Does Acupuncture Affect Labor and Delivery?
Author: Timothy Gorski
Review of: Tempfer, C., Zeisler, H., Heinzl, H., Hefler, L., Husslein, P., Kainz, C.H. Influence of Acupuncture on Maternal Serum Levels of Interleukin-8, Prostaglandin F2alpha, and Beta-Endorphin: A Matched Pair Study. Obstet Gynecol. 1998; 92:245-248.
A Viennese research team reported last year that "prenatal acupuncture treatment significantly reduced the duration of labor and may be proposed as a valuable tool in prenatal preparation." [1] Their claim is based on an unblinded matched case study, which involved comparing 40 women who had four weekly acupuncture sessions in the last month of their pregnancies with another group of women of similar age and parity (number of previous children, if any) who gave birth at about the same time. The authors assessed intrapartum blood pressures, length of the first and second stages of labor, use of oxytocin and analgesics, frequency of episiotomy and/or lacerations, birth weight, and umbilical cord pH in an effort to demonstrate that acupuncture could affect these variables. Also measured were serum levels of interleukin-8 and prostaglandin F2 hormones thought to be important in cervical ripening and preparation for labor, as well as serum levels of endorphin, a demonstrated marker of the pain and stress of human parturition. [2]
The authors reported that there were no differences between the acupuncture and control groups for any of the measured variables except for the length of the first stage of labor. Among those who received the acupuncture, the first stage of labor was said to be an average of more than two hours with P<.001 on a paired t test. This formed the basis of the authors' major conclusion. Several of the same authors of this report published a similar case-control study in which they also purported to show that the acupuncture-treated patients required less oxytocin during labor. [3]
There are several serious difficulties with this report. The most troublesome is the authors' arbitrary selection of 3 cm dilation for the beginning of the first stage of labor.
Human parturition is divided into three stages, the first being that from the onset of labor until complete dilation, the second lasting until delivery of the newborn, and the third ending with delivery of the placenta. All but the first stage have a sharp beginning and endpoint. The onset of the first stage of labor is often judged in retrospect since the definition of labor is rhythmic uterine contractions accompanied by progressive dilation of the cervix. Nor does cervical dilation proceed at a uniform rate. In general, the early part of the first stage of labor is characterized by very slow cervical dilation, which is referred to as the "latent phase." Later, the "active phase" commences, during which cervical dilation proceeds at a more rapid rate of 1 cm an hour or greater, and sometimes considerably greater. Although the majority of parturients will enter the active phase at around 3 cm dilation, it has been known for some time that up to a third may not enter the active phase until as many as 5 cm. [4] Likewise, although complete cervical dilation is a definite endpoint, it must be determined by means of an invasive vaginal examination. As a result, a parturient who is comfortable can reach the beginning of the second stage of labor without this fact being known to her or her attendants. Likewise, it is also well known that a nearly completely dilated cervix can be reduced by the expulsive efforts of the parturient or, sometimes, even by the digital examination of the attending nurse or physician.
These and other difficulties account for the fact that the classic reports of Friedman [5,6] showed wide variability of the length of the first stage of labor and, especially, of the latent phase of labor. The contributing anatomic and physiologic factors accounting for this remain poorly understood. But these uncertainties severely limit the legitimacy of these authors' major conclusion as to the efficacy and utility of acupuncture in the setting of human parturition, particularly in the face of their failure to demonstrate any differences in the serum markers investigated.
Thus, if the patients in the acupuncture group tended to come in later during the latent phase of labor for any reason-as they well might, for example, if they had an expectation of having an easier course-they would have appeared to have had a shortened first stage of labor by the authors' criteria. Likewise, if the patients in the acupuncture group were checked more frequently-and, again, they may very well have been if there was an expectation on their part and the part of their attendants that their labor would progress more rapidly-they may have appeared to have had a shortened labor. These are quite reasonable suppositions given that the approximately two hours difference between the length of the first stage of labor for the authors' two groups is dwarfed by the much larger discrepancies-up to 10 hours-between published measurements of the average duration of labor. [7,8,9,10]
In addition, a large number of other factors may influence the duration of labor, and particularly the latent phase of the first stage of labor. One of these is obstetric analgesia. Although the authors of this report, interestingly enough, do not seem troubled by the fact that the patients who received acupuncture did not have a reduced need for analgesics, these women may have received it later than those in the matched group if there was an expectation that less analgesia would be necessary. This information is, unfortunately, not provided.
The authors do show that, between the two groups of patients, there is good comparability for the average age, parity (previously delivered children), gestational age, blood pressures, and umbilical artery cord pH. But there is a notable difference in the birth weight of the newborns in the two groups, the acupuncture-treated group averaging 3,304 grams and the matched group 3,002 grams, nearly a 3/4-pound difference. The authors do not address this discrepancy. (It seems puzzling why they did not take the opportunity of suggesting that acupuncture treatments may have been responsible for it.) But one could easily speculate that it may have influenced their findings with respect to labor duration. Newborn birth weight, for example, even if it exerted no direct effects, could reflect general maternal nutritional status and condition, use of tobacco and other substances, and other factors.
Also left unmentioned in this study are other variables relating to the exact presentation of the fetuses and other details of the labor courses for the two groups, an omission that also applies to the companion report concerning the use of oxytocin in acupuncture-treated parturients as compared to matched cases. But even leaving all of this aside, it is well known that nonspecific factors influence the duration of labor. The presence of a supportive companion, for example, has been reported to shorten labor by anywhere from an hour [11] to more than 10 hours, [12] and a meta-analysis of 11 such studies was also supportive of this finding. [13] Considering these demonstrated effects of psychosocial support, it is clearly unacceptable that in a study such as this that the control group was not afforded any sort of credible placebo intervention, even without blinding. For even if the major conclusion of the authors can be believed, and if, as seems likely, any beneficial effects of acupuncture in this setting are exerted through nonspecific (psychosomatic or mind-body) means, there are surely more personal, more pleasant, more rational, and perhaps more effective ways of eliciting this response. Psychotherapy, stress reduction education, being afforded additional help at home or time off from work, or any sort of additional, credible intervention designed to raise expectations about parturition could conceivably exert the same effects as the invasive placement of needles for the purpose of altering the flow of mystical "chi energy."
Although the authors of this report do not give credence to the philosophical "chi energy" notions on which acupuncture is founded, they freely speculate on possible neuroendocrinological mechanisms for how acupuncture might shorten labor. Yet, given the above considerations and also the authors' own failure to demonstrate alterations in serum levels of their chosen markers, it would seem more prudent to give consideration to more prosaic explanations including the null hypothesis and/or nonspecific effects. The authors could then directly investigate their ideas about the role of neuroendocrine factors in obstetrics.
Sadly, this work and others like it are predictably seized upon as validating the "chi energy" notions of a prescientific vitalistic philosophy of medicine. This is untrue, of course. But the element of acupuncture in this setting does have the unfortunate effect of distorting, distracting from, and confusing more rational and well-substantiated ideas about the still-to-be-discovered anatomical, physiological, psychological, and biochemical details of human parturition.
