2016 年 124 巻 1 号 p. 63-72
Prognathism in the human facial skeleton is generally divided into two types: alveolar and facial. The degree of alveolar prognathism has conventionally been evaluated using the alveolar profile angle, defined on the basis of the line through the nasospinale and prosthion and the Frankfort horizontal (FH). Previous examinations of Japanese crania from the protohistoric Kofun period (c. 300–700 AD) to the modern period have established that the medieval (c. 1200–1600 AD) Japanese possessed the strongest alveolar prognathism. However, the nasospinale is an ambiguous landmark that is difficult to determine. Moreover, it has been suggested that the alveolar profile angle itself is problematic for quantifying alveolar prognathism. Here, we devised a new method for evaluating each type of prognathism independently and investigated temporal changes in the Japanese population using three-dimensional (3D) coordinate landmark data collected with a 3D contact digitizer. Facial prognathism was quantified using an obtuse angle, termed the midfacial protrusion angle, created by a line that runs through the nasion and subspinale with the FH. Alveolar prognathism was evaluated using a reflex angle, named the alveolar protrusion angle, formed between the nasion–subspinale and subspinale–prosthion lines. An analysis of 66 male crania derived from the northern Kyushu and Yamaguchi (westernmost part of Honshu) region showed no significant change of the alveolar protrusion angle over time from the Kofun period to the modern period. In contrast, the midfacial protrusion angle increased from the Kofun period to the medieval period and then decreased in the modern period. These results suggest that the strongest facial prognathism was exhibited in the medieval period. Previous results, which suggested that alveolar prognathism was strong in the medieval Japanese, may have been significantly affected by facial prognathism.
Prognathism is usually categorized as either alveolar prognathism or facial prognathism. Alveolar prognathism, in which the alveolar arches protrude beyond the jaw bases or a basal bone of the jaw structure, is distinct from facial prognathism, which refers to a prominence of the face in relation to the brain case, regardless of the protrusion of the alveolar arches (Björk, 1951).
In Japan, the degree of alveolar prognathism has been examined in order to consider the population history of the Japanese people. Therefore, a considerable body of data has been collected from craniometric studies on samples from different periods. A pioneering study by Suzuki (1969) revealed temporal changes in the cranial morphology of people in the Kanto region (eastern Japan). The results indicated that alveolar prognathism became more prominent from the neolithic Jomon period to the medieval period (c. 1200–1600 AD) and less prominent from the medieval period to the modern period. Subsequently, a similar tendency has been shown with medieval samples newly unearthed in the Kanto region and in western Japan (e.g. Nakahashi and Nagai, 1985; Sakuma, 1986; Nakahashi, 1993; Matsushita, 2002; Nagaoka et al., 2006). Thus, strong alveolar prognathism is generally accepted as one of the common characteristics of medieval people in mainland Japan.
The degree of alveolar prognathism has traditionally been evaluated by the size of the alveolar profile angle (Martin, 1928), which is defined on the basis of a line passing through the nasospinale and prosthion and the Frankfort horizontal (FH). However, Kaifu (1999) pointed out that this angle is questionable as a parameter of alveolar prognathism: because the FH is adopted as a reference line the resulting data are influenced by various structural components of the cranium. As an alternative method, he employed an upper alveolar line instead of the FH. The upper alveolar line was defined by two points on the border between the lateral and inferior surfaces of the alveolar process at the positions of C/P1 and M1/M2. His analysis indicated that alveolar prognathism has monotonously increased over time up to the modern period. Therefore, Kaifu (1999) described that although the alveolar profile angle in the medieval period was unique, this fact does not necessarily represent the strength of the alveolar prognathism of this population. Thus, the interpretation of alveolar prognathism varies according to the particular reference lines selected.
The two conflicting results described above did not demonstrate how cranial structural components affected the alveolar profile angle. For example, facial prognathism might affect the assessment of alveolar prognathism. Furthermore, both types of prognathism might have been confused. Therefore, temporal changes in alveolar prognathism should be reinvestigated with a new strategy that enables both types of prognathism to be examined independently. In this article, we applied landmark-based three-dimensional (3D) measurements to examine this issue, as this method is suitable for measuring various craniometric angles, including the alveolar profile angle (e.g. O’Higgins and Jones, 1998; O’Higgins, 2000; González-José et al., 2008; Bigoni et al., 2010). Landmark-based studies of cranial samples from the Japanese archipelago have recently increased in popularity (e.g. Makishima and Ogihara, 2009; Fukumoto and Kondo, 2010; Fukase et al., 2012). Compared to the conventional osteometric method used for evaluating alveolar prognathism, landmark-based 3D morphometry can visualize variations in the alveolar region in relation to the whole facial architecture more effectively. Factors related to morphological variations can also estimated from the data.
