Pathophysiological Stable Isotope Fractionation

0 downloads 0 Views 4MB Size Report
(Chowang et al. 1988; Soliman et al. ...... gratitude goes to Dr. Jason Laffoon (VU Amsterdam ...... Sarrai, Mona, Herold Duroseau, Jean D'Augustine, Sabita Mok-.

Bioarchaeology International Volume 2, Number 2: 117–146 DOI: 10.5744/bi.2018.1021

Pathophysiological Stable Isotope Fractionation: Assessing the Impact of Anemia on Enamel Apatite δ18O and δ13C Values and Bone Collagen δ15N and δ13C Values Gina M. A. Carroll,a* Sarah A. Inskip,b and Andrea Waters-­Ristc Department of Anthropology and Archaeology, University of Calgary, 2500 University Rd. NW, Calgary, AB, Canada b McDonald Institute for Archaeological Research, University of Cambridge, UK c Department of Anthropology, Western University, 1151 Richmond St. London, ON, Canada *Correspondence to: Gina Carroll, Department of Anthropology and Archaeology, University of Calgary, 2500 University Rd. NW, Calgary, AB T2N 1N4, Canada e-­mail: [email protected]​­.ca a

ABSTRACT

Within the past quarter century, researchers have taken steps to understand pathophysiological stable isotope fractionation within mammalian tissues more accurately. Biomedically, researchers have demonstrated that pulmonary disease, smoking, organ failure, anemia, anorexia, and changes in metabolic rate all affect the isotopic composition of human tissues and tissue by-­products. This research strongly suggests that a relationship exists between human (patho)physiology and stable isotope biochemistry. Despite the results achieved by these studies, only a small minority of bioarchaeologists have attempted to elucidate these mechanisms in human skeletal and dental tissues. This research presents the results of a pilot study aimed at examining the degree to which bone collagen δ13C and δ15N values and enamel apatite δ18O and δ13C values vary between individuals with and without lesions indicative of a chronic anemia. Consistent with previous research, our results indicate that the enamel apatite of suspected anemics have significantly lower δ18O values relative to their lesion-­free counterparts (U = 4.00, p = 0.05); however, this result was limited to the first permanent molar. Due to the small sample size and the lack of information concerning breast-­feeding and weaning practices in the region during this time, it is not possible to link this variation definitively to the pathophysiology of anemia and/or its sequelae. There was no significant variation in bone collagen δ13C or δ15N values between anemic and lesion-­free juveniles (δ13C: U = 26.00, p = 0.38; δ15N: U = 33.00, p = 0.85) or between anemic and lesion-­free adults (δ15N: U = 2.70, p = 0.26; δ13C: U = 4.57, p = 0.10). A number of intrinsic and extrinsic factors may have contributed to the lack of variation. While sample sizes are small, the data indicate that future analysis is warranted. Keywords: stable isotope analysis; anemia; Spain



En el último cuarto de siglo, los investigadores han tomado medidas para comprender con mayor precisión Fraccionamiento de isótopos estables fisiopatológicos en tejidos de mamíferos. Biomédicamente, los investigadores han Demostró que la enfermedad pulmonar, el tabaquismo, la insuficiencia orgánica, la anemia, la anorexia y los cambios metabólicos. La velocidad de todos los efectos de la composición isotópica de los tejidos humanos y los productos derivados del tejido. Esta investigación fuertemente sugiere que existe una relación entre la fisiología humana (patho) y la bioquímica de isótopos estables. A pesar de los resultados logrados por estos estudios, solo una pequeña minoría de bioarqueólogos ha intentado Elucidar estos mecanismos en los tejidos humanos esqueléticos y dentales. Esta investigación presenta los resultados de un estudio piloto destinado a examinar el

Copyright © 2018 University of Florida Press

Received 8 February 2018 Revised 20 May 2018 Accepted 2 July 2018

Pathophysiological Stable Isotope Fractionation

118

grado en que el colágeno óseo valores δ13C y δ15N, y el esmalte los valores de apatito δ18O y δ13C varían entre individuos con lesiones indicativas de anemia crónica, y aquellos sin De acuerdo con investigaciones anteriores, nuestros resultados indican que el apatito de esmalte de las anémicas han agotado significativamente los valores de δ18O en relación con sus contrapartes libres de lesiones (U = 4.00, p = 0.05), sin embargo esto se limitó al primer molar permanente. Debido al pequeño tamaño de la muestra y la falta de información sobre las prácticas de lactancia materna y destete en la región durante este tiempo, no es posible vincular definitivamente esta variación a la fisiopatología de la anemia y / o sus secuelas. Había no hay variación significativa en los valores de colágeno óseo δ13C o δ15N entre los juveniles anémicos y sin lesiones (δ13C: U = 26.00, p = 0.38; δ15N: U = 33.00, p = 0.85), o adultos anémicos y sin lesiones (δ15N: U = 2.70, p = 0,26; δ13C U = 4.57, p = 0.10). Varios factores pueden haber contribuido a la falta de variación. Se requiere más investigación con tamaños de muestra más grandes y una estrategia de muestreo más refinada.

Until recently, it was widely accepted that the isotopic composition of bone collagen and enamel apatite reflect the isotopic composition of foods consumed during the period of tissue formation (Ambrose and Norr 1993; DeNiro and Epstein 1978; Schwarcz and Schoeninger 2011; van der Merwe and Vogel 1978). Within the past quarter century, however, researchers have successfully demonstrated that the relationship between diet and the isotopic composition of the consumer’s tissues is mediated by a number of intrinsic and extrinsic factors, including chronic and acute alterations in metabolism secondary to disease stress. For instance, biomedical studies have repeatedly demonstrated that anorexia (Deschner et  al. 2012; Fuller et al. 2005; E. R. Vogel et al. 2012), diabetes (Patel et  al. 2014; Petzke 2015), acquired and genetic anemias (Epstein and Zeiri 1988; Reitsema and Crews 2011; Widory 2004; Zanconato et al. 1992), renal and/ or hepatic failure (Petzke et al. 2006), and other pathological processes (e.g., Katzenberg and Lovell 1999; Olsen et al. 2014; C. D. White and Armelagos 1997) significantly alter the isotopic composition of mammalian tissues and tissue by-­products. While the relationship between stable isotope fractionation and human pathophysiology has a long history in biomedical and related sciences, to date the application of stable isotope analyses in bioarchaeology has largely been limited to modeling past subsistence strategies, breast-­feeding and weaning practices, and patterns of migration/transhumance. As a result, there is a dearth of research examining the effects of systemic disease on the isotopic composition of mineralized tissues from archaeological contexts, and the use of stable isotopes as a biomarker for systemic disease has been largely unexplored. This research presents the results of a pilot study aimed at examining the degree to which bone collagen δ13C and δ15N values and enamel apatite δ18O and δ13C values vary between individuals with and without lesions indicative of a chronic anemia. To date, the authors are aware of only a single study examining the

effects of anemia on the fractionation of stable isotopes in skeletal tissues (Reitsema and Crews 2011). Given the systemic nature of anemia and its sequelae, as well as the hypothesized predominance of anemic conditions in the archaeological past, it is imperative to assess whether, and to what extent, anemia affects the isotopic composition of bone collagen and enamel apatite. Anemia Etiology and pathophysiology

