AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 00:00–00 (2014)

Integrating Cortisol and Isotopic Analyses of Archeological Hair: Reconstructing Individual Experiences of Health and Stress Emily C. Webb,1* Christine D. White,1 Stan Van Uum,2 and Fred J. Longstaffe3 1

Department of Anthropology, The University of Western Ontario, London, ON, Canada Department of Medicine, The University of Western Ontario, London, ON, Canada 3 Department of Earth Sciences, The University of Western Ontario, London, ON, Canada 2

KEY WORDS

stress; cortisol; stable isotopes; mummies; hair

ABSTRACT Archeological hair from 14 adults from the Nasca Region, Peru (c. AD1–1000) was analyzed for carbon and nitrogen isotopic compositions and cortisol levels. We investigated the relationship between isotopic compositions, which reflect diet, and cortisol, which reflects biogenic cortisol production and chronic stress. Using a case study approach, we determined that there are consistent changes in cortisol production associated with the rapid dietary change characteristic of local mobility. Moreover, changes in nitrogen- and carbonisotope compositions, when integrated with cortisol levels, enabled inferences to be made about nitrogen metabolism and carbon routing, and elucidated the nature of potential stressors in the months before death. The isotopic and cortisol data suggested a relatively high rate of exposure to stress that is consistent with what is

known about the Nasca Region social and physical environments. Of the 14 adults included in this study, six likely suffered from illness/trauma before death, and a further three experienced stress without an observable associated change in isotopic composition. Five individuals also experienced increased stress related to local mobility, inferred from co-occurring changes in cortisol production and dietary shifting. The integration of cortisol and isotopic data revealed individual characteristics of hidden frailty and risk that would not be apparent using more traditional methods of evaluating health status. This approach will provide a powerful enhancement to the understanding of stress, morbidity, and well-being developed through skeletal analysis. Am J Phys Anthropol 000:000–000, 2014. VC 2014 Wiley Periodicals, Inc.

Hair retains a sequential archive of physical, environmental, and dietary information, and is a valuable resource for both archeological and clinical research investigating behavior, health, and stress. Carbon- and nitrogen-isotope analysis of hair in archeological contexts has been used to reconstruct seasonal dietary change, residential mobility, and geographic relocation, ante-mortem diet and seasonality of death, and to investigate the metabolic impact of trauma and illness (White, 1993; Schwarcz and White, 2004; Knudson et al., 2007; White et al., 2009; Williams et al., 2011; Knudson et al., 2012; Williams and Katzenberg, 2012; Webb et al., 2013). Because hair can be noninvasively sampled, there is also a sizable body of clinical isotopic research focusing on starvation, dietary habits and nutrition, geographic relocation, pregnancy and specific diseases (O’Connell and Hedges, 1999; Bol and Pflieger, 2002; Fuller et al., 2004, 2005; McCullagh et al., 2005; Petzke et al., 2005, 2006; Mekota et al., 2006; Huelsemann et al., 2009; Mekota et al., 2009; Valenzuela et al., 2011). Cortisol, a hormone produced in response to stress, can also be measured in hair. Clinical studies have measured systemic cortisol levels during pregnancy, and explored the impact of chronic stress, perceived stress, and specific diseases on cortisol production (Kalra et al., 2007; Van Uum et al., 2008; Kirschbaum et al., 2009; Thomson et al., 2009; Gow et al., 2010; Dettenborn et al., 2012; Russell et al., 2012). Cortisol has also been successfully extracted from archeological hair samples and has been used to document individual experiences of stress in the months preceding death (Webb et al., 2010).

For this study, hair samples from 14 adults excavated at Cahuachi and near Huaca del Loro in the Southern Nasca Region of Peru (AD1-1000) were analyzed to determine carbon- and nitrogen-isotope compositions (Webb et al., 2013) and cortisol levels. Good biomolecular preservation has been established through a pilot study of Peruvian mummies from several sites, including the Nasca Region (Webb et al., 2010). The Nasca Region remains in particular were chosen for additional study because of the generally low frequency of stress markers on bones and teeth for previously analysed skeletal material from this context (Kellner, 2002; Cagigao, 2009), suggesting that understanding ante-mortem experiences may be critical to elucidating stress and death in

Ó 2014 WILEY PERIODICALS, INC.

Grant sponsor: Social Sciences and Humanities Research Council of Canada, the Natural Sciences and Engineering Research Council of Canada, the Canada Research Chairs Program, the Physicians’ Services Incorporated Foundation, Canada Foundation for Innovation and Ontario Research Fund. *Correspondence to: Emily Webb, Organic Geochemistry Unit, School of Chemistry, Cantock’s Close, University of Bristol, Bristol BS8 1TS, United Kingdom. E-mail: [email protected] Received 4 August 2014; revised 16 October 2014; accepted 12 November 2014 DOI: 10.1002/ajpa.22673 Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com).

2

E.C. WEBB ET AL.

the Nasca Region. Both isotopic and cortisol data are incorporated sequentially into hair as it grows, and reflect approximately the same period of time for each individual. Combining these datasets has the potential to provide unique insight into the often hidden issues of frailty, risk and ante-mortem stress, which are central to our understanding of health, life, and death in ancient societies. Here, we use a case study approach to demonstrate the interpretive potential of integrating isotopic and cortisol data to explore diet, stress, and health, and to describe individual experiences in the months or years preceding death.

