Sie sind auf Seite 1von 13

This article is protected by copyright. To share or copy this article, please visit copyright.com. Use ISSN# 10786791. To subscribe, visit alternative-therapies.com.

meta-analysis

ACUPUNCTURE FOR ESSENTIAL HYPERTENSION

Leo-Wi Kim, MD, PhD; Jiang Zhu, MD, MS

Objective • To assess the efficacy of acupuncture for treatment of essential hypertension and the efficacy of acupuncture using pre- scription adhering to the principles of “syndrome differentiation.” Data Sources • Medline, Embase, Cochrane Central Register, and China National Knowledge Infrastructure (September 2008). Study Selection • Randomized, controlled trials comparing acupuncture with sham acupuncture, antihypertensive drugs, Chinese herbal medicine, or exercise in essential hypertension. Data Extraction • Two reviewers independently assessed trials for inclusion, extracted data, assessed methodological quality, and extracted outcome data on blood pressure. Data Synthesis • Treatment effects were summarized as mean differences with 95% confidence intervals. Twenty trials were included: three trials were relatively rigorous while others were methodologically suboptimal. Acupuncture arms achieved sig- nificant effect modification on blood pressure compared with control arms (19 comparisons: systolic blood pressure [SBP]:

mean difference -4.23 mmHg, 95% confidence intervals -6.47 to -1.99; diastolic blood pressure [DBP]: -2.53, -3.99 to -1.08), with

significant heterogeneity. In high-quality trials, blood pressure was significantly lower in treatments of acupuncture plus anti- hypertensive drug arms than in sham-acupuncture plus hyper- tensive drug arms (two comparisons: SBP: -5.72 mmHg, -8.77 to -2.68; DBP: -2.80, -5.07 to -0.54), with no significant heterogene- ity. As for trials using prescription adhering to the principles of syndrome differentiation, we found a significant blood pressure reduction with acupuncture arms in comparison with control arms (11 comparisons: SBP: -6.46 mmHg, -8.04 to -4.87; DBP:

-3.07, -4.17 to -1.96) with no significant heterogeneity. In con- trast, in trials not using prescription adhering to the principles of syndrome differentiation, we found no significant reduction in blood pressure with acupuncture arms in comparison with control arms (eight comparisons: SBP: -1.55 mmHg, -5.39 to 2.29; DBP: -2.12, -4.97 to 0.73) with significant heterogeneity. Conclusions • Because of the paucity of rigorous trials and the mixed results, these findings result in limited conclusions. More rigorously designed and powered studies are needed. (Altern Ther Health Med. 2009;16(2):e-pub ahead of print.)

Leo-Wi Kim, MD, PhD, is a lecturer in the department of Oriental Medicine Resources, College of Science and Engineering, Far East University, Korea. Jiang Zhu, MD, MS, is a professor in the Departments of Acupuncture and Moxibustion, College of Acupuncture and Moxibustion, Beijing University of Traditional Chinese Medicine, China.

Corresponding author: Leo-Wi Kim, MD, PhD E-mail: leowikim@naver.com

Editor’s note: Due to space limitations, four tables and figures (Tables 2 and 4; Figures 4 and 5) were omitted from the print version of the article. The full version of the article appears here.

T oday, hypertension represents a growing worldwide public health concern. Recent data suggest that 26.4% of the world’s adult population suffers from hyperten- sion. 1 The presence of high blood pressure (BP) dou- bles the risk of ischemic heart disease and increases

the incidence of stroke four-fold. 2 Thus hypertension is an impor- tant public-health challenge both because of its high frequency and because of the concomitant risk of cardiovascular diseases.

Evidence from randomized controlled trials (RCTs) shows that effective medication reduces the risk of cardiovascular morbidity and mortality. 3,4 There is concern that the benefits demonstrated in RCTs of antihypertensive medication are not implemented in everyday clinical practice 5 and that the goals of lowered BP are achieved in only 25% to 40% of the patients who take antihyperten- sive medication. 1,5,6 The recent studies have emphasized the bene- fits of early and good BP control, confirming the requirement for two or three antihypertensive agents to achieve satisfactory con- trol. 7,8 In order to reduce the likelihood of drug toxicity and pill burden and to aid adherence, a combination of established agents ensuring efficacy and using a low dose is required. 9 It is empha- sized that the majority of the hypertensive population will require individualized therapy to achieve the recommended goals and that individualized therapy has potential for reducing the burden of cardiovascular disease, particularly of stroke. 10 In traditional Chinese medicine (TCM), acupuncture plays a central role in a comprehensive system of medicine aimed at maintaining health and correcting disease processes. One of the important concepts of TCM is “syndrome differentiation.” It means to analyze, induce, synthesize, judge, and summarize the clinical data of symptoms and signs collected with the diagnostic

methods of TCM into specific syndromes. The therapeutic meth- ods are then decided according to the result of syndrome differ- entiation. 11 The therapies related to treating hypertension in TCM include Chinese herbs, acupuncture, acupressure, moxibus- tion, and qigong. Acupuncture as a nonpharmacological inter- vention has been used to treat a wide variety of conditions to regulate cardiovascular diseases in the East for centuries. Recently, acupuncture has become one of the most popular com- plementary therapies in the West. 12 Acupuncture therapy is used on patients with mild or borderline hypertension who want to avoid drug therapy or as an alternative therapeutic option to reduce dosages of antihypertensive agents. The aim of this review is to evaluate the efficacy of acupunc- ture in treatment of essential hypertension and the effect of acu- puncture using prescription adhering to the principles of syndrome differentiation and to identify whether acupuncture therapy appeared sufficiently promising as to justify further large-scale RCTs.

METHODS Eligibility Criteria We included all RCTs of acupuncture that were a sole treat- ment or an adjuvant treatment for medication in adult patients with essential hypertension. We included trials with needle inser- tion into traditional meridian points. Trials that used auriculo- therapy, laser acupuncture, or electroacupressure (without needle insertion) were excluded. We excluded trials that evaluat- ed the efficacy of acupuncture only after one treatment as well as trials of secondary hypertension.

Study Identification and Selection We searched electronic databases Medline, Embase, Cochrane Central Register, and China National Knowledge Infrastructure (CNKI) from January 1, 1980, to January 10, 2008, and updated it to September 31, 2008. We combined acupuncture-related terms (acupuncture, acupoint, needling, or moxibustion) with hypertension- related terms (blood pressure, hypertension, or hypertensive). We also manually searched the reference lists from primary articles. We imposed no restrictions on the type or language of publication. We considered older RCTs that were included in previous reviews of acupuncture for lowering blood pressure. 13

Data Extraction Two reviewers independently screened all citations and abstracted data. We extracted data pertaining to quality of the methods, participants, interventions, and outcomes. All the tri- als reported composite outcomes, so they were not appropriate for meta-analysis. Instead, we extracted BP data that were objec- tively defined and were common among trials. All other out- comes were excluded from this review.

Quality Assessment Two reviewers independently assessed the methodological quality of included trials by using an 11-item scale developed by

the Cochrane Back Review Group, 14,15 with disagreements resolved by consensus. We awarded a maximum of 11 points in 11 categories: Was the method of randomization adequate? Was the treatment allocation concealed? Were the groups similar at

baseline? Was the patient blinded to intervention? (We could not

be certain that invasive sham needles were sufficiently credible as

sham control to the treatment being evaluated; therefore, we assigned 0.5 point to invasive sham and 1 point to noninvasive sham.) Was the care provider blinded to intervention? Was the outcome assessor blinded to intervention? Were co-interventions

avoided or similar? Was the compliance acceptable in all groups? Was the dropout rate described and acceptable? Was the timing

of outcome assessment identical and adequate in assessment of

the efficacy acupuncture for treating hypertension in all groups (0.5 point to the identical timing, 0.5 point to the adequate tim- ing)? Did the analysis include an intention-to-treat analysis? We considered a score of 6 or more points to indicate high quality.