Anemia is a multi-­etiological, pathophysiologically diverse class of over 400 hereditary and acquired red blood cell (i.e., erythrocyte) and hemoglobin disorders. Regardless of their underlying etiology, all anemias are characterized by mild to severe decreases in hemoglobin synthesis and/or erythrocyte production (erythropoiesis). According to contemporary epidemiological estimates, between 1.6 and 2.0 billion people (roughly 30% of the global population) are anemic (McLean et  al. 2009; World Health Organization [WHO] 2017). It is estimated that comparable (relative) rates of anemia may have occurred in the archaeological past,1 largely due to the consumption of nutrient-­ deficient and/or nutrient-­inhibiting diets, the exchange of anemic loci due to genetic admixture, human expansion into regions where anemia-­inducing vectors were endemic, and/or the modification of tropical and subtropical landscapes for agricultural purposes (Roberts and Manchester 2007). The most common cause of anemia worldwide, and the most widely cited cause of anemia in the archaeological record, is iron-­ deficiency anemia (IDA) (Miller 2013; WHO 2017). Iron is an enzymatic catalyst and key component of protein synthesis, oxidative energy production, DNA/ RNA replication and repair, and cell cycle progression. 1. From the Neolithic Revolution onward (Ulijaszek 1991).

Carroll et al.

The progressive depletion and exhaustion of iron stores can therefore result in a number of systemic issues, including mild to severe decreases in hemoglobin synthesis and erythropoiesis (i.e., anemia), decreased cellular energy metabolism, defective linear growth, and increased cellular hypoxia and oxidative stress (Chowang et al. 1988; Soliman et al. 2014; Toxqui and Vaquero 2015). Aetiological risk factors for IDA include chronic dietary inadequacy, malabsorption, overt and covert blood loss (e.g., gastrointestinal bleeding, parasitic infection, menstruation), and chronic inflammation (Miller 2013; Özdemir 2015; Theurl et al. 2009). Given their increased iron requirements, the prevalence of IDA is greatest in women of childbearing age (~12 to ~49  years), pregnant women, and preschool-­age children (≥6 months and ≤5 years), although the proportion of affected individuals varies substantially within and between regions, and according to local conditions (Miller 2013; WHO 2017). The two most prevalent hereditary anemias worldwide, and the most commonly reported examples of genetic anemias in the archaeological record, are sickle-­cell disease (SCD) and the thalassemia syndromes. SCD is a group of autosomal recessive disorders caused by a point mutation in the HBB gene. In individuals with sickle-­cell anemia (SCA), the most commonly cited example of SCD, both β-­globin polypeptides are replaced with hemoglobin S (genotype HbSS) (Table  1). Unlike normal adult hemoglobin (variant HbA) (see footnote 1 in Table 1), hemoglobin S contains a valine rather than a glutamic acid at the sixth position of the β-­globin polypeptide chain. Due to the non-­polar, hydrophobic properties of valine, hemoglobin S has a tendency to form insoluble fibrous precipitates when deoxygenated (Barabino et al. 2010; Fronticelli and Gold 1976; Wishner et  al. 1975). This process results in the circulation of deformed, dehydrated erythrocytes that are hypercoagulable and prone to hemolysis (Barabino et al. 2010; Fronticelli and Gold 1976; Wishner et al. 1975) (Table 1). Individuals who inherit a single copy of the mutated SCD gene develop sickle-­cell trait (genotype HbS), a typically benign condition (John 2010; Tantawy 2014) (Table 1). The thalassemia syndromes are a heterogeneous class of autosomal recessive hemoglobin disorders caused by mutations or deletions of one or more of the genes encoding for globin-­polypeptide2 synthesis (Hilliard 2. Globin polypeptides are the protein subunits of hemoglobin. Normally, healthy hemoglobin molecules consist of four prosthetic heme (iron) groups and four globin subunits. In fetuses and neonates, the majority of hemoglobin is composed of two alpha (α) and two gamma (γ) polypeptides (HbF hemoglobin). Over the course of ontogeny, HbF hemoglobin is largely replaced by adult hemoglobin. Roughly 95% of adult hemoglobin

119

and Berkow 1996; Thein 2005). This results in the deficient or absent production of the affected globin polypeptide(s)—­most commonly the alpha (α)-­ or beta (β)-­globin chains (Hilliard and Berkow 1996; Thein 2005). The resulting syndromes range from clinically asymptomatic to incompatible with fetal life, depending on the underlying mutations (Table 1). Diagnosing and differentiating anemias in the archaeological record

Due to the wide variation in clinical expression and the lack of pathognomonic features, diagnosing and differentiating between classes of anemia within the archaeological record can be a challenging endeavor. To date, diagnoses primarily rely on the presence and distribution of characteristic osteological lesions (Table 2). The most commonly diagnosed symptoms of anemia in skeletal populations are cribra orbitalia (CO) (Fig. 1) and porotic hyperostosis (PH) (Fig. 2) (Angel 1966; Balikar et al. 2013; Lagundoye 1970; Sebes and Diggs 1979; Stuart-­Macadam 1987). While debate surrounding their aetiological relationship continues (Rivera and Lahr 2017; Walker et al. 2009), it is widely accepted that both conditions result from long-­term elevations of erythropoietin in response to cellular hypoxia and hypoxemia (McGee et al. 2012). Erythropoietin increases the cycling of hematopoietic stem cells and stimulates the proliferation, differentiation, and maturation of immature erythrocytes (McGee et al. 2012). It also stimulates osteoblastogenesis and osteoclastogenesis by acting on bone marrow stromal cells, mesenchymal stem cells, and hematopoietic progenitor cells (McGee et al. 2012). The persistent overstimulation of erythropoietin, as seen in cases of severe anemia, results in compensatory erythroid hyperplasia, marrow hyperplasia, and/or marrow hypertrophy (Shawky and Kamai 2012). Initially, affected elements3 demonstrate signs of medullary expansion, laminal thinning, and osteopenia/ osteoporosis, typically only visible upon radiographic examination (Glader 1999; Shawky and Kamai 2012; Yochum and Rowe 2005). As the hyperplastic erythroid response becomes increasingly chronic, the trabeculae of the diploë begin to grow radially or is composed of two α-­globin and two β-­globin polypeptides (variant HbA). Approximately 4% of adult hemoglobin is composed of two α-­globin and two delta-­(δ) globin polypeptides (variant HbA2). The remaining ~1% of circulating hemoglobin is HbF hemoglobin (Barabino et al. 2010). 3. Due to the changing distribution of hematopoietic marrow over the course of ontogeny, lesions typically appear in the orbits and on the cranial vault of children and on the vertebrae, sternum, ribs, and axial skeleton of adults (Halvorsen and Bechensteen 2002; Hoffbrand and Lewis 1981; Walker et al. 2009).

