vastus lateralis muscle belly. An electroconductive gel was consistently applied between the electrodes and skin to minimize impedance. We used the following ES parameters: 400-microsecond pulse duration, 10/5second on/off ratio, and 2 different experimental frequencies, randomly presented. Frequencies of 10 and 75Hz were considered representative of low-frequency (LF) ES,12 which is generally used to mimic endurance exercise,13 and highfrequency (HF) ES,14 which is generally adopted to mimic resistance exercise.13 Following electrode positioning and instructions, current amplitude (in mA) was progressively increased by the investigator from zero to the following levels: (1) sensory threshold (current perception), when the subject indicated initial (lowest) perception of stimulus sensation (tingling, itching, heat); and (2) motor threshold, when minimal knee extension torque was produced at the knee joint (usually 5–10Nm). Current amplitude was increased until one of the thresholds was reached at a rate of about 1mA/sec; then it was reduced (approximately 3–5mA) and subsequently reincreased, to accurately determine each threshold level. Rest periods of at least 30 seconds were provided between thresholds. Subjects were consistently asked to relax during ES. A standardized warm-up, consisting of 5 minutes of submaximal ES (frequency: 5Hz; pulse duration: 350 microseconds), to further familiarize the subjects with ES, and approximately 10 submaximal voluntary (isometric) contractions of the knee extensor muscles (intensity range: 10 – 80% of the estimated maximum strength) was completed. Then participants were asked to exert two maximal voluntary contractions (MVCs). Visual feedback and consistent verbal encouragement from the investigator were provided throughout the maximal effort repetitions. Subjects were instructed to produce their maximal force (“hard”) without any concern for the rate of force development. The duration of these contractions was approximately 5 seconds, and 60 seconds of rest were interspersed between trials. A third trial was completed only if the difference between the first two MVCs was greater than 5%. The average of the two highest MVCs was retained. After 10 minutes of passive recovery, the supramotor threshold was determined by progressively increasing current amplitude beyond motor threshold, until the stimulated quadriceps produced a torque level of 10% of knee extension MVC. Considering that there is no consensus on the minimal clinically useful contraction intensity using ES,15 this relatively low force level was selected because it corresponds to the stimulated quadriceps force that can be recorded by almost all subjects (both male and female subjects) at the beginning of their first ES session.15 Immediately after stimulations at motor and supramotor thresholds, subjects were asked to place a vertical mark on a horizontal line (100mm) to rate the pain intensity, that is, Visual Analogue Scale (VAS) for pain.16 The scale therefore ranges from 0 to 100mm, with the zero value representing “no pain” and the 100 mm value representing “worst possible pain.” Statistics After check for normality, differences in age, anthropometric characteristics, and MVC torque between men and women were examined using a Student’s unpaired t test. A corrected ␣ level of p ⱕ 0.006 (0.05/8) was accepted as significant. A two-factor analysis of variance (sex ⫻ frequency) was performed on thresholds (sensory, motor, and supramotor thresholds) and on pain scores at motor and supramotor thresholds. Repeated measures on the last factor were used. If significant main effects or interactions were present, a Tukey’s (honestly significant difference) post hoc test was conducted. After correction for multiple comparisons (0.05/ 5), the significance level was set at p ⱕ 0.01. The Pearson product-moment correlation coefficients were calculated for pairs of variables (thresholds vs anthropometric characteristics), and r values ⱖ 0.60 were considered to be acceptable (with n ⫽ 40; ␣ ⫽ 0.01; ␤ ⫽ 0.05).17 Between-session reliability of ES thresholds and of the associated VAS scores at both HF and LF were assessed in a group of 10 healthy men (age: 34 ⫾ 5 years; body mass: Table 2. Summary of Two-Factor Analysis of Variance Results for Electrical Stimulation Thresholds and Pain Scores Sex Effect p Frequency Effect Threshold F F p Sensory threshold Motor threshold Supramotor threshold VAS motor threshold VAS supramotor threshold 27.7 0.0 7.3 ⬍0.001 22.4 NS 79.6 ⬍0.01 121.4 7.5 ⬍0.01 5.2 NS 3.1 NS 29.3 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 No significant sex ⫻ frequency interaction was found. VAS ⫽ visual analogue scale; NS ⫽ non significant. Fig 1. (A) Current amplitude at sensory threshold, (B) current amplitude at supramotor threshold, and (C) Visual Analogue Scale (VAS) scores at motor threshold by sex. Data (mean and standard error of the mean) are collapsed across frequencies. ***p ⬍ 0.001 different between men (gray bars) and women (white bars). **p ⬍ 0.01 different between men and women. 