SEARCH

SEARCH BY CITATION

Keywords:

  • parity status;
  • pelvis

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

In the 40 years since the phrase ‘scars of parturition’ was coined, studies have attempted to show the relationship between scars on the bony pelvis and parity history. Despite numerous studies, the relationship of parity and scarring remains unclear. The challenge facing these studies is the rarity of skeletal collections of known parity. The alternative study approach is examining relationships between scars and factors other than pregnancy-related strains that may affect their manifestation.

Skeletal remains of 312 individuals were examined for scarring at the dorsal pubic surface, pubic tubercle, preauricular sulcus, interosseous groove and iliac tuberosity. Pelvic and femoral measurements were also taken. Features were compared according to sex and age. Principal components analysis was performed to assess the influence of body and pelvic size on scar manifestation.

Scars occurred in both sexes, although they were more common and more severe in females. Scar severity remained unchanged or increased with age in both sexes. Females had smaller bodies but larger pelves than males. The interspinous and transverse inlet diameters and the femur measurements increased with age. Principal components analysis showed that body and pelvic sizes represented the majority of the observed variation, with scars occurring more commonly in small-bodied individuals with large pelves, most of which were females. Both sexes also show a difference in the magnitude of scarring at the pubis and ilium.

These results suggest that weight-bearing and pelvic stability may be a better explanation for scarring than parturition-related strain. Female pelves are more flexible and require more ligamentous stabilization, causing increased scar formation. The weight-bearing strain on male pelves may sometimes also be sufficiently large to cause similar scars.

Future studies may be able to test this theory on samples of known parity history. Copyright © 2014 John Wiley & Sons, Ltd.

Over the past 40 years, studies have debated the use of ‘parturition scars’ on the bony pelvis to infer parity history of an individual. Some studies propose that scars are the result of strain on the attachment sites of pelvic ligaments due to increased ligamentous laxity caused by hormones secreted during pregnancy, combined with separation of the pelvic joints during parturition (Angel, 1969; Houghton, 1974; Putschar, 1976; Kelley, 1979). Others argue that because scars sometimes also occur in males and nulliparous females, pregnancy and parturition-related strains cannot be the only cause of scarring (Stewart, 1970; Holt, 1978; Snodgrass & Galloway, 2003; Mitra, 2010; Arraiza & Merbs, 2012; Ubelaker & De La Paz, 2012). The only alternative suggestion in the literature is that scarring is the result of ligamentous stabilization of the pelvic girdle (Andersen, 1986). Thus, females with more flexible pelvic architecture would require more ligamentous stabilization and present with more scarring, whereas males with more tightly articulated pelves would have less scarring. Several studies have suggested that body size may be a contributing factor to scar formation, which may explain the occurrence of scarring in males, but this theory has not been sufficiently tested (Galloway et al., 1998; Arraiza & Merbs, 2012; Ubelaker & De La Paz, 2012). Despite the ongoing debate about the causes of scar formation, several anthropologists and archaeologists still use scars to infer parity of skeletal material (Andersen, 1986; Ubelaker & De La Paz, 2012).

The main challenge facing studies of the relationship between scars and parity history is that skeletal collections of known parity are rare. This study employed an alternative approach to examine these ‘parturition scars’ by examining the relationship between scars and non-pregnancy-related factors such as body and pelvic size that might influence scar manifestation in both females and males.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Sample

Skeletal remains of 312 individuals (128 females and 184 males) from the University of Cape Town (UCT) and University of Stellenbosch collections were examined. The sample consisted of individuals of cadaveric origin or of forensic cases brought to UCT for analysis. In both cases, the individuals mostly lived in the Western Cape Province of South Africa within the past 40 years, with the small number of remaining individuals coming from other areas of South Africa. Age and sex data of the individuals of cadaveric origin were obtained from the accession registers of their respective collections. Standard age and sex estimation techniques (İşcan et al., 1993; Buikstra & Ubelaker, 1994; Igarashi et al., 2005) were used for individuals of forensic origin. Selection criteria included skeletal maturity (±20 years of age), relative completeness of the innominates and the absence of visible pathology. Individuals were divided into four age groups, modified from Buikstra & Ubelaker (1994). The groups were as follows: ‘young’ (20–35 years), ‘middle’ (36–50 years), ‘old’ (51–65 years) and ‘very old’ (>65 years). A summary of the sex and age group distribution of the sample is given in Table 1.

Table 1. Sex and age group distribution of individuals examined in this study
 FemaleMaleTotal
Age group
Young202848
Middle353671
Old275582
Very old4665111
Total128184312

Data collection

Older studies, such as those of Holt (1978) and Kelley (1979), used a morphological approach when examining the degree of scarring. More recent studies tend to use a metric approach (Andersen, 1986; Snodgrass & Galloway, 2003). The present study used a combination of the two methods by first measuring scar features and then using these measurements to classify individuals into defined categories, thus providing both improved precision of classification and sufficient descriptive information regarding the appearance of the features (Ebrahim & Sullivan, 1995; Eufinger et al., 1997). No significant differences were detected between bilateral measurements; thus, only measurements of the left side were used in further analyses (Buikstra & Ubelaker, 1994).

