Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iahb20

Download by: [Washington University in St Louis] Date: 15 May 2017, At: 08:18

Annals of Human Biology

ISSN: 0301-4460 (Print) 1464-5033 (Online) Journal homepage: http://www.tandfonline.com/loi/iahb20

Urban and rural mortality and survival in MedievalEngland

Brittany S. Walter & Sharon N. DeWitte

To cite this article: Brittany S. Walter & Sharon N. DeWitte (2017) Urban and ruralmortality and survival in Medieval England, Annals of Human Biology, 44:4, 338-348, DOI:10.1080/03014460.2016.1275792

To link to this article: http://dx.doi.org/10.1080/03014460.2016.1275792

Accepted author version posted online: 22Dec 2016.Published online: 17 Jan 2017.

Submit your article to this journal

Article views: 335

View related articles

View Crossmark data

RESEARCH PAPER

Urban and rural mortality and survival in Medieval England

Brittany S. Walter and Sharon N. DeWitte

Department of Anthropology, University of South Carolina, Columbia, SC, USA

ABSTRACTBackground: Late medieval England underwent intensive urbanisation, particularly in its largest city:London. Urban dwellers were exposed to factors such as high population density, elevated risk ofinfection, unsanitary living conditions and precarious food supplies.Aim: To assess whether the urban environment was more detrimental to health than the rural environ-ment, this study compares risks of mortality and survival, as proxies for health, in medieval urban vsrural England.Subjects and methods: This study uses samples from rural St. Peter’s cemetery in Barton-upon-Humber, Lincolnshire (c. 1150–1500) and urban St. Mary Spital cemetery in London(c. 1120–1539). Cox proportional hazards analysis and Kaplan-Meier survival analysis are used to assessdifferences in mortality and survival between urban and rural environments, including differencesbetween sexes.Results: The results indicate that urban adults faced elevated risks of dying and reductions in survivor-ship. Specifically, urban females faced elevated risks of dying and reductions in survivorship, while therisks for males were similar in both environments.Discussion: These results suggest that the effects of urbanisation in medieval England varied by sex.Deleterious conditions associated with urbanisation in London were hazardous for adults, particularlyfemales who may have migrated into London from rural areas for labour opportunities.

ARTICLE HISTORYReceived 18 August 2016Revised 14 October 2016Accepted 14 November 2016

KEYWORDSPaleodemography; hazardanalysis; bioarchaeology;urbanization; medievalEngland

Introduction

Transitional periods, associated with such phenomena asagricultural intensification and urbanisation, are of particularinterest to bioarchaeologists, as several have concluded thatincreasing social and economic complexity are associatedwith dramatic declines in health (Armelagos & Cohen, 1984;Roberts & Cox, 2007; Steckel & Rose, 2002). The UnitedNations (UN, 1996, 2001a, b) and World Health Organisation(WHO, 1998) have made improving living conditions in citiesexperiencing rapid urbanism a priority, and both call forresearch exploring the health risks associated with the envir-onmental hazards of urbanism. Most of the urban health risksexperienced by populations in the past also affect modernpopulations, including increased levels of infection from over-crowding and inadequate sanitation services and food sup-plies (Moore et al., 2003). Bioarchaeological research onurbanisation provides a unique temporal depth to our under-standing of the hazards of urbanism that clarifies humanadaptation to these changes and can contribute to the devel-opment of effective planning and prevention strategies forurban expansion.

England underwent intensive urbanisation from the 11thto 15th centuries, particularly in its largest city: London(Magnusson, 2013). Most evidence of the medieval Londonenvironment comes from primary sources such as courtrecords, complaints, ordinances or decrees to abate pollution.

Environmental archaeology has also contributed to ourunderstanding of urbanisation in late medieval England, spe-cifically the effects of pollution and urban growth on thelandscape (see Hall & Kenward, 1994; Schofield, 2011).Bioarchaeological research, however, contributes to ourunderstanding of how the environment may have affectedpeople. By examining the skeletal remains of individuals whoexperienced urbanisation, bioarchaeologists can elucidatepatterns of health and mortality associated with urbanism.Analysis of skeletal assemblages of people exposed to urbanenvironmental factors, such as high population densities,potentially elevated risk of infection and unsanitary livingconditions (as described in the Discussion below), can pro-vide unique insights into the effects of urbanisation onhealth and mortality in the past; particularly when comparedto assemblages of people unexposed or less severelyexposed to these factors.

Previous bioarchaeological studies investigating health inthe context of urbanisation primarily assessed raw frequen-cies of pathological lesions in urban and rural skeletal sam-ples, interpreting higher levels of pathologies in urbansamples as evidence for the negative health effects of urban-isation (Cohen, 1989; De La R�ua et al., 1995; Lewis et al.,1995; Storey, 1992). However, skeletal lesions should notnecessarily be interpreted as direct indicators of healthbecause of heterogeneity in frailty and selective mortality

CONTACT Brittany S. Walter walterbs@email.sc.edu Department of Anthropology, 817 Henderson Street, Gambrell Hall 440A, University of South Carolina,SC 29208, USA� 2017 Informa UK Limited, trading as Taylor & Francis Group

ANNALS OF HUMAN BIOLOGY, 2017VOL. 44, NO. 4, 338–348http://dx.doi.org/10.1080/03014460.2016.1275792

(Wood et al., 1992). Although previous studies of urbanisa-tion are potentially informative about broad health changes,they reveal little about underlying heterogeneity withinurban or rural populations. In order to more adequatelyaddress the potentially complex relationship between urban-isation and health, we need to use paleodemographic analy-ses that allow for intra-population variation in patterns ofmorbidity and mortality and are, thus, capable of revealingwhether urbanisation disproportionally affects certain sub-populations (e.g. males vs females or particular age groups).

Most paleodemographic studies of urbanisation haveassessed mortality using traditional approaches, such as com-parisons of mean age-at-death or life tables (Nagaoka &Hirata, 2007; Steckel, 2005; Storey, 1985). However, the recon-struction of the health consequences of urbanisation in thepast is inherently complex and might be limited by smallsample sizes and traditional methods, given the phenomenaof demographic non-stationarity, hidden heterogeneity andselective mortality (Konigsberg & Frankenberg, 2002; Vaupel& Yashin, 1985; Wood et al., 1992). The quantitative modelsapplied in this study, however, allow us to account for select-ive mortality and heterogeneity in frailty. Moreover, the stat-istical approach used in this study, hazards analysis,accommodates missing data without imposing a particularage pattern on skeletal data (Gage, 1988). Further, this studyuses transition analysis, an age estimation technique for skel-etal remains that avoids some of the limitations associatedwith traditional age estimation methods (Boldsen et al.,2002). This age estimation method and the hazards analysisare increasingly being used in paleodemography (e.g. Bullocket al., 2013; DeWitte & Wood, 2008; Wilson, 2014). Theseapproaches, however, have not yet been applied to an inves-tigation of urban–rural mortality and survival differences. Thisstudy uses hazards and survival analysis to assess demo-graphic differences between the medieval urban St. MarySpital cemetery (SRP98) in London (c. 1120–1539) and thecontemporaneous rural St. Peter’s cemetery (BOH) in Barton-upon-Humber, Lincolnshire, England (c. 1150–1500).