The magnitude of the genetic and environmental effects of cranial phenotypic variations currently constitutes an important focus of anthropological study (Guglielmino-Matessi et al., 1979; Varela and Cocilovo, 1999; Jantz and Meadows Jantz, 2000; González-José et al., 2004; Harvati and Weaver, 2006). In particular, genetic factors (e.g. gene flow by migration) are firmly related to either temporal changes or variations in the craniofacial morphology of populations (Lahr, 1996; Sparks and Jantz, 2002; Relethford, 2004; Carson, 2006; Manica et al., 2007; Martínez-Abadías et al., 2009). The dual-structure model proposed by Hanihara (1991) explained that people who immigrated to western Japan from China and the Korean Peninsula after the Yayoi period, and the descendants of these individuals (immigrant Yayoi), have gradually spread and have become mixed with the native Jomon population. He also suggested that the immigrant Yayoi have had a smaller impact on the Kanto region (eastern Japan) than on western Japan. In fact, protohistoric Kofun and medieval samples from the Kanto region have shown complex genetic effects. Nonmetric as well as metric studies on crania or teeth have shown the overwhelming effects of the immigrant Yayoi on the genetics of this population (Yamaguchi, 1987; Dodo and Ishida, 1990, 1992; Ishida, 1992; Matsumura, 1994, 1995). Meanwhile, remnants of Jomon traits are also often observed in the Kofun and medieval periods (Suzuki, 1969; Hanihara, 1987; Mizoguchi, 1988; Brace et al., 1989; Kaifu, 1997). On the other hand, the northern Kyushu and Yamaguchi region, located in western Japan, are in an area that was affected early and heavily by the population who immigrated after the Yayoi period. In fact, from the Kofun to the modern periods, a certain amount of continuity in morphological traits has been observed in this region (Nakahashi, 1993). Therefore, cranial samples from the northern Kyushu and Yamaguchi region seem to be more suitable than are those from the Kanto region for investigating temporal changes without being influenced by population genetic factors.
In this study, using landmark coordinates of the facial skeleton combined with two newly designed craniometric angles, we aimed to re-evaluate temporal changes in the degree of alveolar prognathism from the Kofun period to the modern period with samples from the northern Kyushu and Yamaguchi region. Functional aspects and the effect of facial prognathism on the interpretation of alveolar prognathism are also discussed.
For this study, 66 adult male crania from the northern Kyushu and Yamaguchi region (western Japan) were utilized (Figure 1). All samples are stored in the Kyushu University Museum. The sample names, number of individuals, period, and provenance of each cranial series are shown in Table 1. The Kofun period is equivalent to the protohistoric period in Japan. The Kofun sample consisted of crania excavated from various Kofun graves built between the latter half of the third century and the first half of the seventh century in Fukuoka and Saga Prefectures (Department of Anatomy, Faculty of Medicine, Kyushu University, 1988). Two of 19 specimens are derived from the Takenami site, and the other 17 specimens are derived from Biwanokuma cists, Ishinami tumuli, Uratani tumuli, a Iyama tumulus, Gionyama tumuli, a Hotarugaoka tunnel-tomb, a Nagino tumulus, Miyanomoto tumuli, a Hushuyama tunnel-tomb, a Goya site, Yame cists, a Yashikitayama tunnel-tomb, Tebika tumuli, an Imamachi tumulus, a Roji tumulus, a Maruyama tumulus, and a Taniguchi tumulus. Crania excavated from the Yoshimohama site in Yamaguchi Prefecture were used as the medieval sample (Nakahashi and Nagai, 1985). The Edo period corresponds to the early modern period in Japan. Materials from the Edo period consisted of human skeletal remains from the Tenpukuji site in Fukuoka Prefecture (Nakahashi, 1987). The modern Japanese sample consisted of cadaveric specimens collected by the Department of Anatomy at Kyushu University. Inhabitants from northern Kyushu (Fukuoka, Saga, and Nagasaki Prefectures) were selected. Hereinafter, the terms Kofun, Medieval, Edo, and Modern are used in this paper to represent sample names from the Kofun to modern periods.
