PALEOBIOS ,13 / 2004 / Lyon-France ISSN 0294-12 / Virgin Europe : periosteal reaction prior to the 15 th century and the potential influence of slavery
Virgin Europe : periosteal reaction prior to the 15 th century and the potential influence of slavery
Bruce M. Rothschild (1), Frank J.Ruhli (2), Christine Rothschild (3), Virginia Naples (4),
Israel Hershkovitz (5), Jeno Sebes (6) & Michel Billard (7)
(1): Professor of Medicine at Northeastern Ohio Universities College of Medicine (NEOUCOM);Adjunct Professor of Biomedical Engineering, University of Akron; Research Associate at the Carnegie and University of Kansas Museums of Natural History and Director of The Arthritis Center of Northeast Ohio, Youngstown, OH 44512
(2): Assistant at Institut of Anatomy, research fellow at Institute for the History of Medicine, University of Zurich, Winterthurstr.190, 8057 Zurich, Switzerland
(3): Arthritis Center of Northeast Ohio, 5500 Market, Youngstown, OH 44512 , USA.
(4): Northern Illinois University, De Kalb, IL 60115, USA.
(5): Sackler School of Medicine, Tel Aviv University, Ramat Aviv 69978, Israel.
(6): University of Tennessee, Memphis, TN 38119, USA.
(7): Research fellow at Laboratory of Anatomical Anthropology and Paleopathology, University of Lyon, Rockefeller Av.8, 69373 Lyon, France.
Reprint requests to: B. M. Rothschild, Arthritis center of Northeast Ohio, 5500 Market Street, Suite 119, Youngstown, Ohio 44512, USA. Tel: 330-783-5900; Fax: 330-783-5350; e-mail : email@example.com
Running title : Virgin Europe periostitis
Periosteal reaction has been variably considered as a non-specific sign of trauma and alternatively as having potentially diagnostic implications. Examination of first millennium AD European cemeteries falsifies the non-specific trauma hypothesis, while examination of variation in the early portion of the second millennium AD reveals a new (at least to Europe) population phenomenon. In contrast to isolated bumps and osteomyelitis, the study of periosteal reaction in early Europe is the study of 'virgins.' It is only in the second millennium AD that periosteal reaction becomes widespread, both in popu lation penetrance and in extent of skeleton affected. That pattern, previously documented for the treponematosis yaws, is synchronic with new slave sources. Is the 'corruption' of virgin Europe a product of slave seeking behavior?
Key words : Middle Ages, treponematoses, yaws, slavery, periosteal reaction
Titre : La réaction périostée en Europe avant le XV° et l'influence possible de l'esclavage
L'examen des squelettes de nécropoles européennes du premier millénaire rejette l'hypothèse de non spécificité traumatique de la réaction périostée générale : en dehors des lésions d'ostéomyélite et de choc traumatiques isolés, les périodes anciennes en Europe sont "vierges" de toute réaction périostée généralisée, tandis qu'un phénomène nouveau apparaît (tout au moins en Europe) seulement au début du second millénaire avec l'extension de la réaction périostée parmi les populations et sur le squelette lui-même. Ce modèle de réaction périostée, préalablement documenté comme le pian, est synchrone de nouvelles sources d'approvisionnent d'esclaves: ce commerce serait-il à l'origine de la "corruption" de la virginité européenne?
Mots clés : Moyen Age, tréponematoses, pian, esclavage, réaction périostée
1 - INTRODUCTION
The suggestion that periosteal reaction has specificity as a population phenomenon appears well supported, although it is usually not diagnostic in a single individual or an isolated bone (Rothschild and Martin, 1993; Rothschild and Rothschild, 1996, 1998a). The alternative interpretation has been that periosteal reaction is predominantly a trauma-related event (Cook, 1976, 1984). While isolated bumps can occur secondary to trauma (Resnick, 2002; Rothschild and Martin, 1993), there has been no documentation that diffuse periosteal reaction, even of a single bone, can be caused by trauma - in the absence of complicating osteomyelitis or stress fracture (Resnick, 2002; Rothschild and Martin, 1993). If trauma truly were a source of periosteal reaction, one would expect its frequent notation in Europe, especially in the Middle Ages. European skeletal populations were therefore examined for evidence of periosteal reaction. Further, it was desired to assess the timing of any subsequent transition in presence or character of periosteal reaction. Potential relationship to North African populations and to African 'cultural developments' was also explored.
2 - MATERIALS AND METHODS
The European sites delineated in table 1 were chosen to assess the population frequency and nature, extent and character of any non-focal (e.g., bump) periosteal reaction present. Similar analysis was subsequently performed on North African sites in Mali and Mauritania (Table 1).
