Ppm_june07_kirsch_reprint.qxp

Electromedicine
CES in the Treatment of Depression, Part 2
This second article, of a two-part series on the efficacy of Cranial Electrotherapy
Stimulation (CES) in treating depression, reviews the results of meta-analysis
conducted on CES studies.
By Daniel L. Kirsch, PhD, DAAPM, FAIS and Marshall F. Gilula, MD Cranial electrotherapy stimulation them with three, 30 minute CES sessions in an open clinical trial.
This did not reduce depressive symptoms so stimulation was in- creased to a minimum of five, 30 minute sessions. At this level, vices using microcurrent levels of electri- measurable changes began to be seen. CES is now routinely pre- scribed for depression for a minimum of three weeks of daily treatment followed by a reduced schedule or for use on an as- Some of the early studies are suited to meta-analysis in that ing tiny electrical currents—similar to scores on pre- and post-testing were given, using psychometric instruments such as the Zung Self Rating Depression Scale.
Often other information was added such as the percent of pa- tients who improved at least 50%, those who did not improve at form, moves electrons through the brain at a variety of frequen- all, and those who demonstrated signs of other responses. That cies collectively known as harmonic resonance. This normalizes form of data is a poor fit with meta-analysis.
the electrical activity of the brain as measured by an electroen- Following the often dramatic results published by Rosenthal, cephalogram (EEG). The patient undergoing CES treatment will other researchers began studying psychiatric inpatients. Such often report a pleasant, relaxed feeling of well-being. Improve- patients were all heavily medicated as well. Accordingly, CES ment is usually experienced during treatment, but may be seen would be required to demonstrate effects over and above the ef- hours later, or even the day after treatment. Depression control fects of medications, and that was usually found.
is often experienced after two to three weeks of daily treatment.
Ear clip electrodes, moistened with an appropriate conducting Complications in Some CES Test Protocols
solution, are applied for 20 minutes to an hour or more on an Feighner was one of the early CES research pioneers working initial daily basis for 3-6 weeks, followed by a reduced schedule with psychiatric inpatients. He ran into two complications: pa- of 2 or 3 treatments a week until the depression is resolved, and tients were heavily medicated, and a crossover design was uti- then further reduced to an as-needed (p.r.n.) basis.
lized. Adding to that, the patients were treated 30 minutes daily This article focuses on the meta-analysis of CES studies of de- for only five days.5 Only pre-crossover scores from that study are pression along with a discussion of individual study designs and included in the meta-analysis since they are the only statistical- outcomes. It is important to note that protocols for some CES ly legitimate data as explained in the discussion of meta-analy- studies were poorly designed; inconsistent patient selection and sis in the previous issue of Practical Pain Management.6 concurrent use of other pharmaceutical modalities rendered the Marshall subsequently studied inpatients in a state hospital results inconclusive with regard to CES efficacy in treating de- by, once again, providing CES for 30 minutes a day for only five pression. The authors have carefully selected only valid studies days. Here again, both the treatment and control subjects were to provide the most complete and accurate meta-analysis of CES heavily medicated. The study became invalid when the controls showed a substantial improvement in their depression duringthe course of the study, and there was no control group left Early CES Studies In Treating Depression
against whom the treatment effect of CES could be measured.7 Rosenthal conducted some of the earliest CES studies of depres- CES studies that were invalidated due to a loss of controls were sion when CES was first introduced in the U.S. His work was often published, including the Marshall study and one by Passi- primarily with psychiatric outpatients, although he sometimes ni, who repeated a similar protocol in 1976 with an inpatient sam- used medical staff as controls. He was basically trying to find out ple that included a wide variety of diagnoses including addiction what, if anything, CES treatment would do for his patients, how and psychosis. All patients also received psychotropic medication, many sessions it might require, and what level of current it took and the study showed an improvement in depression following 14 days of daily CES for 30 minutes along with medications. This The early U.S protocols studied patients who had been refrac- study had no controls with which to measure treatment effects and tive to previous antidepressant treatment, but only provided was thereby invalidated, but that did not prevent it from being 2007 PPM Communications, Inc. Reprinted with permission.
