Treatment of Low Back Pain with Specific Acupoint. A Double-Blind Placebo-Controlled Trial

Marco Aurelio Dornelles, MD
Freire Alemao,616/303
90450-060 - Porto Alegre
RS - Brazil


Low back pain is a common disease that injury a lots of people, causing physical impairment, pain and disability. It’s treatment is long lasting and expensive to patient and society. The purpose of this trial is to offer a new model for acupuncture treatment to low back pain Acupuncture treatment of a chronic low back pain was studied in a placebo-controlled double-blind crossover trial completed by 18 patients. Their pain were measured by VASP (Visual Analog Pain Scale ). These patients were randomly selected in two Groups : A and B. The A group received acupuncture needle at points K3 ; UB23 ; UB52 ; UB58 during first phase of treatment which lasted four weeks, and placebo acupoint during the second phase of trial. The B Group ( also 4 weeks ) that started with placebo acupoint ( S41, L9, L13 and BAXIE ) was changed during the second phase of treatment to specific acupoints and lasted four weeks, too. A significant reduction of 43% in the "VASP" occurred in the first phase of treatment for the A Group. On the other hand, the reduction of only 17% in the B Group was not significant. During the second phase of treatment the reduction was 52 % for the A Group and 71 % for the B Group . Finally when both groups were receiving acupuncture in specific points the reduction was significant only Group B . This trial was completed with success using a group of points needling was theoretically; irrelevant sites as placebo group and showed that is necessary specific acupoints for the treatment of low back pain.


Acupuncture; Low back pain ; Acupoint specificity ; Clinical trial ; Controlled




Numerous studies report the ability of acupuncture to reduce the pain in acute and chronic cases (Pommeranz,1987), but most of there are either anecdotal or employ inappropriate methodologies. In fact, scientific data for acupuncture safety and efficacy based on well-controlled studies utilizing acceptable methodologies and pain models is limited (MacDonald,1983).

A key problem with acupuncture research has been the difficulty in designing appropriate control groups (Campbell,1991). In some cases elaborate physical procedures such as mock transcutaneous nerve stimulation ( mTENS ) have been employed (Lewith,1983). However, despite its frequency in the literature, this practice is problematic because TENS in not acupuncture, and its use as a control is questionable. Very few studies used an adequate placebo procedure (Moore,1976). Another control commonly employed is "sham " acupuncture, in which needles are inserted at variables distances from the "real " acupuncture points, and with variable depth (Edlist et al, 1976). One argument suggest that pain attenuation by sham acupuncture may be due to diffuse noxious inhibitory control (Takeshige,1989), and his may confound efforts to test acupuncture effects. Another difficult to choose sham acupuncture is the fact that to show statistical significance in the differences between sham-point needling and true-points needling requires huge numbers of patients ( at least 122 per study ), in accordance with Vicent and Richardsonīs (1986) point of view. In this way, sham acupuncture is not an appropriate control.

Other common problems in acupuncture studies include ambiguous inclusion or diagnostic criteria, non-standard outcome measures, and insufficient follow-up periods (Patel,1987) . In many studies, patients, evaluation procedures, and statistical analyses are often inadequately described (Patel,1989).



Acupuncture can be applied to any of several hundred known acupuncture points all over the body. It has been reported that the application of acupuncture to a particular point is critical for analgesia to develop in a region of body (Croze et al,1976) . Although the old Chinese acupuncture theory of a hypothetical system of energy channels or meridians was considered the basis for acupuncture specificity (Chang,1979), the current scientific explanation over this specificity is conflicting. Chung(1989) demonstrate no evidence for stimulation point specificity for the production of analgesia, although other papers has been showed that acupuncture point stimulation is not effective in causing analgesia in all animals, classifying rats in two types, the responders and nonresponders to acupuncture points (Takeshige,1983) and acupuncture points and nonacupuncture points are distinguished by differing central pathway (Takeshige,1989).

In order to resolve some of the methodological problems apparent in the literature and develop a model for appropriate acupuncture evaluation of low back pain, we have completed a pilot study.

The present study is placebo-controlled and double-blind, which objectives are:

1 - to observe if the use of a group of acupoint selected without effect over the low back pain is an appropriate control group; and

2 - if there are acupoint specific to treat low back pain.




Twenty acupuncture naive patients with chronic low back pain without rhizopathy and had been failed to derive sufficient relief from appropriate conventional methods of treatment were included after they accepted the study design. All patient were referred to us before they were examined by a clinical assistant who blinded regarding what treatment group the patient belonged to. The patients were informed that we studying the effect of two different methods equally efficient against their low back pain. Thus, the patients were informed that placebo treatment will be an efficient new acupuncture treatment.

Before the patients were examined a random process was used to divide the sexes and age equally as possible between the two groups. The acupuncture group had 7 females and 3 males and placebo group ( nonspecific acupuncture points ) had 6 females and 4 males, whith 52 years old of mean age for first group and 54 years old for the last group.


The treatments were given twice a week for eight weeks. All patients were randomly allocated to either of two groups. Group A received specific acupuncture needles at points K3, UB23, UB52 and UB58 bilaterally, and one needle at DU3 and DU4 as described by Traditional Chinese Medicine ( TCM ) (Pommeranz,1987) to treat low back pain during first phase of treatment lasting four weeks. Group B received placebo acupoints defined like acupoints without effects over low back pain, in accordance with TCM at points S51, L9 and LI3, bilaterally and two needles at Baxie acupoints, lasting four weeks as in the Group A. Then after that period the patients crossed over and received the alternate for four weeks. The maximum number of treatments between the two phases was arbitrarily defined as 10 ( ten ).


