PATHOLOGICAL CRANIAL LESIONS IN A JUVENILE CRANIAL COLLECTION
A Thesis submitted to the faculty of San Francisco State University in partial fulfillment of A - the Requirements for -i r the degree M2 Master of Arts In . IsA 53 Anthropology
by Hannah Marie Miller San Francisco, California January 2018 Copyright By
Hannah Marie Miller
2018 CERTIFICATION OF APPROVAL
I certify that I have read Pathological Cranial Lesions in a Juvenile Cranial Collection by Hannah Marie Miller, and that in my opinion this work meets the criteria for approving a thesis submitted in partial fulfillment of the requests for the degree: Master of Arts in Anthropology at San Francisco State University.
Cymhia Wilczak Associate Professor of Anthropology
Associate Professor of Anthropology PATHOLOGICAL CRANIAL LESIONS IN A JUVENILE CRANIAL COLLECTION
Hannah Marie Miller San Francisco, California 2018 The purpose of this study is to examine and analyze the age distributions and co-occurrence of endocranial and ectocranial lesions commonly associated with metabolic disease and inflammatory processes in juveniles. The pathogenic changes studied are porotic hyperostosis, cribra orbitalia and endocranial lesions. The specific etiology of any of these lesions remains uncertain. (Lewis 2004, Janovic et al. 2012, Walker et al. 2009, Wilczak and Zimova Hopkins 2010). While porotic hyperostosis and cribra orbitalia have been subject to multiple studies, to date no one has statistically confirmed a relationship between these lesions. Endocranial lesions have never been systematically studied in conjunction with porotic hyperostosis or cribra orbitalia. As Wilczak and Zimova Hopkins (2010) suggested that not all lesions classified as cribra orbitalia shared a common etiology, classification of porotic hyperostosis and cribra orbitalia were modified for analysis here. Endocranial lesions were classified according to type (Hershkovitz et al. 2002 and Lewis 2004). The co-occurrence of each type of documented lesion was examined with chi-square analysis using the Monte Carlo method to estimate p-values. Analysis found that while there is a statistically significant association between lesions commonly referred to as endocranial lesions and porotic hyperostosis (p= 0.03), there was not an association between porotic hyperostosis and cribra orbitalia (p= 0.2). This finding challenges two ideas that permeate the anthropological literature. First is the idea that there is not an association between endocranial lesions, porotic hyperostosis and cribra orbitalia. This study indicates that there is in fact an association between porotic hyperostosis and endocranial lesions. The results also suggest that not all lesions classified as porotic hyperostosis and cribra orbitalia have the same etiology.
I certify that the Abstract is a correct representation of the content of this thesis. ACKNOWLEDGMENTS
I would like to begin by thanking my thesis committee Cynthia Wilczak Ph.D. and Mark Griffin Ph.D who have guided me through this process. You both have helped refine ideas, re-work projects and have generally acted as sounding boards for the past few years. Having never failed to challenge and test me you have both made me a much better researcher and anthropologist. I would also like to thank Dawn Mulhern Ph.D., for introducing me to Biological Anthropology and beginning the process to make me an independent researcher. Thank you to Dorothy Dechant Ph.D. and Gary Richards Ph.D. who were absolutely invaluable to me during the data collection phase. To the staff of the SFSU NAGPRA program Jeff Fentress Ph.D. and Kathy Wallace, thank you for giving me the opportunity to work with you for the last few years and helping me to gain experience in an area which I still find immensely fascinating. Amanda Price and Chelsea Jordan, I don’t think I could have done this without you two there to talk me off the edge when I started to panic. Most of all I would like to thank my family who has never failed to support me throughout this process.
DEDICATION For George Raymond Betz You were the one who first introduced me to Anthropology, and inspired my love of learning about others. You took the time to enquire about my classes and my research, even when the topic was not of particular interest to you. I will never forget sitting at your bedside and talking about the research for this project. Even at your most difficult time you were enthusiastic about this project, just because I was. I am only sorry you never had the opportunity to see the final results.
v TABLE OF CONTENTS
List of Table...... v
List o f F igures...... vii
List of Appendices...... viii
Introduction...... 1
Literature Review...... 8
Porotic Hyperostosis and Cribra Orbitalia...... 8
The Early Works...... 9
Introduction of Anemia...... 13
The Synergistic Approach...... 16
Maize Dependency Models...... 20
Revisiting the Early Works...... 23
A Positive Adaptive Response...... 28
Challenging Positive Adaptation...... 29
Scurvy as a Cause of Porous Cranial Lesions...... 30
Porotic Hyperostosis as a non-specific indicator of stress...... 32
The Iron Deficiency Anemia Etiology Challenged...... 34
Co-occurrence of Porotic hyperostosis and Cribra orbitalia...... 36
Historical Scoring Methods...... 37
Endocranial Lesions...... 38
Early Works...... 38 SES and Intrathorasic Disease...... 39
Nonspecific endocranial lesions...... 41
Endocranial lesions and Scurvy...... 43
Endocranial lesions and Trauma...... 44
Endocranial lesions and Infectious disease...... 45
M aterials...... 47
M ethods...... 49
Orbital lesions AKA Cribra Orbitalia...... 49
Cranial Vault Lesions AKA Porotic Hyperostosis...... 51
Endocranial lesions...... 54
R esults...... 57
Porotic Hyperostosis and Cribra Orbitalia...... 57
Orbital Lesions and Endocranial Lesions...... 62
Cranial Vault Lesions and Endocranial Lesions...... 63
D iscussion...... 66
Conclusion...... 70
Literature Cited ...... 74
vii LIST OF TABLES
Table Page
1. Age Groupings by major developmental period ...... 48
2. Number of individuals sampled for each lesion type...... 48
3. Description, Classification, and Codes for orbital les...... 50
4. Description, classification and codes for lesions on the cranial vault...... 53
5. Description and codes of endocranial lesions...... 55
6. Distribution of orbital lesions by age range...... 59
7. Distribution of cranial vault lesions by lesion age...... 60
8. Cross tabulation of orbital and cranial lesions for chi square analysis 61
9. Frequency of endocranial lesions by age range...... 61
10. Cross tabulation of orbital and endocranial lesions for chi square analysis...... 63
11. Cross tabulation of endocranial and cranial vault lesions for Chi-square analysis...... 64
viii LIST OF FIGURES
Figures Page
1. Orbital lesions A) Porosity with vascular lesions B) Porosity only C) Hyperostotic Cribra Orbitalia...... 50 2. Cranial Vault lesions A) Porosity only B) Porosity with vascular channels C) Porosity with significant coalescence...... 53
IX APPENDICES
A ppendix
1. Appendix A 1
CHAPTER 1 - INTRODUCTION
Paleopathology was defined by Sir Marc Armand Ruffer (Aufderheide and
Rodriguez- Martin 1998) as the study of diseases whose existence can be demonstrated on the basis of skeletal remains from ancient times. As a subdiscipline of biological anthropology, paleopathology studies abnormal variation in human remains as well as
secondary evidence of disease in ancient times. Use of secondary evidence is vital because diseases that do not affect skeletal remains may only be visible in documents and art from the period. Paleopathology examines not only disease, but also how humanity has adapted to changes in their environment at both an individual and a population level.
Alternatively, by studying the secondary records we gain a greater understanding of how people within a specific cultural framework understood disease. What follows here is a methodological study comparing 3 types of cranial lesions, which may share a common etiology according to the anthropological literature.
Lesions of the endocranium, porotic hyperostosis and cribra orbitalia, are among the most commonly documented pathological changes in the archaeological record (Hill and Armelagos 1990). In contrast, lesions occurring on the endocranial surface have received relatively little recognition in the anthropological literature. Two major methodological studies comprise the bulk of research on endocranial lesions (Hershkovitz et al. 2002, Lewis 2004). However the anthropological literature suggests that there may be common etiologies for lesions occurring on the endocranium and ectocranium, due to 2
the similarities in their differential diagnosis. Early anthropological works suggested that
lesions on both the endocranium and ectocranium could share a common etiology
(Koganei 1912), an idea that does not occur frequently in the modern literature. The goal
of this study was to document the age distribution and understand the relationship
between lesions occurring on the orbital roof, cranial vault and endocranial surfaces in a juvenile cranial collection.
Lesions on the ectocranial surface have been subject to study for over 100 years
(Hill and Armelagos 1990, Owen 1859). The most commonly documented ectocranial
lesions are porotic hyperostosis and cribra orbitalia. These lesions manifest as porotic
lesions on the cranial vault and orbital roof. The earliest publications sought to
understand the etiology of porotic lesions occurring on the orbital roof and cranial vault
(Hrdlicka 1914, More 1929, Owen 1859, Parrot 1879 cited in Hill and Armelagos 1990,
Virchow 1874, Williams 1929). These lesions were generally considered to be caused by
iron deficiency anemia, though other types of anemias have been proposed (Angel 1964,
1966, 1967) The link between anemia, porotic hyperostosis and cribra orbitalia persisted
in the anthropological until 2009 (Walker et al. 2009), when it was suggested that there
were likely many different potential causes of these lesions. It has also been suggested
that the different types of porotic lesions classified as cribra orbitalia may in fact have
different etiologies (Wapler et al. 2004, Wilczak and Zimova Hopkins 2010). 3
Endocranial lesions were first documented in 1912 by Koganei, who linked these
lesions with those occurring on the ectocranial surface. This idea persisted until 1978
(Mensforth et al. 1978) when it was suggested that lesions of the endocranium and ectocranium did not have a common cause. Research on endocranial lesions then shifted to determining potential causes of lesions occurring on the endocranial surface
(Hershkovitz et al. 2002, Mensforth et al. 1978, Lewis 2004, Patterson 1992, Schultz
1989, 2001).
Analysis of the potential causes of porotic hyperostosis, cribra orbitalia and endocranial lesions based on the anthropological literature suggests that the some of the
same disease processes may be responsible for all three types of lesions (Hershkovitz et
al. 2002, Lewis 2004, and Walker et al. 2009). Conditions resulting in inflammatory
reactions in or near the bone as well as any condition resulting in hemorrhage are
postulated as potential causes of lesions of the orbital roof, cranial vault and endocranium
(Hershkovitz 2002, Lewis 2004, Schultz 2001, Walker 2008). The literature on
endocranial lesions generally assumes that they do not have a common etiology with
porotic hyperostosis and cribra orbitalia. However, if all three types of lesions could be
cause by the same types of disease processes and occur in similar areas, why do we not
consider it possible that these three lesions are linked in some cases?
As previously stated, lesions on the endocranium and ectocranium have been
postulated to have the same etiologies. However the majority of the recent 4
anthropological literature indicates that these lesions are not linked. The question that drove this research is if lesions on the orbital roof, cranial vault and endocranium have the same potential causes; would there be a statistical relationship in the co-occurrence of these lesions? Because of the relative dearth of information on endocranial lesions within the anthropological and clinical literature, it is impossible to answer this question based on analysis of the current literature. Therefore it was necessary to conduct a methodological study on the co-occurrence of cribra orbitalia, porotic hyperostosis and endocranial lesions.
Data on all three lesion types was collected from the Spencer R. Atkinson Library of Applied Anatomy at the University of the Pacific School of Dentistry in San Francisco
California. All juvenile crania present in the collection at the time of sampling were analyzed for the presence and severity of the lesions. Data analysis was performed to document the age distributions of lesions and to determine if there is a significant relationship between lesions of the orbital roof, cranial vault and endocranium. The age of these individuals was previously determined based on dental calcification (Richards
2007). The age ranges considered for analysis were based on the ranges used by Lewis
(2004). Age ranges used in statistical analysis are as follows; 0.6-6.5 years, 6.6-10.5 years, 10.6-14.5 years, 14.6-17.0 years and 17.1 - 20 years. Data collection procedures for each lesion type are outlined below. 5
Orbital lesions. In the data collection phase each orbital roof was observed and scored separately. Macroscopically visible changes were documented and scored according to the following scoring guide: 0) no lesion visible, 1) porosity only, 2) porosity with vascular channels, 3) porosity with significant pore coalescence but no evidence of diploic hypertrophy or 4) porosity with significant pore coalescence and diploic expansion. During data analysis these categories were further refined. Both categories, which were characterized with significant coalescence of pores were combined into a new variable which was called hyperostotic cribra orbitalia. Then the scores of both orbits were compared and the most severe manifestation of the orbital lesion was assigned for that individual preceding statistical analysis.
Ectocranial lesions. In the original scoring of ectocranial lesions the methods were adapted from the scoring methodology outlined by Wilczak and Zimova Hopkins (2010) for orbital lesions. Adapting the scoring methods Wilczak and Zimova Hopkins (2010) developed for lesions of the orbital roof allows for direct comparisons of lesions of the orbital and cranial vault lesions. Each bone of the cranial vault was examined and scored according to the following criteria: 0) No lesion, 1) Porosity only, 2) Porosity with vascular channels, 3) porosity with significant coalescence but no evidence of diploic expansion or 4) porosity with significant pore coalescence and diploic expansion.
Preceding statistical analysis these categories were further refined. Lesions in categories
3 and 4 were combined into a new variable, which was classified as porotic hyperostosis. 6
The lesion on the entire endocranial surface with the highest assigned value was then assigned for statistical analysis in that individual.
Endocranial lesions. Data collection for the endocranial lesion was dependent on the condition of each cranium. In cases of an intact cranium, a flexible LED penlight was inserted through the foramen magnum. The endocranial surface was observed through the optic chiasm and the foramen magnum. In cases where the cranial vault had been separated, the endocranial surface was examined by removing the superior portion and observing the endocranium directly. Lesions were scored as: 0) no lesion, 1) porosity only, 2) porous or vascular woven bone deposition, 3) vascular lesions, 4) erosions or 5)
SES (Lewis 2004, Hershkovitz et al. 2002). The lesion with the highest assigned value was then considered in statistical testing.
Within the sample I compared the co-occurrence of lesions and difference in the distributions of each lesion type by age range. Chi square analysis was used to determine if there was statistical significance, with Monte Carlo simulations to estimate p-values due to the uneven distribution of lesions in this sample.
This study was undertaken because comparing lesions with uncertain etiologies for patterns of co-occurrence will help reveal potential links, which may better define the diagnostic criteria and possible etiologies of these lesions. Two main research hypotheses were tested here. First lesions with similar etiologies should have a similar age 7
distributions. Second co-occurrence of similar lesions should occur across all three cranial locations.
The greatest limitations of this study originate with the sample itself. As there is little provenience information on this sample, it is difficult to determine what types of conditions each individual in the sample was susceptible to. This is because the most information available on an individual was their nationality. As children of difference nationalities and socioeconomic status are exposed to different conditions, the lack of provenience information makes it more difficult to determine what types of disease process each individual may have been susceptible to. Additionally there was varied preservation of the individuals, which limited the sample size. 8
CHAPTER 2 - LITERATURE REVIEW
Analysis of the literature on endocranial lesions, porotic hyperostosis and cribra orbitalia reveals that many of the same etiologies have been proposed for these lesions.
That being said, there has been almost no research to determine whether or not these pathological lesions are associated with one another. What follows is a brief review of the anthropological literature. As much of the literature uses the term porotic hyperostosis to refer to lesions more accurately called cribra orbitalia, sometimes not even distinguishing between the two lesion types, the conditions are combined into one section. This section is arranged chronologically and by the major themes of each time period. Following that review is a section on endocranial lesions. Due to the dearth of published studies, this section is arranged chronologically and then by possible etiology.