Methods

Skeletal assemblages

St. Mary Spital (1120–1539)St. Mary Spital was in use during the early urbanisation ofmedieval London. Medieval London was exceptional inBritain because of the degree to which the area urbanisedand how quickly it did so (Thomas, 2002). Urban centres likeLondon attracted poor people, who were presumably morevulnerable to infections and exhibited high death rates com-pared to their rural counterparts (Singman, 1999). St. MarySpital was established for treating the general population ofLondon, and occasionally catered to the religious community(i.e. monks and lay sisters) and some wealthy benefactors.Connell et al. (2012) suggest that, because there are massburials within the cemetery and because the burial popula-tion is larger than the number of individuals that could havebeen treated at the hospital, the burial population was not

limited to hospital inmates and would have also includedindividuals from the general community of London.

SRP98 is one of the largest urban skeletal assemblagesexcavated in Europe to date, with 5387 skeletons availablefor analysis, and is curated by the Museum of London(Connell et al., 2012). Using high-precision Bayesian radiocar-bon dating within a well-defined stratigraphic framework, thecemetery has been classified into four distinct chronologicalphases: 1120–1200 (P14), 1200–1250 (P15), 1250–1400 (P16),and 1400–1539 (P17) (see Sidell et al., 2007 for details regard-ing phasing of the cemetery using Bayesian modelling). Weselected a stratified random sample of 335 skeletons fromamong skeletons that were at least 50% complete and hadthe relevant skeletal elements needed for age estimation andsex determination across each of the four temporal phases;the sample sizes are presented in Table 1.

Barton-upon-Humber (1150–1500)St. Peter’s cemetery at Barton-upon-Humber (BOH),Lincolnshire, England was in use from 1086–1855; excavationof the cemetery yielded 2750 skeletons (Waldron, 2007). BOHincludes individuals from all levels of society living in the vil-lage and its hinterland (Waldron, 2007). Barton was consid-ered a poor rural community during the medieval period(Clapson, 2005). Although Barton is often referred to as asmall town in the literature, it actually never had a charter inthe medieval period or any form of corporate government,attributes that typically differentiate a town from a village(Bryant, 2003). The agricultural village did not experience sig-nificant urbanisation until the late 17th century when facto-ries sprang up along the River Hull (Clapson, 2005), after theperiod of interest for this study; thus BOH is appropriatelyconsidered ‘rural’ for this study.

The cemetery has been divided into five temporal phases(E: 950–1150, D: 1150–1300, C: 1300–1500, B: 500–1700, A:1700–1855) using radiocarbon dating of coffins and skeletalremains, dendrochronology of coffin boards, and by estab-lishing relative chronologies through localised grave sequen-ces (see Bayliss & Atkins, 2011 for details regardingcemetery phasing). This phasing permitted the selection ofa sample that includes only burials that pre-date urbanisa-tion at Barton. For this study, we sampled 149 individualsequally from two phases in the assemblage (D: 1150–1300and C: 1300–1500); the sample sizes are presented inTable 1. This includes all skeletons from those two phasesthat were at least 50% complete. Barton-upon-Humber pro-vides an appropriate rural comparative skeletal assemblagebecause it is large, well preserved, and provides theapproximate temporal phases necessary for comparisonwith St. Mary Spital.

Table 1. Sample sizes for St. Mary Spital (SRP98) and Barton-upon-Humber (BOH) cemeteries

SRP98 BOH SRP98 þ BOH

Adults 335 149 485Males 166 74 240Females 169 75 244

ANNALS OF HUMAN BIOLOGY 339

Skeletal analysis

Sex estimationSex for each individual was determined by examining sexu-ally dimorphic features of the pelvis and skull (Buikstra &Ubelaker, 1994). Skeletal features that were scored include:supraorbital margin, supraortbital ridge, mastoid process,external occipital protuberance, mental eminence, pubic ven-tral arc, subpubic concavity, ischiopubic ramus and greatersciatic notch. When available, features of the pelvis wereweighed more heavily than those of the skull, as the formeryield more accurate sex estimates (Meindl et al., 1985;Walrath et al., 2004). Individuals for whom sex determinationwas not possible or questionable were not included in analy-ses evaluating sex differentials.

Age-at-deathAdult age-at-death was determined using transition analysis(Boldsen et al., 2002) via the ADBOU (AnthropologicalDatabase, Odense University) age estimation software.Individuals for whom age could not be estimated (i.e. thoseadults missing or with poorly preserved age-indicators) werenot included in analyses. Only adults were included in thisstudy (i.e. individuals with an estimated age-at-death of15 years and older).

There are several limitations to and biases associated withestimating age-at-death using skeletal remains, particularlyfor adults. Unlike sub-adult age estimates, adult skeletal ageestimates are typically determined by assessing macroscopicdegenerative changes to the skeleton that are also influ-enced by factors such as environment, genetics, and physicalactivity throughout life (Bello et al., 2006). In addition to theinaccuracy of degenerative change rates, conventional meth-ods of age estimation risk imposing the age distribution of aknown-age reference collection onto the target collection (i.e.age mimicry) (Bocquet-Appel & Masset, 1982). Moreover,traditional age estimation methods assign individuals tobroad, pre-determined age intervals. An individual with anage-at-death on the low or high end of a pre-determinedage interval could potentially be categorised into an incorrectage interval, resulting in inaccurate age-at-death distribu-tions. Traditional methods also tend to yield only broad ter-minal age intervals for the oldest individuals in a sample (e.g.50 and older), limiting assessment of demographic trends atthe oldest adult ages. However, transition analysis, the ageestimation method used for this project, attempts to avoidthese biases by using some of the criteria articulated in theRostock Manifesto (see Hoppa & Vaupel 2002).