Data Synthesis and Analysis We categorized the included trials into predefined subgroups. The subgroup analyses were performed separately, depending on

the quality of trials. Some trials used pre-selected, unified acu- points for all the participants, whereas others grouped all the par- ticipants into several categories by syndrome differentiation and used different pre-selected acupoints according to category. As prescription adhering to the principles of syndrome differentiation

is the most important basic principle of TCM theory, we per-

formed two subgroup analyses for evaluating the effect of prescrip- tion adhering to the principles of syndrome differentiation. We calculated the mean differences (MD) with 95% confi- dence intervals (CI) by using the values of the outcome at the end

of treatment. If any further follow-ups were measured, we per-

formed a separate analysis of each measurement point. The het-

erogeneity between trials was tested with χ 2 statistic (Cochran’s

Q test), 16 computing the square distance of each study from the

combined effect. We also calculated the quantity I2, 17 which describes the percentage variation across studies that is due to heterogeneity rather than chance. The random-effect model was used if the heterogeneity statistic among studies was statistically significant 18 ; otherwise, the fixed-effect model was used. 19 By using a funnel plot, we assessed potential publication bias. 20 All statistical analyses were performed using Review Manager Version 5.0.16 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2008, Copenhagen, Denmark).

RESULTS Search Results This review identified a total of 687 titles and four manual searches that met the search criteria. The full articles of the retrieved 188 trials were read to assess their appropriateness for meta-analysis. Data from 20 articles 21-40 of 47 potentially appro- priate RCTs met the inclusion criteria. Four trials 24, 32, 35, 40 were identified through Medline, Embase, or Cochrane Central Register, four trials 23, 26, 31, 38 were identified through both Medline

and CNKI database, and 12 trials 21, 22, 25, 27-30, 33, 34, 36, 37, 39 were from the CNKI database. We excluded 27 trials: 15 for not presenting sep- arate results for BP, one for insufficient outcome data, eight for one-time treatment, two for duplicate reports, and one trial for not having kept the antihypertensive drug’s dose uniform during the duration of the acupuncture treatment. Figure 1 presents a flow chart of retrieved trials and trials excluded with specified reasons. All trials had been conducted in four different countries and published in English journals (three trials 32,35,40 ), a German journal (one trial 24 ), or Chinese journals (16 trials).

Trial Characteristics A total of 1528 patients from 20 trials were included in this review. Table 1 details additional characteristic of the interven- tions in the included RCTs. Acupuncture was the sole treatment in 11 trials, whereas in nine trials, acupuncture was used as a cooperative treatment for medication. As for control, sham acu- puncture was adopted for control in four trials 24,32,35,40 , whereas 14 trials 21-23,25,26,28,30,31,33,34,36-39 used antihypertensive drugs only as a con- trol, one trial 27 used Chinese herbal medicine plus an antihyper- tensive drug, and one trial 29 used exercise only. In all the trials, the selection of acupoints was designed for the purpose of reducing the BP of hypertensive patients. Eight trials used unified acupoints for all those participants, whereas 11 trials grouped all those participants into three to five catego- ries by syndrome differentiation and used different pre-selected acupoints according to those categories. One article 32 provided two separate active formulas: one was “TCM individualized” (using prescription adhering to the principles of syndrome dif- ferentiation), and the other was “TCM standardized” (not using prescription adhering to the principles of syndrome differentia- tion but using unified acupoints).

Trial Quality Table 2 summarizes the qualities of the 20 included RCTs. The methodological qualities of 17 trials, which were published in Chinese or German journals, were suboptimal. Only three tri- als 32,35,40 published in English journals were relatively rigorous. In terms of randomization, only one trial 32 used an allocation procedure that would be considered as concealed. Two trials 32,40 were double-blinded, and two trials 24,35 were single-blinded. Four trials 24,32,35,40 used sham needles as control, and two trials 32,35 were analyzed on intention-to-treat basis. Lost follow-up was described in four trials 24,32,35,40 and further follow-up occurred in two of these trials. 32,35

Quantitative Data Syntheses Blood Pressure Acupuncture arms achieved significant effect modification on BP compared with control arms (systolic blood pressure [SBP]: MD -4.23 mmHg, 95% CI random -6.47 to -1.99; diastolic blood pressure [DBP]: -2.53, -3.99 to -1.08). We found significant heterogeneity for this outcome (SBP: χ2=69.11, I ² =74%; DBP:

χ ² =59.80, I ² =70%). The funnel plot showed asymmetry consis-

Potentially relevant studies identified and screened for retrieval (N=687):

Medline (n=129) Embase (n=103) Cochrane Central (n=36) CNKI (n=415) Manual search (n=4)

Cochrane Central (n=36) CNKI (n=415) Manual search (n=4) Excluded (n=499): Nonrandomized trials, or review (n=399)

Excluded (n=499):

Nonrandomized trials, or review (n=399) Not an in vivo human trial (n=100)

trials, or review (n=399) Not an in vivo human trial (n=100) Studies retrieved for more detailed

Studies retrieved for more detailed evaluation (n=188)

Studies retrieved for more detailed evaluation (n=188) Excluded (n=141): Search overlap (n=58) Nonrandomized

Excluded (n=141):

Search overlap (n=58) Nonrandomized trials (n=29)

Not essential hypertensive patients (n=28) Protocol manuscript (n=2) Not inserted into tender points (n=18) Preferred inter-acupoints (n=1) Preferred inter-manipulations (n=5)

inter-acupoints (n=1) Preferred inter-manipulations (n=5) Potentially appropriate RCTs to be included (n=47) Excluded

Potentially appropriate RCTs to be included (n=47)

(n=5) Potentially appropriate RCTs to be included (n=47) Excluded (n=27): Not presenting separate results for BP

Excluded (n=27):

Not presenting separate results for BP (n=15) Insufficient outcome data (n=1) One-time treatment (n=8) Duplicate reports (n=2) Not having kept the antihypertensive drug’s dose uniform during the duration of the acupuncture treatment (n=1)

during the duration of the acupuncture treatment (n=1) RCTs included in the reviews (n=20 trials, 1528

RCTs included in the reviews (n=20 trials, 1528 patients) Quality of trials categorized as follows:

High-quality (n=3 trials, 358 patients) Low-quality (n=17 trials, 1170 patients) Acupuncture interventions in trials categorized as follows:

Acupuncture only (n=11 trials) Acupuncture plus medication (n=9 trials) Measurement points as follows:

After ≤5 d of treatment (n=19 trials) After 3 wks of treatment (n=1 trial)

FIGURE 1 Flow of Studies Through Selection Process

TABLE 1 Characteristic of Interventions and Outcomes in Included Trials*

Study ID

Location

Acupuncture rationale

 

Needling Details

(Year)

 

Style of

Rationale for

Prescription adhering to the principles of TCM syndrome differentiation

Unified acupoints

Additional acupoints by syndrome types

Depths of

acupunc-

treatment

(auricular

acupoints)

insertion

ture

 

(cun)

He XW,

China

Chinese

TCM pattern

Used

BL23, LI11, ST40

4 types: hyperactivity of liver yang: BL18, LR03; yin asthenia and yang hyperactivity: SP06, ST36; asthenia of yin and yang: SP06, GV04; asthenia of yang: GV04

NR

1994

21

diagnoses

Yin ZF, 1994 22 China

Chinese

TCM pattern

Used

GB20, LI11, SP06, ST36

3 types: hyperactivity of liver yang: LR03; asthenia of kidney and liver yin: KI03; retention of phlegmatic dampness: ST40

NR

 

diagnoses

Dan Y, 1998 23

China

Chinese

TCM pattern

Used

BL17, GB20, LI04, LI11, LR03

3 types: yin asthenia and yang hyperactivity: KI07, PC06; retention of phlegmatic dampness: SP04, ST40; asthenia of yin and yang: SP06, ST36

NR

 

diagnoses

Kraff K,

Germany

Chinese

TCM pattern

Used

BL18, BL23, GB20, GV20, HT07, KI03, LR02, LR03, SP06

 

0.5 cm

1999

24

diagnoses

 

Chen YF,

China

Chinese

TCM pattern

Not

LI11, ST40

NR

2000

25

diagnoses

Jiang XL,

China

Chinese

NR

Not

LI11, LR03, ST40

 

2003

26

Song YM,

China

Chinese

TCM pattern

Used

BL18, BL23, GB20, LI11, LR03,

LR02 was added to headaches or dizziness, HT07 to insom- nia, PC06 to palpitation.

