AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 132:381–394 (2007) Continuity or Discontinuity of the Life-Style in Central Italy During the Roman Imperial Age-Early Middle Ages Transition: Diet, Health, and Behavior Giovanna Belcastro,* Elisa Rastelli, Valentina Mariotti, Chiara Consiglio, Fiorenzo Facchini, and Benedetta Bonfiglioli Laboratory of Bioarchaeology and Forensic Osteology, Department of Experimental Evolutionary Biology, University of Bologna, 40126 Bologna, Italy KEY WORDS dental pathologies; skeletal health markers; diet ABSTRACT Dento-alveolar pathologies and alterations (dental wear, caries, abscesses, ante mortem tooth loss (AMTL), calculus, hypoplastic defects, and chipping) and skeletal markers of health (cribra orbitalia and periostitis) were analyzed in two skeletal samples from the necropolises of Quadrella (I–IV c. AD) and Vicenne-Campochiaro (VII c. AD) in the Molise region of central Italy. The aim was to determine if the Roman Imperial Age-Early Middle Ages transition characterized by political, socioeconomic, and cultural transformations affected the biology of these populations, particularly their alimentation and health status. The frequencies of caries and AMTL, similar in the two samples, suggest a high consumption of carbohydrates. The higher levels of heavy wear, calculus, and interproximal chipping in the Vicenne population indicate a greater use of fibrous foods (both meat and others), in line with the dietary model of Germanic peoples. Health conditions do not appear to have been good in either period, as shown by the high frequencies of linear hypoplasia and the presence of cribra orbitalia and periostitis. The diet of the individuals buried with horses of the Vicenne population did not differ from that of the rest of the population, whereas there were evident differences in the use of the teeth for nonmasticatory activities among these individuals. Therefore, from the point of view of alimentation and health status, the profound socioeconomic and cultural transformations during the Late Antiquity-Early Middle Ages transition do not seem to have been translated into a true discontinuity of the two Molisan populations. Am J Phys Anthropol 132: 381–394, 2007. V 2006 Wiley-Liss, Inc. The Late Antiquity-Early Middle Ages transition in Europe was a period of profound cultural, political, social, and economic transformations after the fall of the western Roman Empire (dated at 476 AD). The first two centuries of the Roman Imperial Age were characterized by general peace and economic prosperity, but during the III c. AD the socioeconomic and political situation worsened, in line with the general conditions of the Roman State that suffered from political instability, heavier taxes, collapse of trade and agriculture, etc. (Righini, 1976). The following centuries saw the spread of epidemic diseases (plague, smallpox, tuberculosis, leprosy, malaria), famine, and undernourishment, due also to increased contacts with other populations following the wars and barbarian invasions (Kiple, 1999). Although archaeological, historical, and demographic data testify to this process of transformation—for example, the Italian population was halved from eight million (III c. AD) to four million (second half VI–second half IX c. AD) (Del Panta et al., 1996)—there have been few osteoarchaeological and epidemiological studies of these transitional phases and their possible effects on population biology. The purpose of this paper is to document and interpret dental and osseous markers of health in two skeletal series from the Sannio area of Molise, central Italy: one from the Roman Imperial period (Quadrella, Isernia, I–IV c. AD) and the other from the Early Middle Ages (Vicenne-Campochiaro, Campobasso, VII c. AD) (Fig. 1). Specifically, we test the hypothesis that alimentary patterns changed and health status declined, reflecting profound socioeconomic and cultural transformations during the Late Antiquity-Early Middle Ages transition. THE NECROPOLISES C 2006 V WILEY-LISS, INC. C The Roman necropolis of Quadrella (QDR), consisting of over 110 graves, was an extra-urban area of Aesernia, a Roman colony since 263 BC. As soon as Aesernia became a municipium (self-governing centre of the Roman Empire), there was a strong influx of people from Latium, which contributed to the prosperity of the small city. Most graves are of the \cappuccina" type (burial covered with tiles in a double pitch roof) and the few grave goods are uniform and repetitive, indicating \a rather undifferentiated social extraction of the occupants" (Terzani, 1991; Terzani and Matteini Chiari, 1997). Inscriptions on tombstones refer to liberti (freed slaves) (Terzani, 1991). The early medieval necropolis of Vicenne-Campochiaro (VCN), consisting of 167 graves, is situated between Sepino, Bojano, and Isernia and is dated to the Lombard Grant sponsor: Italian Ministry of University and Research (MIUR)–Cofin. *Correspondence to: Giovanna Belcastro, Laboratory of Bioarchaeology and Forensic Osteology, Department of Experimental Evolutionary Biology, University of Bologna, Via Selmi 3, 40126 Bologna, Italy. E-mail: maria.belcastro@unibo.it Received 28 March 2006; accepted 10 October 2006 DOI 10.1002/ajpa.20530 Published online 11 December 2006 in Wiley InterScience (www.interscience.wiley.com). 382 M.G. BELCASTRO ET AL. nomadic populations of the Iron Age and persists in the Age of Migrations (IV–X c. AD) (Kurylev et al., 1989). For the community of Vicenne, therefore, we can talk of a multicultural and perhaps multiethnic context with \Asian," Germanic, and local components (Genito, 1991; Belcastro and Facchini, 2001; Belcastro et al., 2001a,b, 2003). DIETARY PATTERNS IN THE ROMAN IMPERIAL PERIOD AND EARLY MIDDLE AGES Fig. 1. The map refers to the Italian territory during the Lombard-Byzantine period. period (VI–VIII c.). The necropolis refers to a community located in the areas described by the Lombard historian Paolo Diacono, in his History of the Lombards, as extensive uninhabited areas in the plain between Sepino, Bojano, and Isernia (book V, 29) used to control the borders of the Duchy of Spoleto and Benevento. With the current archaeological and historical data, it is not possible to refer the necropolis to a certain town and the graves may have been related to nomadic settlements (De Benedittis, 1988; Ceglia and Genito, 1991). One of the most important elements of the necropolis is the multiethnic context, observed by the archaeologists and historians who studied the materials and the types of burials. In fact, Lombard, local, and Avar materials (jewels, stirrups) were found (Ceglia and Genito, 1991; Genito, 1991). Moreover, the discovery of 12 graves of horse and man, i.e., the contextual burial with a horse equipped with typically Avar harness and stirrups, is the first documentation of this type of burial in Europe (Genito, 1991). It is also important to underline that stirrups have not been found in any other European necropolis dated to before this period. Stirrups revolutionized the military strategies of fighting, since without them a warrior could not fight while riding a horse. These graves, with a human skeleton and a horse skeleton buried together in the same pit, are considered high-status graves, because of the inferred sacrifice of the horse on the death of the master (Ceglia, 2000). This is known elsewhere in Italy only in grave 43 of the Lombard necropolis of San Mauro at Cividale del Friuli (Silva, 2000). This type of burial is not reported for Europe: from protohistoric times until the Alemannic cemeteries, horses were buried separately from humans. The type of burial at Vicenne is quite similar to the numerous ones in Eurasian The typical Roman meal of the lower classes consisted of bread (rich in bran and impurities), wine, vegetables— onions, garlic, and chick peas were typical of the lower classes (Dosi and Schnell, 1990)—lard, some fruit, and olives. In fact, Cato informs us that olives and bread were the basic foods of farmers and the working class (Brothwell and Brothwell, 1969). Vegetables were probably the most frequent accompaniment to bread but they could alternate with other poor dishes like fish and cheese. Meat was generally a luxury, but some types were available to the less well-to-do classes, like sausages, ham, or poultry, which could form part of the diet of urban populations (Neri, 1985). Hunting provided game, while domestic animal