External electric muscle stimulation improves burning sensations and sleeping disturbances in patients with type 2 diabetes and symptomatic neuropathy

P A I N M E D I C I N E
External Electric Muscle Stimulation Improves Burning
Sensations and Sleeping Disturbances in Patients with Type 2
Diabetes and Symptomatic Neuropathy
Per M. Humpert, MD,* Michael Morcos, MD,* Dimitrios Oikonomou, MD, Karin Schaefer,Andreas Hamann, MD, Angelika Bierhaus, PhD, Tobias Schilling, MD, and Peter P. Nawroth, MD Medizinische Klinik 1 und Klinische Chemie, University Hospital, Heidelberg, Germany*PMH and MM contributed equally.
A B S T R A C T
Objective. External muscle stimulation (EMS) of the thighs was previously shown to have beneficialeffects in a pilot study on painful diabetic neuropathy. However, differential effects on specificsymptoms of neuropathy as well as determinants of treatment response have not been described.
Design. Ninety-two type 2 diabetes patients with different neuropathic symptoms were included ina prospective uncontrolled trial. Patients were treated twice a week for 4 weeks. Symptoms weregraded on numeric scales at baseline, before the second and the eighth visit.
Results. Seventy-three percent of the participants reported marked improvement of symptoms.
Subjective treatment response was positively and independently associated with symptom intensitybut independent of disease extent, metabolic factors, age, or gender. Total symptoms graded bypatients on numerical scales decreased significantly after 4 weeks of treatment. Patients in the uppertertile of symptom intensity showed significant improvement of paresthesia, pain, numbness andmost pronounced for burning sensations and sleeping disturbances.
Conclusions. In an uncontrolled setting, EMS seems to be an effective treatment for symptomaticneuropathy in patients with type 2 diabetes, especially in patients with strong symptoms.
Key Words. Type 2 Diabetes; Diabetic Neuropathy; Electric Muscle Stimulation; Treatment Introduction
regarding the influences of EMS on neuronal It was recently shown in a pilot study that exter- functionandmechanismsunderlyingtheobserved nal electric muscle stimulation (EMS) of the treatment effects. Previous studies suggested that thigh causing isometric muscle contraction might electrical stimulation activates the dorsal columns, be an effective treatment for painful diabetic neu- inhibits C-fibers, and consequently leads to a ropathy [1]. Type 1 and 2 diabetes patients were decrease in pain perception [3]. EMS, which can treated with EMS and compared with patients stimulate a large number of nerves in the treated with transcutaneous electric nerve stimu- thigh, might therefore lead to a spinal stimulation that in turn decreases excitability of small nerve reduced the total symptom score significantly, while EMS resulted in a significantly higher It is not known which specific symptoms of response rate [1]. So far, there is no information diabetic neuropathy can be treated using EMS andwhich factors determine the treatment responses.
Hence, we conducted this study in a large group Reprint requests to: Per M. Humpert, MD, Medizinische of type 2 diabetes patients with neuropathic Klinik 1 und Klinische Chemie, Im Neuenheimer Feld symptoms to identify subjects benefiting from 410, 69120 Heidelberg, Germany. Tel: +49-6221-56-38887; Fax: +49-6221-56-4233; E-mail: per.humpert@ this intervention and studied effects on different American Academy of Pain Medicine 1526-2375/08/$15.00/** **–** Baseline characteristics of the treated patients baseline, before the second and eighth treatment session. The duration of treatment effects was estimated by the participants before the eighthsession. Eighty-one participants completed the 4-week protocol while 11 patients were lost to External Electric Muscle Stimulation Patients were treated with EMS for 60 minutes, twice a week, and for 4 weeks (eight treatment sessions). EMS was performed using the HiToP® Germany). As previously described [1], a 20-Hz frequency scan of carrier frequencies between 4,096 Hz and 32,768 Hz was used to generate a deep and comfortable muscle contraction. This application was repetitively modulated with 3 Treatment response individual judgment (%) seconds rest time, 3 seconds rise time, and 3 seconds contraction time. The intensity of the electrical stimulation was adjusted to a tolerable level causing muscle contraction and avoiding pain Statistical ProceduresLogistic regression models were calculated to detect possible