2023 Volume 131 Issue 2 Pages 79-88
Developmental variations in the cranial vault arise during early ontogenesis due to premature or delayed suture closure, non-union between the normal ossification centers, and emergence of additional ossification centers. These variations cause abnormal calvarial partition and are often associated with modifications of the overall cranial configuration. This study aims to evaluate the impact of metopic suture persistence and the development of supernumerary bones in the occipital region on neurocranial morphology. A series of 245 crania of Bulgarian males was investigated. According to the presence or absence of developmental variations in the frontal and occipital bone they were divided into four series: control (n = 115); crania with a retained metopic suture (n = 32); crania with supernumerary bones in the occipital region (n = 67); and metopic crania with supernumerary bones in the occipital region (n = 31). Polygonal 3D models were generated, on which the 3D coordinates of 35 landmarks describing the neurocranium were collected. Geometric morphometric analyses were applied to compare the size and shape of the landmark configurations. Significant size differences were observed only in the frontal and occipital bone configurations between the metopic crania with occipital supernumerary bones and the non-metopic crania. Significant shape differences were found in all comparisons between the metopic and non-metopic groups for all configurations, except the occipital one. Metopism has a significant impact on overall neurocranial morphology, which is intensified by the presence of supernumerary bones in the occipital region.
Calvarial bones are intramembranous in origin and arise after the second month of gestation from definite ossification centers, which further unite to form these bones. The separate calvarial bones are interconnected through sutures (Tubbs et al., 2012). It has been suggested that the initial attachment sites of dura mater to the cartilaginous cranial base, which give rise to the dural reflections, conform to the early recesses in the developing brain and predetermine the location of the major sutures (Smith and Töndury, 1978). Between the 12th and 16th weeks of gestation, intramembranous ossification progresses in the central zones between the major reflective bands of the dura. By the 16th week, the radiating centers of ossification approach the sites of reflective bands in the dura, which remain unossified regions of connective tissue between the outspreading islands of membranous bone (Smith and Töndury, 1978). The formation of the calvarial sutures initiates when the osteogenic fronts of the growing bone plates approximate one another and set up a gradient of growth factor signaling between them. The opposite bone fronts of the midline sutures (sagittal, interfrontal) do not overlap, whereas the transversely situated sutures (lambdoid, coronal) do overlap (Opperman, 2000). Sutures are bone growth sites and allow enlargement of the calvarial bones (Cohen, 1993). Usually, the sutures function and stay patent until early adulthood when they gradually obliterate (Nikolova et al., 2019). The metopic suture is the only exception, inasmuch as it normally closes by the end of the first postnatal year (Weinzweig et al., 2003). Since the sutures play a critical role in cranial morphogenesis, any deviation from the normal timing of suture formation, functioning, and closure results in an abnormal cranial configuration.
Developmental variations in the cranial vault arise during embryogenesis and early postnatal life due to premature or delayed suture closure, non-union between normal ossification centers, and emergence of additional ossification centers. These variations cause abnormal partition of the cranial vault and are often associated with specific modifications of the overall cranial configuration. The unusual obliteration of the squamous suture has been found to be related to slight bilateral differences in the middle and posterior part of the neurocranium (Nikolova et al., 2021a). It has also been found that the condition of a bipartite frontal bone in adult individuals, known as metopism, is related to specific characteristics including frontal sinus underdevelopment (Nikolova et al., 2018a, 2018b), significant delay of sagittal suture closure (Nikolova et al., 2022a), supernumerary bones (Hanihara and Ishida, 2001a; Nikolova et al., 2016, 2020), and modification of cranial morphology (Nikolova et al., 2021b, 2022b). Metopism has been reported to occur in different population groups with varying frequencies from 0% to exceeding 15% (Berry and Berry, 1967; Hauser and De Stefano, 1989; Hanihara and Ishida, 2001a; Zdilla et al., 2018)
The squamous part of the occipital bone is composed of interparietal and supraoccipital parts. The interparietal part ossifies intramembranously during the third and fourth intrauterine months from several ossification centers. Faulty fusion between these centers and to the supraoccipital part gives rise to many variations in this region (Srivastava, 1992; Matsumura et al., 1993; Nikolova et al., 2014a). In such cases, the resultant complete or incomplete interparietal (Inca) bone has been reported to occur with а varying frequency of less than 1% to more than 10% (Kadanoff and Mutafov, 1984; Hauser and De Stefano, 1989; Hanihara and Ishida, 2001b). In addition, the posterior and posterolateral (mastoid) fontanels as well as the lambdoid suture are common sites where additional ossification centers emerge and give rise to wormian bones, which according to their location could be defined as fontanel bones (posterior fontanel bone, ossicle in the asterion) and sutural bones within the lambdoid suture.
