AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 128:734746 (2005)

Comparison of Child Morbidity in Two Contrasting

Medieval Cemeteries From Denmark
P. Bennike,1* M.E. Lewis,2 H. Schutkowski,3 and F. Valentin4
1

Laboratory of Biological Anthropology, Institute of Forensic Medicine, University of Copenhagen, DK 2200

Copenhagen N, Denmark
2
Department of Archaeology, Whiteknights, Reading RG6 6AB, UK
3
Department of Archaeological Sciences, University of Bradford, Bradford BD7 1DP, UK
4
CNRS UMR 7041, MAE, 92023 Nanterre, France
KEY WORDS

preadults; stress indicators; growth; BMC; leprosy; Later Medieval Denmark

ABSTRACT
This study compares associations between
demographic proles, long bone lengths, bone mineral content, and frequencies of stress indicators in the preadult
populations of two medieval skeletal assemblages from
Denmark. One is from a leprosarium, and thus probably
represents a disadvantaged group (Nstved). The other
comes from a normal, and in comparison rather privileged,
medieval community (belholt). Previous studies of the
adult population indicated differences between the two
skeletal collections with regard to mortality, dental size,
and metabolic and specic infectious disease. The two
samples were analyzed against the view known as the
osteological paradox (Wood et al. [1992] Curr. Anthropol.
33:343370), according to which skeletons displaying pathological modication are likely to represent the

healthier individuals of a population, whereas those without lesions would have died without acquiring modications as a result of a depressed immune response. Results
reveal that older age groups among the preadults from
Nstved are shorter and have less bone mineral content
than their peers from belholt. On average, the Nstved
children have a higher prevalence of stress indicators, and
in some cases display skeletal signs of leprosy. This is
likely a result of the combination of compromised health
and social disadvantage, thus supporting a more traditional interpretation. The study provides insights into the
health of children from two different biocultural settings
of medieval Danish society and illustrates the importance
of comparing samples of single age groups. Am J Phys
Anthropol 128:734746, 2005. V 2005 Wiley-Liss, Inc.

It was demonstrated that any attempt to reconstruct

health patterns of past populations from skeletal remains
is paradoxical, as these data can only provide information
about the morbidity and mortality rates of nonsurvivors
(Wood et al., 1992). The frequencies of disease in a skeletal
assemblage would be higher than for the surviving population from which they were derived. Individuals displaying pathological lesions, including stress indicators, in a
skeletal population were traditionally thought to represent the less advantaged members of that society. These
individuals may have had a depressed immune system,
may have been malnourished, and probably lived in an
overcrowded, unhygienic environment that increased
their risk of exposure to a greater number of pathogenic
organisms. In contrast, those without lesions were
thought to represent that part of the population who, perhaps due to nutritional and cultural advantages, encountered fewer episodes of stress until time of death. However, Ortner (1991, p. 10) stated that the presence of skeletal lesions diagnostic of infectious diseases implies a good
immune response and a relatively healthy individual.
Thus, in order to display pathological lesions on the skeleton, the individual must either recover from or adapt to
the stress. Ortner (1991, p. 10) went on to state that such
an immune response is not as effective as the one that successfully rids the body of the infectious organism during
the early stages of the disease. This led Wood et al. (1992)
to question basic and widely accepted assumptions about
past health patterns by pointing out the osteological paradox and to provoke researchers to revisit the methods
and interpretative techniques employed in the discipline.

The critical response by Goodman (1993) highlighted,

among other things, the need to organize samples into age
categories and use multiple indicators of disease and
nutritional stress in order to identify any hidden subgroups within a given sample. In addition, Saunders and
Hoppa (1993) examined the question of linear growth in
survivors and nonsurvivors in order to ascertain whether
skeletal samples accurately represented the population
from which they were derived. Their study revealed that
short stature seems to be disease-dependent.
The use of stress models, such as those developed by Little (1995) and Goodman et al. (1984), illustrates the complex nature of factors contributing to, and resulting from,
physiological stress. Morbidity, mortality, and fertility are
inuenced by the synergistic interaction of environmental
parameters, economics, and cultural restraints, expressed

C
V

2005 WILEY-LISS, INC.

Grant sponsor: EU Program Human Capital and Mobility; Grant

numbers: ERBCHRXCT 930193.
*Correspondence to: Dr. Pia Bennike, Laboratory of Biological
Anthropology, University of Copenhagen, Blegdamsvej 3, DK-2200
Copenhagen N, Denmark.
E-mail: bennike@antrolab.ku.dk
Received 7 January 2004; accepted 22 September 2004.
DOI 10.1002/ajpa.20233
Published online 25 July 2005 in Wiley InterScience
(www.interscience.wiley.com).

735

CHILD MORBIDITY IN DENMARK

for example in birth rates, weaning practices, and food
taboos that may buffer or induce stress in a specic population. The impact of stress relies on host resistance determined by genetic factors, age, and sex (Goodman et al.,
1988; Wall, 1991). An emphasis on the study of multiple
indicators of stress in skeletal material thus represents a
move away from the identication of specic diseases, to
an assessment of nutritional and disease status within a
population (Goodman et al., 1988).
In the light of this discussion and the fact that biocultural studies of childrens skeletons in medieval Europe
are limited (but see Lewis, 2002a,b), two skeletal samples
comprising infants and juveniles (henceforth referred to
as preadults) from medieval Denmark were examined.
The Danish Middle Ages (AD 1050 until the dissolution of
the Roman Catholic Church in 1536) saw a rise in contact
with the rest of Europe due to the introduction of Christianity, increased trading, and warfare. The vast majority
of the medieval population (about 90%) lived in the countryside and worked on the land. During the second half of
the Middle Ages, almshouses provided shelter for the sick
and poor in the cities. However, those who were aficted
with leprosy were ostracized and sequestered into special
hospitals known as leprosaria (Bennike and Brade, 1999;
Arentoft, 1999). Both assemblages are from rural sites;
the rst consists of individuals interred within a leprosarium at Nstved, (Southern) Zealand, during the Late
Medieval period (Mller-Christensen, 1978). The second
sample was contemporary to the rst and derived from a
monastic community at belholt (Northern Zealand) and
the surrounding parish (Mller-Christensen, 1982). Many
of these individuals were expected to have been relatively
healthy in comparison to those at Nstved, and the majority probably entered the mortuary sample as the result of
acute disease or social standing allowing them to afford a
burial. Even though documentation is rather scarce, the
skeletons are supposed to represent not only the prior,
monks, and novices. The monastery provided lodging for
rich people and shelter for the poor; food was free. Some of
the visitors were travellers, while others came to receive
treatment and get cured, as the monastery housed a small
hospital. The number of monks was probably no higher
than 12, including the prior. The monastery employed laymen to carry out various practical tasks, and their numbers may have been considerable in relation to the monks.
Thus, in addition to the brethren and some of the sick who
may have died during their stay, all other skeletons, as
the large numbers of female and childrens skeletons
show, are most likely to represent members of the laymens families, together with members of richer families
who could afford high-status burials inside the church
(Mller-Christensen, 1982).
Previous studies of adult skeletons illustrated differences in the disease patterns of the two populations. More
than 70% of the Nstved skeletons were reported to show
clear evidence of changes due to leprosy (Mller-Christensen, 1978). Fifty-ve percent of them had cribra orbitalia,
as opposed to only 20% at belholt (Mller-Christensen,
1978; Mller-Christensen and Sandison, 1963). Moreover,
the adult mortality prole revealed that many more from
the leprosarium died between 2035 years of age (40% in
Nstved vs. 14% in belholt) and only a few at age 55
and above (2% in Nstved vs. 12% in belholt) (MllerChristensen, 1978). Lunt (1969) demonstrated that the
permanent teeth from Nstved in general were slightly,
yet not signicantly, smaller than those from the belholt
monastery. It therefore can be reasonably assumed that

the preadults from Nstved also belonged to a less privileged community.