BIOCHEMICAL ANALYSIS OF ARCHEOLOGICAL HAIR Hair growth considerations Hair grows at an average rate of 0.35 mm/day, or 1 cm/month, and is metabolically inactive, retaining a sequential archive of biochemical information (O’Connell and Hedges, 1999). Hair growth rate cannot be determined after death, and has been observed to range from 0.6 to 2 cm per month among modern humans (LeBeau et al., 2011). There is a 14-day lag between hair formation and emergence of the mature hair from the scalp, and this creates a blind spot covering the two weeks immediately preceding death (Nakamura et al., 1982; O’Connell and Hedges, 1999). Hairs that have completed growth are periodically shed, and subsequently replaced by new hairs. Hair grows during the anagen phase (lasting two to three years) and it is during this phase that new information is incorporated. The anagen phase is followed by the catagen, or quiescent, phase, which lasts from one to four weeks, and the telogen phase, which lasts from one to three months. The hair is finally shed when a new anagen phase begins in the follicle, and the old hair is pushed out by the new hair. Typically, 80– 85% of scalp hair is in anagen phase, 1–2% in catagen phase, and 10–20% in telogen phase, but these proportions may change seasonally or with illness (Harding and Rogers, 1999; Schwertl et al., 2003). In a mixedphase hair sample, temporal trends may be obscured and intersegment variation in isotopic composition will be attenuated (Williams et al., 2011); thus, using mixedphase samples risks under-representing the true extent of variability in hair biomolecular data. The proportion of actively growing hairs in a random sample of hair can also vary among individuals (Williams et al., 2011). Differences in cortisol levels and isotopic compositions across bundles of hair from different parts of the scalp would hypothetically occur if the proportions of anagen: catagen: telogen hair differed substantially within each bundle. Finally, Sauve et al. (2007) determined that there were no statistically significant differences among cortisol levels from five different parts of the scalp. The cortisol measurements ranged from 50 to 86.5 pg/mg, which is small relative to the scale of variation we observed for cortisol levels in archeological hair. There is no evidence to suggest intrinsic biological variation in isotopic composition of hair from different parts of the scalp. Biomolecules, including cortisol and protein, are incorporated into the hair fiber as it grows. The cortisol level measured in a 1 cm segment of hair reflects the amount of free cortisol circulating in the body during the period of growth (i.e., roughly one month). Similarly, keratin d13C and d15N values reflect the isotopic composition of American Journal of Physical Anthropology

food consumed during the period of growth. Because hair in anagen phase grows continuously, a change in diet can be detected in hair keratin isotopic compositions within a few days. Variation is dampened to some extent, however, by the buffering effect of incorporating amino acids into keratin from the body protein pool, which will still contain carbon and nitrogen from the ‘old’ diet. On the basis of controlled research with human scalp hair, complete isotopic equilibration of the body protein pool with a new diet may take up to 12 months for carbon, and approximately five months for nitrogen (O’Connell and Hedges, 1999). Because hair grows quickly, shifts in dietary isotopic composition are reflected in changing keratin d13C and d15N values incrementally along the hair shaft. A correlative relationship between carbon- and nitrogen-isotope compositions under conditions of rapidly changing diet has been observed (Webb et al., 2013). Further, the relationship between keratin d13C and d15N values becomes weak during equilibration, since the nitrogen- and carbonisotope compositions of the body pool are not adjusting to the new constant diet at the same rate (Webb et al., 2013). It is possible that there is an offset between cortisol and isotopic data as it is archived in hair, wherein measured cortisol levels in hair change more rapidly than isotopic compositions, which may lag behind by a few days. We contend that the temporal offset between cortisol and dietary information is likely mitigated to some extent by the sampling interval of 1 cm used here, but, in the absence of clinical research, we caution that the temporal match may not be exact.

Carbon- and nitrogen-isotope analysis The hair fiber is composed of dead keratinized cells. Keratin d13C and d15N values reflect the isotopic composition of dietary protein when diet is protein sufficient (Ambrose, 1993; Petzke et al., 2005). Keratin protein is composed of essential amino acids, which must be routed from dietary protein, and nonessential amino acids, which may be routed from dietary protein or synthesized from stored carbohydrates, lipids, or proteins (Schoeller, 1999; McCullagh et al., 2005). The d13C values reflect relative proportions of C3 and C4 foods directly consumed or in the diet of consumed animals. If, however, carbohydrate intake is inadequate, lipolysis (breakdown of stored fats) may be accelerated. Although the relative proportions of C3 and C4 foods consumed would not change, the d13C values could decrease in response to the change in carbon source (i.e., from dietary carbohydrates to body lipids) because body lipids are 13Cdepleted relative to diet (Lee-Thorp et al., 1989; Mekota et al., 2006, 2009). Keratin d15N values are used to assess trophic level of consumed protein, and consumer tissues are always 15N-enriched by 3–4& per trophic level relative to diet (Ambrose, 1993, 2000). Recent bioarcheological and clinical research has demonstrated that d15N values can be significantly influenced by the environment, by metabolic and physiological stress (e.g., pregnancy, nutritional inadequacy), and by pathological conditions (e.g., disease, illness, and trauma) (Heaton et al., 1986; Hobson et al., 1993; Katzenberg and Lovell, 1999; Fuller et al., 2004, 2005; Petzke et al., 2006; Mekota et al., 2006, 2009; Williams et al., 2011; Olsen et al., 2013). In healthy individuals, the rate of protein synthesis equals the rate of protein breakdown and loss, and this

