Review of Acupuncture for Schizophrenia


Review of Acupuncture for Schizophrenia


There is a paucity of scientific data on the use of acupuncture or low-power laser treatment in schizophrenia. The authors have conducted a literature search, identifying eight papers on the subject: all were significantly flawed, most seriously so. The papers suggest that acupuncture and lowpower laser treatment may be as effective as chlorpromazine in schizophrenia. However no scientifically sound conclusions can be drawn. Better designed research is required before acupuncture can be considered a suitable treatment for schizophrenia.

Key words

Acupuncture, Chlorpromazine, Low-power laser, Mental illness, Schizophrenia.


Since the resurgence of traditional Chinese medicine (TCM) under Chairman Mao, some Chinese hospitals use acupuncture and herbs alongside antipsychotic medication for the treatment of psychotic disorders. However, the extent of usage and effectiveness of such practice is unclear.


We conducted a literature search on the electronic databases Medline and Psychlit using the search terms: acupuncture, electroacupuncture, traditional Chinese medicine, schizophrenia, mental illness and psychiatry. In addition, the Research Council for Complementary Medicine provided a similar search from their database, CISCOM.

TCM classification of schizophrenia

Schizophrenia is a Western term of recent origin whose meaning has evolved and continues to be controversial. There is no direct translation into TCM terminology. During this century, traditional Chinese doctors have come to use the term in their own practice and relate it to the following traditional classification. The majority of cases are covered by the TCM term Mania in which "Double yang results in mania-schizophrenia characterised by excitation". Depression-insanity in TCM is close to "Schizophrenia characterised by melancholy and tension caused by double Yin...[including]...heart confused by phlegm...[and]...depression of liver energy." In "Paranoia-insanity...worry harms the spleen as spleenasthesia leading to schizophrenia characterised by terror and worry." ( 1).

TCM treatment of schizophrenia

Acupuncture in TCM is one of a number of approaches to improve a patient's condition. The therapy would also include herbs ( 2), productive labour, self-reliance, community ethos and a spiritual dimension, whether Taoist, Confucian or Maoist ( 3). Modern developments in acupuncture involving electrical and laser stimulation of points have now been incorporated into treatments for schizophrenia ( 2).

Putative mechanism of action

Chronic schizophrenics show a heightened level of arousal. There are Chinese studies showing that acupuncture can influence cortical activation and the responsiveness of the reticular activating system to stimuli ( 4). It is possible that acupuncture could reduce levels of arousal and therefore result in a therapeutic effect.

Neurophysiological studies investigating the analgesic effect of acupuncture have shown that acupuncture stimulates Aë sensory fibres with high intensity, low frequency shocks which may be mechanical or electrical. The analgesic effect may be mediated by modulation of neurotransmitters including 5-Hydroxytryptamine ( 5-HT) and dopamine centrally ( 5). There are a number of studies showing changes in 5-HT production and excretion as a result of acupuncture, but there is not yet a clear consensus of the pattern or mechanism ( 6). Han showed in animal experiments that acupuncture is capable of accelerating the synthesis and release of 5-HT and noradrenaline (NA) in the central nervous system ( 7).

Zhang and Shen ( 8) showed that electric stimulation of acupuncture points Yintang (Extra 1), in the midline between the eyebrows, and Baihui (GV. 20), on the crown of the scalp, produce a significantly increased level of brain 5-HT and NA in rats. In a follow-up study Zhang et al. ( 9) showed that electroacupuncture increases urinary excretion of 3-methoxy-4-hydroxyphenylglycol sulphate in unmedicated schizophrenics. This is the main metabolite of NA.

Research findings

One of the better studies was by Liu et al. ( 10) who performed a randomised, controlled trial. Schizophrenics with auditory hallucinations were allocated to three groups: helium-neon laser irradiation of point Ermen (TE.21) just anterior to the ear, mock-irradiation at that point together with chlorpromazine, and irradiation of a non-standard acupuncture point. They found marked improvement in the Ermen laser group, slightly less improvement in the chlorpromazine group, and no improvement in those laser-stimulated at the non-standard point.

Jia et al. ( 11) used helium-neon laser irradiation of the Yamen (GV. 15) acupoint, at the first cervical interspace, compared with a control group using chlorpromazine. They did not randomise the allocation of cases. They used case definition according to criteria stated in the Diagnostic and Statistical Manual of the American Psychiatric Association ( 12). The Brief Psychiatric Rating Scale ( 13), which is valid and reliable, was used as the measured outcome, scored by blinded, independent doctors. The results showed a significant improvement in both groups, with no statistically significant difference between the two, suggesting acupuncture was as effective as chlorpromazine.

