Effects of TENS frequency, intensity and stimulation site parameter manipulation on pressure pain thresholds in healthy human subjects

Pain 106 (2003) 73–80 www.elsevier.com/locate/pain

Effects of TENS frequency, intensity and stimulation site parameter manipulation on pressure pain thresholds in healthy human subjects Linda S. Chestertona,*, Nadine E. Fostera, Christine C. Wrightb, G. David Baxterc, Panos Barlasa a

Department of Physiotherapy Studies, Keele University, Staffordshire ST5 5BG, UK b School of Health and Social Sciences, Coventry University, UK c School of Rehabilitation Sciences, University of Ulster, Jordanstown, UK

Received 6 March 2003; received in revised form 5 June 2003; accepted 8 July 2003

Abstract This study evaluated the effects of varying frequency, intensity and stimulation site, of transcutaneous electrical nerve stimulation (TENS) in an experimental model of pain. In a double-blind design 240 volunteers were randomised to one of six experimental TENS groups, a sham TENS or control (n ¼ 30 per group; gender balanced). Two TENS frequencies (110 or 4 Hz) and two intensities (strong but comfortable or highest tolerable) at a fixed pulse duration (200 ms) were applied at three sites relative to the measurement site (segmentally, extrasegmentally or a combination of these), for 30 min. Pressure pain thresholds (PPT) were measured using a pressure algometer, in the first dorsal interosseous muscle, every 10 min, during stimulation and for a further 30 min. The high frequency, high intensity segmental, and combined stimulation groups, showed rapid onset and significant hypoalgesic effects. This effect was sustained for 20 min post-stimulation in the high frequency segmental group. All other TENS intervention groups showed hypoalgesic responses similar to the sham TENS group, and none of these groups reached a clinically significant hypoalgesic level. Conclusions: The role of TENS frequency, intensity and site are pivotal to achieving optimal hypoalgesic effects, during and after stimulation. Clinical applications of these parameter combinations require further investigations. q 2003 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: TENS; Electrostimulation; Parameters; Mechanical pain; Healthy human

1. Introduction Sensory stimulation in the form of transcutaneous electrical nerve stimulation (TENS) is an established clinical tool for the treatment of pain (Walsh, 1997; Johnson et al., 1992b). Despite 35 years of use and an extensive research base, the debate surrounding the efficacy of different parameter combinations persists. Systematic reviews report a lack of evidence in support of TENS analgesia (Brosseau et al., 2002; Carroll et al., 1997, 2001; Milne et al., 2001; Gadsby and Flowerdew, 2000; Osiri et al., 2000; McQuay and Moore, 1998) however, systematic reviews by their very nature tend to focus upon methodological characteristics of studies perhaps at the expense of * Corresponding author. Tel.: þ 44-1782-584191; fax: þ 44-1782584255. E-mail address: [email protected] (L.S. Chesterton).

evaluating clinically relevant stimulation techniques. A meta-analysis by Brosseau et al. (2002), admits that the findings lacked data on how the site of application, the treatment duration, the frequency or intensity of stimulation affects TENS efficacy. This study also shows heterogeneity of patient pathology, which is often overlooked but likely to affect outcomes (Brosseau et al., 2002). Experimental investigations into the physiological effects of electrical stimulation suggest that responses follow a predictable course, dependent upon these parameter combinations (Huang et al., 2002; Loaiza et al., 2002; Sandkuhler, 2000). Furthermore, experimental studies of analgesic responses to electrical models of pain in humans suggest varied effects are observed with different parameter combinations (Chakour et al., 2000a; Cramp et al., 2000; Sluka et al., 2000; Walsh et al., 1995a, 2000; McDowell et al., 1999; Johnson et al., 1989, 1991). Chesterton et al. (2002), in a previous study, we showed that low frequency,

0304-3959/$20.00 q 2003 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0304-3959(03)00292-6

