Indicators of motherhood? Sacral preauricular extensions and notches in identified skeletal collections

Abstract The sacral preauricular extension (SPE) and sacral preauricular notch (SPN) are morphological changes at the ventral apex of the sacrum. We recently specified their shapes and appearances and suggested a scoring system based on prehistoric Austrian skeletal assemblages. We hypothesized that these specific pelvic changes relate to past pregnancies and parturitions, a hypothesis that we now tested on a subsample of individuals from the Simon Identified Skeletal collection in Geneva (n = 62) and the Christ Church, Spitalfields collection in London (n = 27) linked to historical information on deliveries. We found SPE and SPN in low frequencies and only in female individuals with at least two children in both collections, and a significant association between the emergence of SPE and first births by 25 years. SPN was found only in two females in the Simon collection, but both with a very high number of recorded parturitions including twin births. Based on these results, we are confident in our assumption that at least SPE, and possibly also SPN, result from increased compression forces at the sacroiliac joint, and especially at the ventrosuperior margin, in recurring (complicated) birth events, the interaction of enhanced pelvic joint mobility that is highest up to age 25, and postural changes related to weight gain during pregnancy. Pelvic shape, dimensions, body proportions, biomechanical issues and hormonal levels may also play a role in their emergence.

. The focus of this paper is on the occurrence of extensions or notches at the ventral apex (i.e., ventrosuperior margin of the sacral ala) of the sacrum, described within the framework of the ERC-funded project "The value of mothers to society," aiming to assess prehistoric women's reproductive and social status. The respective features were first noticed in pelvic remains of skeletons excavated at Austrian sites dating from Neolithic to Iron Age, where 87 male and 126 female individuals were analysed in this regard. In this research, we noticed a previously undescribed feature at the ventral sacral apex in a female from Unterhautzenthal and subsequently encountered similar changes in other prehistoric pelvic remains, but only in females Rebay-Salisbury et al., 2018) and named it "sacral preauricular extension" (SPE). It is an osseous expansion at the level of the terminal line, always located at the ventral apex of the sacral ala. Usually, a subtle line delimits the SPE from the auricular joint surface, which is essential for the differentiation from other changes. The thin, ventrally pointing SPE stands in contrast to marginal osteophytes, which bridge or begin to bridge the joint space at their base in coarse shape from the iliac side. More rarely, a loss of convexity at the same location was noticed, termed "sacral preauricular notch" (SPN). The features are sometimes accompanied by corresponding facets (CF) at the ilium, whereby the SPE causes an imprint and the SPN builds a recess. A systematic approach to analyse the features including the development of a detailed recording scheme was started (Pany-Kucera et al., 2019).
Sacral changes possibly associated with pregnancy and parturition events in the area of the attachment site of the sacroiliac ligament have been mentioned in previous papers (Andersen, 1986;Cox, 1989;Cox & Scott, 1992;Houghton, 1974;Kelley, 1979;Maass, 2012;Ullrich, 1975). We found the SPE and SPN only in female skeletons, sometimes occurring with other distinctly expressed pelvic features in female pelves (Pany-Kucera et al., 2021). In collaboration with anatomists from the University of Vienna, a clarification of the causes for the development of the features was pursued. Currently we think that SPE may occur through the mechanism of heterotopic ossification, as it did not seem to be a reparative ossification, whereas SPN probably occurs through epiphyseolysis, both by increased and recurring pressure at this specific location, particularly during complicated labour (Pany-Kucera et al., 2019). But missing parity information on prehistoric individuals made it impossible to draw reliable conclusions between SPE, SPN and pregnancies or parturitions. This led to the decision to study these specific features in identified skeletal collections with known sex, age at death and information on obstetric histories. We hypothesized that SPE and SPN occur in parous females.
The focus of this paper is to test if the incidence and expression of the SPE, the SPN and corresponding facets, may relate to pregnancy and parity in skeletal samples with known obstetric history.