The five areas examined were the dorsal pubic surface, pubic tubercle, preauricular sulcus, interosseous groove and the iliac tuberosity (Figure 1).

image

Figure 1. Medial view of the right innominate (of a 47-year-old female) showing the areas of scarring examined: (a) dorsal pubic surface; (b) pubic tubercle; (c) preauricular sulcus; (d) interosseous groove; and (e) iliac tuberosity. This figure is available in colour online at wileyonlinelibrary.com/journal/oa

Download figure to PowerPoint

The maximum diameter of the largest pit on the dorsal pubic surface was measured and then classified according to Stewart (1970) as ‘absent’ (no pitting), ’trace to small’ (diameter <2.0 mm) or ‘medium to large’ (diameter >2.0 mm).

The pubic tubercle height, defined as the maximum height that the tubercle protrudes from the bone on the ventral side of the pubis (Snodgrass & Galloway, 2003), was measured, and the tubercle extensions was classified as ‘small’ (<1.0 mm), ‘medium’ (1.0 mm to 3.0 mm) or ‘large’ (>3.0 mm).

The maximum depth and width of the preauricular sulcus were measured and used to classify the sulcus as ‘absent/broad-shallow’ (<3.0 mm deep), ‘narrow-shallow’ (3.0–5.0 mm deep), ‘defined’ (>5.0 mm deep but <10.0 mm wide) or ‘complex’ (>5.0 mm deep and >10.0 mm wide).

The maximum depth and width of the interosseous groove were measured and used to classify the groove as ‘shallow’ (<3.0 mm deep, <5.0 mm wide), ‘moderate’ (>3.0 mm deep, 5.0–10.0 mm wide) or ‘developed’ (>3.0 mm deep, >10.0 mm wide).

The thickness of the iliac tuberosity was measured from its highest point to the posterior of the ilium, as described by Andersen (1986), allowing the tuberosity to be classified as ‘no eminence’ (<20.0 mm), ‘depressed’ (20.0–25.0 mm) or ‘pointed’ (>25.0 mm).

The pelvic girdle was re-articulated with elastic bands with a 5.0-mm-thick rubber insert at the pubic symphysis to account for the absence of soft tissue, as well as to improve the stability of the articulation of the girdle (Schroeder et al., 1997; Ridgeway et al., 2011). The following measurements, commonly used in obstetric and gynaecological practice (Kurki, 2007; Cunningham et al., 2010), were taken:

  • Anteroposterior inlet diameter: sacral promontory to pubic crest
  • Anteroposterior outlet diameter: end of fifth sacral segment to lower border of pubic symphysis
  • Transverse inlet diameter: maximum distance between arcuate lines
  • Transverse outlet diameter: posterior of one ischial tuberosity to the corresponding point on the opposite side of the body
  • Interspinous diameter: distance between ischial spines.

The bicondylar length and maximum head diameter of the left femur were measured and used as proxies for stature and body mass, respectively (Ruff et al., 1991). Regression formulae for estimating stature and/or body mass from femoral measurements were not employed, because such calculations have been shown to be unreliable in non-European or American samples such as that of the present study (Kurki et al., 2010).

Observer error

Observer repeatability was assessed using Bland–Altman tests to compare repeated measurements of 30 randomly selected pelves (males and females). Repeat measurements were made by the original observer and an independent observer after completion of the original data set.

Data analysis

Statistical analysis was performed using Statistica® (Statsoft Inc., 2009). Data were submitted to Chi-squared and Fisher-exact testing (significance level p = 0.05) to detect possible associations of scar features to sex or age.

Pelvic and femoral measurements were compared between sexes using two-variable t-tests or Mann–Whitney U-tests, and among age groups using one-way analysis of variance or Kruskal–Wallis tests.

Multivariate analysis

Principal components analysis (PCA) was performed to identify possible size and shape patterns in the data. Data were converted to z-scores to standardize each variable. PCA requires complete individual data sets; thus, individuals with missing data were excluded from the analysis, leaving only the sex samples large enough for the analysis (n = 43 females, n = 68 males).

Eigenvalues were calculated for the 12 possible principal components, and the eigenvalue-one criterion (Kaiser, 1960) was applied to determine the number of components to be retained for further analysis. Eigenvalue coefficients of the 12 variables for each of the retained components were calculated, allowing assessment of the relationships of the variables to these components. Score plots representing the eigenvalue coefficients of each individual examined were then constructed to allow comparison of the sexes according to each component.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Observer error

The differences between the repeated measurements of the 30 randomly selected pelves were not significant for all measurements, with the absolute mean difference between sets of measurements being less than 1.0 mm for both intra-observer and inter-observer error.