74

L.S. Chesterton et al. / Pain 106 (2003) 73–80

high intensity, extrasegmental stimulation produced a large and rapid onset hypoalgesic effect, which was sustained for 30 min post-stimulation. Whilst high frequency, ‘strong but comfortable’ intensity, segmental stimulation produced comparable hypoalgesic levels during stimulation, this effect was not sustained post-stimulation. A combination of stimulation sites using these frequency and intensity parameters also produced comparable (but not greater) hypoalgesic effects but in both cases no post-stimulation effects were observed. Thus stimulation site seems to influence maximal hypoalgesic responses and, more importantly, post-stimulation hypoalgesia. The aim of the current investigation was to extend the previous findings by examining parameter combinations of ‘Intense TENS’ and low frequency/low intensity TENS at different stimulation sites. Mannheimer and Lampe (1988) describe ‘Intense TENS’ as high frequency (100 – 150 Hz), long pulse duration (150 – 250 ms) using the highest tolerable intensity stimulation, producing tetanic muscle contractions. Although clinical use of these parameters is not widely documented, it is suggested that they produce sufficient analgesia to permit minor surgical procedures such as suture removal (Walsh, 1997). Low frequency/low intensity TENS is defined in this study as 4 Hz at a ‘strong but comfortable’ intensity. These stimulation characteristics were applied segmentally, extrasegmentally or a combination of these to the measurement site, as described in the previous study. The current study therefore aimed to determine the effects of a comprehensive range of parameter combinations (frequency, intensity and stimulation site) on an experimental model of pressure pain threshold (PPT) in healthy human volunteers, as a precursor to clinical studies.

2. Method Ethical approval was obtained from the departmental research ethics committee at Keele University. A randomised, double-blind, sham controlled experiment, using repeated PPT measurements, taken over 60 min, was undertaken. The method, including equipment, subject preparation and PPT measurements has previously been described in Chesterton et al. (2002). The only differences in this protocol relate to the parameter combinations of the active experimental groups. An outline of the protocol is as follows. Subjects were seated in a comfortable upright position and pressure pain was induced using a pressure algometer1 with a flat circular metal probe dressed in several layers of lint and measuring 1.1 cm in diameter. The algometer was mounted vertically on a purpose-built calibrated iron stand2 and force was applied perpendicular to the skin, at a controlled and steady rate of 5 N per second. Subjects were

instructed in the application of the algometer, given a demonstration and then underwent practice PPT measurements using their non-dominant hand. Subjects were asked to say ‘stop’ immediately when a sensation of pain, distinct from pressure or discomfort, was felt. At this point the experimenter immediately retracted the algometer. The PPT measurement point was marked in the mid-point of the muscle belly of the first dorsal interosseous muscle. Two measures were taken at 10 min intervals over a 60-min period, giving a total of 14 measures. The first PPT measurement was used as a baseline figure and was taken before switching on the TENS unit. 2.1. Subjects In total, data from 240 subjects were used (120 male, 120 female). For randomisation to the six active TENS groups, 180 volunteer subjects (90 female, 90 male) were recruited from the university student and staff population. Control and sham data (n ¼ 60) were utilised from the previous experiment in this series (Chesterton et al., 2002), where TENS-naive subjects were drawn from a similar sample population and in the same manner. The sample size was calculated so that statistical analysis would be supported by 80% power at a ¼ 0:05 for pair-wise comparison of active groups to the sham TENS group (Cohen, 1992). The mean age of the total sample was 24 years (range 18– 57 years). Subjects were screened for relevant contraindications: neuromuscular or cardiac disorders, peripheral neuropathy, history of trauma or surgery to the dominant hand or leg, current medication and pain, history of epilepsy, diabetes, pregnancy. Four subjects were excluded (n ¼ 1 epilepsy, n ¼ 2 current medication, n ¼ 1 altered sensation) and replacements were recruited. Subjects consented to take part after a full experimental briefing. Randomisation of subjects was achieved using computer generated random number lists and was controlled for gender to achieve equal numbers in each experimental group. One-way analysis of variance (ANOVA) showed no significant differences in PPT, between the groups at baseline (p ¼ 0:142) To control for age, this factor was used as a covariate within the main analysis. 2.2. TENS procedure Electrical stimulation was generated via a calibrated dual channel, portable TENS unit.3 An asymmetrical biphasic waveform was delivered through 5 cm2 self-adhesive carbon rubber electrodes.4 Table 1 provides a summary of the stimulation parameter combinations for all eight groups. Blinding was achieved by using two experimenters, where experimenter 1 was solely responsible for all aspects of 3

1 2

Salter Abbey Weighing Machines Ltd, England. Salter Abbey Weighing Machines Ltd, England.