Individuals from the Simon Identified Skeletal Collection housed at the University in Geneva (Switzerland) and the Christ Church
Spitalfields crypt collection at the Natural History Museum in London (United Kingdom) were analysed. The two selected identified skeletal collections were chosen due to availability of obstetric data for a subset of the females (Abegg & Desideri, 2019;Cox, 1989;Perréard & Eades, 2003;Perréard Lopreno & Brů žek, 2010), which is challenging to find (Maass & Friedling, 2016), even more with females of different age groups.
The Simon Identified Skeletal collection at the University of Geneva comprises 496 individuals of Swiss origin, who had died between the end of the 19th and the early 20th centuries (Abegg & Desideri, 2019;Perréard & Eades, 2003). They stem mostly from a predominantly rural, working-class, pre-industrial period population.
For all of them, data sheets with basic information like sex and age, and main pathological changes (e.g., fractures) are available, details on the obstetrical history are disposable from 99 females. We selected our skeletal subsample of 62 individuals on-site (43 females with documented obstetric history, among them 13 nulliparae, 4 primiparae, 26 multiparae; 19 males), mainly based on the preservation status of the pelvis, and excluded those with pathological conditions or fractures affecting the spine, pelvis and lower limbs mentioned in the data sheets (compare differential diagnoses, Pany-Kucera et al., 2019). The pelves were analysed for the occurrence of SPE, SPN and CF, and only then we obtained information on the number of children, age at first and last birth, birth spacing, and marital status for the female individuals.
The second data acquisition took place at the Natural History Museum in London, and concerned individuals from the Christ Church Spitalfields crypt in East London, who died between 1729 and 1852.
They were largely of French origin (Huguenots), settled in Spitalfields due to the silk industry, and were members of the middle class.
Obstetric data is available for a subsample of 94 female individuals (total number of individuals in the named sample n = 387), and includes number of children, age at first and last birth, birth spacing and marital status (Cox, 1989;Cox & Scott, 1992). We pre-selected 27 well-preserved female individuals with documented obstetric data for our analysis (six nulliparae, one primipara and 20 multiparae) based on the descriptions in the extensive publication by Cox (1989).
Although we selected the best-preserved individuals from both groups, some skeletal elements were damaged or missing, which resulted in limited missing data. The on-site examination was performed blind for the obstetric data. Due to the limited access to the collection, a lack of recording time and the fact that we do have com- We calculated descriptive statistics for both sexes for age and sex, and within the females for parity groups. Results of crosstabs statistical analyses for two independent samples (Monte Carlo confidence level 95%) are shown. Fisher's exact test (in Chi 2 tests) for small sample sizes are performed and are given in the tables in cases of significant results, to look for possible associations of SPE and SPN with sex, age and body height. Individuals were separated into two age groups (17-39, n = 28, and 40+, n = 42), and two stature groups (shorter than 155 cm and taller than 155 cm, n = 34 respectively).
Within the female group (including nulliparae) we tested for the occurrence of SPE and SPN in relation to parity, age at first birth (first birth before or after the age of 25 years), two or more births at 25 years, close birth spacing (≤2 years spacing), and total number of children (three groups: 0-2 children, 3-5 children, 6 or more children).
All calculations were first performed for right and left sacral sides separately, and then for the combined right and left side (r_l) variable, F I G U R E 1 Variants of SPE (a, c-e), CF (b) and SPN (f) in females of the Simon Skeletal Identified Collection, Geneva. Individual number/age at first birth/number of children/ age at death: a and b: BIE 32/21 years/5/34 years, c: LSZ 17/24 years/2/50 years, d: PAM 01/18 years/6/82 years, e: PAM 16/22 years/4/71 years and f: STP 02/20 years/9/76 years consisting of the maximum values from both sides (the latter results are shown in the tables). Counts and expected counts are given in the crosstabs tables, to clarify and strengthen the significant results.

| RESULTS
Descriptive statistic results on age at death and individual numbers in the analysed groups can be found in Table 1. In all the results from the crosstabs analyses, the nulliparous females are included (except for having the first child by 25 years, for obvious reasons), though the results remained significant when excluding them.

| Sacral preauricular extension and corresponding facets
In the analysed subsamples, only females (Table 2), and within those, only parous females (Table 3), who had two or more children presented a SPE. The frequency of occurrence of SPE, however, was rather low (c. 6-10%), and no statistically significant difference between the analysed groups regarding the SPE occurrence was found, neither for right or left sides, nor for the combined right and left variable (Fisher's exact test significance (two-sided) is 0.337 for sex (Table 2) and 0.177 for parity (Table 3)). Corresponding facets at the ilium in the shape of an imprint occurred in both subsamples from the Geneva and London collections with a frequency of c. 4%-6%.
In the whole group of parous females from both sites (Simon and Spitalfields, n = 51), the occurrence of SPE ranged between 3 in 49 (6.1%), right side, and 5 in 51 (9.8%), left side; including one individual with a bilateral expression, therefore, in the combined variable, n = 7. Facets corresponding to SPE at the ilium were found in 2 of 50 females (4%) on the right side, and 3 of 51 (5.9%) on the left side.
A significant association was found in crosstabs calculations between the presence of an SPE and having the first child by 25 years for the combined maximum right and left variable ( Fisher's exact test (two-sided).
Clearly insignificant results were found in statistical tests for an association of SPE in females with close birth spacing, or between age groups (Table 6a) and grouped body height (Table 6b).