NR

2003

27

diagnoses

ST36

Wu QM,

China

Chinese

NR

Not

GV20, LI04, LR03

 

0.5~1

2003

28

Zhao DJ,

China

Chinese

NR

Not

CV04, LR03, SP06, ST36, ST40

 

NR

2003

29,

Hu LH,

China

Chinese

TCM pattern

Used

GB20, GV20, HT07, LI11, SP06 (Heart, Liver, Kidney, Shenmen, Jiangyagou)

3

types: hyperactivity of liver yang: LR03; asthenia of kidney and liver yin: KI03; retention of phlegmatic dampness: ST40

NR

2004

30

diagnoses

Zhang YL,

China

Chinese

TCM pattern

Used

GB20, LI11, SP06, ST36

3 types: hyperactivity of liver yang: LR03; asthenia of kidney and liver yin: KI03; retention of phlegmatic dampness: ST40

1~1.5

2005

31

diagnoses

Macklin EA,

United

Chinese

TCM pattern

Formula 1 used

GB20, GV20, LI04, LI11, LR03 (Heart, Liver, Shenmen, Jiangyagou, Sympathetic nerve)

5

types: hyperactivity of liver yang: GB21, GB34, GB43,

0.3~1.5

2006

32

States

diagnoses

LR02, ST36, ST44, Taiyang; asthenia of kidney yin and hyperactivity of liver yin: BL18, BL23, HT07, KI03, SP06, yin- tang; retention of phlegmatic dampness: BL20, BL64, CV12, PC06, SP06, ST08, ST36, ST40; asthenia of yin and yang:

 

BL23, CV04, CV06, GV04, KI03, SP06, ST36; asthenia of qi and blood and hyperactivity of liver yang: BL18, BL20, BL23, CV04, CV06, HT07, KI03, SP06, ST36

 

Formula 2 not used

GB20, LI11, LR03, SP06, ST36 (Heart, Jiangyagou)

 

Wang C,

China

Chinese

NR

Not

GV20, LI11, LR03, ST36

 

NR

2006

33

Wang LY,

China

Chinese

NR

Not

GB20

0.8~1

2006

34

Flachskampf FA, 2007 35

Germany

Chinese

TCM pattern

Used

BL18,BL23,BL64,CV04,CV06,

4

types: hyperactivity of liver yang, retention of phlegmatic dampness, asthenia of yin and yang, yin asthenia and yang hyperactivity.

 
 

diagnoses

CV12, GV20, GB20, LI04, LI11,

 

LR02,LR03,PC06,SP06,ST36,

ST40,Taiyang

 

Guo YH,

China

Chinese

NR

Not

GV20, KI03, LI11, LR03, SP06, ST36, ST40

 

NR

2007

36

 

Hu LH,

China

Chinese

TCM pattern

Used

GB20, GV20, LI11, SP06 (Wrist-ankle acupuncture)

3

types: hyperactivity of liver yang: LR03; asthenia of kidney and liver yin: KI03; retention of phlegmatic dampness: ST40

NR

2007

37

diagnoses

Huang F,

China

Chinese

TCM pattern

Used

GB20, LI11, LR03, PC06, ST36,

Only enrolled patients with syndrome of phlegmatic reten- tion and blood stasis in Luo.

30mm or

2007

38

diagnoses

ST40

20mm

Wang X,

China

Chinese

TCM pattern

Not

LR03

0.5~0.8

2007

39

diagnoses

Yin C, 2007 40

Korea

Saam acu-

Saam acu- puncture the- ory of Korean acupuncture

Used

3 types: BL25, LI11, ST36 for reinforcement of large intestine meridian energy, BL13, LU09, SP03 for lung, CV04, KI02, KI07 for kidney, GV14, GB20, LI01 for bladder. HT07, PC06 were added when a psychological factor was considered.

NR

 

puncture of

 

Korean

TCM indicates traditional Chinese medicine; BP, blood pressure; ABPM, ambulatory blood pressure monitoring; Chinese-herb, Chinese herbal medicine; ET, end of treatment; NR: Not reported. †A: reducing manipulation by twirling, rotating, lifting and thrusting the needle; B: reducing manipulation by twirling and rotating the needle; C: reinforcing or reducing manipulation by twirling and rotating

TABLE 1 Characteristic of Interventions and Outcomes in Included Trials, continued*

24.

Deqi

feeling

25.

Kraft K, Coulon S. Effect of a standardized acupuncture treatment on complains, blood

pressure and serum lipids of hypertensive, postmenopausal women. A randomized,

controlled clinical study [article in German]. Forsch Komplementarmed. 1999;6(2):74-79.

Needling Details

Needle type

Chen YF, Qian H, Li L, et al. Effects of acupuncture on contents of plasma endothelin

and angiotensin in the patient of hypertension. Zhongguo Zhen Jiu.

retention

stimulation;

Needle

Needle

Co-interventions

Additional

medication

Other

interventions

Duration,

Days

(Sessions, n)

Control

Interventions

Practitioner

Background

Outcomes

Measurement

Points

time

2000;20(11):691-694.

manipulations*

(min)

26. Jiang X. Effects of magnetic needle acupuncture on blood pressure and plasma ET-1

level in the patient of hypertension. J Tradit Chin Med. 2003;23(4):290-291.

27. Song YM, Song D, Du LL. Clinical observation on acupuncture combined with medi-

           

NR

cine in 64 cases of hypertension. Shanxi J Tradit Chin Med. 2003;24(11):1005-1006.

15~20

NR

Manual; F

No

No

30 (daily)

Antihypertensive

NR

BP, BP change magnitude

ET

28.

Wu QM, Feng GX. The correlation between hypotensive effect and plasma Ang after warm acu-moxi on Kaisiguan(Extra) and Baihui(DV20) points. New J Tradit Chin Med.

drugs

2003;35(12):45-47.

42 (21)

Antihypertensive

drugs

NR

BP

After 1 d of treatment

NR

29.

Zhao DJ, Fan QL. Effects of acupuncture on insulin resistance in the patient of hyper- tension. Zhongguo Zhen Jiu. 2003;23(3):165-167.

30

NR

Manual; F

No

No

30.

After

Deqi

Hu LH, Yan W, Chen WG, Zhou GM, Jiang L, Yang YY. Clinical observation on elec-

20~30

NR

Manual; F

No

No

troacupuncture combined with Cizhu sticked to the auricular points and medicine for treatment of hypertension. Chin J Physical Med Rehabil. 2004;26(4):248-249.

21 (daily)

Nifedipine

NR

ABPM

ET

31.