TPN 300, Physio-Med Services, Glossop, Derbyshire, UK. PALS Electrode TPN 40, Physio-Med Services, Glossop, Derbyshire, UK. 4

L.S. Chesterton et al. / Pain 106 (2003) 73–80

75

Table 1 Summary of parameters for the eight experimental groups Group

Frequency (Hz) Pulse duration (ms) Active stimulation site

4 Hz segmental 4 Hz extrasegmental 4 Hz combined

4 4 4

200 200 200

110 Hz segmental 110 110 Hz extrasegmental 110 110 Hz combination 110

200 200 200

Control Sham

– –

– –

Radial nerve only Peroneal nerve (GB34) only Radial nerve (Lu 7 LI6) þ and peroneal nerve (GB34) Radial nerve only Peroneal nerve (GB34) only Radial nerve (Lu 7 LI6) þ and peroneal nerve (GB34) Both inactive Both inactive

TENS applications, and experimenter 2 was responsible for PPT measurements. Subjects and experimenter 2 were blind to group allocation. Experimenter 2 entered the room only to take measurements during which time the TENS unit was switched off. Intact skin sensation was established at each stimulation site using a standard ‘sharp/blunt’ discrimination test. Each site was then prepared with alcohol. These sites were defined according to the innervation of the first dorsal interosseous muscle. Therefore, segmental stimulation was applied over the distribution of the superficial radial nerve in the forearm, which corresponds to acupoints Large Intestine 6 (LI 6) and Lung 7 (LU 7). Extrasegmental stimulation was applied anterior and just inferior to the fibula head over acupuncture point Gall Bladder 34 (GB34) and combination stimulation was applied at both these sites simultaneously. A single channel (two electrodes) was then attached to each site, 1 cm apart with the cathode proximal. All subjects, regardless of group, had electrodes placed at both stimulation sites to maintain blinding. Before stimulation, subjects in experimental groups utilising 4 Hz and ‘strong but comfortable’ intensity were told that they would feel strong ‘pricking or tapping sensations’ beneath some or all of the electrodes and might experience muscle contractions around their thumb, fingers, leg and toes. The intensity was increased until the subject’s verbal report that the sensation was ‘strong but comfortable’. Subjects in 110 Hz ‘to tolerance’ intensity groups were told that they would feel a sharp buzzing or tingling sensation beneath some or all of the electrodes. They were also told that they would experience muscle contractions and might see movement in their fingers, thumbs and toes. Subjects were asked to allow the stimulation intensity to be increased to a level described as the maximum tolerable, without experiencing pain. Subjects in the dead battery sham TENS group, were told that some forms of TENS were imperceptible and, therefore, they may or may not feel any stimulation sensation. Since the subjects were TENS-naive it was assumed there was no sensation expectation. Subjects were also asked at regular intervals if the experimenter could increase the intensity using the dial on the TENS machine. This reinforced

Stimulation site identifier Intensity Segmental Extrasegmental Combination

Strong but comfortable Strong but comfortable Strong but comfortable

Segmental Extrasegmental Combination

To tolerance To tolerance To tolerance

None None

N/A N/A

the suggestion of active treatment. Subjects in the control group were told that, to maintain blinding of experimenter, electrodes had been attached but the machine would not be switched on and no TENS would take place. Each 10-min stimulation period was timed from the point when the intensity of the TENS had reached the appropriate level for the experimental group. The intensity was monitored throughout the period and turned up if necessary to maintain the specified level of the subjects’ verbal report. Stimulation lasted a total of 30 min, with a further monitoring period of 30 min giving a total experimental period of 60 min. 2.3. Data analysis The lower of the two PPT readings at each measurement point was used for analysis, and data were standardised to show a change against the baseline PPT recording. These difference scores ‘PPTdiff’ were calculated using the following equation: PPTdiff ¼ PPTtimeðiÞ 2 PPTbaseline . Hypoalgesic effects are therefore represented by a positive score, with a negative score indicating hyperalgesia. These interval data scores were analysed using a two-way analysis of covariance (ANCOVA) with repeated measures on the dependent variable of PPT using age as a covariate. One-way ANOVA with post hoc multiple comparisons (Bonferroni adjustments) were applied to identify differences between experimental groups. Statistical significance was set at 0.05 (a ¼ 0:05, b ¼ 0:20). All data were analysed using the Statistical Package for Social Scientists (SPSS, Version 10) for Windows.