Transition analysis uses data from a known-age referencesample to obtain a conditional probability Pr(cjja) of a skel-eton exhibiting a certain age indicator stage or a suite of ageindicator stages (cj), given the individual’s known age (a).This conditional probability is then combined with a priordistribution of ages at death using Bayes’ Theorem to esti-mate the posterior probability that a skeleton of unknownage died at a certain age, given that the skeleton displaysa particular suite of age-indicator stages. The ADBOUprogramme uses an informative prior distribution of

age-at-death (the Gompertz-Makeham model, see Woodet al., 2002) estimated from 17th-century Danish rural parishrecords, and the conditional probability of the age indicatorsgiven known age at death estimated from the SmithsonianInstitution’s Terry Collection. The parameter estimates for theprior distribution estimated from the 17th-century Danishrecords are: a1¼0.01273, a2¼0.00002478, and b¼ 0.1060(J Boldsen, personal communication, 3 September 2008). Theskeletal age indicators used in this method include severalfeatures of the pubic symphysis and iliac auricular surface,and cranial sutures (Boldsen et al., 2002).

Statistical analysis

Cox proportional hazards modelThe effect of rural vs urban environments on risk of death isevaluated using the Cox proportional hazards model (Cox,1972) with pooled point estimates of age for adults fromboth cemeteries and modelling ‘urban’ as a covariate(0¼ rural, 1¼urban). The Cox model is a semi-parametricregression model that estimates relative risk of death, anddoes not require the specification of the baseline hazardfunction. The model tests the null hypothesis that the covari-ate has no effect on the hazard, with the reported hazardratio indicating the change in risk of death associated with aunit increase in the covariate. Significant ratios (p< 0.05) thatare greater than 1.0 indicate that the urban covariate is asso-ciated with elevated risk of death. Sex differentials in mortal-ity for the urban and rural environments are also evaluatedusing this approach. To assess whether mortality differencesbetween urban and rural environments are consistentbetween the sexes, we model ‘urban’ as a covariate affectingthe Cox model separately for males and females.

Kaplan-Meier survival analysisThe effect of rural vs urban on survival is evaluated usingKaplan-Meier survival analysis, with a log rank test usingpooled point estimates of age from both cemeteries andmodelling ‘urban’ as a covariate. To assess sex differences inadult survivorship in urban and rural environments, malesand females are also analysed separately.

All analyses were performed using SPSS Version 21. It isimportant to note that our analyses do not take into accountthe errors associated with estimated ages-at-death; thus, thestandard errors produced by the Cox proportional hazardanalyses and the Kaplan-Meier survival analyses may beunder-estimated. Therefore, the results from our analysesshould be viewed as informative in so far as they indicategeneral trends, although the numerical estimates themselvesshould be viewed with caution.

Fertility proxyChanges in fertility (one possible manifestation of demo-graphic non-stationarity) have been found to alter age-at-death distributions, irrespective of trends in age-specificmortality (Milner et al., 1989; Paine, 1989; Sattenspiel &Harpending, 1983). In a population experiencing an increase

340 B. S. WALTER AND S. N. DEWITTE

in fertility, for example, increasing numbers of children willbe born each year, thus increasing the number of childrenwho die each year, even if age-specific mortality rates do notchange. Ultimately, under these circumstances, the resultingcemetery assemblage from this growing population will con-tain an excess of young individuals relative to older individu-als. This phenomenon makes it difficult to infer mortalitypatterns directly from age-at-death data from cemetery sam-ples (Milner et al., 2008) and complicates the comparison oftwo different, contemporaneous assemblages that are poten-tially derived from populations with different fertility rates.

In this study, we control for fertility by examining thenumber of individuals 30 years of age and above divided bythe number of individuals 5 years of age and above (i.e.D30þ/D5þ). There is a strong negative relationship betweenD30þ/D5þand birth rate (Buikstra et al., 1986), and the corre-sponding 95% comparison intervals can be informative aboutwhether birth rates differ significantly across samples.

Although our use of this fertility proxy allows us to inferwhether differences in estimated survivability and mortalitybetween the rural and urban assemblages are artefacts of dif-ferences in fertility between the populations, it does notresolve the potential influence of migration. As is often thecase with bioarchaeological research, we make the assump-tion that the populations under consideration were stable(closed to migration and with constant age-specific fertilityand mortality rates). This assumption is reasonable, as mostpopulations appear to have stable age structures (Gage et al.,2012). Although previous modelling work has shown thatdemographic perturbations (resulting, for example, from crisismortality associated with famine or plague) can have effectson age-at-death distributions that last for several decades(Paine, 2000; Weiss, 1975), substantial effects are relativelyshort-lived. Given the relatively long period of time acrosswhich our samples from both cemeteries are drawn, suchpotential effects of temporary perturbations in demographicpatterns are likely not strong enough to affect our conclusions.

It is possible, however, that migration during this period mightaffect our results. We describe the possible influence of migra-tion on our findings in the Discussion.

Results

Age-at-death distributions

The age-at-death distributions from SPR98 and BOH ceme-teries for all adult ages are shown in Figure 1. The results ofa Kolmogorov-Smirnov test indicate that the two distribu-tions are significantly different (p< 0.001). As shown inFigure 1, there are more young adults (15–20 years of age) inSRP98 compared to BOH. After 20 years of age, the adult dis-tributions are similar until 40 years of age; however, BOH hasa higher proportion of individuals 40–70 years of age, andSRP98 has a slightly higher proportion of individuals abovethe age of 70.

Cox proportional hazards modelThe results of the Cox proportional hazard analyses areshown in Table 1. The estimated hazard ratio, when sex ispooled, is significant and greater than 1.0, with the corre-sponding confidence interval including only values above1.0. These results suggest elevated risks of mortality foradults in the urban environment compared to those in therural environment.

The sex-specific results are also shown in Table 2. Forfemales, the hazard ratio is significant and greater than 1.0,with the corresponding 95% confidence interval including

Figure 1. Age-at-death distributions from St. Mary Spital (SRP98) and Barton-upon-Humber (BOH) cemeteries.

Table 2. Cox proportional hazards results for adults, femalesand males, with ‘urban’ as the covariate.

Exp(B) 95% CI Sig.

Adults 1.225 1.009–1.487 0.040Males 1.104 0.837–1.457 0.482Females 1.349 1.023–1.779 0.034

ANNALS OF HUMAN BIOLOGY 341

only values above 1.0. For males, however, although the haz-ard ratio is greater than 1.0, it is not significant, and the cor-responding 95% confidence interval includes 1.0 and valuesbelow 1.0. Together, these results suggest that females facedelevated risks of dying in the urban environment, while therisks of dying for males may have been similar in bothenvironments.

Kaplan-Meier survival analysisThe results of the Kaplan-Meier survival analysis are shown inTable 3, and the survival curves are shown in Figure 2. Thesurvival functions reveal a significant difference in mean sur-vival time between BOH and SRP98 (Mantel-Cox p¼ 0.038),and the corresponding 95% confidence intervals for the twocemeteries do not overlap. This suggests that urban adultshad lower survivorship compared to rural adults.