breeding was related to the rite of sacrifice: the idea of meat as an offering to the divinity was so strong that to kill an animal only to eat its meat was equivalent to squandering it (Dupont, 1997). In the countryside, the consumption of meat was rarer and generally related to festive occasions. Puls—a porridge of cereals mixed with water, salt, and a little oil (Neri, 1985)—and dairy products were an alternative source of protein, especially in the sylvo-pastoral economy of hilly and mountainous regions, like the Sannio area of Molise. The historical literary sources indicate that the most important products of the region were wheat, olives, wine (Silius Italicus), and figs (Pliny) (White, 1970). However, the principal activity of this region was sheep husbandry, with the consequent production of wool, milk, and cheese, and probably meat. Cheese was eaten freshly made or preserved and was an important ingredient of Roman bread and cakes (Brothwell and Brothwell, 1969). After the fall of the western Roman Empire, the different alimentary traditions were quite evident: the Mediterranean dietary pattern (bread, oil, and wine) characterized the Roman world, while hunting, cattle breeding, and fishing were typical of the Germanic peoples. Among the Lombards, many pacts were ratified on the basis of livestock exchanges, and the numerous remains of ritual meals discovered in the graves attest to the breeding of cattle, pigs, and sheep (Bóna, 1990), as also observed in the necropolis of VCN. Around the V–VI century, however, there was a coming together of the two cultures and a mutual integration in the following centuries (Montanari, 1993). For the Early Middle Ages, the historical-historiographic sources emphasize the importance of both cereals (albeit inferior grains following the collapse of wheat production) and the vegetable and animal products typical of uncultivated areas (game, fish and livestock raised in clearings or in the woods). In most of the population, meat was boiled—roasts were exclusive to the warrior nobility—to attenuate the taste of salt with which the meat was preserved, to tenderize it, and to exploit the nutritious juices dissolved in the broth, which was then reutilized to prepare other dishes (Rösener, 1996; Montanari, 1997). The meal for most populations consisted of a stew cooked with beans, vegetables, a small amount of fresh or salted meat, small pieces of stale bread and flour, American Journal of Physical Anthropology—DOI 10.1002/ajpa 383 ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION TABLE 1. Distribution of adult individuals and teeth in Quadrella (QDR) and Vicenne (VCN) samples Individuals Males Females n.i.a Adult young Adulte middle Adult old Adult Total a Teeth QDR VCN Individuals of VCN buried with horses 26 28 13 19 9 24 15 67 45 40 3 41 14 21 12 88 12 – – 5 2 3 2 12 QDR VCN Individuals of VCN buried with horses 514 514 176 487 163 419 135 1204 1033 679 42 966 311 334 143 1754 276 – – 140 43 68 25 276 n.i., not identified. followed by bread and wine; the composition of the dish changed with the seasons. Hence, a certain amount of animal products seems to have formed part of the diet of the poor classes in this period; this is an anomalous occurrence in the history of alimentation, which previously and subsequently was characterized by the prevalence of cereals (Montanari, 1997). MATERIALS AND METHODS Data concerning oral health for the Quadrella sample (QDR) have already been published (Bonfiglioli et al., 2003), whereas the data for Vicenne (VCN) are unpublished, except for preliminary data on chipping (Bonfiglioli et al., 2004). The composition of the skeletal samples has already been reported (Brasili and Belcastro, 1998; Belcastro and Facchini, 2001). Table 1 presents the adult sample analyzed. The dental features examined were recorded according to the criteria reported in Belcastro et al. (2004). Dental wear on the occlusal surface was scored according to Smith (1984) and the degrees of wear were grouped into mild (degrees 1 and 2), intermediate (degrees 3 and 4), and heavy (degrees 5–8). Here we present the results for heavy wear. Caries were recorded according to Powell (1985): \only those cavities that would admit the tip of a dental explorer were scored as actual caries to eliminate false scoring of discolored but intact enamel." In addition, we calculated the \caries correction factor" (Lukacs, 1992, 1995), which estimates caries in a more appropriate way by relating ante mortem tooth loss (AMTL) to exposure of the pulp cavity due to caries or wear. Abscesses were recorded only in the case of a perforating fistula (Brothwell, 1981) and AMTL was evaluated according to Lukacs (1989). Because calculus can easily be removed during the phases of excavation and restoration, we only scored its presence–absence. Linear enamel hypoplasia (LEH) was scored according to the criteria of the Fédération Dentaire International (1992). We used a three-grade scale to score chipping (Bonfiglioli et al., 2004). A few, but clear, indications help to distinguish ante mortem chipping from post mortem damage. In accordance with Milner and Larsen (1991), \the smoothing of the sharp edges of fractured enamel through continued attrition and the staining characteristic of adjacent broken and intact surfaces will help researchers identify when the damage occurred." Cribra orbitalia were detected macroscopically in individuals with at least one orbital roof preserved. The degree of severity was scored according to the Stuart-Macadam (1982) scale. Periostitis was scored as absent or present when new superficial bone and fine striae were observable on the femur and tibia (Fig. 2). In particular, we only Fig. 2. Presence of periostitis (the slightest grade). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] scored the complete bone, although some small fragments and the proximal and distal ends may have been absent. Side and bilateral presence and, for the tibia, medial and/ or lateral position were recorded. Bones with clear signs of fracture associated with periostitis were excluded. The frequencies of the dental features were calculated both per individual (number of affected individuals/number of observations) and per tooth (number of affected teeth/number of observations), taking into account the jaw and side. For the per individual frequencies of caries, abscesses, AMTL, calculus, and chipping, subjects were eliminated from the analysis if it was not possible to observe at least one complete quadrant of their dentition. For the per individual frequencies of LEH, the subjects were considered \with" or \without hypoplastic defects" only if it was possible to observe at least four teeth (maxillary incisors and mandibular canines), according to Goodman and Rose (1990). Interpopulation and, in some cases, intrapopulation differences were analyzed with the Chi-squared (v2) test, using StatSoft, Inc., 1992–1999, with the level of significance set at P < 0.05. RESULTS Dental features Heavy wear. Heavy wear (Table 2) is more frequent in the Early Middle Ages (VCN) sample than in the Roman Imperial one (QDR) and these differences are particularly evident in the females. In the middle and old adult age- American Journal of Physical Anthropology—DOI 10.1002/ajpa 384 – – 0.0007 – – 62.2 (694) 58.3 (825) 93.4 (603) 38.2 (916) 60.1 (1519) 0.0013 – – – – 0.0138 0.0000 0.0000 0.0272 0.0000 0.0000 – Old adult Middle adult Total Posterior Anterior Total sample ¼ males + females + n.i.; Values in parentheses indicate absolute values; AMTL, ante mortem tooth loss; LEH, linear enamel hypoplasia. Values in parentheses indicate percentages. 