published. The authors concluded that since all of the sham treat- assessed on the National Institutes of Mental Health (NIMH) ed patients improved, improvement from CES treatment could Self Rating Scale and obtained a 73% reduction in depression only be attributed to the placebo effect.8 among the treated patients compared with a 21% improvement Levitt studied six male and seven female psychiatric inpatients, divided into two groups, with diagnoses of schizophrenia, alco- Shealy studied depression in chronic pain patients and con- holism, psychotic depression, mixed neurosis and personality dis- trols. He found that CES therapy yielded a 60% improvement in orders. Two of the CES devices malfunctioned, and this reduced their depression score and there was a significant elevation in his treated group to five who received treatment 30 minutes a serotonin (mean of 33.18 ± 9.33 pre-test to 44.64 ± 9.10 post day for ten sessions over a two week period. They were all on psy- test, P<.0089), and a significant decrease in cholinesterase (mean chotropic and sleep medications. Some sham treated patients im- of 13.82 ± 2.86 pre-test to 10.45 ± 3.04 post test, P<.0067).15 proved as did some CES patients, and some saw their conditions Lichtbroun and, later, Tyers began a series of studies of de- worsen. This was essentially a negative outcome from the point pression in fibromyalgia patients.16-18 They measured depression of view of CES treatment effectiveness. It should be noted how- with the POMS test and found that as the patients’ pain scores ever, that in addition to the medications, Levitt was using an early improved, so did their depression which showed between 26% style of CES electrodes in which saline soaked gauze pads to 35% improvement after three weeks of daily, 60 minute CES wrapped around thin steel plates were placed tightly over the closed eyes. This provided undesirable visual effects such as Two other groups of depressed subjects studied were gradu- blurred vision.9 Subjects in both groups reported the temporary ate students in a business school suffering under the stress of visual disturbances which were later judged to be caused by me- completing an MBA program and patients suffering from life- chanical pressure on the eyes, not the electrical intervention it- time disability due to closed head injuries.19,20 Both were double- self. This electrode method was abandoned over 30 years ago. blind studies in which CES or sham CES was given for one hourdaily, Monday through Thursday for three weeks in the closedhead injured subjects and one hour daily for 21 days in the grad- “…with an effect size of r =.50, CES is much more uate students. The closed head injured subjects achieved a 30% effective than any antidepressant medication, and, improvement in their depression while the graduate studentsimproved 34%.
unlike them, lacks significant adverse effects.” A group of 28 children and adults with attention deficit dis- order (ADD) were studied in an open clinical protocol that Improved Study Methodology
looked at various factors, including depression scales. They were In 1975, a depression study was conducted with 72 inpatient al- given 45 minute daily CES treatments for three weeks. They were coholics who were provided 15 daily CES treatments, 40 minutes retested at 18 months follow up. Their depression improved by a day at a current level just below sensory threshold (sub-senso- 32% at the conclusion of the study, and was maintained at that ry). This length of treatment was chosen because of some of Rosenthal’s earlier difficulties, and because most of the patients Physicians evaluated 500 patients who were treated with CES, were also taking psychotropic medications. Their depression, as 69 of whom carried a primary depression diagnosis with the bal- measured on the Profile of Mood States (POMS) improved 76% ance having comorbid depression. The group improved an av- while the sham-treated patients continued to worsen.10 erage of 71% over varying courses of treatment.22 Another study Krupitsky’s group at Yale studied affective disturbance in 20 examined patients’ own self rating of improvement. This infor- alcoholic patients in 1991, and found an average of 28% im- mation was obtained from surveys of 318 patients who had been provement on two depression measures. They concluded that diagnosed with depression and who had used their Alpha-Stim CES was an effective non-pharmacological method to treat af- CES device for at least three weeks prior to sending in the sur- fective disturbances in alcoholic patients in remission.11 vey. They rated their improvement an average of 58% on a 100 May also studied inpatient addicts and found that 60 minutes point scale. While 12% rated their improvement less than 25%, of CES for 25 days allowed the patients to attain an unprece- more than twice as many (27%) rated their improvement be- dented recovery record, with the group of 14 patients improv- ing 76% on the Multiple Affect Adjective Check List and im- Two additional crossover studies were done, one in which sham proving 77% on the Beck Depression Inventory.12 This was the treated patients were actually given CES at a low current while longest treatment time studied to date. Although based on a suc- treated patients were provided CES at a higher current, after cessful pilot study, the U.S. government has now funded a six which they were crossed over.24 No improvement was noted in month CES treatment study for veterans with spinal cord in- the study, although there were protocol design flaws. Another double-blind crossover design, in which five 30 minute treat- The final addiction study to date that included data on de- ments were given, provided results in the patients prior to the pression was a doctoral dissertation in 1994 by Bianco who stud- crossover, allowing its use in meta-analysis.25 ied 65 inpatient poly-substance abusers. He provided 45 min-utes of CES daily from 6 to 14 days, and found their improve- Studies Collected For Meta-Analysis
ment on the Beck Depression Inventory to be 80%.13 Table 1 presents all of the studies collected for meta-analysis.