Metopism and supernumerary calvarial bones are commonly found together (Hanihara and Ishida, 2001a). Recently, it has been demonstrated that metopism is related to a modification of the overall morphology of the neurocranium (Nikolova et al., 2022b), but without taking into account the presence of supernumerary bones. As far as we know, there is no comparative study on the potential impact of developmental variations in the occipital region on the morphology of the neurocranium. Therefore, this study has been designed to compare the shape and size of the neurocranium in metopic and non-metopic crania with and without variations in the occipital bone. Thus, we aim to assess the particular as well as the combined impact of metopism and supernumerary bones in the occipital region on cranial morphology.
A series of 245 male crania from the Ossuary at the National Museum of Military History (Bulgaria) was investigated. The osteological material stored in the Ossuary belonged to Bulgarian soldiers who participated and died in wars at the beginning of the 20th century. There was no available information on the age at death for all investigated individuals. According to the Museum’s archives, the youngest recruits were 17 years old and the oldest ones were 62 years old. The Human Research Ethics Committee at the Institute of Experimental Morphology, Pathology and Anthropology with Museum, Bulgarian Academy of Sciences approved investigation of the osteological collection (protocol number 7/30.10.2018).
Based on the presence or absence of developmental variations in the frontal and occipital bone, the investigated crania were organized into four series: control crania without variations in the frontal and occipital bone (C; n = 115); metopic crania (MS; n = 32); crania with supernumerary bones in the occipital region (SB; n = 67); crania with metopism and supernumerary bones in the occipital region (MS–SB; n = 31). The occipital region was inspected for the presence of the following variations: an interparietal bone (Inca bone); a preinterparietal bone; a posterior fontanelle bone; an ossicle in the asterion; a large single sutural bone in the lambdoid suture; or three or more small sutural bones in the lambdoid suture (Figure 1). The variations were defined according to the classification suggested by Nikolova et al. (2014a).
Developmental variations in the occipital region: (a) interparietal (Inca) bone; (b, c), bipartite preinterparietal bone; (d) posterior fontanel bone and sutural bones in the lambdoid suture.
A handheld laser scanner (Creaform VIUscan) was used to generate 3D polygonal images of the crania. The scanner accuracy reported by the manufacturer was up to 0.05 mm. The parameters established to be optimal for dry bone scanning were a scanning resolution of 0.40 mm and a texture resolution of 150 DPI (Toneva et al., 2017, 2020), and these were applied to each of the scanning sessions. The generated 3D surface images were postprocessed using the software platform VXelements. Image postprocessing included deletion of the artifacts captured during the scanning. The 3D images were exported in the Wavefront 3D Object (OBJ) file format, which supports texture mapping.
On the generated polygonal models, a total of 35 (11 midsagittal and 12 bilateral) landmarks (Figure 2; Supplemental Table S1) were digitized in MeshLab v. 2016.12 (Cignoni et al., 2008) and their 3D coordinates were recorded. The landmarks were grouped into four sets describing the configuration of the entire neurocranium as well as its parts: the frontal bone, the parietotemporal region, and the occipital bone (Supplemental Table S1).
