2015 Volume 123 Issue 2 Pages 107-119
We investigated aging-related changes in the skulls of Japanese macaques (Macaca fuscata). A total of 145 (70 males, 75 females) skulls from macaques aged 7.0–26.9 years (males) and 7.0–30.7 years (females) were measured for 22 craniometric items. Some skull dimensions increased from young adulthood (7.0 years) to the peak at 13.3–19.0 years in males and at 19.7–22.6 years in females. Some dimensions remained at their peak value right through into very old age whereas others continued increasing during this stage of life. Continued increase of cranial size in adulthood has been also observed in humans, but the magnitude of change was greater in macaques. Facial and mandibular dimensions showed larger and more significant increases than neurocranial dimensions in macaques, as in humans, including facial height, bizygomatic breadth, mandibular body height, and ramus breadth in both sexes. Intertemporal distance and biorbital breadths after 16.0 years of age decreased significantly in males, and cranial and posterior basicranial lengths increased only in males. We suggest that these craniometric changes are associated with the development of the insertion area onto which muscles attach (by accumulation of physical stress). The face and mandible are greatly influenced by tooth loss and/or dental disorders, both of which are evident in humans. In the present study large changes were also found in skulls that had lost several teeth.
Aging involves a decline in the ability to adapt to environmental stress (Bogin, 2001). Senescence is defined by changes that occur primarily in the postreproductive period in humans (around 50 years of age or more; Fedigan and Pavelka, 2001), which overall reduce the functional capacities of the organism and tissues in the decades after menopause (Borkan et al., 1982; Bogin, 2001). Although nonhuman primates maintain their reproductive activity almost until their death (Pavelka and Fedigan, 1999), physical aging advances from young adulthood throughout adult life (Hamada and Yamamoto, 2010).
Elderly adults can be characterized by the extent of aging in the musculoskeletal system, such as by a decrease in stature or kyphosis in the vertebral column (Hamada and Yamamoto, 2010). Bone is an organ that shows deterioration with age in both humans and nonhuman primates (Mazess, 1982; Black et al., 2001). Aging-related changes in human skulls are receiving increased attention and have been studied, for example in forensic research (Albert et al., 2007) and esthetic surgery (Bartlett et al., 1992). Postcranial skeletal changes in aging are obvious in both humans and nonhuman primates. Continuous expansion in the human postcranial skeleton (transverse diameters) has been documented for various bones (Smith and Walker, 1964; Garn et al., 1967; Pfeiffer, 1980; Ruff and Hayes, 1982). It is also seen in nonhuman primates, as shown by macaques (Bowden et al., 1979; Kimura, 1994) and chimpanzees (Morbeck et al., 2002).
There is controversy with respect to changes of the skull with age in adult humans. With sexual maturation, cranial growth in humans has been considered to cease or be insignificant (Tallgren, 1974). However, it has been recognized that cranial growth does not end at the adolescent stage in humans, but continues at a slow rate throughout adulthood (Hrdlička, 1936; Israel, 1977; Ruff, 1980; Forsberg et al., 1991), and it has been suggested that the cranium decreases in size in very old people (Hrdlička, 1936; Israel, 1973b). A small increase in the thickness of the skull during adult life has been suggested (Adeloye et al., 1975), but no statistically significant increase was found (Lynnerup, 2001). However, these conclusions are subject to sampling bias (small sample sizes, samples limited to certain historic anatomical collections or homogeneous populations), different age ranges, the confounding effects of pathology, methodologies in data collection, and the statistical methods adopted (Lynnerup, 2001; Albert et al., 2007).
The phylogenetic closeness and marked biological similarities to humans make macaques and baboons a good model of human skeletal aging (Walker, 1995; Colman and Binkley, 2002). The similarities include posture (Black et al., 2001), reproductive endocrinology (Pavelka and Fedigan, 1999), bone loss following estrogen depletion (Colman and Binkley, 2002), and histomorphometry (Jerome et al., 1994). In contrast, mice lack the osteonal bone remodeling (Jilka, 2013) and the murine cortex is composed mostly of the circumferential lamellae (Enlow and Brown, 1958). The life cycle of macaques is the same as that of humans (Hamada and Yamamoto, 2010), including the elderly stage. In general, macaques age at a rate of 2.5–3.5 times that of humans (Colman and Anderson, 2011; Duncan et al., 2011). Although a good model of human aging, macaques have not been well studied in terms of the effect of aging on the skull.