0.0000 – – 0.0001 0.0000 0.0052 – – – Mandible Young adult b 0.0002 – – – – – – 42.4 (512) 57.4 (666) 51.4 (422) 50.5 (756) 50.8 (1178) – 11.6 (724) 14.8 (977) 5.9 (660) 18.3 (1041) 13.5 (1701) 4.1 (807) 12.5 (385) 31.0 (458) 11.1 (569) 13.5 (791) 6.5 (522) 16.2 (838) 12.5 (1360) 2.3 (429) 14.9 (228) 20.6 (579) 0.0000 6.4 (546) 3.1 (765) 2.4 (453) 5.6 (858) 4.5 (1311) 1.7 (601) 4.9 (326) 9.2 (359) 0.6 (487) 1.8 (676) 0.2 (425) 1.9 (738) 1.3 (1163) 0.6 (351) 1.5 (133) 1.8 (546) 15.1 (794) 15.0 (919) 6.1 (658) 20.7 (1055) 15.1 (1713) 10.1 (950) 16.5 (298) 29.6 (331) 25.1 (793) 20.0 (895) 27.1 (653) 19.4 (1035) 22.4 (1688) 01.5 (946) 35.0 (283) 70.4 (317) Per tooth Maxilla 11.5 (496) 18.6 (655) 24.4 (426) 10.3 (725) 15.6 (1151) 03.2 (468) 17.2 (163) 30.5 (387) 0.0000 16.2 (512) 14.0 (665) 2.8 (427) 21.9 (750) 15.0 (1177) 12.4 (476) 11.6 (164) 19.4 (402) – 83.6 (67) 60.0 (60) 0.0009 48.3 (60) 20.3 (64) – 79.5 (88) 71.6 (67) – 58.1 (86) 63.5 (63)b Per individual a 0.0021 0.0000 0.0000 – 60.8 (441) 57.2 (502) 86.9 (359) 41.6 (584) 58.9 (943) 0.0000 – – 66.7 (69) 59.0 (780) 61.9 (896) 64.8 (648) 57.9 (1028) 60.6 (1676) P VCN 100.0 (83) 95.2 (62) VCN QDR P VCN 100.0 (83) Calculus Caries VCN VCN Wear P QDR Abscesses P QDR QDR P AMTL VCN (%) 0.0001 classes, the frequencies are twice as high in VCN as in QDR. In contrast, the teeth of the young adults are more heavily worn in QDR than in VCN (especially in the males). QDR TABLE 2. Frequencies of the dento-alveolar lesions in the total sample of Quadrella (QDR) and Vicenne (VCN)a P QDR LEH M.G. BELCASTRO ET AL. Caries. The frequencies of caries (Table 2) are the same in both samples (about 15%). This situation remains the same even after application of the caries correction factor: the frequencies of caries are 24.7% in QDR (10.0% for the anterior teeth and 33.2% for the posterior ones) and 24.6% in VCN (10.6% for the anterior teeth and 32.5% for the posterior ones). The anterior teeth, especially, the upper ones, are significantly more carious in the Early Middle Ages sample than in the Roman Imperial one, whereas the posterior ones show more decay in the Roman Imperial males (Table 3). Females do not show significant differences with age (Table 4), while young adult men are more affected in QDR and old men in VCN. Concerning the location and severity of carious lesions (Table 5), the frequency of occlusal caries is higher in QDR, whereas the frequencies of cervical and root caries are higher in VCN. The distribution of the types of caries is similar between the sexes, except for the significantly higher frequencies of cervical caries in the VCN males. In the old adults of both samples, occlusal caries are less numerous while cervical ones are more frequent. Abscesses. The frequencies of abscesses (Table 2) are higher in the Early Middle Ages sample. Because of the relationship between abscesses and pulp cavity exposure due to severe wear or caries (Lukacs, 1989), we counted the alveoli with both abscesses and caries: 73.3% (11/15) of the abscesses in QDR and 69.5% (41/59) in VCN are related to carious teeth. In particular, 53.8% (7/13) of the abscesses in QDR and 59.5% (28/47) in VCN are associated with cavity exposure caries, while, respectively, 14.3% (7/49) and 36.8% (28/76) of the total number of teeth with cavity exposure are associated with abscesses, the latter difference being significant (P ¼ 0.0047). In only one case in QDR and in two cases in VCN, the abscesses are related to pulp exposure due to severe wear. With age, the number of abscesses increases, especially in the male skeletons from VCN (Table 3). Ante mortem tooth loss. AMTL shows similar overall frequencies in the two necropolises (Table 2). However, the teeth of the Early Middle Ages females are more affected by AMTL than the teeth of the Roman Imperial ones, the difference being more evident for the posterior alveoli, whereas the opposite occurs for the teeth of males (Tables 3 and 4). The incidence of AMTL increases with age, but more so in VCN than in QDR. Calculus. Calculus is more frequent in the Early Middle Ages sample (Table 2), both per individual and per tooth, especially in the males (Table 3). For the anterior teeth, the differences between the two samples are significant for both sexes, while for the posterior teeth only for males. Interestingly, while there are few sex differences in QDR, the VCN men have higher values (P ¼ 0.0522) than the VCN women (Tables 3 and 4). Linear enamel hypoplasia. The per individual and per tooth frequencies of this feature are similar in the total samples (and in the males and females) of both necropolises (Tables 2–4). The anterior teeth are much more affected than the posterior ones. Moreover, the anterior teeth of the Early Middle Ages sample are significantly more affected than those of the Roman Imperial one. American Journal of Physical Anthropology—DOI 10.1002/ajpa TABLE 3. Frequencies of dento-alveolar lesions in the males of Quadrella (QDR) and Vicenne (VCN) samplesa Wear Per individual Per tooth Maxilla Mandible Total Young adult Middle adult Old adult a b Abscesses AMTL Calculus LEH VCN P QDR VCN P QDR VCN P QDR VCN P QDR VCN P QDR VCN P 68.0 (25)b 60.5 (43) – 84.6 (26) 71.1 (45) – 26.9 (26) 48.6 (35) – 68.0 (25) 55.3 (38) – 88.5 (26) 100.0 (44) 0.0475 91.3 (23) 100.0 (44) – 9.1 (220) 20.7 (270) 22.2 (176) 11.8 (314) 15.5 (490) 5.2 (211) 28.8 (66) 25.5 (188) 26.2 (481) 15.9 (508) 25.2 (381) 18.3 (608) 31.0 (989) 1.8 (599) 41.7 (163) 71.3 (171) 0.0000 19.7 (229) 17.4 (282) 4.4 (182) 26.1 (329) 18.4 (511) 15.0 (214) 9.0 (67) 24.5 (204) 14.5 (475) 14.5 (530) 7.8 (385) 18.7 (620) 14.5 (1005) 8.8 (599) 17.1 (170) 34.3 (181) – 7.0 (343) 3.5 (455) 2.8 (283) 6.2 (515) 5.0 (798) 1.6 (386) 4.8 (188) 12.0 (209) 0.0003 13.1 (282) 13.1 (352) 7.1 (240) 16.8 (384) 13.1 (634) 2.3 (213) 19.4 (103) 19.5 (282) 9.6 (471) 11.1 (599) 2.7 (402) 15.3 (628) 10.4 (1030) 4.0 (519) 11.9 (226) 23.0 (252) – 43.1 (225) 55.0 (271) 55.7 (176) 46.3 (320) 49.6 (496) 0.0000 61.6 (203) 53.8 (212) 86.2 (152) 41.1 (263) 57.6 (415) – – 0.0112 0.0126 0.0095 0.0457 0.0000 – – 0.0078 – 0.0122 – 0.0234 0.8 (245) 1.6 (309) 0.5 (201) 1.7 (353) 1.3 (554) 1.1 (182) 0.0 (32) 1.6 (314) – 0.0574 0.0014 0.0002 – – 0.0000 – 0.0050 – – 61.1 (470) 64.8 (512) 65.9 (381) 61.2 (601) 63.0 (982) 0.0070 0.0207 0.0000 0.0000 63.9 (427) 53.8 (480) 92.3 (352) 37.1 (555) 58.5 (907) – – 0.0310 – – 0.0012 0.0424 – Values in parentheses indicate absolute values; AMTL, ante mortem tooth loss; LEH, linear enamel hypoplasia. Values in parentheses indicate percentages. Wear Per Individual Per tooth Maxilla Mandible Anterior Posterior Total Young Adult Middle Adult Old Adult 1 Caries Abscesses AMTL Calculus LEH QDR VCN P QDR VCN P QDR VCN P QDR VCN P QDR VCN P QDR VCN P 53.8 (26) 55.0 (40) – 71.4 (28) 87.5 (40) – 14.3 (28) 48.0 (25) 0.0084 60.0 (25) 80.0 (30) – 82.1 (28) 100.0 (36) 0.0129 100.0 (26) 100.0 (37) – 8.7 (206) 13.4 (292) 17.1 (181) 8.2 (317) 22.2 (297) 25.4 (362) 30.0 (257) 20.1 (402) 0.0001 16.4 (207) 12.2 (288) 2.3 (176) 20.4 (319) 16.4 (304) 16.0 (363) 3.9 (258) 24.0 (409) 9.0 (223) 14.7 (361) 6.2 (226) 16.5 (358) 15.4 (253) 20.8 (400) 11.1 (252) 23.4 (401) 0.0225 40.9 (215) 60.9 (297) 49.5 (182) 54.2 (330) 56.9 (295) 59.3 (359) 63.6 (253) 54.9 (401) 11.4 (498) 24.0 (659) 0.0000 13.9 (495) 16.2 (667) 52.5 (512) 58.3 (654) 1.9 (211) 9.3 (97) 24.8 (141) 0.9 (329) 25.8 (120) 67.8 (146) – 12.2 (213) 13.4 (97) 16.7 (138) 12.3 (332) 16.4 (128) 24.0 (150) 0.0001 0.0014 0.0000 0.0013 0.0000 – 0.5 (188) 1.0 (297) 0.0 (176) 1.3 (309) 5.4 (203) 2.6 (306) 1.8 (170) 4.7 (339) 0.0042 – 0.8 (485) 3.7 (509) 0.0023 12.5 (584) 18.7 (653) – 0.0 (154) 2.0 (101) 1.2 (171) 1.9 (212) 5.1 (138) 5.3 (150) – 2.6 (195) 11.2 (125) 24.4 (221) 4.4 (272) 13.2 (159) 40.8 (206) – – – – – – – 0.0118 – 0.0330 0.0291 0.0405 0.0170 0.0029 0.0154 – 0.0022 – – 61.2 (178) 63.3 (226) 86.6 (157) 47.0 (247) 60.1 (253) 65.5 (325) 95.3 (236) 40.9 (342) 62.4 (404) 63.1 (578) – – 0.0020 – – – – 0.0003 Values given in parentheses indicate absolute values; AMTL ¼ ante mortem tooth loss, LEH ¼ linear enamel hypoplasia. 