77 ⫾ 13kg; test-retest interval: 7 days). Average intraclass correlation coefficients for the three thresholds (HF: 0.93; LF: 0.86) and for the two VAS scores (HF: 0.90; LF: 0.91) were high and demonstrated, according to standard definition,18 excellent test-retest reliability. Results Body height, body mass, femur length, skinfold thickness, estimated cross-sectional area, and MVC torque differed significantly between men and women (see Table 1). Electrical Stimulation Thresholds Two-factor analysis of variance results are shown in Table 2. Motor threshold was not significantly affected by sex, whereas sensory ( p ⬍ 0.001; Fig 1A) and supramotor ( p ⬍ 0.01; see Fig 1B) thresholds were lower in women than in men (⫺43 and ⫺17%). Interestingly, the lowest sensory thresholds were observed in those subjects with the largest skinfold thicknesses and vice versa (r ⫽ ⫺0.71; p ⬍ 0.001; Fig 2). The slope of the regression lines did not differ significantly between men and women ( p ⫽ 0.32; 2 ⫽ 0.03, analysis of covariance) probably because of the small sample size and low r value (see Fig 2). Moreover, the lowest sensory currents were observed in those subjects with the smallest cross-sectional area and vice versa (r ⫽ 0.65; p ⬍ 0.001). For the motor and supramotor thresholds, no correlation coefficients greater than 0.6 were found. All the ES thresholds were significantly lower at HF than at LF, that is, ⫺19% for sensory (HF: 3.8 ⫾ 1.7mA; LF: 4.6 ⫾ 1.9mA), ⫺28% for motor (HF: 22.2 ⫾ 5.1mA; LF: 30.6 ⫾ 9.6mA), and ⫺40% for supramotor threshold (HF: 38.8 ⫾ 9.5mA; LF: 64.9 ⫾ 18.9mA). No significant sex ⫻ frequency interaction was observed. Pain Scores VAS score at motor threshold (see Fig 1C) was significantly higher in women than in men (⫹112%). Pain Maffiuletti et al: Electric Current Thresholds 509 Fig 2. The lowest sensory thresholds were observed in those subjects with the largest skinfold thicknesses and vice versa (r ⫽ ⫺0.71; p ⬍ 0.001). The slope of the regression lines did not differ significantly between men (solid circles) and women (open circles) ( p ⫽ 0.32, analysis of covariance). score (range, 0 –100mm) at supramotor threshold was higher at LF (66 ⫾ 28mm) than at HF (44 ⫾ 31mm) (⫹34%; p ⬍ 0.01). No significant sex ⫻ frequency interaction was observed. Discussion The main findings of this study were that electrical sensory and supramotor thresholds were significantly lower in women than in men; that is, female subjects perceived the current (sensory threshold) and produced a predetermined submaximal (10% MVC) torque output (supramotor threshold) with lower current amplitudes. Our findings also showed that self-reported pain at motor threshold was significantly greater in women than in men, and that LF ES evoked more pain than HF ES at supramotor threshold. In this investigation, sensory threshold was defined as the lowest stimulus sensation perceived by the subjects. This painless sensation was commonly described as tingling, itching, and superficial heat. Although these sensations may be mediated by both A␤ and A␦ fibers,19 it has been suggested that detection-threshold levels of stimulation predominantly activate largediameter A␤ afferents.20,21 Therefore, somatic sensory receptors activated at this first threshold would mainly include encapsulated nerve endings (eg, Pacinian corpuscles and Merkel disks) located in cutaneous and subcutaneous tissues.4 Our female subjects had a threefold larger skinfold thickness than men in the thigh area beneath the stimulating electrodes. Interestingly, women demonstrated significantly lower sensory thresholds both at LF and HF, which confirms the greater female sensitivity for electrical,22 mechanical, and thermal stimuli applied at the sensory level.23 The lowest sensory currents were observed in those subjects with the largest skinfold thicknesses (mainly women), 510 Annals of Neurology Vol 63 No 4 April 2008 therefore suggesting a link between subcutaneous adipose tissue mass and sensory excitability. This allows us to speculate that the number and/or sensitivity of cutaneous and subcutaneous receptors activated by detection-threshold current could differ between men and women. Increasing current amplitude beyond the sensory threshold results in the excitation of efferent terminal axon branches,5 in addition to myelinated afferents, which, in turn, triggers visible muscle contractions. Because subcutaneous fat thickness, which limits current propagation between the electrode and the efferent axons, is greater in women than in men,24 we expected lower motor and supramotor excitability in the former subjects. Surprisingly, we observed insignificant differences in motor threshold between sexes, and supramotor threshold was even significantly lower in the female group. Considering that both motor and supramotor threshold assessments were based on torque recordings, we speculate that