Deletion or deactivation of one α-­g lobin gene (chromosome 16)

Deletion or deactivation of two α-­g lobin genes (chromosome 16)

Deletion or mutation of three α-­g lobin genes (chromosome 16)

Deletion or mutation of all four α-­g lobin genes (chromosome 16)

Deletion or mutation of one β-­g lobin gene (chromosome 11)

Deletion or mutation of both β-­g lobin genes (chromosome 11); partial β-­g lobin synthesis remains

Deletion or mutation of both β-­g lobin genes (chromosome 11); no β-­g lobin synthesis

Substitution mutation (valine replaces glutamic acid at sixth position in amino acid sequence) in one β-­g lobin gene (chromosome 11)

Substitution mutation (valine replaces glutamic acid at sixth position in amino acid sequence) in both β-­g lobin genes (chromosome 11)

α-­t halassemia silent carrier

α-­t halassemia trait

HbH (β4) disease

Hb Bart’s (γ4) hydrops fetalis

β-­t halassemia minor (trait)

β-­t halassemia intermedia

β-­t halassemia major

Sickle-­cell trait

Sickle-­cell anemia

HbSS

HbS

βo/βo

β+/β+ βo/β+

β+/β βo/β

-­/-­ -­/-­

-­/-­ -­/α

-­/-­ α/α or -­/α -­/α

-­/α α/α

Genotype

Moderate to severe normochromic, normocytic hemolytic anemia

No anemia or erythrocyte abnormalities, although sickling may occur under a select set of circumstances

Severe microcytic, hypochromic, hemolytic anemia; erythrocyte fragments and abnormalities visible microscopically; HbA2 and HbF elevated, HbA decreased or absent

Moderate microcytic, hypochromic anemia; erythrocyte abnormalities visible microscopically; HbA2 and HbF elevated, HbA decreased or absent

Mild microcytic, hypochromic anemia; erythrocyte abnormalities visible microscopically; HbA2 and HbF elevated

Severe, fatal microcytic, hypochromic anemia; erythrocyte fragments visible microscopically; 100% of hemoglobin is composed of γ-­g lobin tetrameres and hemoglobin

Moderate to severe microcytic, hypochromic anemia; erythrocyte fragments visible microscopically; 5–­30% of hemoglobin is composed of γ-­g lobin tetrameres

Mild microcytic, hypochromic anemia; erythrocyte fragments visible microscopically; 4–­6% of hemoglobin is composed of γ-­g lobin tetrameres

No anemia or erythrocyte abnormalities, although mild microcytosis and hypochromia may be present; 1–­2% of hemoglobin is composed of γ-­g lobin tetrameres

Laboratory Features

Chronic hemolysis, increased erythrocyte coagulation, hypoxia/hypoxemia, chronic inflammation, increased oxidative stress, hypermetabolism, increased protein catabolism and protein turnover

Essentially asymptomatic, although symptoms may develop as a result of rigorous exercise, severe dehydration, or altitude hypoxia

Transfusion-­dependent anemia, chronic hemolysis, iron overload resulting in chronic multi-­organ damage, hepatosplenomegaly, endocrine abnormalities, hypermetabolism, increased protein catabolism, and protein turnover

Symptoms vary according to underlying mutation and degree of β-­g lobin production. Clinical features are more severe than beta-­t halassemia minor and less severe than β-­t halassemia major

Essentially asymptomatic

Severe hypoxia/hypoxemia, chronic hemolysis, pleural and pericardial effusion, ascites, skin edema; death in utero or shortly after birth

Chronic hemolysis, hypoxia/hypoxemia, jaundice, gallstones, hepatosplenomegaly; transfusions may be required

Symptoms generally mimic those of iron-­deficiency anemia

Essentially asymptomatic

Clinical Features

Between the ages of 4 months and 2 years

Between the ages of 2 and 6 years, if any

Between the ages of 4 months and 2 years

Between the ages of 2 and 6 years

Between the ages of 2 and 6 years, if any

In utero/birth

In utero/birth

Birth, if any

N/A

Age at Symptom Onset

≤5 years (±12 months)**

Normal life expectancy

≤3 years (±24 months)

≥2 years and ≤30 years*

Normal life expectancy

Death occurs in-­utero or shortly after birth

≤30 years

Normal life expectancy

Normal life expectancy

Life Expectancy without Medical Intervention

*Life expectancy varies according to the underlying mutation and the severity of the symptoms. **There are no reported life-­expectancy estimates for untreated sickle-­cell anemia. The age estimates for poorly managed sickle-­cell anemia range greatly between studies. Quinn et al. (2004) estimated a median survival age of 5.6 years, Diggs (1973) estimated a median survival age of 14.3 years (with 20% of deaths occurring in the first 2 years of life), and Lobo et al. (2013) estimated a median age of 16 years. The authors consider the lower age range to be more reflective of the life expectancy of children without access to modern medical intervention. References cited: Charache 1990; Galanello and Origa 2010; Thein 2005.

Molecular Basis

Variant

Table 1.  Characteristics of Sickle-­Cell Disease and the Thalassemia Syndromes.

Expansion and thickening of the diploic space; atrophy of the outer table Rodent-­face (depressed or collapsed nasal bridge, zygomatic flaring, maxillary prognathism, dental protrusion, frontal and/or parietal bossing) Increase in the meningeal and/or vascular channels

Reduced pneumatization of the frontal, maxillary, sphenoid and/or mastoid

Expansion and thickening of the diploic space; atrophy of the outer table

Brodie syndrome (mandibular arch telescoped within maxillary arch)

Rodent facies (depressed or collapsed nasal bridge, zygomatic flaring, maxillary prognathism, dental protrusion, malar prominence, frontal and/or parietal bossing)

Dental pathologies (periodontitis, shortened crowns and roots, increased prevalence of dental caries, taurodontism)

Appendicular skeleton

Expansion and thickening of the diploic space

Pneumatization of the paranasal sinuses; marrow hyperplasia in the frontal, temporal and facial bones

Extra-­medullary hematopoiesis Focal and/or systemic osteopenia or osteoporosis Cortical thinning and medullary expansion Premature epiphyseal fusion (limb length discrepancies, growth faltering) Osteitis, osteomyelitis, septic arthrosis and/or thalassemia osteoarthropathy Ischemia and infarction (epiphyseal necrosis, cortical splitting, sclerosis, metaphyseal infarcts)

Extra-­medullary hematopoiesis

Focal and/or systemic osteopenia or osteoporosis

Cortical thinning and medullary expansion

Premature epiphyseal fusion (limb length discrepancies, growth faltering) and/or axial deviation of the limbs

Osteitis, osteomyelitis, septic arthrosis and/or thalassemia osteoarthropathy

Ischemia and infarction (epiphyseal necrosis, cortical splitting, sclerosis, metaphyseal infarcts)

Thinning of the mandibular cortex

Malocclusion (maxillary protrusion, mandibular atrophy, hyperplasia of the anterior maxillofacial structures, premature fusion of occipital sutures)

Hematopoietic marrow hyperplasia and hypertrophy (CO/PH/HE)

Hematopoietic marrow hyperplasia and hypertrophy (CO/PH/HE)

Skull

α-­Thalassemia HbH (β4) Disease15–­22

β-­Thalassemia Intermedia and Major1–­15

Element

Osteitis, periostitis, osteomyelitis and/or septic arthrosis, gout

Osteonecrosis (avascular necrosis or aseptic necrosis; unilateral or bilateral), osteolysis, and/or cartilage destruction

Premature epiphyseal fusion (limb length discrepancies, growth faltering)

Cortical thinning and medullary expansion

Focal and/or systemic osteopenia or osteoporosis

Extra-­medullary hematopoiesis

Dental pathology (necrosis of the dental pulp, periodontitis, increased prevalence of dental caries)

Osteomyelitis of the mandible

Rodent-­face (depressed or collapsed nasal bridge, zygomatic flaring, maxillary prognathism, dental protrusion, frontal and/or parietal bossing)

Expansion and thickening of the diploic space; atrophy of the outer table

Expansion and thickening of the diploic space

Hematopoietic marrow hyperplasia and hypertrophy (CO/PH/HE)

Sickle-­Cell Anemia 1, 15, 23–­26

Table 2.  Characteristic Osteological Manifestations of the α-­ and β-­Thalassemias, Sickle-­Cell Anemia, and Iron-­Deficiency Anemia.