If traditional interpretations were correct, one would
expect that, as the more advantaged group, those from
belholt would have fewer early deaths and a lower frequency of stress indicators. In contrast, those at Nstved
should die earlier on average and display more stress indicators, and their growth should be stunted compared with
their peers from belholt, thus illustrating lifetimes
stricken with illness and all its social ramications. However, if skeletal assemblages do indeed reect a paradox,
then one would expect preadults at belholt to display
more stress indicators as a result of their biological and
cultural advantage and ability to be strong enough to
recover from disease insults.
Age-at-death of individuals with and without stress
indicators was already used to help identify those most at
risk from morbidity and mortality. Goodman and Rose
(1990) associated hypoplasia in Guatemalan children with
a lower mean age-at-death. This nding was conrmed in
a study of prehistoric Native Americans from the Libben
site by Duray (1996). Both Van Gerven et al. (1981) and
Mittler et al. (1992) found a high correspondence of cribra
orbitalia and the probability of dying in infancy and early
adolescence in a Nubian sample, and Cook and Buikstra
(1979) found an association of those with linear enamel
hypoplasia on deciduous teeth with increased mortality.
In her study of a British population from York, Grauer
(1993) reported increased ages-at-death for those displaying remodeled periostitis once adulthood had been
reached, while Mensforth (1985) and Mensforth et al.
(1978) found a decreased mean age-at-death for those with
unremodeled porotic hyperostosis and periostitis in the
Libben sample.
With the exception of Goodman and Rose (1990) and
Cook and Buikstra (1979), all these studies involved using
mean age-at-death for the whole skeletal sample. As will
be demonstrated, this is likely to obscure the more subtle
variability in death rates within the preadult sample,
where these lesions rst appear. It was shown that stress
indicators are associated with decreased age-at-death in
adulthood, but is this also true of the preadult sample?
Are those without any stress indicators outliving the children displaying them or did they succumb to death before
the lesions could appear? By restricting the study to preadults (those under 20 years of age), it is hoped to provide a
more precise picture of the nature of stress indicators in
relation to age patterns. In particular, the study aims to
address whether preadults from the leprosarium of
Nstved were shorter, had lower bone mineral content,
and developed more skeletal modications following episodes of stress than their contemporaneous peers from the
parish community at belholt.

MATERIALS
The morbidity and mortality proles of 259 preadult
skeletons from later medieval Sjlland (Zealand) in Denmark were compared. The rst sample comprised 76 preadult skeletons interred in the cemetery connected to St.
Georges Hospital at Aaderup, Nstved (total number
650), one of 31 leprosaria established in Denmark in the
later medieval period (AD 12501550). The second comprised 183 preadults from the cemetery connected to the
Augustinian monastery at belholt between AD 1175
1550 (total number 756). Both sites are contemporane-

736

P. BENNIKE ET AL.

ous and, being only 150 km apart, from the same geographical area.
Mller Christensen excavated the skeletons at belholt
from 19351965 and uncovered burials within the cloister
ambulatory and garth (Bennike, 2002). The adult skeletons, 60% males and 40% females, together with the 183
children, suggested that some individuals were the married servants of the monastic community and their offspring. Further burials in the nave of the church and the
churchyard to the northwest of the monastery represented
monks and privileged lay persons (Mller Christensen,
1982). Thus, those buried in the cemetery probably represent children of farmers connected to and employed by the
monastery, pupils, and the children of parents who could
afford to have them buried within the foundations. This is
supported by the fact that the age distribution at
belholt is very similar to that of other medieval cemeteries in Denmark.
The discoveries at belholt prompted the excavation of
the cemetery at Nstved (19481968). The hospital at
Nstved was rst mentioned in AD 1261 and is known to
have functioned solely as a leprosarium until AD 1542,
when the Ribe Recess ordered all leprosy hospitals in Denmark to be disestablished and patients to be admitted to
larger general hospitals. The socioeconomic status of those
interred at Nstved would have varied, as inmates were
admitted to the hospital from the town of Nstved and
the surrounding area (Mller Christensen, 1961; Andersen, 1969).
Documentary evidence suggests that the Danish medieval community could recognize the classic signs of leprosy
as most of the documents describe the classic facial
changes (Andersen, 1991). In fact, no clear evidence of leprosy was found in over 300 skeletons excavated from an
almshouse (or Helligandshus) in the center of Nstved
(AD 14001800) (Macey, 1996) or in 60 skeletons of a mass
grave nearby dated to the 14th century (Bennike, 2000).
However, in a study of Danish and Swedish medieval skeletons based on a number of specic and nonspecic criteria used for calculation of specicity and sensitivity of the
diagnosis of leprosy, Boldsen (2001) claimed that a high
proportion of the rural Danish population was infected
with leprosy in the 13th century. This conclusion still has
to be proven.
The preadults at Nstved may have entered the sample
in a number of ways. Firstly, children may have been
admitted to the hospital when they presented soft-tissue
lesions associated with leprosy. Today, some children born
to parents with lepromatous leprosy normally develop
clinical signs of the disease by age 5 years (Melsom et al.,
1982; Lewis, 2002c). Richards (1977) cited many cases of
children being examined and admitted to leprosaria in his
study of the disease in Scandinavia, and it was not
unusual to separate the child from its parents. However,
today, infants displaying single leprous lesions are often
overlooked if they are concealed by clothing or ignored by
the mother, and the lesions may heal spontaneously (Lara
and Nolasco, 1956), only to recur at a stressful time during the childs development, e.g., during puberty. Therefore, it is unlikely that the youngest preadults in this sample (02.5 years) were admitted to the hospital diagnosed
with leprosy; yet congenital contraction of the disease cannot be ruled out.
There were equal numbers of male and female skeletons, and some children may have accompanied their
infected parent or parents into the hospital if they were
suspected of being contaminated or, more likely, to pre-

vent them becoming orphaned. Richards (1977, p. 60)

cited the case of a young woman from Sweden who had
entered the hospital with her parents and, although she
never developed the disease, remained there for 16 years.
Some preadults may represent babies born to women with
lepromatous leprosy or to the healthy wives of inmates.
Early accounts from Nstved referred to womenfolk of
sick brothers and sisters entering the leprosarium to care
for their family (Richards, 1977, p. 36). Husbands and
wives were often segregated or advised to divorce when
they entered a leprosarium (Brody, 1974), and male infertility sometimes caused by the disease would have
decreased the number of children born in the hospital.
Thus, the sample at Nstved provides an unusual opportunity to study the skeletons of preadults exposed to
Mycobacterium leprae.
From historical records, it is known that leprosaria in
Scandinavia were almost completely dependent on the
parishes to which they belonged for food supply, nancial
support, maintenance, and administration (Andersen,
1969; Richards, 1977; Arentoft, 1999). It is also possible
that, in times of hardship in the parish, these duties were
neglected and would have led to poorer hygienic and dietary conditions, which would have taken their toll on
already disease-burdened inmates.

METHODS
Age at death
Data from preadult skeletal material are widely
believed to represent the most demographically variable
and sensitive barometer of biocultural change (Van
Gerven and Armelagos, 1983; Roth, 1992). Patterns of infant and child mortality were shown to have a profound
effect on the crude death rates of a population and, when
coupled with evidence of childhood morbidity, have become
accepted as a measure of population tness and an indicator
of fertility (Mensforth et al., 1978). Although the differential
preservation of fragile preadult skeletons, and selective burial due to cultural factors, may present a serious problem
with underrepresentation, preadult data are relatively free
from the problems of age assignment that plague paleodemographic studies based on adult skeletons. Age estimates
of preadult skeletons are derived using age-specic markers
of growth and maturation. The timing and outcome of these
events are thought to be tentatively linked to a genetic blueprint, as opposed to the degenerative age determinants of
adulthood (Rallison, 1986). For this reason, preadult skeletons allow for more precise skeletal age estimation than is
possible for adults.
Growth studies on modern populations showed that
dental development, characterized by tooth formation
and eruption, is less sensitive to environmental inuences such as malnutrition and infection, and is more
genetically controlled than skeletal maturation and size
(Schour and Massler, 1940; Acheson, 1959; Stini, 1985;
Eveleth, 1986). Dental development is believed to be the
most accurate indicator of age-at-death on the preadult
skeleton (Lewis and Garn, 1960; Smith, 1991, Hillson,
2000).
The age distribution of any given skeletal sample is
determined by the mortality and fertility distribution, and
therefore, growth patterns determined from nonsurvivors
may not provide a true reection of the growth pattern of
surviving individuals in an age group (Johnston, 1962; Van
Gerven and Armelagos, 1983; Jantz and Owsley, 1984;
Saunders and Hoppa, 1993; Wood et al., 1992). On the other

CHILD MORBIDITY IN DENMARK

TABLE 1. Mean ages at death
belholt

Nstved

Age

Mean

SD

Mean

SD

<2.5
2.66.5
6.610.5
10.614.5
14.620.0

43
57
35
17
9

1.17
4.73
8.86
12.20
16.61

0.71
1.13
1.12
0.99
1.87

14
11
17
11
20

1.32
4.98
8.95
12.04
17.62

0.59
0.79
1.20
0.99
2.94

161

6.19

73

9.49

Total mean

hand, when attempting to compare the health status of two

cemetery samples, age-at-death and growth data coupled
with evidence of disease and nutritional stress can provide
useful information on selective mortality in certain cultural
or ecological circumstances.
Skeletons were omitted from the preadult (ca. <20
years) sample when the epiphyseal head had completely
fused to the femoral diaphysis. The femoral head was
chosen to exclude individuals above age 20 years (Buikstra and Ubelaker, 1994). Age-at-death for preadults was
assessed by the crown and root formation of the deciduous
and permanent dentition, as described by Moorrees et al.
(1963a,b) and tabulated by Smith (1991). When no teeth
were preserved, epiphyseal fusion was employed (Buikstra and Ubelaker, 1994). This was necessary in only six
cases. The age of perinates was estimated using diaphyseal long bone lengths (Olivier and Pineau, 1960; Scheuer
and Black, 2000).
Skeletons were divided into ve age categories: newborn
(0)2.5 years, 2.66.5 years, 6.610.5 years, 10.614.5
years, and over 14.6 years. The average age in each category was calculated for both cemeteries (Table 1). By using
broad age categories, it was hoped to reduce the errors
introduced by inter- and intrapopulation variability of
age-related skeletal traits (Lampl and Johnston, 1996).