The sex-specific Kaplan-Meier results are shown in Table 3,and the survival curves are shown in Figures 3 and 4. Forfemales, there is a significant difference in mean survivaltime between BOH and SRP98 (Mantel-Cox p¼ 0.044), withslightly overlapping 95% confidence intervals (�1 year ofoverlap). There is no significant difference in mean survivaltime between urban and rural males, and the male 95%

confidence intervals overlap substantially. These results sug-gest that urban females had reduced survivorship comparedto rural females, but males experienced similar survivorshipin rural and urban environments.

Fertility proxyThe fertility proxies and their corresponding 95% confidenceintervals for both cemeteries are provided in Table 4. TheD30þ/D5þ value for BOH is higher than the D30þ/D5þ valuefor SRP98, which suggests that birth rates may have beenlower in the rural environment compared to the urban envir-onment. However, the comparison intervals for the two cem-eteries substantially overlap, indicating a lack of significantdifference in birth rates between the two.

Discussion

The results of these analyses suggest that, for adults in gen-eral, there were elevated risks of mortality and reductions insurvivorship for individuals in the urban environment com-pared to the rural environment. However, separate analysesof males and females reveal that females faced elevated risksof dying and reductions in survivorship in the urban environ-ment, but the risks for males were similar in both environ-ments. These demographic differences between the twocemeteries do not appear to be an artefact of differences inbirth rates. The inconsistent trends in mortality and survivor-ship suggest that the effects of urbanisation in medievalEngland varied by sex.

The observed elevated mortality and, by inference, com-promised health in the urban environment for adults in thisstudy are consistent with previous bioarchaeological investi-gations of urbanisation (Lewis et al., 1995; Morfin et al., 2002;Storey, 1985). For example, comparison of age-at-deathdistributions between rural Wharram Percy and urban

Table 3. Kaplan-Meier survival analysis results for all adults, females andmales.

SampleMean survivaltime (years) 95% CI Mantel-Cox v2 p-values

AdultsSRP98 31.600 29.799–33.401 4.283 0.038BOH 36.065 33.457–38.672

FemalesSRP98 31.253 28.765–33.741 4.070 0.044BOH 35.999 32.444–39.555

MalesSRP98 34.511 31.740–37.281 0.106 0.745BOH 36.949 33.087–40.812

Figure 2. Kaplan-Meier survivorship curves with 95% confidence intervals for adults of St. Mary Spital and Barton-upon-Humber cemeteries.

342 B. S. WALTER AND S. N. DEWITTE

St. Helen-on-the-Walls in medieval York also indicates thaturban dwellers did not live as long as rural villagers (Robertset al., 1998).

It is traditionally argued that urban living is inherentlyrisky because of high population density, sanitation issues,water and air pollution, high prevalence of infectious diseaseand famine (Moore et al., 2003). Differences in adult mortalityand survivorship between the two cemeteries in this study

might reflect detrimental environmental conditions inLondon compared to Barton-upon-Humber. London was thelargest urban centre in England during the Late MedievalPeriod, and experienced rapid population increase, doublingor even quadrupling from the 13th to the 14th century(Schofield, 2011; Williams, 1963) (estimates of the populationsize at the beginning of the 14th century range from40 000–80 000 (Holt & Rosser, 1990)). With density increasingrapidly in such a short period of time, waste disposal andsubsequent pollution and water contamination were inevit-able problems in London.

Concerns about sanitation in medieval London are evidentin the implementation of legal regulations for waste disposaland basic sanitation systems (Sabine, 1934, 1937). City ordi-nances indicate that London residents were responsible for

Figure 3. Kaplan-Meier survivorship curves with 95% confidence intervals for females of St. Mary Spital and Barton-upon-Humber cemeteries.

Figure 4. Kaplan-Meier survivorship curves with 95% confidence intervals for males of St. Mary Spital and Barton-upon-Humber cemeteries.

Table 4. D30þ/D5þvalues and 95% confidence intervalsfor St. Mary Spital (SRP98) and Barton-upon-Humber(BOH) cemeteries.

D30þ/D5þ 95% CI

SRP98 0.364 0.284–0.444BOH 0.450 0.337–0.563

ANNALS OF HUMAN BIOLOGY 343

dealing with their own waste and that householders andbusinesses were responsible for keeping the areas outside oftheir property clean, properly paved and clear from obstruc-tion (Jørgensen, 2008). In the late 13th century, officials werealso appointed to oversee streets and pavements (Sabine,1937). Several directives for public sanitation were repeatedfrom the 12th to 14th centuries (Sharpe, 1905), which sug-gests that efforts to keep the city clean were not successful.

Unlike London, English villages similar to Barton generallyhad plenty of space to dispose of their refuse away from resi-dences (Hallam, 1981). Food, household and stable wasteswere often taken to a common area at the periphery of thevillage that neighbours shared (Miller & Hatcher, 2014), givento pigs owned by the household (Albarella, 2006) or used foragricultural production in fields (Duby & Postan, 1998; Jones,2011). At the English village of Wharram Percy, a lack of rub-bish pits near households and absence of animal bone nearthe residential area indicate that household waste was cartedto field areas where large quantities of pottery have beenrecovered (Beresford & Hurst, 1991).

Sanitary issues and animal waste in the streets of urbanareas inevitably led to water contamination. Within medievalLondon’s geology, there was a plentiful supply of well andspring water (Keene, 2001). However, most springs wereinconveniently located at the edges of the city, and mostwells were within private properties, only accessible to theowners (Schofield, 2011). Londoners without wells most likelyused suburban streams for their domestic water needs, whichwere polluted as a result of waste drainage (Keene, 2001)and close proximity to latrines (Schofield, 2011)

Water-borne infections are sustained through pollution ofdrinking water, which was not prevented in England untilthe 19th century (Waldron, 1989). Waterleaders sold waterfrom the Thames to residents (Ekwall, 1947), even though italso served as the ultimate method of waste disposal forLondon (e.g. latrines were built over running waterways thateventually emptied into the Thames (Keene, 2001)). Boyd’s(1981) archaeological study of sediments at the Thames andFleet rivers indicate increasing pollution during the sametime that Edward III ordered the mayor of London to cleanthe area in the early 14th century. In 1237, constructionbegan on the first conduit system, which provided taps atintervals throughout the city (Sharpe, 1899). However,because the Conduit could not meet the demand of theLondoners and fee for its use was eventually forced, the poormost likely used the more easily accessible, albeit polluted,Thames as their main source of drinking water (Gummer,2009).