two mechanisms could explain, at least in part, the observed results. First, central contributions to contractions evoked by ES, such as ESspecific brain activation (dose–response) in specific sensorimotor regions (possibly related to pain),25 and/or spinal motoneuron recruitment via electrically evoked sensory volleys (which is consistent with the development of persistent inward currents in spinal motoneurons or interneurons),26 would have been greater in women, therefore promoting central in addition to peripheral torque development.26 Second, because electrode area was the same for all subjects (25cm2), whereas quadriceps cross-sectional area was significantly larger in men (approximately 77cm2) than in women (approximately 53cm2), the relative area of stimulated muscle was inevitably greater in the female sample. This would mean that, for a given dose of electric current, a larger portion of the quadriceps muscle (relative to total muscle area) can be activated in women, particularly in those subjects presenting the smallest muscle cross-sectional areas. The contribution of these mechanisms would have been greater at higher current levels, therefore accounting for sex-related differences at supramotor but not at motor threshold. Women reported significantly higher pain scores than men at the motor threshold. Because of the large interindividual variability, sex differences at the supramotor level did not reach significance ( p ⫽ 0.086). However, four female subjects (20%), but no men, were unable to attain supramotor threshold at LF because of intolerable pain. These findings confirm that noxious electrical stimuli are perceived as more painful by healthy women than by healthy men, in line with previous research.27 For example, male subjects have been shown to have higher pain thresholds than their female peers and greater electrically induced pain tolerance,28 even if some investigators did not observe these sex differences.27 Although pain was not the main focus of this study, it is interesting to remember that thermal29 and acute pain19 sensations associated to surface ES are mediated by medium-diameter, lightly myelinated fibers (A␦) that innervate the skin (free nerve endings).20,21 Interestingly, greater epidermal nerve fiber density has recently been observed in women when compared with their male counterparts,30,31 which would explain, at least in part, their greater sensitivity to painful stimuli. Our findings add to those obtained in recent skin biopsy and cadaveric studies,30,31 and favor a neurophysiological rather than a psychological or sociocultural explanation for more pronounced pain perception in female individuals. We demonstrated significant influences of frequency modulation on ES thresholds and on the associated self-reported pain for both men and women. The finding that motor and supramotor thresholds were significantly lower at 75Hz (fused tetanus) than at 10Hz (unfused tetanus) may be explained by the fact that torque is essentially derived from a single pulse in the latter condition. This inevitably resulted in significantly greater pain scores at LF than at HF for the supramotor threshold. At the motor threshold, frequency effect was significant only before correction for multiple comparisons ( p ⫽ 0.028). The fact that pain response was reduced with increasing stimulation frequency in both sexes weighs against C-fiber involvement (“windup” phenomenon)32 in surface ES.20,21 Pain levels associated with single shocks need to be determined to rule out the possibility of windup occurrence with this ES paradigm. Further research is also required to investigate the present ES thresholds: (1) at torque levels greater than 10% MVC, (2) on muscles other than the quadriceps femoris, (3) on male and female subjects having wide ranges of subcutaneous fat, as well as the effect of (4) electrode size, (5) skin temperature, (6) anthropometrical factors,33 and (7) menstrual cycle34 on electric current thresholds. Our findings may help clinicians and practitioners better understand the sex-specific response to surface ES (current thresholds and pain) and to design more rational stimulation strategies. Because the choice of optimal current amplitudes is paramount to the effectiveness of ES treatments, neurorehabilitation settings in particular, we recommend that clinicians consider that women may present lower current thresholds than men, particularly for sensory level stimulation. Our results also provide further insights into the neurophysiological mechanisms underlying ES of intact human muscles. Within the limits of this study, female quadriceps femoris muscle appears to be more electrically excitable than male quadriceps, contrary to what is generally believed. References 1. Sheffler LR, Chae J. Neuromuscular electrical stimulation in neurorehabilitation. Muscle Nerve 2007;35:562–590. 2. Sluka KA, Walsh D. Transcutaneous electrical nerve stimulation: basic science mechanisms and clinical effectiveness. J Pain 2003;4:109 –121. 