Location of the landmarks: (a) frontal view; (b) lateral view; (c) occipital view; (d) basilar view. The full name, the abbreviation, and the description of the landmarks are given in Supplemental Table S1.
Geometric morphometric (GM) analyses were applied to compare the size and shape of the landmark configurations in the investigated series. GM analyses are based on landmark coordinates and represents a set of methods to analyze data that capture the geometry of morphological structures (Bookstein, 1991; Adams et al., 2004). It enables detailed examination of morphological structures and processes and visualization of the statistical results as shapes or forms (Zelditch et al., 2004; Mitteroecker and Gunz, 2009).
SizeThe size of each configuration was measured by centroid size, i.e. the square root of the sum of squared distances of all landmarks in the configuration to their centroid (Bookstein, 1991). The size differences between the investigated four groups were tested by one-way ANOVA with Tukey’s post-hoc test for multiple pairwise comparisons. The normality and homoscedasticity of the centroid size data were tested in advance using the Shapiro–Wilk normality test and Levene’s equal variance test, respectively.
ShapeGeneralized Procrustes Analysis (GPA) was applied to the landmark data to remove the effects of the location, scale, and orientation of each configuration. GPA involves superimposing a set of landmark configurations by centering/translating, scaling, and rotating the configurations around the centroid until the sum of the squared Euclidian distances between the homologous landmarks is minimal, where the rotation is an iterative step (Gower, 1975; Rohlf and Slice, 1990). Procrustes shape coordinates of each configuration were computed as a result of the GPA. The data for the symmetric component of shape were used in the subsequent analyses. The effect of allometry was evaluated by applying multivariate regression of the Procrustes coordinates on the log-transformed centroid size. The regression separates the component of shape variation that is predicted by size from the residual component of variation, which is uncorrelated with size (Klingenberg, 2016). The multivariate regression was applied to the data of each configuration and the proportion of total variation for which the regression accounted and the P-value of the associated permutation test (number of permutation rounds, 10000) were reported. If the allometric relationship between size and shape was statistically significant, the residuals from that regression were used further as shape variables.
Principal Component Analysis (PCA) was applied to explore the shape variation in the sample and for dimensionality reduction, since a smaller number of uncorrelated variables (principal components, PCs) were produced for further analysis. The distribution of the specimens in shape space was demonstrated by plotting the first two PCs. The shape configurations associated with the extreme positive and negative PCs (reconstructed at 0.1 Procrustes units) were visualized using wireframe graphs. The GPA, multivariate regression, PCA, and the visualization of the associated wireframe configurations were performed in MorphoJ, v. 1.07a (Klingenberg, 2011).
The shape differences between the groups were evaluated based on the PCs retaining 95% of the shape variance of the original data. Thus, we reduced the number of dependent variables to a smaller number than the size of the smallest subgroup in our study. The statistical significance of the shape differences between the four groups were tested applying one-way PERMANOVA with Bonferroni correction for pairwise comparisons. PERMANOVA is a non-parametric test used for testing significant differences between two or more groups, based on a definite distance measure (Anderson, 2001). We used the Euclidean distance as a type of intersample proximity computation. The significance test in PERMANOVA was calculated by permutation of group membership with 9999 replicates. The multivariate normality and equality of covariance matrices were tested. The multivariate normality was evaluated by Mardia’s multivariate skewness and kurtosis tests (Mardia, 1970) and the Doornik and Hansen omnibus test (Doornik and Hansen, 1994). If even one of these tests showed a significant P-value, the distribution was considered significantly non-normal. The equality of covariance matrices was tested by Box’s M test. PERMANOVA was used in our study, because of the violation of the assumptions of multivariate normality in some of the groups and/or equality of covariance matrices.
Canonical Variate Analysis (CVA) was applied to the sets of PCs describing 95% of the variance to visualize the separation of the groups for each configuration after maximizing the differences between the groups. CVA tries to find dimensions in shape space that maximize the ratios of the between-groups to the within-groups distances. The multivariate normality and Box’s M tests, PERMANOVA, and CVA were performed in PAST, v. 4.07b (Hammer et al., 2001).