Little is known about aging-related changes in the skulls of nonhuman primates. Exceptionally, some studies noticed a decrease in the intertemporal distance in male Japanese macaques in wild populations (Mouri et al., 2004), an enlargement of the medullary cavity of bones that border the facial suture in pig-tailed macaques (Kokich et al., 1979), and endocranial suture closure in rhesus macaques (Wang et al., 2006).
In the present study we investigated aging-related changes in skulls of Japanese macaques (Macaca fuscata), and compared them with those in human skulls. The objectives of this study were to determine whether skulls of nonhuman primates increase or decrease in size and/or change shape with age, whether changes in skull with age differ between sexes, whether magnitude of size or shape change in macaques is comparable to those in humans, and whether facial dimensions are influenced by tooth loss in macaques.
A total of 145 skulls from adult Japanese macaques (M. fuscata) of known age, comprising 70 males and 75 females (Table 1), were used. In general, the adult stage is considered to start at the age when body growth stops, represented by such whole-body dimensions as stature (or its proxy, trunk length in nonhuman primates) (Hamada and Yamamoto, 2010). However, the age at which body growth stops has not been determined for Japanese macaques because body weight fluctuates within a given individual and shows great variation among individuals after adolescence (Hamada and Yamamoto, 2010); and the growth in linear dimension is too slow to identify the exact age at which growth ceases (Hamada and Yamamoto, 2010). Thus 7 years of age, when the permanent teeth have fully erupted (Iwamoto et al., 1987), has been regarded as the age demarcating the adult stage.
All the subject macaques were being fed, at least in the period before their death, at the Primate Research Institute of Kyoto University. All skulls were stored at the institute. No macaques suffered from serious disease or received serious experimental treatment. The ages of the macaques are expressed in centesimal age. Although the dates of birth of 11 individuals were not known, their years of birth and dates of death were recorded. For the ages of these macaques we used the date June 15 for their birth, as Japanese macaques have a definite birth season, from March to August (Nozaki and Oshima, 1987). The years of birth of the individuals ranged from 1966 to 2001, and the years of death ranged from 1976 to 2009. Ages of male and female subjects ranged from 7.0 to 26.9 years and from 7.0 to 30.7 years, respectively (Table 1).
Twenty linear dimensions and two angles of skull and mandible were measured on each specimen (Table 2, Figure 1) (Mouri, 1994; Mouri et al., 2004): 7 neurocranial dimensions (cranial length, cranial base length, cranial height, cranial breadth at the nuchal line, posterior basicranial length, postorbital breadth, and intertemporal distance), 10 facial dimensions (facial length, facial height, bizygomatic breadth, zygomatic height at temporozygomatic suture, orbital breadth, orbital height, biorbital breadth, nasal length, maxilloalveolar breadth, and maxillary angle), and 5 mandibular dimensions (ramus breadth, mandibular body height, mandibular body thickness, bicondylar breadth, and mandibular angle). Measurements were made in triplicate, and a median value was used for analysis. All measurements of the skull were made to the nearest 0.01 mm using a digital sliding caliper (Mitutoyo Corp., Japan). Maxillary angle was calculated using the cosine rule of a triangle with three linear sizes of sides. Since the three linear sizes (ZO–ZR, ZR–EAM, and EAM–ZO; Table 2) of the triangle are not on the sagittal plane, the maxillary angle is also oblique to the sagittal plane. The mandibular angle was obtained photogrammetrically (Figure 1e). The photographic setup consisted of a tripod (Dolly pod DP-3D, Velbon Tripod Co., Ltd.) holding a digital camera (FinePix HS20EXR, Fujifilm Co., Ltd.). The tripod was adjusted to maintain the optical axis of the lens horizontal and at the height of the subject. The lateral view of the left side of the mandible was positioned approximately 3 m from the objective lens of the camera and was adjusted so that the sagittal plane of the mandible was perpendicular to the optical axis of the lens. After raising the optical axis of the lens to the alveolar process at the middle of the second mandibular molar, photographs were taken. The mandibular angle was measured from images with Image J 1.46 (Ferreira and Rasband, 2013). The items measured are showed in Table 2 and Figure 1. The status of dentition and of the alveolar bone were recorded for each specimen.
