Developmental processes are extremely complex processes that lead to the creation of a final or current phenotype. Due to their extreme complexity, a perfectly regular and symmetrical process is almost impossible to achieve. The balance of all processes with the interaction of external environmental factors implies the achievement of full functionality of the system with primary or secondary morphological construction. Such construction in its course leaves on the morphology of the organism small markers that can serve as an indicator of certain factors of development, both environmental and genetic (1). The growth and development of the human craniofacial system is under significant control of genetic factors, i.e. under the influence of certain groups of genes whose activation and expression are well programmed. Control genetic mechanisms are largely non-specific in nature, such as regulatory genes that act not only on the craniofacial system, but on the whole organism. On the other hand, there are very specific control patterns of development that establish a species-specific physiognomy with relatively small morphological variations. Although human face variations appear to be large, they are actually minimal given the complexity of the development cycle (2). Different parts of the face have different development in terms of time, speed and direction of growth and development (3). A significant role in the duration of these processes, as well as in lifelong changes is played by different hormonal interactions and the time of action and concentration of individual hormones, which, each with its own complete or additive effect, contributes to the change of morphological characteristics of the face. Asymmetry is a deviation from the identical shape or size of two sides of the body. It is one of the most common facial features. Due to the lack of use of the mid-sagittal plane, mirror object modelling can provide the most accurate assessment of craniofacial asymmetry. It is widely believed that increased asymmetry indicates the failure of the organism to maintain developmental homeostasis and the overcoming of environmental disorders over genetic control mechanisms (4). Developmental stability means the body's resistance to various disorders that can affect the body during growth and development, and we can express it by slight random deviations from bilateral symmetry (5). Since fluctuating asymmetry is caused by the inability of the organism to allow the left and right sides to develop equally, it is often used in research as a measure of developmental instability that occurs during the development of the organism (6, 7). If an environmental factor changes, the result is a change in directional asymmetry because this stress acted on both the left and right sides of the body, as a developmental response to environmental factors, depending on positional signals regulating the differences between left and right. Therefore, the variation of directional asymmetry with developmental factors is a special example of a reaction norm, something that affects both sides of the body. Considering that the significant role of heredity in facial development has been pointed out so far, a study of the developmental stability of the facial phenotype would indicate the relationship and interaction of genetic and environmental factors during the growth and development of the craniofacial system. Despite the overall low level of differentiation in the human population, local factors such as geographical and cultural isolation can greatly affect genetic discontinuity. Clearly distinguished genetic isolates are very valuable for mapping rare and elucidating the mechanism of occurrence of complex genetic diseases (8). Therefore, this study of the asymmetry and shape of the craniofacial skeleton was conducted on an isolated population of Croatian Dalmatian islands and on a panmictic (non-isolated) population of the city of Split. Aim of the study The aims of the study were: 1. To calculate the amounts of total asymmetry of the entire craniofacial skeleton, cranium and mandible and determine whether they are grouped by age and sex and study them in a group of subjects from the area of Korčula, Vis and Split. 2. To determine whether there is direct asymmetry of the entire craniofacial skeleton, cranium and mandible in the group of subjects from the area of Korčula, Vis and Split and, if so, whether there is a difference between these groups and what is its direction. 3. To determine whether there is a fluctuating asymmetry of the entire craniofacial skeleton, cranium and mandible in the group of subjects from the area of Korčula, Vis and Split and, if so, whether there is a difference between these groups and what is its direction. 