| Sacral preauricular notch and corresponding facets
Like the SPE, the SPN was only found in females of the two samples, and only in multiparous females. A SPN occurred with a frequency of 2 in 51 (3.9%) female individuals who had more than one child by 25 years, at the left ala sacralis, and only in the Simon collection.
A statistically significant result was obtained in crosstabs calculations when testing for left side SPN and total number of children (

| Details on the subsample from the identified skeletal Simon collection
Of the 43 females analysed from this collection, 30 have had children; four were primiparae, six females have had two children, and 20 have had more than two children (Table 1, nine females had five or more, and two of them had more than 10, maximum number 15 children).  No relevant pelvic changes such as SPE, SPN or corresponding facets were found in the males (Table 2).

| Details on the subsample from the Christchurch Spitalfields collection
The 21 parous females out of the 27 analysed individuals (Table 1) from the Christchurch subsample included one primipara, one woman with two children, and 19 who have had more than two children (seven had five or more, four had 10 or more, maximum number 15 children). The mean number of children in the selected female group was 3.9 (nulliparae included in calculation), mean age at first The SPE occurred in three multiparous females from the Spitalfields collection, the frequency of occurrence is 2 of 21 (9.5%) for each side, with one bilateral expression. The bilateral SPE occurred in a female who had given birth to three children between the age of 21 and 23 (CAS 2327, age at death 23 years (Figure 2a,b)). The individual with the right sided extension has had five children at the age of 19-27 years (CAS 2070, age at death 35 years, Figure 2c), and the individual with the left sided SPE has had four children between the age of 20-38 years (CAS 2368, age at death 45 years, Figure 2d).
While the right-and the left-sided SPEs (1 of 21 at each side) showed a corresponding facet at the ilium, the iliac bones of the female with the bilateral SPE did not show these changes. The right femur of CAS 2070 was 3 cm longer than the left one. The oldest of these females showed slight degenerative changes at lumbar vertebrae four and five.
A SPN was not found in the Spitalfields subsample.

| DISCUSSION
The results of this study on subsamples from the Simon and the Spitalfields Identified Skeletal collections corroborate our hypothesis that the SPE occurs in relation to multiparity. We found significant associations for SPE within the female group, more precisely for those women who had their first births up to the age of 25 years, and especially those who had minimum two children at this age. No association with close birth spacing or total number of children was detected. We observed no evidence of SPE in nulliparous females or in males. We found no indications that SPE would be more frequent in the age group 40+, or stand in any relationship to body height.   (Figure 3, Pany-Kucera et al., 2019). The line is also found in the analysed 19-20th century groups, and is especially noticeable in Figures 2a, 2d and 2e, and 3a-c.
In some cases, the line is more difficult to discern (Figures 2c and 3d).
Degenerative changes at the sacral auricular facet can make it more difficult to discern SPE from marginal exostoses. However, SPE and SPN exclusively occur at the ventral sacral apex. In rare cases, SPE occurs with a bilateral expression, whereas so far, this was not found for SPN.
We argue that the described features are caused by the conjunction of increased pressure on the ventral sacral apex, weight gain, and increased motion ranges in the sacroiliac joint over the course of multiple gestations and (presumably complicated) parturitions (Pany-Kucera et al., 2019). After puberty, females have significantly increased levels of oestrogen compared to men. Oestrogen directly influences muscle tissue, tendons, and ligaments, increasing their laxity after puberty (Brunner et al., 1991;Hansen, 2018;Vleeming et al., 2012), particularly up to the age of 25 years, where the highest degree of movement occurs in the sacroiliac joint (Brooke, 1924, Snodgrass & Galloway, 2003, Suchey et al. 1979).
This and a different surface architecture allow for a higher range of movement in female sacroiliac joints (Hansen, 2018;Vleeming et al., 2012;Vleeming & Schuenke, 2019). Moreover, increases in hormonal levels, especially of oestrogen, occur in every menstrual cycle with great inter-individual variability, which also interrelate with nutritional and genetic factors (Hansen, 2018). Increasing relaxin levels in early pregnancy, with a peak in the 12th week, further raise motion ranges in the sacroiliac joints (Borg-Stein & Dugan, 2007). A change in pelvic posture is induced by weight gain and a shift of the centre of gravity during pregnancy (also initiating gait pattern changes, Ribeiro et al., 2013), which may be maintained after parturition, as levels of oestrogen and relaxin decrease rapidly (Aldabe et al., 2012;Hansen, 2018

| CONCLUSION
In this study, we found indications that the occurrence of the SPE and, to a lesser degree, the SPN, emerge with multiple pregnancies and births up to 25 years. We found a statistically significant association between the occurrence of SPE and the age at first births, and giving birth to two or more children up to the age of 25 years. We frequently found the features in young females, and so far, we found no conclusive evidence for a causal relationship of SPE with close birth spacing, age or body height. However, the changes occured in low frequencies here (maximum 10%) and there are multiparae who had their children before or around 20 years that do not show these changes.
We therefore conclude that factors such as pelvic shape and dimensions, body proportions, as well as biomechanical and hormonal issues are probably involved in the emergence of the SPE and SPN.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.