Bone length, bone mineral content, nonspecic

stress, and disease
Skeletons were examined for bone length, bone mineral
content (BMC), and indicators of stress including cribra
orbitalia and dental enamel hypoplasia. Evidence of nonspecic infections (maxillary sinusitis, endocranial lesions, and periostitis) was also recorded, as was evidence
of leprosy in the Nstved preadult sample.
Today, child health studies rely on anthropometric
measurements such as height for age, weight for height,
and arm circumference; a decit in these measurements
may increase the individuals risk of death (WHO Working Group, 1986; Saunders and Hoppa, 1993). A lower
height for age and slowed maturation in a given population are associated with malnutrition, disease, and physiological stress (Scrimshaw et al., 1959; Eveleth and Tanner, 1990; Mensforth, 1985). During a stressful period,
the undernourished child slows down and waits for better times (Tanner 1990, p. 130). The level of slowed
growth relates to the severity of the episode, with the
fetal stage and the rst 3 years of life shown to be a particularly susceptible period for growth retardation (Eveleth and Tanner, 1990; Saunders and Hoppa, 1993). Stini
(1985) referred to growth retardation as developmental
adaptation, as this response should increase an individuals chance of survival due to fewer nutritional requirements. Seckler (1982) took this interpretation one step
further and suggested that small is healthy, but since
growth retardation is correlated with negative health fac-

737

tors, it is attributed to an adjustment to environmental

stress factors, rather than adaptation per se (Scrimshaw
and Young, 1989, cited in Saunders and Hoppa, 1993;
Beaton, 1989).
Once equilibrium has been restored after an episode of
stress, children have the ability to catch up with their
healthy peers and resume growth at a rate that will enable them to reach the target growth curve. This ability
to increase the velocity of growth until the optimum level
is reached is impaired with age and cannot occur during
or after puberty (Rallison, 1986; Eveleth and Tanner,
1990). Mensforth (1985) correlated growth retardation
with iron-deciency anemia in the Libben Late Woodland
sample, and attributed it to nutritional stress in the weaning period. Due to its association with malnutrition and
disease, growth retardation is expected to be associated
with an increase in the frequency of stress indicators in
our samples. Infection and malnutrition were also shown
to affect growth and maturation in modern populations
(Acheson and Macintyre, 1958; Stini, 1985; Duncan, 1980;
Cole, 1989; Wall, 1991). Therefore, it is expected that preadult individuals at Nstved are smaller and have lower
BMC values than their counterparts from belholt.
Only a handful of studies examined bone cortical thickness as an indicator of childhood stress in skeletal populations (e.g., Huss Ashmore, 1978; Martin and Armelagos,
1979; Cook, 1984), and to the knowledge of the authors,
even fewer examined bone mineral content in preadult
skeletal samples. Such data are widely accepted to serve
as standards for bone mineralization (Thompson, 1980;
Malina and Bouchard, 1991). They are generally considered to reect the prevailing status of physical activity
and nutritional supply as well as (for adults) hormonal
inuences connected with mineralization, growth, and
remodeling of bone (Carlson et al., 1976; Thompson, 1980;
Ruff and Hayes, 1984; Bennike et al., 1993). As malnutrition, or more specically protein calorie deciency, was
shown to affect linear growth and maturation, a change in
the appositional growth of bone laid down in this period
should also be expected.
Mean diaphyseal length was plotted against dental age
to produce skeletal growth proles for the humerus,
femur, and tibia, and BMC measurements were performed
on femora and humeri. Specimens were rejected for BMC
measurements if there were visible signs of erosion at
midshaft. In addition, radiographs were taken in order to
exclude bones with evidence of soil in the medullary cavity. BMC measurements were performed on a Gammatec
GC 50 Dual Photon scanner (for a detailed description of
the method, see Bennike et al., 1993). The region of interest was dened as being 61 cm around the diaphyseal
midshaft, corresponding to 10 scans, each with a breadth
of 2 mm. This scanning breadth was applied to all specimens, irrespective of age. Quality control included a check
of the scanner parameters against a marble standard at
the beginning and end of a measurement set. Precision
was checked by 21 replicate measurements of one
archaeological bone specimen, a 331-mm femur from an
11.4-year-old child. The measurement results of the marble standard (n 36) were obtained with an error of
0.31%, and the mean mineral content (6.50 6 0.047 units/
cm) lay within the known 95% condence interval of the
standard (6.496.50 units/cm). Precision measurements
from the control femur revealed a small measurement
error (coefcient of variation) of 0.89%.
Cribra orbitalia is thought to be indicative of childhood
malnutrition and/or pathogen load resulting in iron-de-

738

P. BENNIKE ET AL.

ciency anemia (Mensforth et al., 1978, Weinburg, 1992;

Stuart-Macadam and Kent, 1992 ). It was also suggested
that acute and chronic infections may stimulate the
immune system to withhold iron from invading microorganisms as a defense mechanism. In some studies, cribra
orbitalia was thought to be a useful indicator of dietary
deciency of iron in past populations (El-Najjar et al.,
1979; Cohen and Armelagos, 1984; Gilbert, 1985). However, clinical studies and bone chemical analyses questioned the validity of this association or suggested a link
between metabolic deciencies and severity of lesions
(Reinhard, 1992; Schutkowski and Grupe, 1997; Subira
et al., 1992). Ortner and Ericksen (1997) and Ortner et al.
(2001) suggested that vitamin C deciency (scurvy)
caused cribra orbitalia, while in a recent study of adult
skeletons, Wapler et al. (2004) showed that the microscopic appearance of cribra orbitalia may be due to several pathological causes (including anemia in 3040%) as
well as postmortem erosions (2 %).
In this study, lesions were graded according to the
scheme by Stuart-Macadam (1991, p. 109), but lesions
described as type 1, which show very weak capillary-like
impressions, were not recorded, as these lesions are considered too mild to represent serious health problems. We
also found that the very mild cases could easily be mistaken for postmortem changes and could lead to considerable inter- and intraobserver bias. This exclusion obviously means that only the tip of the iceberg was registered, which is common in most skeletal studies.
Furthermore, active and healed cases of cribra orbitalia
were not registered separately, as they proved to be hard
to distinguish from one another macroscopically.
Dental enamel hypoplasia was identied as grooves or
pits on the dental enamel surface of the deciduous and
permanent dentition. They represent a prenatal or childhood growth disturbance due to disease or malnutrition,
and are associated with decreased longevity (Larsen,
1997). Measurements were taken from the proximal margin of the defect to the cemento-enamel junction at the
anterior aspect of the deciduous and permanent teeth.
Enamel hypoplasia was considered present when more
than two teeth on opposite sides of the jaws were affected,
and absent when four or more anterior teeth were available for examination but did not show any lesions. Premolars were included in this number as, in some individuals
of the leprosy sample, the central incisors may have been
lost due to atrophy of the maxillary alveolar bone, and
may have caused the defects to be under-recorded. The
age of defect formation in 1-year age categories was determined using a conversion chart (Goodman and Rose,
1991, p. 288). The mean age of formation and the number
of lines in each sample were calculated.
New bone formation on the endocranial surfaces was
recorded morphologically as either layers of new bone on
the original cortical surface, hair-on-end extensions of
the diploe, or as vascular impressions extending into the
inner lamina of the cranium (Schultz, 2001). New bone
formation on the postcranial skeleton was recorded as
healed (lamellar bone formation) or active (immature new
bone) (Ortner, 2003). Lewis (2004) discussed the etiology
of endocranial lesions and suggested that they should be
referred to as infection or nonspecic indicators of hemorrhage, which may be caused by vitamin C deciency or
child abuse.
Periostitis on the tibiae and bulae is common in individuals with leprosy, and in these cases was considered nonspecic in its etiology. Maxillary sinusitis was recorded accord-