influence factors on treatment One hundred type 2 diabetes patients with symp- response. In these models, patients were divided tomatic diabetic neuropathy were included in this into responders and non-responders according to study after giving written informed consent. As a self-report of improvement of symptoms and in consequence of a lack in suitable placebo treat- patients completing the 8-week protocol accord- ments, the study was performed in a prospective ing to a minimum of 30% mean decrease in symp- uncontrolled design. The study protocol was toms according to the graded symptom scores approved by the University of Heidelberg Ethics Committee. Patients were eligible if they reported A total symptom score (TSS) was calculated by any symptoms of diabetic neuropathy (i.e., pain, addition of the individual scores on the numeric paresthesia, burning sensations, numbness, sleep- scales, leading to a maximum score of 50. To study ing disturbance). Patients with implanted pace- influences of EMS on the specific symptoms, makers or defibrillators were excluded from the patients were divided into tertiles of the TSS and study. Eight patients were excluded from the analy- the respective symptoms (i.e., paresthesia, pain, ses as other causes of neuropathy or symptoms burning sensation, numbness, sleeping distur- could not be excluded (i.e., excessive alcohol con- bance). Tertiles of patients and the respective sumption, peripheral artery disease and ischemia, symptom scores for equally powered groups were chemotherapy, amyloidosis). The detailed charac- as follows: paresthesia (1–4/5/6–10), pain (1–3/4– teristics of the patients included are shown in 6/7–10), burning sensation (1–3/4–6/7–10), sleep- ing disturbance (1–3/4–5/6–10), and numbness(1–3/4–6/7–10). Treatment effects were studied in the entire group and in patients in the highest At baseline, diabetic neuropathy was graded using tertile of the respective score. Significant differ- the established neuropathy disability scores (NDS) ences in symptom scores between the follow-up and neuropathy symptom scores (NSS) [4]. In measurements were determined using two-tailed addition, patients had to grade their symptoms paired t-tests compared with the baseline. All sta- on 10-point numeric scales (1 = no symptoms, tistical analyses were performed using SPSS 10 = worst possible) as previously reported [1] at Electric Muscle Stimulation for Symptomatic Neuropathy Logistic multivariate analyses of variables Sixty-seven (~73%) of the 92 participants reported subjective improvement of neuropathic symptoms.
Eighty-one patients completed the protocol, while 11 patients did not continue after the fourth treat- ment session; all of these patients reported no treatment response. Forty-seven percent of the participants completing the protocol had an Ն30% on the graded symptom scales. Although patients were not remunerated for participation or * P = 0.02; ** P = 0.08.
travel expenses, adherence to the protocol was Response to treatment is judged as reported by the individual patient (left 100% in the patients reporting improvement of column) and in patients with a Ն30% reduction of mean symptoms on the feetafter 4 weeks of treatment (right column, N = 81).
symptoms. The mean duration of symptomatic Treatment response is the dependent variable and coded numerically as relief was 31 Ϯ 21 hours; the maximum duration 1 = non-responder and 2 = responder.
reported was 80 hours. Fifty-four patients (~59%)were previously or currently treated with medica- P = 0.05) and highly significant on visit 8 tion for neuropathic symptoms and participants (26.0 Ϯ 10.4 vs 18.2 Ϯ 10.4, P < 0.001). In self- currently treated continued medication; the reported non-responders, TSS did not change sig- number of patients with an improvement of the nificantly on visit 2 (26.3 Ϯ 10.3 vs 26.0 Ϯ 10, mean symptom score by Ն30% was similar in P = ns) and visit 8 (24.2 Ϯ 11.2 vs 23.5 Ϯ 9.7, these patients compared with previously untreated P = ns). When all patients were divided into ter- participants (41% vs 51%, P = ns). The only side tiles of TSS, the second and third tertiles showed effect of EMS reported was mild muscle soreness improvements of TSS on visits 2 and 8, while the in the thighs on the day after treatment.
lower tertile showed significant improvement of TSS on visit 8 only (Figure 1a). In self-reported responders were studied at baseline to identify responders, all three tertiles of TSS showed sig- possible influence factors on treatment responses.