Zhang YL, Li CP, Peng M, Yang HS. Effects of acupuncture combined with medicine on

 

After

neuropeptide Y in the patient of hypertension [article in Chinese]. Zhongguo Zhen Jiu.

30

NR

Manual

No

No

84 (24)

Invasive sham

NR

ABPM

ET

Deqi

2005;25(3):155-157.

acupuncture

 

32.

Macklin EA, Wayne PM, Kalish LA, et al. Stop Hypertension with the Acupuncture

 

After

Research Program (SHARP): results of a randomized, controlled clinical trial.

15~30

NR

Manual; A or E

No

No

14 (daily)

Nifedipine

NR

BP

ET

Deqi

Hypertension. 2006;48(5):838-845.

33.

After

Deqi

Wang C, Cheng ZQ. Clinical effective valuation and its mechanical analysis of acupunc-

30

NR

obese

ture

2006;33(10):1327-1328.

on

hypertensive

Manual; NR

patients.

Captopril

Tradit

Liaoning

J

Chin

No

Med.

20 (daily)

Captopril

NR

BP

After 1 d of treatment

NR

34.

Wang LY, Chen BG. Clinical study on therapeutic effect and adjustment to plasma ET

NR

NR

No

Manual; F

Nifedipine,

and serum TNF-α in the patient of essential hypertension with acupuncture at Fengchi

captopril, aspirin

30 (daily)

Nifedipine,

captopril, aspirin

NR

BP

ET

point. J Hubei College of Tradit Chin Med. 2006;8(1):8-10.

enteric-coated,

enteric-coated,

35.

Flachskampf FA, Gallasch J, Gefeller O, et al. Randomized trial of acupuncture to lower blood

Chinese herb

 

Chinese herb

pressure. Circulation. 2007;115(24):3121-3129.

Captopril

 

After 5d of treatment

After

36.

Deqi

Guo YH. Clinical observation on effect of acupuncture on insulin resistance in hyper-

Huatuopian needles,

tension. Acta Chin Med Pharmacol. 2007;35(6):51-53.

20

No 38, 1.5 cun,

Manual; A

No

Moxibustion

30 (daily)

NR

BP

37.

Chana

Hu LH, Yan W, Zhou GM, Chen SQ, Wu YP. Clinical observation on wrist-ankle acu-

 

After

Deqi

38.

puncture combined with medicine for treatment of hypertension. Chin J Cardiovasc

20

NR

Manual; E

No

Behavior

therapy

Rehabil Med. 2007;16(2):184-185.

Huang F, Yao GX, Huang XL, Liu YN. Clinical observation on acupuncture for treat-

40

(30)

Behavior therapy

NR

BP

ET

After

ment of hypertension of phlegm-stasis blocking collateral type [article in Chinese].

30

0.3x40 mm

Electrical

Amlodipine

Aricular

24

(daily)

Amlodipine

NR

ABPM

ET

Deqi

Zhongguo Zhen Jiu. 2007;27(6):403-406.

(2-3Hz); E

acupuncture

39.

Wang X, Wu HL, Li SQ. An assessment of antihypertensive effect of acupuncture at

 
   

After

Deqi

40.

Taichong(LR 3) with ambulatory blood pressure monitoring. J New Chin Med.

2007;39(11):21-22.

30

No 28, 2 cun, Hua-

Manual; C

Nifedipine

No

Yin C, Seo B, Park HJ, et al. Acupuncture, a promising adjunctive therapy for essential

tuopianneedles,China

20 (daily)

Nifedipine

NR

BP

ET

After

hypertension: a double-blind, randomized, controlled trial. Neurol Res. 2007;29 Suppl

30

0.16~0.3x13~5 mm

Manual; NR

No

Auricular acu-

42-56 (≤12)

Invasive sham acupuncture, 0.2x15, 25, 30 mm needles, 16 mm depth

Described

 

After 3 wk of treatment (after 10 wk of random assignment), after 4, 6, 9, 12 mo of random assignment

1:S98-S103.

Deqi

41.

stainless steel, Seirin

Kaptchuk TJ, Stason WB, Davis RB, et al. Sham device v inert pill: randomised con-

Kasei needles, Japan;

trolled trial of two placebo treatments. BMJ. 2006;18(7538):391-397.

Carbo needles, China

puncture

42. Park J, White AR, Ernst E. New sham method in auricular acupuncture. Arch Intern Med. 2001;161(6):894; author reply 895.

 

BP change mag- nitude

43. Streitberger K, Kleinhenz J. Introducing a placebo needle into acupuncture research. Lancet.

 

1998;352(9125):364-365.

44. White P, Lewith G, Hopwood V, Prescott P. The placebo needle, is it a valid and con-

 

vincing placebo for use in acupuncture trials? A randomised, single-blind, cross-over pilot trial. Pain. 2003;106(3):401-409.

 

45. Park J. Sham needle control needs careful approach. Pain. 2004;109(1-2):195-196.

After

46.

30

0.35x50 mm

Electrical; NR

Benazepril

Deqi

Streitberger K, Vickers A. Placebo in acupuncture trials. Pain. 2004;109(1-2):195;

author reply 197-199.

Exercise, dietet-

ic treatment

56

(32)

Benazepril, exercise,

dietetic treatment

NR

BP

ET

47.

After

Deqi

Kaptchuk TJ, Kelley JM, Conboy LA, et al. Components of placebo effect: randomised

30

0.25x40 mm

Manual; A

No

No

controlled trial in patients with irritable bowel syndrome. BMJ.

stainless steel

28 (daily)

Metoprolol

NR

BP, BP change magnitude

After 5 d of treatment

2008;336(7651):999-1003.

NR

48.

Dennehy EB, Webb A, Suppes T. Assessment of beliefs in the effectiveness of acupunc-

Shenzhou needles,

ture for treatment of psychiatric symptoms. J Altern Compl Med. 2002;8(4):421-425.

30

0.25x25~50 mm,

Manual; NR

Used or not

No

42 (22)

Invasive sham

acupuncture

Described

ABPM, BP

change

ET, after 3, 6 mo of treatment

49.

china

Pittler MH, Abbot NC, Harkness EF, Ernst E. Location bias in controlled clinical trials

 

magnitude

of complementary/alternative therapies. J Clin Epidemiol. 2000;53(5):485-489.

 

50.

After

Deqi

Pham B, Klassen TP, Lawson ML, Moher D. Language of publication restrictions in sys-

tematic reviews gave different results depending on whether the intervention was con- ventional or complementary. J Clin Epidemiol. 2005;58(8):769-776.

30

NR

Manual; E

No

No

30 (daily)

Enalapril maleate

NR

BP

After 5d of treatment

NR

51.

MacPherson H, White A, Cummings M, Jobst K, Rose K, Niemtzow R; STandards for

30

0.3x40 mm

Manual; E

Amlodipine

Wrist-ankle

Reporting Interventions in Controlled Trails of Acupuncture. Standards for reporting

acupuncture

20

(daily)

Amlodipine

NR

ABPM, BP change magnitude

ET

   

After

Deqi

interventions in controlled trials of acupuncture: The STRICTA recommendations.

STandards for Reporting Interventions in Controlled Trails of Acupuncture. Acupunct

Med. 2002;20(1):22-25.

30

0.3x40 mm

Manual; A

Captopril

No

28 (daily)

Captopril

 

NR

BP

ET

After

Deqi

20

No. 28, Huatuopian needles, China

Manual; B

No

No

7 (daily)

Captopril

NR

ABPM

ET

 

After

No reten-

NR

Manual; G

Used

Auricular acu-

56

(17)

Noninvasive sham acupuncture (Park’s sham needle), antihypertensive drugs, exercise

NR

BP, BP change magnitude

ET, after 4 wk of intervention

Deqi

tion

puncture, exer-

cise

 

the needle; D: twirling, rotating, lifting and thrusting the needle; E: uniform reinforcing-reducing manipulation; F: reinforcing or reducing manipulation; G: twirling the needle.