INTEGRATING CORTISOL AND ISOTOPIC DATA FROM HAIR homeostatic state is reflected in body nitrogen balance. This balance responds to changes in amino acid availability (Waterlow, 1999). The metabolic amino acid pool is composed of amino acids derived from dietary protein and catabolized body protein. The body is in positive nitrogen balance when synthesis exceeds breakdown (e.g., during tissue repair after trauma, pregnancy), and newly synthesized proteins will be 15N-depleted relative to diet. In negative nitrogen balance, catabolized body protein amino acids make up a greater proportion of the amino acids available for synthesis, leading to progressive 15N enrichment of the metabolic amino acid pool. Negative nitrogen balance is induced by stress (e.g., infection, fever, anxiety, inadequate diet; Hobson et al., 1993; Schoeller, 1999). The progressive enrichment of the metabolic amino acid pool during negative nitrogen balance occurs because nitrogen from catabolized body proteins is 15Nenriched relative to diet. Isotopic fractionation (differential partitioning of 15N vs. 14N) is thought to occur during deamination and transamination reactions as nitrogen is transported to the liver for elimination. Isotopically-light nitrogen (14N) is favored for excretion in the form of urea, resulting in an overall enrichment of the metabolic amino acid pool (Hobson et al., 1993; Schoeller, 1999). This is the basis for both the trophic level effect previously described, and the ability to detect metabolic changes associated with stress via nitrogen-isotope analysis.

Cortisol The glucocorticoid hormone cortisol is produced in response to real and perceived stress, and it plays a key role in mediating the body’s physiological stress response (O’Connor et al., 2000; Davenport et al., 2006; Miller et al. 2007; Van Uum et al., 2008; Novak et al., 2013). Clinical research has demonstrated that hair cortisol levels reflect stress-induced changes in cortisol production by the hypothalamic-pituitary-adrenal axis (Davenport et al. 2006; Kalra et al., 2007; Van Uum et al., 2008; Kirschbaum et al., 2009). In clinical contexts, cortisol in blood plasma, saliva, or urine is typically measured at specific points in time over short experimental time periods, whereas hair, as a continuously growing tissue, is wellsuited to longer term retrospective studies of stress. The primary metabolic effect of cortisol is to promote gluconeogenesis, which mobilises stored energy from adipose and body proteins when dietary sources and glucose reserves are depleted. Cortisol also accelerates both proteolysis and lipolysis, and modulates the immune system by directing the duration and magnitude of the inflammatory response and lymphocyte maturation (Sapolsky et al., 2000). Cortisol can therefore be seen as an important mechanism through which experienced physiological and psychosocial stresses influence the physical body (Miller et al., 2007). Cortisol levels in hair reflects systemic cortisol levels in the body, and have been widely used in clinical research to retrospectively examine exposure to stress and hormone production (e.g., Kirschbaum et al., 2009; Dettenborn et al., 2012). The preservation of biogenic patterns of cortisol production in archeological hair has also been demonstrated (Webb et al., 2010). Free cortisol is incorporated into hair follicles via the bloodstream, as are other steroid hormones, drugs, and biomolecules. Although there is no evidence to suggest that cortisol, once incorporated, migrates through the hair shaft (Novak et al., 2013), there are several factors that can impact the detec-

3

tion of biogenic cortisol levels, such as leaching, contamination or modification of biogenic cortisol. External contamination from the burial environment is unlikely, since cortisol is only produced by mammals. A possible source of exogenous cortisol may be animal fats used in the manufacture of soaps or shampoos but, if this were occurring, a systematic distribution of contaminant cortisol along the hair shaft or obliteration of intersegmental differences would be expected. The most likely form of diagenetic alteration in an archeological context is leaching. Cortisol is most effectively removed from hair by repeated or prolonged exposure to a solvent (e.g., water or alcohol, Davenport et al., 2006; Novak et al., 2013; Thomson, personal communication), particularly at elevated temperatures. In modern studies, for example, the amount of cortisol retained in hair further than 3 cm from the scalp decreases sharply, which was attributed to loss of cortisol through washing with alcohol-containing shampoos (Kirschbaum et al., 2009). In archeological hair samples, changing levels of cortisol along individual hair bundles therefore most likely reflect biogenic patterns of cortisol production, even if the absolute amount of cortisol may be somewhat lower due to post-mortem loss.