(continued )

Unbalanced bone turnover (increased bone resorption)

Cortical thinning and medullary expansion

Focal and/or systemic osteopenia or osteoporosis

Extramedullary hematopoiesis

Expansion and thickening of the diploic space; atrophy of the outer table

Focal and/or systemic osteopenia or osteoporosis

Hematopoietic marrow hyperplasia and hypertrophy (CO/PH/HE)

Iron-­Deficiency Anemia 27–­29

Focal and/or systemic osteopenia or osteoporosis Osteomyelitis Pitting, thickening, and generalized enlargement of the ribs Cortical thinning Unbalanced bone turnover (increased bone resorption)

Radiographic rib-­w ithin-­rib appearance

Costal osteoma

Cortical thinning and medullary expansion

Unbalanced bone turnover (increased bone resorption)

Radiographic bone-­w ithin-­bone appearance

Bone infarcts

Anterior vertebral vascular notching

Spinal pathology (e.g. Tower, H-­shaped, and/or fish-­mouth vertebral deformities, vanishing vertebrae, kyphosis)

Sub-­periosteal new bone growth and/ or thickening

Pathological fracturing and/or osteochondritis

Dactylitis and/or marked carpal deformities

Unbalanced bone turnover (increased bone resorption)

Sickle-­Cell Anemia 1, 15, 23–­26

Paraspinal extra-­medullary hematopoiesis

Unbalanced bone turnover (increased bone resorption)

Focal and/or systemic osteopenia or osteoporosis

Evidence of endocrine dysfunction (including diabetes, hypogonadism, and thyroid/parathyroid dysfunction)

Evidence of endocrine dysfunction (including diabetes and thyroid/parathyroid dysfunction)

Pitting, thickening, and generalized enlargement of the ribs

Subperiosteal new bone growth and/or thickening

Sub-­periosteal new bone growth and/or thickening

Pitting, thickening, and generalized enlargement of the ribs

Radiographic evidence of musculoskeletal iron deposition

Radiographic evidence of musculoskeletal iron deposition

Vertebral growth disturbances

Pathological fracturing and/or osteochondritis

Pathological fracturing and/or osteochondritis

Bone infarcts

Biconvexation of the metaphyses and/or epiphyses

Biconvexation of the metaphyses and/or epiphyses

Spinal pathology (intervertebral disc degeneration, ring apophysis, platyspondyly, fracturing)

Unbalanced bone turnover (increased bone resorption)

Delayed osteoid maturation and mineralization; Unbalanced bone turnover (increased bone resorption)

Spinal pathology (e.g.. scoliosis, kyphosis, vertebral collapse, intervertebral disc degeneration, fracturing)

α-­Thalassemia HbH (β4) Disease15–­22

β-­Thalassemia Intermedia and Major1–­15

Unbalanced bone turnover (increased bone resorption)

Cortical thinning

Focal and/or systemic osteopenia or osteoporosis

Iron-­Deficiency Anemia 27–­29

Notes: With the possible exception of mummified fetal and neonatal remains analyzed using high-­resolution computed tomography, it is extremely unlikely that diagnosable cases of α-­t halassemia hydrops fetalis will be visible in archaeological skeletal assemblages. As such, it was not included in this table. In the majority of instances, α-­t halassemia silent carrier, α-­t halassemia trait, β-­t halassemia minor, and sickle cell are relatively asymptomatic. When symptoms do occur, they resemble those of iron-­deficiency anemia. References cited: 1. Hershkovitz et al. 1997; 2. Ortner 2003; 3. Tyler et al. 2006; 4. Lagia et al. 2007; 5. Musallam et al. 2012; 6. Salama et al. 2006; 7. Almeida and Roberts 2005; 8. Karakas et al. 2016; 9. Javid and Said-­A l-­Naief 2015; 10. DeLong and Burkhart 2013; 11. Mahachoklertwattana et al. 2003; 12. Morabito et al. 2004; 13. Perisano et al. 2012; 14. Voskaridou et al. 2009; 15. Yochum and Rowe 2005; 16. Glader 1999; 17. Shawky and Kamai 2012; 18. Galanello and Cao 2011; 19. Kharsa 2008; 20. Origa et al. 2005; 21. Riyanti 2008; 22. Yochum and Rowe 2005; 23. Anand and Glatt 1994; 24. Onwubalili 1983; 25. Serjeant 1993; 26. Embury 1986; 27. Toxqui and Vaquero 2015; 28. Díaz-­Castro et al. 2012; 29. Aksoy et al. 1966.

Axial skeleton

Element

Table 2.  Characteristic Osteological Manifestations of the α-­ and β-­Thalassemias, Sickle-­Cell Anemia, and Iron-­Deficiency Anemia. (continued)

Carroll et al.

123

Figure 1. Orbital lesions consistent with cribra orbitalia. Left: ID #4238, estimated age at death: 18 months ( ±6 months) –2 years (±8 months); middle: ID #5255, estimated age at death: 3 years (±12 months); right: ID #2030, estimated age at death: 2 years (±8 months) –3 years (±12 months) (photographs by G. M. Carroll).

Figure 2.  Cranial lesions consistent with porotic hyperostosis. Left: ID #10957, estimated age at death: ≥35 years–­≤ 50 years; right: ID #10021, estimated age at death: ≥18 years–­≤ 25 years. Photographs by G. M. Carroll.

perpendicularly and the cranial vault and/or orbital roofs become markedly thickened and osteoporotic (Figs. 1–2) (Glader 1999; Ortner 2003; Shawky and Kamai 2012; Stuart-­Macadam 1987; Yochum and Rowe 2005). There may also be evidence of medullary expansion, cortical thinning, and osteopenia/osteoporosis in the axial and/or appendicular skeleton (Fig.  3) (Glader 1999; Ortner 2003; Shawky and Kamai 2012; Stuart-­Macadam 1987; Yochum and Rowe 2005). In certain cases, marrow will extend through the cortex of the Harversian canals, elevating the periosteum and creating fine, radiating linear extensions known as hair-­on-­end trabeculae (Fig. 4) (Ortner 2003). While these hair-­on-­end lesions have been noted in individuals with chronic IDA (Britton et al. 1960; Lanzkowsky 1968), HbH disease (Glader 1999; Shawky and