In contrast to London, acquiring fresh water was not anissue at Barton-upon-Humber, because several streams trav-ersed the area, flowing into the marshes of the RiverHumber, and were readily accessible to residents (Rodwell &Atkins, 2011). Residents also used The Beck, which was fedby an artesian spring, to acquire their drinking water (Lyman,2004; Rodwell & Atkins, 2011). Water supplies at Barton werelikely not at a risk for contamination, because rural areas hadmore space for latrines and waste disposal.

High population density paired with unhygienic conditionsmade urban dwellers susceptible to infection. Thus, the

prevalence of infectious disease may have been higher inLondon compared to rural areas. Common diseases noted inlate medieval London include leprosy, tuberculosis and syph-ilis (Roberts & Cox, 2003). BOH exhibits remarkably few (skel-etally diagnosable) cases of infectious diseases, such astuberculosis, compared to contemporaneous urban areas inmedieval England (Waldron, 2007). Waldron (2007) suggeststhat Barton may have been immune from the generalupsurge in the disease; although it is possible that diseasesaffecting the population may have been particularly virulent,causing death before the skeleton was affected.

It is possible that the elevated risks of dying and reduc-tions in survivorship in the urban environment reflect theeffects of the 14th century Black Death and subsequent out-breaks of plague. Although all of England experienced out-breaks of plague, it might have been more problematic inurban areas, as urban dwellers often fled to rural areas toavoid it (Schofield & Vince, 2003). However, there are conflict-ing views on whether towns or rural villages experiencedplague differently. Some scholars argue that higher popula-tion density caused elevated mortality rates in towns(Britnell, 1994; Wrigley, 1969), while others argue that ruralareas experienced higher plague mortality (Benedictow, 2004,2005). Evidence of plague and other short-term fluctuationsin mortality are not evident in the BOH skeletal assemblage(e.g. there is no evidence of multiple deaths during a shortperiod of time), but are apparent in parish records (Waldron,2007). In 1593, an outbreak of plague in Barton resulted in203 deaths (�26% of the population) and was recorded inparish records (Waldron, 2007). This level of mortality is con-sistent with estimates of plague mortality in 17th centuryLondon (Cummins et al., 2015), which suggests that plaguemortality might not explain the urban vs rural differencesobserved in this study.

Healthcare may have been more accessible at Barton thanin London, as several doctors and apothecaries are includedin the parish registers (Watts, 2013). However, these parishregisters did not begin until the 16th century. Care receivedat SRP98 would have involved spiritual care from theAugustinian canons and nursing care from the lay sisters,with little medical intervention, at least until the 15th century(Thomas, 2000). There is some evidence of physicians at thehospital in the late 15th and early 16th century, but it is notconclusive (Thomas, 2000).

During this time, England experienced several faminesresulting from changing climate conditions. These adverselyaffected food production in neighbouring rural areas fromwhich London acquired most of its food (Farr, 1846; Keyset al., 1950). Rural areas like Barton that were dependent onagriculture, however, may have been better buffered fromthe famines that occurred during this time. In London, only afew urban residences had land for small gardens (Connellet al., 2012), while village residences could turn to their ownfields and livestock rather than selling them for sustenanceduring times of poor harvest. Thus, the effect of famine couldhave exacerbated the potentially detrimental effects ofurbanisation, making nutritional differences and, thus, poten-tially health differences between urban and rural populationsmore pronounced.

344 B. S. WALTER AND S. N. DEWITTE

Rural-to-urban migration is one of the most commontypes of human movement in all periods of recorded history(Boyce, 1984), allowing cities with high mortality rates to sus-tain high population densities (McNeill, 1980; Wrigley, 1969).Migrants made up a large proportion of the population inLondon (Wrigley, 1969). Immigrants to urban environmentsmight have faced elevated risk of infection with diseasesthey did not encounter during their childhoods in rural areas.Some studies find that migrants from rural to urban environ-ments in modern populations have elevated risk of diseasecompared to urban residents (Baker, 1984; Velimirovic, 1979;Way, 1976). Moreover, rural-to-urban migrants can beexposed to disease during the migration process (Prothero,1977). Thus, hazardous conditions experienced by rural immi-grants to London may have contributed to the elevated riskof urban adult mortality for SRP98 observed in this study.

Migration into London from neighbouring rural areas wasparticularly common among adolescents and young adults,especially females, because of greater economic opportuni-ties (Dyer, 2002; Hanawalt, 1995). Evidence for this pattern ofrural-to-urban migration is evident in the higher proportionof young adults in the age-at-death distribution of SRP98compared to BOH (Figure 1). According to tax documents,urban cities had more female servants than male servants(Kowaleski, 2014). After the Black Death (1348–1349), thenumber of young females in urban areas increased (Lewis,2016). Female servants were charged with sex-specific foodpreparation tasks, which could have increased their exposureto zoonotic diseases (Smith, 1997). Osteological evidence ofthe strain of domestic labour on urban females compared torural females is evident in medieval English skeletal assemb-lages as well (Lewis, 2016). A significant proportion offemales in medieval London also suffered from poverty(Leyser, 1995), often living in over-populated and unsanitaryconditions, which would have increased their risk of infection(Taylor, 1983). Moreover, the progression from infection todisease and risk of fatality for reproductive aged females ishigher compared to males, at least for some diseases such astuberculosis (WHO, 2003), which was increasing in Englandprior to the Industrial Revolution (Chalke 1962). Through ananalysis of 151 sites in Medieval England, Lewis (2016) foundthat young urban females showed elevated frequencies ofdiseases compared to urban and rural male and rural femaleadolescents (aged 6.5–25 years), which she attributes toyoung urban females being more vulnerable to disease thanother groups. All of these factors could have contributed toincreased risk of mortality for young females migrating fromsurrounding rural areas to London, which is consistent withthe elevated mortality and reduced survivorship of urban vsrural females in this study. Similarly, at St. Helen-on-theWalls, an English urban cemetery, there was a larger propor-tion of females aged 25–34 compared to males (Dawes,1980). Similar patterns were also found in St. Mark’s, Lincolnand York Minster, York and were attributed to female-ledmigration (Grauer, 1991). It is important to emphasise thepossibility that the elevated mortality and reduced survivor-ship of females from SRP98 might actually be a reflection ofa higher proportion of young females migrating into Londonfrom surrounding areas rather than elevated mortality of

females after they migrated to London. With respect to therural cemetery, this sex-specific migration pattern could haveresulted in a lower proportion of young females in BOHbecause they have migrated into urban areas like Londonand, thus, would have been buried in urban cemeteries. Notall females were permanent immigrants, however, as it wasnot uncommon for females to return to their rural villages tomarry after working in an urban area as an adolescent (Bitel,2002; Goldberg, 2004).