Intraobserver errorThe intraobserver error of digitization was evaluated for all landmarks based on their raw coordinates. For that purpose, 25 (13 non-metopic and 12 metopic) crania were digitized three times by one observer. The digitization error of each landmark was estimated in millimeters based on the landmark standard deviation. The standard deviation was calculated using the Euclidean distances measured from the different placements of the landmark to its centroid (von Cramon-Taubadel et al., 2007). The average of the landmark standard deviations calculated for all 25 crania was used as a measure of the intraobserver error of digitization. Values of the interobserver error within 1 mm were considered acceptable; values between 1 and 2 mm were found useful, while those of more than 2 mm were regarded as inappropriate for further analyses (Lagravère et al., 2010).
Summarized data for the types and distribution of the observed variations in the occipital region are presented in Table 1.
| Variations/series | Non-metopic (C and SB, n = 182; SB, n = 67) | Metopic (MS and MS–SB; n = 63; MS–SB, n = 31) | ||||
|---|---|---|---|---|---|---|
| n | %* | %** | n | %* | %** | |
| Interparietal (Inca) bone | 2 | 1.1 | 3.0 | 4 | 6.3 | 12.9 |
| Preinterparietal bone | 6 | 3.3 | 9.0 | 6 | 9.5 | 19.4 |
| Posterior fontanel bone | 15 | 8.2 | 22.4 | 6 | 9.5 | 19.4 |
| Ossicle in the asterion | 17 | 9.3 | 25.4 | 1 | 1.6 | 3.2 |
| Sutural bones in the lambdoid suture | 54 | 29.7 | 80.1 | 27 | 42.9 | 87.1 |
*Percentage of the total number of non-metopic/metopic crania. **Percentage of the non-metopic/metopic crania with variations in the occipital bone.
The digitization errors of most of the landmarks were less than 1.0 mm (Supplemental Table S2). More than half of the landmarks had values of less than 0.5 mm. The only three landmarks that reached errors of more than 1.0 mm were the opisthocranion and the right and left eurion. The digitization errors obtained for all landmarks were assessed as acceptable for the purposes of the study.
Centroid size differencesSignificant intergroup differences in the centroid size were observed only in the frontal and occipital bone configurations. The frontal bone configuration was considerably larger in the MS–SB crania compared with the C and SB series. The occipital bone configuration was significantly smaller in MS–SB compared with C (Table 2, Table 3).
| Landmark configuration | C | SB | MS | MS–SB | ||||
|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| Neurocranium | 457.73 | 11.24 | 457.48 | 10.46 | 455.24 | 9.60 | 455.63 | 11.39 |
| Frontal bone | 221.39 | 7.18 | 221.26 | 6.49 | 224.31 | 6.07 | 226.10 | 7.24 |
| Parietotemporal region | 295.78 | 7.49 | 297.15 | 6.82 | 293.17 | 7.41 | 294.40 | 8.99 |
| Occipital bone | 146.47 | 5.42 | 146.12 | 5.62 | 144.96 | 5.03 | 143.45 | 5.41 |
| Landmark configuration | One-way ANOVA | Pairwise comparisons (Tukey’s post-hoc test), P-value | ||||||
|---|---|---|---|---|---|---|---|---|
| F | P | C:SB | C:MS | C:MS–SB | SB:MS | SB:MS–SB | MS:MS–SB | |
| Neurocranium | 0.544 | 0.653 | NS | NS | NS | NS | NS | NS |
| Frontal bone | 5.315 | 0.001* | NS | NS | 0.003* | NS | 0.006* | NS |
| Parietotemporal region | 2.317 | 0.076 | NS | NS | NS | NS | NS | NS |
| Occipital bone | 2.906 | 0.035* | NS | NS | 0.028* | NS | NS | NS |
* Statistically significant difference at P < 0.05. NS, non-significant.