ing to the criteria of Boocock et al. (1995), with the exception

of large pits that are believed to represent growth changes
in the antra (Lewis et al., 1995). For this reason, it was only
recorded in individuals from age 2.6 years onward, when
sinus development is advanced enough to allow more effective
visualization of the cavities (Maresh, 1940). Maxillary sinusitis has many etiologies, including environmental air pollution, upper respiratory tract infections, and dental disease
(Wald et al., 1981; Lewis et al., 1995; Merrett and Pfeiffer,
2000). More specically, maxillary sinusitis is associated with
inammation of the rhinomaxilla in leprosy (Wright, 1979).
Chi-square tests for 2
2 and 2
c tables, with Yates
continuity correction (Kirkwood, 1988), were carried out to
test the differences in prevalence of stress indicators
between sites. Wilcoxons rank correlations were performed
to test for differences in long bone lengths. Mean age-atdeath of those individuals with and without stress indicators was tested using the Kolmogorov-Smirnov statistic.
Leprosy was diagnosed when the characteristic lesions,
distributed on the maxilla (frontal alveolar absorption,
porotic areas on the palate, resorption of the nasal spine,
and remodeling of the nasal area), hands and feet (mainly
resorption and remodeling), or lower limbs (periosteal
reaction), were recognized (Lewis, 2002c). Lepromatous
leprosy causes bilateral neuropathy, and in older individuals can cause osteopenia in males due to reduced testosterone levels (Jopling and McDougall, 1988). Therefore,
BMC measurements were compared between the two
samples.

RESULTS
Age distribution
In order to assess whether the preadult age distribution at belholt was attritional, and thus comparable to
other contemporary Danish populations, a comparison
was made with a sample of preadult skeletons (n 252)
from a completely excavated medieval rural cemetery at
Tirup (Kieffer Olsen, 1993). Despite a slight difference in
the ranges of age groups, the demographic pattern at
belholt was similar to that of Tirup. There were 60%
and 61% of preadults between 06.5 years at Tirup and
belholt, respectively, with 33% of the sample making up
children between 6.614.5 years, and 6% in the 14.6
20.0-year age categories, with a similar distribution at
both sites. This similarity in age distribution with a normal medieval cemetery seems to indicate that the preadult skeletons at belholt were not a special, selected
group, e.g., one derived from the hospital section of the
monastery. Contrary to this pattern, we found that when
the mortality proles of Nstved and belholt were compared, the two populations had signicantly different distributions in all age groups (
2 28.3, df 4, P < 0.01)
(Fig. 1). For example, at Nstved, a signicantly lower
number of individuals was in the 2.66.5-year age group
(
2 9.5, P < 0.01), representing only 15% of the total
sample (belholt, 35 %), and a signicantly higher number of 14.620.0-year-olds (
2 21.7, P < 0.01), accounting for 27% of the Nstved population (Table 1). Calculations of average age in each age group were rather similar.

Indicators of bone length, BMC, and

nonspecic stress
Growth proles for both sites followed a similar pattern
until the 10.614.5-year age category, where the Nstved

739

CHILD MORBIDITY IN DENMARK

Childhood leprosy at Nstved

Fig. 1. Age distribution.

children were signicantly shorter than their peers from

belholt by up to 3.8 cm for femoral diaphyseal lengths
(P < 0.05) (Table 2, Fig. 2). The BMC-for-age data are presented in Table 3 and Figure 3 and reveal congruent patterns, which seem to support the linear growth data.
The average prevalence of stress indicators at belholt
was lower than in preadults from Nstved, who had signicantly higher rates of enamel hypoplasia and maxillary
sinusitis (Table 4, Fig. 4). In both samples, the prevalence of
endocranial lesions peaked in the 02.5-year age category
and was signicantly higher at Nstved than belholt in
the 2.66.5-year age category. When the prevalence of stress
indicators was compared within samples, cribra orbitalia
peaked in the 2.66.5-year age group at belholt, whereas
at Nstved, cribra orbitalia was most prevalent in the 14.5
20.0-year age category (Table 4). The age of hypoplasia formation, divided into 1-year periods, showed that the majority
of defects were formed between 2.63.5 years of age in both
groups, with 55% (84 of 153 defects in 23 individuals) affected
at belholt and 41% (68 of 167 defects in 21 individuals) at
Nstved (Fig. 5).
Mean ages at death were calculated for individuals with
and without stress indicators. The results suggested that
children with stress indicators in general were living longer than those without (Table 5). At belholt, children
with enamel hypoplasia were 10.8 years old on average,
i.e., 4.5 years older than those not displaying the lesion
(K 2.0, P 0.01, where K Kolmogorov-Smimov). Those
with cribra orbitalia also lived longer than those without
(on average, 2.8 years), but the difference was not signicant. Those with maxillary sinusitis were associated with
signicantly increased longevity (K 1.43, P 0.05), as
they were living 2.4 years longer than those without. At
Nstved there was a similar trend, although not signicant: those with hypoplasia, periostitis, or rhinomaxillary
syndrome (pathognomonic of leprosy) survived longer
than those who did not display the lesions.
However, in the case of endocranial lesions, the pattern
was reversed. At belholt, the presence of endocranial
lesions was associated with decreased longevity. Those
displaying lesions died around age 1.6 years, and those
without lived signicantly (K 1.78, P 0.05) longer and
attained a mean age-at-death of 6.8 years. At Nstved,
children without endocranial lesions also lived signicantly (K 1.84, P 0.05) longer (7.5 years) than those
with the lesions (3.4 years).

At Nstved, 16 of 60 children that could be examined

(27%) had evidence of leprosy. Fifteen of those (25%) were
diagnosed with lepromatous leprosy (LL) on the basis of
characteristic changes to the rhinomaxilla. Thirteen of 16
had postcranial bones preserved as well, and of those,
seven (54%) displayed additional postcranial lesions. One
further individual, whose maxilla was missing, was diagnosed on the basis of bilateral periostitis on the tibiae and
bulae and destructive changes to the feet. The changes
associated with lepromatous leprosy were found to
increase in older individuals. The three youngest children
belonged to the 6.610.5-year age category, representing
18% of that group. When the prevalence of stress indicators was compared between those with and without visible
signs of leprosy, there was a higher prevalence of periostitis and maxillary sinusitis (
2 9.13 and 7.81, respectively; P 0.005) in children diagnosed with leprosy. However, with the exception of shorter humeri in the Nstved
adolescents (14.620.0 years) with leprosy compared to
those at belholt, the growth proles of leprous and nonleprous individuals at Nstved did not differ. Due to the
possibility of regional anesthesia or disuse, the leprous
children were expected to display lower bone length and
BMC in comparison to the rest of the sample. However,
although all the leprosy children were generally found to
have shorter long bones and lower BMC values than the
other children at both sites, none of these differences was
signicant.