Skeletal remains were subjected to visual examination of all articular and cortical surfaces - to identify all occurrences of alterations throughout each skeleton, specify the types of bony alterations at each occurrence, and map the distribution of occurrences in each skeleton.
* : Pattern characteristic of hypertrophic osteoarthropathy
3 - RESULTS
Examination of 1476 individuals from 8000 to 800 years before present in Europe revealed evidence of isolated bumps and occasional cases of osteomyelitis, but only a single case of diffuse periosteal reaction (Table 1). The latter from Gandaillat, dated at 2500 ybp, affected tibiae, fibulae, femora, ulnae, radii, metacarpals, metatarsals and ribs in the classic distal diaphyseal pattern of hypertrophic osteoarthropathy (Rothschild and Rothschild, 1998b). While bone thickening and cloaca were noted in isolated bones with osteomyelitis, no sabre shin alterations were observed in almost 1200 skeletons.
The only studied site with diffuse periosteal reaction was Winchester in England (Tables 1,2). The osseous distribution of treponemal disease in that site was indistinguishable between individual and ossuary burials from that site, allowing that data to be combined. As bilaterality and pauci-ostotic/poly-ostotic distribution cannot be assessed in ossuaries, that portion of the analysis was performed only on actual individual burials (from that site). Eight of 34 individuals had periosteal reaction. Two had sabre shin deformity, with prominent surface periosteal reaction. Tibial involvement was invariable bilateral. Disease was predominantly poly-ostotic (>3 bone groups affected). Hand and foot involvement was common (25%). Twenty percent of juveniles were affected.
Examination of 91 individuals from 8000 to 3000 years before present in North Africa revealed no periosteal reaction (Table 1).
4 - DISCUSSION
4.1 - Virgin State
Absence of periosteal reaction (unassociated with osteomyelitis and exclusive of isolated bumps) from pre-15th century Europe falsifies the hypothesis that generalized periosteal reaction is trauma-related. It does document a 'virginity' for Europe, with respect to entities that commonly cause periosteal reaction. It has been suggested that periosteal reaction occurring as a population phenomenon (e.g., more 1% of population with diffuse involvement of more than one bone) has only a very limited number of etiologies: Treponematoses and hypertrophic osteoarthropathy (Resnick, 2002; Rothschild and Martin, 1993). While hypertrophic osteoarthropathy (as a complication predominantly of intrathoracic disease) is actually quite rare in unselected populations (Resnick, 2002; Rothschild and Rothschild, 1998b) and as noted in the populations studied herein, treponemal disease has a population ‘signature’ if it is present. That population signature is in the form of periosteal reaction, affecting 2-13% of skeletons with syphilis and 20-40% of skeletons with yaws or bejel. Absence of periosteal reaction in 1194 individuals precludes presence of treponemal disease (Chi square <0.0001).
4.2 - New Disease
This changed, at least in England, somewhere between 500 and 800 years ago. What happened? A new pattern presented, a pattern indistinguishable from what is seen in yaws (Table 2). Recognition of bone involvement in 1/3 of the skeletal population suggests that essentially the entire population was afflicted, a phenomenon typical for yaws (Helfet, 1944; Hudson, 1958; Hunt and Johnson, 1923; Moss and Bigelow, 1922). This disease is easily distinguished from the more pauci-ostotic syphilis (Chi square = 3.973, p<0.05), in which hand and foot and subadult affliction are so rarely observed in skeletal populations (Rothschild and Rothschild, 1994, 1995a,b); Rothschild et al., 1995a). Other evidence for syphilis (in the form of complete sabre shin surface remodeling and unilateral tibial disease) was also lacking. Disease in the Winchester cemetery was also easily distinguished from the more pauci-ostotic bejel, which infrequently affects hands and feet (Hershkovitz et al., 1995; Rothschild and Rothschild, 1995b).
Interpretation of the historical record also suggested that the non-venereal treponemal disease - yaws - was present in the British Isles (Cockburn, 1995; Spirov, 1991). The disease sibbens in Scotland appears to represent what is today recognized as yaws.
How does this contrast with previous speculation? The observations in 13th century England confirm those of Goodman (Goodman et al., 1988) that periosteal reaction is very common in such pre-Columbian skeletal populations. Their interpretation (that this is simply an indicator of stress) is clearly rebutted by comparisons with English populations. Stirland (1995) also reported poly-ostotic disease in the St. Margaret Fyebridgegate cemetery (usage 1254-1468). Her description (in selected skeletons) fits that of yaws and appears indistinguishable from the findings reported above for the same time period. While the reports by Hurley et al (1994) and Power (1992) of a 14th century (period III) individual aged 12-13 suggested syphilis, examination of the actual skeleton revealed the changes unlike those of treponemal disease, but actually representative of histiocytosis.
