Introduction

For more than 100 years, obstetricians have recognized the pivotal role that the bony pelvis plays in childbirth. The size and shape of the bony pelvis has been classified [1] and its effect on labor and delivery has been studied extensively. Anthropometry, the measurement of human individuals for the purposes of understanding human physical variation, has also been applied to the bony pelvis, predominantly in regards to parturition. For instance, the majority of anthropometric studies find that height positively correlates with vaginal delivery success [2–4].

Data that are available demonstrate that differences in the bony pelvis may contribute to the pathophysiology of pelvic organ prolapse. CT and MRI studies of women with and without pelvic organ prolapse demonstrate wider transverse pelvic inlets and shorter obstetric conjugates in women with prolapse [5, 6]. Furthermore, women with pelvic organ prolapse have less vertically oriented pelvises and less lumbar lordosis [7, 8]. Research also demonstrates that pelvic floor areas differ between races. The posterior pelvic floor is significantly smaller in African Americans compared to European Americans, contributing to a total pelvic floor area that is 5% smaller [9]. The data gleaned from these studies do not indicate whether bony pelvis dimensions and areas are related to height or anthropometric measurements and race.

The objective of this study is to examine the relationship between height and other anthropometric measurements and the size of the adult female bony pelvis, including the true conjugate, interspinous distance, intertuberous distance, and pelvic inlet and outlet areas. A secondary goal is to describe the differences between African American and European American women.

Materials and methods

This study was approved by the Cleveland Museum of Natural History and was exempt from Institutional Review Board review as it does not involve obtaining information about living individuals. Ninety-six disarticulated pelvises were selected from the Hamann–Todd collection at the Cleveland Museum of Natural History. This collection consists of more than 3,100 human skeletons, with documentation of race, age, height, weight, and multiple other anthropometric measurements. Specimens less than 18 years old at death, less than 90 lb at death, and those whose height was greater or less than 2 standard deviations from the mean were excluded. Damaged specimens were also excluded. Pelvises were randomly selected from a list of available specimens, matching the general height distribution of the entire female collection. Initially, 100 specimens were measured; however, four were excluded due to missing data. Forty-seven African American and 49 European American pelvises were selected and reassembled using a technique that has been standardized for this collection. Specifically, the sacrum was articulated with the innominate bones at the sacroiliac joint and the innominate bones were articulated at the pubic symphysis. Once articulated, the three bones were held in place by elastic bands. Given the absence of soft tissue, the pubic symphysis was separated by a 5 mm insert, a distance that has been demonstrated on MRI studies [10]. Architectural bony landmarks were determined as per Standards for Data Collection From Human Skeletal Remains: Proceedings of a Seminar at the Field Museum of Natural History [11]. Three-dimensional points of all pertinent bony landmarks were obtained using the MicroScribe G2 3D Digitizer (Immersion Corporation, San Jose, CA, USA) by one author (BR). The MicroScribe G2 3D Digitizer uses digital optical sensors and has accuracy up to 0.02 cm. The equipment and measuring techniques were evaluated for reliability prior to recording measurements by taking multiple measurements of premeasured distances in a standardized fashion and assuring reliability and accuracy. Additionally, a small sample of distances between pelvic landmarks were taken with the MicroScribe G2 3D Digitizer and calipers which demonstrated accuracy.

Using the three-dimensional points obtained by the MicroScribe G2 3D Digitizer, distances and areas were calculated using mathematical equations. The true conjugate was defined as the distance between the sacral promontory and the superior posterior pubic symphysis. The interspinous distance was defined as the distance between the ischial spines and the intertuberous distance was defined as the distance between the ischial tuberosities. The pelvic inlet area was calculated using the transverse and anteroposterior diameters. Pelvic outlet area was calculated using the anterior urogenital triangle area, the area between the ischial tuberosities, and the posterior triangle area (Fig. 1). Anthropometric measurements were performed post-mortem in a standardized fashion and included height, suprasternal height, hip height, leg length, symphyseal height, torso length, foot length, and iliac crest diameter (Fig. 2). Race was assigned at the time of autopsy.

Fig. 1
figure 1

Diagrams of the bony pelvis with the true conjugate, interspinous distance, intertuberous distance, pelvic inlet area, and pelvic outlet area demonstrated. The pelvic inlet area was calculated using transverse and anteroposterior distances. The midpelvis distance was calculated using the intraspinous diameter. The pelvic outlet area was calculated using the anterior urogenital triangle, the area between the ischial tuberosities, and the posterior triangle

Full size image
Fig. 2
figure 2

This homunculus demonstrates the height measurements taken post-mortem. 1 Torso length, 2 greater trochanter length, 3 iliac crest diameter, 4 pubic symphyseal height, 5 foot length, 6 suprasternal height, 7 anterior superior iliac spine height, 8 height

Full size image

JMP 7.0 was used for statistical calculations. The relationship between pelvic sizes and anthropometric measurements was evaluated using Pearson’s correlation coefficient (r). Given that this study required multiple testing of eight related variables, therefore increasing the chance of type I error, Bonferroni’s correction [12] was used to determine statistical significance of p ≤0.0021. Linear regression was used to evaluate independent associations of anthropometrics and race to pelvic size measurements.