385 American Journal of Physical Anthropology—DOI 10.1002/ajpa TABLE 4. Frequencies of dento-alveolar lesions in the females of Quadrella (QDR) and Vicenne (VCN) samples1 ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION Anterior Posterior Caries QDR 386 Total QDR Occlusal Total Anterior Posterior Young adult Middle adult Old adult Coronal Total Anterior Posterior Young adult Middle adult Old adult Cervical Total Anterior Posterior Young adult Middle adult Old adult Root Total Anterior Posterior Young adult Middle adult Old adult Destructive Total Anterior Posterior Young adult Middle adult Old adult a Males VCN QDR Females VCN QDR VCN N n % N n % P N n % N n % P N n % N n % P 176 12 164 59 19 78 33 0 33 22 2 6 18.8 0.0 20.1 37.3 10.5 7.7 258 40 218 96 50 98 35 0 35 24 6 3 13.6 0.0 16.1 25.0 12.0 3.1 – – – – – – 94 8 86 32 6 50 16 0 16 11 0 4 17.0 0.0 18.6 34.4 0.0 8.0 146 30 116 53 29 62 15 0 15 12 2 0 10.3 0.0 12.9 22.6 6.9 0.0 – – – – – 0.0371 69 4 65 26 13 23 15 0 15 11 2 2 21.7 0.0 23.1 42.3 15.4 8.7 108 10 98 41 21 36 19 0 19 12 4 3 17.6 0.0 19.4 29.3 19.0 8.3 – – – – – – 176 12 164 59 19 78 22 5 17 8 3 7 12.5 41.7 10.4 13.6 15.8 9.0 258 40 218 96 50 98 48 5 43 31 7 7 18.6 12.5 19.7 32.3 14.0 7.1 – 0.0246 0.0087 0.0066 – – 94 8 86 32 6 50 13 3 10 5 2 5 13.8 37.5 11.6 15.6 33.3 10.0 146 30 116 53 29 62 23 3 20 14 4 5 15.8 10.0 17.2 26.4 13.8 8.1 – – – – – – 69 4 65 26 13 23 8 2 6 3 1 2 11.6 50.0 9.2 11.5 7.7 8.7 108 10 98 41 21 36 24 2 22 16 3 2 22.2 20.0 22.4 39.0 14.3 5.6 – – 0.0216 0.0134 – – 176 12 164 59 19 78 40 2 38 8 7 29 22.7 16.7 23.2 13.6 36.8 37.2 258 40 218 96 50 98 92 18 72 17 16 54 35.7 45.0 33.0 17.7 32.0 55.1 0.0026 0.0399 0.0227 0.0002 – 0.0132 94 8 86 32 6 50 22 1 21 0 1 19 23.4 12.5 24.4 0.0 16.0 38.0 146 30 116 53 29 62 59 17 42 11 12 35 40.4 56.7 36.2 20.8 41.4 56.5 0.0045 0.0311 – 0.0036 – 0.0396 69 4 65 26 13 23 14 1 13 0 3 8 20.3 25.0 20.0 0.0 23.1 34.8 108 10 98 41 21 36 33 3 30 6 4 19 30.6 30.0 30.6 14.6 19.0 52.8 – – – 0.0451 – – 176 12 164 59 19 78 1 0 1 0 0 1 0.6 0.0 0.6 0.0 0.0 1.3 258 40 218 96 50 98 15 3 12 2 4 7 5.8 7.5 5.5 2.1 8.0 7.1 0.0025 – 0.0066 – – – 94 8 86 32 6 50 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 146 30 116 53 29 62 7 1 6 2 2 3 4.8 3.3 5.2 3.8 6.9 4.8 0.0291 – 0.0339 – – – 69 4 65 26 13 23 1 0 1 0 0 1 1.4 0.0 1.5 0.0 0.0 4.3 108 10 98 41 21 36 7 2 5 0 2 4 6.5 20.0 5.1 0.0 9.5 11.1 – – – – – – 176 12 164 59 19 78 44 5 39 19 4 27 25.0 41.7 23.8 32.2 21.1 34.6 258 40 218 96 50 98 55 11 44 18 13 23 21.3 37.5 20.2 18.8 26.0 23.5 – – – 0.0443 – – 94 8 86 32 6 50 25 4 21 4 2 17 26.6 50.0 24.4 12.5 33.3 34.0 146 30 116 53 29 62 33 8 25 11 7 15 22.6 26.7 21.6 20.8 24.1 24.2 – – – – – – 69 4 65 26 13 23 16 1 15 6 2 7 23.2 25.0 23.1 23.1 15.4 30.4 108 10 98 41 21 36 21 3 18 6 6 8 19.4 30.0 18.4 14.6 28.6 22.2 – – – – – – Total ¼ males + females + n.i.; N ¼ total number of the teeth observed; n ¼ total number number of the teeth affected. M.G. BELCASTRO ET AL. American Journal of Physical Anthropology—DOI 10.1002/ajpa TABLE 5. Per tooth frequencies of location and severity of caries in the Quadrella (QDR) and Vicenne (VCN) samplesa 0.0396 0.0432 0.0059 – 0.0068 0.0366 0.0068 – 39.2 35.7 45.8 32.0 37.3 27.1 43.9 54.5 111 119 109 121 230 89 47 67 283 333 238 378 616 329 107 123 47.7 43.9 59.5 37.1 45.5 35.7 62.8 53.5 Total ¼ males + females + n.i.; N ¼ total number of the teeth observed; n ¼ total number number of the teeth affected. a 197 262 173 286 459 199 86 129 0.0058 0.0004 – 0.0000 0.0000 0.0388 0.0240 – 43.9 37.0 57.3 29.7 40.2 30.1 50.7 66.4 190 184 203 171 374 175 77 97 433 497 354 576 930 582 152 146 55.5 50.9 65.2 46.1 53.0 37.9 67.3 68.9 117 119 105 131 236 77 37 111 211 234 161 284 445 203 55 161 0.0011 0.0003 0.0031 0.0000 0.0000 0.0094 0.0014 – 42.3 36.1 51.8 30.9 38.9 28.4 47.9 61.0 309 307 314 302 616 264 124 164 731 851 606 976 1582 928 259 269 51.9 45.6 61.3 40.8 48.4 35.3 64.5 61.9 247 263 241 269 510 159 91 205 476 577 393 660 1053 450 141 331 387 Chipping. Chipping is present in practically all individuals of both necropolises (Table 6). However, the total frequencies of chipping are significantly higher in the Roman Imperial teeth than in the Early Middle Ages ones, even when the sexes and age-classes are considered separately. The frequencies increase from young to middle to old adults. Regarding intrapopulation differences, the anterior teeth are always more chipped than the posterior ones in both necropolises (QDR, P ¼ 0.0000; VCN, P ¼ 0.0000), and the males are more affected than the females, with significant differences only in QDR (P ¼ 0.0241). The frequencies of lingual and multiple chipping (a tooth with several chips in different positions) are higher in the Roman Imperial sample, whereas interproximal chipping is more frequent in the Early Middle Ages one, especially in the females (Table 7). Significant sex differences only occur in the Early Middle Ages sample: for buccal chipping (more frequent in males) and interproximal chipping (more frequent in females) (P ¼ 0.0099). Grade 1 (small) chipping is more frequent in VCN and Grade 2 in QDR. There is no difference in the distribution of large chips (Grade 3). 94 115 103 106 209 71 54 69 – 97.3 36 37 100.0 26 44 25 64 Per individual Per tooth Maxilla Mandible Anterior Posterior Total Young adult Middle adult Old adult 64 100.0 84 82 97.6 – 25 100.0 44 100.0 – 26 n N n N P % n N % n N N n % N n % P VCN Males QDR VCN Total QDR TABLE 6. Frequencies of chipping in Quadrella (QDR) and Vicenne (VCN) samplesa QDR % Females VCN % P ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION Bone features Cribra orbitalia. Unfortunately, the data on cribra orbitalia refer to small samples (QDR: 8/35; VCN: 9/48) and interpopulation differences are not significant. However, in both necropolises, we only observed cases of cribra orbitalia corresponding to the first degree of the StuartMacadam scale. Periostitis. The Early Middle Ages sample shows a higher prevalence of femoral periostitis (Table 8), also in the bilateral condition (total: P ¼ 0.0084; males: P ¼ 0.0450). The tibia is more frequently affected than the femur, the medial side more than the lateral, and the periostitis is usually bilateral in both samples. There are no significant differences between the frequencies of the subjects that present both cribra orbitalia and LEH. Individuals buried with horses Table 9 reports the per tooth frequencies of dento-alveolar pathologies in the individuals buried with horses versus the males of VCN and versus the males of QDR. The teeth of the individuals buried with horses are significantly more worn and more affected by LEH than those of the men of both necropolises; the higher frequencies of calculus and abscesses than in the QDR males reflect the differences between the VCN and QDR men. The frequency of chipping is significantly higher in the individuals buried with horses than in the VCN males, especially, for the anterior teeth, the buccal position, and Grade 3 (Table 9). They also present more buccal chipping than the QDR men. It was possible to examine cribra orbitalia and periostitis in only a few of these individuals: cribra orbitalia are absent in all six of the skeletons examined (individuals buried with horses: 0/6 vs. males: 5/23). Since the surface of many lower limb bones of the individuals buried with horses are decorticated due to post mortem degradation, we can only comment briefly on periostitis. Of six femora observed from five individuals, only one presents periostitis. For the tibia, our observations refer to five bones from five individuals and only one shows periostitis. American Journal of Physical Anthropology—DOI 10.1002/ajpa 388 Total QDR Buccal Total Young Middle Old Lingual Total Young Middle Old Interproximal Total Young Middle Old Multiple Total Young Middle Old Grade 1 Total Young Middle Old Grade 2 Total Young Middle Old Grade 3 Total Young Middle Old a Males VCN QDR VCN Females VCN QDR N n % N n % P N n % N n % P N n % N n % P 510 159 91 205 235 91 39 90 46.1 57.2 42.9 43.9 616 264 124 164 311 150 58 85 50.5 56.8 46.8 51.8 – – – – 236 77 37 111 110 49 13 45 46.6 63.6 35.1 40.5 374 175 77 97 202 103 41 55 54.0 58.9 53.2 56.7 – – – 0.0142 209 71 54 69 96 34 26 35 45.9 47.9 48.1 50.7 230 89 47 67 108 47 17 30 47.0 52.8 36.2 44.8 – – – – 510 159 91 205 87 26 19 30 17.1 16.4 20.9 14.6 616 264 124 164 75 26 16 25 12.2 9.9 12.9 15.2 0.0201 0.0356 – – 236 77 37 111 38 14 11 10 16.1 18.2 29.7 9.0 374 175 77 97 48 19 9 13 12.8 10.9 11.7 13.4 – – 0.0195 – 209 71 54 69 36 11 8 12 17.2 15.5 14.8 17.4 230 89 47 67 26 7 7 12 11.3 7.9 14.9 17.9 – – – – 510 159 91 205 112 13 13 41 22.0 8.2 14.3 20.0 616 264 124 164 186 71 41 40 30.2 26.9 33.1 24.4 0.0110 0.0000 0.0012 – 236 77 37 111 46 13 5 24 19.5 16.9 13.5 21.6 374 175 77 97 95 44 20 19 25.4 25.1 26.0 19.6 – – – – 209 71 54 69 47 21 8 11 22.5 29.6 14.8 15.9 230 89 47 67 81 27 21 21 35.2 30.3 44.7 31.3 0.0023 – 0.0009 – 510 159 91 205 76 4 20 44 14.9 2.5 22.0 21.5 616 264 124 164 44 17 9 14 7.1 6.4 7.3 8.5 0.0000 – 0.0018 0.0005 236 77 37 111 42 1 8 32 17.8 1.3 21.6 28.8 374 175 77 97 29 9 7 10 7.8 5.1 9.1 10.3 0.0002 – – 0.0007 209 71 54 69 30 5 12 11 14.4 7.0 22.2 15.9 230 89 47 67 15 8 2 4 6.5 9.0 4.3 6.0 0.0053 – 0.0067 – 510 159 91 205 210 73 32 77 41.2 45.9 35.2 37.6 616 264 124 164 325 160 55 77 52.8 60.6 44.4 47.0 0.0001 0.0022 – 0.0000 236 77 37 111 84 33 16 33 35.6 42.9 43.2 29.7 374 175 77 97 195 104 35 44 52.1 59.4 45.5 45.4 0.0001 – – – 209 71 54 69 93 37 16 31 44.5 52.1 29.6 44.9 230 89 47 67 123 56 20 33 53.5 62.9 42.6 49.3 0.0371 – – – 510 159 91 205 182 64 31 69 35.7 40.3 34.1 33.7 616 264 124 164 160 62 31 44 26.0 23.5 25.0 