Kamai 2012), SCA (Baker 1964; Diggs 1973; Hollar 2001; Lagundoye 1970; Moseley 1974; Sebes and Diggs 1979; Williams et  al. 1975), and β-­thalassemia intermedia (Balikar et  al. 2013), the general consensus among clinicians is that hair-­on-­end trabeculae are primarily symptomatic of β-­thalassemia major (Baker 1964; Hollar 2001; Moseley 1974). Differential diagnoses for PH, CO, and hair-­on-­end lesions are noted in Table 3. It is interesting to note that all conditions listed in Table 3 are characterized by or associated with an increased risk of anemia. The diagnosis of SCD and the thalassemia syndromes in skeletal populations is also predicated on the contemporary distribution of sicklic and thalassemic polymorphisms and the historic distribution of malaria, both of which are thought to reflect the relative distribution of SCD and the thalassemia syndromes in the past. It is now widely accepted that SCD and the thalassemia syndromes are maintained in populations who are regularly exposed to malaria as a result of balancing selection (Haldane 1949; Hedrick 2011; Montalento 1949). Individuals who inherit one copy of the defective α and/or β gene (Table 1) proliferate functional erythrocytes with properties that decrease the survivability of malarial parasites and increase the relative fitness of the host (Ferreira et al. 2011; López et al. 2010; Yuthavon and Wilairat 1993). Individuals who inherit two copies of the defective α and/or β genes develop severe anemia and rarely survive into adolescence without modern medical intervention (Table 1) (Diggs 1973; Lobo et al. 2013; Quinn et al. 2004). Since carriers have an increased fitness advantage, the genotype is positively selected for and both polymorphisms are maintained within the affected population. Consequently, with the exception of the Americas, there is extensive overlap between the historic geographic distribution of malaria and the geographic distribution of sicklic and thalassemic polymorphisms (Haldane 1949; Hedrick 2011; Montalento

124

Pathophysiological Stable Isotope Fractionation

Figure 3.  Postcranial lesions consistent with chronic hematopoietic marrow hyperplasia and hypertrophy. Top and middle left: ID #4238, estimated age at death: 18 months (±6 months) to 2 years (±8 months); top right, bottom left, and bottom right: ID #5920, estimated age at death: ≥18 years–­≤ 25years. Photographs by G. M. Carroll.

Figure 4.  Cranial lesions consistent with hair-­on-­end trabeculae, a symptom of chronic marrow hyperplasia and hypertrophy. Left and middle: ID #2134, estimated age at death: 2 years (±8 months) to 3 years (±12 months); right: 11321: 11321, estimated age at death: 2 years (±8 months) to 3 years (±12 months). Photographs by G. M. Carroll.

Carroll et al.

125

Table 3.  Differential Diagnoses for Porotic Hyperostosis, Cribra Orbitalia, and Hair-­on-­End Trabeculae.

Condition

Description

Association with Anemia

Chronic renal failure

Gradual loss of renal function

Commonly results in anemia of inflammation; anemia of inflammation is a risk factor in the development of non-­ hereditary iron-­refractory iron-­deficiency anemia 1–­4

Acute and chronic leukemia

Replacement of healthy lymphoid or myeloid cells with malignant cells

In modern clinical practice up to 90% of patients develop anemia due to defective or deficient RBC production; acquired aplastic anemias may evolve into leukemia6–­10

Polycythemia vera (primary polycythemia)

Chronic myeloproliferative neoplastic disorder characterized by excessive erythrocyte production

Uncommonly associated with microcytic, hemolytic, and megaloblastic anemias; development of polycythemia-­induced myelofibrosis significantly increases risk of anemia 11–­12

Secondary polycythemia

Excessive erythrocyte production secondary to hypoxia, obstructive sleep apnea, obesity hypoventilation syndrome, obstructive pulmonary disease, uterine leiomyomas, renal cysts, meningioma, and certain malignancies (renal cell cancer, cerebellar hemangioblastoma, hepatocellular cancer, parathyroid meningioma)

Uncommonly associated with microcytic and megaloblastic anemias; development of polycythemia-­induced myelofibrosis significantly increases risk of anemia 11–­12

Cyanotic congenital heart disease

Group of congenital cardiac defects resulting in hypoxemia

Associated with an increased risk of anemia as a result of chronic or acute blood loss, arteriovenous malformations or collateral vessels, hemolysis, or reduced hematopoiesis13

Thrombotic thrombocytopenic purpura

Acute, fulminant disorder characterized by thrombocytopenia and microangiopathic hemolytic anemia

Characterized by microangiopathic hemolytic anemia 14

Pancytopenia

Development of anemia, leukopenia, and thrombocytopenia

Characterized by the development of anemia; megaloblastic anemia is a common cause of pancytopenia 15

Chronic protein deficiencies (marasmus & kwashiorkor)

Severe deficiency in dietary proteins, carbohydrates, and lipids (marasmus); severe deficiency in dietary proteins (Kwashiorkor)

Commonly associated with the development of protein-­ deficiency anemia; chronic protein deficiency results from similar proximate causes as iron-­deficiency anemia and megaloblastic anemias, and one or both may co-­occur16

Chronic hypovitaminosis D

Depletion or exhaustion of vitamin D stores (body pool ≤20 ng/mL)

Associated with an increased risk of anemia, particularly anemia of inflammation; anemia of inflammation is a risk factor in the development of non-­hereditary iron-­refractory iron-­deficiency anemia 2–­4, 17

Chronic ascorbic acid deficiency (scurvy)

Depletion or exhaustion of ascorbic acid (vitamin C) stores (body pool ≤350 mg)

Ascorbic acid is required for erythrocyte production and intestinal iron absorption; in modern clinical practice ~80% of patients develop a normocytic, macrocytic, normoblastic, and/ or macronormoblastic anemia 18–­21

Chronic infection/ inflammation*

Prolonged inflammatory response characterized by concomitant tissue destruction and repair

Chronic inflammation/infection results in anemia of inflammation; anemia of inflammation is a risk factor in the development of non-­hereditary iron-­refractory iron-­deficiency anemia 2–­4

*Particularly of the frontal, maxillary, and/or ethmoid sinuses, lacrimal glands, scalp, or cranial vault.22 References cited: 1. Babitt and Lin 2012; 2. D’Angelo 2013; 3. Nairz et al. 2016; 4. Nemeth and Ganz 2014; 5. Alter 2014; 6. Crowley et al. 1985; 7. Matloub et al. 1993; 8. Majumder et al. 2006; 9. Steele and Narendran 2012; 10. Lichtiger and Huh 1985; 11. Strati et al. 2014; 12. Guo et al., 2017; 13. Dimopoulos et al. 2009; 14. Shenkman and Einav 2014; 15. Makheja et al. 2013; 16. Bernát 1983; 17. Smith and Tangpricha 2015; 18. Cox et al. 1962; 19. Cox et al. 1967. 20. Greenham 1989; 21. Hallberg et al. 1989; 22. Ortner 2003.