INTEGRATING ISOTOPIC AND CORTISOL DATA Dietary shifting and psychosocial stress For an essentially healthy individual (i.e., not suffering significant physiological and/or metabolic perturbation) consuming a nutritionally adequate diet, carbon-, and nitrogen-isotope compositions may vary minimally or seasonally. Among ancient Peruvians, this variation is typically less than 2.7& in d13C over the length of the hair sample, or is in a fashion consistent with acquiring foods from multiple production zones, either by movement of food or local mobility (Williams et al., 2012; Webb et al., 2013). Diurnal variations in cortisol production, including age- and sex-related differences (Van Cauter et al., 1996; Kudielka et al., 2004), are subsumed within the sampling resolution for this study. Clinical research suggests that there may be seasonal variation in cortisol production (Hansen et al., 2001; Matchock et al., 2007), but the extent to which such patterning may be detectable in hair is unknown. Psychosocial and physical stress could cause cortisol production to increase without an associated change in isotopic composition from one month to the next. Social stress, particularly threats to physical integrity, as well as low controllability of stressors and individual psychological responses to stress can all influence cortisol production (Miller et al., 2007). Exploiting multiple production zones through local mobility involves significant movement across the landscape. In the Andes, this movement would have involved the extra effort required for travel up and down altitudinal gradients. Psychosocial and physical stressors related to travel are also likely to contribute to experienced stress as reflected in cortisol levels. Exposure to new pathogens, adaptation to a different social or physical environment, or leaving a stressful environment to settle in a less stressful locale (or vice versa) are all potential stressors that may be directly associated with residential mobility.

Pregnancy Clinical and bioarcheological research have demonstrated that pregnancy can be inferred with considerable American Journal of Physical Anthropology

4

E.C. WEBB ET AL.

certainty using the isotopic composition of hair keratin. During pregnancy, dietary intake, particularly protein intake, should increase in order to meet the nutritional requirements associated with maintaining maternal health and foetal growth and development (Mojtahedi et al., 2002). Under nutritionally adequate conditions, nitrogen balance is positive, and increasing nutritional demands are met by normal routing of dietary amino acids for tissue synthesis. The d13C values of tissues synthesized during this time are expected to reflect diet, and the d15N values are expected to decrease (by 1– 1.5&; Fuller et al., 2005). If there are periods of nutritional stress during pregnancy (e.g., morning sickness) or if diet is at times nutritionally inadequate, this will induce negative nitrogen balance, and body stores will be mobilized to support tissue synthesis. Under these conditions, the d15N values are expected to increase (by 0.5–1&; Fuller et al., 2005; Williams et al., 2011). During pregnancy, systemic cortisol levels increase to approximately two to three times above individual baseline values in the third trimester, and return to the baseline within days after birth (Kirschbaum et al., 2009).

Malnutrition and starvation Isotopic studies of malnutrition and starvation have used anorexia nervosa as a clinical analogy for understanding changes in metabolism associated with calorierestricted and/or nutritionally inadequate diets. When the diet is inadequate, fat stores are progressively depleted and body proteins are mobilized. As malnutrition progresses to starvation, either because of decreased food intake or decreased intestinal absorption, fat stores are exhausted and body protein catabolism continues (Mekota et al., 2006, 2009). Since catabolized body proteins will make up an increasingly greater proportion of nitrogen-containing compounds in the metabolic amino acid pool, additional 15N-enrichment relative to diet (by 1–2&) will be observed in synthesized tissues (Hobson et al., 1993; Mekota et al., 2006, 2009). Mobilization of fat stores may also cause a decrease in d13C values (LeeThorp et al., 1989; Mekota et al., 2006, 2009). In cases where increased nutritional intake leads to recovery from malnutrition and the return of a healthy metabolism, it is expected that d15N values will initially decrease (transition from 15N-enriched stored protein to dietary protein), and then equilibrate to the nitrogenisotope composition of diet (Mekota et al., 2006, 2009). Clinically, weight loss via calorie restriction leads to minimal increase in plasma cortisol levels (i.e., consuming a low calorie, but nutritionally adequate, diet). Severe malnutrition and starvation may; however, induce marked increase in plasma cortisol, likely caused by both increased secretion of cortisol and hypometabolism (Fichter et al., 1986; Støving et al., 1999; Johnstone et al., 2004; Manary et al., 2006). Malnutrition and starvation are often accompanied by infection or illness, which cause increased cortisol production that may not be directly associated with restricted diet.

Trauma and illness During tissue repair after fracture or trauma, nutritional state influences nitrogen balance. The body will be in positive nitrogen balance if increased nutritional demands are met by diet or in negative nitrogen balance if diet is inadequate or pain, infection, or fever co-occurs. American Journal of Physical Anthropology

The d13C values of hair keratin synthesized during the period of interest do not change uniformly in response to the injury and will reflect diet (Williams et al., 2011). Changes in d15N values of 61–1.5& associated with trauma, pain and reinjury have also been observed (see also Katzenberg and Lovell, 1999; Williams et al., 2011; Olsen et al., 2013). Systemic infection, fever, diarrhea, and/or viral illness all induce negative nitrogen balance (i.e., a catabolic response), and could thus potentially cause an increase in d15N values (Beisel, 1972; Scrimshaw and San Giovanni, 1997). This catabolic response begins at the onset of infection or illness, and persists into convalescence (Beisel, 1972). If malnutrition becomes a complication, the d13C values may also decrease (section “Malnutrition and Starvation”). Combined physiological and psychological stress associated with trauma and/or infection typically lead to higher average cortisol levels (Sapse, 1997; Beishuizen et al., 2001; Ball, 2008). In clinical contexts, patients undergoing treatment for multiple traumas or systemic infection had high cortisol levels that gradually decreased during recovery (Beishuizen et al., 2001). Although longterm high levels of cortisol in the body suppress immune function, cortisol does play an important role in immune response to injury and illness over the short-term.