Porotic Hyperostosis and Cribra Orbitalia
Porotic hyperostosis and cribra orbitalia are among the most common pathologies documented in skeletal remains. Their presence is often scored together because much of the published research assumes a common etiology for the two lesions. Porotic hyperostosis is generally viewed as a porotic reaction on the cranial vault, most frequently affecting the parietals and the occipital. Cribra orbitalia is commonly seen as porotic lesions affecting the anteriosuperior orbital roof although similar lesions may affect any area of the orbit. 9
Both types of lesions are typically documented as active lesions in juveniles, while they are usually partially remodeled in adults (Stuart-Macadam 1985, 1992a,
1992b, 1995). This is typically understood to be the result of age-related changes. In juveniles red blood cell production is centered in the cranial bones. In adults the production of red blood cells shifts to the hematopoietic marrow centers of the axial skeleton (Stuart-Macadam 1992c). This shift in the hematopoietic centers has led to the assumption that both cribra orbital and porotic hyperostosis are representative of conditions occurring in childhood, while the lesions occurring in adults are an artifact of the childhood condition (Stuart-Macadam 1992a, 1992b). It should be recognized that the term porotic hyperostosis is sometimes used to refer to lesions occurring on the orbital roof. For the purpose of this study I instead use the terms orbital roof lesions and cranial vault lesions when referring to cribra orbitalia and porotic hyperostosis respectively to clearly distinguish them and to avoid any connotations about the etiology or morphological expression of the lesions. The terms used in the following section are based on the terminology utilized by the authors of the literature.
The Early Works
The earliest reference to lesions consistent with porotic hyperostosis or cribra orbitalia date back to the 19th century. Welcker (1888, Cited in Hill and Armelagos 1990) is generally considered to be the first researcher to study cribra orbitalia, primarily due to his coining the term although others had described the condition (Nathan and Hass 1966). 10
Welcker incorrectly attributed cribra orbitalia to population differences and considered it
a racial marker. However he did make several notes on the lesions that are still
considered canon today. Among them are the symmetrical nature of the lesions and the
similar lesions that occur on the cranial vault (Hill and Armelagos 1990).
Publications on porotic hyperostosis occur even earlier with publications in 1874
(Virchow 1874), 1879 (Parrot 1879, cited in Hill and Armelagos 1990) and 1859 (Owen
1859). Virchow (1874) was the first to assume a common etiology for both porotic hyperostotic vault lesions and lesions consistent with cribra orbitalia. Parrot (1879, cited
in Hill and Armelagos 1990) was the first to offer a possible cause of these lesions, though he incorrectly linked them with syphilis. However it was Welker’s (1888) work, which would influence researchers for the next two decades.
Subsequent studies on the remains of ancient Egyptians (Oetteking 1909), Ainu
(Koganei 1912), African (Toldt 1886), and Japanese (Adachi 1904a, 1904b) populations all accepted Welker’s (1888) conclusion that cribra orbitalia and porotic hyperostosis were indeed racial traits. Koganei (1912) also proposed that inflammation of the periosteum due to infection or compression of the lacrimal gland by neoplasms could be the cause of these lesions.
Hrdlicka’s 1914 study documented porotic hyperostosis in ancient Peruvians.
Hrdlicka used the term symmetric osteoporosis at the time, but the lesions that he describes clearly share all the same characteristics as porotic hyperostosis. These lesions 11
were typically documented in the remains of infants and young children from different populations, which is consistent with the manifestation of porotic hyperostosis. Many of
Welkler’s (1888, Cited in Hill and Armelagos 1990) observations were corroborated by
Hrdlicka. However Hrdlicka (1914) made several observations that have influenced subsequent studies. The lesions Hrdlicka described were limited to external surfaces of the bone, avoided points of muscle attachment and sutures, Hrdlicka was the first to suggest that these lesions were formed in infancy or during childhood. Many of these observations are part of the modern diagnostic criteria for porotic hyperostosis and cribra orbitalia. Hrdlicka incorrectly attributed that these lesions were most likely the result of systemic toxic disorder.
The first studies to actively search for an etiology for porotic hyperostosis were published in 1929 (Moore 1929; Williams 1929). Moore compared the crania of ancient
Mayan individuals exhibiting hyperostosis to the radiograph of a modern individual diagnosed with sickle cell anemia. Radiographic analysis of both the modern and ancient crania presented an expansion of the diploe, hair on end trabeculae, and the obliteration of the ectocranium with a largely unaffected endocranium. The similarities in the radiographs of the ancient and modem crania led the hypothesis that sickle cell anemia, or a similar disease, was potentially the cause of the cranial changes (Moore 1929).
Williams (1929) compiled all the reported cases of porotic hyperostosis as well as his own data on the lesions from his work with Native American remains. Using his own 12
data, Williams concluded that porotic hyperostosis occurred most frequently on the
frontal, parietal and occipital bones. These findings re-affirmed the assertions made by
both Hrdlicka and Welcker. In comparing radiographs of several archaeological
specimens to modern clinical cases, Williams noted that the lesions were similar to clinical cases of sickle cell and von Jaksch’s anemia, rickets and scurvy. Bone sectioning was used to compare the porotic and hyperostotic lesions with those from individuals with rickets. Rickets was eliminated as a potential cause as it was believed that Native
American populations would have had sufficient exposure to sunlight to prevent the onset of rickets, even though sectioning revealed similarities between the porotic hyperostotic
lesions and those associated with rickets (Williams 1929). Williams (1929) was also the
first researcher to propose that cranial deformation could play a role in the development of the lesions, though it was postulated that these changes would only occur in conjunction with a disease process. He suggested that the evidence supported metabolic conditions as the most likely disease type, with anemias being the most likely. Williams
(1929) was also the first to suggest a relationship between the domestication and farming of maize and porotic hyperostosis. This idea would later be refined to link the development of iron deficiency anemia with a shift in subsistence patterns, which would
later become the predominate explanation for cribra orbitalia and porotic hyperostosis. 13
Introduction of Anemia
The next major research trend in porotic hyperostosis and cribra orbitalia began in the 1960’s. In 1961 Henschen set out to disprove Welcker’s assertion that porotic hyperostosis could be used as a trait to indicate ethnicity, using the remains from two medieval cemeteries on the same Danish island. Despite the similarities in the samples, the rates of porotic hyperostosis were radically different in the two cemeteries. The cemetery to the south, which consisted of primarily “poor, undernourished and mostly leprous individuals” (Henschen 1961: 729) had much higher frequencies of porotic hyperostosis then the other cemetery. This pattern led Henschen to reject Welcker’s assertions that cribra orbitalia was a racial marker. Instead Henschen offered the hypothesis that differences in the populations, specifically in the nutritional and chronic infection patterns of the two communities, was the cause of the differing rates of porotic lesions. He also used these factors to explain the lack of porotic hyperostosis in the crania he examined that were more recently deposited, as the nutritional and health conditions in modern populations were superior to those during the medieval period, which would result in far lower frequencies of porotic hyperostosis in modem remains (Henschen
1961).
In 1966 a highly influential article was published by Nathan and Hass. The authors developed a scoring guide for cribra orbitalia and applied it to 718 crania from ancient samples from the Judean desert; native populations from North, South and 14
Central America; Eskimos and modern peoples from India. At least one of the scored
lesions was present in every individual sampled, though the most severe manifestations
were more frequent in children and were seen more frequently in adult females then adult
males. It was the manifestation of cribra orbitalia in all individuals, despite the diverse
sample, which indicated that the lesions were most likely associated with nutritional
deficiency (Nathan and Hass 1966). While the scoring methods were developed
specifically for cribra orbitalia, the authors noted that porotic lesions of the cranial vault
seemed to follow the same characterizations. Three types of porous lesions were
classified as: 1) porotic which was characterized by small isolated pores, 2) cripeotic in which the pores are larger and more densely distributed without pore coalescence, and 3) trabecular in which the pores have become coalesced and resulting in the exposure of
bone trabeculae. This scoring system was used in the development of later scoring guides
still in use today (Buikstra and Ubelaker 1994; Stuart- Macadam 1985; Wilczak and
Zimova Hopkins 2010).
J. Lawrence Angel (1964, 1966, 1967) published the first reference to porotic
hyperostosis as we know to it today. Angel’s work is vital to understanding porotic hyperostosis because he is the first researcher to concisely define what characterizes a
lesion classified as porotic hyperostosis. Angel indicated that porotic hyperostosis was not only a visible porous lesion on the outer table of the bone, but also an increase in the
spongy bone. Porotic hyperostosis was originally used to describe only lesions located on the cranial vault. In more recent years the literature has seen the substitution of the term 15
porotic hyperostosis for lesions on the orbital roof, which should, more accurately, be referred to as cribra orbitalia. While Angel provided a concise definition of lesions that could be called porotic hyperostosis, the literature has adopted a scoring system more consistent with what Nathan and Haas (1966) proposed for cribra orbitalia.
The second reason his work is so vital is that Angel provides the first coherent and convincing argument linking porotic hyperostosis to anemia. Using research from previous publications, Angel noted that the descriptions of porotic hyperostosis better fit the clinical descriptions of bony change associated with hemolytic anemia rather than other nutrient deficiencies like rickets or infantile scurvy (Angel 1964). He further developed the idea that a hereditary anemia could be the cause of porotic hyperostosis in his Bronze Age Greek population due to the endemic nature of thalassemia within the descendent population. This idea was supported when Angel noted a higher frequency of porotic hyperostosis in Mediterranean populations who lived in marshy environments, which were the preferred habitat of the Anopheles mosquito known to carry Falciparum malaria. The presence of the Anopheles mosquito was linked with the endemic nature of thalassemia within Modern Greek populations. Thalassemia, like sickle cell anemia, confers a degree of immunity against malaria caused by Falciparum malaria. The overlap in areas with populations containing high frequencies of porotic hyperostosis, the range of mosquitos which host the Falciparum malaria protozoan and the endemic nature of thalassemia in the descendent groups of his study populations formed the basis of 16
Angel’s conclusion that thalassemia was the most likely anemic condition to cause porotic hyperostosis within the samples from Greece (Angel 1964, 1966).
However Angel noted that thalassemia could not be applied universally as the cause of porotic hyperostosis due to the high frequency of the latter worldwide. Angel proposed that sickle cell anemia was the probable cause of porotic hyperostosis in areas of the world where the disease is endemic today. In the pre-contact New World the proposed cause of these defects was iron deficiency anemia, which could result from “prolonged lactation, or of a severely restricted childhood diet, or of severe dysentery”, as there were no known hereditary anemias within the Americas (Angel 1966:761).
The Synergistic Approach
During the 1970’s research into porotic hyperostosis continued and the theory that anemia is the predominate etiology for these lesions became popular. Researchers generally emphasized iron deficiency anemia as a more significant agent then genetic anemias in the formation of porotic hyperostosis due to the worldwide distribution of these lesions. It was the opinion of the time that the worldwide distribution of porotic hyperostosis could not be the result hereditary anemias as their presence is too restricted.
Once a consensus on the primary etiology of the lesions was reached, the focus of research shifted. Now researchers wanted to know what factors were responsible for the development of iron deficiency anemia within their populations. 17
Hengen (1971) is perhaps the first work we see that reflects the debates characterizing this decade of research. He excluded systemic toxic disorders, protein deficiency, scurvy,
A-acitaminosis and panthothenic acid deficiency as potential causes of hyperostotic cranial lesions. It was argued that porotic hyperostosis represented hyperactivity of the diploe and that any causal condition must have this as a central feature (Hengen 1971).
Anemia, which can produce bony changes associated with marrow expansion, was once again proposed as the most likely etiology. Given the rarity of hereditary anemias on a global scale, iron deficiency anemia was postulated as the most likely etiology. He proposed that while a diet poor in iron could be a cause of iron deficiency anemia, parasitic infection leading to iron deficiency anemia was the likely cause of porotic hyperostosis. Noting that porotic hyperostosis was more common in equatorial environments, in children and in the historical record, Hengen (1971) argued that parasitic infections followed these same trends and that high parasitic loads could be the cause of the anemias resulting in hypertrophic lesions.
Using prehistoric Nubian remains, Carlson et al. (1974) set the precedence for what would later develop into the maize dependency model. The authors hypothesized that iron deficient anemia was the cause of porotic hyperostosis within their sample population rather than a genetic anemia. The authors believed that an iron-poor, cereal based diet combined with maternal iron stress and parasitic infection interacted in infants causing weanling diarrhea. This condition led to infants developing iron deficient anemia, 18
thereby explaining the high frequency of porotic hyperostosis in the children of their sample. Perhaps more important than their conclusion on the etiology of porotic hyperostosis in their sample was the cautious tone of the piece. The authors stressed that the unique environmental, cultural and biological elements of a population needed to be considered as it would inform the paleopathological diagnosis.
Three years later a study was published exploring the interaction of diet and infection in the pre-Colombian Midwestern United States and the effect that they may have on the development of iron deficiency anemia (Lallo et al. 1977). Data from individuals under the age of 15 years were collected from two sites and three distinct time periods. Remains from Dickson Mounds site in west central Illinois made up the majority of the sample.
This site included individuals whom experienced three distinct subsistence patterns: hunting and gathering, transitional substance and an agricultural subsistence pattern. A fourth sample was taken from the Eiden site in Ohio as it was considered to be comparable to the agricultural subsistence sample from Dickson Mound. Within the groups that had agriculturally based subsistence patterns, the authors noted both a greater frequency of porotic hyperostosis and more severe lesions when compared to the other samples. When examining the diets of both agricultural groups, it was found that both agricultural populations were highly dependent on maize. Maize contains little bioavailable iron and contains phytates, which bind iron. This combination reduces the amount of iron available through ingestion of maize (Lallo et al. 1977). Porotic 19
hyperostosis was found in much higher frequency in the youngest individuals, which was attributed to these individuals already experiencing iron stress due to the demands of rapid growth (Lallo et al. 1977). It was further hypothesized that a nonspecific infectious disease may have played a role in the development of iron deficiency anemia. This was based on the observation that 73.8% of individuals with porotic hyperostosis also displayed indications of postcranial infection (Lallo et al. 1977). The authors explained the high frequency of porotic hyperostosis in the agriculturalists from the Dickson Mound and Eiden sites as a synergistic effect of children experiencing iron stress due to the demands of growth, the lack of iron in the maize-dominated weanling diet and infectious disease which heightened their vulnerability to developing iron deficiency anemia (Lallo et al. 1977).
Mensforth and colleagues (1978) continued researching the interaction of diet and disease in the development of iron deficiency anemia. Studying the remains of individuals under the age of ten from the Late Woodland Libben site in Ohio, the authors included far more factors that may have interacted with diet and disease in the development of iron deficiency anemia. Among these factors the authors considered iron levels at birth, birth weight and rate of growth. It was concluded that these factors could impact the development of iron deficiency anemia. For example, an individual with low iron levels at birth would be expected to develop iron deficiency anemia earlier than an individual who was had high iron levels at birth from the same population. They also 20
believed that diet was intrinsic enough in the development of iron deficiency anemia that porotic hyperostosis could be considered a marker of nutritional stress (Mensforth et al.
1978). It should be noted, however, that Mensforth would later revise his opinion and instead postulate that infections disease transmitted though mosquitos was the major contributing factor in the development of anemia at the Libben site (Mensforth 1991).
Not all research attributed porotic hyperostosis solely to iron deficiency anemia.
Researchers like Cybulski (1977) focused on the role that the genetic anemias could play in the development of porotic hyperostosis and cribra orbitalia. While studying remains from costal British Colombia, Cybulski noted high frequencies of cribra orbitalia in early historic Haida populations when compared to other contemporaneous populations. He postulated that this might be due to the presence of a genetic anemia, though he did not dismiss iron deficiency anemia as a potential cause.