1949). In contemporary populations, the α-­thalassemias are most common among individuals of Southeast Asian descent, with up to 40% of individuals expressing the carrier genotypes (Lal et  al. 2011; Lau et  al. 1997; D. Li et al. 2006). It is also common among individuals living in the Mediterranean and the Middle East (Weatherall and Clegg 2000). The frequency of the β-­thalassemia genotypes is highest in the Mediterranean basin,4 the Indian subcontinent, Southeast Asia, the Middle East,5 Melanesia, the Pacific Islands, 4. E.g., Spain, Italy, Tunisia, Portugal, Cyprus, Greece, Serbia, Montenegro, Bulgaria, Sardinia. 5. E.g., Qatar, Iran, Iraq, Jordan, Kuwait, Syria, Turkey, Saudi Arabia, Lebanon, Azerbaijan.

and parts of Africa6 (De Sanctis et al. 2017; Kotila et al. 2009; Weatherall and Clegg 2000), while the frequency of SCD is highest in sub-­Saharan Africa, the Middle East, India, and parts of the Mediterranean (Allison 2002; Piel et al. 2010, 2014). It is assumed that the contemporary distribution of these polymorphisms is similar to their historic distribution.7 Given the challenges associated with diagnosing and differentiating between classes of anemia in skeletal samples, as well as the proposed predominance of anemia in the archaeological record, it is imperative to 6. E.g., Egypt, Kenya, Nigeria, Liberia, Ghana, and Ivory Coast. 7. With the exception of the Americas, where there have been no confirmed cases of pre-­contact thalassemia or SCD.

Pathophysiological Stable Isotope Fractionation

126

develop methods to more accurately identify affected individuals. Stable Isotope Fractionation in Biological Systems under Disease Stress

Isotopes are variants of a particular chemical element (e.g., oxygen, nitrogen, carbon) that have the same atomic number and nearly identical chemical behaviors but with different atomic masses and physical properties. Isotopes with more atomic mass due to the presence of one or more additional neutrons are referred to as heavy isotopes, while those with a lower atomic mass are referred to as light isotopes. Both heavy and light isotopes participate freely in biochemical reactions, but because the chemical bonds and attractive forces within heavier isotopes are stronger than those within lighter isotopes, heavier isotopes react less quickly than lighter isotopes. This results in a distinct discrimination effect, known as isotopic fractionation. The fractionation of isotopes is measured using delta (δ) values in parts per mil (‰) using the equation δRsample = {(Rsample/Rstandard) –­1} × 1,000 where R represents the ratio of heavy to light isotopes. The discriminating factor is measured using the equation α = Rproduct/Rreactant. A number of in-­depth reviews of isotopic fractionation are available (e.g., Hayes 1982; Schoeller 1999; J. C. Vogel 1980). Accurately modeling isotopic fractionation in complex multicellular organisms can be a challenging endeavor, as biological systems are affected by a number of intrinsic and extrinsic factors that can alter the rate at which isotopic fractionation occurs. In general, however, since healthy individuals are able to auto-­ regulate and maintain their internal environment through homeostatic processes, they are assumed to be in a relatively steady state (Schoeller 1999). A number of studies have shown, however, that the metabolic and pathophysiological changes associated with chronic and acute illness can alter this steady state, resulting in different fractionation patterns between healthy and diseased individuals (Deschner et al. 2012; Fuller et al. 2005; Gaye-­Siessegger et al. 2007; Mekota et al. 2006; E. R. Vogel et al. 2012). For instance, research has shown that during starvation the metabolic nitrogen pool of affected consumers is enriched with 15N relative to the diet due to shifts in protein and amino acid metabolism and the increased reliance on amino acid catabolization for gluconeogenesis (McCue 2010; Mekota et al. 2006). Over time, the bulk δ15N of these consumers becomes heavily enriched

relative to their non-­nutritionally stressed counterparts (Deschner et al. 2012; Fuller et al. 2005; E. R. Vogel et al. 2012). Since starvation also alters carbohydrate metabolism and increases the reliance on lipid reserves (which are depleted in 13C), the bulk δ13C values of nutritionally stressed consumers have been shown to be enriched relative to their diets (Doi et al. 2017; Gaye-­Siessegger et al. 2004a, 2004b). Similar discrimination effects have been noted within and between the tissue and tissue by-­products of consumers with renal and/or hepatic failure (Petzke et al. 2006), diabetes (Patel et al. 2014; Petzke 2015), osteoporosis (C. D. White and Armelagos 1997), non-­specific bone infection and inflammation (Katzenberg and Lovell 1999; Olsen et  al. 2014), osteoarthritis (Olsen et  al. 2014), bone atrophy (Katzenberg and Lovell 1999), lung disease (Epstein and Zeiri 1988), and anemia (Epstein and Zeiri 1988; Reitsema and Crews 2011; Widory 2004; Zanconato et al. 1992). Proposed mechanisms of anemia-­induced fractionation in skeletal tissues: Bone collagen δ15N and δ13C values

In healthy, non-­anemic individuals, the anabolism and catabolism of bone and bone proteins (i.e., collagen) are at equilibrium (Katzenberg and Lovell 1999). Since there are no significant increases or decreases in tissue growth, maintenance, or repair, δ15N values are balanced and should therefore primarily reflect the isotopic composition of dietary proteins (δ15N of diet plus ~3–­5‰) (Ambrose and Norr 1993; Katzenberg and Lovell 1999; Tieszen and Fagre 1993). Clinical and epidemiological research has demonstrated repeatedly, however, that SCD, the thalassemia syndromes, IDA, and other nutritional anemias (e.g., pernicious anemia) decrease bone matrix formation, increase bone resorption, and adversely affect collagen maturation, thereby increasing the risk of osteopenia and osteoporosis (Díaz-­Castro et al. 2012; Rossi et al. 2014; Sadat-­Ali and Elq 2007; Sarrai et al. 2007; Toxqui and Vaquero 2015; Voskaridou and Terpos 2008). Consequently, we should expect to see a shift from a predominantly dietary-­based composition (i.e., steady state) to a mixed dietary-­diseased composition. This idea is supported by research of C. D. White and Armelagos (1997), who demonstrated that the bone collagen of females with osteopenia were significantly enriched with 15N. The authors suggested that the enrichment effect may have been a result of underlying differences in urea excretion and/or renal function (C. D. White and Armelagos 1997), both of which are expected to be altered in individuals with anemia-­ induced osteopenia and osteoporosis. In addition to pathophysiological changes in bone turnover, SCA,

Carroll et al.