26.8 0.0004 0.0002 – – 236 77 37 111 88 33 11 40 37.3 42.9 29.7 36.0 374 175 77 97 90 43 18 21 24.1 24.6 23.4 21.6 0.0002 0.0032 – 0.0165 209 71 54 69 72 24 20 24 34.4 33.8 37.0 34.8 230 89 47 67 67 19 13 23 29.1 21.3 27.7 34.3 – – – – 510 159 91 205 118 22 28 59 23.1 13.8 30.8 28.8 616 264 124 164 131 42 38 43 21.3 15.9 30.6 26.2 – – – – 236 77 37 111 64 11 10 38 27.1 14.3 27.0 34.2 374 175 77 97 89 28 24 32 23.8 16.0 31.2 33.0 – – – – 209 71 54 69 44 10 18 14 21.1 14.1 33.3 20.3 230 89 47 67 40 14 14 11 17.4 15.7 29.8 16.4 – – – – Total ¼ males + females + n.i.; N ¼ total number of the teeth observed; n ¼ total number number of the teeth affected. M.G. BELCASTRO ET AL. American Journal of Physical Anthropology—DOI 10.1002/ajpa TABLE 7. Per tooth frequencies of position and grade of chipping in Quadrella (QDR) and Vicenne (VCN) samplesa – – – – a N ¼ total number of the teeth observed; n ¼ total number number of the teeth affected; FP, femoral periostitis; TP, tibial periostitis; CO, cribra orbitalia; CO + LEH, individuals with both cribra orbitalia and linear enamel hypoplasia. – – – – – – – 52.2 35.0 10.0 10.0 75.0 0.0 22.2 11 9 1 1 5 1 3 48 35 35 35 26 26 48 26 19 2 5 13 1 7 41 28 26 28 26 26 30 63.4 67.9 7.9 17.9 50.0 3.8 23.3 25 15 2 2 17 2 10 52.1 42.9 5.7 5.7 65.4 7.7 20.8 – – – – – – – 22 22 22 22 14 14 15 14 9 2 4 7 0 4 63.6 40.1 9.1 18.2 50.0 0.0 26.7 26 16 16 16 15 15 30 15 8 0 2 8 2 6 57.7 50.0 0.0 12.5 53.3 13.3 20.0 – – 18 17 17 17 11 11 13 61.1 52.9 5.9 5.9 45.4 9.1 23.1 23 20 20 20 12 12 18 12 7 2 2 9 0 4 – – – – 21.7 18.8 0.0 0.0 0 0 0 0 51 34 34 34 4 1 0 2 FP Per individual Bilateral Right femur Left femur TP Per individual Bilateral Right tibia Left tibia Mediale surface Lateral surface CO + LEH 47 34 34 34 8.5 2.9 0.0 5.9 13 10 0 1 25.5 29.4 0.0 2.9 0.0266 0.0084 – – 24 18 18 18 4 1 0 2 16.7 5.6 0.0 11.1 26 18 18 18 8 7 0 1 30.8 38.9 0.0 5.6 0.0422 0.0450 – – 20 15 15 15 0.0 0.0 0.0 0.0 23 16 16 16 5 3 0 0 P % n n N N N n % n % P N N % N n % P QDR VCN Males QDR VCN Total QDR TABLE 8. Frequencies of periostitis in the total, males and females of Quadrella (QDR) and Vicenne (VCN) samplesa % N Females VCN ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION 389 DISCUSSION Dental features Wear. The higher frequency of heavy wear in the Early Middle Ages sample, particularly in the posterior teeth, suggests a higher consumption of hard fibrous food requiring vigorous mastication. However, the differences between the two populations could also be explained by the different cleaning and preparation of certain foods (especially vegetables) to soften them and to remove abrasive elements, as hypothesized for the Roman Imperial sample (Bonfiglioli et al., 2003). The heavier wear on the anterior teeth of the VCN females than the QDR ones could be related to the higher frequency of AMTL in the posterior teeth of the VCN women, which shifts the masticatory load anteriorly. Caries. The lack of differences between the two samples in the frequencies of caries suggests similar alimentary habits, with abundant carbohydrates. The caries frequencies on the posterior teeth suggest that the VCN men had a diet less rich in carbohydrates than the VCN women and perhaps than the QDR men. However, some authors (Maat and Van der Velde, 1987; Kerr et al., 1990; Hillson, 2001) have reported a negative correlation between dental wear and caries following \cancellation" of the sulci and fossae related to the natural morphology of the tooth or to removal of the carious tissue. The posterior teeth of the Early Middle Ages males are much more worn than those of the Roman Imperial males and this could explain the lower frequency of caries, which may have been removed by the wear. The number of caries increases with age in both samples, but there are higher frequencies in the older adults of the Early Middle Ages sample and this difference could be related to a greater increase of cervical caries. This may have been due to more severe retraction of the alveolar border, exposing the cement enamel junction and part of the root to the action of caries. In fact, the frequencies of severe alveolar bone loss (>4.00 mm) are significantly higher in VCN (69.7%, 23/33) than in QDR (38.1%, 8/21) (P ¼ 0.0223). In turn, this retraction could be related to the greater calculus accumulation in VCN, which may have led to inflammation, and to greater continuous eruption of the teeth in VCN, probably related to higher wear and thus greater mechanical stress on the teeth. However, significant differences in different age-classes are not observed in the females: it is possible that the higher incidence of AMTL in the teeth of the Early Middle Ages women eliminated the carious teeth, thus leveling possible differences in caries between the women of the two necropolises. Abscesses. The abscesses presented by the two samples appear to be related to the presence of caries with pulp cavity exposure. The significant differences between the two necropolises in the per tooth frequencies in both sexes and the absence of significance in the per individual frequencies in males could be related to the fact that many VCN males (41.2%, 7/17) present more than two abscesses. The higher number of abscesses in VCN does not reflect a similar distribution of caries, whose frequencies are similar in the two populations. Yet, in VCN there is a higher number of caries with pulp exposure that led to abscesses, even though the frequencies of caries with pulp exposure are similar in the two samples. Therefore, it is possible that in VCN the inflammation spread more easily along the pulp canal, causing long-term chronic periapical American Journal of Physical Anthropology—DOI 10.1002/ajpa 390 M.G. BELCASTRO ET AL. TABLE 9. Per tooth frequencies of dento-alveolar lesions in the individuals buried with horses and in the males of Vicenne (VCN) and Quadrella (QDR)a Individuals buried with horses Heavy wear Caries Abscesses AMTL Calculus LEH Total chipping Anterior Posterior Buccal Lingual Interproximal Multiple Grade 1 Grade 2 Grade 3 VCN males QDR males N n % N n % P (Ibh/VCNm) N n % P (Ibh/QDRm) 260 259 221 285 245 216 231 96 135 109 109 109 109 109 109 109 73 41 10 30 153 152 109 64 45 71 10 21 7 46 30 33 28.1 15.8 4.5 10.5 62.4 70.4 47.2 66.7 33.3 65.1 9.2 19.3 6.4 42.2 27.5 30.3 729 746 577 745 737 668 699 258 441 265 265 265 265 265 265 265 134 105 30 74 466 367 265 139 126 131 38 74 22 149 60 56 18.4 14.1 5.2 9.9 63.2 54.9 37.9 53.9 28.6 49.4 14.3 27.9 8.3 56.2 22.6 21.1 0.0001 – – – – 0.0000 0.0127 0.0000 – 0.0038 – – – 0.0093 – 0.0412 490 511 554 634 496 415 445 161 284 236 236 236 236 236 236 236 79 94 7 83 246 239 236 105 131 110 38 46 42 84 88 64 16.1 18.4 1.3 13.1 49.6 57.6 53.0 65.2 46.1 46.6 16.1 19.5 17.8 35.6 37.3 27.1 0.0001 – 0.0051 – 0.0010 0.0006 – – 0.0085 0.0010 – – 0.0027 – 0.0479 – a N ¼ total number of the teeth observed; n ¼ total number of the teeth affected; AMTL, ante mortem tooth loss, LEH, linear enamel hypoplasia; Ibh, Individuals buried with horse; VCNm, Vicenne males; QDRm, Quadrella males. inflammation, apparently the only type that leads to loss of bone material (Hillson, 2001). Ante mortem tooth loss. The exact cause of AMTL cannot be precisely determined, since it is a nonspecific indicator of pathologies resulting from numerous factors (Lukacs, 1989; Clarke, 1990; Hillson, 2001). Periodontal disease and calculus are closely correlated with ante mortem tooth loss but it is not always possible to identify either the type of periodontopathy (or its aetiology) by examination of the dry bone (Molnar and Molnar, 1985) or the actual relationship with calculus because of problems of underestimation of this character. In both samples, the distribution of AMTL conforms rather well to the pattern of caries, even though the differences are significant for AMTL but not for caries in the females. In the Early Middle Ages women, severely carious teeth, with substantial pulp exposure, may have been extracted, thus increasing the frequencies of AMTL and decreasing those of caries. Indeed, methods of tooth extraction are described in medical texts from both periods (Micheloni, 1976). The increasing frequencies of caries and AMTL in the different age-classes can be explained by the greater time of exposure to cariogenic factors, with consequent aggravation and multiplication of the lesions and subsequent tooth loss. However, the considerable masticatory stress on the teeth of the elderly VCN