Hearst studied 28 psychiatric outpatients who were on less There were a total of 23 suitable CES studies of depression, rep- medication than a typical inpatient sample. Because the study resenting some 1,075 subjects studied. It bears emphasis in this took place in 1974, the early protocol of five, 30 minute treat- day of black box warnings on SSRI’s that none of the CES stud- ment or sham treatment sessions was followed. The patients were ies found any significant negative side effects. CES is known to 2007 PPM Communications, Inc. Reprinted with permission.
TABLE 1. LIST AND DESCRIPTION OF DEPRESSION STUDIES
Blinding
Primary Diagnosis
Therapist
Assessor
Study Design
OutcomeMeasure
a Zung’s Self Rating Depression Scaleb Profile of Mood Statesc Multiple Affect Adjective Check Listd Montgomery and Asberg Depression Rating Scale produce skin irritation at the electrode site in people with light jects to use for subsequent crossover sham treatment. That leaves skin and may cause an occasional headache. Such side effects 20 studies involving 937 subjects that are considered valid for Table 2 presents the studies shown in Table 1, with three of the studies removed. In one, the study was invalidated when the Secondary Analysis of Studies
sham treated patients also improved.8 The other two studies had Some studies reported more than one measure of depression.
crossover designs, and the investigators did not report the treat- Feighner reported two measures, as did Krupitsky, May and ment results prior to the crossover. The sham patients in a Rosenthal (in three different studies), while Moore reported crossover design who had active CES during the initial arm of three. In order to limit the input of error variance from any given the study typically continue to improve, making them unfit sub- study, each study was represented with only one score, and to be 2007 PPM Communications, Inc. Reprinted with permission.
TABLE 2. AN INITIAL ANALYSIS OF STUDIES SHOWN IN TABLE 1
Number of Patients
Controls
Statistic Reported
Zr Scorea
a From Fisher Tables of r to zr transformation26 b Percent change equals r, from the binomial effect size distribution. From Wolf27 equitable, means of all the scores given were computed and uti- size obtained is more than sufficient to show that CES is a very ef- lized. Since percentages can not be legitimately averaged, they fective treatment for depression. In fact, with an effect size of r were converted to Zr scores and then those scores were aver- =.50, CES is much more effective than any antidepressant med- aged. The mean Zr score was then converted back into a per- ication, and, unlike them, lacks significant adverse effects.6,29 cent score. The results of this for CES in the treatment of de- To estimate the outer limits of the effect size to be expected in any future meta-analyses of studies of CES for depression, the The effect size from the 20 studies analyzed is r =.50, which is confidence interval of the effect size needs to be derived. That considered a strong effect size. While there was a wide disparity is calculated from the standard deviation, divided by the square of number of subjects in the various studies, an N weighted effect root of the number of studies in the analysis, yielding the stan- size of r =.51 was obtained, showing that the number of people dard error of the mean. The resulting score indicates that if 15 appearing in a given study was relatively unimportant. The effect additional meta-analyses of 21 studies each is performed in the 2007 PPM Communications, Inc. Reprinted with permission.