Conclusion

The results of this study suggest that the effects of urbanisa-tion in medieval England varied by sex. Comparison of theurban St. Mary Spital cemetery in London and ruralSt. Peter’s cemetery in Barton-upon-Humber via hazards andsurvival analysis of pooled-sex samples indicate that urbanadults, in general, experienced elevated mortality andreduced survivorship. These results are consistent with mostprevious bioarchaeological studies comparing urban andrural environments. Environmental factors that are character-istic of urban centres such as high population density, ele-vated risk of disease, poor sanitary conditions and faminemay have contributed to the detrimental effect of the urbanenvironment on health, and thus mortality and survival.

Analysis by sex, however, suggests that females in theurban sample experienced elevated risks of mortality andreduced survival compared to rural females, whereas malesfaced equal risks in both environments. These results may bea consequence of a higher proportion of females migratingfrom rural to urban centres in search of work opportunities.These females may have suffered from poverty, famine andexposure to pathogens during immigration or upon theirarrival in the city, all of which could have increased their riskof mortality.

This study considers mortality and survival trends amongmales and females in urban and rural cemeteries, allowing usto examine potential heterogeneity within the urban andrural environments. The elevated mortality and lower survivalfor urban females compared to rural females was apparentonce the sexes were analysed separately, and may have con-tributed to the overall mortality differences of adultsbetween the two cemeteries. Although general comparisonsof cemeteries are informative about broad demographic dif-ferences or similarities, it is important to use approaches thatallow for the examination of heterogeneity in past popula-tions so that intra-population variation in mortality patternscan be identified and potential disproportionate effects ofurbanisation on sub-populations may be revealed.Additionally, this study contributes a unique bioarchaeologi-cal line of evidence for the investigation of differential mor-tality in the context of urbanisation beyond primarydocuments that often exclude the poor and migrants, whichcomprised the bulk of the population of Late MedievalLondon.

Finally, urban towns and cities have their roots in rural vil-lages, making the relationship between town and countryinherently complex. Migrants contribute to demographic

ANNALS OF HUMAN BIOLOGY 345

patterns within urban centres, but are often invisible fromprimary historical documents. Future studies using stable iso-tope analysis to identify migrants in urban centres can offervaluable information regarding where migrants were comingfrom, who these migrants were (e.g. young females) andhow these migrants experienced and adapted to the environ-mental factors characteristic of urbanism. The integration ofbiochemical research (e.g. stable isotope analyses) withpaleodemography is essential for disentangling the compli-cated links between mortality, fertility and migration to bet-ter understand medieval life in the face of a changingclimate.

Acknowledgements

We would like to thank Rebecca Redfern and Jelena Bekvalac at theMuseum of London’s Centre for Human Bioarchaeology for providingaccess to the St. Mary Spital collection. We would also like to thankSimon Mays, Kevin Booth and staff of St. Peters Church at EnglishHeritage for providing access to the Barton-upon-Humber skeletal collec-tion and for providing facilities to collect data. Finally, we would like tothank the reviewers for their encouraging comments and helpfulsuggestions.

Disclosure statement

The authors report no conflicts of interest. The authors alone are respon-sible for the content and writing of the paper.

Funding

This work was supported by the National Science Foundation underGrant BCS-1540208 and Grant BCS-1261682; Walker Institute, Universityof South Carolina; SPARC Graduate Research Grant, University of SouthCarolina.

ORCID

Brittany S. Walter http://orcid.org/0000-0002-3246-0461Sharon N. DeWitte http://orcid.org/0000-0003-0754-8485

References

Albarella U. 2006. Pig husbandry and pork consumption in medievalEngland. Oxford: Oxford University Press.

Armelagos GJ, Cohen MN. 1984. Paleopathology at the origins of agricul-ture. Orlando (FL): Academic Press.

Baker P. 1984. Migrations, genetics, and the degenerative diseases ofSouth Pacific Islanders. In: Boyce AJ, editor. Migration and Mobility.London: Taylor and Francis. p 209–239.

Bayliss A, Atkins C. 2011. Specialist Studies, 1: Interpreting Chronology.In: Rodwell W, Atkins C, editors. St. Peter's Barton-upon-Humber,Lincolnshire: A Parish Church and its Community, History,Archaeology, and Architecture Part 2. Oxford: Oxbow Books. p753–788.

Bello SM, Thomann A, Signoli M, Dutour O, Andrews P. 2006. Age andsex bias in the reconstruction of past population structures. Am JPhys Anthropol 129:24–38.

Benedictow Ø. 2004. The Black Death, 1346–1353: the complete history.New York: Boydell and Brewer.

Benedictow Ø. 2005. The Black Death: The Greatest Catastrophe ever.History Today 55:42–49.

Beresford MW, Hurst JG. 1991. Wharram Percy: Desserted MedievalVillage. New Haven, CT: Yale University Press.

Bitel L. 2002. Women in Early Medieval Europe 400–1100. Cambridge:Cambridge University Press.

Bocquet-Appel J-P, Masset C. 1982. Farewell to paleodemography. J HumEvol 11:321–333.

Boldsen JL, Milner GR, Konigsberg LW, Wood JW. 2002. Transitionanalysis: a new method for estimating age from skeletons. In:Hoppa R, Vaupel JW, editors. Paleodemography: age distributionsfrom skeletal samples. Cambridge: Cambridge University Press:73–106.

Boyce AJ. 1984. Migration and mobility: biosocial aspects of humanmovement. Synposia of the Society for the Study of Human BiologyVol 23. London: Taylor and Francis Ltd.

Boyd P. 1981. The paleoecology of estuarine deposits associated witharchaeological sites, with particular reference to the City of London.In: Brothwell D, Dimbleby G, editors. Environmental Aspects of Coastsand Islands. B.A.R. International Series, 94. p 87–88.

Britnell R. 1994. The Black Death in English Towns. Urban History21:195–210.

Bryant G. 2003. The Later History of Barton-on-Humber Part One: theChurch in Late Medieval Barton-on-Humber. Gainesborough: Workers'Educational Association.

Buikstra JE, Konigsberg LW, Bullington J. 1986. Fertility and the develop-ment of agriculture in the prehistoric Midwest. Amer Antiq51:528–546.

Buikstra JE, Ubelaker DH. 1994. Standards for data collection from humanskeletal remains. Fayetteville, Arkansas: Arkansas ArchaeologicalSociety.

Bullock M, M�arquez L, Hern�andez P, Ru�ız F. 2013. Paleodemographicage-at-death distributions of two Mexican skeletal collections: a com-parison of transition analysis and traditional aging methods. Am JPhys Anthropol 152:67–78.

Chalke HD. 1962. The impact of tuberculosis on history, literature andart. Med Hist 6:301–318.

Clapson R. 2005. The later history of Barton-upon-Humber: Part Four. In:Barton and the River Humber 1086–1900. Belton, Gainesborough:Worker's Education Association.