individuals, shown by the higher frequencies of heavy wear, could be an additional cause of AMTL, since by destabilizing the tooth it could have increased its mobility and the likelihood of its loss. Calculus. The consumption of high-protein foods contributes to increased alkalinity in the mouth, favoring precipitation of minerals in the oral fluids. However, poor oral hygiene and the consumption of carbohydrates are other important aspects of calculus formation, since the initiation of mineralization is related to the amount of plaque and thus to factors that increase its accumulation (Ånerud et al., 1991; Hillson, 1996). In addition, the microorganisms in dental plaque act by destroying the inhibitors of mineralization (Scheie, 1989). The high frequencies of caries and calculus in QDR suggest that carbohydrates formed a large part of the diet of the Roman Imperial population (Bonfiglioli et al., 2003) and also indi- cate a lack of regular oral hygiene. Nevertheless, in the Early Middle Ages sample from QDR, the higher frequencies of calculus, which do not correspond to higher frequencies of caries, may indicate a greater consumption of high-protein foods, as well as poor oral hygiene. This may also have been true for males with respect to females, suggesting some dietary differences within the population. The higher protein consumption could be related to a greater consumption of meat: in the Lombard era, cattle, pig, and sheep breeding was the fundamental wealth of Germanic peoples. Indeed, the presence of bony remains of meals is documented in graves of Lombard necropolises (Bóna, 1990). This practice is also known for several graves in the VCN necropolis in which Ovis bones were found (5 graves of horsemen and 12 other graves). Linear enamel hypoplasia. Enamel defects related to disturbances during the secretory phase of amelogenesis (Sarnat and Schour, 1941; Goodman et al., 1980) may be the result of an interrelationship between malnutrition and infectious diseases (Solomons and Keusch, 1981). The high frequencies of LEH in both samples suggest the presence of strong metabolic stresses during early childhood. According to Reid and Dean (2000), hypoplastic enamel defects occur after the first year of life on the surfaces of all anterior tooth types. This period corresponds to the time of weaning in Roman and Early Middle Ages societies, as indicated by historical literary sources. According to Soranus (98–117 AD), an infant should be fed only with breast milk until the age of 6 months, after which solid food (for example bread and milk, cereal, moist porridge) could be gradually introduced into the diet. Galen (130–200 AD) believed that solid food (bread, vegetables, meat, milk) could be given to an infant when he had cut his first tooth, and complete weaning occurred at about 3 years of age. The foods and times of weaning were similar in the following centuries because Early Middle Ages medical literature on infant feeding reflected the ideas of Soranus and Galen (Fildes, 1986). In our samples, nutritional deficits and serious diseases, related to the introduction of a solid diet deficient in protein content and perhaps not very sanitary, could have led to strong disturbances of amelogenesis. Nevertheless, in the American Journal of Physical Anthropology—DOI 10.1002/ajpa ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION 391 Fig. 3. Chipping in the posterior teeth of an individual (T. 27B, old male) of the Quadrella necropolis. [Color figure can be viewed in the online issue, which is available at www.interscience. wiley.com.] Fig. 4. Buccal chipping in the anterior teeth of an individual buried with horses (T. 66, young adult male) of the Vicenne necropolis. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] Early Middle Ages sample, the higher frequencies of LEH in the anterior teeth appear to reflect worse living conditions, especially during infancy and early childhood. Poor living conditions are also indicated by the infant mortality: about 27% and 25% for QDR and VCN, respectively (Brasili and Belcastro, 1998; Belcastro and Facchini, 2001; Belcastro et al., 2003). This agrees with other reports for Italy and Europe in that period (Del Panta et al., 1996). groups, in which the dentition is used in stone tool making to sharpen the cutting edge by biting off small flakes (Gould, 1968) or to manipulate certain rudimentary tools with the teeth (Lous, 1970; Molnar, 1972). Chipping. Chipping can be related to both masticatory and nonmasticatory activities (Turner and Cadien, 1969; Molnar, 1972; Milner and Larsen, 1991). It can result from the presence of contaminants (particularly grains of sand) in foods like roots and tubers (Robinson, 1963) or other hard resistant foods, but it can also be due to paramasticatory (breaking hard shells, shattering bones to extract the marrow, chewing of roots or leaves, etc.), nonmasticatory, and work activities (Iannetti et al., 1984) or accidental events. The presence of chipping in almost all individuals of the two necropolises and the high per tooth frequencies seem to indicate that the teeth were subjected to repeated traumas, especially the anterior ones. The higher frequencies of teeth with chipping and with multiple chips in the Roman Imperial individuals could be related to more frequent and/or more severe contacts with hard objects or to masticatory traumas (Fig. 3). The more frequent lingual chipping in the Roman Imperial sample seems to be due to masticatory traumas. The higher frequencies of interproximal chipping in the Early Middle Ages sample, especially in the posterior teeth, suggest the presence of contaminants in the diet. However, in VCN the fact that buccal chipping is more frequent in men and interproximal chipping in women may reflect greater nonmasticatory use of the teeth by males and higher masticatory stress in females, perhaps due to harder foods, as indicated by the higher values of wear and AMTL. Thus, the diet of the Early Middle Ages population may have contained extraneous elements associated with hard fibrous foods, whose consumption is supported by the higher frequencies of heavy wear. The high frequencies of buccal chipping in both samples and of multiple chipping in QDR suggest that the teeth were used as tools in daily activities, particularly by the men. This is documented in other human Bone features Cribra orbitalia. The aetiology of cribra orbitalia could be due to several causes. In our series, the absence of severe cribra orbitalia lesions and the relatively uniform distribution of the lesions in the two populations suggest substantially similar dietary and hygienic-sanitary habits. Periostitis. Many aetiological factors can be considered possible causes of periostitis (cf. Larsen, 1997, 2002). We found a higher frequency of femoral periostitis and of its bilateral condition in VCN, especially in the male sample. In our samples, the tibia is more affected by periostitis than the femur, but there are no significant interpopulation differences. Although it is difficult to assess the aetiology of this feature, the bilaterality of the tibial periostitis can be considered a pathognomonic marker of specific infection. In VCN, the bilaterality of the tibial lesions and the presence of facial lesions allowed us to diagnose leprosy in three individuals (Belcastro et al., 2005). Individuals buried with horses The 12 contextual graves of man and horse and some burial goods suggest a special role of these individuals in the Early Middle Ages community (Ceglia and Genito, 1991; Arslan, 2000). They and the other males of VCN present similar frequencies of some dento-alveolar pathologies and alterations (caries, abscesses, AMTL, calculus), probably indicating a homogeneous diet. Nevertheless, the differences in heavy wear (particularly in the anterior teeth) and chipping (more frequent in the individuals buried with horses even though 5/11 of them are young adults) suggest nonmasticatory uses of the teeth (leather softening, tool making, repairing, etc.). It is possible that the rank and activities of these subjects, probably also involved in fighting, resulted in repeated traumas. The most affected teeth are the anterior ones, which are more exposed to bumps and blows; the buccal chipping and the American Journal of Physical Anthropology—DOI 10.1002/ajpa 392 M.G. BELCASTRO ET AL. TABLE 10. Per tooth and per individual frequencies of Quadrella, Vicenne, and other Roman Imperial and Early