TABLE 2 Quality Assessment in Included Trials

Study ID, Year

Quality Assessment Items* (Composite score)

He XW, 1994 21

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Yin ZF, 1994 22

NR,NR,NR,NR,NR,NR,1,NR,NR,1,NR (2)

Dan Y, 1998 23

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Kraff K, 1999 24

NR,NR,NR,0.5,NR,NR,1,1,1,1,0 (4.5)

Chen YF, 2000 25

NR,NR,NR,NR,NR,NR,1,NR,NR,1,NR (2)

Jiang XL, 2003 26

NR,NR,NR,NR,NR,NR,1,NR,NR,1,NR (2)

Song YM, 2003 27

NR,NR,NR,NR,0,NR,1,NR,NR,1,NR (2)

Wu QM, 2003 28

1,NR,1,NR,NR,NR,1,NR,NR,1,NR (4)

Zhao DJ, 2003 29

1,NR,1,NR,NR,NR,1,NR,NR,1,NR (4)

Hu LH, 2004 30

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Zhang YL, 2005 31

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Macklin EA, 2006 32

1,1,1,0.5,0,1,1,1,1,0.5,1 (9)

Wang C, 2006 33

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Wang LY, 2006 34

1,NR,1,NR,NR,NR,1,NR,NR,1,NR (4)

Flachskampf FA, 2007 35

1,NR,1,0.5,0,0,1,1,1,1,1 (7.5)

Guo YH, 2007 36

1,NR,1,NR,NR,NR,1,NR,NR,1,NR (4)

Hu LH, 2007 37

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Huang F, 2007 38

1,NR,1,NR,NR,NR,1,NR,NR,1,NR (4)

Wang X, 2007 39

NR,NR,1,NR,NR,NR,1,NR,NR,1,NR (3)

Yin C, 2007 40

1,NR,1,1,NR,1,1,1,1,1,NR (8)

*NR indicates not reported. Eleven categories: randomization, allocation con- cealment, similarity at baseline, blinding for participants, blinding for care provider, blinding for outcome assessors, similarity of co-interventions, com- pliance in all groups, dropout, timing of outcome assessment, intention-to- treat analysis.

tent with publication bias (Table 3, Figure 2). Acupuncture arms achieved significant effect modification on BP change magnitude compared with control arms (SBP: MD -5.98 mmHg, 95% CI random -9.48 to -2.47; DBP: -3.95, -5.19 to -2.72) with a significant heterogeneity (SBP: χ ² =67.39, I ² =94%; DBP: χ ² =9.65, I ² =59%). There was an indication of publication bias (Table 3, Figure 3).

High-quality Trials BP was significantly lower with acupuncture plus antihyper- tensive drugs than with sham acupuncture plus hypertensive drugs (SBP: MD -5.72 mmHg, 95% CI fixed -8.77 to -2.68; DBP:

-2.80, -5.07 to -0.54). We found no significant heterogeneity for this outcome (SBP: χ² =0.38, I ²=0%; DBP: χ²=0.66, I² =0%) (Table 3, Figure 2). Acupuncture plus antihypertensive drugs achieved signifi- cant effect modification on BP change magnitude compared with sham acupuncture plus hypertensive drugs (SBP: MD -8.38

mmHg, 95% CI fixed -9.16 to -7.60; DBP: -4.54, -5.08 to -4.00)

with no significant heterogeneity (SBP: χ ² =1.21, I ² =17%; DBP:

χ²=0.69, I² =0%) (Table 3, Figure 3). At the treatment’s completion of one trial, 32 there was no BP measurement; however, the first measurement occurred 3 weeks after the treatment’s completion. Acupuncture achieved no sig- nificant effect modification on BP change magnitude compared

with sham acupuncture (SBP: MD 0.20 mmHg, 95% CI -0.74 to

1.14; DBP: 0.00, -1.96 to 1.96) (Table 3). Another trial 35 measured BP at 3 months and 6 months after the treatment’s completion. Acupuncture plus antihypertensive drugs achieved no significant effect on BP compared with sham acupuncture plus hypertensive drug.

Low-quality Trials Five subgroups were analyzed based on methodological variables of acupuncture arms and control arms. The BP decreased significantly from baseline with acupuncture plus

antihypertensive drugs than with antihypertensive drugs. The BP-lowering effect of acupuncture plus Chinese herbal medi-

cine and antihypertensive drug was also significantly higher

than that of Chinese herbal medicine plus antihypertensive drug, whereas, compared with antihypertensive drugs, acu- puncture statistically showed no significant effect modifica- tion with statistically significant heterogeneity. Compared with sham acupuncture, acupuncture showed no significant effect modification. Compared with the exercise arm, acupunc-

ture plus exercise arm also showed no significant effect (Table

3, Figure 2). In the analysis of BP change magnitude, the efficacy of acu- puncture was not significant when compared with antihypertensive drugs. Also, compared with antihypertensive drugs, the efficacy of acupuncture plus antihypertensive drug was not significant.

Effect of Acupuncture Using Prescription Adhering to the Principles of Syndrome Differentiation in TCM Diagnosis As for trials using prescription adhering to the principles of

syndrome differentiation, we found a significant reduction in BP with acupuncture arms in comparison with control arms (SBP:

MD

-6.46 mmHg, fixed -8.04 to -4.87; DBP: -3.07, -4.17 to -1.96,

with

no significant heterogeneity [SBP: χ²=11.37, I ²=12%; DBP:

χ² =15.72, I ² =36%]). There was no indication of publication bias on inspection of funnel plots for asymmetry (Table 4, Figure 4).

In contrast, in trials not using prescription adhering to the

principles of syndrome differentiation, we found no significant reduction in BP with acupuncture arms in comparison with con-

trol arms (SBP: MD -1.55 mmHg, random -5.39 to 2.29; DBP:

-2.12, -4.97 to 0.73), with significant heterogeneity (SBP:

χ² =39.24, I² =82%; DBP: χ²=43.64, I ² =84%). There was an indica-

tion of publication bias (Table 4, Figure 5).

Adverse Effects of Acupuncture In only three trials 35,39,40 were adverse effects discussed. In those trials, no serious adverse effects associated with acupunc-

TABLE 3 Efficacy of Acupuncture on Blood Pressure

     

Systolic Blood Pressure (mmHg)

Diastolic Blood Pressure (mmHg)

Comparisons

Participants

Mean difference†

P

Heterogeneity

Mean difference†

P

Heterogeneity

Sub-groups

 

(n)

(n)

(95% CI, Fixed)

value‡

P value‡

(95% CI, Fixed)

value‡

P value‡

Blood pressure 21,22,23,24,25,26,27,28,29,30,31,33,34,35,36,37,38,39,40

19

1336

-4.23[-6.47,-1.99]§

<.001

<.001

-2.53[-3.99,-1.08]§

<.001

<.001

High-quality trials

ACHD vs SAHD 35,40

2

170

-5.72[-8.77,-2.68]

<.001

.54

-2.80[-5.07,-0.54]

.02

.42

Low-quality trials

AC vs SA 24

1

10

-5.00[-12.80,2.80]

.21

-0.50[-4.59,3.59]

.81

AC vs HD 21,22,23,25,28,34,36,39

8

596

0.35[-2.75,3.45]§

.83

.003

-0.05[-2.03,1.94]§

.96

.02

ACHD vs HD 26,30,31,33,37,38

6

404

-7.48[-9.44,-5.52]