Schematic illustration of most of the measurements taken on the skull (see Table 2 for abbreviations): (a) skull in ventral view; (b) skull in anterior view; (c) skull in lateral view; (d) skull in dorsal view; (e) mandible in lateral view; (f) mandible in transverse view.
Teeth in subjects were almost all retained until death, but 9 subjects had lost several teeth (Table 1): 4 males (ages from 23.7 to 26.0 years) and 5 females (ages from 25.2 to 30.0 years) had lost 8–11 teeth and 6–18 teeth, respectively. Tooth loss is an important factor of change in facial dimensions (Goldstein, 1936; Bartlett et al., 1992). Therefore, subjects that had lost many teeth were analyzed separately (4 males and 5 females) and then were compared with those subjects of comparable ages with all teeth retained, which comprised of 6 males (ages from 23.3–26.9 years) and 6 females (ages from 25.3–30.7 years). The effect of tooth loss on facial dimensions could thus be evaluated.
Statistical analyses were performed using functions of Excel (Microsoft Co. Ltd.). To describe age change patterns, Loess smoothing, locally weighted scatterplot smoothing (Past version 2.17 software; Hammer et al., 2001), was used. We compared craniometric dimensions and angles at start, peak size, and end of the adult stage. The age at peak size was derived from the Loess smoothing function. For measurements that did not show an increase–peak–decrease age-change pattern (continuous increase or decrease), sizes at the start and end were compared. The sizes at start (young adult, 7.0 years) and peak in each measurement were represented by medians of data of 7.0–9.0 years in both sexes, and those of peak age ±1 year in both sexes, respectively. Because of the small sample size of data from very old subjects, end size was represented by the median of the data of the five oldest subjects, i.e. 24.0–26.9 years for males and 25.5–30.7 years for females. Percent differences (median) between sizes at start, peak, and end were obtained and differences of angle measurements were calculated by subtracting the two values to create an absolute value. Differences between start, peak, and end sizes (medians) and between sexes were tested by Mann–Whitney U-test using Past version 2.17 software (Hammer et al., 2001). Differences between lost-teeth and retained-teeth subjects and between sexes were also tested by Mann–Whitney U-test.
As cranial variables are not independent of one another, principal component analysis (PCA) was applied to both males and females using Past (Hammer et al., 2001). The relationship between scores of principal components (PCs) and ages was determined by the Loess smoothing function.
Some dimensions and angles showed aging-related patterns. Skull measurements at start (young adulthood), peak, and end (very old age) are presented in Table 3, Table 4, and Table 5, and Figure 2. Changes with age were significant (P < 0.05) in 14 and 7 cranial dimensions in males (Table 3, Table 5) and females (Table 4, Table 5), respectively.
Scatter plots showing age change in all skulls (without teeth loss) with Loess smoothing. Male, ○ and dashed line; female, ● and solid line.
Scatter plots showing age change in all skulls (without teeth loss) with Loess smoothing. Male, ○ and dashed line; female, ● and solid line.
The four neurocranial dimensions, i.e. CL, PBL, CBL, and ITD, changed significantly with age in males but not in females (Table 4). CL, PBL, and CBL in males peaked at 16.1 years with an increase of 4.55% (P < 0.01), at 16.0 years with 5.19% (P < 0.05), and at 14.5 years with 6.25% (P < 0.05), respectively. After the peaks these dimensions did not decrease at all. PBL also showed a significant increase between the start age and the end age in males by 4.66% (P < 0.05). ITD decreased greatly with age in males by 68.89% from the start to the end (P < 0.001, Figure 2). Other dimensions, such as CBN and POB in males, remained stable or changed slightly, but not statistically significantly (Table 3).
Facial cranial dimensions tended to show statistically significant changes with age, but several facial dimensions showed no or slight age changes, including OB, OH, NL, and MXA in both sexes. FH continuously increased through very old age in both sexes with increases of 13.83% (P < 0.001) in males and 12.10% (P < 0.01) in females. FL increased significantly with age in males and peaked at approximately 16.0 years of age with an increase of 13.44% (P < 0.001). After the peak, FL in males decreased slightly. On the other hand, this dimension increased only slightly in females. BZB and BOB also increased significantly with age in both sexes. After the peaks, these dimensions decreased slightly or not at all in both sexes; however, exceptionally BOB in males decreased significantly (P < 0.01). MAB increased significantly with age in both sexes, peaked at 13.3 years of age in males with an increase of 4.24% (P < 0.05), and increased 3.96% from the start (7.0 years) to the end (30.7 years) in females (P < 0.01). ZH increased significantly with age only in males, and reached a peak at 19.0 years of age with an increase of 16.06% (P < 0.05), whereas this dimension changed slightly in females, but not statistically significantly.