TABLE 3. A SECONDARY ANALYSIS OF STUDIES SHOWN IN TABLE 1
depression scores of the controls in eachstudy. Table 4 presents an abbreviated ver- Number of Patients
sion of those results, from which the ef-fect size for fluoxetine treatment of de- Controls
Total Statistic Reported Results
Zr Scorea
group by the change in the treated group,then subtracted that score from 100 to get over and above that of the placebo pa-tients.
fect of fluoxetine over and above that of the placebo patients in the five studies wasonly 8%. In study 25, the placebo group ed patients. That figure rose to 11% whenthe studies were corrected (weighted) for sample size. Kirsch also evaluated otherantidepressant drug studies that were sent searchers added subjects to their studies based on the Hamilton Depression Scale,a psychometric paper and pencil test that either the patient or the researcher cancomplete. Tests with similar validity and reliability were used in the CES studies.
So from the standpoint of enlisting sub-jects, identical or similar diagnostic de- vices were used for both the pharmaceu-tical and CES research. would not be expected to improve as dra-matically in the placebo condition as did the placebo patients in all the pharma-ceutical studies reported to the FDA. Or to put it another way, one would not an-ticipate that 89% of placebo patients with weeks or less as they did in the fluoxetine studies. Most were recruited by advertise- Effect Size Confidence Limits, p<.01=.32-.68 ments in newspapers and other media.
a From Fisher Tables of r to zr transformation26 b Percent change equals r, from the binomial effect size distribution. From Wolf27 c From Rosenthal28d The first two columns do not add to this figure due to subjects in the crossover studies appearing twice future (more than 300 additional studies), there is a 99% likelihood that the effect care (i.e., antidepressant medications).
size obtained will fall within an effect size were either more serious cases of depres- Comparison To Efficacy of
proval of fluoxetine (Prozac).30 Five stud- cals were typically studied over a longer Antidepressant Fluoxetine
ies were submitted, which Kirsch analyzed to determine how the effectiveness of CES pression scores experienced by the treat- 2007 PPM Communications, Inc. Reprinted with permission.
Summary of CES Modality For Depression
TABLE 4. EFFECTIVENESS OF FLUOXETINE OVER AND ABOVE
The following presents a brief synopsis of the discussion in parts1 and 2 of this series.
PLACEBO EFFECT IN TREATING DEPRESSION
• Cranial Electrotherapy Stimulation (CES) can occasionally be a single, time-limited treatment of many mild depres- sions with or without concomitant medication.
• Meta-Analysis is a valid way to assess the effect size of • Meta-Analysis of effect size has shown that CES, with and without concomitant medication, compares very favorably with the effect size of medication treatment.
• Although CES is nearly free of significant adverse effects, there is a spectrum of usually mild cutaneous irritativeeffects at the electrode site which can limit treatment • Depressive disorders require competent medical evalua- tion to rule out a primary or comorbid substance-related effect or a primary or comorbid treatable medical illness.
• Neither CES nor antidepressants should be employed for treatment without continuing and competent healthcaresupervision because of emerging suicidality as some Conclusion
Regardless of the manner in which one analyzes CES studies of • CES should always be considered as an add-on to medica- depression, a moderate to strong effect size is revealed, which tions before considering the more invasive Vagal Nerve exceeds the results of antidepressant drug studies submitted to Stimulator (VNS) or Deep Brain Stimulation (DBS) the FDA for marketing approval (see Table 4). because it is much cheaper and potentially as efficacious, With moderately severe and severe depressions, CES should definitely be considered as an add-on modality because of the 2007 PPM Communications, Inc. Reprinted with permission.
potential for (a) synergizing the efficacy tionale Ärztegesellschaft für Energiemedizin, 10. Smith RB and O’Neill L. Electrosleep in the man-agement of alcoholism. Biological Psychiatry. 1975.
of the drug (s), and (b) reducing the over- Austria 2000, in German. Best known for de- all adverse effects of psychopharmaceuti- signing the Alpha-Stim CES and MET line of 11. Krupitsky EM, Burakov AM and Karandashova GF medical devices, Dr. Kirsch is Chairman of et al. The administration of transcranial electric treat- Electromedical Products International, Inc. of ment for affective disturbances therapy in alcoholic pa-tients. Drug and Alcohol Dependence. 1991. 27:1-6.