Differences in scarring between sexes

The occurrence of dorsal pubic pitting (χ2 = 53.30; p < 0.01), pubic tubercle extension (χ2 = 13.34; p < 0.01), preauricular sulcus appearance (χ2 = 65.02; p < 0.01) and the interosseous groove appearance (χ2 = 129.63; p < 0.01) were significantly associated with sex (Table 2). In all cases, except pubic tubercle extension, the female sample presented with more frequent and more severe scarring than the male sample; however, there was a large overlap in scar appearances between the sexes. In the case of the pubic tubercle, males tended to have larger tubercles than females. The iliac tuberosity appearance was not significantly associated with sex (χ2 = 3.45; p = 0.18).

Table 2. Comparison of the appearance of scarring between sexes
 FemaleMale
n%n%
Dorsal pubic pitting
Absent8266.717697.2
Medium/large4133.352.8
Pubic tubercle
Small2126.977.3
Medium5266.77679.2
Large56.41313.5
Preauricular sulcus
Absent/broad-shallow2418.98848.3
Narrow-shallow3729.16837.4
Defined2721.32111.5
Complex3930.752.8
Interosseous groove
Shallow3127.416290.0
Moderate1815.9126.7
Developed6456.763.3
Iliac tuberosity
No eminence3528.74324.0
Depressed mound6049.210759.8
Pointed mound2722.12916.2

Differences in pelvic and femur measurements between sexes

All pelvic measurements were significantly larger in females, whereas both femoral measurements were significantly larger in males, with p < 0.01 in all cases (Table 3).

Table 3. Comparison of pelvic and femur measurements between sexes
 Female, mm (n)Male, mm (n)
Anteroposterior inlet diameter111 ± 10.8 (96)102 ± 9.8 (158)
Transverse inlet diameter124 ± 9.5 (97)115 ± 9.4 (158)
Interspinous diameter105 ± 7.6 (93)86 ± 8.3 (156)
Anteroposterior outlet diameter118 ± 9.3 (92)107 ± 8.6 (152)
Transverse outlet diameter124 ± 10.0 (89)103 ± 12.2 (158)
Femur length419 ± 2.2 (117)450 ± 2.5 (174)
Femur head diameter41 ± 2.7 (117)46 ± 3.3 (169)

Differences in scarring among age groups

Only the pubic tubercle and iliac tuberosity differed significantly with age (Table 4). The difference in tubercle extension was only significant in the pooled sample (χ2 = 15.77; p = 0.02), with individuals of the two older groups having larger tubercles than those of the two younger groups. Similarly, the iliac tuberosity of older individuals was more developed than those of the younger individuals (p < 0.03 in all three samples).

Table 4. Comparison of the appearance of scarring among age groups
 Female, % (n)Male, % (n)Pooled sample, % (n)
 YoungMiddleOldVery oldYoungMiddleOldVery oldYoungMiddleOldVery old
Dorsal pubic pitting
Absent73.7 (14)64.7 (22)70.4 (19)62.8 (27)96.4 (27)100 (35)98.2 (54)95.2 (60)87.2 (41)82.6 (57)89.0 (73)82.1 (87)
Medium/large26.3 (5)35.3 (12)29.6 (8)37.2 (16)3.6 (1)0 (0)1.8 (1)4.8 (3)12.8 (6)17.4 (12)11.0 (9)17.9 (19)
Pubic tubercle
Small42.9 (6)39.1 (9)21.4 (3)11.1 (3)7.7 (1)13.3 (2)8.3 (3)3.1 (1)25.9 (7)28.9 (11)12.0 (6)6.8 (4)
Medium42.9 (6)60.9 (14)71.4 (10)81.5 (22)92.3 (12)86.7 (13)75.0 (27)75.0 (24)66.7 (18)71.1 (27)74.0 (37)78.0 (46)
Large14.2 (2)0 (0)7.2 (1)7.4 (2)0 (0)0 (0)16.7 (6)21.9 (7)7.4 (2)0 (0)14.0 (7)15.2 (9)
Preauricular sulcus
Absent/broad40.0 (8)14.3 (5)7.7 (2)19.6 (9)67.8 (19)55.5 (20)49.1 (27)34.9 (22)56.2 (27)35.2 (25)35.8 (29)28.4 (31)
Narrow30.0 (6)42.9 (15)34.6 (9)15.2 (7)28.6 (8)30.6 (11)43.6 (24)39.7 (25)29.2 (14)36.6 (26)40.7 (33)29.4 (32)
Defined10.0 (2)11.4 (4)23.1 (6)32.6 (15)3.6 (1)13.9 (5)7.3 (4)17.5 (11)6.3 (3)12.7 (9)12.4 (10)23.9 (26)
Complex20.0 (2)31.4 (11)34.6 (9)32.6 (15)0 (0)0 (0)0 (0)7.9 (5)8.3 (4)15.5 (11)11.1 (9)18.3 (20)
Interosseous groove
Shallow46.7 (7)25.8 (8)33.3 (8)18.6 (8)92.9 (26)91.2 (31)90.9 (50)87.3 (55)76.7 (33)60.0 (39)73.4 (58)59.4 (62)
Moderate13.3 (2)16.1 (5)4.2 (1)23.3 (10)7.1 (2)5.9 (2)3.6 (2)9.5 (6)9.3 (4)10.8 (7)3.8 (3)15.1 (16)
Developed40.0 (6)58.1 (18)62.5 (15)58.1 (25)0 (0)2.9 (1)5.5 (3)3.2 (2)14.0 (6)29.2 (19)22.8 (18)25.5 (27)
Iliac tuberosity
No eminence52.6 (10)36.4 (12)30.8 (8)11.4 (5)42.3 (11)34.3 (12)20.8 (11)13.9 (9)46.7 (21)35.3 (24)24.0 (19)12.8 (14)
Depressed47.4 (9)42.4 (14)46.1 (12)56.8 (25)46.2 (12)57.1 (20)66.0 (35)61.5 (40)46.7 (21)50.0 (34)59.5 (47)59.6 (65)
Pointed0 (0)21.2 (7)23.1 (6)31.8 (14)11.5 (3)8.6 (3)13.2 (7)24.6 (16)6.6 (3)14.7 (10)16.5 (13)27.6 (30)

