Ideally, science is supposed to be self-correcting. Scientists test their hypotheses, and after repeated testing and replication, hypotheses that lack support should be rejected. But sometimes based on limited evidence, scientists accept a hypothesis, and it becomes dogma before there’s sufficient support. This is especially true for “Just-so” stories, hypotheses that explain biological phenomena through compelling evolutionary explanations. These hypotheses make intuitive sense and are sometimes used to frame further research without sufficient testing. Biological anthropologist Holly Dunsworth studied one such hypothesis, the “Obstetrical Dilemma,” and found little evidence for it. She’s posed an alternative hypothesis and cautions us that continued acceptance of the Obstetrical Dilemma causes harm. The evolutionary narrative of overly large fetal heads constrained by overly narrow birth canals leads to unnecessary interventions in childbirth, which may be harmful to mothers and infants.

The Obstetrical Dilemma was proposed to account for a presumed conflict in human biology. Due to our bipedal anatomy, humans have narrow pelvic openings compared to chimpanzees. In chimpanzees, the pelvis is shaped like a bony column with parallel sides; in humans, our pelvis is shaped more like a bowl, wide at the top and sloping inwards at the bottom. This bowl shape centers our torso directly above our legs, but it also means that the pelvic inlet is narrower compared to chimpanzees. Additionally, due to our relatively large brain sizes, we also have babies with large heads. The evolution of our bipedal pelvis and large brains considered two hallmarks of human evolution. Human women have a difficult time giving birth to our large-brained babies, and the size of babies’ heads and women’s pelvis can be a tight squeeze. The obstetrical dilemma holds that our pelvis size is a constraint on fetal growth, so human infants are born relatively “early” before their brains become too big to fit through. According to this hypothesis, if not for our narrow pelvic openings, human fetuses would have continued gestating in their mothers’ wombs for longer time periods. The obstetrical dilemma seems to neatly explain why childbirth in humans is difficult, as well as why human babies seem helpless and underdeveloped. This hypothesis led to viewing the first few months of babies’ lives as the “fourth trimester,” and influenced obstetricians’ decisions about interventions such as inductions and c-sections.

Dunsworth explains that there’s just one problem. We don’t have the evidence to support it. Her initial research into this hypothesis set out to support it, but the further she studied, the less evidence she found. First, human pregnancy is just slightly longer than other apes’ pregnancies, whereas the obstetrical dilemma poses that it is “shortened” compared to the hypothetical gestation length we “should” have if not for our narrow pelvis. Chimpanzees are born with a “relative brain size” of 40%, meaning that at birth, an infant chimpanzee’s brain is about 40% of the average adult brain size. Humans are born with only 30%, which led researchers to conclude that human pregnancy “should” be about 16 months. But they assume that chimpanzees have the ideal primate gestation length and relative brain size—even though other primates, like capuchin monkeys, are born with larger relative brain sizes. Second, we lack evidence that wider pelvises would constrain women’s ability to walk or run. Most of us learn in introductory classes that men’s narrower hips are better adapted to walking and running, whereas women’s wide, childbearing hips are less athletically inclined. However, the evidence thus far does not support this presumption. Dunsworth points out that men’s pelvises tend to be wider than women’s, due to their larger body size. However, women’s pelvises are “roomier” in that their shape allows for a wider birth canal and pelvic inlet. The assumption that men’s pelvises are better adapted to walking and running is based on general sex differences in athletic performance. However, research on men and women indicates that hip width is not related to running efficiency. Furthermore, although men run faster than women at shorter distances, women are able to out-perform men at ultramarathon distances.

Dunsworth proposes an alternative hypothesis to explain why human infants are born at 9 months. Drawing on comparative evidence of metabolism in primates, Dunsworth and colleagues developed the “Energetics of Gestation and Fetal Growth” hypothesis. Rather than being evicted from the womb before their heads are too big, this hypothesis posits that human babies are born when their growth rates become too costly for their mothers’ metabolism to support.

Dunsworth warns us that the unsupported Obstetrical Dilemma has already influenced medical interventions in human childbirth. If babies’ heads are presumed to be too big for their mothers’ birth canal, that can lead to higher rates of inductions before babies are full-term, or unnecessary c-sections. Dunsworth points out how the presumptions of the Obstetrical Dilemma can bias research, such as one study that compared head circumference in babies born via unplanned c-section or instrumental delivery versus vaginally. Babies born via c-section had head circumferences about 1 cm larger. But that one-centimeter difference may simply be a consequence of the compression experienced during vaginal birth, rather than the cause of the interventions. As Dunsworth cautions, the presumptions of the obstetrical dilemma can cause doctors to assume that interventions are necessary when they are not. She notes that “Framing nature as a ‘dilemma’ pits women’s bodies as problems to be solved, rather than as humans to be cared for and assisted.” Recognizing the shaky ground that the Obstetrical Dilemma is founded on, and proposing new ways of viewing human childbirth, may lead to more better outcomes for both mothers and infants.


Dunsworth, H.M. 2018. There is no ‘Obstetrical Dilemma’: Toward a braver medicine with fewer childbirth interventions. Perspectives in Biology and Medicine 61(2): 249-263.

Dunsworth, H.M. 2016. “The ‘Obstetrical Dilemma’ Unraveled.” In Costly and Cute: Helpless Infants and Human Evolution, ed. W. Trevathan and K. Rosenberg, 29-50. Santa Fe: School for Advanced Research.

Dunsworth, H.M., et. al, 2012. Metabolic hypothesis for human altriciality. Proceedings of the National Academy of Sciences 113(25): 6816-6818.

Lipshuetz, M, et. al, 2015. A large head circumference is more strongly associated with unplanned cesarean or instrumental delivery and neonatal complications than high birthweight. American Journal of Obstetrics and Gyncecology 213(6): 833.e1–833.e12.

Warrener, A.G. et al, 2015. A wider pelvis does not increase locomotor costs in humans with implications for the evolution of childbirth. PLoS One 10(3): e0118903.



Published On: September 27, 2018

Michelle Rodrigues

Michelle Rodrigues

Michelle A. Rodrigues is a Postdoctoral Fellow with the Beckman Institute and the Department of Anthropology at University of Illinois.


  • Peter van den Engel says:

    Pretty well stated. I guess human babies need a longer time absorbing first impressions via visual and auditive senses for learning than apes because their lives are more complicated involving language and therefore do not need to be physically active from start. They are allowed to be born earlier and the birth canal has adapted to that, whether the mother cannot sustain its physical body growth any longer is not really the issue. It has adapted to that timing too.

  • Holly Dunsworth says:

    There is no existing evidence that human babies are born early. We are born with the absolute biggest infant brain of all primates. And, the larger the adult brain, the smaller the proportion of it that’s present at birth. That’s a pattern across primates.

  • Dwight Read says:

    Dunsworth argues that human birth occurs when the metabolic requirement of the growing fetus reaches the metabolic energy that can be provided by the mother. That this may be the triggering event for childbirth is certainly plausible, but if the growth rate of the fetus is independent of the fetal head size and the architecture of the pelvis, then it is must be coincidental that the fetal head size is pushing the limits of the architecture of the pelvis at time of childbirth. A more plausible argument would be that the length of pregnancy and the fetal growth rate are jointly under selection for the maximum possible fetal development before childbirth, given the constraints posed by the pelvic architecture.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.