β-­thalassemia major, and β-­thalassemia intermedia are commonly associated with malnutrition and anorexia (Fung 2010, 2016; Reid 2013; Tienboon et  al. 1996). As a result, the isotopic values of affected individuals exhibiting these symptoms should mirror those experiencing non-­anemia-­induced starvation. We therefore hypothesize that the bone collagen of individuals with severe manifestations of anemia will be enriched with 15N and 13C relative to their non-­ anemic counterparts. We also hypothesize that with the possible exception of HbH disease and β-­thalassemia intermedia, heterozygotic variants of SCD and the thalassemia syndromes will not have a significant impact on bone collagen δ15N or δ13C values, since they do not result in severe anemia and are not associated with the types of chronic sequelae that may alter isotopic fractionation. Proposed mechanisms of anemia-­induced fractionation in dental tissues: Enamel apatite δ18O and δ13C values

The authors propose that both the rate of cellular metabolism and the rate of total respiration may affect the fractionation of stable oxygen isotopes in anemic dental enamel. Total respiration involves a number of active and passive processes that result in the preferential metabolism of the lighter stable 16O isotope (Hoefs 2004; Widory 2004). This results in a distinct intra-­and inter-­tissue fractionation effect. The slower the fractionating reaction rate, the greater the fractionation effect (i.e., the lighter the isotopic value) (Hoefs 2004; Widory 2004). The first point of isotopic oxygen fractionation during the respiratory process occurs as inhaled oxygen diffuses through the alveolar membranes and into the pulmonary capillaries. During this stage, the rate of 16O diffusion is ~3% greater than the rate of 18O diffusion, leading to an enrichment of 16 O isotopes in the vascular system (Epstein and Zeiri 1988; Heller et al. 1994). The second point of oxygen fractionation occurs as oxygen is fixed by hemoglobin, which preferentially uptakes the 16O isotope (Epstein and Zeiri 1988; Pflug and Schuster 1989). The third point of fractionation occurs once oxygen is released from hemoglobin for cellular respiration, as mitochondria metabolize 16O isotopes ~1.3% faster than 18O isotopes (Feldman et  al. 1959; Heller et  al. 1994). The cumulative effect of isotopic oxygen fractionation during the process of total respiration results in the δ18O values of respired aliquots of air (i.e., oxygen used by the body) to be lighter than that of the atmosphere. The δ18O value of enamel apatite is primarily a function of the isotopic composition of drinking water

127

and the water bound to food (Fricke et al. 1995; C. D. White et al. 1998; Wright and Schwarcz 1998). Unlike drinking water, which is not fractionated upon absorption, water bound to food is only freed following catabolism of the food product (Bryant and Froelich 1995). The rate at which this occurs depends on the overall efficiency of the gastrointestinal tract as well as the composition of the food itself (e.g., complex vs. non-­complex carbohydrates, proteins vs. plant products) (Bryant and Froelich 1995). Once catabolized, the oxygen of foodstuffs is released into the bloodstream as carbon dioxide and water, where they are fractionated based on the aforementioned principles. Other factors that influence the rate of fractionation include active and basal metabolic rates, gastrointestinal health, urea output, internal body temperature, and rate of evaporative water loss (Bowen et al. 2009; Bryant and Froelich 1995; Kohn 1996). Decreases in the rate at which these reactions occur increases stable oxygen isotope fractionation, resulting in an enrichment of 16O. The contribution of any given fractionation factor to the overall isotopic value of tissues and tissue by-­ products during the process of total respiration can be exemplified by the formula

α1 α2 O2 → [O2 ] + Hb → Hb [O2 ], k1 k2 where k 1 is representative of the rate at which oxygen diffuses through the pulmonary membranes and k 2 is representative of the rate at which oxygen is fixed to hemoglobin (Epstein and Zeiri 1988). The fractionation factor (i.e., the degree to which fractionation of stable oxygen isotopes will occur) for k 1 is noted as α1, and for k 2 as α2 (Epstein and Zeiri 1988). At homeostasis, the rate of oxygen diffusion (k1) and oxygen fixation (k 2) are more or less comparable. As such the δ18O value of tissues and tissue by-­products represent the combination of α1 and α2 (Epstein and Zeiri 1988). When homeostasis is disrupted, as is the case with anemia, the overall fractionation of stable isotopes will be dictated by either the rate of oxygen diffusion or oxygen fixation (Epstein and Zeiri 1988). Given the pathophysiology of anemias, the rate of oxygen fixation is more often the inhibited process because anemics have fewer optimally functioning hemoglobin molecules and/or fewer overall erythrocytes. The slower the rate of oxygen fixation, the greater the fractionation effect (Epstein and Zeiri 1988; Widory 2004). As such, we can expect that individuals with lower concentrations of hemoglobin will have overall lower δ18O values relative to their non-­ anemic counterparts, because the rate at which they fix oxygen is significantly decreased. This has been

128

confirmed by previous research (Epstein and Zeiri 1988; Widory 2004; Zanconato et al. 1992). We can therefore also hypothesize that the more severe the anemia, the more significant the fractionation effect. A second mechanism that may cause or contribute to lower δ18O values in anemics is the rate of cellular metabolism. Traditionally, it has been assumed that, since intracellular water is in osmotic equilibrium with extracellular water, the isotopic composition of intracellular water will be isotopically identical to that of the extracellular water (Chikaraishi et  al. 2004; Sessions et al. 1999), which is itself a function of the isotopic composition of drinking water and water bound to food (Fricke et al. 1995; C. D. White et al. 1998; Wright and Schwarcz 1998). Research by Kreuzer-­ Martin et al. (2005) has demonstrated, however, that the intracellular water of E. coli is isotopically depleted relative to the water of the medium in which they were grown. Their results indicate that up to 70% of intracellular water may be derived from metabolic processes and that the degree to which the isotopic composition of intracellular and extracellular water vary depends on the metabolic state of the organism (Kreuzer-­Martin et  al. 2005). This research is supported by previous work demonstrating the effects of metabolic rate, body size, and/or activity patterns on isotopic oxygen fractionation (e.g., Bryant and Froelich 1995; Epstein and Zeiri 1988; Feldman et al. 1959; Heller et al. 1994; Lane and Dole 1956; Pflug and Schuster 1989; Schuster et  al. 1994; van Dam et  al. 2004; Widory 2004; Zanconato et al. 1992). Clinical research has repeatedly demonstrated that SCD and the thalassemia syndromes (particularly SCA, β-­ thalassemia major, β-­thalassemia intermedia, and/or HBH disease) induce states of chronic hypermetabolism (Badaloo et al. 1989; Barden et al. 2000; Borel et al. 1998a, 1998b; Buchowski et al. 2001; Salman et al. 1996; Singhal et al. 2002; Vaisman et al. 1995). If the isotopic composition of intracellular water is indeed a function of metabolic activity, as research suggests, this strengthens our hypothesis that the tissues of individuals with SCA, β-­thalassemia major, β-­thalassemia intermedia, and/or HBH disease will exhibit depleted δ18O values relative to their non-­anemic counterparts. It is widely accepted that the δ13C value of enamel apatite is a function of dissolved inorganic carbon derived primarily from carbohydrates and lipids and to a lesser extent from dietary proteins (Hedges 2003; Tieszen and Fagre 1993; Zazzo et al. 2010). As mentioned previously, the δ13C values of severely anemic individuals is expected to be more enriched in 13C relative to their non-­anemic counterparts due to the increased reliance on lipid reserves and altered carbohydrate metabolism. However, in order for enrichment to occur, anemic individuals would have had to