3. Results Mean PPT difference scores (^ standard error of the mean) for all experimental groups at each time point are shown in Table 2. These mean PPT values illustrate important differences for groups 110 Hz segmental and 110 Hz combined stimulation when compared with all other experimental groups, up to the 50 min time point. The ANCOVA with repeated measures revealed significant differences for the main effect of group (p , 0:0005

76

L.S. Chesterton et al. / Pain 106 (2003) 73–80

Table 2 Mean PPT difference scores (standard error of mean) for each group at each time point (n ¼ 30 per group) Group

Stimulation

4 Hz segmental 4 Hz extrasegmental 4 Hz combination 110 Hz segmental 110 Hz extrasegmental 110 Hz combination Control Sham TENS

Post-stimulation

10 min

20 min

30 min

40 min

50 min

60 min

2.55 (1.71) 4.23 (1.72) 2.68 (1.71) 12.13 (1.73) 2.48 (1.72) 10.2 (1.71) 20.40 (1.04) 3.72 (1.28)

3.19 (1.87) 4.87 (1.89) 5.72 (1.87) 15.5 (1.89) 3.69 (1.87) 12.7 (1.88) 20.047 (1.37) 4.02 (1.84)

4.87 (2.11) 4.32 (2.12) 6.35 (2.11) 16.7 (2.13) 2.71 (2.11) 16.8 (2.11) 20.32 (1.49) 1.38 (1.76)

0.69 (2.17) 4.53 (2.19) 4.31 (2.18) 12.5 (2.20) 2.49 (2.18) 8.54 (2.18) 0.13 (1.68) 2.00 (1.52)

1.04 (1.96) 4.41 (2.05) 3.83 (1.99) 12.7 (2.01) 1.52 (1.99) 5.31 (1.99) 20.64 (1.41) 2.83 (1.64)

2.14 (2.02) 4.89 (2.03) 4.58 (2.02) 9.83 (2.04) 20.22 (2.02) 7.82 (2.02) 2.02 (1.61) 3.27 (1.68)

All scores are expressed in Newtons.

be consistent with the way you present p values, sometimes you use p ¼ 0:12 and others p ¼ 0:12), and for the interaction effect between group and time (p , 0:0005). No significant difference was shown for time or for the covariate of age. One-way ANOVA identified that significant differences occurred between the experimental groups at every time point with values ranging from p , 0:0005 to p ¼ 0:01. Post hoc multiple comparisons identified groups showing significant differences with either the control or sham TENS group. These are summarised in Table 3 and show that two groups demonstrated hypoalgesic effects, 110 Hz segmental and 110 Hz combined, both achieving statistically significant differences with the control and sham TENS groups. Significant differences with the high frequency/high intensity/segmental and combination groups were also observed at each time point throughout the stimulation period (p , 0:005 to p ¼ 0:031) for the low frequency/low intensity stimulation groups and the high frequency/high intensity/extrasegmental group. Poststimulation these differences continued between the high frequency/high intensity/segmental group and high

frequency extrasegmental and low frequency segmental groups (p ¼ 0:002 to p ¼ 0:017). Fig. 1 summarises the mean PPT difference scores over time for all the experimental groups. All values above the horizontal axis represent an increase in PPT and therefore a hypoalgesic effect. A rapid and increasing hypoalgesic effect is observed for both the 110 Hz segmental and 110 Hz combined groups, reaching a maximum mean PPT value of 16.7 and 16.8 N, respectively. This effect then declined in the post-stimulation period for both groups. However, the hypoalgesic effect in the 110 Hz segmental group remained above a suggested clinically relevant level (defined as 10 Newtons; Chesterton et al., 2003). The profile of hypoalgesic effects for all other intervention groups is similar to that of the sham TENS group.

4. Discussion The results of the present study show that high frequency, high intensity stimulation at both segmental and combined

Table 3 Summary of post hoc tests for statistical comparison between active TENS, control and sham groups at each time point Phase

Time (min)

Group

Comparison group

Mean difference between groups

Standard error

p Value

Stimulation

10

Control

Stimulation

20

Sham Control

110 Hz segmental 110 Hz combined 110 Hz segmental 110 Hz segmental 110 Hz combined 110 Hz segmental 110 Hz combined 110 Hz segmental 110 Hz combined 110 Hz segmental 110 Hz combined 110 Hz segmental 110 Hz segmental 110 Hz segmental 110 Hz segmental No significant differences

2 12.2 210.49 28.08 215.2 212.9 210.7 28.42 216.4 216.9 214.7 215.2 212.1 2 10.2 213.0 29.52

2.42 2.42 2.42 2.68 2.68 2.68 2.68 3.01 3.01 3.01 3.01 3.08 3.08 2.82 2.82

,0.0005 0.001 0.028 ,0.0005 ,0.0005 0.003 0.054 ,0.0005 ,0.0005 ,0.0005 ,0.0005 0.003 0.031 ,0.0005 0.025

Sham Stimulation

30

Control Sham

Monitoring

40

Monitoring

50

Monitoring

60

Control Sham Control Sham Control Sham

All values are expressed in Newtons.