A map showing the distribution of samples used in this study. The gray area represents the northern Kyushu region and the westernmost part of Honshu. Yoshimohama and Tenpukuji, which are indicated by stars, are sites where samples were obtained from the medieval and Edo periods, respectively.
| Sample name | n | Period | Provenance |
|---|---|---|---|
| Kofun (protohistoric) | 19 | c. 300–700 AD | Various archaeological sites in Fukuoka and Saga Prefectures |
| Medieval | 12 | c. 1200–1600 AD | Yoshimohama site in Yamaguchi Prefecture |
| Edo (early modern) | 14 | c. 1600–1900 AD | Tenpukuji site in Fukuoka Prefecture |
| Modern | 21 | c. 1900–1950 AD | Cadaveric specimens from Fukuoka, Saga, and Nagasaki Prefectures |
Specimens whose alveolar process had been markedly reduced due to antemortem tooth loss were not used. The sex and age of the samples from archaeological sites were judged according to the estimation methods proposed by Ubelaker (1999).
A 3D contact digitizer (Microscribe G2X, Immersion Corp., San Jose, CA) was used to obtain 3D coordinates. According to Stephen et al. (2015), the standard deviation values for the digitized x, y, z coordinates did not exceed 0.23 mm (16 locations with different arm configurations: each 100 trials) and neither workspace area differences nor arm orientation differences affected measurement accuracy. All operations were performed by K.O.
To assess the measurement error of the apparatus, distances between bilateral frontmalare orbitale and zygomaxillare anterior (zma) were measured by the contact digitizer and compared with published craniometric data measured by osteometric calipers (Department of Anatomy, Faculty of Medicine, Kyushu University, 1988) by means of a t-test. The distance between bilateral frontmalare orbitale was larger than the published record only by 0.05 mm (n = 27) with no significant difference although our measurement of zma–zma distance was significantly smaller than the previous measurement by 0.81 mm (n = 23). We considered this difference as an interobserver (definitional) measuring error rather than a bias caused by the instrument because significant differences appeared for only one of these two similar facial measurements.
Four craniofacial landmarks (nasion, subspinale, prosthion, and zma) were selected and used for analysis (Figure 2). All landmarks, except the prosthion, followed definitions outlined in the standard craniometric method (Martin and Saller, 1957; Martin and Knussmann, 1988). For the prosthion, a definition by Howells (1973) was adopted according to indications by Dodo (2000).
Craniofacial landmarks used for the analysis. n = nasion; ss = subspinale; pr = prosthion; zma = zygomaxillare anterior.
We did not consider cranial asymmetry as an important factor for our objective and utilized a symmetricalizing treatment of the landmark coordinates (Ogihara et al., 2006). That is, landmarks defined to be located within the midsagittal plane were replaced so that they were all on the same approximate plane as determined by least-squares fitting, and landmarks defined to be bilateral were treated as being symmetrical with respect to the midsagittal plane. According to Makishima and Ogihara (2009), the symmetricalization of landmark coordinate data well reproduces the results of multivariate analysis based on scalar quantities obtained by conventional cranial measurements, even though subjects are slightly asymmetrical.
We chose cranial samples with comparatively sufficient preservation. However, if either side of the bilaterally defined landmarks was not measurable, the coordinates from the intact side were used to represent the landmark on the broken side by symmetricalizing. Furthermore, in order to compare the facial profiles of each time period, the landmark coordinates of all samples were aligned so that they were oriented with the FH and positioned with the nasion. All of these processes were performed using Microsoft Excel 2010 (Microsoft Corp., Redmond, WA).
Setting of new angles to evaluate facial prognathism and alveolar prognathismWe newly introduced two craniometric angles in the present study (Figure 3). The midfacial protrusion angle (MPA), which is intended to evaluate facial prognathism, is an obtuse angle created by a line that runs through the nasion and subspinale with the FH. The alveolar protrusion angle (APA), for evaluating alveolar prognathism, is defined by a reflex angle between the nasion–subspinale and subspinale– prosthion lines.
The two protrusion angles and length used in this analysis were the midfacial protrusion angle (MPA); alveolar protrusion angle (APA); and minimum distance between the nasion and the zygomaxillare anterior (zma) –zma lines projected on the Frankfort horizontal (DNZ). n = nasion; ss = subspinale; pr = prosthion; FH = Frankfort horizontal.