4. To determine the existence of sex and age differences in the form of the craniofacial skeleton and depending on the place of residence of the subjects (Korčula, Split, Vis). Sample and methods The studied population of the islands of Korčula and Vis is a rare and valuable example of a geographically and reproductively isolated population. On the other hand, the population of the city of Split is an example of a panmictic population. Since it shares most of the environmental factors with the population of Vis and Korčula, the population of the city of Split was used as a control population. This group is representative of a mixed and nonisolated population. A total of 110 CBCT images were used for this study. Out of the total number, 46 images belong to the inhabitants of the island of Vis, 15 to the inhabitants of Korčula, and 49 images belong to the population of the city of Split. All images were taken with the same CBCT device (Scanora 3D, Soredex, Finland) and belong to the Project database. All images are of the same dimensions, field size (FOV) 10 x 14.5 cm, isotropic voxel size 0.25 mm with a total of 300 projection images per subject. The images were stored in the Digital Imaging and Communication in Medicine (DICOM) format. Three-dimensional data reconstruction and three-dimensional anthropometric measurements were performed in Viewbox 4 software (dHAL, Greece). After loading the three-dimensional image and showing the craniofacial skeleton, the region of interest for each subject was determined individually. It included the entire craniofacial complex, the cranium from the upper bone edge of the left and right orbits to the lower edge of the maxilla, and the entire mandible. The following anthropometric points in the mid-sagittal plane were used: prosthion (pr), anterior nasal spine (sna), posterior nasal spine (snp), incisive foramen (fi), linguale (lng) and infradentale (id). The paired points were: orbital (or), pyriform aperture (ap), zygoorbital (zo), zygomaxillare (zm), mandibular foramen (fma), mental foramen (fme) and coronoid (cn). After digitization, the anthropometric points were determined by their coordinates (x, y, z) and analysed by geometric morphometrics in Mathematica 4.1 (Wolfram Research, Inc., United Kingdom). The following methods were used to analyse the shape and symmetry of the craniofacial skeleton: General Procrustean method of superposition (for standardization of object size), modelling of an image of an object, and principal components analysis (PCA) (9). Mathematical procedures for determining the total asymmetry were performed according to Mardia et al. (10). The degree of individual morphological asymmetry was calculated as the difference in shape (Procrustean distance) between the configuration of the face and its mirror reflection. The first determined total asymmetry (total asymmetry - TA) of the sample was divided into directional asymmetry (DA) and fluctuating asymmetry (fluctuation asymmetry - FA). The principal components analysis (PCA) indicated the distribution of dominant shape variations within different populations and presented the main differences between them (11). Differences in asymmetry amounts were checked by analysis of variance (ANOVA). Results Total asymmetry The total asymmetry for the area of the entire craniofacial skeleton as well as for the cranium and mandible expressed in units of Procrustes distance is not grouped according to age or sex. ANOVA showed that there was no statistically significant difference (R 2= 0.094) in the amount of total asymmetry neither by age and sex of the subjects, nor by place of residence (Vis and Korčula), nor by the amount of inbreeding. Subjects from the area of Korčula differ statistically significantly in the amount of total asymmetry for the area of the mandible compared to the subjects from Vis and Split. Directional asymmetry The difference in asymmetry between the sexes and between populations depending on the place of residence has not been established. The amount of directional asymmetry is the smallest for the mandible for all three groups of subjects. Subjects from the area of Vis have the highest amount of directional asymmetry for the entire area of the craniofacial skeleton and for the area of the cranium, and subjects from the area of Korčula have the highest amount of directional asymmetry of the mandible. Changes in the direction of directional asymmetry after the age of 55 are visible in all subjects in the orbital and nasal area, while changes in the mandibular area are noticeably small. Changes in the direction of directional asymmetry do not intensify with age. The largest changes are shown by the pyriform aperture point in subjects younger than 55. As the aperture rotates counter clockwise, the right one moves laterocaudally and the left one moves mediocranially. Fluctuating asymmetry There was no statistically significant difference in the amount of fluctuating asymmetry, neither by age and sex of the subjects, nor by place of residence (Vis and Korčula), nor by the amount of inbreeding. Changes in the direction of fluctuating asymmetry are visible in all subjects on the entire craniomandibular skeleton. Changes in the direction of fluctuating asymmetry are more pronounced in subjects older than 55. The largest changes are visible in the zygoorbital, nasal, and coronoid areas. Shape analysis An F test showed that there was a statistically significant difference in the shape of the craniofacial skeleton between female and male subjects (p= 0.025). Image 1. Dependence of the change in the shape of the craniofacial skeleton in relation to age in male and female subjects Changes in the shape of the craniofacial skeleton are greater in female subjects than in male subjects. After the age of 55 (Image 1), the curve of the change in the shape of the craniofacial skeleton in female subjects, which is age-dependent, has a pronounced ascending trajectory, after