Maize Dependency Models
At the same time that researchers were studying the interaction of disease and diet in the formation of porotic hyperostosis, a much more simplistic model was being developed to explain the prevalence of porotic hyperostosis. This line of research, called the Maize Dependency model, would become the predominate explanation for porotic hyperostosis in the Americas. 21
The works of El-Najjar and colleagues (1975, 1976) would provide the foundation for the maize dependency model. The researchers sampled two groups of Ancient
Puebloan peoples from New Mexico and Arizona. The first groups represented four skeletal populations of individuals who lived in the canyon bottoms of the American
Southwest. These populations relied primarily on maize agriculture. The second group included two skeletal populations of individuals who lived on the sage plains of the
American Southwest. These individuals did not rely as heavily on maize and instead
supplemented their diet with animal products (El-Najjar et al. 1975, El-Najjar et al.
1976). The populations who lived on the canyon bottoms had a greater frequency of porotic hyperostosis than those who occupied the sage plains. The authors (1975, 1976)
claimed that the different frequency of porotic hyperostosis between the two habitation
zones was the result of iron deficiency anemia due to different diets. As parasites known to cause blood loss, and therefore cause anemia, were unknown in the prehistoric
American Southwest, the authors dismissed any contribution of parasites to the
development of anemia in these populations. Instead they focused on the iron-poor nature
of maize, stating that the heavier reliance on maize among those who lived on the canyon
bottom caused the higher prevalence of anemia and therefore the higher frequencies of
porotic hyperostosis (El-Najjar et al. 1975, El-Najjar et al. 1976).
This particular model of anemia development became popular due to its
simplicity. A lack of iron-rich foods as a cause of iron deficiency anemia is much simpler 22
than considering the interaction of many different variables. Additionally the association between increases in the frequency of porotic hyperostosis and the transition to agriculture had been noted in several distinct geographic regions (Cohen and Armelagos
1984). This association between agriculture and porotic hyperostosis and the simplicity of the theory led to the Maize Dependency Model becoming a major explanation for high rates of porotic hyperostosis for nearly 30 years. That is not to say that the maize dependency theory was not challenged well before it fell out of favor.
The first major critique of the maize dependency model was published ten years after the works of El-Najjar and colleagues. Walker (1986) studied the prehistoric populations from the Channel Islands of California. These populations relied primarily on the abundant marine resources surrounding them. Despite this iron-rich diet, the Channel
Island populations exhibited a high frequency of porotic hyperostosis. This finding directly contradicts the basic assumption of the maize dependency model. The maize dependency model assumes that populations with iron-rich diets will exhibit lower frequencies of porotic hyperostosis compared to populations with a diet low in iron. It was examining the terrain of the different Channel Islands and comparing it to the frequency of porotic hyperostosis that provided Walker the most likely cause of the anemias on the Channel Islands. The areas with the highest frequencies of porotic hyperostosis had little fresh water. The island with the greatest supply of fresh water had the lowest rates of porotic hyperostosis. This led Walker to the conclusion that waterborne bacteria leading to weanling diarrhea was the most likely cause of iron 23
deficient anemia within the Channel Islands (Walker 1986). He also postulated that
“prolonged breast feeding, helminth infections contracted through eating raw fish and sea mammal meat, and protein-calorie malnutrition during periods of low marine productivity” could have contributed to iron deficiency his study populations (Walker
1986: 354).
In addition to the direct evidence supplied from the Channel Islands research, Walker
(1986) also found flaws in the maize dependency model when examining lesion frequencies in maize dependent populations of the New World. Several of the maize dependent populations that he reviewed showed low frequencies of porotic hyperostosis.
If a diet consisting largely of maize was the primary cause of iron deficiency anemia, then these populations, which were known to have relied on maize agriculture, did not fit the model. Walker argued that the relationship between maize dependency and high incidence of porotic hyperostosis was not necessarily cause and effect. Walker instead proposed that an increased risk of disease load from contaminated water supply was likely what was similar between the disparate populations with high frequencies of porotic hyperostosis (Walker 1986).
Revisiting the Early Works
In the 1980’s Patricia Stuart-Macadam began publishing on porotic hyperostosis and cribra orbitalia combining clinical and anthropological data, and re-testing early observations that had been largely forgotten or ignored in the literature. It was the 24
integration of clinical data with the early anthropological literature that allowed Stuart-
Macadam to view porotic hyperostosis through a new lens.
Stuart- Macadam research began testing Hrdlicka’s 1914 conclusion that “the cases [of porotic hyperostosis that are seen] in adults represent merely remnants of the lesions of childhood” (Hrdlicka 1914; 62). Using the remains of 206 juveniles and 546 adults from the Poundbury Camp site in Great Britain Stuart- Macadam conducted a thorough study of porotic hyperostosis (Stuart-Macadam 1985). Individuals were assessed for the presence and severity of porotic hyperostosis as well as the presence of enamel hypoplasia and metopism. Statistical testing indicated an association between porotic hyperostosis and both enamel hypoplasia and metopism. This implied that all three of these condition occurred during a similar developmental period. Additionally only juveniles exhibited the most severe form of porotic hyperostosis, characterized as an
“outgrowth in trabecular structure from the normal contour of the outer bone table”
(Stuart-Macadam 1985, 392).
In order to support this finding, Stuart-Macadam (1985) integrated clinical data,
bone and bone marrow physiology as well as the anthropological data. The bony changes associated with anemia are the result of diploic expansion to create more space for hematopoietic marrow in response to increased blood production due to a lack of oxygen- carrying hemoglobin. The expansion of the diploe places pressure on the outer table of the skull, which results in the loss of the outer table and exposure of the diploe. The 25
marrow cavities of children contain high concentrations of hematopoietic marrow, while as adults the amount of hematopoietic marrow contained in the marrow cavities significantly decreases. In adults an increase in blood production simply requires an increase in the volume of hematopoietic marrow, which does not require diploe expansion, as there is sufficient room in the marrow cavities for the increased hematopoietic marrow, thus no lesions form. In addition to the physiological data
supporting her conclusions, Stuart-Macadam (1985) indicated that the clinical and anthropological data also supported her conclusions on the childhood origins of porotic hyperostosis.
Furthermore Stuart-Macadam (1985) stated that if porotic hyperostosis did not represent a childhood condition then distribution of porotic hyperostosis in ancient populations should have a similar distribution to modern cases of iron deficiency anemia.
Therefore if porotic hyperostosis is not representative of childhood conditions then the
adult cases of porotic hyperostosis should be skewed with more women being affected then men. However, when reviewing the anthropological literature, she found that the
majority of the studies showed either very small or no differences between the sexes
(Stuart-Macadam 1985). Thirteen years later she revisited the literature review and
included 28 studies that include data on the distribution of cribra orbitalia by sex. She
found that 24 of these showed no statistically significant differences between the sexes
(Stuart-Macadam 1998). The author concluded that Hrdlicka’s (1914) observation that
porotic lesions are representative of childhood conditions was accurate. 26
In her publication addressing the link between porotic hyperostosis and anemias,
Stuart-Macadam revisited techniques previously used by Moore (1929) and Williams
(1929) employing radiographic comparisons of skulls with porotic hyperostosis and those from clinical cases of anemia. Stuart-Macadam sought to improve on these earlier studies by increasing the sample size and using a systematic way of comparing the bony changes. Radiographs were compared on seven criteria: “1) Hair-on-end” pattern of trabeculation; 2) outer table thinning or disappearance; 3) texture changes; 4) diploic thickening; 5) orbital roof thickening; 6) orbital rim changes; 7) frontal sinus development” (Stuart-Macadam 1987c: 511-512). One hundred forty-four crania with and without porotic hyperostosis were used for analysis. Those with porotic hyperostosis showed a range in the severity and location of the lesions. Comparison of radiographs from these individuals to those of individuals with clinical cases of anemia revealed many of the same bone changes were observable in both groups. Furthermore, within the archaeological collection, the radiographic changes occurred with a greater frequency on crania with porotic hyperostosis than on crania without. This difference in frequency was found to be statistically significant, suggesting that there was indeed an association between porotic hyperostosis and the anemias.
However Stuart-Macadam did not only rely on radiographic evidence to support the association between anemia and porotic hyperostosis. The same year that Stuart-
Macadam published her radiographic study (1987c), she also published a multi-level analysis (1987b) used to support the anemia theory. Combining data from macroscopic, 27
microscopic and demographic data, Stuart-Macadam stated that there were striking similarities between clinical anemia studies and anthropological studies of porotic hyperostosis. Macroscopically both groups presented with porotic lesions of the outer table of the cranium, diploic expansion, symmetrically distributed lesions and lesions occurring primarily on the frontal, parietal and occipital bones. Microscopically both groups showed thinning of the outer table and expanded cancellous bone.
Demographically both anemia and porotic hyperostosis were most common during childhood, though both conditions were relatively rare in individuals under six months of age. Once again Stuart- Macadam concluded that porotic hyperostosis most likely occurred as the result of the bony changes associated with anemias (1987b), while confirming the findings on lesion distribution of Welcker (1888, Cited in Hill and
Armelagos 1990).
Stuart- Macadam would next revisit the work of Virchow (1874), examining the relationship between cribra orbitalia and porotic hyperostosis (Stuart-Macadam 1989b).
Virchow (1874) had previously proposed a common etiology for cribra orbitalia and porotic hyperostosis based on the similarity of the lesions. Once again Stuart-Macadam used a comparative approach. Anthropological and clinical data was compared at the macroscopic, microscopic, demographic and radiographic level. Due to the similarities found in the orbital and vault lesions in all types of comparison Stuart-Macadam concluded that cribra orbitalia and porotic hyperostosis shared a common etiology
(1989b). 28
While the work of Stuart-Maeadam in this section did not examine a new set of questions or generate a new theory about porotic hyperostosis, it is included for the same reason that the research in this thesis was conducted. It is necessary for pathologists to question the findings of their predecessors as well as their colleagues. Without doing so we rely on information that may be outdated or even inaccurate. By revisiting the early works and statistically testing the observations of her predecessors, Stuart-Macadam provided evidence and statistical analysis supporting many of the observations which modern diagnostic criteria are based on.
A Positive Adaptive Response
While revisiting the early works on porotic hyperostosis Stuart-Macadam was also developing a new theoretical frame for the study of porotic hyperostosis. In this new framework she proposed that porotic hyperostosis was a positive adaptive response and not a sign of dietary stress (Stuart-Macadam 1987a, 1988, 1989a, 1992a, 1992b, 1992c).
As many pathogenic microbes rely on their host for the iron supplies needed for reproduction, low levels of iron in the host are detrimental to many pathogens. Therefore
Stuart-Macadam suggested that low circulating iron levels may be a host response against pathogens, which may in turn lead to iron deficient anemia (Stuart Macadam 1988). With this in mind she argued that the lesion associated with porotic hyperostosis were in fact the results of a positive adaptative response, not dietary stress. 29
Stuart-Macadam stated that “except in cases of outright malnutrition, diet plays a minor role, if any, in the development of iron deficiency anemia” (1992c:40). Stuart-
Macadam stated that as much as 90% of the human body’s supply of iron is recycled from old red blood cells, it would take a diet that was extremely poor in iron to cause iron deficient anemia (Stuart-Macadam 1992a). She argued that the apparent correlation of porotic hyperostosis and the introduction of cereal grains might be due to an increase pathogen load, which accompanied the shift into an agricultural subsistence pattern
(Stuart-Macadam 1992a).
Challenging Positive Adaptation
Goodman (1994) would be the first to challenge Stuart-Macadam assertions on the positive adaptive nature of porotic hyperostosis. Goodman asserts that this theory utilized an overly simplistic understanding of physiology, which discounted the interconnectivity of the bodies systems. He argues that this led to Stuart-Macadam’s view that “signs of stress are seen as adaptation for no other reason than that they exist in stressed but surviving organisms.” (Goodman 1994:164). Furthermore Goodman points out that, while the human body may withhold iron in response to a high pathogen load, this iron deficiency will always have a negative functional impact and should therefore be viewed as an adjustment not an adaptation (Goodman 1994).
Echoing the concerns of Goodman (1994), Holland and O’Brien (1997) also challenged the adaptive response theory. Holland and O’Brien took issue with the attempt 30
to view anemia as an adaptive response as it minimalized or ignores the health consequences that can be present in even minor, short-term periods of iron deficiency.
Furthermore the authors questioned how much protection low iron levels confer against parasitic infection. Recognizing that high incidence of porotic hyperostosis are often linked to agricultural societies, the authors believed that both the maize dependency theory and the adaptive response theory were too simplistic to explain the etiology of porotic hyperostosis. Instead they proposed multi-causal models that include both diet and disease as factors likely to affect the development of diseases, like those proposed by
Lallo et al. (1977), Mensforth et al. (1978) and Walker (1986).
Scurvy as a Cause of Porous CraniaI Lesions
While anemia, either acquired or genetic, has been generally accepted as the cause of porotic hyperostosis and cribra orbitalia since the 1960s, it is by no means the only proposed etiology. Other researchers began to argue, quite convincingly, that anemia was not the only disease process that could cause porosity on the cranium. In multiple publications Ortner has linked porotic lesions to infantile or childhood scurvy (Ortner and
Ericksen 1997; Ortner et al. 1999; Ortner et al. 2001). Ortner asserts that porous lesions in the superior orbits, greater wing of the sphenoid, squamous portion of the temporal, the mandible and maxilla are unlikely to be the result of anemia as they have limited diploic space that would not be affected by increased hemopoesis (Ortner 1984). The presence of these porous areas are instead attributed to the inflammatory response caused by 31
hemorrhaging (Ortner and Erieksen 1997; Ortner et al. 1999; Ortner et al. 2001) due to weakened blood vessels associated with vitamin C deficiency (Ortner and Erieksen
1997). A hemorrhage between the periosteum and the bone surface will stimulate production of woven bone, and any hemorrhage can result in bone porosity as the body attempts to diffuse blood accumulation from the hemorrhage.
While the anthropological literature frequently reports cases of porotic cranial lesions on the parietal bones as porotic hyperostosis, Ortner (1984) never intended for these lesions to be linked. Due to the differing methods of creation, Ortner considered the porotic cranial lesions exhibited in scorbic individuals as unique pathologies (Ortner and
Erieksen 1997; Ortner et al. 1999; Ortner et al. 2001). The lesions created on the endocranium due to vitamin C deficiency are caused by atypical factors affecting the lamellar surface. By contrast porotic hyperostosis occurs as cancellous bone affects areas that are typically filled with lamellar bone.
While Ortner’s observations on the association between porous lesions of the cranium and scurvy have generally been accepted in the paleopathological community, it has been challenged. In particular Melikian and Waldron (2003) compared Ortner’s diagnostic lesions for scurvy against known clinical cases of scurvy. Melikian and
Waldron (2003) studied archaeological remains from both Peru and Great Britain for lesions matching Ortner’s description of porotic scorbic lesions. These archaeological remains were then compared to the remains of four infants with known cases of scurvy 32
housed at the Wellcome Museum of Anatomy and Pathology. The clinical remains were all described as having “large areas of new bone [formation] on the frontal, parietal or occipital bones, sometimes symmetrical and in one case, with new bone on the roof of the orbit and the greater wing of the sphenoid. (Melikian and Waldron 2003:210). The authors felt that the patterning and expression of the lesions differed enough among the archaeological and clinical samples that Ortner’s diagnostic criteria for scurvy was questionable. Melikian and Waldron’s interpretations were in turn challenged by
Brinkley and Ives (2006) who noted that clinical cases often exhibit the most extreme manifestation of a disease while archaeological cases often represent milder forms of the disease.