Differences in pelvic and femur measurements among age groups

There was no significant difference among age groups regarding the anteroposterior inlet diameter or the anteroposterior and transverse outlet diameters in either the female, male or pooled samples, with p > 0.07 in all cases (Table 5).

Table 5. Comparison of pelvic and femur measurements among age groups
 Anteroposterior inlet diameter, mm (n)Transverse inlet diameter, mm (n)Interspinous diameter, mm (n)Anteroposterior outlet diameter, mm (n)Transverse outlet diameter, mm (n)Femur length, mm (n)Femur head diameter, mm (n)
Female sample
Young108 ± 7.8 (11)119 ± 8.0 (12)99 ± 5.9 (11)115 ± 8.1 (12)121 ± 6.2 (11)408 ± 21.3 (18)40 ± 2.6 (19)
Middle109 ± 10.4 (28)120 ± 9.3 (29)104 ± 7.4 (29)118 ± 10.5 (28)122 ± 9.1 (27)419 ± 19.9 (33)41 ± 3.0 (32)
Old113 ± 13.0 (23)125 ± 7.2 (23)107 ± 7.3 (23)119 ± 8.2 (20)123 ± 9.0 (21)424 ± 24.0 (21)41 ± 2.4 (22)
Very old111 ± 10.4 (34)127 ± 10.3 (33)106 ± 7.7 (30)118 ± 9.5 (32)127 ± 11.8 (30)421 ± 21.5 (45)42 ± 2.1 (44)
Male sample
Young101 ± 7.5 (26)111 ± 9.0 (26)84 ± 7.5 (26)108 ± 6.1 (25)100 ± 11.6 (26)451 ± 29.5 (28)45 ± 3.3 (28)
Middle101 ± 10.7 (28)113 ± 7.8 (30)85 ± 8.1 (29)106 ± 8.0 (28)101 ± 9.8 (29)442 ± 20.7 (34)45 ± 2.9 (34)
Old102 ± 9.8 (48)115 ± 9.1 (48)85 ± 7.4 (47)106 ± 8.5 (46)102 ± 10.9 (47)454 ± 21.1 (50)46 ± 2.8 (46)
Very old102 ± 10.4 (56)119 ± 9.5 (54)88 ± 9.3 (54)108 ± 9.8 (53)107 ± 13.9 (56)450 ± 26.9 (62)47 ± 3.3 (61)
Pooled sample
Young103 ± 8.3 (37)114 ± 9.4 (38)89 ± 9.7 (37)110 ± 7.6 (37)106 ± 14.0 (37)434 ± 33.6 (46)43 ± 3.9 (47)
Middle105 ± 11.2 (56)116 ± 9.4 (59)94 ± 12.3 (58)112 ± 11.3 (56)111 ± 14.4 (56)430 ± 23.3 (67)43 ± 3.6 (66)
Old105 ± 12.2 (71)118 ± 9.8 (71)93 ± 12.5 (70)110 ± 10.3 (66)109 ± 14.0 (68)445 ± 25.8 (71)45 ± 3.6 (68)
Very old105 ± 11.1 (90)122 ± 10.6 (87)95 ± 12.2 (84)112 ± 10.8 (85)114 ± 16.3 (86)438 ± 28.4 (107)45 ± 3.8 (105)

Significant differences among age groups were found with regard to the transverse inlet diameter in all three samples (p < 0.01), the interspinous diameter of the female sample (p = 0.04), femur length of the pooled sample (p = 0.02) and femur head diameter in all three samples (p < 0.01 in all cases). In all of these cases, the measurements increased with age.