Zhang ( 14) used scalp acupuncture in 296 patients with hallucinations due to multiple causes, mostly schizophrenia. He used points Houding (GV. 19) and Baihui (GV. 20) with auxiliary points Zhengying (GB.17), Muchang (GB.16), Luxi (TE.19), Yifeng (TE.17), Touqiayin (GB.11), Tienzhu (BL.10) and Chengguang (BL.6) with point-through-point needling. This was an uncontrolled study in which 70.6% were considered cured, "complete disappearance without relapse", after 10 to 20 sessions.

Kane and Di Scipio ( 15) reported 3 patients who were used as their own controls with a pseudo-acupuncture technique. Stimulation of Hegu (LI.4), Taichong (LR.3) and Renzhong (GV. 26) alternated weekly with Taiyang (KI.1), Neiguan (PC.6) and Sanyinjiao (SP.6) compared to a control treatment using non-traditional acupuncture points. Blinded assessors used well-designed outcome measures, but there were no generalisable results because of the small number of patients involved. None the less, all three showed a clear improvement while receiving traditional acupuncture, which was lost during control treatment.

Shi and Tan ( 1) looked at 500 cases of schizophrenia treated with acupuncture. Those with mania had reducing manipulation of Dazhui (GV. 14), Qiangjia (GV. 18), Renzhong (GV. 26)and Jiuwei (CV. 15). Those with depressive type schizophrenia had reinforcing manipulation of Juque (CV. 14), and Tangzhong (CV. 17). Those with paranoia type had reinforcing manipulation of Zhongwen (CV. 12), Tangzhong (CV. 1 7), Shenting (GV. 24), Shenmen (HT. 7), Sanyinjiao (SP.6), Neiguan (PC.6) and Gongsun (SP.4). Cure was reported in 55% ("all symptoms disappeared"), 16.6% were "remarkably improved" and 11.6% showed no improvement. This study was flawed: with unclear selection of cases, no controls, and non-blind raters using unreliable measures. In particular there was an unchecked use of chlorpromazine.

Zhang et al. ( 16, 17) performed a study of treatment of schizophrenia which compared: herbs alone, electroacupuncture with herbs, electro-acupuncture alone, and chlorpromazine. He used G7401 electroacupuncture apparatus at a frequency of 120Hz with a 500 micrometer pulse-width to points: Yifeng (TE.17), Tinggong (S1.19), Chengling (GB.18), Toulinqi (GB.15), Baihui (GV. 20). The herb used was Dinaj Gui Cheng Qi Tang 50ml twice daily. The dose of chlorpromazine was 300-600mg per day. The results were: "marked improvement" in 61.6%, 64.0%, 30.4% and 36.0% respectively. The study showed that TCM is "less effective for indifference or apathy." It seems to demonstrate a good efficacy of the herb used, with little additional benefit attributable to acupuncture. This study suffers from selection bias, poor outcome measures and a short study period in which neither acupuncture nor chlorpromazine were given time to be effective before excluding subjects.

Shi ( 18) compared the treatment of hallucination with auricular acupuncture alone, auricular and body acupuncture, and auricular acupuncture with chlorpromazine. The auricular point was either Naodian (external auricle) or Shenmen, and Yangwei (internal auricle). The body points were Tinggong (S1.19), Tinghui (GB.2), Yifeng (TE.1 7), Yanggu (SI.5), Taiyuan (LU.9), Yangxi (LI.5) and Yinbai (SP. 1). The dose of chlorpromazine was <200mg per day. In all cases, the more frequent the attacks and the more long-standing the disease, the less effective the treatment. Out of 120 cases, 58.3% claimed that their hallucinations disappeared, 22.5% showed alleviation of symptoms and 19.2% showed no improvement. There were no differences between treatment groups. This study is of little value since there is no randomisation, no control group and no clear outcome measures.

Esser et al. ( 3) aimed to investigate the idea that acupuncture works by restoring homoeostasis; for example that it would decrease over-arousal and increase under-arousal. They measured physiological variables such as blood pressure and pupil diameter of a group of psychiatric outpatients, mostly schizophrenic, in China. Their method was inadequate, with no controls, no psychological or social measures, non-blind observers and only 17 subjects, with vague inclusion criteria. The only conclusion of note is that the acupuncture produced a heightened sense of well being in all but one of the subjects which lasted for approximately a week after treatment.