Cohen MN. 1989. Health and the rise of civilization. New Haven, CT: YaleUniversity Press.

Connell B, Gray Jones A, Redfern R, Walker D. 2012. A bioarchaeologicalstudy of medieval burials on the site of St. Mary Spital: excavations atSpitalfields Market, London E1:1991–2007. London: Museum ofLondon Archaeology.

Cox DR. 1972. Regression models and life tables. J Roy Statist Soc Ser BMeth 34:187–220.

Cummins N, Kelly M, �O Gr�ada C. 2015. Living standards and plague inLondon, 1560-1665. Econ History Rev 69:3–34.

Dawes JD. 1980. The human bones. In: Dawes JD, Magilton J, editors.The cemetery of St Helen-on-the-walls, Aldwark: the archaeology ofYork. London: Council for British Archaeology. p 19–120.

De La R�ua C, Izagirre N, Manzano C. 1995. Environmental stress in aMedieval population of the Basque Country. Homo 45:268–289.

DeWitte SN, Wood JW. 2008. Selectivity of Black Death mortality withrespect to preexisting health. Proc Natl Acad Sci USA 105:1436–1441.

Duby G, Postan C. 1998. Rural economy and country life in the medievalWest. Philadelphia (PA): University Of Pennsylvania Press.

Dyer C. 2002. Making a living in the Middle Ages: the people of Britain;850–1520. New Haven (CT): Yale University Press.

Ekwall E. 1947. Early London personal names. Lund: CWK Gleerup.Farr W. 1846. The influence of scarcities and of the high prices of wheat

on the mortality of the people of England. J Stat Soc London9:158–174.

Gage TB. 1988. Mathematical hazard models of mortality: an alternativeto model life tables. Am J Phys Anthropol 76:429–441.

Gage TB, DeWitte S, Wood JW. 2012. Demography part 1: mortality andmigration. In: Stinson S, Bogin B, O'Rourke D, editors. Human biology:an evolutionary and biocultural perspective. Hoboken, NJ: Wiley-Blackwell: 695–756.

346 B. S. WALTER AND S. N. DEWITTE

Goldberg P. 2004. Migration, youth and gender in later medievalEngland. In: Goldberg P, Riddy F, editors. Youth in the Middle Ages.Rochester, MN: York Medical Press: 85–99.

Grauer AL. 1991. Patterns of life and death: the palaeodemography ofmedieval York. In: Bush H, Zvelebil M, editors. Health in past societies.Oxford: BAR International Series. p 67–80.

Gummer B. 2009. The Scourging Angel: The Black Death in the BritishIsles. London: Bodley Head.

Hall AR, Kenward HK. 1994. Urban-rural connexions: perspectives fromenvironmental archaeology. Oxford: Oxbow Books Ltd.

Hallam HE. 1981. Rural England, 1066-1348. San Francisco (CA): Harper.Hanawalt BA. 1995. Growing up in medieval London: the experience of

childhood in history. Oxford: Oxford University Press.Holt R, Rosser G. 1990. The English Medieval Town: a reader in

English urban history, 1200–1540. Boston (MA): Addison-WesleyLongman Ltd.

Hoppa RD, Vaupel J. 2002. The Rostock Manifesto for paleodemography:the way from stage to age. In: Hoppa RD, Vaupel J, editors.Paleodemography: age distributions from skeletal samples. New York:Cambridge University Press. p 1–8.

Jones R. 2011. Elemental theory in everyday practice: food disposal inthe later medieval english countryside. In: Kl�ap�ste J, Sommer P, edi-tors. Processing, storage, distribution of food food in the medievalrural environment, Ruralia VIII, Brepols: Brepois Publishers. p 145–154.

Jørgensen D. 2008. Cooperative sanitation: managing streets and guttersin late medieval England and Scandinavia. Technol Culture49:547–567.

Keene D. 2001. Issues of water in medieval London to c. 1300. UrbanHistory 28:161–179.

Keys A, Bro�zek J, Henschel A, Mickelsen O, Taylor HL. 1950. The biologyof human starvation. Minneapolis, MN: University of Minessota Press.

Konigsberg LW, Frankenberg SR. 2002. Deconstructing death in paleo-demography. Am J Phys Anthropol 117:297–309.

Kowaleski M. 2014. Medieval people in town and country: new perspec-tives from demography and bioarchaeology. Speculum 89:573–600.

Lewis ME. 2016. Work and the adolescent in Medieval England AD900–1550: the osteological evidence. Mediev Archaeol 60:138–171.

Lewis ME, Roberts CA, Manchester K. 1995. Comparative study of theprevalence of maxillary sinusitis in later Medieval urban and ruralpopulations in northern England. Am J Phys Anthropol 98:497–506.

Leyser H. 1995. Medieval women: social history of women in England,450–1500. Oxford: Oxford University Press.

Lyman T. 2004. Barton-upon-Humber Conservation Area AppraisalAdopted Document. [cited 2015 Jan 10]. Available from: www.north-lincs.gov.uk/EasySiteWeb/GatewayLink.aspx?alId=5716

Magnusson RJ. 2013. Medieval urban environmental history. HistoryCompass 11:189–200.

McNeill WH. 1980. Migration patterns and infection in traditional soci-eties. In: Stanley NF, Joske RA, editors. Changing disease patterns andhuman behaviour. London: Academic Press. p 28–36.

Meindl RS, Lovejoy CO, Mensforth RP, Carlos LD. 1985. Accuracy and dir-ection of error in the sexing of the skeleton: implications for paleo-demography. Am J Phys Anthropol 68:79–85.

Miller E, Hatcher J. 2014. Medieval England: rural society and economicchange 1086–1348. London: Routledge.

Milner GR, Humpf DA, Harpending HC. 1989. Pattern matching of age-at-death distributions in paleodemographic analysis. Am J PhysAnthropol 80:49–58.

Milner GR, Wood JW, Boldsen JL. 2008. Paleodemography. In: KatzenbergM, Saunders S, editors. Biological anthropology of the human skel-eton. New York: Wiley-Liss. p 561–600.

Moore M, Gould P, Keary BS. 2003. Global urbanization and impact onhealth. Int J Hyg Environ Health 206:269–278.

Morfin LM, McCaa R, Storey R, Angel A. 2002. Health and nutrition inpre-Hispanic Mesoamerica. In: Steckel R, Rose J, editors. The Backboneof History: Health and Nutrition in the Western Hemisphere.Cambridge: Cambridge University Press. p 307–342.

Nagaoka T, Hirata K. 2007. Reconstruction of paleodemographic charac-teristics from skeletal age at death distributions: perspectives fromHitotsubashi, Japan. Am J Phys Anthropol 134:301–311.