Multivariate analysis

Only the first three components had eigenvalues greater than 1.0 and accounted for approximately 60% of the total variance observed (Table S1, Supporting information).

Table 6 shows the eigenvalue coefficients of the first three principal components (PC1–PC3). PC1 accounted for 34.4% of the total variance and primarily contrasted body size variables (strongest negative coefficients) with pelvic size variables (strong positive coefficients). All of the scar features were positively correlated to this component, except the pubic tubercle extension, which showed weak negative correlation.

Table 6. Eigenvalue coefficients of the first three principal components
VariableEigenvalue coefficients
Principal component 1Principal component 2Principal component 3
  1. Sample sizes for PCA (by age group): female sample – young = 4, middle = 10, old = 10, very old = 19; male sample – young = 11, middle = 11, old = 24, very old = 22.

Dorsal pubic pitting0.52−0.28−0.35
Pubic tubercle−0.100.58−0.04
Preauricular sulcus0.540.190.51
Interosseous groove0.650.010.39
Iliac tuberosity0.270.470.45
Anteroposterior inlet diameter0.520.40−0.43
Transverse inlet diameter0.620.50−0.23
Interspinous diameter0.870−0.08
Anteroposterior outlet diameter0.620.210.01
Transverse outlet diameter0.840.02−0.15
Femur length−0.480.70−0.11
Femur head diameter−0.570.65−0.04

PC2 accounted for 16.9% of the variance and shows strong contrast of the pubic tubercle, iliac tuberosity, both pelvic inlet diameters and both femoral measurements with dorsal pubic pitting (only negative coefficient, −0.28).

PC3 accounted for 8.5% of the variance and primarily contrasted scarring at the posterior pelvic girdle with scars on the dorsal pubic surface and the anteroposterior inlet diameter.

Figures 2-4 are scatterplots of the individual component scores for the first three principal components, showing the distribution of the scores for each component for males and females. The sexes were clearly separated according to the first principal component, showing very little overlap between males and females, but not according to the second or third components, which show large areas of overlap between the sexes.

image

Figure 2. Score plot of individual scores of male and female samples for the first and second principal components.This figure is available in colour online at wileyonlinelibrary.com/journal/oa

Download figure to PowerPoint

image

Figure 3. Score plot of individual scores of male and female samples for the first and third principal components. This figure is available in colour online at wileyonlinelibrary.com/journal/oa

Download figure to PowerPoint

image

Figure 4. Score plot of individual scores of male and female samples for the second and third principal components. This figure is available in colour online at wileyonlinelibrary.com/journal/oa

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Differences in scarring between sexes

The occurrence and severity of scarring at all sites, except the iliac tuberosity, were significantly associated with sex. Females presented with more severe scarring than males at the dorsal pubic surface, preauricular sulcus and interosseous groove. Considerable numbers of males, however, also presented with scarring, albeit to a lesser degree, which indicates that parturition is not the only cause of scar formation. Angel (1969) suggested that scarring in males is likely related to disease or trauma. This may have been true for the small number of males with dorsal pubic pitting (n = 5), but the large number of males with more severe scarring at the preauricular sulcus and interosseous groove does not support this claim. The occurrence of scarring in both sexes suggests that similar strains act on the ligaments attached to these sites. The more severe appearance of the scarring in females, however, suggests that the magnitude of this strain was larger than in males. Despite the majority of females presenting with more severe scarring, large numbers of females also had less severe scarring similar to that of males. Angel (1969), Houghton (1974) and Kelley (1979) reported similar results, suggesting that females with more severe scarring were parous and those with less severe scarring were nulliparous. Because the parity histories of the individuals examined in this study were not available, this claim could unfortunately not be tested.

Pubic tubercle extension was the only scar feature that was larger in males than in females. Although the majority of both sexes had moderate tubercle extension, the majority of remaining males had large tubercles, whereas the majority of remaining females had small tubercles. This indicated that the strain of the rectus abdominis muscle on its attachment site at the tubercle was larger in males than in females, possibly due to males generally being more muscular and performing more activities involving trunk flexion and carrying of heavy loads (Ruff, 1987; Abitbol, 1996; Puoane et al., 2002). The large overlap in tubercle appearance between sexes could thus simply be a reflection of differences in body size and activity patterns between sexes. Alternatively, the similarity may also have been due to extension of the tubercle in females during pregnancy, causing it to become extended to a similar degree to that of males, as suggested by Cox & Scott (1992).

Scarring was present in both sexes, and thus, pregnancy and parturition-related strains were not the only cause of scar formation. Andersen (1986) suggested that scar formation was the result of excess movement allowed by the more flexible architecture of female pelves and that this increased flexibility caused increased strain on the pelvic ligaments, resulting in remodelling of the ligamentous attachment sites and the formation of more severe scars in females. It thus appears that weight-bearing strain on the pelvic ligaments may be one of the causes of scar formation, with the capacity of the ligaments to absorb this strain determining the severity of the resulting scars. The fact that pregnancy-related strain is essentially an extreme form of weight-bearing could explain the association of scar formation and parity history reported in the literature.