4.3 - Differential Diagnosis
Non-treponemal disorders associated with periosteal reaction do not occur with sufficient population frequency (Resnick, 2002; Rothschild, 1982; McCarty and Koopman, 1993) to merit serious consideration. Further, they do not cause sabre shin reaction and have very different patterns of skeletal affliction. Hypertrophic osteoarthropathy is predominantly a disease of distal diaphyses (Resnick, 2002; Rothschild, 1982; Rothschild and Martin, 1993). It typically spares the epiphyses and often affects ribs and clavicles. Thyroid acropachy spares the proximal appendicular skeleton, predominantly producing hand and foot bone periosteal reaction (Resnick, 2002; Rothschild, 1982; Rothschild and Yoon, 1982). Infantile cortical hyperostosis is a disorder afflicting clavicles, scapulae, and ribs (Resnick, 2002). Hypervitaminosis A is predominantly an enthesial disease, and fluorosis produces highly characteristic trabecular alterations (Resnick, 2002; Rothschild and Martin, 1993; Seawright and English, 1967).
4.4 - Why
If a new disease was introduced, why did this happen and why at this time? Most earlier European interactions were with Northern Africa. North Africa was free of treponemal disease from 3000-7000 ybp in our skeletal examinations. Further, Ruffer (1921) reported absence of treponemal disease in 25,000 Egyptian skeletons that he examined . Interaction with Northern Africa would therefore seem an unlikely source of the new disease. What about the Middle East? What was the impact of the Crusades? Timing again is critical. There appears to have been no introduction of a new disease (e.g, treponematosis) by the Crusades. Absence of treponemal disease, until at least the 12th century in England, presents a difficult to justify time lapse (if the Middle East were a source). However, the pattern of periosteal reaction recognized in the Middle East is also different from that noted in England. The Middle East pattern of pauciostotic disease (averaging 2 bones affected) and sparing of hands and feet (bejel pattern) (Hershkovitz et al, 1995) is easily distinguished from the more polyostotic pattern observed in England. Even if one ignores the timing, the pattern of bone involvement is incompatible with the disease found in the Middle East.
The timing is perhaps the critical question. Could occurrence of the new disease in England be related to the Islamic slave trade? Slavery practices, permitted by the Holy Quran, defined in the Hadith (traditions of the Prophet Mohammed’s lifetime) and codified in the Hadith or sacred law codes of the Ulama (Alexander 2001) appear similar to that followed by the Romans - with a significant exception: Islam had great tolerance for Christianity and Judaism. Muslims were apparently under obligation to expand their territory and their 'congregation.' Non-Muslims apparently had the option of what was called 'voluntary' conversion or slavery (Alexander, 2001). Muslim influence first expanded in Africa in the 7th century. While much of North Africa professed Christianity (Alexander, 2001), desert-dwellers, predominantly from Egypt, apparently remained animist. Saharan Africa also provided similar slave-taking opportunities. Thus, whatever slaves were taken in that time period were from areas which (on the basis of absence of periosteal reaction in their skeletons) apparently lacked treponemal disease.
It was only in the 12th and 13th centuries that the Islamic slave trade expanded into the West African Forest Zone and into Somalia, Tanzania and Kenya (Alexander, 2001), where yaws apparently had its origins (Rothschild et al 1995b). Notation of absent 'land bridge' between north and east Africa (Alexander, 2001) supports the hypothesis that the catchment area for yaws had been segregated from the source of slaves prior to the 12th century. The timing matches with introduction of the new disease into England. It also matches with the timing of social changes in Europe. Slavery had long been part of European History (Taylor, 2001), the very name derived from 'Slav', the... resource population for Viking-Arab trade (Taylor, 2001, p. 35). What social factors altered local availability of trade and apparent greater dependence upon slaves drawn from yaws-endemic areas?
Coincidence? It will be intriguing to test this hypothesis by examination of more recent sites in Europe. Absence in 13th century France, Hungary and Switzerland may reflect differential spread or simply the wide time interval represented by the Winchester site, the latter precluding precise timing of the introduction event.
Appreciation is expressed to Drs. Bardas Chorcai, Cristina Cattaneo, Alessandra Mazzucchi, Olivier Dutour, J.L. Heim, Bruno Kaufmann, Ildiko Pap, Catryn Power, T Molleson, Michael Schultz and the Instituto di Antropologia dell'Universita di Firenze, Italy; British Museum of Natural History, London; Cork Corporation, University College, Ireland; Hungarian National Museum, Budapest; University of Gottingen, Germany; University of Milano, Italy; Université de la Mediterranee, Aix en Provence, France; Anthropologisches Forschungsinstitut, Aesch, Switzerland; Musee de l'homme, Paris, France and Museum Testut Latarjet de Medecine et d’Anatomie de Lyon, France.