L.S. Chesterton et al. / Pain 106 (2003) 73–80

77

Fig. 1. Mean PPT difference scores (^standard error bars) for each experimental group ( p indicates significant difference with the control group p , 0:05, þ indicates significant difference with the sham TENS p , 0:05 and horizontal line at 10 N indicates clinically important level of change).

sites produced the largest change in mean PPT with the similar rapid onset response profiles during the stimulation period. The significant maximal hypoalgesic levels (at 16.7 and 16.8 N) are suggested to be, above clinically important levels of 10 N (Chesterton et al., 2003). The groups differ post-stimulation, where the combination group showed a rapid decline in PPT, which fell below clinically relevant levels within the first 10 min. The exact mechanism for this is not known, however it could be that the combined stimulation produces both facilitatory and inhibitory effects (Millan, 2002), the net balance of which is hypoalgesia during stimulation, which is not, maintained post-stimulation. However, the high frequency, high intensity segmental group maintained a significant and clinically important effect for 20 min. All low frequency, low intensity, stimulation groups and the high frequency, high intensity, extrasegmental group showed PPT response profiles similar to that of the sham TENS and did not achieve significant differences with the control group nor clinically relevant changes. This finding may further highlight the importance of intensity as a parameter of stimulation. Despite being of potential clinical value, studies of ‘Intense TENS’ as described by Mannheimer and Lampe (1988) have only been reported by a few authors employing models of experimental pain on humans. However, it is apparent from many other relevant, well circumscribed studies, that a symbiotic relationship between parameters (particularly that of frequency and intensity) exists such that neurophysiological effects are likely to depend on combinations as opposed to any individual parameter (Walsh et al., 1995a,b, 1998; Craig et al., 1996; Johnson et al., 1989,

1991; Golding et al., 1986; Jette, 1986; Ashton et al., 1984; Andersson and Holmgren, 1975). 4.1. The effect of stimulation intensity With respect to the particular effects of high intensity, high frequency stimulation observed in this study, several authors have reported pertinent findings. In an animal model, Sjolund (1985) found that stimulation at high frequency (80 Hz) and high intensity (10 £ sensory threshold) of the plantar and sural nerves, produced maximal suppression of the C fibre evoked flexion response in a rat. Frequencies above 100 Hz required a higher intensity to induce a decreased response. Chakour et al. (2000b) corroborate Sjolund (1985) and report high frequency (80 Hz), maximal intensity stimulation as more effective at raising mean laser pain threshold than 2, 5, or 2000 Hz at either maximal or sub-maximal intensity. Indeed Chakour et al. (2000a) suggest that subjects who fail to respond to sub-maximal intensity stimulation show significant increases in pain threshold when later stimulated at maximal intensity, and therefore high intensity stimulation generally produces greater hypoalgesic effects regardless of other parameter selections. Whilst these findings were based on a study with low statistical power (only n ¼ 10 per group) and used only 5 min of TENS stimulation, the main finding is consistent with the present study. Perhaps of more interest, due to its comparable methodology, is the work of Johnson et al. (1992a). In this study ‘Intense TENS’ (80 Hz, to tolerance intensity) was tested against other parameter combinations (Conventional, Burst and Acupuncture-like TENS) every 10 min over a 60-min timescale in a cold pressor pain experimental model. One of the main findings

