We believe the medical literature is frankly inconclusive on whether acupuncture helps women to deliver a healthy baby. Acupuncture is inherently hard to study and as a result we’ve seen a spectrum of conflicting results from trials. That said, it appears that acupuncture is:
Effective at mitigating patient anxiety during treatment
Useful compared to “no acupuncture” but not versus a “sham” or “placebo”
“Dose dependent” and more likely to be effective the more often it’s administered
Before we dive into whether acupuncture improves a woman's chances to have a baby, we should get familiar with the reasons acupuncture is hard to study according to western best practices. We’ll refer back to this when we consider what each study tells us or doesn’t tell us. Here are the issues in order of importance.
Confusion Around Control Arm (Most Important):
In assessing whether acupuncture “works,” we need to compare it to an alternative. Compared to doing no acupuncture (“usual care”), acupuncture generally rates favorably. But compared to something patients perceive to be acupuncture (but is not), known as a “sham” or “placebo” acupuncture, acupuncture often looks ineffective. There is real debate as to which should be our comparator arm.
Inconsistent Timing & Volume of Treatment:
During an IVF cycle, treatment can be performed at many phases (e.g. leading up to a transfer, around the time of transfer, in the weeks after) and lining up the clinical data based on when treatment occurred is both important and hard. Most studies involve a total of 2 treatments performed on the day of embryo transfer, which is less treatment than many experts consider necessary.
Not "Real-World" Treatment:
Acupuncturists typically have a lot of leeway in where they place their needles. In the majority of these studies, the acupuncture treatment points were placed as per the specific protocol the investigators chose. Fixed protocols do not take into account the unique needs of an individual patient and thus do not reflect the full impact of a tailored approach patients would receive outside of a study.
The available data is old and comes from an era when IVF success rates were lower. That matters because when acupuncture shows a benefit, it’s often in studies where IVF patients not given acupuncture had poor results. Today, many of those same patients might have fared better making it harder for acupuncture to show additional benefit.
We believe patients should be focused on whether doing acupuncture helps them have a live birth. Unfortunately, many acupuncture studies only look at whether acupuncture improves “pregnancy rates” (or some version thereof), which is useful but falls short of what we all truly care about: taking home a baby.
While the impact of acupuncture has been widely studied (over 40 studies and 10 meta-analysis) we do not believe it’s been well-studied (for the reasons mentioned above). With the data we currently have, this is how we’d summarize the findings:
Acupuncture clearly lowers patient anxiety and duress during fertility treatment
Greater volumes of treatment are more like to improve outcomes rather than treatment just around embryo transfer
Acupuncture is more likely to show benefit versus “no acupuncture” but not versus a “sham” or “placebo” approach
Undergoing acupuncture during fertility treatment clearly mitigates stress. This is true regardless of whether acupuncture was administered just around the time of embryo transfer or during a longer duration of time. Below we list the better run studies on the subject and most come to a similar conclusion.
Anxiety and stress and can be excruciating and addressing this alone may make acupuncture “effective.” Psychological anxiety also impedes a patient’s ability to get more treatment.
This can limit a patient's likelihood of eventually deliverying, especially if she discontinues treatment at an early age or in the first few cycles.
One single-center study measured the impact of doing acupuncture solely around the time of embryo transfer and compared that with results from patients who underwent 12 or more treatments (in additional to herbal therapy). Investigrators noted significantly higher live birth rates associated with greater treatment. While this study has challenges (single center, not prospectively randomized) we think its endpoint (live birth), sample size (500+ patients) and level of statistical significance (P < .03) make it noteworthy.
The data generally shows acupuncture has a benefit compared with “usual care” but not when compared to “sham” or “placebo” (treatments that mimic acupuncture).
We think the best analysis comes from a 2013 meta-analysis by Manheimer who analyzed 16 studies, which between them showed major disagreement, and came to the below conclusion.
This raises a fundamental question about whether one should compare acupuncture against “usual care” or “sham / placebo.”
Acupuncture advocates prefer to compare it to “usual care” with the following rationale, which in our minds is credible:
By going to see a practitioner and relaxing in a quiet setting, patients who undergo “sham” treatment likely receive some benefit. As a result, this group is in fact receiving part of the therapy of acupuncture and so they are not truly a “placebo” or “sham” comparator.
In the real world, patients only choose between getting acupuncture or not getting it. Designing a trial to compare acupuncture against a non-practical option (e.g. receiving “sham” acupuncture) produces irrelevant information.
Of the 40-plus studies done the subject, we believe the ones below are the most credible.
Trying To Conceive, No IVF
Women with PCOS often suffer from ovulation infrequency and one small, but credible, study showed that women who underwent 10 - 13 weeks of acupuncture saw ovulation improvement versus a control group.
However, the study stopped short of looking at women actually trying to conceive. Two studies went a step further and neither showed acupuncture had an impact.
Trying To Conceive, Using IVF
We have two studies to work with here and both are imperfect. The best run trial by Rashidi showed that acupuncture patients had a higher ongoing pregnancy rate, but the difference was not large enough to meet statistical significance, likely due to the study’s small size.
A meta-analysis by Jo pools multiple studies on the subject (including Rashidi’s) and while this beefs up the sample size we have to work with, the variability in the groups studied creates chaos in the data. Jo’s study showed no difference in live birth rate between the acupuncture and non-acupuncture groups.
There is little published literature on the effects of acupuncture on men and its ability to improve their chances of fathering a child. Unfortunately, the best data we have addresses whether acupuncture improves a man’s semen parameters which does not necessarily translate into a higher live birth rate.
What’s more, for men suffering from non-obstructive Azoospermia (those who produce no sperm), we believe acupuncture is unlikely to make any difference.
In one small, but relatively well-run study (12 treatments, prospective study with good randomization) in patients with extremely low sperm concentrations (<1 million per mL) acupuncture showed a statistically significant decrease in sperm concentration while having a slight improvement on the percentage of sperm that are able to swim (motility). In populations with such low sperm volumes, neither change is likely to impact the chances of a couple’s ability to conceive.
When Chinese investigators performed a meta-analysis on the studies addressing this topic more broadly (and not just in those men who have severely low concentrations) they concluded the evidence was “insufficient” on whether acupuncture improves semen parameters.
While the data is scant on whether acupuncture helps women with endometriosis conceive, there is reason to believe that endometriosis patients benefit from acupuncture in the form of pain relief. One small, randomized control trial out of China noted that women with endometriosis experienced lower pain rates related to Dysmenorrhoea on a 15 point scale by a wide, 5 point margin.