Mandibular dimensions in general showed significant changes with age (Table 3, Table 4, Table 5). Exceptionally, however, BCB changed slightly but not significantly. MBH and RB increased significantly with age in both sexes, and then tended to remain stable or decrease not at all. MBT in males increased continuously with age by 9.67% from the start to the end (P < 0.01), whereas this dimension changed in females, but not statistically significantly. MA showed a significant increase with age in both sexes by 8.07° in males (P < 0.05) and 5.50° in females (P < 0.01).
In a PCA using 20 linear cranial measurements and based on a correlation matrix, the first three principal components accounted for 58.98%, 8.18%, and 5.05% of total variation, respectively (Table 6). The first principal component (PC1), representing size change, showed that all dimensions apart from ITD had positive loadings. The PC1 scores showed age-change patterns similar to many dimensions (Figure 3a). Males had higher PC1 scores than females. Age at the peak PC1 score was 16.0 and 20.2 years in males and females, respectively.
Plots of (a) first principal component scores, (b) second principal component scores, and (c) third principal component scores against age with Loess smoothing. Male, ○ and dashed regression line; female, ● and solid regression line.
The second principal component (PC2), describing shape change with age, showed that females had higher scores than males (Figure 3b). OH (0.476), ITD (0.429), POB (0.360), and OB (0.298) were positively loaded, whereas PBL (−0.237), MAB (−0.208), and MBT (−0.250) were negatively loaded (Table 6). The PC2 scores showed a rather rapid decrease from 7.0 to 11.0 years, remained stable from 11.0 to 17.0 years, and then decreased again in males, although remained almost unchanged from 7.0 to 30.7 years in females (Figure 3b).
The third principal component (PC3) represented shape change, with NL (−0.511), OH (−0.372), and FH (−0.366) negatively loaded and CH (0.395) positively loaded (Figure 3c, Table 6). PC3 showed gradual decreases in both males and females (Figure 3c).
Thirteen facial and mandibular dimensions of subjects that had lost several teeth were compared with those subjects of comparable ages with all teeth retained (Table 7). The results of the comparison in this study showed that two dimensions of lost-teeth subjects differed significantly from retained-teeth subjects in both sexes. FH and MBH in males were 16.75% (P < 0.01) and 17.29% (P < 0.01) smaller than the medians for male macaques with all teeth retained, respectively, whereas these measurements in females were 11.16% (P < 0.05) and 15.45% (P < 0.01) smaller than female subjects with all teeth retained, respectively.
Cross-sectional aging studies suffer from secular trends or environmental variability that masks ontogeny (Baer, 1956). Although we studied macaque skulls cross-sectionally, all of them originated from captive monkeys fed under the same conditions during their lifespans. Thus, secular trends or environmental factors may not have had any effect on cranial dimensions in this study.
The results of the present study indicate that some of the cranial dimensions showed aging-related changes, with a pattern increasing from young adulthood (7.0 years) to mid-adulthood (13.3–19.0 years in males and 19.7–22.6 years in females) and unchanged from mid-adulthood to very old age (26.9 years or more) or increasing from young adulthood to very old age. It is probable that the increase from 7.0 years to mid-adulthood or very old age comprises both total physical growth, which ends at 10.0 or 15.0 years of age (Hamada, 1994; Hamada and Yamamoto, 2010) and specific increases in the head and face. Maximal stages of skull size in some measurements, including bizygomatic, biorbital, and ramus breadths and mandibular body height, tended to be attained later in females than in males. This corresponds with the data that trunk length and epiphyseal unions in postcranial skeletons tended to occur later in females than in males in Japanese macaques (Kimura, 1994; Hamada, 2008). Intertemporal distance and biorbital breadth (after 16.0 years) decreased significantly in males. Facial and mandibular dimensions tended to show greater and more significant change with age than neurocranial dimensions, including facial height, bizygomatic breadth, mandibular body height, and ramus breadth in both sexes. Furthermore, as for proportional changes in the face, facial height increased more than bizygomatic breadth in both sexes, making the face relatively longer with age in macaques. In addition, proportions changed, i.e. orbital height to orbital or biorbital breadth.