The results of the multivariate regression of the Procrustes coordinates on centroid size showed that size has a significant impact on shape in all investigated configurations (Table 4). The size of the neurocranium and parietotemporal configurations accounted for a very small percentage of the shape variation (<1%), while the percentage of the shape variation explained by size for the occipital bone configuration was >4%. Since all landmark configurations showed a significant effect of size on shape, the regression residuals, i.e. the allometry-corrected shape variables of each configuration, were used as input variables for further analyses.
| Landmark configuration | % predicted shape variation | p-value (permutation test) |
|---|---|---|
| Neurocranium | 0.94 | 0.002 |
| Frontal bone | 2.14 | <0.001 |
| Parietotemporal region | 0.83 | 0.035 |
| Occipital bone | 4.09 | <0.001 |
All landmark configurations, except that of the occipital bone, showed significant shape differences between the groups. The whole neurocranium, the frontal bone, and the parietotemporal region were considerably different between the series: C and MS; C and MS–SB; SB and MS; SB and MS–SB (Table 5).
| Landmark configuration | Total sum of squares | Within-group sum of squares | F | p-values | Pairwise comparisons (Bonferroni corrected p-value) | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| C:SB | C:MS | C:MS–SB | SB:MS | SB:MS–SB | MS:MS–SB | |||||
| Neurocranium | 0.782 | 0.730 | 5.697 | <0.001* | ns | <0.001* | <0.001* | <0.001* | <0.001* | ns |
| Frontal bone | 1.040 | 0.971 | 5.660 | <0.001* | ns | <0.001* | <0.001* | <0.001* | <0.001* | ns |
| Parietotemporal region | 0.918 | 0.835 | 8.008 | <0.001* | ns | <0.001* | <0.001* | <0.001* | <0.001* | ns |
| Occipital bone | 1.476 | 1.450 | 1.403 | 0.134 | ns | ns | ns | ns | ns | ns |
ns—non-significant
*—statistically significant difference at p < 0.05
PCA of the neurocranium configuration in the investigated series indicated variation from a rounded overall shape with a protruding metopion region at the extreme negative PC1 values to a narrow and prolonged shape with a sloping frontal bone at the extreme positive ones (Figure 3a). There was no clear separation between the investigated groups. However, most of the metopic crania and mainly those with supernumerary bones in the occipital region (MS–SB) were placed in the negative part of the PC1 axis, whereas the majority of the non-metopic crania (C and SB) were in the positive part of the axis.
Scatterplots of the shape variations in PC1 vs PC2 and the related wireframes showing the shape configurations at PC1 and PC2 extremes: (a) neurocranium; (b) frontal bone; (c) parietoparietal region; (d) occipital bone. Legend: blue squares, control series (C); pink asterisks, crania with supernumerary bones in the occipital region (SB); green dots, metopic crania (MS); brown diamonds, metopic crania with supernumerary bones in the occipital region (MS–SB); light blue wireframes, mean shape; dark blue wireframes, extreme shape configurations along the PC axes.
The major shape variations in the frontal bone configuration were expressed by changes in the relative bone width and the protrusion of the metopion region. The investigated series largely overlapped, although there was some separation along the PC2 axis, since most of the MS and MS–SB crania had positive values corresponding to a relatively shorter, wider, and more projecting frontal squama at the metopion with a flat glabellar region. The C and SB crania were more dispersed along the PC2 axis, but most of them had negative PC2 scores indicating a longer, narrow, and more inclined frontal squama with a projecting glabellar region (Figure 3b).
The shape changes in the parietotemporal region were associated with its relative width and the position of both landmarks eurion. The investigated series were not clearly separated along the first two PCs. However, with few exceptions, the MS and MS–SB crania were placed in the positive part of the PC1 axis, corresponding to a relatively wider parietotemporal region with more anteriorly and inferiorly located eurions. The C and SB crania were more dispersed along the PC1, but most of them were located in the negative part of the PC1 axis associated with the reverse configuration (Figure 3c).