Differences in pelvic and femoral measurements between sexes

Similar to previous studies (Tague, 1989; Kurki, 2007), all of the pelvic measurements were significantly larger in females than in males, despite males having larger body sizes, as reflected in the significantly larger femur length and head diameter measurements. The differences in the pelvic measurements are likely due to differences between sexes regarding the functional requirements of the pelvic girdle. Males only have the requirements of bipedalism that restrict pelvic size in order to optimize the efficiency of weight transmission through the girdle, whereas females have the added requirements for obstetric adequacy (Rosenberg, 1988; Bruzek & Murail, 2006; Kurki, 2007).

Differences in scarring among age groups

Only the pubic tubercle extension and iliac tuberosity appearance were significantly associated with age. In both cases, the scars showed an increase, rather than the expected decrease, with age, suggesting that the strains acting on these sites increased with age. In the case of the iliac tuberosity, the association was present in both female and male samples, as well as in the pooled sample. This suggests that similar changes in the magnitude of strain acting on the ligaments attached to the tuberosity occur in both sexes. In the case of the pubic tubercle, though the association with age was only present in the pooled sample. The female and male samples showed a similar increase in tubercle extension with age, although the association did not reach statistical significance. It is possible that this lack of significance in the female and male samples may be due to the small sample size of the young and middle adult age groups, thus preventing statistical detection of this association. In both the case of the pubic tubercle and the iliac tuberosity, it is also possible that the increase with age is simply due to osteophytic development.

The increase in pubic tubercle extension and development of the iliac tuberosity, as well as the lack of change in scar manifestation in the other scar sites examined, suggest that the strain on the ligaments or muscles attached to these sites either increases or remains fairly constant with increasing age. This is the opposite of what is expected if pregnancy and parturition-related strains were the only cause of scar formation, since such strains are transient in nature, resulting in a reduction of scarring over time. These results are similar to those of Suchey et al. (1979), who found that females that died more than 15 years after the last birth event tended to have larger scars than those who died within a shorter interval. Also, the same changes, or lack thereof, are observed in both sexes, suggesting that the factors that are responsible are common to both sexes. These factors may thus be related to age-related changes in activity patterns and/or body mass (Deurenberg et al., 1991).

Differences in pelvic and femur measurements among age groups

The observed lack of differences in the anteroposterior inlet diameter and the anteroposterior and transverse outlet diameters was expected and are in agreement with the results of Moerman (1982). The lack of change also indicates that age-related changes in body mass and physical activity did not significantly affect these dimensions.

The transverse inlet and interspinous diameters were shown to increase with age. These differences were unexpected, because pelvic measurements tend not to change significantly after skeletal maturity (Moerman, 1982). Inspection of the means and standard deviations for each age group shows that, despite the statistical difference detected, there is considerable overlap in the ranges for the different groups. It thus appears that the detected difference is likely due to sampling bias.

The mean femur length and head diameter increased with age in the pooled sample, which was also unexpected. It is likely that the observed difference is due to the male sample, which has been shown to have a larger mean than the female sample, constituting a larger proportion of the pooled sample, thus causing the difference among age groups to be detected as significant.

Multivariate analysis

The first three principal components accounted for approximately 60% of the observed variance.

PC1 could be interpreted as representative of the relative sizes of the body and pelvis to each other and represented approximately a third of the observed variance. All of the scar features, with the exception of the pubic tubercle, were positively loaded on this component, suggesting that scar development tended to occur in small-bodied individuals with large pelves. The negative loading of the pubic tubercle on this component suggests that tubercle extension tended to occur in large-bodied individuals with small pelves. Separation of the sexes according to PC1 (Figures 2 and 3) supports the results of the univariate analyses, that is, females have smaller bodies with proportionally larger pelves, as well as a greater occurrence of scarring, whereas man have larger bodies with proportionally smaller pelves and less scarring. Cox (1989) reported a similar relationship between scarring and pelvic size and suggested that increased need for ligamentous stabilization of the pelvic girdle is the cause of scar formation. It is, however, also possible that those females that show more scarring are the ones that did give birth, whereas those that are similar to males are nulliparous, but this could not be tested because of the lack of known parity history.

PC2 contrasted scarring at the pubic tubercle and iliac tuberosity, the pelvic inlet diameters and femoral measurements (positive coefficients) with scarring on the dorsal pubis (negative coefficient). All of the features with positive coefficients, except the anteroposterior inlet diameter, were found to be correlated to age (see aforementioned discussion). The results of PC2 may thus simply be the result of the same sampling bias, supporting the results of the univariate analysis in showing that pubic tubercle and iliac tuberosity extension are age related. The sexes could not be separated according to PC2 (Figures 2 and 4), suggesting that the relationship of these scars to age is similar in the female and male samples. Further analyses of the role of age could not be tested in the PCA sample, because sample sizes of the age groups would be too small and would prevent confident interpretation of such results.