Porotic Hyperostosis as a non-specific indicator of stress
While many researchers do not feel as though there is enough data to attribute one specific etiology for porotic hyperostosis, others suggest that porotic hyperostosis cannot be linked to any one specific etiology. In 2001 Shultz published his histological analysis on porotic hyperostosis. Believing that neither macroscopic nor radiological analysis of porotic cranial lesions could reliably provide a diagnosis, Shultz emphasized the need to use histopathology in order to accurately determine the cause of porotic cranial lesions.
His analysis of the histopathology of porotic hyperostosis led Shultz to conclude that infection, metabolic disease, and neoplasms could all cause porotic cranial lesions. 33
Three years later Shultz co-authored a study (Wapler et al. 2004) in which the authors conducted a microscopic and histological study utilizing crania from Nubia to determine the frequencies of different causes of cribra orbitalia. Bone samples were collected from affected orbital roofs and prepared into slices, which were viewed using microscopy between xl0.5 and x360. Anemia-induced cribra orbitalia was identified when marrow spaces in the diploe were wider than typical. This would indicate hypertrophy of the marrow. In severe cases diagnosis related to anemia were recognized when the external lamina of the bone was open. The widening of the marrow and lamina were not identifiable macroscopically. The study found that of the 333 individuals exhibiting orbital roof lesions in their study, 56% showed no hypertrophy of the red bone marrow (Wapler et al. 2004). This finding eliminates iron deficient anemia as the cause of orbital roof lesions in over half of the study population. Furthermore extrapolating from that idea, it is possible that a number of the cases that have been associated with iron deficiency anemia are actually the result of another condition (Wapler et al. 2004).
Furthermore another 20% of identified cases of cribra orbitalia were in fact taphonomic, not pathologic in origin (Wapler et al. 2004). As there was no one cause of cribra orbitalia within the population, the point originally made by Shultz (2001), that a single cause for porotic cranial lesions seems unlikely, was reaffirmed. 34
The Iron Deficiency Anemia Etiology Challenged
While the theme of the last two sections was proposing etiologies other than anemia, the theme of this section is those articles that challenge or rebut the assumption that porotic cranial lesions could be ever be caused by iron deficiency anemia. This discourse began in 2009 when Walker and colleagues published “The Causes of Porotic
Hyperostosis and Cribra Orbitalia: A Reappraisal of the Iron-Deficiency-Anemia
Hypothesis” (Walker et al. 2009). Citing modern clinical research, which suggested that iron deficiency anemia results in the decrease and early death of red blood cells, the authors suggested that iron deficiency anemia could no longer be considered a possible cause of porotic hyperostosis. They came to this conclusion because the marrow hyperplasia that causes porotic hyperostosis results from an increase in red blood cell production. If iron deficiency anemia results in a decrease in red blood cells than marrow hypertrophy could not be associated with that particular anemia. Instead they suggested hemolytic and megaloblastic anemias as a potential cause of the marrow hypertrophy.
This proved somewhat problematic as the most common types of hemolytic anemias are hereditary anemias like sickle cell anemia and thalassemia. These hereditary anemias had already been eliminated as potential causes of most porotic hyperostosis cases due to the high prevalence of porotic hyperostosis and lack of known hereditary anemias in the
Americas. Instead the authors offered B12 or folic acid deficiencies resulting in megaloblastic anemia as the primary cause of porotic hyperostosis. Even so the authors 35
stated that “the synergistic effects of nutritionally inadequate diets, poor sanitation, infectious disease, and cultural practices related to pregnancy and breastfeeding provide a plausible explanation for the high rates of porotic hyperostosis found in many prehistoric populations” (Walker et al. 2009:7).
After reviewing the clinical literature on anemia on their own Oxenham and
Cavill (2010) concluded that Walker and colleagues (2009) had misunderstood “the clinical literature concerning the various anemias and associated hematopoietic responses or consequences thereof.” (Oxenham and Cavill 2010: 200). Oxenham and Cavill found that the only form of anemia that was incapable of producing marrow hypertrophy resulting in porotic hyperostosis would be what the authors called the anemia of chronic disease.
Shortly after the publication of Walker and colleagues (2009) “The Causes of
Porotic Hyperostosis and Cribra Orbitalia: A Reappraisal of the Iron-Deficiency-Anemia
Hypothesis”, Wilczak and Zimova Hopkins (2010) presented a poster that questioned whether all lesions commonly classified as cribra orbitalia had the same etiology. The authors suggested that only cases where there is “porosity with significant coalescence and/ or trabeculated hyperostotic expansion” (Wilczak and Zimova Hopkins 2010) could be linked to anemias based on macroscopic examination. It was hypothesized that areas of porosity with vascular impressions are likely the result of isolated inflammatory reactions, though advanced stage remodeling of hyperostotic lesions was not eliminated 36
as a potential cause of vascularized porotic lesions. Potential etiologies of porotic lesions were considered to be numerous. However this is the first work to try and classify and score lesions based on their macroscopic appearance.
As briefly outlined above, historically, the etiology of cribra orbitalia and porotic hyperostosis of the cranium has been linked to various anemias. More recent publications have debated this point, instead indicating that these conditions are caused complex interactions between malnutrition, sanitation, infectious disease and/ or cultural practices
(Holland 1997; Oxenham and Cavill 2010; Peckman 2003; Rothschild 2012; Rothschild
Manzi and Salvadei 2002; Schultz 2001; Wapler et al. 2004; Walker et al. 2009), where iron deficient anemia may be secondary.
Co-occurrence of Porotic hyperostosis and Cribra orbitalia
“Virchow suggested as early as 1974 that porotic skull lesions of the orbit and vault were related and were the result of the same underlying disorder; however, this remains a controversial issue. Although many researchers recognize a relationship, others prefer to consider orbital and vault lesions as separate conditions with their own etiologies” (Stuart-Macadam 1989b, 187). Studies have been done comparing the theoretical aspects of an association between porotic hyperostosis and cribra orbitalia, but the results of these studies are conflicted (Stuart -Macadam 1989b, Walker et al. 2009).
Other studies have had the data on the co-occurrence of these lesions but have not statistical tested this association. In multiple studies researchers have noted varying 37
frequencies of porotic hyperostosis and cribra orbitalia (Guichon 1994, Keeenleyside and
Panayotova 2006, Lovell 1997, Perez-Perez and Lalueza- Fox 1992, Stuart- Macadam
19998, Suby and Guichon 2010) between and within populations. Due to the conflicting information about the co-occurrence of these lesions, one would expect that there had been multiple studies testing the co-occurrence of porotic hyperostosis and cribra orbitalia. However to the best of my knowledge this study is the first to directly test whether or not these two lesions have a statistically significant co-occurrence.
Historical Scoring Methods
Standard methodology for scoring porotic hyperostosis records three types of porous lesions. Original classification of the orbital lesions were: 1) porotic which was characterized by small isolated pores; 2) cripeotic in which the pores are larger and more densely distributed without pore coalescence and 3) trabecular in which the pores have become coalesced and resulting in the formation of bone trabeculae (Nathan and Hass
1966). The basic scoring criteria were then adapted by Stuart Macadam (1985) to score the cranial lesions. The first stage remains largely unchanged, the second and the third were changed significantly. The stage two was considered to have larger pores that are more densely distributed, but in Stuart- Macadam’s (1985) version stage two may also have coalescence of the pores. Stage 3 was defined by the presence of diploic expansion.
In both these recording methods the score value corresponds to increasing severity of the same disease process. The scoring methodologies used by Stuart-Macadam (1985) and 38
Nathan and Hass (1966) have set the standards for recording porotic lesions on the orbital roof and cranial vault (Buikstra and Ubelaker 1994).
Endocranial Lesions
The term endocranial lesion is used here to refer to any lesion occurring on the endocranial surface of the cranium. It does not include lesions that affect both the inner and outer surfaces of the cranium.
Early Works
The history outlined above details the manifestation of porotic lesions on the outer surface of the cranium. Cranial defects on the interior surface of the cranium were also being analyzed in this time period. First described in 1912, endocranial lesions caused by the deposition of new bone were classified as cribra cranii (Koganei \9\2). Noting ‘web like’ new bone formation of the frontal, parietal and occipital bones, Koganei (1912) found these lesions occurred more often in adults than in children. These endocranial lesions were linked with porotic lesions on the ectocranium. The link between endocranial and ectocranial lesions resulted in some researchers proposing a common etiology. Two primary etiologies were proposed; nutritional deficiencies (Henschen
1961) or inflammatory processes (Moller-Christensen 1961). Over a decade after the two publications hypothesizing a common etiology for endocranial and ectocranial lesions 39
Mensforth et al. (1978) proposed that the two types of lesions did not have a common etiology. Instead the lesions documented on the endocranium were viewed as the result of an inflammatory reaction (Mensforth et al. 1978). Trauma, with resulting epidural hematoma and non-specific meningitis and inflammation of the meninges were also hypothesized to cause endocranial lesions (Shultz 1989, 2001). Viruses, fungal agents, tumors and lead poisoning (Patterson 1992) as well as various conditions resulting in secondary infections, malnutrition and childhood infections (Hutchinson and Moncrieff,
1944) are all proposed etiologies for endocranial lesions.
SES and Intrathorasic Disease
Research into endocranial lesions had stagnated but was revived when Hershkovitz and colleges (2002) published a paper on serpens endocrania symmetrica, henceforth referred to as SES. SES is a type of endocranial lesion that forms as a result of simultaneous bone deposition and reabsorption, which causes the characteristic maze-like channels. Hershkovitz et al. (2002) used a documented skeletal collection in their study of SES, and therefore had at least partial records for each set of remains containing information like age, sex and cause of death. Over 78% of the individuals who exhibited
SES had a documented cause of death that was either listed as tuberculosis or as related to tuberculosis (Hershkovitz, 2002; 214). The other causes of death were listed as syphilitic pneumonia, bronchopneumonia, gastric carcinoma and myocarditis, of which 40
the former two were more common. Therefore the authors concluded that there is an association between SES and intrathoracic disease, possibly tuberculosis in particular.
Subsequent to the publication of “Serpens Endocrania Symmetrica (SES): A New
Term and Possible Clue for Identifying Intrathoracic Disease in Skeletal Populations”
(Hershkovitz et al. 2002), most research on SES focuses on the co-occurrence of the endocranial lesions and tuberculosis. While cranial involvement in tuberculosis is relatively rare, it is documented far more frequently in children then adults (Lewis, 2011).
Even in children the endocranial lesions tend to occur primarily on the frontal and parietal bones, where there is a greater amount of cancellous bone, and therefore more opportunity for infiltration by the Mycobacterium tuberculosis (Dawson and Robson
2012, Lewis 2011).
However, the endocranial lesions associated with tuberculosis are not limited to
SES. The literature also indicates that large lytic lesions, which penetrate both tables of the cranium, occur with tuberculosis. In children, large areas of bone loss, at least 20mm in diameter, with a moth eaten appearance, have been documented (Dawson and Robson
2012, Lewis 2011). The lesions typically occur in isolation, and are usually larger in diameter on the endocranial surface than the ectocranial surface. Interestingly, while
Hershkovitz et al (2002) identified the relationship between endocranial lesions and tuberculosis, the distinctive SES lesions are not typically the focus of discussion about tuberculosis and endocranial lesions. Instead, the literature tends to focus on the large 41
lytic lesions, often with brief statements like “[a]reas of bone formation and resorption termed endocranial lesions or serpens endocrania symmetrica are also thought to be a possible indication of tuberculosis and especially linked to tuberculosis meningitis in children” (Dawson and Robson 2012, 34).
The literature has reported a significant link between SES and tuberculosis, though no statistical significance testing has been done to prove a definitive link. The journals Tuberculosis, Pathobiology, Clinical Microbiology, Infection, Journal o f
Archaeological Science, International Journal o f Paleopathology, and the International
Journal o f Osteoarchaeology (Donoghue H. 2011, Donoghue et al. 2009, Hershkovitz et al. 2015, Janovic A. et al. 2015, Lewis 2011, Minozzi S. 2012, Mo et al. 2015 , Palfi G. et al 2015, Roberts C. et al. 2009, Schultz and Schmidt-Schultz 2015, Teschler et al.
2015) have all published articles heavily influenced by Hershkovitz et al. (2002).
However, these papers associate endocranial lesions with only tuberculosis. Hershkovitz et al. (2002) considered SES to be associated with intrathoracic diseases, one of which was tuberculosis. The connection between SES endocranial lesions and many different types of intrathoracic infections has been ignored. However, Hershkovitz et al. (2002) also influenced another path of research into endocranial lesions.
Nonspecific endocranial lesions
Two years after Hershkovitz et al. (2002), Lewis (2004) published a piece proposing etiologies for multiple types of endocranial lesions. Instead of focusing on a 42
single lesion type Lewis (2004) collected data on porotic lesions, new bone deposition, capillary lesions or vascular impressions and hair on end lesions, where bone lesions appear as small stiff vertical protrusions. In doing so Lewis offered the bioarchaeological community a new diagnostic tool. Lewis stated that all of these different types of lesions were the result of hemorrhage caused by different conditions, including infectious disease, malnutrition, trauma, tumors, and deviance of the circulatory system. Lewis also discerned that deposits of new bone or the diffuse woven bone could be considered not to be pathological in children. When the bone deposits fell along a major growth line, or
consisted of new bone growth in an individual less than 7 years of age, the lesions could
be considered normal. This is due to the massive development of the brain and
endocranium during early childhood, which may cause hemorrhage as a result of strain
during growth. Lewis not only established what lesions were pathological, but also
proposed causes for each type of lesion, which could then be integrated into the current
bioarchaeological record.
“Endocranial Lesions in Non-adult Skeletons: Understanding their Aetiology”
(Lewis 2004) spawned a rise in research incorporating endocranial lesions into the
diagnosis of other conditions. Far more frequently publications have begun documenting
endocranial lesions in individuals with various pathological conditions. 43
EndocraniaI lesions and Scurvy
As Lewis (2004) stated that hemorrhage as the result of metabolic disease is a possible cause of endocranial lesions, it is no surprise that the bulk of the literature surrounding metabolic disease and endocranial lesions details cases of scurvy.
In 2014 more publications on scurvy documented endocranial lesions then have been published in the past two decades (Bourbou 2014, Brown and Ortner 2011, Crandall and Haagen 2014, Crist and Sorg 2014, Halcrow et al. 2014, Klauss 2014 ,Mays 2008,
Petersone-Gordina, Gerhards, and Jakob 2014, Stark 2014). The pathognomonic signs of scurvy have always been associated with hemorrhage, making it no surprise that researchers have linked scurvy to endocranial lesions. A chronic deficiency in vitamin C negatively affects collagen production. Capillary walls in particular are weakened by the poor production of collagen fibers. Any trauma the capillary experiences, even minor traumas produced by regular use, like chewing, could cause a hemorrhage. “The human body targets the contents of the circulatory system when it is outside of its natural environment. The body naturally treats free blood as inflammatory agent and targets it for removal. This results in an inflammatory reaction in any bone in contact with the hemorrhage, typically resulting in the development of porosity of the bone” (Haagen et al. 2014, 35). It is this process that causes the porous lesions on the greater wing of the sphenoid, which is considered characteristic of scurvy. This underlying weakness of the capillary wall is not limited to any one part of the body, the defect in collagen fibers 44
affect all areas of the body. The same capillary weakness occurs within the vessels of the brain and dura, causing similar lesions on the endocranial surface. Any behavior that results in the rapid movement of the cranium could result in a hemorrhage on the interior of the skull. “As arteries in the dura rupture and leak into surrounding tissue space such that the hematoma separates the dura and periosteum from the bone and tears bridging vessels between the arachnoid and dura layers of the meninges” (Haagen 2014, 35).