The observed variation in the occipital bone shape among the investigated crania concerned mainly the position of the lambda, opistocranion, and inion. Distribution of the investigated crania did not show any separation between the predefined series according to the occipital bone shape (Figure 3d).
The PCA outputs based on the regression residuals of the investigated configurations are given in Supplemental Table S3.
Canonical variate analysisThe scatterplots of the first two canonical variates (CVs) for each configuration are presented in Figure 4a–d. Considering the shape of the whole neurocranium in our sample, it was obvious that the C and SB groups covered one another entirely, whereas the MS and MS–SB groups were clearly distinguishable one from another (along the CV2 axis) as well as from the overlapping C and SB series (along the CV1 axis) (Figure 4a).
CVA scores for the landmark configuration: (a) neurocranium; (b) frontal bone; (c) parietoparietal region; (d) occipital bone. Legend: blue squares, control series (C); pink asterisks, crania with supernumerary bones in the occipital region (SB); green dots, metopic crania (MS); brown diamonds, metopic crania with supernumerary bones in the occipital region (MS–SB).
Frontal bone shape did not clearly separate the investigated series, since all groups had overlapping regions. However, along the CV1 axis there was some separation between the non-metopic (C and SB) and the metopic crania (MS and MS–SB) (Figure 4b).
The shape of the parietotemporal region also showed an overlap between the groups. Once again, here could be seen some separation along the CV1 axis between the non-metopic (C and SB) and metopic (MS and MS–SB) crania (Figure 4c).
Considering the occipital bone configuration, there was no separation between the investigated groups, since all of them significantly overlapped along the CV1 and CV2 axes (Figure 4d).
The CVA results for the investigated configurations are given in Supplemental Table S4.
The main cranial morphogenetic processes occur in the early ontogenetic stages. There are different mechanisms causing abnormal calvarial partition and formation of supernumerary bones, which are briefly considered here. In metopism, a normal suture formation occurs, but the time of its functioning is unusually prolonged, which results in a non-typical partition of the vault. Supernumerary calvarial bones could arise as a result of fragmentation or non-fusion of the normal ossification centers. Such variations frequently occur in the intramembranous portion of the occipital squama (Srivastava, 1992; Matsumura et al., 1993; Koyabu et al., 2012). The interparietal part ossifies intramembranously and occupies the area above the highest nuchal line to the triangular space between the two parietal bones. According to Koyabu et al. (2012), the interparietal consists of four basic elements: one medial pair and one lateral pair. In humans, the medial elements initially fuse with each other, whereas the lateral elements remain separate from the medial element, producing a tripartite interparietal, which unifies further as a single entity. Non-fusion between the ossification centers gives rise to numerous variations in this part of the skull, such as the interparietal or Inca bone (Kadanoff and Mutafov, 1984; Hanihara and Ishida, 2001b) and the preinterparietal bone (Matsumura et al., 1993). Wormian bones are supernumerary bones, which have no regular relationship with the normal ossification centers. They are inconsistent but of frequent occurrence and may be located anywhere in or between the normal sutures and fontanels (Parker, 1905). Nevertheless, they appear more frequently in the posterior sutures, in particular within the lambdoid suture (Sanchez-Lara et al., 2007; Nikolova et al., 2014b). A fontanel bone with an irregular or rounded shape may arise from an additional ossification center in the area of the posterior fontanel (Kadanoff and Mutafov, 1984). From the fourth fetal month onwards, additional ossification centers may also develop in the posterolateral fontanels (Hauser and De Stefano, 1989). If they do not fuse with the neighboring bones, they give rise to supernumerary fontanel bones.
A frequent co-occurrence of a persistent metopic suture and supernumerary bones has previously been observed (Hanihara and Ishida, 2001a; Nikolova et al., 2020). Comparing the developmental variations in the occipital region in the metopic and non-metopic groups in our study, a few general tendencies could be observed (Table 1). First of all, the total frequency of supernumerary bones in the occipital region is higher in metopic crania than in non-metopic ones. Secondly, among all observed variations, those due to a non-fusion between the ossification centers of the interparietal part of the occipital squama are more frequent in metopic than in non-metopic crania. Finally, variations arising due to the presence of additional ossification centers are proportionately distributed in both series, excluding the ossicle in the asterion, which is a relatively common finding in non-metopic crania and very rare in metopic crania.