78

L.S. Chesterton et al. / Pain 106 (2003) 73–80

was that ‘Intense TENS’ produced a rapid and powerful hypoalgesic effect that was stronger than all other groups during the stimulation period. These results reflect the findings in the present study, however Johnson and colleagues additionally showed a rapid fall in this hypoalgesic effect during the first 10 min post-stimulation, which differs from the current study. Here a fall in hypoalgesic effect is observed in the combined site group post-stimulation, but not in the segmental group, which sees a sustained high level of hypoalgesia. This mirrors the results from our previous study (Chesterton et al., 2002) where only high intensity, low frequency extrasegmental stimulation produced a prolonged hypoalgesic effect with combined site stimulation using the same frequency and intensity combinations showing a rapid hypoalgesic fall off post-stimulation. The physiological explanation(s) for the differences observed between segmental and combination stimulation are unclear but could possibly be attributed to changes in the fine balance of inhibitory and facilitatory effects brought about by the extrasegmental stimulation parameters however, it would appear that high levels of intensity may be key in producing prolonged hypoalgesia. 4.2. The effect of stimulation site With regard to the effect of varying stimulation site in applications of ‘Intense TENS’, few previous studies are available for comparison. Andersson and Holmgren (1975) stimulated the cheek, the hand and a combination of these whilst measuring electrical pain threshold in the teeth. In agreement with the present study, they also found that segmental and combined stimulation produced similar increases in pain threshold whilst extrasegmental stimulation produced no change. Although this study was not blinded nor controlled and the results were based on only six subjects, the results are directly comparable to those observed in the current study. A larger (n ¼ 14) but still uncontrolled study by Janko and Trontelj (1980) induced pain using electrical stimuli on the distal phalanx of the middle finger. When TENS was applied at variable frequency and pulse duration of 0.1 –3 ms, pain was most effectively diminished with noxious stimulation to the same finger. The effect was reduced with noxious stimuli on two neighboring fingers and further weakened with noxious stimulus applied to the contralateral hand. This is consistent with the results seen for both the segmental and extrasegmental high frequency stimulation results observed in the current study and, therefore, it would appear that segmental stimulation must be combined with high frequencies to achieve hypoalgesic effects. 4.3. Physiological mechanisms of high frequency, high intensity stimulation The physiological mechanisms that underlie the effects of high frequency, high intensity stimulation have been

investigated by several authors and may explain the effects observed especially when different stimulation sites are used. In an animal model of primates, Chung et al. (1984) concluded that high frequency (0.5 –20 Hz), high intensity peripheral stimulation applied segmentally produced powerful in spinothalamic tract (STT) cell inhibition. An alternative electroanalgesic mechanism of peripheral blockade of both A-alpha and A-delta fibres was proposed by Campbell and Taub (1973). The authors investigated the effect of percutaneous-electrical-stimulation (100 Hz, 50 V and 0.5 ms pulse duration on the compound action potential (CAP) recorded from the human median nerve). Results showed that stimulation, applied to a single finger decreased CAP amplitude and increased latency of the A-alpha wave. Subjects also reported a reduction in pain. More recently in double-blind controlled trials, Walsh et al. (1995a, 1998) concluded that TENS-mediated analgesia might be a consequence of peripheral effect. They recorded PPT and nerve conduction in the right superficial radial nerve of healthy humans, in response to TENS of different stimulation parameters (combinations of 110 or 4 Hz; 200 or 50 ms; at strong but comfortable intensity). Stimulation at 110 Hz and 200 ms, applied directly over the course of the nerve produced a significant increase in peripheral nerve conduction latency, which lasted for at least 30 min post-stimulation. Furthermore, the authors reported a high correlation (r ¼ 0:9) between increases in PPT and negative peak latency for this TENS setting. This body of evidence in addition to the results from the current study suggests that parameter combinations producing a local intense stimulation effect are important in achieving maximal hypoalgesic effects with high frequency, high intensity, segmental stimulation repeatedly identified as showing greatest effects. 4.4. Low frequency and low intensity stimulation A number of well circumscribed experimental studies investigating low frequency, low intensity stimulation, where low intensity is defined, as ‘strong but comfortable’ have been reported. Of particular interest is a study by Andersson and Holmgren (1975) where stimulation at segmental, extrasegmental or a combination of sites showed no effect on pain threshold. The findings in relation to segmental stimulation are repeated in several studies using PPT, cold pressor pain and spinal reflex experimental protocols, by Walsh et al. (1995a,b, 1998, 2000) and Foster et al. (1996). Whilst all of these investigations have low statistical power (n # 12 in each group) the results are consistent with those observed in the present study. Interestingly Chakour et al. (2000a) suggest that low frequency (2 Hz) sub-maximal TENS should be used as physical yet non-active sham treatment as an improvement over dead battery placebo methodologies. The results of this study would support this view, since the hypoalgesic response profiles for these groups (at least in this

L.S. Chesterton et al. / Pain 106 (2003) 73–80

experimental model of pain) are very similar to that of the ‘dead battery’ sham condition. In conclusion, results from the present study show high frequency, high intensity TENS applied segmentally may be an effective form of electrostimulation analgesia providing post-stimulation effects. Results also demonstrate that low frequency, low intensity TENS is unlikely to be effective, regardless of stimulation site. These results confirm and add to our previous findings that careful selection of frequency, intensity and site is required for achieving optimal hypoalgesic effects. This has important implications for the interpretation of systematic review evidence where the impact of parameter selection is often overlooked. Therefore before concluding TENS to be an ineffective clinical modality, it would be prudent to complete further systematic clinical investigations of parameter manipulation, based upon studies suggesting effective combinations in both human and animal experimental models of pain.