Methodological issues

Definition of cases

Most of the studies reviewed have had an unclear definition of the disease. The problem has not been that there is a clash between Western and Eastern classification, but simply that the definitions have been unstated. There are internationally agreed, operational, valid and reliable diagnostic criteria, such as ICD-10, the World Health Organisation's international classification of diseases ( 19), which should form the basis for case definition. Some studies have tested acupuncture treatment for symptoms rather than by broad diagnosis ( 18). This is acceptable, provided the symptoms are well defined and measured.

Selection of cases

How applicable the results of a clinical trial are to the group being studied depends upon the random selection of cases from a representative population, with clear inclusion and exclusion criteria to avoid bias. This has been very poorly done in the studies we have reviewed. For example, there are few attempts at excluding co-morbidity ( 1). Zhang et al. incorporated a selection bias because cases were included only if "They experienced marked improvement or cure at the end of the previous course of treatment." ( 17).


In general the studies use diffuse clinical outcome measures such as "hallucination, pseudohallucination or talking to self" ( 18) or "lucidity was partially recovered", which are not well defined, valid or reliable and are very open to observer bias. The commonly used Chinese definitions of outcome are: cured, markedly improved, improved or no improvement. This would not be a reliable or repeatable measure in a Western setting, especially where the observers are not blind to the treatment allocation. However, differences between these categories are large enough and sufficiently familiar to the Chinese practitioners in their clinical work to have some public meaning, thus these results should not be dismissed.

Placebo controls

According to Western interpretation of traditional Chinese theory, acupuncture should only be effective when performed at particular points. Thus, sham acupuncture ( 20), where needles are inserted at non-standard points, has been thought to provide a good control for all the therapist and other components of the placebo effect due to acupuncture. Unfortunately, this is not necessarily so. Lewith and Vincent ( 21) noted that sham acupuncture appeared to have an analgesic effect in 40-50% of patients in comparison with 60% for classical acupuncture. This means either that most of the acupuncture effect is placebo or that it is active at non-classical points. For the purpose of testing schizophrenia treatment, sham acupuncture may thus not be the best placebo.

Minimal acupuncture attempts to move on from this by inserting the needles only 1-2mm and with no stimulation. Its clinical effect is likely to be less than that of classical acupuncture, while maintaining most of the benefits of a placebo, unless the subject is well acquainted with classical acupuncture or unless the therapist changes any other factor inadvertently. Both these methods probably have a physiological effect, albeit a reduced one. Possible true placebos include deactivated mock TENS, mock laser and acupuncture needle guide tubes tapped without needle insertion; however these may not always be considered truly equivalent to acupuncture by the patients.


Because acupuncture requires some training, it is difficult for the therapist to be blind to the treatment he is giving. One can only strive to ensure that in giving placebo acupuncture consistency is maintained in all the other factors to avoid altering the placebo or therapist effect on the patients. However, it is perfectly possible for the selection, random allocation, and pre- and post-treatment measurements to be carried out by independent assessors with no clinical involvement.

Acupuncture method

According to traditional Chinese theory, acupuncture treatment is necessarily individualised, even if there is a general pattern. There is great variety in the points available, and in the depth and period of stimulation. The modern development of electroacupuncture and the use of lasers at acupuncture points have added to the complexity of treatment variation. However, most of the studies reviewed here have tested a set method of treatment for a given group.

Protocol violations

There are many examples of bias introduced into the reviewed studies due to changing treatment or group allocation during a study, or not accounting for cases who dropped out during a course of treatment or were lost to follow-up. For example, "For acupuncture-shy patients, small doses of such sedatives as chlorpromazine were given..." ( 1): the amounts were not recorded. And, "Adjustment of the treatment depends on the condition of the patient. If after 20 days, there was no response, the patients were withdrawn from the study and treatment changed." ( 17).

Statistical power

The main statistical limitations have been due to the poor methodology outlined above and due to small trial numbers. Where statistics have been used, it is often on demographic details such as the sex of the patients rather than comparing the modes of treatment.