3. Lund I, Lundeberg T, Kowalski J, Svensson E. Gender differences in electrical pain threshold responses to transcutaneous electrical nerve stimulation (TENS). Neurosci Lett 2005;375:75– 80. 4. Purves D, Augustine GJ, Fitzpatrick D, et al. Neuroscience. 3rd ed. Sunderland, MA: Sinauer Associates, 2004. 5. Hultman E, Sjoholm H, Jaderholm-Ek I, Krynicki J. Evaluation of methods for electrical stimulation of human skeletal muscle in situ. Pflugers Arch 1983;398:139 –141. 6. Bergman BC, Martin DT, Wilkinson JG. Knee extensor torque and perceived discomfort during symmetrical biphasic electromyostimulation. J Strength Cond Res 2001;15:1–5. 7. Reed B. The physiology of neuromuscular electrical stimulation. Pediatr Phys Ther 1997;9:96 –102. 8. Morrissey MC. Electromyostimulation from a clinical perspective. A review. Sports Med 1988;6:29 – 41. 9. Lee SJ, Janssen I, Heymsfield SB, Ross R. Relation between whole-body and regional measures of human skeletal muscle. Am J Clin Nutr 2004;80:1215–1221. 10. Bax L, Staes F, Verhagen A. Does neuromuscular electrical stimulation strengthen the quadriceps femoris? A systematic review of randomised controlled trials. Sports Med 2005;35:191–212. 11. Housh DJ, Housh TJ, Weir JP, et al. Anthropometric estimation of thigh muscle cross-sectional area. Med Sci Sports Exerc 1995;27:784 –791. 12. Deley G, Kervio G, Verges B, et al. Comparison of lowfrequency electrical myostimulation and conventional aerobic exercise training in patients with chronic heart failure. Eur J Cardiovasc Prev Rehabil 2005;12:226 –233. 13. Atherton PJ, Babraj J, Smith K, et al. Selective activation of AMPK-PGC-1alpha or PKB-TSC2-mTOR signaling can explain specific adaptive responses to endurance or resistance training-like electrical muscle stimulation. Faseb J 2005;19:786 –788. 14. Theurel J, Lepers R, Pardon L, Maffiuletti NA. Differences in cardiorespiratory and neuromuscular responses between voluntary and stimulated contractions of the quadriceps femoris muscle. Respir Physiol Neurobiol 2007;157:341–347. 15. Alon G, Smith GV. Tolerance and conditioning to neuromuscular electrical stimulation within and between sessions and gender. J Sports Sci Med 2005;4:395– 405. 16. Downie WW, Leatham PA, Rhind VM, et al. Studies with pain rating scales. Ann Rheum Dis 1978;37:378 –381. 17. Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med 1998;17:101–110. 18. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420 – 428. 19. Julius D, Basbaum AI. Molecular mechanisms of nociception. Nature 2001;413:203–210. 20. Collins WF Jr, Nulsen FE, Randt CT. Relation of peripheral nerve fiber size and sensation in man. Arch Neurol 1960;3: 381–385. 21. Sang CN, Max MB, Gracely RH. Stability and reliability of detection thresholds for human A-Beta and A-Delta sensory afferents determined by cutaneous electrical stimulation. J Pain Symptom Manage 2003;25:64 –73. 22. Lautenbacher S, Rollman GB. Sex differences in responsiveness to painful and non-painful stimuli are dependent upon the stimulation method. Pain 1993;53:255–264. 23. Torgen M, Swerup C. Individual factors and physical work load in relation to sensory thresholds in a middle-aged general population sample. Eur J Appl Physiol 2002;86:418 – 427. Maffiuletti et al: Electric Current Thresholds 511 24. DeVahl J. Neuromuscular electrical stimulation in rehabilitation. In: Gersh M, ed. Electrotherapy in rehabilitation. Philadelphia: Davis, 1992:218 –226. 25. Smith GV, Alon G, Roys SR, Gullapalli RP. Functional MRI determination of a dose-response relationship to lower extremity neuromuscular electrical stimulation in healthy subjects. Exp Brain Res 2003;150:33–39. 26. Collins DF. Central contributions to contractions evoked by tetanic neuromuscular electrical stimulation. Exerc Sport Sci Rev 2007;35:102–109. 27. Riley JL 3rd, Robinson ME, Wise EA, et al. Sex differences in the perception of noxious experimental stimuli: a meta-analysis. Pain 1998;74:181–187. 28. Walker JS, Carmody JJ. Experimental pain in healthy human subjects: gender differences in nociception and in response to ibuprofen. Anesth Analg 1998;86:1257–1262. 512 Annals of Neurology Vol 63 No 4 April 2008 29. DiVincenti FC, Moncrief JA, Pruitt BA Jr. Electrical injuries: a review of 65 cases. J Trauma 1969;9:497–507. 30. Gøransson LG, Mellgren SI, Lindal S, Omdal R. The effect of age and gender on epidermal nerve fiber density. Neurology 2004;62:774 –777. 31. Mowlavi A, Cooney D, Febus L, et al. Increased cutaneous nerve fibers in female specimens. Plast Reconstr Surg 2005;116: 1407–1410. 32. Mendell LM. Physiological properties of unmyelinated fiber projection to the spinal cord. Exp Neurol 1966;16:316 –332. 33. Umapathi T, Tan WL, Tan NC, Chan YH. Determinants of epidermal nerve fiber density in normal individuals. Muscle Nerve 2006;33:742–746. 34. Riley JL 3rd, Robinson ME, Wise EA, Price DD. A metaanalytic review of pain perception across the menstrual cycle. Pain 1999;81:225–235.