Acknowledgements Funding for the experiment and equipment was provided by the Department of Physiotherapy Studies, Faculty of Health, Keele University, England.

References Andersson SA, Holmgren E. On acupuncture analgesia and the mechanism of pain. Am J Chin Med 1975;3:311–34. Ashton H, Ebenezer I, Golding JF, Thompson JW. Effects of acupuncture and transcutaneous electrical nerve stimulation on cold-induced pain in normal subjects. J Psychosom Res 1984;28:301–8. Brosseau L, Milne S, Robinson V, Marchand S, Shea B, Wells G, Tugwell P. Efficacy of the transcutaneous electrical nerve stimulation for the treatment of chronic low back pain: a meta-analysis. Spine 2002;27: 596–603. Campbell JN, Taub A. Local analgesia from percutaneous electrical stimulation. A peripheral mechanism. Arch Neurol 1973;28:347– 50. Carroll D, Moore RA, McQuay HJ, Fairman F, Tramer M, Leijon G. Transcutaneous electrical nerve stimulation (TENS) for chronic pain. Cochrane Database Syst Rev 2001; [CD003222]. Carroll D, Tramer M, McQuay HJ, Nye B, Moore RA. Transcutaneous electrical nerve stimulation in labour pain: updated systematic review. Br J Obstet Gynaecol 1997;104:195–205. Chakour M, Gibson SJ, Khalil Z, Helme RD. The influence of TENS parameters upon experimental pain perception in healthy pain free subjects. Phys Ther Rev 2000a;5:23–7. Chakour M, Gibson SJ, Neufeld M, Khalil Z, Helme RD. Develpoment of an active placebo for studies of TENS treatment. In: Devor M, Rowbotham MC, Wiesenfeld-Hallin Z, editors. Ninth World Congress on Pain. Seattle, WA: IASP Press; 2000b. p. 987–92. Chesterton LS, Barlas P, Foster NE, Baxter GD, Wright CC. Gender differences in pressure pain threshold in healthy humans. Pain 2003; 101:259–66. Chesterton LS, Barlas P, Foster NE, Lundeberg T, Wright CC, Baxter GD. Sensory stimulation (TENS): effects of parameter manipulation on mechanical pain thresholds in healthy human subjects. Pain 2002;99: 253–62. Cohen J. A power primer. Psychol Bull 1992;112:155– 9.