There was rarely any mention of adverse reactions to the acupuncture or Chinese herbs used in the studies, and there was no mention of compliance with the treatments used. Both adverse reactions and compliance are significant problems that face psychiatrists in the use of neuroleptic medication. A number of serious adverse effects have been attributed to acupuncture ( 22).

All eight papers identified in our literature search on the use of acupuncture or low-power laser treatment for schizophrenia were significantly flawed. Although no reliable conclusions can be drawn from them, their evidence suggests that acupuncture or low-power laser may be effective in the treatment of schizophrenia. However, better designed, randomised controlled studies are necessary before acupuncture could be advocated as a treatment for schizophrenia.

(1.) Shi ZX, Tan MZ (1986) An analysis of the therapeutic effect of acupuncture treatment in 500 cases of schizophrenia. Journal of Traditional Chinese Medicine. 6(2): 99-104

(2.) Wu F (1995) Treatment of schizophrenia with acumoxibustion and Chinese medicine. Journal of Traditional Chinese Medicine. 15(2): 106-9

(3.) Esser AH, Botek ST, Gilbert C (1976) Acupuncture tonification: adjunct in psychiatric rehabilitation. American Journal of Chinese Medicine. 4(1): 73-9

(4.) Rubic B (1995) Can Western science provide a foundation for acupuncture? Alternative Therapies. 1(4): 41-7

(5.) Thompson JW. Neuropharmacology of acupuncture and electroacupuncture. Intractable Pain Society Proceedings. 5(1): 11-3

(6.) Starr M (1994) 5HT: the story so far. Acupuncture in Medicine. 12(2): 100-2

(7.) Han JS (1986) Electroacupuncture: an alternative to antidepressants for treating affective diseases? International Journal of Neurosciences. 29:79-92

(8.) Zhang WH, Shen YC (1981) Change in levels of monoamine neurotransmitters of rat brains after electric acupuncture stimulation. International Journal of Neurosciences. 15:147-9

(9.) Zhang WH, Luo HC, Shen YC (1981) The effect of electric acupuncture treatment on urinary MHPT-sulphate excretion in unmedicated schizophrenics. International Journal of Neuroscience 14:179-82

(10.) Liu Z, Wang Y, Zhang S, He A, Chen Y (1986) Therapeutic effect of HE-NE laser irradiation of point Erman in schizophrenic auditory hallucination. Journal of Traditional Chinese Medicine. 6(4): 253-6

(11.) Jia YK, Luo HC, Zan L, Jia TZ, Yan M (1987) A study of the treatment of schizophrenia with He-Ne laser irradiation of acupoint. Journal of Traditional Chinese Medicine. 7(4): 269-72

(12.) APA (1987) Diagnostic and statistical manual of mental disorders, 4th Edition. American Psychiatric Association, Washington

(13.) Overall JE, Gorham DR (1962) Brief Psychiatric Rating Scale. Psychological Reports. 10:799-812

(14) Zhang MJ (1988) Treatment of 296 cases of hallucination with scalp acupuncture. Journal of Traditional Chinese Medicine 8(3): 193-4

(15.) Kane J, Di Scipio WJ (1979) Acupuncture treatment of schizophrenia: report on three cases. American Journal of Psychiatry. 136(3): 297-302

(16.) Zhang L, Xu S, Tang Y, Zhu W (1990) A comparative study of the treatment of schizophrenia with electric acupuncture, herbal decoction and chlorpromazine. American Journal of Acupuncture. 18(1): 11-5

(17.) Zhang LD, Tang YH, Zhu WB, Xu SH (1987) Comparative study of schizophrenia treatment with electroacupuncture, herbs and chlorpromazine. Chinese Medical Journal. 100(2): 152-7

(18.) Shi ZX (1988) Observation on the therapeutic effect of 120 cases of hallucination treated with auricular acupuncture. Journal of Traditional Chinese Medicine. 8(4): 263-4

(19.) WHO (1992) International Statistical Classification of Diseases, loth Revision. World Health Organisation, Geneva

(20.) Lewith GT, Machin D (1983) On the clinical effects of acupuncture. Pain 16:111-27

(21.) Lewith GT, Vincent C (1995). Evaluation of the clinical effects of acupuncture: a problem reassessed and a framework for future research. Pain Forum 4(1): 29-39

(22.) Rampes H, James R (1995) Complications of acupuncture. Acupuncture in Medicine. 13(1): 26-33

The British Medical Acupuncture Society.


By Nicholas Beecroft and Hagen Rampes

Share this with your friends