Endocranial lesions and Trauma
Though Lewis (2004) first identified traumas as a potential cause of endocranial lesions, Catherine Gaither has further developed our understanding of trauma and endocranial lesion. Gaither published two papers (Gaither 2012 and Gaither and Murphy
2011) that addressed both endocranial lesions and trauma. Lewis (2004) noted that bilateral endocranial bone deposition was frequently seen in cases of modern child abuse, particularly with cases known colloquially as shaken baby syndrome. As a child is shaken their head rapidly jerks forwards and backwards, causing damage to the brain and vascular system. Any associated hemorrhage would cause an inflammatory reaction leading to changes in the bone. This typically manifests as porotic reaction or a vascular impression on the bone. Gaither (2012) and Gaither and Murphy (2011) used this one feature to interpret violence. Wanting to determine if there was a significant shift in the types of violence that children experienced, endocranial lesions were examined to determine type of violence. In all the samples only a small percentage of individuals had 45
bilateral endocranial lesions considered to be characteristic of child abuse. However there were varying rates of traumatic injuries, not typical of childhood, which would suggest that children of the Maya community were subject to a new source of violence after
European contact.
Endocranial lesions and Infectious disease
Most of the literature on endocranial lesions and infectious disease has been published on tuberculosis (e.g. Dawson and Brown 2012, Lewis 2011, Wilbur et al.
2009). However these publications typically use the term SES, classifying it as a type of endocranial lesion. Because we already described the relationship between endocranial lesions and SES lesions, we will not do so here. Instead we will discuss one of the only other infectious diseases discussed in conjunction with SES type lesions, leprosy. In
“Paleopathological and Molecular Study on Two Cases of Ancient Childhood Leprosy from the Roman and Byzantine Empires” (Rubini et al. 2012) the authors describe a 4-5 month individual found with significant reactive new bone formation on the endocranial surface of the frontal and occipital bones. Within the differential diagnosis for this individual, the authors reiterated Lewis’s (2004) contention that reactive new bone formation on the frontal and occipital bones should be considered normal in individuals less than six months of age.
As summarized above, the etiologies and the association between orbital and cranial vault lesions remains uncertain within the anthropological literature. Since no 46
clear answers are available through the literature 1 decided to test the association between cranial vault and orbital roof lesions to further clarify this. One would expect a significant association between two lesions that have the same etiology. Alternately if there is not a significant association between these two lesions then it suggests that they do not have the same etiology. Furthermore, the method of scoring based on the different macroscopic appearances of lesions, does not assume that all lesions in similar locations, the orbital roof, ectocranial vault or endocranium, have the same etiology. 47
Chapter 3: Materials
The crania used in this project were from The Spencer R. Atkinson Library of
Applied Anatomy at the University of the Pacific School of Dentistry in San Francisco
California. Originally a private collection, the crania were donated by Dr. Spencer R
Atkinson and George Hollenback. The collection consists of over 1400 crania, over 400 of which represent individuals under 18 (Dechant 2000). Provenience information for individual crania is largely unknown. This is because the collection was purchased through anatomical supply companies, collected while abroad, and from Atkinson receiving crania as gifts from colleagues (Dechant 2000, 18). What provenience information does exist indicates that there are crania from six continents and fourteen countries (Dechant 2000, 16).
Because cribra orbitalia and porotic hyperostosis are considered the remains of childhood stress, only individuals under 20 were considered. Initially 400 individuals were examined however the final sample consisted of 306 individuals that were scored for both endocranial and ectocranial lesions. Over bleaching of the bone or adherent tissue on the ectocranial surface, which made scoring of lesions difficult, eliminated 76 of the original 400 individuals. An additional 18 individuals were excluded from analysis because adherent materials obscured portions of the endocranial surface (Table 2). This was typically viewed as a stabilizing matrix that had been applied to the endocranial sutures. Due to the lack of records associated with this collection, it is difficult to 48
determine what this matrix is. Ages for all individuals were provided based on dental calcification (Richards, 2007).
Table 1 Age groupings by major developmental stages Valid Cumulative Frequency Percent Percent Percent Valid .6-2.5 years 43 12.9 12.9 12.9 2.6-6.5 years 140 41.9 41.9 54.8 6.6-10.5 years 88 26.3 26.3 81.1 10.6-14.5 years 27 8.1 8.1 89.2 14.6-17 years 19 5.7 5.7 94.9 17.1-20 years 17 5.1 5.1 100.0 Total 334 100.0 100.0
Table 2 Number of individuals sampled for each lesion type
Lesions of the Type of Type of endocranial cranial vault orbital lesions lesions
N Valid 324 324 306
Missing 0 0 18 49
Chapter 4: Methods
In order to collect information on cribra orbitalia the orbital roof was examined macroscopically and with 20x magnification and an alternate light source. The cranial vault and face were similarly examined for evidence of porotic hyperostosis. The endocranial surface was examined in one of two ways. In the case of intact crania a flexible penlight was inserted through the foramen magnum to illuminate the endocranium. The surface was then examined through the foramen magnum and orbit.
As both cribra orbitalia and porotic hyperostosis are considered to be representative of childhood conditions, age was another variable which was considered in this analysis. Ages provided by the University of the Pacific Dental School were in 1/10 of a year intervals. As there is no provenience information ages were determined by x-ray analysis of dental calcification (Richardson 2007). Since these ages would provide multiple cell counts with fewer than 5 individuals in a cell, the ages were combined into ranges based on major developmental periods. Age ranges for analysis are: .6-6.5 years,
6.6-10.5 years, 10.6- 14.5 years, 14.6-17 years and 17.1-20 years (Table 1).
Orbital Lesions AKA Cribra Orbitalia
Instead of scoring all orbital lesions as cribra orbitalia, each type of lesion was scored separately. Each orbit was observed and scored: 0) no lesion, 1) porosity only, 2) porosity with vascular channels, 3) porosity with significant pore coalescence but no evidence of diploic expansion or 4) porosity with significant pore coalescence and diploic 50
expansion. These categories were further refined for statistical testing. The last 2 categories were combined and considered to be hyperostotic cribra orbitalia (Table 3).
As each orbit was scored separately, the data had to be combined to create a new variable for statistical analysis. This was done by comparing the lesions present in each orbit and using the lesion with the highest assigned value as these were considered more severe manifestation. For example if the left orbit presented with porosity only and the right orbit had porosity with vascular channels then the reported lesion would be porosity with vascular channels (Figure 1).
Table 3 Description, classification, and codes for orbital lesions
Lesion description Lesion classification Abbreviation Numerical code
No lesion No lesion NL 0
Porosity only Porosity only PO I
Porosity with vascular Porosity with vascular PVC channels channels
Porosity with significant pore coalescence but no Hyperostotic cribra HCO evidence of diploic orbitalia expansion
Porosity with significant pore coalescence and Hyperostotic cribra HCO evidence of diploic orbitalia expansion Figure 1 Orbital lesions: A) Porosity with vascular lesions; B) Porosity only; C) Hyperostotic cribra orbitalia
As previously stated the literature refers to all lesions on the orbit as cribra orbitalia. Due to the classification system developed for this project it is inaccurate to refer to any orbital lesion as cribra orbitalia. Therefore when lesions of the orbital vault are referred to they will generally be called orbital lesions. When referring to a specific lesion or lesion type they are referred to with the classification from analysis.
Cranial Vault lesions AKA Porotic Hyperostosis
Wilczak and Zimova Hopkins (2010) format for scoring orbital lesions was adapted to the cranial surface. While the predominate methodology for scoring porotic hyperostosis was developed by Stewart-Macadam (1982), adapting the scoring methods
Wilczak and Zimova Hopkins (2010) developed for lesions of the orbital roof allows for direct comparisons of lesions of the orbital and cranial vault lesions. Therefore each bone of the cranial vault were examined and scored. Lesions on the cranial vault were scored as 0.) no lesion, 1.) porosity only, 2.) porosity with vascular channels, 3.) porosity with significant coalescence but no evidence of diploic expansion or 4.) porosity with 52
significant pore coalescence and diploic expansion. After data collection the scores were further refined. Porosity with significant pore coalescence with and without evidence of hyperostotic expansion where combined into one new classification for data analysis, referred to as porotic hyperostosis (Table 4).
Because each bone of the cranial vault was scored separately the data had to be combined into a new variable before statistical tests could be performed. This was done by comparing all lesions on the ectocranium and using the most severe lesion present on the cranial vault. For example if the frontal, left and right parietal bones had no lesions, but the occipital bone had a porous lesion with significant coalescence and no evidence of marrow hypertrophy the entire cranial vault would be classified as porosity with significant coalescence but no evidence of diploic expansion (Figure 2). 53
Table 4 Description, classification and codes for lesions on the cranial vault
Lesion description Lesion classification Abbreviation Numerical code
No lesion No lesion NL 0
Porosity only Porosity only PO 1
Porosity with vascular Porosity with vascular PVC 2 channels channels
Porosity with significant pore coalescence but no Porotic hyperostosis PHO 3 evidence of diploic expansion
Porosity with significant pore coalescence and Porotic hyperostosis PHO 3 evidence of diploic expansion
' f ' * V c
Figure 2 Cranial vault lesions: A) Porosity only; B) Porosity with vascular channels; C) Porosity with significant coalescence
As previously stated the literature refers to all lesion of the cranial vault as porotic hyperostosis. With the classification system developed for this project referring to all lesions on the cranial vault as porotic hyperostosis would be misleading. Therefore when 54
addressing general lesions on the cranial vault, the lesions will be called cranial vault lesions. Specific lesions on the cranial vault will be referred to as the lesion for analysis.
Endocranial lesions
Observations on endocranial lesions were made in one of two ways. When making observations on an intact cranium a flexible LED penlight was inserted through the foramen magnum to illuminate the endocranial surface. The endocranial surface of the vault was then examined through the foramen magnum. The cranial floor was observed by looking through the optic chiasm while the endocranium was illuminated through the foramen magnum. In cases where a transverse sectioning occurred the superior portion would be removed to expose the endocranial surface. Both sections were examined for endocranial lesions. To determine what lesions to score, the methods from
Lewis (2004) and Hershkovitz et al. (2002) were combined. For this project endocranial lesion were scored as follows; no lesion, porosity only, porous or vascular woven bone deposition, vascular lesions, erosions or SES (Table 5). 55
Table 5 Description and codes of endocranial lesions
Description of endocranial lesions Abbreviation Numerical code
No lesion NL 0
Porosity only PO 1
Porous or vascular woven bone deposition WB 2
Vascular lesions VL 3
Erosion E 4
Data was collected for the entire endocranial surface. The most common or severe lesion on the endocranial surface was recorded for that individual. Due to this method of data collection no recombination of scoring categories was necessary before statistical analysis.
Data analysis occurred in SPSS Version 22. To avoid transcribing errors all data was entered directly into SPSS during data collection. Frequency analysis was run to determine patterns in the distribution of lesions by age range. As all values for data analysis were assigned to a descriptive appearance of the lesion, this data is considered to be categorical and therefore chi-squared analysis was used to test for associations between lesion types. Where a Pearson’s chi-squared p-value was less than .05, it was assumed that there was a significant association between the lesion types. However, multiple cells had an expected value less than 5. This indicates that the assumptions necessary for the standard calculation of significance level for chi square analysis had not 56
been met, necessitating alternative calculations for p-values. In order to calculate these results, Monte Carlo simulations were used. In cases where the data set is too large for an exact p-value to be calculated, but they do not meet the criteria for asymptotic analysis, the Monte Carlo method provides an estimate of the exact p-value. It does this through repeated sampling. In this case Monte Carlos test was run using a 95% confidence level with a default value of 10,000 samples. Cramer’s V values were used to measure effect size using Monte Carlo estimates of significance. The strength of the association was interpreted based on the guidelines of Rea and Parker (1992). 57
Chapter 5: Results
As the primary focus of this study is to examine the co-occurrence and association between orbital lesions, ectocranial lesions and endocranial lesions, chi- square testing is the most suitable test available. Because the variables were input directly into SPSS 21, no transcription was necessary. Instead statistical testing was performed on previously coded data.
As cribra orbitalia and porotic hyperostosis are typically considered to be representative of a childhood condition, age was added in as a comparative variable.
Ages provided by the University of the Pacific Dental School were in 1/10 of a year intervals. Since this would provide multiple cell counts with fewer than 5 individuals in a cell, the ages were combined into ranges based on major developmental periods. Age ranges for analysis are: .6-6.5 years, 6.6-10.5 years, 10.6- 14.5 years, 14.6-17 years and
17.1-20 years (Table 1).
Porotic hyperostosis and Cribra orbitalia
The first statistical analysis performed was a frequency test for orbital and endocranial lesions by age. Orbital lesions all decrease in occurrence as age increases
(Table 6). Age related patterns of ectocranial lesions are not as clear (Table 7). All lesions decrease in occurrence until 14.6 years of age. At 14.6 years of age there is an increase in frequency of purely porotic lesions. The frequency of this lesion type 58
decreases again in individuals over 17 years old. Similarly the only individual that exhibited woven bone deposition on the ectocranial surface belonged in the 14.6-17 year old age category. Because there was only one individual with woven bone deposition on the ectocranium this lesion type was eliminated from further analysis. Furthermore the lesion frequencies indicate that multiple lesions have fewer than 5 individuals affected in each age category. Since chi- square analysis assumes that there are at least 5 individuals in each cell any results from a standard chi-square analysis would be invalidated.
Therefore in addition to the Chi-Squared analysis, Monte Carlo methods were used to determine whether or not there is a statistically significant associations between different types of orbital and ectocranial lesions (Table 8). 59
Table 6 Distribution of orbital lesions by age range
Type of orbital lesions Age Range CO PVC PO TotalNL .6-2.5 years Count 17 4 9 9 39 % 43.6% 10.3% 23.1% 23.1% 100.0% Std. Residual 0.8 -1.4 1.2 -0.6 2.6-6.5 years Count 34 24 26 36 120 % 28.3% 20.0% 21.7% 30.0% 100.0% Std. Residual -1.4 -0.2 1.7 0.5 6.6-10.5 Count 25 27 9 23 84 years % 29.8% 32.1% 10.7% 27.4% 100.0% Std. Residual -0.9 2.3 -1.1 -0.1 10.6-14.5 Count 11 4 0 9 24 years % 45.8% 16.7% 0.0% 37.5% 100.0% Std. Residual 0.8 -.4 -1.9 0.9 14.6-17 Count 11 3 0 5 19 years % 57.9% 15.8% 0.0% 26.3% 100.0% Std. Residual 1.6 -.5 -1.7 -0.1 17.1-20 Count 10 1 3 2 16 years % 62.5% 6.3% 18.8% 12.5% 100.0% Std. Residual 1.8 -1.3 0.3 -1.2 Total Count 108 63 47 84 302 % 35.8% 20.9% 15.6% 27.8% 100.0% 60
Table 7 Distribution of cranial vault lesions by age
Age ranges based on development periods. Frequency %
.6-6.5 years No lesion present 120 67.4
Porotic hyperostosis 2 1.1
Porosity with vascular 2 1.1 impressions
Porosity only 54 30.3
Total 178 100.0
6.6-10.5 years No lesion present 64 72.7
Porosity with vascular 1 1.1 impressions
Porosity only 23 26.1
Total 88 100.0
10.6-14.5 years No lesion present 20 83.3
Porosity only 4 16.7
Total 24 100.0
14.6-17.00 years No lesion present 11 57.9
Porosity with vascular 1 5.3 impressions
Porosity only 6 31.6
woven bone deposition 1 5.3
Total 19 100.0
17.1-20 years No lesion present 13 81.3
Porosity only 3 18.8
Total 16 100.0 61
/
Results of Monte Carlo estimations indicate that there is a statistically significant association between orbital lesions and cranial vault lesions at both the 95 and
99 percent confidence level (Table 8). The association between ectocranial and orbital lesions was significant (p = 0.006, n = 324), but weak (Cramer’s V = 0.160). There is a positive association between no lesions occurring on either surface (resid = 6.9), orbital porosity with no cranial vault lesions (resid.=7) and orbital porosity with vascular channels occurring with porosity only on the cranial vault (resid.=6.9). Negative associations could be seen in cases where there were no lesions on the orbital roof and porosity only on the cranial vault (r=-6.6), cribra orbitalia occurring with no cranial vault lesions occurring (resid.=-7.7), porotic orbital lesions with vascular impressions occurring with no cranial vault lesions (resid.=-6.1).