<.001

.70

-4.39[-6.57,-2.20]§

<.001

.02

ACCHHD vs CHHD 27

1

96

-10.50[-14.75,-6.25]

<.001

-8.30[-11.82,-4.78]

<.001

AC plus exercise vs Exercise 29

1

60

-11.00[-28.00,6.00]

.20

-7.50[-15.03,0.03]

.05

High-quality trials:

 

3

months after treatment

ACHD vs SAHD 35

1

140

-2.00[-5.98,1.98]

.32

0.00[-3.01,3.01]

1.00

6

months after treatment

ACHD vs SAHD 35

1

140

2.00[-2.31,6.31]

.36

0.00[-2.82,2.82]

1.00

Blood pressure change magnitude 21,34,35,37,40

5

338

-5.98[-9.48,-2.47]§

<.001

<.001

-3.95[-5.19,-2.72]§

<.001

.05

High-quality trials

ACHD vs SAHD 35,40

2

170

-8.38[-9.16,-7.60]

<.001

.27

-4.54[-5.08,-4.00]

<.001

.41

Low-quality trials

AC vs HD 21,34

2

108

-4.41[-10.39,1.56]§

.15

<.00001

-2.27[-6.94,2.41]§

.34

.009

ACHD vs HD 37

1

60

-2.30[-6.14,1.54]

.24

-2.70[-6.13,0.73]

.12

High-quality trials:

 

3

weeks after treatment

AC vs SA 32

2

188

0.20[-0.74,1.14]

.68

1.00

-0.00[-1.96,1.96]§

1.00

<0.001

*AC indicates acupuncture; SA, sham acupuncture; HD, antihypertensive drug; ACHD, acupuncture plus antihypertensive drug; SAHD, sham acupuncture plus anti- hypertensive drug; CHHD, Chinese herbal medicine plus antihypertensive drug; ACCHHD, acupuncture plus Chinese herbal medicine and antihypertensive drug. †Values<0 favor acupuncture; values>0 favor control. ‡ANOVA P value for difference across groups. §Random-effect model.

ture were reported. The minor adverse effects observed were temporary: spot bleeding, feeling of pain in the skin-puncture site, and some patients, temporary giddiness.

DISCUSSION Main Findings Blood Pressure–lowering Effect Although limited rigorous trials preclude definitive conclu- sions, this meta-analysis found that acupuncture therapy is asso- ciated with BP-lowering benefits in essential hypertension. One trial 35 had a significant BP-lowering effect at the endpoint; how- ever, the effect had disappeared at 3 months after the treatment’s completion. In another active trial, 32 the BP measured 3 weeks after the treatment’s completion was indistinguishable from that associated with sham treatment. The benefit potentially could not be explained by durability effect as hypertension is typically a chronic and discrete condition. In addition, the treatment application of one term was too short, and sessions were too infrequent for conclusion.

Effect of Prescription Adhering to the Principles of Syndrome Differentiation in TCM Diagnosis The process of diagnosis in TCM is based on an understand- ing of the imbalance between the forces of yin and yang and syn- drome differentiation. 11 This meta-analysis found that the acupuncture therapy using prescription adhering to the princi-

ples of syndrome differentiation lowered BP in essential hyper- tension, whereas the acupuncture therapy not using prescription adhering to the principles of syndrome differentiation did not. These findings have limitations, however.

Strengths and Weaknesses of Included Trials The majority of trials evaluated failed to specify many fac- tors: whether allocation was concealed, randomization was ade- quate, outcome assessors were blinded, lost follow-up was described completely, or whether data were analyzed on an intention-to-treat basis. Only four trials 24,32,35,40 described the methods of blinding and randomization used. One trial had different clinicians perform diagnosis, acupuncture treatment, and outcomes assessment. 32 Another trial separated patients based on a language barrier between clinicians and patients. 35 Although these two trials sought isolating methods that may have overcome any potential bias attributable to clinician expectations, doing so is problemat- ic. There is only a slim chance that the care provider may be blinded to treatment assignments in acupuncture clinical trials because the acupuncturist can generally distinguish whether he or she is using sham needles and nonacupoints. Acupuncture is known to be associated with behaviors embedded in medical ritu- als. 41 In addition, we think that a clinician should be motivated primarily to perform a successful procedure for all patients rather than to show that acupuncture is an effective adjuvant procedure

FIGURE 2 Efficacy of Acupuncture on Blood Pressure

Systolic Blood Pressure (SBP)

Systolic Blood Pressure (SBP) Acupuncture Control Mean Difference Mean Difference Study or Subgroup Mean SD
Systolic Blood Pressure (SBP)
Acupuncture
Control
Mean Difference
Mean Difference
Study or Subgroup
Mean
SD
Total
Mean
SD
Total
Weight
IV, Random, 95% Cl
IV, Random, 95% Cl
1.1.1
ACHD vs SAHD (high-quality trials)
Flachskampf FA, 2007
Yin C, 2007
Subtotal (95% CI)
125.0
12.0
72
130.0
11.0
68
6.70%
122.1
6.6
15
129.1
7.5
15
5.80%
87
83
12.50%
-5.00 [-8.81, -1.19]
-7.00 [-12.06, -1.94]
-5.72 [-8.77, -2.68]
Heterogeneity: Tau²=0.00; chi²=0.38, df=1 (P=.54); I²=0%
Test for overall effect: Z=3.69 (P=.0002)
1.1.2
AC vs SA
Kraft K, 1999
Subtotal (95% CI)
Heterogeneity: Not applicable
Test for overall effect: Z=1.26 (P=.21)
142.0
7.0
5
147
5.5
5
4.10%
5
5
4.10%
-5.00 [-12.80, 2.80]
-5.00 [-12.80, 2.80]
1.1.3
AC vs HD
Chen YF, 2000
Dan Y, 1998
Guo YH, 2007
He XW, 1994
Wang LY, 2006
Wang X, 2007
Wu QM, 2003
Yin ZF, 1994
Subtotal (95% CI)
135.6
12.4
35
129
11
35
5.50%
128.1
12.8
26
128
12
26
4.70%
138.7
9.0
40
133.4
9.6
40
6.50%
144.2
31.1
18
143.7
15.4
30
1.70%
162.2
3.5
30
164.4
3.6
30
7.80%
116.3
11.9
65
119
13.6
63
6.20%
145.7
16.5
40
144
17
40
4.40%
136.5
11.9
48
142
13
30
5.40%
302
294
42.20%
6.60 [1.11, 12.09]
0.10 [-6.64, 6.84]
5.30 [1.22, 9.38]
0.50 [-14.89, 15.89]
-2.20 [-4.00, -0.40]
-2.70 [-7.13, 1.73]
1.70 [-5.64, 9.04]
-5.50 [-11.24, 0.24]
0.35 [-2.75, 3.45]
Heterogeneity: Tau²=11.60; chi²=21.72, df=7 (P=.003); I²=68%
Test for overall effect: Z=0.22 (P=.83)
1.1.4
ACHD vs HD
Hu LH, 2004
Hu LH, 2007
Huang F, 2007
Jiang XL, 2003
Wang C, 2006
Zhang YL, 2005
Subtotal (95% CI)
129.0
8.0
45
135.0
9.0
45
6.90%
131.8
9.6
30
137.3
9.8
30
5.90%
128.9
13.1
30
140.0
18.0
30
4.00%
137.6
16.1
30
145.1
9.2
30
4.80%
126.1
7.7
30
134.7
6.5
29
6.80%
127.8
13.3
45
137.7
14.1
30
5.00%
210
194
33.40%
-6.00 [-9.52, -2.48]
-5.50 [-10.41, -0.59]
-11.10 [-19.07, -3.13]
-7.50 [-14.14, -0.86]
-8.60 [-12.23, -4.97]
-9.90 [-16.27, -3.53]
-7.48 [-9.44, -5.52]
Heterogeneity: Tau²=0.00; chi²=3.02, df=5 (P=.70); I²=0%
Test for overall effect: Z=7.49 (P<.00001)
1.1.5
ACCHHD vs CHHD
Song YM, 2003
Subtotal (95% CI)
Heterogeneity: Not applicable
126.8
9.0
64
137.3
10.5
32
6.40%
64
32
6.40%
-10.50 [-14.75, -6.25]
-10.50 [-14.75, -6.25]
Test for overall effect: Z=4.84 (P<.00001)
1.1.6
AC plus exercise vs Exercise
Zhao DJ, 2003
Subtotal (95% CI)
Heterogeneity: Not applicable
Test for overall effect: Z=1.27 (P=.20)
Total (95% CI)
129.0
36.0
30
140.0
31.0
30
1.40%
30
30
1.40%
-11.00 [-28.00, 6.00]
-11.00 [-28.00, 6.00]
698
638
100%
-4.23 [-6.47, -1.99]
Heterogeneity: Tau²=15.84; chi²=69.11, df=18 (P<.00001); I²=74%
-20
-10
0
10
20
Favors
Favors
Test for overall effect: Z=3.70 (P=.0002)
acupuncture
control