Huskisson (1974) described an visual analogue scales for subjective measurement of pain. A horizontal line 100 mm in length, with 0, "no pain at all "and 10 ", pain as severe as it could possibly be’’ at the left and right extremes, respectively, scored by measuring the distance of patient’s mark before each treatment phase by an independent Clinical Assistant. Initial assessment was followed by a one-week baseline observation period during which the patients recorded pain severity using the Visual Analogue Scale of Pain ( VASP ) and that assessment had served for control in each group.(TABLE I ).



Acupuncture was performed according to TCM (Pommeranz,1987) by a physician trained at the Chinese Traditional medical Research Institute in Pekin (Author). With the patient in lateral decubitus, 10 (ten) needles were inserted close the spinal at tenderness points and at distance acupoints defined as efficient to complete the treatment. The needles were left in situ for 30 minutes with no further stimulation.

The placebo group consisted in the use of local and distant acupuncture points withouth relation with low back pain, in accordance with TCM. Ten needles were used at each treatment session and left in situ for 30 minutes with no further stimulation in the same way like acupuncture group.


The t-Test, at significant level for p < .05 , was used to compare the specific acupoints against nonspecific acupoints and whether that nonspecific acupoints may be considered an adequated placebo procedure.



Results are presented for the 18 patients who completed the trial. Two patients failed to complete the first phase of group B ( placebo ). The patients reported that the pain was the same at beginning of treatment and they not continue the trial.

Table I compares decrease in Visual Analogue Scales of Pain ( VASP ) in the patients groups before and after treatment sessions and the pain reduction observed and the baseline observation. A significant reduction in the VASP occurred in the first treatment phase for Group A and, inversely, in the Group B the reduction was not significant. The reduction was 43% for Group A receiving acupuncture, and 17 % for Group B receiving placebo. Further reduction in the VASP occurred in both groups during the second phase of treatment. The reduction was 52 % for Group A, now receiving placebo, and 71 % for Group B, now receiving acupuncture. This reduction was significant for only Group B .





( 1 WEEK )






49.2 %

48.9 (2.8)

27.5 (2.4)

43 % ( S )


43.4 %

44.2 (2.7)

36.6 (3.6)

17 % (NS)




27.5 (2.4)

13 (1.3)

52 % (NS)



36.6 (2.6)

10.6 (1.2)

71 % (S)

S = Significant ( p <.05 )

NS = Not Significant



Because patients entered the study on voluntary basis, the group was self-selected and randomised trial was defined. This may be important because the patient’s expectant attitude to acupuncture could affect the response to treatment (Mendelson et al,1983) . The same way, the clinical observer that assessed all patients at the beginning and the end of each course of treatment and the physician that applied the needles were blinded one each other. Therapist expectant effects upon patient assessement and other methodological artifacts may equally well create the appearance of trends in data (Patel,1987;Patel,1989) which do not deserve to be with this design. This trial was double-blind as it was not possible to be sure that the patients would not describe, during the second assessment, the methods of treatment that had been employed.

Although, in this trial was not used a rest period between the phases this is not important, because the follow-up periods have been variable with the effects of acupuncture persisting since 40 hours (Fox,1976) until 12 months (Carlson,1993). The relative long-term follow-up data is especially disappointing with chronic pain patients where the need to demonstrate lasting benefits from any new treatment is of paramount importance (Richardson,1986).

The overall significance of point location is unclear as yet. The majority of studies investigating this question have compared acupuncture at classical sites with needle insertion at random or theoretically irrelevant locations (Patel,1987). Of eleven controlled clinical trials in this category five find classical acupuncture to be significantly more effective than the control group and the remaining find non-significant differences in favour of acupuncture. In this trial were tested with success the use of a group of points needling at theoretically irrelevant sites as a placebo group.

Satisfactory pain relief was defined by Beecher (1955) as a 50 percent or greater reduction of pain severity. More recently, a reduction of 33 percent in pain severity has been accepted as significant (Moertel,1976). Based on this criterion, in the first phase of the trial acupuncture led to satisfactory pain relief (43 % ), whereas placebo did not ( 17 % ). This was reversed during the second phase, with the placebo group reporting 52 % pain reduction against with 71 % for the acupuncture group. However, the both groups had reduction on VASP during the trial, when the acupuncture was used the results were more significant than placebo acupuncture group. Group A showed a significant reduction on the first phase of trial and although that effect can be observed on the second phase of the trial this was lesser than Group B, now using acupuncture points. Possibly, this effects is attributed to a physiological or psychological states induced by the first treatment and persisting during the second phase of the trial (Price et al,1984) or an increase in the endorphin level in the cerebrospinal fluid (Sjolund et al,1977), caused for any stimulation on the body skin (Han, 1989).

This study have been concluded the following :

a- this pilot study provides an effective model for evaluation of efficacy on low back pain.

b- the present trial successfully established a valid nonspecific group of acupoints as a control group (placebo ).

c- specific acupoints to treat low back pain, in accordance with TCM, is more effective than nonspecific acupoints used like placebo.