which it becomes discontinuous with unpredictable ascending and descending variations. The trajectory curves of men and women are approximately parallel until approximately 55 years of age when in women, the pattern of aging changes and changes in the shape of the craniofacial skeleton caused by aging occur. Changes in the shape of the craniofacial skeleton in female subjects are more intense than changes in male subjects. In contrast to the intensity, the directions of changes in the shape of the craniofacial skeleton in relation to age in male and female subjects are similar. Changes in the coronoid, anterior nasal spine, and posterior nasal spine points caused by increasing age occur only in female subjects. Although the subjects are not grouped into clusters according to their place of residence (Korčula, Split, Vis) and there are many overlaps in the shape of the craniofacial skeleton, the F test showed that there is a statistically significant difference in the shape of the craniofacial skeleton between Korčula and Vis. Discussion The results of this study showed that total asymmetry is not grouped by age or by sex. This is in contrast to studies (12, 13) that have shown that the asymmetry of the human face differs between the sexes, its size is age-dependent and is greatest earlier in life. These findings are in agreement with the Rotterdam study (14) in which total and fluctuating asymmetry are significantly related to age and sex. The reasons for these contradictory findings, and taking into account that the asymmetry decreases with age, may lie in the different age structure of the examined populations, i.e. the older age structure of the island population that was the subject of this study. The results of this study showed that the amount of direct asymmetry is the smallest by several times for the area of the mandible for all subjects, which leads us to the conclusion that the influence of genotype in these subjects on the area of the mandible is the smallest and that all causes of fluctuating asymmetry of the mandible in this population should be considered as environmental impact. As the amount of fluctuating asymmetry of the mandible is the largest for the area of Split, it is inferred that the environmental impact of the city of Split is strong. The amounts of both total and fluctuating asymmetries of the mandible are smaller than that of the cranium, which generally indicates that the environment has less of an impact on asymmetry of the mandible and that the impact on the cranium is much stronger. The total asymmetry of the mandible is statistically significantly lower in the subjects from the Korčula area, and at the same time, the direct asymmetry in this population is the largest, which means that the amount of the total asymmetry is actually directional and that the asymmetry of the mandible in the Korčula area is almost not at all under environmental impact and that the environmental impact on the onset of asymmetry the mandible in subjects from the area of Vis and Split is much higher. As the largest amounts of Directional asymmetry in the entire craniofacial area, area of the cranium and mandible are divided between island populations, we can think in the direction of a stronger genetic influence on the phenotype of both populations, by comparing them with the population of the city of Split. As the Functioning asymmetry of the mandible is the largest for the area of Split, environmental impact is the greatest and this is in complete contrast to the impact on the asymmetry of the mandible in subjects from the area of Korčula. Functioning asymmetry for the entire craniofacial skeleton and the cranium is the largest in subjects from the area of Vis which also shows a strong environmental impact. This environmental impact, which is obviously stronger in the area of Vis and Split in relation to the more isolated population of Korčula, mutually competes and leads to the total asymmetry of the cranium and mandible being the largest in the area of Vis, but is greatest for the entire craniofacial area in subjects from the area of Split. All this leads us to the following conclusions: the environmental impact on the formation of asymmetry is strong and it has made the populations of Split and Vis similar despite their genetic differences. Microclimate, the specificities of environmental factors, i.e. their weak influence and isolation in the area of Korčula made the specificity of the genotype manifest itself as the smallest total asymmetry of the mandible. It is correct to view these results through a prism of age, since it is an older population, when the influence of genetics, although still effective, decreases and the influence of environmental factors increases. As the influence of environmental factors is obviously small on Korčula, all mandibular asymmetry, although statistically significantly small, should be viewed in the light of purely genetic influence. Observing the direction of directional asymmetry in the subjects of this study, larger changes in the orbital and nasal area are visible in all three groups (Korčula, Split, Vis), while changes in the mandible are noticeably small, which is expected compared to numerical data. Furthermore, comparing the difference in the amount of directional asymmetry with aging, we see that there are no significantly different