Limitations Although both clinical and bioarcheological research have examined cortisol production and variability in keratin isotopic composition under various physiological, metabolic, and psychosocial conditions, there has not been a clinical study where both isotopic and cortisol data were collected from the hair of the same individuals. Moreover, modern clinical data cannot be applied uncritically to individuals from archeological contexts. Clinical data addressing, for example, metabolic changes associated with disease processes, will involve individuals undergoing treatment and consuming a balanced diet who, for the most part, were likely to have been fairly healthy before becoming ill. By comparison, although ancient people in the Nasca Region most likely responded to and cared for illness or injury, treatment was ultimately unsuccessful for the individuals examined in this study. Finally, the overall state of health prior to illness or trauma (i.e., stress history extending further back than the length of the hair sample), is often not accessible for individuals from archeological contexts, although recent stress experiences and evidence of childhood and long-term stress can sometimes be observed. The proposed expectations for cortisol and isotopic data and their relationship to various stressors are summarized in Table 1.

ARCHEOLOGICAL CONTEXT Nasca society The Nasca polity inhabited the Rio Grande de Nasca Drainage, Peru, during the Early Intermediate Period (AD1–750; Fig. 1). During the Early Nasca period (AD1– 450), settlements emerged in the upper valleys of the drainage, and construction and use of the ceremonial center, Cahuachi, reached its apogee (Silverman, 1993; Vaughn, 2009). Cahuachi is thought to have functioned primarily as a pilgrimage centre with a small elite population described as a “centralized theocratic authority” (Orefici, 2006, p. 184). Feasting and rituals focused on

INTEGRATING CORTISOL AND ISOTOPIC DATA FROM HAIR

5

TABLE 1. Summary of expectations for cortisol and isotopic data Isotopic data

Cortisol levels

13

Seasonal diet

62.7& change in d C Some change in d15N values

Low stress: minimal variability Psychosocial stress: increased cortisol

Travel

62.7& change in d13C. Possible increase in d15N values with infection

Physical stress: increased cortisol. Exposure to new pathogens: increased cortisol

Malnutrition and starvation

1–2& increase in d15N values. Probable decrease in d13C values. Recovery indicated by decrease in d15N values followed by equilibration with diet

If severe, cortisol production will increase

Trauma and illness

Good nutrition: some decrease in d15N values. Poor nutrition: some increase in d15N values (61–1.5&). Recovery indicated by equilibration of d15N values with diet. Minimal change in d13C values.

High cortisol levels that decrease during recovery

cyclical agricultural events are suggested to have drawn participants from the surrounding valleys (Silverman, 1993). The transitional Middle Nasca period (AD450– 550) is characterized by a dramatic reworking of iconography and the cessation of major construction at Cahuachi (Silverman, 1993; Schreiber and Lancho Rojas, 2003). During this period, the typically unstable Andean climate became even more unpredictable as the result of severe droughts (Thompson et al., 1985). From c.AD250 until the end of the Middle Horizon (c.AD1000), climate in the Nasca Region became increasingly arid (Eitel et al., 2005; Eitel and Machtle, 2009; Bird et al., 2011). Because of the low river volume and inconsistent availability of surface water, the Nasca people developed a system of wells and irrigation canals (puquios) to access more reliable subterranean water and thereby increase agricultural production in the dry middle valleys (Schreiber and Lancho Rojas, 2003). Kantner and Vaughn (2012) hypothesize that the theocratic elite at Cahuachi may have lost social and ideological power because of their inability to mitigate the effects of the drought. The populace coalesced into a few large settlements during the Late Nasca period (AD550–750; Schreiber and Lancho Rojas, 2003; Schreiber, 2005). This population aggregation was concomitant with increased social complexity and greater representation of conflict-related themes on pottery. Towards the end of the Early Intermediate Period, evidence of damaging El Ni~ no effects has been observed, including flooding at Cahuachi (Orefici and Drusini, 2003; Beresford-Jones et al., 2009). The Wari polity expanded into Nasca c. AD750, marking the beginning of the Middle Horizon (locally the Loro Period, AD750–1000). Wari settlements were established throughout the Southern Nasca Region, and Wari ceramics are found in Nasca cemeteries and at habitation sites along with the local Loro style (Strong, 1957; Schreiber, 2001, 2005). The local population of the northern river valleys decreased as Nasca people moved south to the Las Trancas river valley where a large local center, Huaca del Loro, was established (Schreiber, 2001; Conlee and Schreiber, 2006). A recent archeological survey (Edwards, 2010) further indicates that the Wari made important changes to the landscape of the upper and mid-valleys of the SNR, establishing several settlements and constructing roads that linked the head-

Fig. 1. Map of Rio Grande de Nasca Drainage indicating location of sampled burials.

waters of the Nasca river tributaries with the coastal plains.