Table 8 Cross tabulation of orbital and cranial lesions for chi square analysis.
Lesion Type NLPOPVCPO
NL Count 92 27 1 0 residual 6.9 -0.4 -0.5 -6.6 PO Count 71 20 0 0
residual 7 -5.7 -1.1 -0.3
PVC Count 27 20 0 0
residual -6.1 6.9 -0.6 -0.1
HCO Count 38 23 3 1
residual -7.7 4.9 2.2 0.8 62
Orbital lesions and endocranial lesions
As previously stated there were fewer individuals with endocranial lesions or cranial vault lesions scored than those with orbital lesions. For the purpose of this analysis individuals who were scored for either orbital lesions or cranial vault lesions but not for endocranial lesions were eliminated from analysis. This reduced the sample size to 306. Initial observations were made from frequency analysis. All types of endocranial lesions generally decrease in frequency as age increases, with the exception of erosions
(Table 9). SES type lesions occur only in the youngest individuals.
Table 9 Frequency of endocranial lesions by age range
Type of endocranial lesions present
Age grouping PO WBF Erosion SES TotalNL
.6-2.5 years 21 0 3 0 3 37
2.6-6.5 years 73 5 7 5 1 115
6.6-10.5 years 66 2 2 2 0 81
10.6-14.5 years 11 3 0 6 0 22
14.6-17 years 8 3 1 2 0 17
17.1-20 years 8 0 0 5 0 17
Total 187 13 13 20 4 289 63
Chi-square cross tabulations indicated that multiple cells had 5 or fewer cases
(Table 10). This invalidates standard chi-square analysis and therefore required the use of
Monte Carlo methods to determine if there was an association between lesions of the orbital roof and those of the endocranium. Results of statistical analysis indicate that there is no statistically significant association between orbital lesions and endocranial lesions at either the 1% or 5% confidence interval. (p=0.19, n=306).
Table 10 Cross tabulation of orbital and endocranial lesions for chi square analysis.
Lesion type NLPO WBFVL
NL Count 81 5 4 17
Residual 2.9 0.1 -0.5 -2.3
PO Count 61 2 1 12
Residual 4.5 -1.6 -2.3 -2.0
PVC Count 25 2 4 7
Residual -3.9 0.2 2.3 -0.2
HCO Count 39 4 3 15
Residual -3.4 1.3 0.5 4.5
Cranial Vault Lesions and Endocranial Lesions
As stated in the previous section the number of individual’s scored for cranial vault lesions and endocranial lesions differed. More individuals were scored for cranial vault lesions than endocranial lesions. Individual which were not scored for endocranial 64
lesions, yet scored for cranial vault lesions were eliminated from analysis. Conclusions based on the frequency of the endocranial lesions are described in previous sections and therefore not re-asserted here to avoid redundancy.
Table 11 Cross tabulation of endocranial and cranial vault lesions for Chi-square analysis
NL POWBF VL Erosions
NL Count 149 11 8 32 13
Residual 4.9 1.9 -0.4 -3.7 -1.0
PO Count 57 2 3 16 5
Residual -0.9 -1.7 -0.4 1.7 -0.6
PVC Count 0 0 0 3 1
residual -2.7 -0.2 -0.2 2.3 0.7
PH Count 0 0 1 0 0
Residual -0.7 0 1.0 -0.2 -1
Analysis of the cross tabulations produced by chi-square tests indicate that multiple cells have a value of less than 5, therefore invalidating a basic chi-square analysis (Table 11). Therefore chi-square analysis with Monte Carlo’s exact test was used to determine if there is a statistically significant association between endocranial lesions and lesions on the cranial vault. Results of the Monte Carlo simulations indicate that there is a statistically significant association between endocranial lesions and cranial vault lesions at the 95% confidence level (p=0.027, n=306), however that association is weak (Cramer’s V = 0.213). Positive associations were found for no lesions occurring in 65
either location (r=4.9) and vascular lesion on the endocranial surface occurring with porous lesions with vascular impressions on the cranial vault. A negative association between porous lesion with vascular impressions on the cranial vault and a lack of endocranial lesions (r=-2.7) was also found. 66
Chapter 6: Discussion
Analysis of orbital and cranial lesions of both types in this project raised some interesting questions when compared to the findings within the anthropological literature.
Frequency analysis of orbital lesions and cranial vault lesions indicated that all lesions types typically decline as age increases. However when analyzing the age distribution of endocranial lesions, it was not so clear. All lesions increase in frequency until 10.6 years of age, when lesion frequency declines. Based on these findings, it suggests endocranial lesions may not have a common etiology with orbital and cranial vault lesions.
Extrapolating from there, we would not expect a strong association between endocranial lesions and orbital or cranial vault lesions, but we would expect that there would be a strong association between orbital and cranial vault lesions.
Analysis of the relationship between orbital lesions and cranial vault lesions is not consistent with the amended hypothesis above. While statistical testing indicates that there is a relationship between these lesions, the pattern of distribution of these lesions calls into question whether or not the same condition is responsible for the lesions. The findings indicate two possible explanations for the distribution of these lesions. If the same condition is responsible for forming both types of lesions than this condition affects different ages groups in different ways. It has been argued that the higher frequency of orbital and cranial vault lesions in children is the result of physiological differences with adults. However, if the same condition is responsible for forming these two lesions, the 67
pattern distributions should be more similar for the two lesions. Alternatively, one may conclude that each manifestation of these lesions has a different etiology. The literature suggests that there are many different conditions that may result in both orbital and cranial vault lesions. One possible conclusion that is not articulated in most of the literature is that these different etiologies may have different lesion manifestations. If this is the case then this further complicated the diagnosis of a condition based on these lesions. Wilczak and Zimova Hopkins (2010) suggested that porous orbital lesions with vascular impressions may be the result of localized inflammatory reactions, which is inconsistent with the findings of this study. In order to clarify the possible conditions causing these lesions further study should be conducted. I would suggest that studies either documented skeletal collections with health records should document the types of orbital and cranial vault lesions. This would allow researchers to look for associations between different lesion manifestations and health conditions.
The findings of this study indicate that the relationship between lesions of the cranium is not as clear as previously assumed. If we reevaluate the relationship between orbital lesions and cranial vault lesions documenting and analyzing the different manifestations of these lesions in collections with health records, it is possible that we would find associations between individual health conditions and a particular type of
lesion. This in turn could potentially revolutionize the documentation of orbital and cranial vault lesions, as well as the way that we diagnose conditions based on these
lesions. If there are associations found between a health condition and a particular type of 68
orbital or cranial vault lesion than this would refine the possible etiologies for that lesion type in the archaeological record. In doing so it would improve the diagnostic standards for orbital and cranial vault lesions in the archaeological record. As stated earlier the findings here suggest the same condition is not responsible for all orbital and cranial vault lesions. By linking specific manifestations of these lesions to health conditions, we may be able to diagnose conditions resulting in orbital and cranial vault lesions in the archaeological record with much more accuracy.
Analysis between the relationships between orbital, cranial vault and endocranial lesions are more complicated. This study indicated that there is not an association between orbital lesions and endocranial lesions, yet there is an association between cranial vault lesions and endocranial lesions. Within this population there is no evidence indicating that the same condition causes the orbital, cranial vault and endocranial lesions. However because this study is the only one to date looking at the relationship between these three lesion types, I would suggest further study using documented skeletal collections in order to determine if these findings are consistent. This is particularly important because if one assumes that the same condition may form any of these lesions then we must question why there would be a statistically significant relationship between cranial vault lesions and endocranial lesions, but not orbital lesions and endocranial lesions. 69
The fact that we see a statistically significant relationship between endocranial lesions and cranial vault lesions, but not between endocranial lesions and orbital lesions is particularly interesting. There is a statistically significant relationship between cranial vault lesions and orbital lesions. Therefore we would expect that same type of relationship between endocranial lesions, cranial vault lesions and orbital lesions.
However the findings of this study indicates the opposite is true. It has been previously stated that different manifestations of cranial vault lesions and orbital lesions may in fact have different etiologies, which may explain the discrepancies in the co-occurrence of endocranial lesions, orbital lesions, and cranial vault lesions. 70
CHAPTER 7: CONCLUSION
The results of this study suggest that the relationship between porotic hyperostosis, cribra orbitalia and endocranial lesions are less clear than the anthropological literature would suggest. Frequency analysis of all three lesion types aligns with the literature. All lesions generally decrease as age increases, indicating that these lesions are likely the result of childhood conditions. However the discrepancies in the age distribution of these lesions suggest that either the same etiology affects different age groups in different ways or that they have different etiologies.
The relationship between orbital lesions and cranial vault lesions has been re asserted throughout the anthropological literature. As such it has been assumed that these two types of lesions have the same etiology. While the specific etiology has been debated, it is generally assumed that these lesions share a common etiology. However the findings of this study concur with the findings of Wilczak and Zimova Hopkins (2010), suggesting that there may be different etiologies for the different manifestations of these lesion types. However Wilczak and Zimova Hopkins (2010) suggested that orbital porosity with vascular channels was likely the result of a localized inflammatory reaction.
This study partially supports Wilczak and Zimova Hopkins suggestion as we do see a strong association between orbital porosity with capsular channels and no endocranial lesions. This would support the supposition that orbital porosity with vascular channels may be caused by localized inflammatory reactions. However in this study we also see 71
that there is a strong association between porosity on the endocranial surface and porosity with vascular channels in the orbits. That suggests that the same condition may be responsible for both types of lesions. The porosity on the endocranial surface could be a less sever manifestation of a inflammitory process that began in or around the orbit.
However if this is the case I hesitate to call this a localized inflamitory process as
Wilczak and Zimova Hopkins (2010) stated
The anthropological literature indicates that there is not an association between endocranial lesions, porotic hyperostosis and cribra orbitalia. In order to accurately test this porotic hyperostosis and cribra orbitalia were treated as two unrelated variables.
Testing these for a co-occurrence with endocranial lesions was particularly illuminating.
Due to the longstanding association between cribra orbitalia and porotic hyperostosis it was assumed that the relationships between these two lesion types and endocranial lesions would be the same. Analysis of the co-occurrence indicates the exact opposite.
There was a statistically significant association between endocranial lesions and cranial vault lesions, but no association between endocranial lesions and orbital lesions. This finding further supports the conclusion that different etiologies may be responsible for the formation of these lesions.
As there is a statistically significant association between cribra orbitalia and porotic hyperostosis one would assume that they have a similar etiology. If that is the case than any pathological lesion, which is associated with one, should be associated with 72
the other. While one etiology may form lesions on the orbital roof and cranial vault, it cannot be said that one pathology is responsible for all orbital roof and cranial vault lesions. The small number of orbital and cranial vault lesions, which exhibit evidence of diploic expansion in particular makes analysis difficult. Based on the anthropological and clinical data, it seems that anemia is still a likely etiology. However, based on the data, it does not seem likely that anemia is the common etiology for all types of cranial vault and orbital lesions.
As previously stated the dearth of information about endocranial lesions creates some difficulty when addressing specific etiologies. A study similar to that conducted by
Hershkovitz and colleges (2002), using documented skeletal collections, may indicate associations between particular health conditions and the different manifestations of endocranial lesions. This would clarify which conditions may cause each type of endocranial lesion. As endocranial lesions are not currently included in the diagnostic standards of any pathological condition, determining what conditions are associated with endocranial lesions may explain the discrepancies seen in the associations between orbital, cranial vault and endocranial lesions.
This study challenges two long held beliefs about lesions of the cranium. From the data here it seems likely that the etiologies of orbital and cranial vault lesions is more complex than we have generally assumed. However limitations in the sample make it difficult to determine what etiologies may be responsible for these lesions. In particular 73
the lack of provenience information about the individuals sampled here makes it quite difficult to determine what kinds of diseases or environmental conditions that these individuals may have been exposed to. In order to get more information and clarify possible etiologies for these lesions, the scoring methods used here should be applied to large documented skeletal collections. In particular it would be interesting to use a collection with medical record associated with each individual.
The possibility of using skeletal collection with medical documentation is appealing as we then have the ability to compare not only the co-occurrence between the lesions themselves, but also the relationship these lesions have to specific medical conditions. In doing this we may gain a better understanding of what conditions may be associated with each lesion type. After all it was comparing radiographs of archaeological cases of marrow hypertrophy to clinical cases of anemia that first provided evidence of an association between porotic hyperostosis and anemia. As the individuals in this study are all sub adults this could be challenging. 74
LITERATURE CITED Adachi, B. 1904a. Die Orbitae und die Hauptmasse des Schadels der Hapaner. Z. Morphol. Anthropol. 7, 379.
Adachi B. 1904b. Die Porositat des Schadeldaches. Z. Morphol.Anthropol 7, 373.
Angel JL
1964. Osteoporosis: Thalassemia? American Journal of Physical Anthropology 22: 369- 374.
1966. Porotic Hyperostosis, Anemias, Malarias and Marshes in the Prehistoric Eastern Mediterranean. Science 153: 760-763.
1967. Porotic Hyperostosis or Osteoporosis Symmetrica. In Diseases in Antiquity. D Brothwell and A Sandison. Springfield, Charles C Thomas.
Angel JL, Kelley JO, Parrington M, and Pinter S. 1987. Life stresses of the Free Black community as represented by the First African Baptist Church, Philadelphia, 1823-1841. American Journal of Physical Anthropology 74: 213-229.
Bourbou C. 2014. Evidence of childhood scurvy in a middle Byzantine Greek population from Crete, Greece (11th-12 centuriesA.DD.) International Journal of Paleopathology 5: 87-94.
Brinkley M and R Ives. 2006. Skeletal Manifestation of Infantile Scurvy. American Journal of Physical Anthropology 129:163-172.
Brown M and D Ortner. 2011. Childhood scurvy in a medieval burial from Macvanska Mitrovica, Serbia. International Journal of Osteoarchaeology 21:197-207.
Buichin R, Buikstra J, Stone A, Harkins K, Suby J, Massone M, Iglesias AP, Wilbur A, Constantinescu F and CR Martin. 2015. Pre-Columbian tuberculosis in Tierra del fuego? Discussion of the paleopathological and molecular evidence. International Journal of Paleopathology 11:92-101.