According to previous studies targeting facial flatness, the distance between the bilateral zma has been designated as the zygomaxillary chord, and the zygomaxillary index has been calculated as the percentage of subspinale subtense to the zygomaxillary chord (e.g. Yamaguchi, 1973, 1980; Dodo, 1986; Hanihara et al., 1999; Dodo and Kawakubo, 2002; Kawakubo, 2007; Kawakubo et al., 2009). Since Coon (1971) noted that the zygomaxillary index is affected by the prognathism, and Kawakubo (2007) regarded facial breadth as an important factor in facial flatness measurements, we investigated the relationship between the anteroposterior position of the maxilla on the basis of nasion and facial breadth. That is, using 3D coordinates for the subspinale and bilateral zma, mean values of the zygomaxillary chord and the minimum distance between the nasion and the zma-zma lines projected on the FH (here termed the DNZ) were calculated (Figure 3).
Statistical analysesFor the values of MPA, APA, zygomaxillary chord, and DNZ, the null hypothesis for distributional normality was not rejected at the 5% level by the Kolmogorov–Smirnov test. Moreover, Levene’s test for homogeneity of variances showed no statistical significance at the 5% level for any of these measurements within the sample. Thus, we used the t-test to compare the measurements, with significance set at 5%. SPSS 16.0 J for Windows (SPSS Inc., Chicago, IL) was used for the statistical tests, and Microsoft Excel 2010 was used for the other calculations.
The MPA was significantly increased from Kofun to Medieval (P < 0.05), but was significantly decreased from Medieval to Edo (P < 0.05) and from Medieval to Modern (P < 0.01), with the largest value occurring in Medieval (Table 2). On the other hand, although the APA tended to decrease from Kofun to Medieval and to increase from Medieval to Modern, there were no statistically significant differences (Table 2).
| Kofun | Medieval | Edo | Modern | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | n | Mean | SD | n | Mean | SD | |
| MPA (°) | 19 | 94.9 * | 2.68 | 12 | 97.4 | 3.55 | 14 | 94.5 * | 2.66 | 21 | 93.1 ** | 3.23 |
| APA (°) | 19 | 190.4 | 4.58 | 12 | 187.7 | 3.25 | 14 | 188.7 | 5.90 | 21 | 191.1 | 6.44 |
MPA = midfacial protrusion angle; APA = alveolar protrusion angle.
In order to render the temporal changes in the MPA and APA intelligible, the coordinates of the subspinale and prosthion (two-dimensional coordinates of projections and height) were plotted in lateral view relative to the FH and nasion, and then the nasion, subspinale, and prosthion were connected by lines (Figure 4). Compared to the subspinale and prosthion in Kofun, those in Medieval were located more anteriorly (Figure 4a). For Edo and Modern, as compared with Medieval, the subspinale and prosthion were plotted more posteriorly (Figure 4b).
Temporal changes in the positions of the subspinale (ss) and prosthion (pr) in the lateral view of the midsagittal plane from the Kofun to the modern periods. The three-dimensional coordinates of the ss and pr are superimposed in the Frankfort horizontal based on the nasion (n).
Table 3 shows the mean values of the zygomaxillary chord and the DNZ. The former represent facial breadth, while the latter is regarded as a parameter of the anteroposterior position of the maxilla. The DNZ was significantly different between Medieval and Modern (P < 0.001). The zygomaxillary chord was significantly shorter (P < 0.01) in Medieval than in Kofun (Table 3). However, no significant difference was observed between Medieval and Edo or between Medieval and Modern. To represent the changes of the position and breadth of the maxilla schematically, the subspinale and bilateral zma coordinates (two-dimensional coordinates of projections and width) were superimposed in the FH and connected by lines (Figure 5). The nasion was used as the reference for the superimposition. Since the symmetricalizing treatment was performed (as noted in the Methods), only the left side was shown. Compared with that for Kofun, the triangle for Medieval was located more anteriorly (Figure 5a). Compared with that for Medieval, the triangle for Edo was located posteriorly, and the triangle for Modern was located more posteriorly than that for Edo (Figure 5b).
| Kofun | Medieval | Edo | Modern | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | n | Mean | SD | n | Mean | SD | |
| Zygomaxillary chord (mm) | 19 | 104.5 ** | 4.19 | 12 | 99.0 | 6.21 | 14 | 100.9 | 4.52 | 21 | 99.0 | 6.66 |
| DNZ (mm) | 19 | 15.9 | 2.73 | 12 | 14.0 | 3.56 | 14 | 16.6 | 3.67 | 21 | 18.2 *** | 2.73 |
DNZ = minimum distance between nasion (n) and the zygomaxillare anterior (zma)–zma lines projected on to the Frankfort horizontal.