PC3 contrasted scarring at the posterior pelvic girdle with scarring at the anterior of the girdle. The inverse relationship between the ligaments of these two pelvic joints suggest that the ligamentous stabilization at these sites complement each other; that is, if one of the sites has well-developed ligaments (and thus increased scarring), the need for stabilization by the ligaments (and thus the scar formation) at the other site is reduced. This suggests that the causes of scarring at the sacroiliac joint (SIJ) region and at the pubic region differ. Scarring at the SIJ is likely linked to strains related to weight-bearing, whereas those at the dorsal pubic surface are due to stretching of the pubic ligaments in order to maintain stability of the pubic symphyseal joint (Houghton, 1975). It is possible that the changes at the pelvic joints are the result of changing strains on the girdle as a result of pregnancy and parturition; however, the fact that the sexes could not be separated according to PC3 (Figures 3 and 4) suggests that other factors are responsible for the modifications at the ligamentous attachment sites. One such factor may be body posture. Stability of the pelvic girdle is vital for efficient weight transfer from the trunk to the lower limbs, especially through the SIJ (Fraser, 1958). Should an individual change their posture to shift the transfer of weight more anteriorly (i.e. by assuming a lordotic posture or performing activities involving repeated trunk flexion), the strain on the anterior of the girdle would increase while that of the posterior girdle would decrease because of the compensating rotations of the ilia (Grant, 1958; DonTigny, 1979). Such changes could potentially result in an increase in scarring at the pubic symphysis but a decrease in scarring at the SIJ. It may be argued that the changing strains on the pelvic girdle in pregnant females might be the cause of scarring, but the fact that males also present with scarring, albeit to a lesser degree than in females, suggest that pregnancy-related strains are not the only causes of scar formation. It appears that weight-bearing strain is the common cause of scarring in both sexes, but that pregnancy-related strains amplify this strain in some individuals, hence the weak correlation between scarring and parity demonstrated in previous studies.

Conclusion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

The results of this study suggest that weight-bearing strains on the pelvic girdle and the associated ligamentous stabilization requirements may be a better explanation for scar development than pregnancy and/or parturition. The presence of scarring in both sexes and the lack of the expected reduction in scarring with age suggest that the factors influencing scar manifestation are common to both sexes and are possibly associated with age-related changes in activity patterns and/or body composition. The difference in occurrence of scarring between sexes may thus be the result of differences in body shape and size and the associated weight-bearing and stabilization requirements of the pelvic girdle between sexes. It is possible that the relationship between scarring and parity reported in the literature may have been due to changes in weight-bearing associated with pregnancy. Future studies may be able to test the conclusions of this study on samples of known parity history.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