Middle Ages necropolisesa Wear (%) Isola Sacra (I–III c. AD) Lucus Feroniae (II–IV c. AD) Rimini (II–IV. AD) Ravenna (I–IV c. AD) Quadrella (I–IV c. AD) Vicenne (VII c. AD) La Selvicciola (VII c. AD) a 15.6 22.4 Caries (%) Abscesses (%) Per tooth AMTL (%) Calculus (%) LEH (%) 4.0 6.1 0.2 0.6 6.8 12.4 33.9 26.9 35.5 46.0 15.0 15.1 12.6 1.3 4.5 0.8 12.5 13.5 18.2 50.9 60.6 27.1 58.9 60.0 23.1 Per individuals Cribra orbitalia (%) 19.4 30.8 59.3 22.8 18.7 23.5 AMTL, ante mortem tooth loss; LEH, linear enamel hypoplasia. severe (degree 3) type are the most frequent conditions (Fig. 4). In three of six cases, anomalous wear matches the \lingual surface attrition masticatory anterior teeth" lesion described in the literature (Turner and Cheuiche Machado, 1983; Irish and Turner, 1987, 1997; Comuzzie and Steele, 1988); therefore, our cases can be attributed to nonmasticatory activities (Belcastro et al., 2001a, 2003). The poor preservation of the bones of many of the individuals buried with horses prevents us from commenting on the total and almost total absence of cribra orbitalia and periostitis, respectively. COMPARISON WITH OTHER ROMAN IMPERIAL AND EARLY MIDDLE AGES ITALIAN SAMPLES Our results were compared with those of other series from Roman Imperial and Early Middle Ages necropolises near Rome: Isola Sacra and Lucus Feroniae (I–III c. AD) and the medieval sample of La Selvicciola (VII c. AD) (Manzi et al., 1999; Salvadei et al., 2001). These are the largest samples from Italy for those periods and they provide useful indications about how the transition occurred on the Tyrrhenian coast. It is interesting that the frequencies of caries, calculus, and LEH are more similar between the two series from Molise than between either of the Molise series and the coeval ones from Latium (Table 10). In particular, the caries (and AMTL) indicate similar widespread carbohydrate consumption and we can hypothesize that an unbalanced diet and relatively poor nutrition and health were present during both the Roman and medieval periods in central Italy. This is also suggested by the higher frequencies of LEH in Quadrella and Vicenne than in the other samples. Our data suggest that protein consumption was higher in Vicenne than in both the Roman Latium samples and the medieval one, probably reflecting group-specific cultural characteristics and dietary habits. Regarding the cribra orbitalia, we did not observe any significant differences between our data and those for Lucus Feroniae and La Selvicciola. In Molise, ecological factors seem to have conditioned the economic models more than the cultural, social, and economic changes following the fall of the western Roman Empire and contacts with new ethnic groups. Molise is a mountainous region, characterized even in pre-Roman times by an economy based on agriculture, practiced in inframontane basins fed by temporary fluctuating streams (since the mountains do not reach the limit of perennial snow, there are no rivers), and above all on the transhumance of sheep. In contrast, the supply of primary foodstuffs was probably not a problem for the Roman populations of Latium. Portus Romae, which contained the Isola Sacra necropolis, was an important port for trade with Mediterranean countries, while Lucus Feroniae was a somewhat poorer rural centre, inhabited mainly by war veterans, slaves, and liberti (free slaves), whose geographic proximity to Rome attenuated possible food shortages and moments of crisis (Manzi et al., 1999). We compared the data of cribra orbitalia with those from the north-eastern Adriatic area—Ravenna (I–IV c. AD) and Rimini (II–IV c. AD)—recorded with the same methods used in the present study (Facchini et al., 2004) (Table 10). The Adriatic series have higher frequencies of cribra orbitalia than QDR and VCN. In particular, the Ravenna series has the highest values and the differences with respect to QDR and VCN are significant (P < 0.0000). These differences could be explained by a different sociogeographic environment due to the marshland surrounding those Adriatic regions (Uggeri, 1975) and consequently a worse health status, probably also due to the wider diffusion of malaria. Therefore, the different climatic and territorial conditions in the north-eastern Adriatic region, resulting in poorer hygienic-sanitary conditions, probably produced biological differences between those populations and the ones of the Sannio area. CONCLUSIONS The results of our study indicate that there was substantial continuity of the dietary habits and probably of the ways of exploiting the territorial resources (high production and consumption of carbohydrates) in Molise from the I to the VII century. However, there was a higher protein intake (probably due to greater consumption of meat) in the Early Middle Ages, perhaps related to the strong exploitation of sylvo-pastoral resources typical of the Germanic peoples. According to the demographic data and the indicators examined in our study, the general health status of the populations of the Sannio area was precarious and they had poor living conditions already in the Roman Imperial era, and these conditions persisted in the following centuries. A further decline is shown by the presence of infectious diseases like leprosy in the Vicenne population. In the populations of Latium, there seems to have been a clear discontinuity from the biological point of view, with an evident worsening of living conditions in the medieval period. However, the profound social and economic transformations, which were translated into a true cultural discontinuity, during the passage from Antiquity to the Middle Ages do not seem to have affected the biology of the Molise populations under study. The differences with respect to the Roman populations living around Rome, which probably enjoyed good living conditions, show that the transition occurred at different times and in different ways in Italy. American Journal of Physical Anthropology—DOI 10.1002/ajpa ROMAN IMPERIAL AGE-EARLY MIDDLE AGES TRANSITION ACKNOWLEDGMENTS We are extremely grateful to the Editor of the Journal and three anonymous reviewers for the valuable comments and suggestions, which allowed us to improve the paper. LITERATURE CITED Ånerud Å, Löe H, Boysen H. 1991. The natural history and clinical course of calculus formation in man. J Clin Periodontol 18: 160–170. Arslan E. 2000. L’anello del cavaliere e il Duca. La tomba 33 di Campochiaro-Vicenne (CB). Quaderni Ticinesi di Numismatica e Antichità Classiche 29:333–356. Belcastro MG, Bonfiglioli B, Facchini F. 2001a. Unusual wear dental patterns in the horsemen of medieval necropolis of Vicenne-Campochiaro (Molise, Italy). In: Proceedings of the 13th European Meeting of the Paleopathological Association, Chieti, Italy, September 18–21, 2000. p 31–35. Belcastro MG, Bonfiglioli B, Mariotti V. 2003. Il popolamento del territorio di Campochiaro in epoca altomedievale. I dati antropologici della necropoli di Vicenne. In: I Longobardi dei ducati di Spoleto e Benevento. Atti del XVI Congresso internazionale di studi sull’alto medioevo. Spoleto: Fondazione Centro Italiano di Studi sull’Alto Medioevo. p 1009–1030. Belcastro MG, Facchini F. 2001. Anthropological and cultural features of skeletal sample of horsemen from the Medieval necropolis of Vicenne-Campochiaro (Molise, Italy). Coll Antropol 25:387–401. Belcastro MG, Facchini F, Neri R, Mariotti V. 2001b. Skeletal markers of activity in the Early Middle Ages necropolis of Vicenne-Campochiaro (Molise, Italy). J Paleopatol 13:9–20. Belcastro MG, Mariotti V, Facchini F, Bonfiglioli B. 2004. Proposal of a data collection form to record dento-alveolar features—application to two Roman skeletal sample from Italy. Coll Antropol 28:161–177. Belcastro MG, Mariotti V, Facchini F, Dutour O. 2005. Leprosy in a skeleton from the century necropolis of Vicenne-Campochiaro (Molise, Italy). Int J Osteoarchaeol 15:1–18. Bóna I. 1990. I Longobardi in Pannonia. In: I Longobardi. Milano: Electa. p 14–19. Bonfiglioli B, Brasili P, Belcastro MG. 2003. Dento-alveolar lesions and nutritional habitus of a Roman Imperial age population (I–IV c.): Quadrella (Molise, Italy). Homo 54:36–56. Bonfiglioli B, Mariotti V, Facchini F, Belcastro MG, Condemi S. 2004. Masticatory and non-masticatory dental modifications in the epipaleolithic necropolis of Taforalt (Morocco). Int J Osteoarcheaol 14:448–456. Brasili P, Belcastro MG. 1998. La necropoli di Quadrella (Isernia) (I–IV