DISCUSSION
Is it logical to assume that a group of skeletons displaying pathological lesions and other stress indicators represent the disadvantaged of the society from which they
were derived? Wood et al. (1992) argued that this basic
assumption, commonly used to measure past population
morbidity, is paradoxical. In order to display a stress indicator such as enamel hypoplasia or cribra orbitalia, it is
necessary to recover from the stress that resulted in a cessation of growth (as these lesions only become visible
when normal growth is resumed), or some kind of longstanding condition has to be present. Thus it follows that
only those with a strong immune system and perhaps cultural advantages are likely to survive episodes of stress.
In fact, should not weaker or culturally disadvantaged
individuals enter the mortality sample with no or fewer
skeletal markers of stress and disease because of their
inability to recover from an episode of stress? When all
preadults were combined in each of the two samples, this
assumption could not be conrmed because the less
advantaged group from the leprosy cemetery displayed
more stress-related skeletal modications throughout
(Fig. 3). However, when the various age groups of the two
samples were compared separately, a much more complex
pattern emerged (Table 5) that demonstrates the importance of subdividing a sample into age groups in comparative studies. Contrasting the groups with and without
skeletal stress indicators revealed higher average ages of
those with lesions than those without, except for endocranial lesions. Thus, at least in terms of longevity, yet contra
other parameters, overcoming episodes of stress does
mean surviving them by a measurable margin, which
lends some support to Wood et al. (1992).
In order to assess those most at risk from stress, it is
important to be aware of the environmental and cultural

740

P. BENNIKE ET AL.
TABLE 2. Diaphyseal lengths for long bones (mm)
Diaphyseal lengths (mm)
Humerus

Age (years)

belholt
02.5
2.66.5
6.610.5
10.614.5
14.620.0

26
42
23
9
6

Total

95.8
159.0
208.3
255.1
302.5

6 19.9
6 21.6
621.2
6 27.9
6 24.4

Femur
Range

75.9115.7
137.4180.6
187.1229.5
227.2283.0
278.1326.9

24
39
23
8
5

106

Nstved
02.5
2.66.5
6.610.5
10.614.5
14.620.0

6
6
12
8
13

Total

45

Mean

Tibia

Mean

Range

6
6
6
6
6

82.5144.7
179.5245.1
264.3322.6
319.4399.2
387.3449.9

16
31
20
9
3

113.6
212.3
293.9
359.3
418.6

31.1
32.8
29.6
39.9
31.3

99
99.6
140.8
209.4
228.5
275.6

6
6
6
6
6

13.6
36.9
15.4
24.2
29.6

86.0113.2
103.9177.7
194.0224.8
204.3252.7
246.0305.2

5
4
7
7
10

Mean

Range

96.5
169.9
225.4
292.4
375.0

6
6
6
6
6

24.3
22.6
22.9
33.3
22.6

72.2120.8
147.3192.5
202.5248.3
259.1325.7
352.4397.6

97.5
174.0
235.0
261.0
315.0

6
6
6
6
6

16.0
33.2
16.4
20.4
24.3*

81.5113.5
140.8207.2
218.6251.4
240.6281.4
290.7339.3

79
114.8
224.5
299.7
330.4
380.9

6
6
6
6
6

28.7
39.8
20.7
26.5
29.3*

86.1143.5
184.7264.3
279.0320.4
303.9356.9
351.6410.2

33

3
4
6
6
7
26

Samples differ signicantly at P < 0.05 (Wilcoxon rank sum test).

TABLE 3. Bone mineral content (g/cm)
Age (years)

belholt
02.5
2.66.5
6.610.5
10.614.5
14.620.0

25
38
22
9
6

Total

Fig. 2. a: Femur diaphyseal length (cm). b: Humerus diaphyseal length (cm).

factors that may buffer or expose the population to physiological disruption. Goodman (1993, p. 284) argued that
evidence from developing countries refutes the idea of the
healthiest individuals having more stress indicators and

Humerus
Mean and SD
0.26
0.64
0.94
1.37
1.85

6
6
6
6
6

0.28
0.23
0.16
0.36
0.79

100

Nstved
02.5
2.66.5
6.610.5
10.614.5
14.620.0

6
5
12
7
13

Total

43

N
24
37
23
8
4

Femur
Mean and SD
0.40
1.04
1.74
2.65
3.56

6
6
6
6
6

0.17
0.29
0.35
0.67
1.64

0.37
0.91
1.91
1.91
2.68

6
6
6
6
6

0.11
0.47
0.28
0.27
0.84

96
0.25
0.55
1.09
1.12
1.39

6
6
6
6
6

0.05
0.28
0.18
0.20
0.46

6
4
7
7
10
34

states: I know of no situation in which a clearly advantaged group, living or past, has more hypoplasia than a
disadvantaged group. He suggested that by taking the
mean age-at-death, the weakest individuals would be
identied as those who died at an earlier age. Although
stress indicators are generally unrelated to the nal cause
of death, the process causing the lesion may have left the
individual more susceptible. Therefore, individuals with
stress indicators may have been prone to die at an earlier
age, something that Saunders and Hoppa (1993) termed
selective mortality. Individuals with certain stress indicators on their skeletons should be entering the mortality
record earlier than those who do not display the lesions.
At belholt, the mortality curve followed a pattern similar to that of rural Tirup, with a high number of infant
deaths and mortality decreasing as children reach adolescence. Here, preadults represent 24% of the total sample,
whereas at Nstved, preadults represent only 12% of the
total sample, and mortality rates increase with age. In
contrast, the group of preadults represented most at
Nstved is between 14.620.0 years of age (27%). This
age distribution cannot be considered an accurate representation of the living population from which the sample
was derived, because normally only sick individuals were
selected for admission, and due to the incubation period of

CHILD MORBIDITY IN DENMARK

Fig. 3. a: Bone mineral content of midshaft femur (g/cm).

b: Bone mineral content of midshaft humerus (g/cm).

leprosy (ca. 7 years), only older individuals would show

the most obvious signs of the disease. The skewed demographic prole continues into adulthood, with the groups
of youngest and oldest adults from Nstved comprising
46% and 2%, respectively, whereas belholt shows a
much more balanced pattern, with 12% and 13%, respectively. These differences were plausibly attributed to the
overall poor health of adult leprosarium inmates (Mller
Christensen, 1978).
Due to the association between chronic infections and
growth retardation in modern populations (Eveleth and
Tanner, 1990), it was expected that preadults from the
leprosarium at Nstved would be consistently shorter for
age than their peers from belholt. The growth proles
demonstrate this to be the case, with deviations most pronounced in the 10.614.5-year and 14.620.0-year age categories. The BMC results followed a similar pattern,
although to some extent, sexual dimorphism may be
responsible for the differences between samples. In a living population, boys and girls generally follow a similar
course of skeletal growth during infancy and childhood
(Falkner and Tanner, 1986; Eveleth and Tanner, 1990).
After the adolescent growth spurt, which girls generally
experience 2 years ahead of boys, boys surpass girls in
body size (Malina and Bouchard, 1991). Females would
probably have started their growth spurt earlier than
males, around age 1214 years. However, by age 1416
years, the boys, with a greater growth velocity, should

741

have surpassed the mean lengths of bones of females. At

Nstved, a greater number of females might be expected,
as modern studies indicate that women develop leprosy at
an earlier age than men. In India, 50% of women with leprosy developed the disease before age 20 years, compared
to 30% of men (Duncan et al., 1981). This pattern is perhaps linked to endocrine activity and a depressed cellmediated immunity due to pregnancy (Jopling and
McDougall, 1988). Conversely, the fact that the belholt
sample was derived from a monastery may have resulted
in a surplus number of males in the sample of adolescents.
However, in the total skeletal sample, adult males
accounted for only 60% of all adults. The more pronounced
difference in the BMC-for-age curve seen in the oldest age
category of preadults may be the result of sexual dimorphism, with males making up the majority at belholt
and shorter females dominating the sample at Nstved.
However, there were no sex-related differences in tooth
size between the two populations (Lunt, 1969). Rather, the
lower values for bone mineral content in the Nstved preadults resulted from reduced physical activity of the
inmates, damage to the peripheral nerves, and an imbalance in osteoclastic/osteoblastic equilibrium that could
have resulted in paralysis of the limbs and subsequent
osteopenia.
Age-related differences in long bone lengths between the
two samples may also be due to the impact of leprosy on
growth in the Nstved adolescents. That the differences
between belholt and Nstved appear to be more pronounced after 10.5 years can be explained by the fact that by
this age, the prevalence of the disease in children increases,
and that catch up growth can no longer disguise the differences between the two groups (Rallison, 1986).
Periostitis is an almost constant feature of lepromatous
leprosy (LL) caused by neuropathy and secondary infection. Therefore, an association with higher frequencies of
leprosy is not surprising. Maxillary sinusitis is known in
modern leprosy patients to be associated with facial
changes due to the concentration of bacilli in the anterior
parts of the maxilla, palate, and nasal region (Soni, 1989;
Hauhnar et al., 1992; Boocock et al., 1995). Therefore, the
higher prevalence of these infections in the LL group is
considered to be the result of secondary infections associated with leprosy, rather than indicating general exposure
to nonspecic stress factors. The only lesions signicantly
associated with decreased longevity were those on the
endocranial surfaces. One of the major causes for these
lesions is believed to be chronic meningitis caused by various pathogens and resulting from numerous childhood
infections. In this study, 10% (9) of children between ages
010.5 years at belholt and 38% (11) of these children at
Nstved displayed new bone formation on the endocranial
surfaces. The prevalence rate at Nstved is high, even
compared with an Illinois Woodland population, where
Cook and Buikstra (1979) found 27% of endocranial
lesions in children between ages 010 years. It is possible
that the young children from Nstved were susceptible to
meningitis due to frequent childhood illnesses, overcrowded conditions, or perhaps as a result of leprosy.
Although nothing has been published about meningeal
infections and leprosy, another mycobacterium, M. tuberculosis, is known to result in inammation of the
meninges in children (Schultz, 2001).
Mller-Christensen and Sandison (1963) and MllerChristensen (1978) found an association between leprosy
and cribra orbitalia in the adult Nstved sample, where
63% had evidence of lesions, compared to 20% of the

742

24 (13)*
54

(1)
(0)
(2)
(3)
(7)
8
0
15
23
54
8
8
12
8
18

17 (11)
64

(6)
(4)
(1)
(0)
(0)
54
36
9
0
0
13
10
6
8
17

Fig. 4. Prevalence of stress indicators.