Pathophysiological Stable Isotope Fractionation

have survived long enough to develop anorexia or other nutritional issues. The authors propose that hemolysis may also contribute to the pathophysiological fractionation of stable carbon and oxygen isotopes in the enamel apatite of individuals with hemolytic anemias. On average, sicklic and thalassemic red blood cells (RBCs) lyse 90 to 100 days earlier than healthy RBCs, resulting in a reduction of up to 50% in carbon dioxide elimination (Bidani and Crandall 1982; Koury 2014). This, in turn, increases the concentration of dissolved carbon dioxide and bicarbonates in the blood, resulting in acidosis and decreased oxygen fixation (Bidani and Crandall 1982). According to Epstein and Zeiri (1988), this process should result in lower δ18O values. The contribution of hemolysis to stable oxygen fractionation has been largely unexplored. Using a modified Epstein and Zeiri (1988) equation for blood carbons, it is possible to deduce that the overall fractionation of stable carbon isotopes during the process of total respiration will be controlled by (1) the relative rate of blood carbon anabolism and catabolism, (2) the relative rate of blood carbon fixation by hemoglobin, or (3) the rate of all reactions at equilibrium. Since hemolysis increases the rate of carbon dioxide fixation, we hypothesize that the overall fractionation effect (during the process of total respiration) will be determined by the speed of that reaction. The authors hypothesize, however, that this effect will not be significant enough to alter any underlying dietary signature. Materials and Methods Site selection

All individuals in this study were excavated from Plaza de España, an Islamic cemetery located in Écija, Spain. The cemetery contained the remains of over 4,500 men, women, and children, distributed over seven intercut layers. Historical documents and cultural artifacts indicate that the majority of the burials date back to the Umayyad period (eighth to eleventh centuries A.D.). This site was chosen because it contains a relatively high proportion of individuals (both sexes and all age ranges) with clear evidence of severe cranial and postcranial hematopoietic marrow hyperplasia/ hypertrophy (Figs. 1 to 4). As mentioned, hematopoietic marrow hyperplasia/hypertrophy is strongly associated with anemia, particularly IDA, SCA, thalassemia major, β-­thalassemia intermedia, and HBH disease. In addition to the osteological evidence, Plaza de España was chosen because of its location. The town of Écija is located in the Genil River Valley, in modern-­ day Seville. Up until 1961, malaria was endemic to the

Carroll et al.

129

Spanish southwest, and Seville in particular experienced a number of outbreaks during the early to mid-­ twentieth century (Sousa et al. 2009, 2014). The authors are currently unaware of any prevalence estimates for malaria in Seville during the Umayyad period, but given the number of wetlands in the area, the climate, and the compendium of historic documents describing malarial-­like illnesses throughout the country (Freer 1857; Palmero and Vega 1988; Pometti 2016), it is likely that rates of malarial infection were as high or higher than those noted in the twentieth century. This, in addition to the contemporary distribution of sicklic and thalassemic polymorphisms in Spain (Amselem et al. 1988; Arends et al. 1985; Cela et al. 2017; Pereira et al. 2009), provide indirect evidence for the presence of SCD and/or the thalassemia syndromes at the site during this time. There is also strong support for the presence of IDA in this population. Historical sources indicate that Écija produced large quantities of cereals and legumes, particularly wheat, barley, sorghum, millet, lentils, and chickpeas (Burton and Price, 1990; García Sánchez 1992, 1995, 2002; Imamuddin 1981). The consumption of these products is supported by isotopic dietary reconstructions conducted elsewhere in Spain around this time period (Alexander et al. 2015; Guede et al. 2017; Mundee 2009). Legumes and cereals are typically high in phytates, oxalates, and polyphenols (Chai and Liebman 2005; Reddy et al. 1982; Salunkhe et al. 1982), compounds that are known to reduce intestinal iron absorption

and increase the risk of IDA if consumed in large quantities (Brune et al. 1989; Ma et al. 2010). Finally, the inferred prevalence of malaria strongly suggests that malarial anemia would have been present in at least a subset of the population during this time. Overall, the combination of characteristic osteological lesions, the extreme likelihood of endemic malaria, the current distribution of sicklic and thalassemic polymorphisms within Spain, and the dietary practices of the time strongly suggests that one or more types of anemia were present in this population during the Umayyad period. Differential diagnoses and their presumed likelihood of occurrence are noted in Table 4. The authors hypothesize that even if one or more of the individuals with lesions indicative of anemia have been misdiagnosed, affected individuals should still demonstrate lower δ18O values and higher δ15N values relative to the control cohort, since the presence and severity of the lesions indicate that some type of systemic marrow hyperplastic, hypertrophic response was occurring, regardless of its underlying etiology. Sample selection

Previous research determined that aDNA preservation within the collection is poor; as such, it was not possible to diagnose any of the individuals in the study genetically. Consequently, the selection and categorization of individuals was based entirely on the

Table 4.  Differential Diagnoses for the Hyperplastic, Hypertrophic Marrow Response Noted in the Écijan Skeletal Collection.

Condition

Presumed Likelihood

Chronic renal failure

Low in children, medium to high in adults; the risk of renal failure and chronic kidney disease increases with age

Acute and chronic leukemia

Medium to high in children, low in adults; leukemia is the most common cancer among children and adolescents

Polycythemia vera (primary polycythemia)

Low in children, medium in adults; risk of development increases after the age of 40

Secondary polycythemia

Low in children, medium in adults; risk of development increases with age

Cyanotic congenital heart disease

Low, particularly in the adults; affected individuals are unlikely to have survived into childhood or adolescence without modern medical intervention

Acquired and hereditary thrombotic thrombocytopenic purpura

Low; the condition is extremely rare and aggressive; affected individuals are unlikely to have survived long enough to have developed the severe hyperplastic/hypertrophic lesions noted

Pancytopenia

Medium, depending on the underlying condition

Chronic protein deficiencies (marasmus and kwashiorkor)

Low; historic documents support the consumption of mutton and other meat products, and there are no reports of famine in the area during this time

Chronic hypovitaminosis D

Low; sun exposure is expected to have been sufficient to prevent hypovitaminosis D; no evidence of rickets or osteomalacia as observed in the skeletal collection

Chronic ascorbic acid deficiency (scurvy)

Low; dietary practices would have likely been sufficient to prevent ascorbic acid deficiency

Chronic infection/inflammation

Medium to high; malaria, which is presumed to have been prevalent in this population, invokes a systemic inflammatory response; non-­malarial cranio-­facial infections could not be ruled out

Pathophysiological Stable Isotope Fractionation

130

Table 5.  Estimated Age at Death and Bone Collagen δ 15N and δ 13C Values for Juveniles with Lesions Indicative of a Severe Childhood Anemia, and

Those Without. Sample Code

Cohort

Estimated Age at Death

4238 2527 2772 11989 2030 11321 4776 4574 5255 12283 1869 2170 6447 10206 5601 4421 5462

Severe childhood anemia Severe childhood anemia Severe childhood anemia Suspected homozygote Suspected homozygote Severe childhood anemia Severe childhood anemia Severe childhood anemia Severe childhood anemia Severe childhood anemia Lesion-­f ree Lesion-­f ree Lesion-­f ree Lesion-­f ree Lesion-­f ree Lesion-­f ree Lesion-­f ree

18 months (±6 months) –­2 years (±8 months) 18 months (±6 months) –­2 years (±8 months) 2 years (±12 months) –­3 years (±12 months) 2 years (±8 months) –­

Suggest Documents