Department of Community Medicine
Mount Sinai Medical Center
1 Gustave Levy Place
New York, NY 10029
The weight of evidence now suggests that physical activity during pregnancy is safe and even beneficial in some circumstances. As a result the American College of Obstetricians and Gynecologists encourages pregnant women to develop an activity regimen in conjunction with their obstetricians. 1
Just when clarity and consensus begin to emerge, along comes a study from Denmark 2 that raises new questions, mainly because it focuses on a measure that is different from those examined previously. The study concentrates on peak activity levels at critical times in pregnancy. The data appear to show that physical strain around implantation increases the risk of spontaneous abortion. Perhaps surprisingly, the risk from a proportional hazards analysis for miscarriage at 5 or more gestational weeks was found to be strong (RR = 4.8, 95% CI = 2–11.4) while the evidence for an association with miscarriage closer to the time of implantation was quite weak (1.4, 95% CI = 0.6–3.4).
Implantation involves a sequence of events that helps establish a viable pregnancy. It has long been recognized as a period in which sensitivity to insult is high. During this time, the dividing conceptus moves down the fallopian tube to the uterine cavity. As it divides, it forms first the morula and then the blastocyst. Concurrently, the uterus changes from a nonreceptive to a receptive environment in response to changing concentrations of progesterone and estrogen. At an early phase of implantation, the blastocyst attaches to the uterine wall, initiating processes that lead to establishment of a placenta. Implantation occurs between 6 to 9 days after fertilization, when a vascular connection is made between the embryo and the maternal circulation that will serve to maintain the embryo’s viability. 3
The design of the Danish study suggests that the principal focus was on the period around implantation. Subjects were enrolled when they stopped birth control and were asked to collect urine samples for analysis of the hCG patterns that indicate pregnancy or subclinical loss. Such a protocol is challenging and would certainly not have been adopted unless very early loss was a primary endpoint. The fact that information on physical strain was only collected until a pregnancy was diagnosed again suggests a focus on early loss. Data on physical strain are presented up to 30 days after ovulation, but follow-up of exposure did not continue as the pregnancy progressed.
Because the association that emerged was with later miscarriage, the failure to collect information on physical strain after implantation is regrettable. Many women who reported physical strain at implantation may have continued the same activities after their pregnancy was detected. The authors could be attributing an association to physical strain at the time of implantation that is partly accounted for by physical strain at a later stage of pregnancy. Although no other critical period was found in the 30 days post-ovulation, this result does not rule out the existence of additional critical periods for physical strain later in gestation.
To my ears, the term “physical strain” has a diagnostic ring. It implies excessive demands and injurious consequences. There is, however, no good basis for making such judgments. Respondents are asked to rate their level of strain, following a set of examples. The examples of very high physical strain provided in the questionnaire include long distance running, a tennis match, and frequent lifting of heavy loads—strenuous activities, certainly, but not necessarily productive of strain in those conditioned to them. Yet if respondents follow the questionnaire examples religiously, then their reports of strain will not be influenced by their fitness level.
In the definition of physical strain, exercise and work activities are combined, although they have different characteristics and, seemingly, different effects. Workplace demands have greater potential to produce strain, because they may be sporadic but intense and are usually involuntary. Nevertheless, previous studies of workplace activity have not found any compelling evidence of an association with fetal loss. 4–7 Studies of leisure time activity, including a study that considered vigorous exercise in the period around implantation, 8 actually show a reduction in risk of spontaneous abortion with exercise. 8,9
The suggestion that strenuous physical activity at implantation may lead to spontaneous abortion is intriguing. The association reported here, with miscarriage in the late first trimester and after, is strong but also variable because it is based on a rather small number of clinical spontaneous abortions. Beyond sample size is the issue of coherence. It is not easy to envision how strenuous activity could act to cause a suboptimal implantation that ends in a loss 5 or more weeks later, without significantly raising the risk for loss in the early post-implantation period. The authors suggest the lag is explained by the high proportion of chromosomal abnormalities in early abortions. This is an interesting speculation but not convincing in the absence of actual data on anomalies in very early loss, which we do not yet have. Inferences from data on losses at later periods are not a satisfactory substitute for direct observation. Hypotheses need to be formulated about the nature of the insult that might cause strenuous physical activity to interfere with implantation or other functions critical to survival and development of the embryo (eg, changes in hormone concentrations). Delayed implantation might explain the timing, but is physical strain a likely cause?
The role of physical activity in promoting health and well being is increasingly recognized. 10 It has many benefits for a woman and prospective mother. Its safety during pregnancy took years to establish, and that sense of security should not be given up lightly. Thus, while the Danish investigators are to be commended for the many innovative aspects of their study – the notion that in the case of spontaneous abortion, it could be peak values rather than averages that are determinant; the exploration of implantation as a critical period; the inclusion of the full spectrum of fetal loss – we must nevertheless be careful about interpreting the results too strongly. There are internal inconsistencies with respect to the timing of the effect that are not fully explained. The results also disagree with the published studies finding that exercise reduces the risk of spontaneous abortion. Until all of this is resolved, if I were a woman wanting to get pregnant, I would not forego my favorite exercise or shun the physical demands of work.
1. American College of Obstetrics and Gynecology: Exercise during pregnancy and the postpartum period (technical bulletin #189). Washington, DC, ACOG Press, 1994.
2. Hjolland NHI, Jensen TK, Bonde JPE, Henriksen TB, Andersson A-M, Kolstad HA, Ernst E, Giwercman A, Skakkebæk NE, Olsen J. Spontaneous abortion and physical strain around implantation: A follow-up study of first-pregnancy planners. Epidemiology 2000; 11:18–23.
3. Susser M, Stein Z, Kline J. Conception to Birth. New York: Oxford University Press, 1989; 44–46.
4. Eskenazi B, Fenster L, Wight S, English P, Windham GC, Swan SH. Physical exertion as a risk factor for spontaneous abortion. Epidemiology 1994; 5:6–13.
5. Fenster L, Hubbard AE, Windham GC, Waller KO, Swan SH. A prospective study of work-related physical exertion and spontaneous abortion. Epidemiology 1997; 8:66–74.
6. Florack EIM, Zielhuis GA, Pellegrino JEMC, Rolland R. Occupational physical activity and the occurrence of spontaneous abortion. Int J Epidemiol 1993; 22:878–884.
7. Ahlborg G, Bodin L, Hogstedt C. Heavy lifting during pregnancy - a hazard to the fetus? A prospective study. Int J Epidemiol 1990; 19:90–97.
8. Clapp JR. The effects of maternal exercise on early pregnancy outcome. Am J Obstet Gynecol 1989; 161:1453–1457.
9. Latka M, Kline J, Hatch M. Exercise and spontaneous abortion of known karyotype. Epidemiology 1999; 10:73–75.
10. American College of Sports Medicine: Guidelines for Exercise Testing and Prescription. Philadelphia: Lea & Febiger, 1991.