Buikstra Je and DH Ubelaker, Eds. 1994. Standards for Data Collection from Human Skeletal Remains. Fayetteville, Arkansas Archaeological Survey Research.
Carlson DS GJ Armelagos and DP van Gerven. 1974. Factors Influencing the Etiology of Cribra Orbitalia in Prehistoric Nubia. Journal o f Human Evolution 3: 405-410.
Cohen MN and GJ Armelagos. 1984. Paleopathology at the Onset of Agriculture: Editors’ Summation. Paleopathology at the Origins o f Agriculture. MN Cohen and GJ Armelagos. New York, Academic Press INC: 585-601. 75
Crandall J and KD. Haagen. 2014. Advancements, challenges and prospects in the paleopathology of scurvy: Current perspectives on vitamin C deficiency in human skeletal remains. International Journal of Paleopathology 5: 1-8.
Crist TA and MH Sorg. 2014. Adult scurvy in New France: Samuel De Champlain’s “Mai de la terre” at Saint Croix Island, 1604-1605 International Journal of Paleopathology 5: 95-105.
Cybulski J.S. 1977. Cribra Orbitalia, a Possible Sign of Anemia in Early Historic Native Populations of the British Columbia Coast. American Journal of Physical Anthropology 47: 31-40.
Dawson H and K Robson Brown. 2012. Childhood tuberculosis: A probable case from late mediaeval Somerset, England. International Journal of Paleopathology 2:31- 35.
Dechat D. 2000. Our Best Kept Secret: The Spencer R. Atkinson Library of Applied Anatomy Skeletons in the closet. Contactpoint: University of the Pacific School of Dentistry. 80(3): 14-19.
Donoghue H. 2011. Insights gained from paleomicrobiology into ancient and modern tuberculosis. Clinical Microbiology and Infection. 17: 6.
Donogue H., Hershkovitz I., Millikin D., Besra G., Lee O., Galili E., Greenblatt C., El- Najjar, M.Y., B Lozoff and DJ Ryan. 1975. The Paleoepidemiology of Porotic Hyperostosis in the American Southwest: Radiological and Ecological Considerations. American Journal of Roentgenology, Radium Therapy and Nuclear Medicine 25:918-924.
El-Najjar M.Y., DJ Ryan, CG Turner and B Lozoff. 1976. The Etiology of Porotic Hyperostosis among the Prehistoric and Historic Anasazi Indians of Southwestern United States. American Journal of Physical Anthropology 44:477-488.
Fairbanks VF, and Beutler E. 1972. Erythrocyte disorders- Anemias related to disturbances of hemoglobin synthesis. In Hematology. W.J. Wiliams, E. Buetler, A.J. Erslev and R.W. Rundles (Eds). New York: McGraw-Hill pp. 305-326.
Furuyama K, Haneko K, Vargas PDT. 2007. Heme as a magnificent molecule with multiple missions: heme determines its own fate and goberns cellular homeostasis. Tohoku Journal of Experimental Medicine 2013: 1-16.
Gaither C. 2012. Cultural conflict and the impact on non-adults at Puruchuco- Huaquerones in Peru: The case for refinement of the methods used to analyze 76
violence against children in the archeological record. International Journal of Paleopathology 2:2-3.
Gaither C and M Murphy. 2011. Consequences of conquest? The analysis and interpretation of subadult trauma at Puruchuco-Huaquerones, Peru. Journal of Archaeological Science 39: 467-478.
Goodman AH, Thomas RB, Swedlund Ac, Armelagos GJ. 1988. Biocultural Perspectives on Stress in Prehistoric, Historical and Contemporary Population Research. Yearbook o f Physical Anthropology 31:169-202.
Goodman AH, 1994. On the interpretation of health from skeletal remains. Current Anthropology 34: 281-288.
Haagen S. Klaus 2014 Subadult scurvy in Andean South America: Evidence of vitamin C International Journal of Paleopathology 5, 34-45.
Halcrow SE, Harris NJ, Beavan N and HR Buckley. 2014. First bioarchaeological evidence of probable scurvy in Southeast Asia: Multifactoral etiologies of vitamin C deficiency in a tropical environment. International Journal of Paleopathology 5: 63-71.
Hengen, O. 1971. Cribra Orbitalia: Pathogenesis and Probable Etiology. Homo 22:57-75.
Henschen F. 1961. Cribra cranii, a Skull Condition Said to be of Racial or Geographical Nature. Pathology and Microbiology 23: 724-729.
Hershkovitz I., Donoghue H., Minnikin D., May H., Lee O., Feldman M., Galili E., Spigelman M., Rothschild B., Bar-Gal G. 2015. Tuberculosis origin: The Neolithic scenario. Tuberculosis. 95: SI22.
Hershkovitz I, Greenwald CM. Latimer B, Jellema SW-B, Eshed V, Dutour O, Rothschild BM. 2002.Serpens Endocrania Symmetrica (SES): a new term and a possible clue for identifying intrathoracic diseases in skeletal populations. American Journal of Physical Anthropology 118(3): 201-216.
Hill, MC and GJ Armelagos. 1990. Porotic Hyperostosis in Past and Present Perspective. A Life in Science: Papers in Honor of J Lawrence Angel. JE Buikstra, Center for American Archaeology. Scientific Papers 6: 52-63.
Holland, TD and MJ O’Brien. 1997. Parasites, Porotic Hyperostosis and the Implication of Changing Perspectives. American Antiquity 62(2): 183-193. 77
Hrdlica, A. 1913. The most ancient skeletal remains of man. Annual Report o f the Smithsonian Institution 1913. Pp. 491-552. Government Printing Office: Washington D.C.
Hutchinson R, Moncrieff A. 1944. Lectures on Diseases of Children. Edward Arnold and Co.: London.
Janovic A, Milovanovic P., Sopta J. Rakocevic Z., Filipovic V., Nenezic D., and M. Djuric.2012. Short Report Intracranial Arteriovenous Malformations as a Possible Cause of Endocranial Bone Lesions and Associated Neurological Disorder. International Journal of Osteoarchaeology. 25(1): 88-97.
Keenleyside A. and Panayotova K. (2006) Cribra orbitalia and porotic hyperostosis in a Greek colonial population (5th to 3rd centuries BC) from the Black Sea. International Journal of Osteoarchaeology, 16: 373-384.
Klaus HD. 2014. Subadult scurvy in Andean South America: Evidence of vitamin C deficiency in the late pre-Hispanic and Colonial Lambayeque Valley, Peru. International Journal o f Paleopathology 5:34-45.
Koganei H. 1912. Cribra cranii and cribra orbitalia. Mitt Med Fak Toyko 10: 113-154.
Lewis ME.2000. Non-Adult Paleopathology. Current Status and Future Potential. In: Osteology in Archaeology and Forensic Sciences. Cox M and Mays S (eds). p. 39-57.
2004 Endocranial Lesions in Non-adult Skeletons: Understanding their Aetiology. International Journal of Osteoarchaeology 14:82-97.
2011. Tuberculosis in the non-adults from Romano-British Poundbury Camp, Dorset, England. International Journal of Paleopathology 1: 12-23.
Lallo, J.W. Armelagos and R.P Mensforth. 1977. The role of diet, disease and physiology in the origin of porotic hyperostosis. Human Biology 49:471-483.
Lovell N.C. (1997) Anaemia in the Ancient Indus Valley. International Journal of Osteoarchaeology, 7: 115-123.
Mays, Simon 2008 A Likely Case of Scurvy from Early Bronze Age Britain. International Journal of Osteoarchaeology 18: 178-187.
Melikian M and T Waldron. 2003. An Examination of Skulls from Two British Sites for Possible Evidence of Scurvy. International Journal o f Osteology 13:207-212. 78
Mensforth RP. 2002. Vitamin D deficiency mortality: Impaired immune response in infants and elevated cancer risk in adults. American Journal of Physical Anthropology 34(Supplement):l 12.
Mensforth RP, Lovejoy OC, Lallo JW, Armelagos GJ.1978. The role of constitutional factors, diet and infectious disease in the etiology of porotic hyperostosis and periosteal reactions in prehistoric infants and children. Medical Anthropology 2: 1-59.
Minozzi S., Catalano P., Caldarini C., Fornaciari G. 2012. Palaeopathology of Human Remains from the Roman Imperial Age. Pathobiology. 79: 5.
Moller-Christensen V. 1961. Bone Changes in Leprosy. Munksgaard: Copenhagen.
Molnar E., Donoghue H., Lee O., Wu H., Besra G., Mannikin D., Bull I., Llewellyn G. Williams C., Spekker O., Palfi G. 2015. Morphological and bimolecular evidence for tuberculosis in 8lh century AD skeletons from Belmegyer-Csomoki tomb, Hungary. Tuberculosis. 95: S35.
Moore, S. 1929. The Bone Changes in Sickle Cell Anemia with Note on Similar Changes Observed in Skulls of Ancient Maya Indians. Journal of Missouri State Medical Association. 26: 561-564.
Nathan H and N Hass 1966. “Cribra Orbitalia”, A Bone Condition of the Orbit of Unknown Nature. Israel Journal of Medical Science 2:171-191
Oetteking B. 1909. Kraniologische Studien an Altagyptern. Arch. Anthropol. NF8, 1-90
Olivares JL, Fernandez R, Fleta J, Ruiz MY, Clavel A. 2002. Vitamin B12 and folic acid in children with intestinal parasitic infection. The Journal of the American College of Nutrition 21: 109-113.
Ortner, DJ.
1984. Bone Lesions in a Probable Case of Scurvy from Melatavik, Alaska. MASCA 3:79- 81.
1998. Male-Female Immune Reactivity and its Implications for Interpreting Evidence in Human Skeletal Paleopathology. In Sex and Gender in Paleopathological Perspective. A Grauer and P Stuart-Macadam. New York, Cambridge University Press: 79-92.
2003a. Identification o f Pathological Conditions in Human Skeletal Remains. New York, Academic Press. 79
2003b, Metabolic Disorders. In Identification o f Pathological Conditions in Human Skeletal Remains. New York, Academic Press.
2003C. Methods Used in the Analysis of Skeletal Lesions. Identification o f Pathological Conditions in Human Skeletal Remains. DJ Ortner. New York, Academic Press: 45-64.
2009a. Involvement of the Lateral Wall of the Orbit in Cases of Scurvy. JM Bauder. Binghamton, NY
Ortner, DJ, W Butler, J Cafarella And L Millgan. 2001. Evidence of Probable Scurvy in Subadults From Archaeological Sites in North America. American Journal of Physical Anthropology 114:343-351.
Ortner DJ and Mays. 1998. Dry-bone manifestations of rickets in infancy and early childhood. International Journal of Osteoarchaeology 8:45-55.
Ortner DJ, EH Kimmerle and M Diez. 1999. Probable Evidence of Scurvy in Subadults from Archaeological Sites in Peru. American Journal of Physical Anthropology 108:321-331.
OrtnerD, Butler W, Cafarella H and Milligan L. 2001. Evidence of probable scurvy in subadults from archaeological sites in North America. American Journal of Physical Anthropology 114:343-351.
Ortner DJ and G Turner-Walker.2003. The Biology of Skeletal Tissues. In Identification o f Pathological Conditions in Human Skeletal Remains. DJ Ortner. New York, Academic Press.: 11-35.
Orter DJ and MF Eriksen. 1997. Bone Changes in the Human Skull Probably Resulting from Scurvy in Infancy and Childhood. International Journal of Osteoarchaeology 7:212-200.
Owen R. 1859. Report on a Series of Skills of Various Tribes of Mankind Inhabiting Nepal, Collected and Presented to the Museum by B.H. Hodgson. Report British Association, London.
Oxenham Mf and I Cavill. 2010. Porotic Hyperostosis and Cribra Orbitalia: The Erythropoietic Response to Iron-Deficiency Anaemia. Anthropological Science 118(3): 199-200.
Palfi ., Maixner F., Maczel M., Molnar E., Posa A., Kristof L., Marcsik A., Balazs J., Masson M., Paja L., Palko A., Szentgyorgyl R., Nerlich A., Zink A., Dutour O. 2015. Unusual Spinal tuberculosis in an Avar Age skeleton (Csongrad- Felgyo, 80
Urmostanya, Hungary): A morphological and biomolecular study. Tuberculosis. 95: S29. ’
Parrot J. 1879. The osseous lesions of hereditary syphilis. Lancet 1: 696-698.
Patterson KD. 1992. Meningitis. In The Cambridge World History o f Human Disease, Kiple K (ed.). Cambridge University Press: Cambridge; 875-880.
Peckman T. 2003. Possible Relationship Between Porotic Hyperostosis and Small Pox Infection in Nineteenth-Century Populations in the Northern Frontier South Africa. World Archaeology 35(2)289-305.
Perersone-Gordina E, Gerhards G and T Jakob. 2014. Nutrition-related health problems in a wealthy 17-18th century German community in Jelgava, Latvia. International Journal of Paleopathology 3: 30-38.
Perez-Perez A. and Lalueza Fox C. (1992) Indicadores de presion ambiental en aborigenes de Fuego-Patagonia. Reflejo de la adaptation a un ambiente adverso. Anales del Instituto de la Patagonia, 21: 99-108
Rea, L.M., and Parker, R.A.1992 Designing and Conducting Survey Research. San Francisco: Jossey-Bass
Resnick, D. 2002. Diagnosis of bone and joint disorders. 2nd ed. Philadelphia: Saunders.
Richards GD. 2007. Basicranial Ontogeny and Evolution in Homo Sapiens. Dissertation. Dept, of Anthropology, UC Berkeley.
Roberts C., Pfister L., Mays S. 2009. Letter to the editor: Was tuberculosis present in Homo erectus in Turkey? American Journal of Physical Anthropology 139:3.
Rothschild Bruce. 2012. Extrapolation of the Mythology That Porotic Hyperostosis is Caused by Iron Deficiency Secondary to Dietary Shift to Maize. Advances in Anthropology 2(3): 157-160
Rothschild B FR Manzi and L Salvadei. 2002. Letter to the Editor Porotic Hyperostosis as a Marker of Health and Nutritional Conditon. American Journal of Human Biology 14(4):417-420.
Rubini M, Erdal Y, Spigelman M, Zaio P and H Donoghue. 2014. Paleopathological and Molecular Study on Two Cases of Ancient Childhood Leprosy from the Roman and Byzantine Empires. International Journal of Osteoarchaeology 24(5): 570-582 81
Shultz M. 1989. Causes and frequency of diseases during early childhood in Bronze Age populations. In Advances in Paleopathology, Capasso (ed.). Marino Solfanelli Editore: Chieti; 175-179.
Schultz, M. 2001. Paleohistopathology of Bone: A New Approach to the Study of Ancient Disease. Yearbook o f Physical Anthropology 44(33): 106-147
Schultz M., Schmidt-Schultz TH.2015. Is it possible to diagnose TB in ancient bone using microscopy? Tuberculosis. 95: S42.
Stark R. 2014. A proposed framework for the study of paleopathological cases of subadult scurvy. International Journal of Paleopathology 5:18-26
Stuart-Macadam P.
1985. Porotic Hyperostosis: Representative of a Childhood Condition. American Journal of Physical Anthropology 74:521-526.
1987a. Nutrition and Anemia in Past Human Populations. Diet and Subsistence: Current Anthropological Perspectives, Calgary. Proceedings of the 19th Annual Conference of the Archaeological Association of the University of Calgary.
1987b. Porotic Hyperostosis: New Evidence to Support the Anemia Theory. American Journal of Physical Anthropology 74:521-526.