Temporal changes in the shape of the maxilla from the Kofun to the modern periods. The three-dimensional coordinates of the subspinale (ss) and bilateral zygomaxillare anterior (zma) are regarded as being representative of maxillary shape, and they are superimposed on the Frankfort horizontal based on the nasion (n). Only one side of the isosceles triangle is shown.
This study re-evaluated the temporal changes in prognathism in western Japan by using two newly introduced angles, MPA and APA, to interpret facial prognathism and alveolar prognathism separately on the basis of landmark coordinate data.
To evaluate alveolar prognathism, previous studies used the nasospinale to measure the alveolar profile angle. However, this landmark is difficult to determine because it is largely masked by the nasal spine (Ohsako, 2000). This leads to definitional or interobserver error of the angle, which is one of the major problems in craniometry (Ohsako et al., 2000). Therefore, we used the subspinale for determining the MPA and APA in the present study. The landmark subspinale represents the anterior surface of the maxillary alveolar process in the median sagittal plane and has been used in facial flatness measurements (e.g. Yamaguchi, 1973, 1980; Dodo, 1986; Hanihara et al., 1999; Dodo and Kawakubo, 2002; Kawakubo, 2007; Kawakubo et al., 2009). Hence, it is considered to be highly reliable as a landmark. When we calculated correlation coefficients between the MPA and the equivalent angle substituted by the nasospinale, a high correlation (r = 0.93) was observed (raw data not shown).
The changes in MPA (Table 2) and the plots of the subspinale in the lateral view (Figure 4) indicated the temporal variations in the anteroposterior positions of the maxilla. From Kofun to Medieval, the MPA significantly increased, while it significantly decreased from Medieval to Modern. Since the MPA represents the extent of protrusion of the whole maxilla, it conforms to facial prognathism, which is a prominence of the face on the base of the neurocranium, regardless of the protrusion of alveolar arches (Björk, 1951). Therefore, our findings suggest that facial prognathism was strengthened from the Kofun to the medieval periods and weakened from the medieval to the modern periods. On the other hand, since there were no significant differences in APA, the temporal change in the degree of inclination of the maxillary alveolar process was not discernible. Thus, we conclude that the degree of alveolar prognathism was relatively constant over time in the sample from western Japan.
Since previous studies based on the alveolar profile angle indicated a lowest value of the angle in the medieval period, those studies concluded that individuals in the medieval period had marked alveolar prognathism (e.g. Suzuki, 1969; Nakahashi and Nagai, 1985). However, in the present study, no evidence was obtained regarding temporal changes in alveolar prognathism. Instead the largest MPA value was demonstrated in the medieval period. These results suggest that the strength of alveolar prognathism evaluated using the alveolar profile angle has been greatly affected by facial prognathism. Figure 6 shows two examples of a facial skeleton with the same alveolar profile angle. As can be observed by the angles between the nasion–nasospinale line and the FH, the right cranium has obviously stronger facial prognathism compared to that in the left cranium. Thus, the same alveolar profile angle will not always provide the same facial profile. Information on the relative anteroposterior position of the maxillary alveolar region seems to be indispensable in the assessment of alveolar prognathism.
Two crania with different facial inclinations have the same alveolar profile angle. n = nasion; ns = nasospinale; pr = prosthion; FH = Frankfort horizontal.
Our findings on the temporal changes of alveolar prognathism were not consistent with Kaifu (1999), in which the upper alveolar line was used as a reference. That is, he showed a monotonous increase in the prognathism from the Kofun period to the modern period, whereas no significant change was detected in the present study. Since Kaifu used cranial samples from the Kanto region, regional differences might have contributed to the inconsistency. To clarify these factors, further investigation of the MPA and APA of the samples from eastern Japan is required. Another possibility for the discrepancy is due to the different reference lines. Although the FH is a useful reference line often employed in craniometric researches, including this study, a few anthropologists have raised doubts about its usage. Suwa (1980) demonstrated a variety of anteroposterior inclinations of the FH due to differences in the natural head position among Japanese populations and American and European Caucasians. He suggested that the natural head position may be affected by non-biological factors, such as cultural customs or occupational habits. Kaifu (1999) also pointed out that the FH is influenced by various structural components of the cranium. Indeed, the FH of different time periods may be unsuitable as an absolute reference line, even in the same population. The results of the present study suggest that the nature of prognathism should be reconsidered.