Bioarcheological perspective on Nasca stress and health status Elevated rates of accidental trauma and trauma associated with violence have been observed in skeletal samples from the Las Trancas and Palpa river valleys (Kellner, 2002; Cagigao, 2009). The incidence of trauma indicative of interpersonal conflict (e.g., cranial and cranio-facial trauma, parry fractures) is relatively high during all periods compared to samples from other regions globally (Kellner, 2002). Interpersonal violence is often motivated by changing social and physical environmental conditions, such as uncertain resource availability or conflict with expanding polities (Tung, 2007). Tung (2007) observed that Wari imperial rule was associated with considerable presence of trauma consistent with interpersonal violence in peripheral Wari settlements American Journal of Physical Anthropology

6

E.C. WEBB ET AL. TABLE 2. Stable isotope and cortisol data

Sample ID

Sex

CAH493a

Male

Segment # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

CAH496

Male 1 2 3 4

CAH523

Male 1 2 3 4 5 6 7 8

CAH536

CAH539a

CAH540a

Female 1 2 3 4 5 6 7 8 9 10 11 12 13 Male 1 2 3 4 5 Male 1 2 3 4 5 6 7 8 9 10 11 12 13 14

CAH553 1 2 3

C/N ratios

d13Cker (&, VPDB)

d15Nker (&, AIR)

4.1 6 0.1 4.3 4.1 4.0 4.0 4.1 4.1 4.0 4.1 4.0 4.1 4.0 4.1 4.1 4.1 4.1 4.0 6 0.1 4.0 3.9 4.0 4.0 3.9 6 0.1 4.0 3.9 3.8 3.9 4.0 3.9

215.6 6 1.3 216.3 214.7 214.5 214.2 213.8 213.6 215.3 217.3 217.0 217.5 216.7 215.1 214.9 216.1 217.1 216.6 6 0.1 216.5 216.6 216.7 216.5 214.6 6 1.2 213.1 213.4 214.4 215.1 215.6 215.9

8.6 6 0.9 8.8 9.9 10.5 9.9 9.3 8.9 8.4 7.8 8.0 7.6 7.6 8.3 8.3 8.1 8.3 11.7 6 0.3 11.5 11.5 12.1 11.5 7.6 6 0.4 7.7 7.8 7.0 7.3 7.8 8.2

4.0 6 0.1 4.1 3.9 4.0 4.0 3.9 3.9 4.0 3.8 4.0

214.6 6 0.2 214.8 214.6 214.4 214.3 214.4 214.4 214.8 214.8 214.9

9.2 6 0.5 9.8 9.7 9.6 9.5 9.4 9.3 8.7 8.6 8.3

3.8 6 0.1 3.8 3.7 3.8 3.8 3.8 4.0 6 0.2 3.9 3.7 3.8 3.9 3.8 3.9 3.9 3.9 4.2 4.1 4.1 4.1

212.7 6 0.7 212.1 213.3 213.5 211.8 212.7 213.9 6 2.5 217.7 217.4 216.4 216.1 215.0 211.5 211.1 211.0 213.2 212.4 212.4 212.7

7.9 6 0.8 8.2 6.9 7.3 8.5 8.5 8.3 6 0.5 9.0 8.5 9.0 8.6 7.4 8.5 8.2 7.9 7.7 8.0 8.4 8.8

3.8 6 0.1 3.9 3.9 3.8

215.3 6 2.0 218.0 218.1 216.8

6.6 6 0.7 6.5 6.1 6.6

American Journal of Physical Anthropology

d13Ccol (&, VPDB)

d15Ncol (&, AIR)

Cortisol (ng/g) 471 6 153 128 760 624 587 492 444 459 448 288 366 454 417 595 530

215.5

9.0

215.0

8.7

213.6

9.3

1195 6 178 1022 1185 1377 2392 150 6 21 125 139 153 185 158 138 251 320 249 6 64 299 275 176 151 174 194 237 248 307 316 351 300 203 289 6 202 165 578 425 149 127 143 6 45 260 106 105 87 111 120 121 120 143 147 147 156 185 188 218 6 155 707 425 232

7

INTEGRATING CORTISOL AND ISOTOPIC DATA FROM HAIR TABLE 2. Continued

Sample ID

Sex

CAH558a

Segment #

C/N ratios

d13Cker (&, VPDB)

d15Nker (&, AIR)

4 5 6 7 8 9 10 11 12 13 14 15

3.8 3.8 3.7 3.8 3.8 3.8 3.7 3.7 3.7 3.8 3.8 3.9 4.0 6 0.0 3.9 3.9 4.0 4.0 3.9 3.8 6 0.1 3.7 3.7 3.7 3.7 3.7 3.7 3.7 3.8 3.8 3.7 3.7 3.8 3.8 3.8 3.8 3.8 3.8 3.8 3.9 3.8 3.8 3.8 3.9 3.9 3.7 6 0.1 3.9 3.7 3.6 3.6 3.6 3.8 3.6 3.6 3.7 3.7 3.7 3.7 3.7 3.7 3.8 3.8 3.7 3.7 3.7 6 0.0 3.7 3.7 3.7 3.6 3.6 3.7

213.3 211.3 212.1 214.1 215.8 216.0 215.7 215.9 215.9 215.5 215.0 216.8 216.2 6 2.9 218.1 218.3 217.7 215.6 211.4 213.8 6 0.9 214.7 215.3 214.6 212.6 212.0 213.2 214.8 214.7 214.6 214.0 213.7 213.3 213.4 213.7 212.9 213.0 212.8 212.1 213.4 214.4 214.6 214.4 214.5 214.5 213.9 6 1.4 215.9 215.6 215.7 215.6 214.8 214.7 215.0 214.9 214.0 213.5 212.9 212.6 212.6 212.2 212.4 212.5 212.6 212.6 213.5 6 0.5 213.8 214.5 214.0 213.5 212.9 212.6