Mineral Wells, Texas, USA with additional of- 12. May B and May C. Pilot project using the Alpha- fices in Europe and Asia. Dr. Kirsch can be Stim 100 for drug and alcohol abuse. In: Kirsch, tabolizing SSRI’s, other antidepressants, Daniel L. (Ed) The Science Behind Cranial Elec-trotherapy Stimulation. Edmonton, Alberta, Canada.
Marshall F. Gilula, M.D. is a Diplomate of Medical Scope Publishing. 2002. p 51.
the American Board of Psychiatry and Neurol- 13. Bianco F. The efficacy of cranial electrotherapy It’s important to stress that adjunctive ogy and a Diplomate of the American Board of stimulation (CES) for the relief of anxiety and depres- Medical Electroencephalography. He is also a sion among polysubstance abusers in chemical de-pendency treatment. Ph.D. dissertation, The Universi- drug can often prevent the need for using board-certified Instructor in Biofeedback and Neurotherapy (NBCB). In 1978 he was a US- 14. Hearst ED, Cloninger CR, Crews EL and Cadoret quently the case in the currently accept- USSR NIMH Exchange Scientist working with RJ. Electrosleep therapy: a double-blind trial.
Archives of General Psychiatry. 1974. 30(4):463-466.
cranial electrotherapy stimulation and gener- 15. Shealy CN, Cady RK, Wilkie RG, Cox R, et Al. De- can prove increasingly cost-effective com- al psychophysiology techniques at the P.K. pression: a diagnostic, neurochemical profile and Anokhin Institute, Soviet Academy of Medical therapy with cranial electrical stimulation (CES). Jour-nal of Neurological and Orthopaedic Medicine and SSRI’s. CES can be an ideal treatment for Sciences, Moscow. In 1983 Dr. Gilula was the Surgery. 1989. 10(4):319-321.
first Motoyama-Ben Tov Fellow at the Institute 16. Lichtbroun AS, Raicer MMC and Smith RB. The of Life Physics, Tokyo (Mitaka-shi), Japan and treatment of fibromyalgia with cranial electrotherapy sizes having the patient take the initiative researched neuroelectric methodology and the stimulation. Journal of Clinical Rheumatology. 2001.
7(2):72-78.
on a daily basis. This involves several be- EEG of altered states with Professor Hiroshi 17. Tyers S and Smith RB. A comparison of cranial Motoyama. Dr. Gilula has had four years of electrotherapy stimulation alone or with chiropractic residency and postdoctoral fellowship training therapies in the treatment of fibromyalgia. The Ameri-can Chiropractor. 2001. 23(2), 39-41.
more education of clinicians and their pa- in psychiatry and over seven years of postdoc- 18. Tyers S and Smith RB. Treatment of fibromyalgia tients about the modality—especially due toral training in neurology (neurophysiology with cranial electrotherapy stimulation. The Original to an initial aversion to electric stimula- and epilepsy). He has 40 years of experience tion because of a mental association with in clinical psychiatry, and was in the Depart- 19. Matteson MT and Ivancevich JM. An exploratoryinvestigation of CES as an employee stress manage- “Electric Shock” (Electroconvulsive Ther- ment of Neurology at the University of Miami ment technique. Journal of Health and Human Re- School of Medicine from 1999 through 2003. source Administration. 1986. 9:93-109.
cially because of the less onerous econom- Dr. Gilula was a Senior Fellow, Miami Center 20. Smith RB, Tiberi A and Marshall J. The use of cra- ics involved and the superior safety fac- for Patient Safety, Department of Anesthesiol- nial electrotherapy stimulation in the treatment ofclosed-head-injured patients. Brain Injury. 1994.
ogy, University of Miami from 2003 through 2005. Dr. Gilula is President and CEO of the 21. Smith RB. Cranial electrotherapy stimulation in Life Energies Research Institute in Miami. He the treatment of stress related cognitive dysfunction,with an eighteen month follow up. Journal of Cogni- tions of the Vagal Nerve Stimulator (VNS) can be reached at [email protected]. tive Rehabilitation. 1999. 17(6):14-18.
22. Kirsch DL. Postmarketing survey of Alpha-Stim References
CES patients. In Kirsch, DL, (Ed.) The Science Be- 1. Rosenthal SH, and Wulfsohn, NL. Electrosleep: A hind Cranial Electrotherapy Stimulation. Edmonton, Al- Daniel L. Kirsch, PhD, DAAPM, FAIS is an preliminary communication. Journal of Nervous and berta, Canada. Medical Scope. 2002. p 43.
internationally renowned authority on elec- Mental Disease. 1970. 151(2):146-151.