%A, percentage affected of total individuals observed; %b, percentage affected in each age group.
Signicantly higher at Nstved compared to belholt (
2, P 0.005).
1

60 (36)
60
Total

35 (14)
39
42 (23)*
55

(2)
(0)
(9)
(5)
(7)
9
0
39
22
30
10
7
12
9
17
19
19
17
17
28
13
10
15
7
15
Nstved
02.5
2.66.5
6.610.5
10.614.5
14.620.0

53 (57)
108
Total

(7)
(7)
(6)
(6)
(10)

54*
70
40
86
67

123

17 (21)

20
0
75
75
56

(2)
(1)
(2)
(2)
(7)
14
7
14
14
50

8 (7)
88

7
5
9
7
11

28
20
22
28
64

112

8 (9)

46
40*
6
0
0

8.5 (7)
82

12
0
17
37
39

0
0
20
27
0
(0)
(0)
(4)
(3)
(0)
0
0
57
43
0
11
36
20
11
4
77
22
0
0
0
25
45
23
12
7
0
8
19
50
0
(0)
(3)
(2)
(2)
(0)
0
3
2
50
0
19
37
21
5
3
0
0
35
44
28
(0)
(0)
(11)
(8)
(2)
19
48
31
18
7
21
58
64
92
60
(6)
(21)
(16)
(11)
(3)
29
37
25
12
5
belholt
02.5
2.66.5
6.610.5
10.614.5
14.620.0

10
37
28
19
5

%b
Age (years)

%A (N)

0
0
52
38
9

%b

%A (N)

(7)
(2)
(0)
(0)
(0)

28
4
0
0
0

%A (N)
%b
%A (N)

N with
sinuses
%A (N)

Periostitis

%b

Endocranial lesions

N with
skulls
N with
tibiae
N with
teeth

Enamel hypoplasias
Cribra orbitalia

N with
orbits

TABLE 4. Percentage of stress indicators, divided into age categories (prevalence)1

Maxillary sinusitis

%b

P. BENNIKE ET AL.

Fig. 5. Developmental age of dental hypoplasia.

adults at belholt. It may be that in order to develop cribra orbitalia, a threshold level of health-associated factors
(such as infectious diseases or nutrirional deciencies)
needs to be reached. Despite the lack of a signicant peak
in cribra orbitalia in the preadult sample at Nstved, the
frequency of this lesion is still higher than at belholt. At
Nstved, 61% of preadults have cribra orbitalia, compared to 54% at belholt. The similar pattern of cribra
orbitalia of the preadult and adult population of Nstved
may reect the same pathogen load and hygienic and
nutritional factors, while the much lower prevalence in
adults at belholt is thought to represent the expected
distribution, with the presence of most cribra orbitalia in
the preadult group compared to adults (Stuart-Macadam,
1991).
The absence of rhinomaxillary changes in the majority
of Nstved preadults does not necessarily conrm that
they were free from leprosy. While they may have suffered
from leprosy at the lower end of the immunity scale, the
youngest individuals may never have developed leprosy.
Although there is evidence of lepromatous leprosy in 3 of
23 individuals from the 6.610.5-year age group, their
numbers are probably too small for any differences to
affect the BMC and long bone length curves (Fig. 3). It is
unlikely that the youngest members of the sample (06.5
year) entered the site by virtue of having leprosy, which

743

CHILD MORBIDITY IN DENMARK

1

TABLE 5. Mean ages at death for individuals with and without stress indicators
With

Without

Stress indicator

Mean

SD

Mean

SD

Difference
(years)

belholt
Cribra orbitalia
Enamel hypoplasia
Periostitis
Endocranial lesions
Sinusitis

7.62
10.8
8.1
1.62
8.0

3.69
2.53
3.03
0.77
3.49

4.79
6.27
5.6
6.8
5.57

4.4
4.06
4.22
4.34
4.24

2.8
4.5**
2.5
5.1*
2.4*

Nstved
Cribra orbitalia
Enamel hypoplasia
Periostitis
Endocranial lesions
Sinusitis
Rhinomaxillary syndrome

9.91
12.7
13.8
3.4
12.45
14.4

6.28
5.03
5.48
3.02
6.52
4.29

9.3
9.16
9.83
10.9
9.27
8.89

6.57
6.21
5.86
5.89
5.16
5.49

0.6
6.5
3.9
7.5*
3.1
5.5

Statics
(KolmogorovSmimov)

K 2.0
K 1.78
K 1.43

K 1.84

K, Kolmogorov-Smimov.
Signicant at P < 0.05.

**

would suggest that they contracted the disease in utero,

or shortly after birth. Such infections are rarely recognized today, and even less so in the medieval period
(Alford et al., 1975). The small number of fetal and infant
remains (n 14) at Nstved suggests that few babies
were born in the hospital sample. Those born to mothers
with lepromatous leprosy today are very weak, grow
slowly, and die early if untreated. Therefore, if babies
were regularly born in the leprosarium, a higher number
of infant deaths would be expected than at belholt, even
though the fertility rate was lower.
It is more probable that children who entered the leprosarium with their parents eventually developed the disease. In South India, a study of children living in a leprosarium showed that, after long exposure to the disease
through parents and patients, children did contract the
disease, but that 75% of them had complete remission
within 6 years (Berreman, 1984). It is possible that some
of these children relapsed with the onset of puberty.
Children may also have contracted the disease in the village and surrounding area and been brought into the leprosarium during later childhood, around age 610 years.
Modern studies show an increase in the prevalence of leprosy in children by the age 7 years, and indeed in the
Nstved sample, the youngest skeleton displaying rhinomaxillary changes is 8 years old. The increase in number of
preadults older than 6.6 years may indicate that after this
age, children were being identied as leprous and entered
the sample from outside. The greater number of children
between ages 617 in the leprosarium would of course have
increased the number of children who died and entered the
mortuary sample. This further increase in number of individuals is consistent with the observation that the disease
is most prominent in the late teens (Duncan, 1985).

CONCLUSIONS
Compared with children interred at the monastery site
of belholt, those from the leprosarium at Nstved can
be considered socially disadvantaged and suffering from
compromised health. They displayed higher frequencies of
stress indicators, which is likely an illustration of the
higher exposure to risk and greater susceptibility of these
individuals to infection and malnutrition.

While the demographic prole at belholt was similar

to an ordinary medieval parish cemetery, Nstved had
rather few young children but high numbers in their late
teens. This pattern is clearly a product of the articial
nature of the Nstved population, reecting the late age
of admission to the leprosarium and an altered reproductive pattern of the inmates. Also, the adult population
from Nstved displays a skewed prole, with high gures
for young adults and a sparse representation of old adults.
Thus the nonattritional demographic prole at Nstved is
characteristic of the entire population. It indicates the
socially imposed selection of the living population and the
inevitably resulting unusual mortality prole in the skeletal assemblage. Leprosy and its concomitant pathological
changes only exacerbated a situation in which social segregation was already a high price to pay.
The distribution of stress indicators of preadults at
belholt and Nstved related to age bears a vague indication that, in general, stress indicators were more prevalent in older age groups. However, this may in part be
inuenced by both the age distribution of the samples and
the skeletal modications themselves. Yet there is a clear
trend toward gradual age-related accumulation of stress
indicators in the leprosarium sample. The only lesions
that were signicantly associated with decreased longevity in both groups were those on the endocranial surfaces.
By age 10 years, long bone growth of the Nstved children fell behind that of belholt, which is most likely
attributable to growth retardation resulting from effects
of leprosy. This is corroborated by a drop in bone mineral
content values during adolescence, demonstrating the
usefulness of this noninvasive method in providing data
on appositional growth and mineralization.
Studies of preadult samples are benecial to skeletal
studies because they are not stricken with the inherent
methodical aws of adult aging, and are thus able to provide a sensitive indicator of health for past societies by
examining those particularly susceptible to disease and
most at risk. However, these studies are limited to those
individuals who failed to adapt to the environmental and
cultural factors inuencing their community. At Nstved,
preadults were selected to enter the leprosarium, and
later the mortuary sample, because they were suffering
from a disease, and therefore their members are not representative of the society from which they were drawn. The

744

P. BENNIKE ET AL.

study shows that it is imperative to be aware of the cultural, historical, and economic context from which a population is derived before the morbidity and mortality patterns of a past society can be reconstructed.
This becomes evident, in particular, when single age
groups rather than bulk categories are examined. While
the combined preadult samples create patterns that favor
traditional views on the meaning of skeletal modications,
only a differential analysis of single age groups reveals the
paradoxical relationship of health, morbidity, and osseous
manifestations that is advocated as typical of skeletal
assemblages. Thus the characteristic distribution of stressrelated skeletal modications across various preadult age
groups lends itself to a biocultural analysis that is able to
identify the more subtle effects of social disadvantage and
compromised health.