changes in directional asymmetry with regards to age, which is consistent with the fact that the largest changes in directional asymmetry are seen at early stages of development. Changes in the direction of individual points as a consequence of directional asymmetry can be interpreted as genotype responses to age-related changes in the craniofacial skeleton. The greatest changes are shown by the pyriform aperture point, with the aperture rotating counter clockwise, the right one moving laterocaudally, and the left one mediocranially. These changes lead to a distortion of the nasal section to the right. As this one, most previous studies on fluctuating asymmetry is based on a comparison of endogamous, often isolated populations with more exogamous (15, 16, 17). The degree of heterozygosity is associated with lower levels of organizational asymmetry and reflects an increased ability to alleviate genetic and environmental stressors. The higher the degree of heterozygosity in genes encoding proteins or at associated loci (non-random association), the more a better adapted individual is protected from the disorder. In this context, three hypothetical mechanisms are proposed: 1) a “direct effect” effect according to which heterozygosity increases the ability of the multilocus and may result in functional dominance at the locus per se (18, 19). It is also possible that heterozygotes possess enzymes with different catalytic properties and are therefore more biochemically efficient than homozygotes (20). 2) the “local effect hypothesis” suggests heterozygosity as an additional dominance when there is a genetic link (link imbalance) between the neutral marker and the marker under selection (18, 21). And 3) the “general effect” hypothesis states that the advantage of heterozygotes is such that the markers that are detected, which arise from the homozygosity interaction, are loci distributed throughout the genome. A prerequisite for this is that the marker and loci are in the service of identity imbalance which is recorded as the inbreeding coefficient of individuals. A higher inbreeding coefficient means higher marker homozygosity and locus binding ability, while a lower coefficient is associated with heterozygosity and poorer marker-locus link (22,23). As the high degree of inbreeding between subjects in this population was shown in several studies within the Project, and as facial fluctuating asymmetry cannot be related to kinship, it seems that the correlation of heterozygosity and locus ability in this study cannot be explained by these three models and requires additional identification. The same is the case with the Rotterdam study. One possible explanation is Chapman et al. who showed that even if there are correlations in heterozygosity and locus binding ability, they are very small (usually less than 1% of phenotypic variance) (22). Furthermore, the degree of heterozygosity may, on average, be low enough to significantly affect developmental pathways. It is also reasonable to assume that the environment of an island population does not fluctuate enough to make homozygosity unfavourable. The third reason relates to the relative genetic effect on facial asymmetry compared to environmental stressors that accumulate over time. Conclusions The conclusions of this study are: 1. The total asymmetry of the entire craniofacial skeleton, cranium and mandible expressed in does not depend on age, sex or place of residence (Korčula, Split, Vis). The amounts for total asymmetry are generally considerably higher for the area of the mandible than for the area of the cranium and for total asymmetry. 2. There is directional asymmetry of the entire craniofacial skeleton, cranium, and mandible in all three groups of subjects. Vector analysis of directional asymmetry showed larger changes in the orbital and nasal areas, while changes in the mandibular area were noticeably small. There is no statistically significant difference in the amount of directional asymmetry for all three observed areas between the groups of subjects from the area of Korčula, Vis and Split. Subjects from the area of Vis have the highest amount of directional asymmetry for the entire craniofacial skeleton and cranial area. Subjects from the area of Korčula have the highest amount of directional asymmetry of the mandible. The amount of directional asymmetry is the smallest for the mandible for all three groups of subjects. 3. There is fluctuating asymmetry of the entire craniofacial skeleton, cranium and mandible in all three groups of subjects (Korčula, Vis, Split), but there is no statistically significant difference between them as in relation to age or sex. The amounts of fluctuation asymmetry for the area of the cranium are greater than the fluctuating asymmetry of the mandible. Vector analysis of the direction of fluctuating asymmetry in relation to age shows that changes in direction are more pronounced in the group of subjects older than 55, and predominantly in the naso-orbital area. 4. Age-related changes in the shape of the craniofacial skeleton are greater in female subjects than in male subjects. They have a constant upward change, but after 55 years the curve of the change in the shape of the craniofacial skeleton as a function of age in female subjects becomes discontinuous with unpredictable upward and downward variations.