nificant improvements of TSS on visit 8 only Responders according to self-judgment had more intense symptoms as given by the NSS scores Changes in specific symptoms were studied in (7.8 Ϯ 1.2 vs 7.2 Ϯ 1.5, P = 0.04), and there was a all participants followed up to the eighth visit and trend toward older age in non-responders (65 Ϯ 8 patients in the upper tertile of the respective vs 69 Ϯ 8, P = 0.07). When patients were divided symptom score on visits 2 and 8 (Figure 2, into responders and non-responders by a mini- mum of 30% decrease in mean symptom score as was previously suggested [5], there were no signifi- (Figure 2c), burning sensations (Figure 2d), sleep- cant differences in the baseline characteristics (not ing disturbances (Figure 2e), and even numbness shown). Logistic multivariate models including (Figure 2f) improved significantly on visit 2 and biometrical data and the classical risk factors visit 8. The treatment effects were strongest on revealed the NSS to be the only variable indepen- visit 8 for burning sensations (8.5 Ϯ 1.2 vs 4.9 Ϯ 2.5, P < 0.001, Table 3) and sleeping distur- response (b = -0.47, P = 0.02, Table 2). In partici- bances (7.9 Ϯ 1.4 vs 4.6 Ϯ 2.8, P < 0.001, Table 3) pants with a minimum decrease of 30% in mean as documented on the 10-point scale. The relative symptom score after 4 weeks of treatment, there decrease in severity of symptoms for these scores was only a trend for an association with the NSS When all participants in this study were ana- lyzed, the TSS significantly improved on visit 8 Discussion
(25.7 Ϯ 10.5 vs 19.2 Ϯ 10.4, P < 0.001), while thechange in TSS on visit 2 was not significant This is the first study on treatment effects of EMS (26.0 Ϯ 10.3 vs 24.8 Ϯ 10.6, P = ns). In the self- in a large group of type 2 diabetes patients. The reported responders, TSS improved marginally data show significant improvement of total symp- significant on visit 2 (25.9 Ϯ 10.4 vs 24.4 Ϯ 10.7, toms and each specific quality of symptoms. Sig- relevant improvement, which was previously shown in a meta-analysis of studies comparing All Participants
placebo-controlled data on pain reduction detected using numerical scales and in patients with different Multivariate analyses revealed that response to treatment was independently and positively associ-ated with the NSS, but not with the NDS. This suggests, that type 2 diabetes patients at all ages and stages of disease might benefit from EMS. Addi-tional metabolic factors such as BMI or even HbA1c, as a marker of long-term glucose control, did not influence the response rate significantly (Table 2). There were no significant differences in response rates between patients previously treated with pharmacological interventions and treatment- naive participants. This indicates that EMS might even be effective in patients that are not sufficiently Baseline
treated using conventional medications such as Responder
anticonvulsants and antidepressants.
Pain and paresthesia in our study were reduced by ~31% and ~35%, respectively, effects that canalso be considered clinically relevant [6]. It is likely that a significant part of the treatment effects canbe attributed to placebo effects; the reduction of numbness was calculated at ~24% and is likely torepresent the placebo effect in our study. Yet defi- mptom S 25
nite differentiation of placebo and treatment effects need to be clarified in future placebo- Our data support the previously published high response rates of EMS in pilot studies and patients with symptomatic diabetic and uremic neuropathy [1,7] and further define the duration of treatment effects as well as specific symptoms that can be Baseline
treated efficiently. The mean duration of the treat-ment effect was limited to 31 hours and varied Figure 1 (A) Total symptom score (TSS) in all participants
strongly between participants. The underlying of the study by tertile of TSS. Patients in the upper and reasons for this large variation remain unclear, and middle tertile of TSS showed significant reductions of symp-toms on the second and the eighth visit. Patients in the it seems important to decipher the physiological lower tertile showed significant reductions on visit 8 only.
changes induced by EMS to understand this * P < 0.05, ** P < 0.01, *** P < 0.001 vs TSS at baseline. (B) finding. Although not an endpoint of this study, it When responders were analyzed separately, highly signifi- seems likely that for example the significant reduc- cant reductions in total symptoms were observed on visit 8 tion of sleeping disorders reported by the partici- only. This was true for type 2 diabetes patients in all threetertiles of the TSS.
pants leads to a major improvement in quality oflife. This improvement and the lack of severeadverse events seem to be reflected by the strong nificant improvement was found in all tertiles of the adherence to the study protocol (100% in self- TSS; however, the treatment effects were most reported responders) over a period of 4 weeks and pronounced in the upper tertile of patients affected eight treatment sessions. Future controlled studies by burning sensations and sleeping disturbances will be needed to study cost-effectiveness and after 4 weeks of treatment. In these patients, symp- clinical efficacy of EMS in comparison to pharma- toms were reduced by ~42%. This reduction in ceutical interventions, especially in consideration symptoms can be considered a strong and clinically of missing relevant side effects. In addition, study Electric Muscle Stimulation for Symptomatic Neuropathy Mean symptom score
Paresthesia
mptom score
mptom score
Change in
Change in
Burning sensations
mptom score
mptom score
Change in
Change in
sleeping disturbance
Numbness
ptom score -2
mptom score
Change in
Change in
Figure 2 Treatment effects of external muscle stimulation (EMS) on symptoms of diabetic neuropathy were analyzed in
participants in the upper tertile of the respective scores to study patients that are significantly affected by the specific
symptom. (A) Mean symptom score, (B) paresthesia, (C) pain, (D) burning sensation, (E) sleeping disturbances, and (F)
even numbness were significantly improved on visit 2 and visit 8. Treatment effects were strongest for the reduction of
burning sensations and sleeping disturbances (D, E). * P < 0.05, ** P < 0.01, *** P < 0.001 vs the respective symptom score
at baseline. Data are given as the mean change in symptom score Ϯ SD.
Symptom scores at baseline and on the eighth visit for all patients followed up (N = 81) and patients in the upper tertile of the respective baseline symptom score Patients in the Upper Tertile of Symptom Scores * P < 0.01; ** P < 0.001 as given by paired t-test.