ACKNOWLEDGMENTS
The research was funded by grant ERBCHRXCT
930193 to P.B. under EU Program Human Capital and
Mobility and carried out at the Museum of Medical History and Laboratory of Biological Anthropology, University of Copenhagen, Denmark. We are most grateful to
Inger Kjr,Verner Alexandersen and Birgitte Sejersen for
helpful discussions.

LITERATURE CITED
Acheson RM. 1959. Effects of starvation, septicaemia and chronic
illness on the growth cartilage plate and metaphysis of the
immature rat. J Anat 9:123134.
Acheson RM, MacIntyre MN. 1958. The effects of acute infection and
starvation on skeletal development. Br J Exp Pathol 39:3745.
Alford CA, Stagno S, Reynolds W. 1975. Diagnosis of chronic
perinatal infections. Am J Dis Child 129:455463.
Andersen JG. 1969. Studies in the medieval diagnosis of leprosy
in Denmark: an osteological, historical and clinical study.
Copenhagen: Costers Bogtrykkeri.
Andersen JG. 1991. The medieval diagnosis of leprosy. In: Ortner
DJ, Aufderheide AC, editors. Human paleopathology: current
synthesis and future options. Washington, DC: Smithsonian
Institution Press. p 205208.
Arentoft E. 1999. De spedalskes hospital. Fynske studier 18.
Odense: Odense Bys Museer.
Beaton GH. 1989. Small but healthy? Are we asking the right
question? Hum Organisation 48:3039.
Bennike P. 2000. A Medieval mass grave from Denmark (1300
1350 AD). Am J Phys Anthropol [Suppl]30:106 [abstract].
Bennike P. 2002. Vilhelm Mller-Christensen: his work and legacy. In: Roberts CA, Lewis ME, Manchester K, editors. The
past and present of leprosy. BAR international series 1954.
Oxford: Archaeopress. p 135144.
Bennike P, Brade AE. 1999. Middelalderens sygdomme og
behandlingsformer i Danmark [Disease and treatment in medieval Denmark]. Copenhagen: Museum of Medical History.
Bennike P, Bohr H, Toft T. 1993. Determination of mineral content and organic matrix in bone using dual photon absorptiometry. Int J Anthropol 8:111116.
Berreman JM. 1984. Childhood leprosy and social response in
South India. Soc Sci Med 19:853865.
Boldsen J. 2001. Epidemiological approach to the paleopathological diagnosis of leprosy. Am J Phys Anthropol 115:380387.
Boocock PA, Roberts CA, Manchester K. 1995. Prevalence of
maxillary sinusitis in leprous individuals from a medieval leprosy hospital. Int J Lepr 63:265268.
Brody SN. 1974. The disease of the soul: leprosy in medieval literature. Ithaca: Cornell University Press.
Buikstra JE, Ubelaker DH. 1994. Standards for data collection
from human skeletal remains.Arkansas Archaeological Survey
research series44. Fayetteville: Arkansas Archaeological Survey.

Carlson DS, Armelagos GJ, Van Gerven DP. 1976. Patterns of

age-related cortical bone loss (osteoporosis) within the femoral
diaphysis. Hum Biol 48:295314.
Cohen MN, Armelagos GJ. 1984. Paleopathology at the origins
of agriculture. New York: Academic Press, Inc.
Cole TJ. 1989. Relating growth rate to environmental factors
methodological problems in the study of growth-infection interaction. Acta Paediatr Scand Suppl 350:1420.
Cook DC. 1984. Subsistence and health in the Lower Illinois
Valley: osteological evidence. In: Cohen MN, Armelagos GJ,
editors. Paleopathology at the origins of agriculture. Orlando,
FL: Academic Press, Inc. p 235269
Cook DC, Buikstra JE. 1979. Health and differential survival in
prehistoric populations: prenatal dental defects. Am J Phys
Anthropol 51:649664.
Duncan ME. 1980. Babies of mothers with leprosy have small
placentae, low birth weights and grow slowly. Br J Obstet
Gynaecol 87:471479.
Duncan ME. 1985. Leprosy in young children-past, present and
future. Int J Lepr 53:468473.
Duncan ME, Melsom R, Pearson JMH, Ridley DS. 1981. The
association of pregnancy and leprosy. Lepr Rev 52:245262.
Duray SM. 1996. Dental indicators of stress and reduced age of
prehistoric Native Americans. Am J Phys Anthropol 99:275286.
El-Najjar M, Ryan DJ, Turner C, Lozoff B. 1979. The aetiology
of porotic hyperostosis among the historic and prehistoric
Anasazi Indians of southwestern United States. Am J Phys
Anthropol 44:477488.
Eveleth PB. 1986. Population differences in growth: environmental and genetic factors. In: Falkner F, Tanner JM, editors.
Human growth: a comprehensive treatise. New York: Plenum
Press. p 221239.
Eveleth PB, Tanner JM. 1990. Worldwide variation in human
growth. Cambridge: Cambridge University Press.
Falkner F, Tanner JM. 1986. Human growth: a comprehensive
treatise. New York: Plenum Press.
Gilbert RI. 1985. Stress, paleonutrition, and trace elements. In:
Gilbert RI, Mielke JM, editors. The analysis of prehistoric
diets. Orlando, FL: Academic Press, Inc. p 339356.
Goodman AH. 1993. On the interpretation of health from skeletal remains. Curr Anthropol 34:281288.
Goodman AH, Rose JC. 1990. Assessment of systemic physiological perturbations from dental enamel hypoplasia and associated histological structures. Yrbk Phys Anthropol 33:59110.
Goodman AH, Rose JC. 1991. Dental enamel hypoplasia as indicators of nutritional status. In: Kelley MA, Larsen CS, editors. Advances in dental anthropology. New York: Wiley and
Liss Inc. p 279293.
Goodman AH, Martin DL, Armelagos GJ, Clark G. 1984. Indicators of stress from bone and teeth. In: Cohen MN, Armelagos
GJ, editors. Paleopathology at the origins of agriculture.
Orlando, FL: Academic Press, Inc. p 1349.
Goodman AH, Thomas RB, Swedlund AC, Armelagos GJ. 1988. Biocultural perspectives on stress in prehistoric, historical, and contemporary population research. Yrbk Phys Anthropol 31:169202.
Grauer AL. 1993. Patterns of anaemia and infection from Medieval York, England. Am J Phys Anthropol 91:203215.
Hauhnar CZ, Mann SBS, Sharma VK, Mehta S, Radota BD.
1992. Maxillary antrum involvement in multibacillary leprosy: a radiological, sinuscopic and histological assessment.
Int J Lepr 60:390395.
Hillson S. 2000. Dental anthropology. Cambridge: Cambridge
University Press.
Huss-Ashmore R. 1978. Nutritional determination in skeletal
populations. Am J Phys Anthropol 48:407.
Jantz RL, Owsley WD. 1984. Long bone growth variation among
Arikara skeletal populations. Am J Phys Anthropol 63:1320.
Johnston FE. 1962. Growth of the long bones of infant and young
children at Indian Knoll. Am J Phys Anthropol 20: 249254.
Jopling WH, McDougall AC. 1988. Handbook of leprosy. Oxford:
William Heinemann Medical Books.
Kieffer-Olsen J. 1993.Grav og gravskik i det middelalderlige
Danmark. 8 kirkegardsudgravninger. Unpublished Ph.D. thesis. Aarhus: Aarhus University.