Paine RR. 2000. If a population crashes in prehistory, and there is nopaleodemographer there to hear it, does it make a sound? Am J PhysAnthropol 112:181–190.

Paine RR. 1989. Model life table fitting by maximum likelihoodestimation: a procedure to reconstruct paleodemographic charac-teristics from skeletal age distributions. Am J Phys Anthropol79:51–61.

Prothero RM. 1977. Disease and mobility: a neglected factor in epidemi-ology. Int J Epidemiol 6:259–267.

Roberts C, Cox M. 2007. The impact of economic intensification andsocial complexity on human health in Britain from 4000 BC (Neolithic)and the introduction of farming to the mid 19th century AD andindustrialisation (post-Medieval period). In: Cohen MN, Crane-KramerGMM, editors. Ancient health: skeletal indicators of agricultural eco-nomic intensification. Gainesville (FL): University Press of Florida.p. 149–163.

Roberts CA, Cox M. 2003. Health and disease in britain: from prehistoryto the present day. Gloucestershire, UK: Sutton Publishing.

Roberts CA, Lewis ME, Boocock P. 1998. Infectious disease, sex, and gen-der: the complexity of it all. In: Grauer AL, Stuart-Macadam P, editors.Sex and gender in paleopathological perspective. Cambridge:Cambridge University Press. p 93–113.

Rodwell W, Atkins C. 2011. St Peter's Barton-upon-Humber, Lincolnshire:a Parish Church and its Community. Oxbow: Oxbow Books.

Sabine EL. 1934. Latrines and cesspools of mediaeval London. Speculum9:303–321.

Sabine EL. 1937. City cleaning in mediaeval London. Speculum 12:19–43.Sattenspiel L, Harpending H. 1983. Stable populations and skeletal age.

Amer Antiq 48:489–498.Schofield J. 2011. London, 1100–1600: the archaeology of a capital city.

London: Equinox Publishing Limited.Schofield J, Vince AG. 2003. Medieval towns: the archaeology of British

towns in their European setting. London: A & C Black.Sharpe RR. 1899. Calendar of letter-books of the city of London: A: 1275-

1298. London: John Edward Francis.Sharpe RR. 1905. Calendar of letter-books of the city of London: G: 1352-

1374. London: John Edward Francis.Sidell J, Thomas C, Bayliss A. 2007. Validating and improving

archaeological phasing at St. Mary Spital, London. Radiocarbon49:593–610.

Singman JL. 1999. Daily life in medieval Europe. London: GreenwoodPress.

Smith S. 1997. Geographical diversity in the resort to marriage in latemedieval Europe: work, reputation, and unmarried females in thehousehold formation systems of northern and southern Europe. In:Goldberg P, editor. Women in medieval english society. Oxford:Sutton Publishing. p 16–59.

Steckel RH. 2005. Health and nutrition in pre-Columbian America: theskeletal evidence. J Interdiscip History 36:1–32.

Steckel RH, Rose JC. 2002. The backbone of history: health andnutrition in the western hemisphere. Cambridge: CambridgeUniversity Press.

Storey R. 1985. An estimate of mortality in a pre-Columbian urban popu-lation. Amer Anthrop 87:519–535.

Storey R. 1992. Preindustrial urbanlifestyle and health. In: Huss-AshmoreR, Schall J, Hediger M, editors. Health and lifestyle change.Philadelphia: University Museum of Archaeology and Anthropology:33–42.

Taylor CE. 1983. Synergy among mass infections, famines, and poverty.J Interdisciplin History 14:483–501.

Thomas C. 2000. The priory and hospital of St. Mary Spital, London.Medieval Life 14:21–25.

Thomas C. 2002. The archaeology of medieval London. Stroud, UK:Sutton.

United Nations (UN). 1996. United nations center for human settlements:an urbanizing world. London; Sterling, VA: Earthscan.

United Nations (UN). 2001a. The State of the Worlds Cities 2001. Nairobi:United Nations Human Settlements Program (UN-Habitat).

United Nations (UN). 2001b. World Urbanization Prospects, the 1999 revi-sion. New York: United Nations Population Division.

ANNALS OF HUMAN BIOLOGY 347

Vaupel JW, Yashin AI. 1985. Heterogeneity's ruses: some surprising effectsof selection on population dynamics. Am Statist 39:176–185.

Velimirovic B. 1979. Forgotten people-health of the migrants. Bull PanAm Health Org 13:66–85.

Waldron T. 1989. The effects of urbanism on human health: the evidencefrom skeletal remains. In: Serieantson D, Waldron T, editors. Diet andcrafts in towns: the evidence of animal remains from Roman to thePost-Medieval periods. Oxford: British Archaeological Reports BritishSeries, 199. p. 55–73.

Waldron T. 2007. The St. Peter's Barton-upon-Humber, Lincolnshire: aParish church and its community. Oxford: Oxbow Books.

Walrath DE, Turner P, Bruzek J. 2004. Reliability test of the visual assess-ment of cranial traits for sex determination. Am J Phys Anthropol125:132–137.

Watts R. 2013. Childhood development and adult longevity in an arch-aeological population from Barton-upon-Humber, Lincolnshire,England. Intl J Paleopath 3:95–104.

Way AB. 1976. Morbidity and postneonatal mortality. In: Baker PT, LittleMA, editors. Man in the andes: a multidisciplinary study of high alti-tude Quechua. Stroudsburg, PA: Dowden, Hutchinson, and Ross:147–160.

Weiss KM. 1975. Demographic disturbance and the use of life tables inanthropology. Memoirs of the Society for American Archaeology No30, Population Studies in Archaeology and Biological Anthropology: ASymposium. 46–56.

Williams G. 1963. Medieval London: from commune to capital. London:Athline Press.

Wilson JJ. 2014. Paradox and promise: research on the role of recentadvances in paleodemography and paleoepidemiology to the study ofhealth in Precolumbian societies. Am J Phys Anthropol 155:268–280.

Wood J, Milner G, Harpending H, Weiss K. 1992. The osteological para-dox: problems of inferring prehistoric health from skeletal samples.Curr Anthrop 33:343–370.

Wood JW, Holman DJ, O’Connor KA, Ferrel RJ. 2002 Mortality models inpaleodemography. In: Hoppa RD, Vaupel JW, eds. Paleodemography:age distributions from skeletal samples. Cambridge: CambrdigeUniversity Press. p. 129–168.

World Health Organziation (WHO). 1998. Life in the 21st Century: a visionfor all. Geneva.

World Health Organziation (WHO). 2003. Gender and tuperculosis. Geneva.Wrigley EA. 1969. Population and history. London: Weidenfeld and

Nicolson.

348 B. S. WALTER AND S. N. DEWITTE