(N) is the number of patients in the upper tertile of the respective baseline symptom score that were followed up to visit 8.
Data are given as mean Ϯ standard deviation.
protocols will have to include measures of depres- Acknowledgments
sion as additional endpoint which is known to have Parts of this study were supported by the Lautenschläger important associations with symptomatic neur- Diabetes Foundation (PMH, PPN) and the Juvenile Dia- betes Research Foundation (AB, PPN). gbo Medizintech- In the lack of a control group, numbers needed nik AG supplied the treatment devices for EMS and to treat can be estimated utilizing placebo data from previously published studies on tricyclic antide-pressants and traditional anticonvulsants in which amoderate relief of pain by 30% was considered as References
the cut-off for treatment response [5]. In these 1 Reichstein L, Labrenz S, Ziegler D, Martin S.
published studies [9–12], 4 out of 81 patients on Effective treatment of symptomatic diabetic poly- placebo treatment experienced a moderate Ն30% neuropathy by high-frequency external muscle symptom relief compared with 38 out of the 81 stimulation. Diabetologia 2005;48:824–8.
2 Kumar D, Marshall HJ. Diabetic peripheral neur- patients treated with EMS in this study. This would opathy: Amelioration of pain with transcutaneous correspond to a number needed to treat of 2.4 in electrostimulation. Diabetes Care 1997;20:1702–5.
our study and argue for EMS being a very effective 3 Watkins ES, Koeze TH. Spinal cord stimulation treatment option. However, these considerations remain speculative and treatment effects need to be 4 Young MJ, Boulton AJ, MacLeod AF, Williams DR, clarified in future placebo-controlled trials. A spe- Sonksen PH. A multicentre study of the prevalence cific placebo device that causes superficial skin of diabetic peripheral neuropathy in the United stimulation without influences on muscle and Kingdom hospital clinic population. Diabetologia nerves was not available for this study and will have 5 Wong MC, Chung JW, Wong TK. Effects of treat- In conclusion, EMS was shown to be of some ments for symptoms of painful diabetic neuropathy:Systematic review. BMJ 2007;335:87.
benefit in ~40–70% of the type 2 diabetes patients 6 Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, treated depending on the definition of treatment Poole RM. Clinical importance of changes in response (i.e., Ն30% reduction in mean symptom chronic pain intensity measured on an 11-point score or self-reported), is free of significant side numerical pain rating scale. Pain 2001;94:149– effects, and led to an accentuated improvement in burning sensations and sleeping disorders. Forty- 7 Klassen A, Di Iorio B, Guastaferro P, et al. High- one percent of patients currently or previously tone external muscle stimulation in end-stage renal treated with pain medication had an improvement disease: Effects on symptomatic diabetic and uremic of mean symptoms by Ն30%. This data have to be peripheral neuropathy. J Ren Nutr 2008;18:46– judged in view of the results obtained in clinical trials of pharmaceutical interventions that resulted 8 Vileikyte L, Leventhal H, Gonzalez JS, et al. Dia- betic peripheral neuropathy and depressive symp- in numbers needed to treat of ~2–4 and led to toms: The association revisited. Diabetes Care significant side effects [13–16]. This study was performed in an uncontrolled design. Future 9 Kochar DK, Jain N, Agarwal RP, et al. Sodium val- controlled trials including pharmaceutical inter- proate in the management of painful neuropathy in ventions will be needed to clearly define the clini- type 2 diabetes—A randomized placebo controlled cal and economic efficacy of this treatment option.
study. Acta Neurol Scand 2002;106:248–52.
Electric Muscle Stimulation for Symptomatic Neuropathy 10 Max MB, Culnane M, Schafer SC, et al. Amitrip- 13 Saarto T, Wiffen PJ. Antidepressants for neuro- tyline relieves diabetic neuropathy pain in patients pathic pain. Cochrane Database Syst Rev 2005: with normal or depressed mood. Neurology 1987; 14 Wiffen P, Collins S, McQuay H, et al. Anticonvul- 11 Max MB, Kishore-Kumar R, Schafer SC, et al. Effi- sant drugs for acute and chronic pain. Cochrane cacy of desipramine in painful diabetic neuropathy: A placebo-controlled trial. Pain 1991;45:3–9, dis- 15 Wiffen PJ, McQuay HJ, Edwards JE, Moore RA.
Gabapentin for acute and chronic pain. Cochrane 12 Rull JA, Quibrera R, Gonzalez-Millan H, Lozano Castaneda O. Symptomatic treatment of peripheral 16 Wiffen PJ, McQuay HJ, Moore RA. Carbamazepine diabetic neuropathy with carbamazepine (Tegretol): for acute and chronic pain. Cochrane Database Syst Double blind crossover trial. Diabetologia 1969;5:

Source: http://www.dr-med-may.de/download/hochtontherapie/Heidelberg-Studie-2008.pdf