The mechanisms and factors related to temporal changesLifestyle differences are thought to have a significant influence on cranial morphology (e.g. Lieberman, 2011; von Cramon-Taubadel, 2011; Noback and Harvati, 2015). Mastication loading and accompanying changes in the hardness of a diet are especially important environmental factors that may have affected the facial development of individuals (Corruccini et al., 1985; Larsen, 1997; González-José et al., 2005; Lieberman, 2008). Some experimental studies in nonhuman mammals have revealed that a soft diet delays growth of the lower part of the face (Corruccini and Beecher, 1982; Beecher et al., 1983; Lieberman et al., 2004).
In Japan, some reports have suggested a relationship between decreased mastication and variations in cranial morphology. Kaifu (1997) investigated temporal morphological changes in the mandible, primarily those from the Kanto region. He found that although masticatory muscle attachment was well developed until the medieval period, notable reduction occurred from the Edo period. This change was explained by weakened mechanical stimulation against the mandible along with changes in diet. Similar trends in the mandibular measurement values (i.e. a decrease in the least anteroposterior ramus length and an increasing mandibular angle) have been observed in the northern Kyushu and Yamaguchi region (Nakahashi and Nagai, 1985). In addition, Kawakubo (2007) suggested that the reduced flatness of the zygomaxillary region in eastern Japanese sample is due to a retreating zygomatic bone induced by weakening of the masticatory system. These previous studies implicate changes in masticatory muscle attachment regions and contact areas. The present study revealed that the anterior surface of the maxillary process, as represented by the subspinale and prosthion, retreated gradually from medieval to modern times (Figure 4b), suggesting that facial prognathism became weaker. In addition, the relative position of the zma gradually retreated from the medieval to modern periods (Table 3, Figure 5b). These changes are illustrated in Figure 7. The recessing region, which includes the zma, is the masseter muscle attachment region. The posterior part of the zygomatic bone constitutes an anterolateral wall of temporal fossa where the temporal muscle is tightly packed. Therefore, it is reasonable to conclude that decreased chewing activity led to the reduction in the inferior part of the maxillary zygomatic process and the zygomatic bone. A recent study has also suggested significant correlations between the shape of temporal fossa and diet (Noback and Harvati, 2015).
Schematic diagram of the temporal changes in the maxilla and zygomatic bones from the medieval to the modern periods.
The maxillary alveolar base has been shown to become reduced toward the modern period in the Kanto region, while the axial inclination of the maxillary incisors began to increase toward the modern period (Kamegai et al., 1982). Since this tendency is expected to show decreasing MPA and increasing APA, we calculated correlation coefficients between the MPA and APA within each group in order to examine the relevance of the degree of facial prognathism and alveolar prognathism. A significant negative correlation (r = −0.46, P = 0.036) was only shown in Modern (data not shown). This result suggests that modern Japanese individuals in the northern Kyushu and Yamaguchi region share a similar facial profile (i.e. weak facial prognathism) with those in the Kanto region.
In this study, we found that facial prognathism was strengthened from the Kofun to the medieval period, while at the same time, the zygomaxillary chord was significantly reduced (Table 3, Figure 5a). The changes that occurred during this period are illustrated in Figure 8. However, the cause of this phenomenon is presently unknown.
Schematic diagram of the temporal changes in the maxilla from the Kofun to the medieval periods.
Dolicocephaly has been assumed to be one of the peculiar characteristics of medieval people, along with strong alveolar prognathism (e.g. Suzuki, 1969; Nakahashi and Nagai, 1985). However, correlation coefficients between the cranial index and the inclination angle of the upper central incisor were not significant for any period (Kaifu, 1999). As revealed by the present study, alveolar prognathism is not a distinctive feature in the medieval period, so it seems to be irrelevant to the cranial index. Instead, strengthened facial prognathism may affect the index in the medieval period. This is a subject for future examination.
Landmark-based studies enable visualization of craniofacial morphology and permit more detailed analysis of morphological variations. Using this method, we re-evaluated cranial morphology and provided additional evidence to support an association between regression of the maxilla and morphological variations in the alveolar region from the medieval period to the modern period. On the other hand, the finding of increased facial prognathism from the Kofun period to the medieval period suggested that complicated mechanisms underlie the variations in maxillofacial morphology.
We wish to thank Dr. Y. Dodo (Tohoku University School of Medicine), who greatly contributed to the preparation of the manuscript. We also thank Drs. S. Iwanaga, K. Funahashi, and S. Yonemoto (Kyushu University Museum) for permission to examine cranial samples under their care.