23. Smith RB. Is microcurrent stimulation effective in tromedicine with 34 years of experience in the 2. Rosenthal SH and Wulfsohn NL. Studies of elec- pain management? An additional perspective. Ameri- trosleep with active and simulated treatment. Current electromedical field. He is a board-certified can Journal of Pain Management. 2001. 11(2):62-66.
Therapeutic Research. 1970. 12(3):126-130.
Diplomate of the American Academy of Pain 24. Frankel BL, Buchbinder R and Snyder F. Ineffec- 3. Rosenthal SH and Wulfson NL. Electrosleep: A clini- tiveness of electrosleep in chronic primary insomnia.
Management, Fellow of the American Institute cal trial. American J of Psychiatry 1970. 127(4):175-176.
Archives of General Psychiatry. 1973. 29:563-568.
of Stress, Member of the International Society 4. Rosenthal SH. Electrosleep: A double-blind clinical 25. Moore JA, Mellor CS, Standage KF and Strong of Neuronal Regulation, and a Member of study. Biological Psychiatry. 1972. 4(2):179-185.
HA. A double-blind study of electrosleep for anxiety 5. Feighner JP, Brown SL and Olivier JE. Electrosleep and insomnia. Biological Psychiatry. 1975. 10(1):59-63.
Inter-Pain (an association of pain manage- therapy: A controlled double-blind study. Journal of 26. Edwards AL. Statistical Methods for the Behavioral ment specialists in Germany and Switzerland). Nervous and Mental Disease. 1973. 157(2):121-128.
Sciences. Hold, Rinehart and Winston. NY. 1966. p503.
He served as Clinical Director of The Center 6. Kirsh DL and Gilula MF. CES In The Treatment of 27. Wolf FM. Meta-Analysis; Quantitative Methods for for Pain and Stress-Related Disorders at Co- Depression–Part 1. Pract Pain Mgmt. May 2007. 7(4): Research Synthesis. Sage. Newbury Park. 1986. p 33.
lumbia-Presbyterian Medical Center, New York 28. Rosenthal R. Meta-analytic Procedures for Social 7. Marshall AG and Izard CE. Cerebral electrothera- Research. Newbury Park:Sage. 1991. pp 70-72.
City, and of The Sports Medicine Group, Santa peutic treatment of depressions. Journal of Consult- 29. Gilula MF and Kirsch DL. Cranial electrotherapy Monica, California. Dr. Kirsch is the author of ing and Clinical Psychology. 1974. 42(1):93-97.
stimulation review: a safer alternative to psychophar- two books on CES titled, The Science Behind 8. Passini FG, Watson CG and Herder J. The effects of maceuticals in the treatment of depression. Journal of cerebral electric therapy (electrosleep) on anxiety, de- Cranial Electrotherapy Stimulation, 2nd Ed. pression and hostility in psychiatric patients. Journal of 30. Kirsch I, Moore TJ, Scorboria A and Nicholls SS.
published by Medical Scope Publishing Corpo- Nervous and Mental Disease. 1976. 163(4):263-266.
The emperor’s new drugs: An analysis of antidepres- ration, Edmonton, Alberta, Canada in 2002; 9. Levitt EA, James NM and Flavell P. A clinical trial of sant medication data submitted to the FDA. Preven- and Schmerzen lindern ohne Chemie CES, die electrosleep therapy with a psychiatric inpatient sam- tion and Treatment. 2002. www.journals.apa.org/pre- ple. Australian and New Zealand Journal of Psychiatry.
vention/volume5/pre0050023a.html Accessed May Revolution in der Schmerztherapie, Interna- 2007 PPM Communications, Inc. Reprinted with permission.

Source: http://www.cesultra.cn/pdf/CES%E5%9C%A8%E6%8A%91%E9%83%81%E7%97%87%E6%B2%BB%E7%96%97%E4%B8%AD%E7%9A%84%E5%BA%94%E7%94%A8(EN).pdf