FIGURE 3 Efficacy of Acupuncture on Blood Pressure Change Magnitude

3.1 Systolic Blood Pressure (SBP)

Systolic Blood Pressure (SBP)

Acupuncture

 

Control

 

Mean Difference

Mean Difference

Study or Subgroup

Mean

SD

Total

Mean

SD

Total

Weight

IV, Random, 95% Cl

IV, Random, 95% Cl

2.1.1

ACHD vs SAHD (High quality trials)

 

Flachskampf FA, 2007 Yin C, 2007 Subtotal (95% CI)

-6.5

2.3

72

1.8

2.5

68

22.50%

-8.30 [-9.10, -7.50] -10.80 [-15.19, -6.41] -8.58 [-10.13, -7.03]

(95% CI) -6.5 2.3 72 1.8 2.5 68 22.50% -8.30 [-9.10, -7.50] -10.80 [-15.19, -6.41] -8.58
   

-14.8

5.5

15

-4.0

6.7

15

16.80%

 

87

83

39.30%

Heterogeneity: Tau2=0.54; chi2=1.21, df=1 (P=.27); I2=17%

 

Test for overall effect: Z=10.86 (P<.00001)

 

2.1.2

AC vs HD

He XW, 1994 Wang LY, 2006 Subtotal (95% CI)

-27.2

3.8

18

-19.7

3.0

30

21.10%

-7.50 [-9.56, -5.44] -1.40 [-2.98, 0.18] -4.41 [-10.39, 1.56]

 
Subtotal (95% CI) -27.2 3.8 18 -19.7 3.0 30 21.10% -7.50 [-9.56, -5.44] -1.40 [-2.98, 0.18]

-13.7

2.4

30

-12.3

3.7

30

21.80%

 
 

48

60

42.80%

 
  48 60 42.80%  

Heterogeneity: Tau2=17.73; Chi2=21.26, df=1 (P<.00001); I2=95%

 

Test for overall effect: Z=1.45 (P=.15)

 

2.1.3

ACHD vs HD

 

8.6

30

-13.3

6.4

30

17.90%

-2.30 [-6.14, 1.54] -2.30 [-6.14, 1.54]

 

Hu LH, 2007 Subtotal (95% CI) Heterogeneity: Not applicable Test for overall effect: Z=1.18 (P=.24) Total (95% CI)

-15.6

 

30

30

17.90%

 
  30 30 17.90%  

165

173

100%

-5.98 [-9.48, -2.47]

165 173 100% -5.98 [-9.48, -2.47]

Heterogeneity: Tau2=14.07; chi2=67.39, df=4 (P<.00001); I2=94% Test for overall effect: Z=3.34 (P=.0008)

-20

-10

0

10

20

Favors

Favors

acupuncture

control

3.2 Diastolic Blood Pressure (DBP)

Diastolic Blood Pressure (DBP)

Acupuncture

 

Control

 

Mean Difference

Mean Difference

Study or Subgroup

Mean

SD

Total

Mean

SD

Total

Weight

IV,Random, 95% Cl

IV, Random, 95% Cl

 

ACHD vs SAHD (High quality trials)

 

2.2.1

   

Yin C, 2007 Flachskampf FA, 2007 Subtotal (95% CI)

-6.9

3.7

15

-1.1

4.7

15

11.90%

-5.80 [-8.83, -2.77] -4.50 [-5.05, -3.95] -4.54 [-5.08, -4.00]

 
(95% CI) -6.9 3.7 15 -1.1 4.7 15 11.90% -5.80 [-8.83, -2.77] -4.50 [-5.05, -3.95] -4.54
 

-3.8

1.6

72

0.7

1.7

68

38.70%

 

87

83

50.60%

Heterogeneity: Tau2=0.00; chi2=0.69, df=1 (P=.41); I2=0%

 

Test for overall effect: Z=16.52 (P<.00001)

 

2.2.2

AC vs HD

He XW, 1994

-16.4

2.3

18

-12.0

1.8

30

29.40%

-4.40 [-5.64, -3.16] 0.40 [-2.98, 3.78] -2.27 [-6.94, 2.41]

 
He XW, 1994 -16.4 2.3 18 -12.0 1.8 30 29.40% -4.40 [-5.64, -3.16] 0.40 [-2.98, 3.78]

-6.8

7.2

30

-7.2

6.1

30

10.10%

   
 

Wang LY, 2006 Subtotal (95% CI)

 

48

60

39.50%

 
Wang LY, 2006 Subtotal (95% CI)   48 60 39.50%    
 

Heterogeneity: Tau2=9.83; chi2=6.84, df=1 (P=.009); I2=85%

 

Test for overall effect: Z=0.95 (P=.34)

 

2.2.3

ACHD vs HD

 

-9.5

8.3

30

-6.8

4.8

30

9.90%

-2.70 [-6.13, 0.73] -2.70 [-6.13, 0.73]

 
   

Hu LH, 2007 Subtotal (95% CI) Heterogeneity: Not applicable Test for overall effect: Z=1.54 (P=.12) Total (95% CI)

 

30

30

9.90%

 
CI) Heterogeneity: Not applicable Test for overall effect: Z=1.54 ( P =.12) Total (95% CI)  

165

173

100.00%

-3.95 [-5.19, -2.72]

165 173 100.00% -3.95 [-5.19, -2.72]

Heterogeneity: Tau2=0.95; chi2=9.65, df=4 (P=.05); I2=59% Test for overall effect: Z=6.27 (P<.00001)

-20

-10

0

10

20

Favors

Favors

acupuncture

control

TABLE 4 Effect of Acupuncture Using Prescription Adhering to the Principles of “Syndrome Differentiation” in TCM Diagnosis

 

Systolic Blood Pressure (mmHg)

Diastolic Blood Pressure (mmHg)

 

Comparisons

Participants

Mean difference*

 

Heterogeneity

Mean difference* (95% CI, Fixed)

 

Heterogeneity

Subgroups

(n)

(n)

(95% CI, Fixed)

P value†

P value†

P value†

P value†

Prescription adhering to the principles of syndrome differentiation

 

Blood

pressure 21,22,23,24,27,30,31,35,37,38,40

11

739

-6.46 [-8.04,-4.87]

<.001

.33

-3.07 [-4.17,-1.96]