Continued increase of cranial size in adulthood has been observed in humans (Goldstein, 1936; Garn et al., 1967; Nasjeleti and Kowalski, 1975; Israel, 1973a, 1977; Susanne, 1977; Ruff, 1980; Bartlett et al., 1992). Human skulls attain maximal size at about 70 years of age (Israel, 1973b). In macaques, ages at maximal skull size are estimated (by PCA) to be 16.0 ± 3 and 20.2 ± 3 years in males and females, respectively, i.e. 48.0 and 60.6 year equivalents in human age.
The magnitude of extensive development after completion of tooth eruption in skull dimensions is much greater in macaques than in humans. For example, facial height increases by 3.66% in men from 20.0 to ≥60.0 years (Nasjeleti and Kowalski, 1975), or 1.49% in men and 1.28% in women aged from 18.0 to 42.0 years of age (Formby et al., 1994). In contrast, facial height in male macaques increased by 13.83% from young adulthood to 26.9 years and by 12.10% in females from 7.0 to 30.7 years. Similarly, facial length in male macaques is larger than that of men (Ruff, 1980) and bizygomatic breadth is also larger in both male and female macaques than that of men and women (Susanne, 1977; Bartlett et al., 1992). Neurocranial dimensions, such as cranial length, show smaller contrasts between humans (Israel, 1977) and macaques than facial dimensions.
The magnitude of age change in skull dimensions, which differs between humans and macaques, may be associated with mechanical stress from mastication, which is likely to stimulate an increase in facial dimensions and a change in neurocranial dimensions. Previous studies have demonstrated that hard, tough, and unprocessed diets play an important role in increasing the general sizes of the skull and face (Carlson and Van Gerven, 1977; Sardi et al., 2006), and thus the softer and more processed foods in human diets are considered to reduce masticatory stress in the craniofacial skeleton (Carlson and Van Gerven, 1977; Lieberman et al., 2004; Sardi et al., 2006; Paschetta et al., 2010), resulting in a smaller increase in facial size than in macaques.
Sex differences in aging-related changes in cranial dimensions are not large in humans (Ruff, 1980). However, sex differences in these changes in macaques are rather large. Various degrees of sex difference were found in macaques, large in facial and small in neurocranial dimensions. Human–macaque differences in sex difference and sex differences in macaques may be associated with the stress of mastication (Wang et al., 2006). Larger masticatory forces are applied to the face in male macaques than in females (Dechow and Carlson, 1990).
Expansion of the cranium may be associated with the development of bones as a response to physical stress from masticatory and/or postural muscles. The increase in cranial and posterior basicranial lengths in male macaques may be associated with the development of the nuchal crest or insertion processes (tubercles) on which the nuchal muscles attach. The large decrease in intertemporal distance and postorbital breadth may be the result of development or stress from temporalis (Figure 4).
The differences in temporal line and bone around inion between old males (a) and females (b) in Japanese macaques.
Facial crania and mandibles are greatly influenced by tooth loss and/or dental disorders in humans (Israel, 1973c; Bartlett et al., 1992; Merrot et al., 2005). In the present study subject macaques that had lost many teeth exhibited a reduction of the alveolar bone and also large changes in the face and mandible. These changes were especially true for mandibular body height and facial height compared with those of subjects of comparable ages with all teeth retained (Figure 5, Figure 6). Similar findings have been reported for humans (Israel, 1973c; Bartlett et al., 1992) and for other nonhuman primates; alveolar resorption occurred in captive rhesus macaques (Lapin et al., 1979) and wild chimpanzees (Kilgore, 1989; Morbeck et al., 2002).
Atrophy of alveolar process of maxilla (a) and mandible (b) in some Japanese macaques that have lost teeth; lateral view.
Scatter plots showing how tooth loss causes large decreases in facial and mandibular body height in Japanese macaques. ○ male, ♦ male with tooth loss, ● female, ▵ female with tooth loss.
We wish to thank the staff and researchers who collected skeletons of known age and life history at the Primate Research Institute (PRI) of Kyoto University. We also would like to thank Dr. Eishi Hirasaki of the PRI for his valuable suggestions. We are deeply grateful to the anonymous reviewers for improving the quality of this paper.