Results

The average (±standard deviation) height for the entire sample was 161.8 ± 8.0 cm with a range from 145.8 to 177.9 cm. The average anthropometric measurements for the African American and European American specimens are shown in Table 1. African Americans were significantly taller compared to European Americans, which is consistent with the entire Hamann–Todd collection. The suprasternal height, anterior superior iliac spine height, symphyseal height, and foot length were also significantly greater for African Americans compared to European Americans. However, there were no differences in leg length, true conjugate, interspinous distance, intertuberous distance, or pelvic areas between races.

Table 1 Average anthropometric measurements for African American (AA) and European American (EA) specimens, in centimeters
Full size table

All vertical anthropometric measurements were significantly correlated with the true conjugate and pelvic inlet area as demonstrated in Table 2. However, no vertical measurement was significantly correlated with interspinous distance. Multiple anthropometric measurements were also significantly correlated with the pelvic outlet area but not intertuberous distance. The strongest association was seen with height, suprasternal height, and iliac crest distance. Linear regression was used to control for race, given the baseline differences in height (the African American cohort was significantly taller). The statistically significant relationships between true conjugate and pelvic inlet and outlet areas and anthropometric measurements persisted after controlling for race. Despite being significantly taller, African American women did not have significantly larger true conjugates or inlet and outlet areas or wider interspinous and intertuberous distances compared to European American women indicating that height, not race, has an association with these variables in both groups.

Table 2 Relationship of anthropometric measurements to pelvic inlet, midpelvis width, and pelvic outlet
Full size table

Discussion

Multiple anthropometric measurements were significantly correlated with the true conjugate and pelvic inlet and outlet areas, but not with interspinous or intertuberous distances. Though the interspinous and intertuberous widths were not statistically significantly correlated to anthropometric measurements, it does not imply that these distances were contracted. Overall height had a greater correlation with bony pelvis measurements than any other anthropometric measurement considered and this relationship persisted after controlling for race. These findings are potentially applicable to the fields of pelvic floor dysfunction and obstetrics.

The results of this study add to the current body of obstetric literature. Multiple studies using clinical outcomes such as successful vaginal delivery have demonstrated that women who have normal spontaneous vaginal deliveries are significantly taller than those who require cesarean section for cephalopelvic disproportion, supporting the finding that height is correlated with pelvic inlet and outlet areas [2–4]. Additionally, studies have associated foot length or shoe size with mode of delivery [4] and smaller pelvic inlet [2], as well as Michaelis sacral rhomboid area with labor dystocia [3]. However, another study of 350 pregnant patients found conflicting results, with height, weight, and shoe size failing to identify inadequate pelvises determined by CT pelvimetry [13]. These studies lend support to the findings of this investigation.

Additionally, the pelvic area findings apply to the existing pelvic floor literature. In a study of 563 Caucasian patients, height was shown to correlate with pelvic anteroposterior inlet and transverse diameter on radiographic pelvimetry [2]. On a multi-ethnic population excluding patients of African descent, height was shown to correlate with levator hiatus length, transverse diameter of the pelvic inlet, sagittal inlet, and sagittal outlet on MRI [14]. Though these findings are consistent with our results, these studies cannot draw conclusions about height, bony pelvis measurements, and race. Our findings demonstrate that the relationship between height and bony pelvis measurements persists after controlling for race.

Racial differences in bony pelvis dimensions are supported by the literature. Previous MRI data indicate that Caucasian women are noted to have longer transverse pelvic diameters of the pelvic inlet and longer anteroposterior diameters of the pelvic outlet as well as longer anteroposterior and transverse levator hiatus diameters when compared to other races [14]. In this study, no women of African descent were included and although the Caucasian women were significantly taller than all other race groups, this was not controlled. Furthermore, African Americans have been found to have narrower pelvises and significantly smaller posterior pelvic floors compared to European Americans [9]. This finding is accentuated after correcting for height. Though these findings are different than our results, they do not directly diverge as these investigators studied distinct landmarks (pelvic floor area) and do not report inlet and outlet areas. Additionally, their findings support the influence that height has on the pelvic floor area.

The large number of specimens examined, collection of data by one individual, and reliable data collection methods add to the strength of this study. However, as this is a study on the bony pelvis and as these pelvises had been disarticulated, we cannot account for differences in soft tissue which may contribute to bony pelvis dimensions. This collection did not have medical history or information on parity or pelvic floor disorders, so we cannot comment on the relationship between these findings and clinical outcomes. Furthermore, some of these bony pelvises had mild degeneration which could potentially affect measurements, especially the distance between the ischial spines. Finally, anthropometric data were collected and race was assigned by the original researchers in the 1920s. Race was assigned as African American or European American and complete descriptions of ancestry were not available.

To conclude, height and other anthropometric measurements are moderately correlated with the true conjugate and pelvic inlet and outlet areas, but not with interspinous or intertuberous distances. Despite differences in height, there were no significant differences in the true conjugate, interspinous and intertuberous distances, and inlet and outlet areas between races. These findings augment previous data and propose that height is an important predictor for bony pelvis dimensions, and therefore, potentially obstetrics and pelvic floor disorders.