Article Text


Somatosensory stimulation and assisted reproduction
  1. David Carr
  1. Correspondence to Dr David Carr, UCL Institute for Women's Health, University College London, 86–96 Chenies Mews, London WC1E 6HX, UK; davidcarr{at}

Statistics from

The role of somatosensory stimulation (acupuncture and related techniques) as an adjunct to assisted reproductive technology (ART) has been hotly debated over the past decade. In the last 6 years there have been no fewer than 12 meta-analyses of 34 randomised controlled trials (table 1). Systematic and narrative reviews of the literature often report conflicting findings and opinions. Despite this controversy, subfertility remains one of the most common reasons that women consult an acupuncturist.47 Meta-analysis of acupuncture trials is often challenging due to methodological diversity, and this is especially true for in vitro fertilisation (IVF) due to marked differences in timing (treatment given during the follicular phase, at the time of oocyte retrieval and/or embryo transfer (ET), and/or as luteal phase support) and the variety of interventions including manual acupuncture, electroacupuncture (EA), transcutaneous electrical acupuncture point stimulation (TEAS) and laser acupuncture. Various outcome measures have been used ranging from biochemical pregnancy rate to live birth rate (LBR), depending on the duration of follow-up. LBR (‘take home baby rate’) is considered to be the ideal primary outcome measure and is clearly most important to patients. Interestingly, concordance between studies with respect to point location is relatively high. The acupuncture points targeted generally correspond anatomically to the segmental innervation of the uterus and ovaries, irrespective of whether the primary approach to point selection is traditional or neurophysiological.

Table 1

Randomised controlled trials and meta-analyses of acupuncture as an adjunct to assisted reproduction

The paper by Shuai et al48 used TEAS rather than needle acupuncture and examined its efficacy and mechanisms of action in women undergoing frozen-thawed ET. TEAS significantly increased the clinical pregnancy rate (CPR) and LBR relative to a sham intervention (mock TEAS). To date there has been one other (positive) trial of TEAS in IVF43 and a third is currently underway.49 Regarding the underlying mechanism, it has been hypothesised that acupuncture-like interventions improve endometrial receptivity, based on the observations of Stener-Victorin et al50 and Ho et al34 that EA reduces impedance to flow in the uterine arteries. It is often inferred, rightly or wrongly, that enhanced uterine blood flow equals improved endometrial blood flow. In support of this assumption, acupuncture has been shown to increase endometrial thickness compared with sham acupuncture and standard care.44 However, Shuai et al48 are the first to have shown positive effects on contemporary ultrasound markers of endometrial receptivity (three-dimensional endometrial and subendometrial vascularisation index and presence of a triple-line pattern) plus HOXA10 protein expression, a novel molecular marker. Interestingly, all these studies have used serial treatments that incorporate the follicular phase of the cycle. The most recent systematic review by Shen et al12 only found evidence of benefit when follicular phase treatment was combined with treatment around ET, and not when timed around ET alone. It is argued that the latter approach, which typically consists of a single treatment before and one after ET (variations of the Paulus protocol14), is insufficient in duration and ‘dose’ to have any major effect on endometrial receptivity and/or blood flow, and that putative effects on pregnancy rates in this context may be due to inhibition of post-transfer uterine contractions51 and/or reduction of patient anxiety,52 both of which may adversely affect pregnancy rates.53 ,54

The efficacy of TEAS suggests that penetration of the skin or muscle is not required in order to effect endometrial receptivity and pregnancy rates, and this is supported by experiments on anaesthetised rats, wherein a measurable increase in uterine blood flow was observed during noxious pinching and innocuous brushing of the hind limb and perineum (regions innervated by the sacral segments).55 This calls into question the validity of sham controls that elicit strong cutaneous stimuli, such as the Streitberger needle, which is not generally considered to be physiologically inert.56 In their recent meta-analysis, Zheng et al8 performed subgroup analyses comparing pregnancy rates of patients receiving verum acupuncture versus non-Streitberger controls and demonstrated a significant increase in CPR (OR 1.34, 95% CI 1.08 to 1.67) and LBR (OR 1.63, 95% CI 1.16 to 2.30). When trials using Streitberger controls were included the effects on LBR disappeared and the effects on CPR were smaller, albeit still statistically significant (OR 1.22, 95% 1.01 to 1.47). The relevance of sham-controlled trials in IVF research has been debated at length (see Manheimer57 for a comprehensive overview), particularly in view of the fact that pregnancy is an objective outcome measure for which blinding is less critical.58–61

When sham-controlled trials are included, the three latest systematic reviews demonstrate no pooled benefit of acupuncture10–12 although Shen et al12 did find evidence of benefit with follicular phase treatment plus treatment around ET. Manheimer et al11 reported that acupuncture treatment appears to be more successful in studies with lower baseline pregnancy rates. Their subgroup analysis appears to have been heavily influenced by one particular negative (and grossly underpowered) trial by Craig et al,28 ,46 in which the baseline pregnancy rate in the untreated control group was extraordinarily high at 69.6% while the rate following acupuncture (Paulus protocol) was similar to previous studies using the same methodology (43.8% vs 42.5%,14 43.0%17 and 43.8%32). However, assuming that the relationship is a genuine one, there are some potential explanations. Hypothetically, if an IVF unit could maximise its pregnancy rate from a biological perspective, there would be limited scope for acupuncture to increase it further as it is unlikely to have supraphysiological effects. However, in the presence of a ‘performance gap’ between actual and achievable pregnancy rates, there may be a greater potential for a shift. Alternatively, this observation may represent a statistical artifact as the lower the baseline pregnancy rate, the smaller the sample size required to show a difference (figure 1). Advanced maternal age has a powerful influence on baseline pregnancy rates; IVF success rates range from 1.9% in the over-45s to 32.2% in the under-35s (Human Fertilisation and Embryology Authority 2010 data).

Figure 1

Effect of baseline pregnancy rate on sample size calculations. The graph shows the relationship between the baseline pregnancy rates in clinical trials and the number of patients required per group to show a significant difference (at 90% power and significance level of p<0.05) at three fixed effect sizes of 5%, 10% and 15%, respectively. Sample size calculation was performed using freely available power calculators for binary outcome superiority trials (

Ultimately, further research is required before somatosensory stimulation techniques such as acupuncture or TEAS can be widely adopted for ART. Meanwhile, agreement is needed about the validity of sham controls, which may otherwise obscure the evidence.


View Abstract


  • Twitter Follow David Carr at @IntegratedOBGYN

  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.