The financial assistance of the National Research foundation (DAAD-NRF) towards this research is hereby acknowledged. Opinions expressed and conclusions arrived at are those of the authors and are not necessarily to be attributed to the DAAD-NRF.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Abitbol MM. 1996. The shapes of the female pelvis: Contributing factors. Journal of Reproductive Medicine 41(4): 242250.
  • Andersen BC. 1986. Parturition scarring as a consequence of flexible pelvic architecture. PhD Thesis, Department of Archaeology, Simon Fraser University, Canada.
  • Angel JL. 1969. The bases of paleodemography. American Journal of Physical Anthropology 30: 427437.
  • Arraiza B, Merbs CF. 2012. Evidence of childbirth in the pelvis of prehistoric Andean women. Estudios de Antropología Biológica 5(1): 6579.
  • Bruzek J, Murail P. 2006. Methodology and reliability of sex determination from the skeleton. In Forensic Anthropology and Medicine, A Schmitt, E Cunhan, J Pinheiro (eds.). Humana Press: New York; 225245.
  • Buikstra JE, Ubelaker DH. 1994. Standards for data collection from human skeletal remains. Arkansas Archaeological Survey Research Series No 44.
  • Cox MJ. 1989. An evaluation of the significance of “scars of parturition” in the Christ Church Spitalfields sample. Unpublished PhD Thesis, University College London, Great Britain.
  • Cox M, Scott A. 1992. Evaluation of the obstetric significance of some pelvic characters in an 18th century British sample of known parity status. American Journal of Physical Anthropology 89: 431440.
  • Cunningham FG, Leveno KJ, Bloom SL, Hauth JC, Rouse DJ, Spong CY. 2010. Williams Obstetrics (23rd Ed). McGraw-Hill Medical: New York; 2933.
  • Deurenberg P, Weststrate JA, Seidell JC. 1991. Body mass index as a measure of body fatness: Age- and sex-specific prediction formulas. British Journal of Nutrition 65: 105114.
  • DonTigny RL. 1979, Dysfunction of the sacroiliac joint and its treatment. Journal of Orthopaedic and Sports Physical Therapy 1: 2335.
  • Ebrahim GJ, Sullivan KR. 1995. Mother and Child Health Research Methods (2nd Ed). Book Aid: London; 186207.
  • Eufinger H, Gellrich NC, Sandmann D, Dieckmann J. 1997. Descriptive and metric classification of jaw atrophy. International Journal of Oral and Maxillofacial Surgery 26: 2328.
  • Fraser JE. 1958. Anatomy of the Human Skeleton (5th Ed). Churchill: London; 124136.
  • Galloway A, Snodgrass JJ, Suchey J. 1998. Markers of childbirth? Effect of body size and pubic morphological change. American Journal of Physical Anthropology 26(supplement): 102.
  • Grant JCB. 1958. A Method of Anatomy: Descriptive and Deductive (6th Ed). Williams & Wilkins: Baltimore.
  • Holt CA. 1978. A re-examination of parturition scars on the human female pelvis. American Journal of Physical Anthropology 49: 9194.
  • Houghton P. 1974. The relationship of the pre-auricular groove of the ilium to pregnancy. American Journal of Physical Anthropology 41: 381390.
  • Houghton P. 1975. The bony imprint of pregnancy. Bulletin of the New York Academy of Medicine 51(5): 655661.
  • Igarashi Y, Uesu K, Wakebe T, Kanazawa E. 2005. New method for estimation of adult skeletal age at death from the morphology of the auricular surface of the ilium. American Journal of Physical Anthropology 128: 324339.
  • İşcan MY, Loth SR, Wright RK. 1993. Casts of age Phases from the Sternal end of the rib for White Males and Females. France Casting: Bellvue, Colorado.
  • Kaiser HF. 1960. The application of electronic computers to factor analysis. Educational and Psychological Measurement 20: 141151.
  • Kelley MA. 1979. Parturition and pelvic changes. American Journal of Physical Anthropology 51: 541546.
  • Kurki H. 2007. Protection of obstetric dimensions in a small-bodied human sample. American Journal of Physical Anthropology 133: 11521165.
  • Kurki HK, Ginter JK, Stock JT, Pfeiffer S. 2010. Body size estimation of small-bodied humans: Applicability of current methods. American Journal of Physical Anthropology 141: 169180.
  • Mitra R. 2010. Osteitis condensans ilii. Rheumatology International 30: 293296.
  • Moerman ML. 1982. Growth of the birth canal in adolescent girls. Obstetrics and Gynecology 143: 528532.
  • Puoane T, Steyn K, Bradshaw D, Laubscher R, Fourie J, Lambert V, Mbananga N. 2002. Obesity in South Africa: The South African demographic and health survey. Obesity Research 10(10): 10381048.
  • Putschar WGJ. 1976. The structure of the human symphysis pubis with special consideration of parturition and its sequelae. American Journal of Physical Anthropology 45: 589594.
  • Ridgeway B, Arias BE, Barber MD. 2011. The relationship between anthropometric measurements and the bony pelvis in African American and European American women. International Urogynecology Journal 8: 10191024.
  • Rosenberg K. 1988. The functional significance of Neanderthal pubic length. Current Anthropology 29: 595617.
  • Ruff CB. 1987. Sexual dimorphism in human lower limb bone structure: relationship to subsistence strategy and sexual division of labor. Journal of Human Evolution 16(5): 391416.
  • Ruff CB, Scott WW, Liu AYC. 1991. Articular and diaphyseal remodelling of the proximal femur with changes in body mass in adults. American Journal of Physical Anthropology 86: 397413.
  • Schroeder CF, Schmidtke SZ, Bidez MW. 1997. Measuring the human pelvis: A comparison of direct and radiographic techniques using a modern United States-based sample. American Journal of Physical Anthropology 103(4): 471479.
  • Snodgrass JJ, Galloway A. 2003. Utility of dorsal pits and pubic tubercle height in parity assessment. Journal of Forensic Sciences 48: 12261230.
  • Statsoft Inc. 2009. STATISTICA® data analysis software, version 9. Tulsa, OK.
  • Stewart TD. 1970. Identification of the scars of parturition in the skeletal remains of females. In Personal Identification in Mass Disasters, TD Stewart (ed.). Smithsonian Press: Washington, DC; 127135.
  • Suchey JM, Wiseley D, Green RF, Noguchi TT. 1979. Analysis of dorsal pitting in the os pubis in an extensive sample of modern American females. American Journal of Physical Anthropology 51: 517540.
  • Tague RG. 1989. Variation in pelvic size between males and females. American Journal of Physical Anthropology 80: 5971.
  • Ubelaker DH, De La Paz JS. 2012. Skeletal indicators of pregnancy and parturition: A historical review. Journal of Forensic Sciences 57(4): 866-872.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
oa2402-sup-0001-Supplementary.docWord document38KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.