1987c. A Radiographic Study of Porotic Hyperostosis. American Journal of Physical Anthropology 74:511-520.
1988. Porotic Hyperostosis: Relationship Between Orbital and Vault Lesions. American Journal of Physical Anthropology 80: 187-193.
1989a. Nutritional Deficiency Disease: A Survey of Scurvy, Rickets and Iron-Deficiency Anemia. In Reconstruction o f Life from the Skeleton. MY Iscan and KA Kennedy. New York, Alan R Liss, Inc.: 201-222.
1989b. Porotic Hyperostosis: Relationship Between Orbital and Vault Lesions. American Journal of Physical Anthropology 80:187-193.
1992a. Anemia in Past Human Populations. In Diet, Demography and Disease: Changing Perspectives in Anemia. P Stuart-Macadam and S Kent. New York, Aldine De Gruyter:151-170.
1992b. Anemia Reevaluated: A Look to the Future. In Diet, Demography and Disease: Changing Perspectives on Anemia. P. Stuart-Macadam and S Kent. New York, Aldine De Gruyer: 261-268. 82
1992c. Porotic Hyperostosis: A New Perspective. American Journal of Physical Anthropology 87:39-47.
1995. Porotic Hyperostosis: Representative of a Childhood Condition. American Journal of Physical Anthropology 66: 391-398.
1998. Iron Deficiency Anemia: Exploring the Difference. In Sex and Gender in Paleopathological Perspective. A Grauer and P Stuart- Macadam. Cambridge, Cambridge University Press: 45-63.
Suby J.A. and Guichon R.A. (2009) Diet, nutrition and paleopathology in Southern Patagonia. Some experiences and perspectives. International Journal of Osteoarchaeology, 19: 328-336.
Teschler- Nicola M., Novotny F., Spannagi-Steiner M., Stadler P., Prohaska T., Irrgeher J., Zitek A., Daubi B., Haring E., Rumpelmayr K., Wild E. 2015. The Early mediaeval manorial estate of Gars/Thunau, Lower Austria: An enclave of endemic tuberculosis?. Tuberculosis 95: S51.
Toldt. C 1886. Uber Welker’s cribra orbitalia. Mitt. D, Anthropol. Gesell. Wein. 16,20- 24.
Virchow H. 1874. Althpatagonische, altohilenische und modern pampas schadel. Verhandlung der Berliner Gesellschaft fur Anthropologie, Ethnologie, und Urgeschichte 6:51 -64.
Walker PL. 1986. Porotic Hyperostosis in a Marine-Dependent California Indian Population. American Journal of Physical Anthropology 69: 345-354.
Walker PI, RB Bathurst, R Richman, T Gjerdrum and VA Andrushko. 2009. The Causes of Porotic Hyperostosis and Cribra Orbitalia: A Reappraisal of the Iron- Deficiency-Anemia Hypothesis. American Journal of Physical Anthropology 139:109-125.
Wapler U, E Crubezy and M Schults. 2004. Is Cribra Orbitalia Synonymous with Anemia Analysis and Interpretation of Cranial Pathology in Sudan. American Journal of Physical Anthropology 123(4):333-339.
Wilbur A, Bouwman A, Stone A, Roberts C, Pfister L, Buikstra J, and Brown, T. 2009. Deficiencies and challenges in the study of ancient tuberculosis DNA. Journal of Archaeological Science 39:9. 1990-1997. 83
Wilczak CA and Veronika Zimova Hopkins. 2010. Patterns of Inflammatory Orbital Lesions in a Modem Anatomical Sample. Presentation for the 37th Annual Meeting of the Paleopathology Association Albuquerque, NM April 13/14.
Williams, H. 1929. Human Paleopathology with some Original Observations on Symmetrical Osteoporosis of the Skull. Archive of Anthropology 84
APPENDIX A
Specimen Age of Age Category Orbital Ectocranial Endocranial individual Lesions Lesion Lesions Score Score A43 1 1 0 4 0 A68 1.3 1 0 0 0 A121 2 1 4 4 0 A133 2 1 4 0 0 B276 2.2 1 4 0 0 D29 1.9 1 0 0 0 D32A 1.9 1 0 0 3 A34 1 1 0 0 0 A36 2 1 4 0 0 A39 1.3 1 4 0 0 A40 1 1 0 0 0 A45 1.4 1 0 0 3 A52 1.6 1 0 0 0 A53 1.8 1 4 0 3 A54 2.1 1 0 0 0 A41 1 1 3 0 0 A47 1.8 1 3 0 0 A49 1.6 1 4 0 0 A50 1.8 1 0 0 3 B242 2 1 0 0 0 B156 1.9 1 4 0 0 B243 1 1 0 0 0 D60 2.1 1 0 0 3 E26 1.3 1 3 0 2 A74 2.4 1 0 0 3 B112 2.4 1 3 0 2 F17 1.8 1 3 0 3 D89 2 1 1 2 2 A51 1.4 1 1 3 3 B52 2.1 1 0 4 0 A55 2 1 0 4 0 A42 2 1 0 4 0 A63 1.1 1 0 4 0 A68 1.3 1 0 4 3 A44 1.9 1 4 4 0 A73 2.3 1 1 4 5 A48 1 1 3 4 0 85
Al 55 1 1 3 4 0 B248 2.2 1 0 4 0 D61 2.4 1 3 4 0 D86 1.7 1 3 4 5 A31 1 1 4 4 0 FI 279 2 1 1 4 3 B53 1 1 4 4 0 B246 2.8 1 0 0 0 D47 4.2 1 4 0 3 A76 2.9 1 0 0 0 A76 2.9 1 3 0 0 B278 3.1 1 0 4 0 B247 3.6 1 0 4 0 D33 4.2 1 0 4 0 A138 3.9 1 0 4 3 A110 4.9 1 4 4 0 A 105 5 1 1 4 5 A127 5.3 1 3 0 0 A207 5.4 1 3 4 0 A l 17 5.8 1 4 0 0 A 142 5.6 1 0 0 3 A145 6.5 1 1 4 0 A l 13 6.2 1 1 3 3 A62 4.1 1 4 0 0 A68 3.3 1 0 0 0 B246 2.8 1 1 0 1 D39 4.1 1 4 0 1 D41 3 1 4 0 0 D270 4.3 1 0 0 0 A57 2.9 1 4 0 4 A 107 3.3 1 0 0 0 A 120 3 1 4 0 0 A64 2.8 1 0 0 3 A67 2.9 1 0 0 0 A130 3.4 1 0 0 0 A69 2.8 1 0 0 1 Al 32 2.9 1 0 0 0 A71 3 1 0 0 0 A 198 3.4 1 4 0 0 A206 3.2 1 0 0 5 A75 2.8 1 4 0 3 B154 3 1 1 0 3 B154 3 1 0 0 1 86
B245 3 1 1 0 0 D28 3 1 0 0 0 D38 3 1 0 0 3 ’ D271 2.6 1 3 0 0 E40 3 1 4 0 0 A238 4.1 1 0 0 0 A312 4 1 0 0 0 A66 3.9 1 0 0 0 M l 3.6 1 4 0 0 M l 4.1 1 0 0 0 D33A 4.2 1 4 0 0 A86 4.1 1 4 0 0 D59 3.8 1 3 0 . 4 Al 35 4 1 0 0 0 A137 4 1 3 0 1 D85 4.2 1 4 0 0 D247 4.2 1 4 0 0 D256 3.5 1 3 0 2 D272 4.2 1 4 0 0 D276 4.2 1 4 0 2 D278 4.2 1 3 0 3 F2 4.2 1 1 0 0 A81 5.2 1 3 0 3 A82 4.7 1 0 0 2 A239 5.3 1 4 0 0 A85 4.4 1 0 0 0 A261 4.4 1 0 0 0 A91 5.1 1 4 0 0 A308 5.3 1 4 0 0 A314 5.3 1 4 0 0 A l 18 5.4 1 1 0 3 D46 5.1 1 4 0 3 D58 4.6 1 4 0 0 A 129 5.1 1 4 0 0 D87 4.5 1 0 0 0 D249 4.5 1 4 0 4 D261 4.8 1 1 0 0 D262 5.1 1 1 0 0 D274 5.1 1 4 0 4 F4 5.2 1 3 0 3 FI 1 4.2 1 1 0 3 F14 4.2 1 3 0 0 f 18 5 1 4 0 0 87
A89 5.5 1 3 0 0 A97 5.9 1 4 4 3 A311 6.1 1 3 0 0 Al 34 5.6 1 0 0 0 B275 5.7 1 1 0 0 D45 5.6 1 1 0 0 A 143 5.9 1 1 0 3 A 146 6 1 4 0 0 D248 5.3 1 4 0 0 D285 6 1 1 0 0 D288 5.8 1 3 0 0 E34 6.1 1 4 0 0 F8 3.2 1 0 0 3 F10 3.5 1 0 0 2 F12 4.2 1 1 0 0 F20 5.7 1 1 0 0 F1281 6 1 0 0 0 A96 6.5 1 0 0 3 A 149 6.5 1 3 0 0 D299 6.4 1 0 0 0 A80 5.4 1 4 0 3 A92 6.5 1 4 0 0 A94 6.3 1 0 0 0 A95 5.7 1 0 0 0 A98 6 1 0 0 0 AlOO 6.4 1 3 0 0 A61 3.9 1 3 2 0 A272 6 1 3 2 2 AMO 5.3 1 1 4 3 D42 4 1 0 4 0 D26 3 1 1 4 0 A38 3.3 1 3 4 0 A76 2.9 1 4 4 3 A58 2.6 1 3 4 2 A79 3 1 0 4 0 A59 3.5 1 4 4 0 A128 2.6 1 1 4 0 B274 3.1 1 3 4 0 D273 3 1 3 4 0 D48 4.1 1 4 4 0 D84 4.2 1 3 4 3 D266 4 1 4 4 0 E39 3.6 1 1 4 0 88
FI 4.2 1 3 4 0 F3 3 1 3 4 0 A136 4.9 1 1 4 2 D19 5.2 1 3 4 0 D63 5 1 3 4 3 D269 5 1 0 4 0 F19 4.2 1 1 4 0 A110 5.7 1 1 4 0 A310 6.2 1 1 4 3 A 126 5.6 1 1 4 0 A320 5.3 1 3 4 4 D40 5.9 1 4 4 3 D26 6.3 1 3 4 0 D44 7.8 2 4 0 0 A93 6.8 2 0 0 0 Al 12 7.9 2 4 0 0 A 114 8.5 2 0 0 0 B260 9 2 0 4 0 Al 16 7.2 2 3 0 1 A 104 8.5 2 0 0 0 A 147 6.8 2 4 0 3 A 109 7.1 2 3 0 0 A l 81 7.1 2 0 0 0 A 197 6.6 2 4 0 0 A205 7.1 2 0 0 0 Al 15 7.5 2 4 0 0 B127 7.1 2 1 0 0 A53 6.7 2 0 0 0 D258 7 2 1 0 2 D263 6.6 2 1 0 0 D279 7 2 1 0 0 D280 6.6 2 3 0 0 A ll 7.5 2 4 0 0 Al 11 8 2 0 0 3 A 148 7.3 2 1 0 0 A178 7.9 2 0 0 0 A82a 7.3 2 0 0 0 A193 7.9 2 4 0 0 A 199 7.2 2 4 0 0 A200 7.7 2 0 0 0 A307 8.1 2 4 0 0 A99 6.6 2 1 0 0 D14 7.3 2 0 0 0 89
D55 8 2 1 0 0 D64 8.1 2 4 0 0 D250 8 2 4 0 0 D251 8 2 1 0 0 D252 7.6 2 1 0 0 m u 7.7 2 1 0 1 D281 8 2 1 0 0 D286 8 2 0 0 0 F6 7.4 2 0 0 0 A139 8.8 2 1 0 0 A210 8.3 2 0 0 0 A217 9 2 1 0 0 A222 9.2 2 3 0 0 Al 54 8.6 2 4 0 0 A 163 9 2 0 0 0 MIA 9.2 2 0 0 4 Al 82 9 2 4 0 0 A285 9.1 2 4 0 0 A183 8.3 2 4 0 0 A309 8.8 2 0 0 0 A191 8.9 2 1 0 0 D67 9.1 2 1 0 0 D70 9.3 2 1 0 0 D284 7.7 2 4 0 0 D287 9 2 0 0 0 D289 8.8 2 1 0 3 F53 7.2 2 0 0 0 F58 7.3 2 4 0 0 F1278 8.2 2 0 0 0 A153 10.2 2 1 0 0 A219 9.5 2 0 0 3 A168 10.3 2 4 0 0 A 196 10 2 4 0 0 B262 10 2 4 0 0 D43 10 2 1 0 3 A 186 8.3 2 1 3 3 A 123 6.6 2 3 4 4 A101 7 2 1 4 3 A 144 6.6 2 3 4 0 A 102 8.1 2 3 4 0 D69 7 2 3 4 0 D283 7.1 2 1 4 0 D292 7 2 1 4 0 90
A306 7.2 2 1 4 0 D74 7.2 2 0 4 o ; A151 8.9 2 0 4 0 A233 9.1 2 0 4 0 D75 9 2 1 4 0 D291 9 2 0 4 0 F7 8.9 2 0 4 0 Al 56 10.1 2 4 4 0 A269 9.9 2 1 4 0 A281 9.4 2 0 4 2 A 180 10 2 4 4 0 A321 9.4 2 1 4 3 D184 10.2 2 4 4 0 D294 9.5 2 4 4 0 F16 10.4 2 4 4 0 A 124 12 3 0 0 0 A213 12 3 0 ! 0 1 Al 57 10.6 3 4 0 4 A313 11.5 3 4 0 0 A319 11.4 3 0 0 0 B44 11.2 3 0 0 0 A211 10.9 3 0 0 0 C l 96 11.8 3 0 0 3 D54 11.5 3 1 0 0 D239 11.3 3 0 0 0 D282 10.9 3 0 0 0 D296 11 3 4 0 0 B265 12 3 0 0 4 B286 13 3 4 0 4 C14 13.5 3 4 0 4 C253 12 3 4 0 1 B158 13 3 0 0 0 B204 12 3 0 0 0 D65 12.8 3 4 0 0 E19 13 3 1 0 0 A273 10.6 3 1 4 1 A158 10.8 3 4 4 4 D32 11.1 3 1 4 0 A215 13.5 3 4 4 4 A 192 15 4 4 0 0 A 194 15 4 0 0 0 A237 16.5 4 0 0 0 A284 15 4 0 0 2 91
C274 15 4 1 0 0 B185 15 4 4 0 0 D205 16 4 0 0 0 D207 15 4 0 0 1 D236 15 4 0 0 0 D267 15 4 1 0 1 D275 15 4 4 0 0 C68 15 4 0 3 4 A 160 15 4 4 4 3 C281 15 4 4 4 3 C291 15 4 1 4 0 D56 15 4 0 4 0 D268 15 4 0 4 1 D297 15 4 0 4 3 C292 15 4 0 5 4 C78 18.2 5 0 0 0 B 113a 18 5 4 0 0 B177 18 5 3 0 4 C25 18 5 0 0 4 C70A 18 5 4 0 3 C83a 18 5 0 0 0 Cl 10 18 5 0 0 0 C l 13 18 5 0 0 0 Cl 34 18 5 1 0 3 D20 18 5 0 0 4 D142 18 5 3 0 0 D227 18 5 3 0 3 E38 18 5 0 0 4 B54 18 5 0 4 4 C36 18 5 0 4 0 D223 18 5 0 4 3