8.0 8.2 7.0 5.9 5.7 6.3 6.5 6.4 6.4 6.5 6.3 5.9 6.9 6 1.2 6.2 5.9 6.3 7.6 8.7 9.4 6 1.4 8.4 8.2 8.5 10.6 11.1 9.2 7.9 8.3 8.5 9.2 9.8 10.2 10.0 10.4 11.5 11.7 11.2 11.8 9.2 7.8 7.8 8.2 8.1 8.0 7.6 6 0.3 7.3 8.0 8.4 7.9 7.4 7.0 7.1 7.1 7.5 7.5 7.8 7.6 7.7 7.7 7.6 7.8 7.7 7.6 9.2 6 0.2 9.2 9.2 9.1 9.3 9.1 9.2

1 2 3 4 5 LTR590

LTR593a

Female 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Male 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

LTR616

Male 1 2 3 4 5 6

d13Ccol (&, VPDB)

d15Ncol (&, AIR)

Cortisol (ng/g) 211 163 112 154 153 206 147 171 123 186 133b 133b 955 6 137 1038 850 1118 988 782 805 6 316 965 1261 1228 1073 742 552 462 532 515 524 569 743 1375 728

636 6 127 811 685 493 575 536 592 575 525 624 532 613b 613b 613b 897b 897b 897b 807b 807b 596 6 322 1653 775 626 670 483 457

American Journal of Physical Anthropology

TABLE 2. Continued

Sample ID

LTR628

Sex

Segment #

C/N ratios

d13Cker (&, VPDB)

d15Nker (&, AIR)

7 8 9 10 11 12 13 14

3.7 3.7 3.7 3.6 3.7 3.7 3.7 3.8 3.7 6 0.2 4.0 3.7 3.7 3.8 3.9 3.6 3.7 3.6 3.5 3.5 3.5 3.6 3.7 6 0.0 3.7 3.7 3.7 3.6 3.6 3.7 3.8 3.7 3.7 3.7 3.7 3.7 3.7 3.7 4.1 6 0.2 4.1 4.9 4.0 4.0 4.1 4.0 4.1 4.2 4.1 4.2 4.2 4.1 4.0 4.0 4.0 4.0 4.0 4.0 3.9 3.8 4.0 3.9 3.9 3.9 4.0 4.0 4.1 4.1 4.1 4.1 4.1 4.0 4.0 4.0

213.1 213.3 213.3 213.3 213.3 213.6 213.3 213.9 217.3 6 0.2 217.5 217.6 217.3 217.2 217.3 217.4 217.6 217.3 217.1 216.9 216.9 217.2 213.8 6 0.7 213.9 213.8 213.9 213.5 214.3 214.4 214.1 213.5 213.0 212.7 213.2 213.2 215.2 214.7 214.1 6 1.0 213.7 213.3 213.2 213.0 213.2 213.0 213.2 212.8 213.5 213.9 213.9 213.7 213.1 213.1 213.0 213.0 213.0 213.4 214.5 214.0 214.4 214.3 215.0 215.5 215.3 214.9 216.0 215.6 214.6 215.5 214.8 215.0 214.9 215.1

9.4 9.3 9.2 8.9 8.8 8.8 9.4 9.6 8.2 6 0.3 9.1 8.4 8.3 8.2 7.8 8.0 7.8 8.1 8.2 8.3 8.1 8.0 9.9 6 0.4 10.7 10.4 9.8 9.6 9.6 9.4 9.8 9.9 10.2 10.2 10.1 10.1 9.3 9.5 10.7 6 1.5 11.4 10.4 10.5 10.6 10.8 11.1 11.0 11.1 11.8 12.0 12.0 12.0 12.5 13.0 13.4 13.5 13.0 12.3 11.0 11.3 11.6 11.1 9.8 9.0 9.0 9.1 8.6 8.8 9.5 8.8 9.0 8.5 8.8 8.5

Male 1 2 3 4 5 6 7 8 9 10 11 12

LTR631

Male 1 2 3 4 5 6 7 8 9 10 11 12 13 14

LTR649

Female 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

a b

Crania modified as Nasca trophy heads. Sampled in 3 cm segments.

d13Ccol (&, VPDB)

d15Ncol (&, AIR)

Cortisol (ng/g) 502 387 496 489 457 431 469 450 514 6 95 624 512 565 705 556 478 393 396 522 437 467 405 6 76 439 337 296 298 353 447 440 490 479 469

343 6 67 323 333 272 313 299 269 250 282 258 452 484 412 350 443 339 354 324 417 373 375