sec. d.C.) e il popolamento dell’Italia Centrale in epoca romana: aspetti paleodemografici. Riv di Antropol (Roma) 76: 171–182. Brothwell D, Brothwell P. 1969. Food in antiquity. London: Thames and Hudson. Brothwell DR. 1981. Digging up bones. Oxford: Oxford University Press. Ceglia V. 2000. Campochiaro (CB). La necropoli di Vicenne. In: Arslan E, Buora M, editors. L’oro degli Avari: Popolo delle steppe in Europa. Milano: Inform Edizioni. p 212–221. Ceglia V, Genito B. 1991. La necropoli alto-medioevale di VicenneCampochiaro. In: Capini S, Di Niro A, editors. Samnium. Archeologia del Molise. Roma: Casa Editrice Quasar. p 229– 234. Clarke NG. 1990. Periodontal defects of pulpal origin: evidence in early man. Am J Phys Anthropol 82:371–376. Comuzzie AG, Steele DG. 1988. Maxillary lingual anterior wear in prehistoric hunter-gathers from Texas coast. Am J Phys Anthropol 75:197. De Benedittis G. 1988. Di alcuni materiali altomedioevali provenienti dal Molise centrale ed il problema topografico della necropoli di Vicenne. In: Atti del convegno: La necropoli di Vicenne 393 nella piana di Bojano—Il Sannio tra tardo impero e altomedioevo. Campobasso: La Rapida Gradedit. p 103–107. Del Panta L, Livi Bacci M, Pinti G, Sonnino E. 1996. La popolazione italiana dal Medioevo ad oggi. Bari: Editori Laterza. Dosi A, Schnell F. 1990. A tavola con i Romani Antichi. Roma: Nuova Editrice Romana. Dupont F. 1997. Grammatica dell’alimentazione e dei pasti romani. In: Flandrin J, Montanari M, editors. Storia dell’alimentazione. Roma-Bari: Editori Laterza. p 145–160. Facchini F, Rastelli E, Brasili P. 2004. Cribra orbitalia and cribra cranii in Roman skeletal remains from Ravenna area and Rimini (I–IV century AD). Int J Osteoarchaeol 14:126–136. Fédération Dentaire International. 1992. A review of the developmental defects of enamel index (DDE Index). Int Dent J 42: 411–426. Fildes VA. 1986. Breast, bottles and babies: A history of infant feeding. Edinburgh: Edinburgh University Press. Genito B. 1991. Tombe con cavallo a Vicenne. In: Capini S, Di Niro A, editors. Samnium. Archeologia del Molise Roma: Quasar. pp 335–338. Goodman AH, Armelagos GJ, Rose JC. 1980. Enamel hypoplasias as indicators of stress in three prehistoric populations from Illinois. Hum Biol 52:515–528. Goodman AH, Rose JC. 1990. Assessment of systemic physiological perturbations from dental enamel hypoplasias and associated histological structures. Yrbk Phys Anthropol 33:59–110. Gould RA. 1968. Chipping stones in the outblack. Nat Hist 77: 42–49. Hillson S. 1996. Dental anthropology. Cambridge: Cambridge University Press. Hillson S. 2001. Recording dental caries in archaeological human remains. Int J Osteoarchaeol 11:249–289. Iannetti G, Maggiore C, Ripari M, Grassi P. 1984. Studio statistico sulle lesioni traumatiche dei denti. Minerva Stomatol 33: 933–943. Irish JD, Turner CG II. 1987. More lingual surfaces attrition of the maxillary anterior teeth in American Indians: Prehistoric Panamanians. Am J Phys Anthropol 73:209–213. Irish JD, Turner CG II. 1997. Brief communication: first evidence of LSAMAT in non-native Americans: historic Senegalese from West Africa. Am J Phys Anthropol 102:141–146. Kerr NW, Bruce MF, Cross JF. 1990. Caries in Mediaeval Scots. Am J Phys Anthropol 83:69–76. Kiple KF. 1999. The Cambridge world history of human disease. New York: Cambridge University Press. Kurylev VP, Pavlinskaya LR, Simakov GN. 1989. Harness and weaponry. In: Basilov VN, editor. Nomads of Eurasia. Los Angeles: Natural History Museum of Los Angeles County in association with Denver Museum of Natural History. p 137–151. Larsen CS. 1997. Bioarchaelogy. Interpreting behaviour from the human skeleton. UK: Cambridge University Press. Larsen CS. 2002. Bioarchaelogy: The lives and lifestyles of past people. J Archaeol Res 10:119–165. Lous I. 1970. The masticatory system used as a tool. Dent Abstr 15:457–458. Lukacs JR. 1989. Dental paleopathology: methods for reconstructing dietary patterns. In: Iscan MY, Kennedy KAR, editors. Reconstruction of life from the skeleton. New York: Alan R. Liss. p 261–286. Lukacs JR. 1992. Dental pathology and agricultural intensification in South Asia: New evidence from Bronze Age Harappa. Am J Phys Anthropol 87:133–150. Lukacs JR. 1995. The \caries collection factor:" a new method of calibrating dental caries rates to compensate for ante mortem loss of teeth. Int J Osteoarchaeol 5:151–156. Maat GJR, Van der Velde EA. 1987. The caries–attrition competition. Int J Anthropol 2:281–292. Manzi G, Salvadei L, Vienna A, Passarello P. 1999. Discontinuity of life conditions at the transition from the Roman Imperial Age to Early Middle Ages: example from Central Italy evaluated by pathological dento-alveolar lesions. Am J Hum Biol 11:327–341. Micheloni P. 1976. Storia dell’odontoiatria. Roma: TEI, Tipografi Editrice Italia. American Journal of Physical Anthropology—DOI 10.1002/ajpa 394 M.G. BELCASTRO ET AL. Milner GR, Larsen CS. 1991. Teeth as artefacts of human behaviour: Intentional mutilation and accidental modification. In: Kelley MA, Larsen CS, editors. Advances in dental anthropology. New York: Wiley-Liss. p 357–378. Molnar S. 1972. Tooth wear and culture: A survey of tooth functions among some prehistoric populations. Curr Anthropol 13: 511–526. Molnar S, Molnar I. 1985. Observations and dental diseases among prehistoric populations of Hungary. Am J Phys Anthropol 67:51–63. Montanari M. 1993. La fame e l’abbondanza. Storia dell’alimentazione in Europa. Roma-Bari: Editori Laterza. Montanari M. 1997. Romani, barbari, cristiani. Agli albori della cultura alimentare europea. In: Flandrin J, Montanari M, editors. Storia dell’alimentazione. Roma-Bari: Editori Laterza. p 213–233. Neri V. 1985. L’alimentazione povera nell’Italia romana. Atti del convegno \L’alimentazione nell’antichità". Parma 2-3 maggio 1985. Parma: Archeoclub di Parma dell’Archeoclub d’Italia. Powell ML. 1985. The analysis of dental wear and caries for dietary reconstruction. In: Gilbert RI, Mielke JH, editors. The analysis of prehistoric diets. Orlando: Academic Press. p 307– 338. Reid DJ, Dean MC. 2000. Brief communication: the timing of linear hypoplasias on human anterior teeth. Am J Phys Anthropol 113:135–139. Righini V. 1976. Profilo di storia economica in \Storia dell’Emilia-Romagna". Bologna: Bologna University Press. Robinson JT. 1963. Australopithecines, culture and phylogeny. Am J Phys Anthropol 21:598. Rösener W. 1996. Peasants in the middle ages. Cambridge: Polity Press. Salvadei L, Ricci F, Manzi G. 2001. Porotic hyperostosis as a marker of health and nutritional conditions during childood: studies at the transition between Imperial Rome and the Early Middle Ages. Am J Hum Biol 13:709–717. Sarnat BG, Schour I. 1941. Enamel hypoplasia (chronic enamel hypoplasia) in relationship to systemic diseases: a chronological, morphological and etiological classification. J Am Dent Assoc 28:1989–2000. Scheie AA. 1989. The role of plaque in dental calculus formation: a review. In: ten Cate JM, editor. Recent advances in the study of dental calculus. Oxford: IRL Press. p 47–55. Silva AI. 2000. Cividale del Friuli. Necropoli di San Mauro. Tomba n. 43 di cavallo e cavaliere. In: Arslan E, Buora M, editors. L’oro degli Avari: popolo delle steppe in Europa. Milano: Inform Edizioni. p 198–205. Smith BH. 1984. Patterns of molar wear in hunter-gatherers and agriculturalists. Am J Phys Anthropol 63:39–56. Solomons NW, Keusch GT. 1981. Nutritional implications of parasitic infections. Nutr Rev 39:149–161. Stuart-Macadam P. 1982. A correlative study of paleopathology of the skull, Ph.D. thesis, Department of Physical Anthropology, University of Cambridge. Terzani C. 1991. Aesernia. In: Capini S, Di Niro A, editors. Samnium, archeologia del Molise. Roma: Casa Editrice Quasar. p 225–228. Terzani C, Matteini Chiari M. 1997. Isernia. In: Terzani C, Matteini Chiari M, editors. Isernia. La Necropoli Romana in Località Quadrella. Roma: Gangemi Editore. p 13–15. Turner CG II, Cadien JD. 1969. Dental chipping in Aleuts, Eskimos and Indians. Am J Phys Anthropol 31:303–310. Turner CG II, Cheuiche Machado LM. 1983. A new dental wear pattern and evidence for high carbohydrate consumption in a Brazilian archaic skeletal population. Am J Phys Anthropol 61:125–130. Uggeri G. 1975. La Romanizzazione dell’antico delta padano. Ferrara: Atti e Memorie SATE. White KD. 1970. Roman farming. London: Thames and Hudson. American Journal of Physical Anthropology—DOI 10.1002/ajpa