Considering the centroid size, significant differences are found only in the frontal and occipital bone configurations. The frontal bone in the non-metopic crania (C and SB) is significantly smaller, but only compared to the MS–SB. At the same time, the occipital configuration in the MS–SB is significantly smaller compared with the control. It has been established that the size of the neurocranium in metopic crania does not differ significantly compared with control series, but the size of its separate parts does. The modification is expressed by a significant enlargement of the frontal bone and reduction of the parietotemporal region and the occipital bone size (Bryce, 1917; Nikolova et al., 2022b). This study shows that this modification is considerable only in metopic crania with supernumerary bones in the occipital region, since the comparisons between the non-metopic crania (C and SB) and the MS series do not show any significant size differences.
The shape differences are significant in all comparisons between the metopic and non-metopic groups for the configurations of the whole neurocranium, the frontal bone, and the parietotemporal region. The supernumerary bones in the occipital bone do not seem to make any meaningful contribution to the shape differences considering comparisons within the metopic (MS, MS–SB) and non-metopic (C, SB) crania.
CVA indicates that the non-metopic crania (C and SB) entirely overlap one another and could not be separated by shape. A separation between the MS and MS–SB could be seen in the shape of the whole neurocranium. However, the local configurations do not show such a division. Except for the occipital bone, which does not show any separation between the groups, the other configurations indicate some separation between the non-metopic and metopic crania. All this indicates that the main factor contributing to shape differences in cranial morphology between the investigated series is metopic suture persistence. The supernumerary bones in the occipital region do not impact the shape in the non-metopic crania, but they contribute to the shape differences within the metopic crania. Nevertheless, CVA is sensitive to violations of the assumptions for normality and equality of covariance matrices, which is observed in some of our groups, and this should be kept in mind.
The obtained results clearly show that the investigated developmental variations in the occipital region are not associated with any modifications of cranial morphology, since no significant size and shape differences are found within the non-metopic and metopic groups. This finding is quite surprising because a functioning suture allows bone growth in a direction perpendicular to the suture line. In metopic crania, the prolonged period of metopic suture functioning is linked primarily, but not exclusively, to a modification in frontal bone morphology. It seems, however, that the formation and persistence of active sutures in the occipital bone region is not related to any significant size and shape changes. Nevertheless, it cannot be excluded that the limited sample size and the inclusion of various types of supernumerary bones in the SB and MS–SB series have had an impact on the obtained results.
Metopism and supernumerary bones in the occipital region separately do not appear to be associated with significant size modifications in neurocranial morphology. Only the crania combining metopism and supernumerary bones in the occipital region (MS–SB) show considerably larger frontal bones compared with the non-metopic crania (C and SB) and significantly smaller occipital bones compared with the control. The shape of the whole neurocranium, the frontal bone, and the parietotemporal region, differs significantly in all comparisons between the metopic and non-metopic groups, showing the important link between metopic suture persistence and the shape of the neurocranium. The insignificant differences within the groups of the non-metopic (C and SB) and metopic crania (MS and MS–SB) indicate that supernumerary bones alone are not linked to any specific changes in shape. It can be concluded that metopism has a significant impact on overall neurocranial morphology, which is intensified by the presence of developmental variations in the occipital region.
This study was supported by the Bulgarian National Science Fund, Grants DN11/9-15.12.2017 and КП-06-H51/4–11.11.2021.
Authorship contribution statementSilviya Nikolova: conceptualization, methodology, investigation, data curation, visualization, project administration, funding acquisition, writing—original draft. Diana Toneva: conceptualization, methodology, formal analysis, investigation, data curation, validation, visualization, project administration, funding acquisition, writing—review and editing. Nikolai E. Lazarov: writing—review and editing.