INTEGRATING CORTISOL AND ISOTOPIC DATA FROM HAIR outside the Nasca drainage, and Wari-affiliated individuals in the Las Trancas skeletal sample had a similarly high incidence of cranial trauma (Kellner, 2002). Finally, Cagigao (2009) noted that, among Early Intermediate Period Nasca individuals from sites in the Palpa river valley, the frequencies of accidental trauma (e.g., limb fractures) and trauma associated with violence were 50% and 38%, respectively. Combined data from these studies indicate that both physical and psychosocial stress caused by interpersonal violence and accidental trauma were likely significant during the Early Intermediate Period and Middle Horizon. Bioarcheological analysis of a skeletal sample excavated from the Las Trancas river valley cemeteries revealed an overall decline in health status from Early Nasca to the Loro period, but, compared to other archeological skeletal samples, the Nasca were not under excessive stress during any period (Kellner, 2002). Nasca individuals experienced physiological stress during childhood (indicated by the presence of linear enamel hypoplasias), localized and systemic infections (osteoperiostitis) and anaemia (potentially caused by parasitic infection, infectious disease and/or chronic diarrhea) (Kellner, 2002). Increased population density often leads to a decline in health because of increased pathogen transmission and waste disposal problems. Stress may be alleviated by improving socioeconomic conditions, but social development may aggravate and/or generate stress as well (Goodman et al., 1988). Exposure to various stressors does not mean that there will be a harmful impact on an individual’s health. Impact is a function of duration and severity as well as an individual’s ability to cope with a given stressor (Sapolsky, 2004). Typically, prolonged stress episodes, and many infectious agents and soft tissue diseases/injuries, will not leave clear skeletal evidence. Moreover, individuals with lesions may be “paradoxically” healthier that those without, since, for skeletal traces to become evident, the individual must survive the stress episode and/or live for a significant amount of time with a disease (Wood et al., 1992).

METHODS Sampling Hair samples were collected at the State Collection for Anthropology and Palaeoanatomy, Munich, Germany in February and September 2008. The remains are from Heinrich Ubbelohde-Doering’s 1932 field season, during which he excavated 50 graves in the Rio Grande de Nasca drainage (Ubbelohde-Doering, 1958, 1966). Hair sampled from fourteen adults from Cahuachi (CAH) and from cemeteries south of Huaca del Loro in the Las Trancas river valley (LTR) was analyzed for cortisol levels and carbon- and nitrogen-isotope compositions. This group of samples includes five crania that have been modified as Nasca trophy heads (Tello, 1918; Proulx, 1989; Browne et al., 1993; Silverman, 1993; Verano, 1995; Proulx, 2001; Verano, 2003; Knudson et al., 2009). Hair samples were collected with care taken to maintain orientation and alignment of individual hair strands within each sample bundle. Each bundle of hair was sectioned into 1 cm segments, with each segment representing approximately one month of growth based on the average growth rate of 0.35 mm/day for scalp hair. Hair lengths in this study ranged from three to 34 cm, resulting in individual tem-

9

poral records ranging from approximately three months to nearly three years.

Laboratory procedures For carbon- and nitrogen-isotope analysis, each hair sample bundle was cleaned of loose particulate matter by wiping gently with ethanol. Hair bundles were then sectioned and each segment sample was placed in a glass vial. Lipids and decomposition fluid residues were removed by soaking hair segments in a 2:1 chloroform: methanol solution for 24 h. Each segment sample was then rinsed with fresh solution and, when necessary, this process was repeated. Segment samples were air-dried for 12 to 24 h, and then finely minced and weighed (510 6 10 lg) into tin capsules for isotopic analysis. Isotopic analysis of keratin was performed using a Costech Elemental Analyzer interfaced with a Thermo Finnigan DeltaPlus XL mass spectrometer. The d13C values were calibrated to VPDB (Coplen, 1994, 1996) using IAEA-CH-6 and IAEA NBS-22, and the d15N values were calibrated to AIR (Mariotti, 1983; Hoefs, 2004) using IAEA-N1 and IAEA-N2. Analytical precision was determined through duplicate analyses of samples, and was 60.2& for d13Cker, and 60.1& for d15Nker. Accuracy was assessed using a laboratory keratin standard, which gave an average d13C value of –24.06 6 0.05&, and an average d15N value of 16.26 6 0.09&, which compare well with its accepted values (–24.04& and 16.36&, respectively). Isotopic analyses were performed at The University of Western Ontario, London, Canada, in the Laboratory for Stable Isotope Science. The methodology used to measure cortisol abundance in hair is described in detail in Van Uum et al. (2008). At least 10 mg of hair per segment is required for cortisol analysis. Hair segments were placed in glass vials and the hair was very finely minced with surgical scissors. One millilitre of methanol (>98%) was added to extract the steroid. The vials were then sealed and incubated for 16 h, while shaking at 100 rpm and heating at 50 C. The methanol was then removed from the vial, transferred to test tubes and evaporated by heating to 40 C under a nitrogen stream. The residue was reconstituted in 250 ml of phosphate-buffered saline (PBS) at pH 8.0. The PBS mixture was then analyzed using a commercially available salivary enzyme immunoassay kit (ELISA, ALPCO Diagnostics). All measurements were done in duplicate, and the average difference between duplicate analyses was 643 ng/g. For this procedure, the typical intraday coefficient of variation is 7.2% in a hair sample with 64 ng/g of cortisol and 6.0% at 629 ng/g. The interday variation is 10.6% and 7.6% for 49 and 548 ng/ml, respectively. Cross-reactivities of steroids in the salivary ELISA kit are as follows: cortisol 100%, corticosterone 31%, progesterone