79

Chung JM, Lee KH, Hori Y, Endo K, Willis WD. Factors influencing peripheral nerve stimulation produced inhibition of primate spinothalamic tract cells. Pain 1984;19:277–93. Craig JA, Cunningham MB, Walsh DM, Baxter GD, Allen JM. Lack of effect of transcutaneous electrical nerve stimulation upon experimentally induced delayed onset muscle soreness in humans [see comments]. Pain 1996;67:285–9. Cramp AF, Gilsenan C, Lowe AS, Walsh DM. The effect of high- and lowfrequency transcutaneous electrical nerve stimulation upon cutaneous blood flow and skin temperature in healthy subjects. Clin Physiol 2000; 20:150–7. Foster NE, Baxter F, Walsh DM, Baxter GD, Allen JM. Manipulation of transcutaneous electrical nerve stimulation variables has no effect on two models of experimental pain in humans. Clin J Pain 1996;12:301–10. Gadsby JG, Flowerdew MW. Transcutaneous electrical nerve stimulation and acupuncture-like transcutaneous electrical nerve stimulation for chronic low back pain. Cochrane Database Syst Rev 2000; [CD000210]. Golding JF, Ashton H, Marsh R, Thompson JW. Transcutaneous electrical nerve stimulation produces variable changes in somatosensory evoked potentials, sensory perception and pain threshold: clinical implications for pain relief. J Neurol Neurosurg Psychiatry 1986;49:1397–406. Huang C, Wang Y, Han JS, Wan Y. Characteristics of electroacupunctureinduced analgesia in mice: variation with strain, frequency, intensity and opioid involvement. Brain Res 2002;945:20 –5. Janko M, Trontelj JV. Transcutaneous electrical nerve stimulation: a microneurographic and perceptual study. Pain 1980;9:219– 30. Jette DU. Effect of different forms of transcutaneous electrical nerve stimulation on experimental pain. Phys Ther 1986;66:187 –93. Johnson MI, Ashton CH, Bousfield DR, Thompson JW. Analgesic effects of different frequencies of transcutaneous electrical nerve stimulation on cold-induced pain in normal subjects. Pain 1989;39:231–6. Johnson MI, Ashton CH, Bousfield DR, Thompson JW. Analgesic effects of different pulse patterns of transcutaneous electrical nerve stimulation on cold-induced pain in normal subjects. J Psychosom Res 1991;35: 313– 21. Johnson MI, Ashton CH, Thompson JW. Analgesic effects of acupuncture like transcutaneous electrical nerve stimulation (TENS) on cold induced pain (cold pressor pain) in normal subjects. Eur J Pain 1992a;13:101 –8. Johnson MI, Ashton CH, Thompson JW. The clinical use of TENS. J Orthop Med 1992b;14:3–12. Loaiza LA, Yamaguchi S, Ito M, Ohshima N. Electro-acupuncture stimulation to muscle afferents in anesthetized rats modulates the blood flow to the knee joint through autonomic reflexes and nitric oxide. Auton Neurosci 2002;97:103–9. Mannheimer JS, Lampe GN. Electrode placement techniques. In: Mannheimer JS, Lampe GN, editors. Clinical transcuatneous electrical nerve stimulation. Philadelphia, PA: F.A. Davis Company; 1998. McDowell BC, McCormack K, Walsh DM, Baxter DG, Allen JM. Comparative analgesic effects of H-wave therapy and transcutaneous electrical nerve stimulation on pain threshold in humans. Arch Phys Med Rehabil 1999;80:1001– 4. McQuay HJ, Moore RA. An evidence based resource for pain relief. Oxford: Oxford University Press; 1998. [chapter 25, p. 207 –211]. Millan MJ. Descending control of pain. Prog Neurobiol 2002;66:355–474. Milne S, Welch V, Brosseau L, Saginur M, Shea B, Tugwell P, Wells G. Transcutaneous electrical nerve stimulation (TENS) for chronic low back pain (Cochrane Review). Cochrane Database Syst Rev 2001;2. Osiri M, Welch V, Brosseau L, Shea B, McGowan J, Tugwell P, Wells G. Transcutaneous electrical nerve stimulation for knee osteoarthritis. Cochrane Database Syst Rev 2000; [CD002823]. Sandkuhler J. Long lasting analgesia following TENS and acupuncture: Spinal Mechanisms beyond gate control. In: Devor M, Rowbotham MC, Wiesenfeld-Hallin Z, editors. Ninth World Congress on Pain. Seattle, WA: IASP Press; 2000. p. 359–69.

80

L.S. Chesterton et al. / Pain 106 (2003) 73–80

Sjolund BH. Peripheral nerve stimulation suppression of C-fiber-evoked flexion reflex in rats. Part 1: parameters of continuous stimulation. J Neurosurg 1985;63:612 –6. Sluka KA, Judge MA, McColley MM, Reveiz PM. Low frequency TENS is less effective than high frequency TENS at reducing inflammation-induced hyperalgesia in morphine-tolerant rats. Eur J Pain 2000;4:185–93. Walsh DM. TENS clinical applications and related theory. New York, NY: Churchill Livingstone; 1997. Walsh DM, Foster NE, Baxter GD, Allen JM. Transcutaneous electrical nerve stimulation. Relevance of stimulation parameters to neurophysiological and hypoalgesic effects. Am J Phys Med Rehabil 1995a;74: 199– 206.

Walsh DM, Liggett C, Baxter D, Allen JM. A double-blind investigation of the hypoalgesic effects of transcutaneous electrical nerve stimulation upon experimentally induced ischaemic pain. Pain 1995b;61:39–45. Walsh DM, Lowe AS, McCormack K, Willer JC, Baxter GD, Allen JM. Transcutaneous electrical nerve stimulation: effect on peripheral nerve conduction, mechanical pain threshold, and tactile threshold in humans. Arch Phys Med Rehabil 1998;79:1051–8. Walsh DM, Noble G, Baxter GD, Allen JM. Study of the effects of various transcutaneous electrical nerve stimulation (TENS) parameters upon the RIII nociceptive and H-reflexes in humans. Clin Physiol 2000;20: 191 –9.