<.001

.11

Blood pressure change

 

magnitude 21,35,37,40

4

278

-7.37 [-9.60,-5.14]‡

<.001

.01

-4.48 [-4.97,-3.99]

<.001

.62

Prescription not adhering to the principles of syndrome differentiation

 

Blood

pressure 25,26,28,29,33,34,36,39

8

597

-1.55

[-5.39,2.29]‡

.43

<.001

-2.12 [-4.97,0.73]‡

.15

<.001

Blood pressure change magnitude 34

1

60

-1.40 [-2.98,0.18]

.08

0.40 [-2.98,3.78]

.82

*Values<0 favor acupuncture; values>0 favor control. †ANOVA P value for difference across groups. ‡Random-effect model.

in the field. Therefore, it seems better in reality that the diagnos- ing procedure for selection of the acupuncture formula is left up to the clinician who will conduct the acupuncture. Three trials 24,32,35 used an invasive sham acupuncture tech- nique that involved the insertion of a similar number of needles at

a similar depth but at nonacupoints. On the other hand, one trial 40

used noninvasive sham acupuncture with a needle device that works like a retractable magic sword: the needle appears to be pen- etrating the skin, and the patient sees and feels a sensation of nee- dle penetration, but the needle is actually retracted up the needle shaft. 42,43 In the majority of acupuncture clinical research, sham acupuncture differs only minimally from active acupuncture and may expose subjects to a degree of risk similar to that of the active procedure. 44 The invention and application of sham-acupuncture devices have greatly enhanced the quality of acupuncture clinical research. Although many debates still exist on whether such nee- dles are the best control for an RCT of acupuncture, 45,46 noninvasive sham acupuncture seems to be more distinguishable from genuine acupuncture than invasive sham acupuncture. 42,43,47 The frequency of sessions in acupuncture therapy varies according to cultural and geographical differences in addition to issues of time and cost. Almost all trials published in China had treatments delivered daily, whereas trials in other locations deliv- ered treatment with lesser frequency.

One trial 48 assessing the efficacy of acupuncture for both physical and psychiatric symptoms and conditions demonstrated that beliefs or expectations could exert a powerful influence on

treatment effects. One 40 of the included trials assessed this using

a subjective measurement with a value on the anticipation/satis-

faction scale. This increased only in real acupuncture therapy, corresponding to the period when the BP-lowering effect of acu- puncture was prominent. This study shows that even if psycho- social benefits have existed in the treatment, it is still based on the efficacy of acupuncture.

Twenty-four-hour ambulatory blood pressure monitoring (APBM) devices were used for measurement in six out of the 20 trials. In the measurement of BP, the improvements for future study include using ABPM devices to obtain more accurate recordings of dynamic changes within a 24-hour period.

Limitations This systematic review is to assess the BP-lowering effect of acupuncture in treatment of essential hypertension with a com- prehensive search and is the first to assess the difference between the effects of prescription adhering to the principles of syndrome differentiation usage and nonusage. The main limitation of this review was the paucity of high- quality RCTs. Three rigorous trials were published in English journals, whereas most of the trials using inadequate methodolo- gy were published in Chinese journals. In the complementary and alternative medicine field, 1 study 49 found that trials pub- lished in low–impact factor or non–impact factor journals were more likely to report positive results than those published in high-impact mainstream medical journals and that the quality of the trials was also associated with the journal of publication. Poor-quality trials generally had less precise estimates of effect and tended to overestimate the effect. High-quality trials have been defined as research that is carried out in a way that allows readers to trust the results 50 with subgroup analyses being per- formed separately. Another limitation is “included heterogeneity.” Substantial heterogeneity is indicated by the fact that the control interven- tions differed across trials. This heterogeneity is probably caused by differential formula of sham needles or differential selection of antihypertensive medications across trials. The selected acu- points were not similar across trials, and the number of used acupoints differed greatly across trials. Also, the duration of treatment and frequency of sessions differed across trials.

FIGURE 4 Effect of Acupuncture Using Prescription Adhering to the Principles of “Syndrome Differentiation” in TCM Diagnosis FIGURE 4-1 Systolic Blood Pressure (SBP)

Systolic blood pressure (SBP) Study or Subgroup

Acupuncture

   

Control

 

Mean Difference IV, Random, 95% Cl

 

Mean Difference IV, Random, 95% Cl

Mean

SD

Total

Mean

SD

Total

Weight

3.1.1

Blood pressure

He XW, 1994

144.2

31.1

18

143.7

15.4

30

0.20%

0.50 [-14.89, 15.89] -5.50 [-11.24, 0.24] 0.10 [-6.64, 6.84] -5.00 [-12.80, 2.80] -10.50 [-14.75, -6.25] -6.00 [-9.52, -2.48] -9.90 [-16.27, -3.53] -5.50 [-10.41, -0.59] -7.00 [-12.06, -1.94] -5.00 [-8.81, -1.19] -11.10 [-19.07, -3.13] -6.46 [-8.04, -4.87]

   
   

Yin ZF, 1994

136.5

11.9

48

142

13

30

1.30%

   
 

Dan Y, 1998

128.1

12.8

26

128

12

26

0.90%

 
 
 

142

7

5

147

5.5

5

0.70%

 
   

Kraft K, 1999 Song YM, 2003 Hu LH, 2004

126.8

9

64

137.3

10.5

32

2.40%

 

129

8

45

135

9

45

3.50%

Zhang YL, 2005 Hu LH, 2007

127.8

13.3

45

137.7

14.1

30

1.10%

131.8

9.6

30

137.3

9.8

30

1.80%

     

Yin C, 2007

122.1

6.6

15

129.1

7.5

15

1.70%

 

125

12

72

130

11

68

3.00%

   
 

Flachskampf FA, 2007 Huang F, 2007 Subtotal (95% CI)

128.9

13.1

30

140

18

30

0.70%

Flachskampf FA, 2007 Huang F, 2007 Subtotal (95% CI) 128.9 13.1 30 140 18 30 0.70%
 

398

341

17.10%

Heterogeneity: chi 2 =11.37, df=10 (P=.33); I 2 =12% Test for overall effect: Z=7.99 (P<.00001)

 

3.1.2

Blood pressure change magnitude

 

He XW, 1994

-27.2

3.8

18

-19.7

3

30

10.10%

-7.50 [-9.56, -5.44]

-15.6

8.6

30

-13.3

6.4

30

2.90%

-2.30 [-6.14, 1.54] -10.80 [-15.19, -6.41] -8.30 [-9.10, -7.50] -8.06 [-8.78, -7.34]

 

Hu LH, 2007 Yin C, 2007 Flachskampf FA, 2007

-14.8

5.5

15

-4

6.7

15

2.20%

-6.5

2.3

72

1.8

2.5

68

67.60%

-6.5 2.3 72 1.8 2.5 68 67.60%  
 
 

135

143

82.90%

533

484

100.00%

-7.78 [-8.44, -7.13]

Subtotal (95% CI) Heterogeneity: hi 2 =10.79, df=3 (P=.01); I 2 =72% Test for overall effect: Z=21.94 (P<.00001) Total (95% CI) Heterogeneity: chi 2 =25.43, df=14 (P=.03); I 2 =45%

Test for overall effect: Z=23.28 (P<.00001)

-20

-10

0

10

20

Favors

Favors

acupuncture

control

FIGURE 4-2 Diastolic Blood Pressure (DBP)

Diastolic blood pressure (DBP) Study or Subgroup

Acupuncture

   

Control

 

Mean Difference IV, Random, 95% Cl

Mean Difference IV, Random, 95% Cl

Mean

SD

Total

Mean

SD

Total

Weight

3.2.1

Blood pressure

 

He XW, 1994