CHILD MORBIDITY IN DENMARK

Kirkwood BR. 1988. Essentials of medical statistics. London:
Blackwell Scientic Publications.
Lampl M, Johnston FE. 1996. Problems in the aging of skeletal
juveniles: perspectives from maturation assessments of living
children. Am J Phys Anthropol 101:345355.
Lara CB, Nolesco JO. 1956. Self-healing or abortive, and residual forms of childhood leprosy and their probable signicance.
Int J Lepr 24:245263.
Larsen CS. 1997. Bioarchaeology. Interpreting behavior from
the human skeleton. Cambridge: Cambridge University Press.
Lewis AB, Garn SM. 1960. The relationship between tooth formation and other maturational factors. Angle Orthod 30:7077.
Lewis M. 2002a. Urbanisation and child health in medieval
and post-medieval England. BAR British series 339. Oxford:
Archaeopress.
Lewis M. 2002b. The impact of industrialization: comparative
study of child health in four sites from medieval and postmedieval England (8501859 AD). Am J Phys Anthropol
119:211223.
Lewis M. 2002c. Infant and childhood leprosy: clinical and
palaeopathological implications. In: Roberts C, Lewis M,
Manchester K, editors. The past and present of leprosy.BAR
international series1954. Oxford: Archaeopress. p 163170.
Lewis M. 2004. Endocranial lesions in non-adult skeletons:
understanding the aetiology. Int J Osteoarchaeol 14:8297.
Lewis ME, Roberts CA, Manchester K. 1995. Comparative study
of the prevalence of maxillary sinusitis in later medieval
urban and rural populations in Northern England. Am J Phys
Anthropol 98:497506.
Little MA. 1995. Adaptation, adaptability and multidisciplinary
research. In: Boaz NT, LD Wolfe, editors. Biological anthropology: the state of the science. Corvallis: Oregon State University Press. p 121147.
Lunt DA. 1969. An odontometric study of Medieval Danes. Acta
Odontol Scand [Suppl]55:27.
Macey JF. 1996. Demography and disease of the Nstved Helligandshus collection. MA thesis. Memorial University of
Newfoundland, St. Johns.
Malina RM, Bouchard C. 1991. Growth, maturation and physical activity. Champaign: Human Kinetics Publishers Inc.
Maresh MM. 1940. Paranasal sinuses from birth to adolescence.
Am J Dis Child 60:5578.
Martin D, Armelagos GJ. 1979. Morphometrics of compact bone:
an example from Sudanese Nubia. Am J Phys Anthropol
51:571578.
Melsom R, Harboe M, Duncan ME. 1982. IgA, IgM and IgA
anti-M-leprae antibodies in babies of leprosy mothers during
the rst 2 years of life. Clini Exp Immunol 49:532542.
Mensforth RP. 1985. Relative tibia long bone growth in the Libben and Bt-5 prehistoric skeletal populations. Am J Phys
Anthropol 68:247262
Mensforth RP, Lovejoy CO, Lallo JW, Armelagos GJ. 1978. The
role of constitutional factors, diet and infectious disease in
the aetiology of porotic hyperostosis and periosteal reactions
in prehistoric infants and children. Med Anthropol 2:159.
Merrett DC, Pfeiffer S. 2000. Maxillary sinusitis as an indicator
of respiratory health in past populations. Am J Phys Anthropol 111:301318.
Mittler DM, Van Gerven DP, Sheridan SG, Beck R. 1992. The
epidemiology of enamel hypoplasia, cribra orbitalia, and subadult mortality in an ancient Nubian population. J Palaeopathol 2:143150.
Mller-Christensen V. 1961. Bone changes in leprosy. Copenhagen: Munksgaard.
Mller-Christensen V. 1978. Leprosy changes of the skull.
Odense: Odense University Press.
Mller-Christensen V. 1982. belholt kloster. 2nd ed. Copenhagen: Nationalmuseet.
Mller-Christensen V, Sandison AT. 1963. Usura orbitae (cribra
orbitalia) in the collection of crania in the Anatomy Department
of the University of Glasgow. Pathol Microbiol 26:175183.
Moorrees CFA, Fanning EA, Hunt EE. 1963a. Formation and
resorption of three deciduous teeth in children. Am J Phys
Anthropol 21:205213.

745

Moorrees CFA, Fanning EA, Hunt EE. 1963b. Age variation of

formation stages for ten permanent teeth. J Dent Res 42:
14911502.
Olivier G, Pineau H. 1960. Nouvelle determination de la taille
ftale dapre`s les longueurs diaphysaires des os longs. Ann
Med Legale 40:141144.
Ortner DJ. 1991. Theoretical and methodological issues in paleopathology. In: Ortner DJ, Aufderheide AC, editors. Human
paleopathology: current synthesis and future options. Washington,
DC: Smithsonian Institute Press. p 511.
Ortner DJ. 2003. Identication of pathological conditions in
human skeletal remains. London: Academic Press.
Ortner DJ, Ericksen MF. 1997 Bone changes in the human skull
probably resulting from scurvy in infancy and childhood. Int
J Osteoarchaeol 7:212220.
Ortner DJ, Butler W, Cafarella J, Milligan L. 2001 Evidence of
probable scurvy in subadults from archaeological sites in
North America. Am J Phys Anthropol 114;343351.
Rallison ML. 1986. Growth disorders in infants, children and
adolescents. New York: John Wiley and Sons, Inc.
Reinhard KJ. 1992. Patterns of diet, parasitism and anaemia in
prehistoric west North America. In: Stuart-Macadam P, Kent S,
editors. Diet, demography and disease: changing perspectives of
anaemia. New York: Aldine de Gruyter. p 219260.
Richards P. 1977. The medieval leper and his northern heirs.
Cambridge: D.S. Brewer.
Roth EA. 1992. Applications of demographic models to paleodemograpy. In: Saunders SR, Katzenberg MA, editors. Skeletal
biology of past peoples: research methods. New York: WileyLiss, Inc. p 175188.
Ruff CB, Haynes C. 1984. Bone mineral content in the lower
limb. J Bone Joint Surg [Am] 66:10241031.
Saunders SR, Hoppa RD. 1993. Growth decit in survivors and
non-survivors: biological mortality bias in subadult skeletal
samples. Yrbk Phys Anthropol 36:127151.
Scheuer L, Black S. 2000. Development: juvenile osteology. London: Academic Press.
Schour I, Massler M. 1940. Studies in tooth development: the
growth pattern of human teeth. J Am Dent Assoc 27:1778
1793, 19181931.
Schultz M. 2001. Palaeohistopathology of bone: a new approach to
the study of ancient diseases. Yrbk Phys Anthropol 44:106147.
Schutkowski H, Grupe G. 1997. Correlations between cribra
orbitalia, archaeometric analyses and habitat conditions.
Anthropol Anz 55:155166.
Scrimshaw NS, Young WR. 1989. Adaptation to low protein and
energy intakes. Hum Organisation 48:2030.
Scrimshaw NS, Taylor CE, Gordon JE. 1959. Interactions of
nutrition and infection. Am J Med Sci 237:367403.
Seckler D. 1982. Small but healthy? Some basic problems in the
concept of malnutrition. In: Sukhatme PV, editor. Newer
concepts in nutrition and their implications for policy. Pune,
India: Maharashtra Association for the Cultivation of Science.
p 139148.
Smith HB. 1991. Standards of human tooth formation and dental age assessment. Adv Dent Anthropol 8:143168.
Soni NK. 1989. Antroscopic study of the maxillary antrum in
lepromatous leprosy. J Laryngol Otol 103:502503.
Stini WA. 1985. Growth rates and sexual dimorphism in evolutionary perspective. In: Gilbert RI, Mielke JH, editors. The
analysis of prehistoric diets. Orlando: Academic Press. p 191
222.
Stuart-Macadam P. 1991. Anaemia in Roman Britain.BAR international series199. Oxford Archaeopress. p 101113.
Stuart-Macadam P, Kent S. 1992. Diet, demography and disease: changing perspectives of anaemia. New York: Aldine de
Gruyter.
Subira M, Alesan A, Malgosa A. 1992. Cribra orbitalia and
nutritional deciency. Studies of trace elements. Munibe
[Suppl]8:153158.
Tanner JM. 1990. Foetus into man: physical growth from conception to maturity. Cambridge, MA: Harvard University Press.
Thompson DD. 1980. Age changes in bone mineralization, cortical
thickness, and Haversian canal area. Calcif Tissue Int 31:511.

746

P. BENNIKE ET AL.

Van Gerven DP, Armelagos GJ. 1983. Farewell to paleodemography? Rumours of its death have been greatly exaggerated.
J Hum Evol 12:353360.
Van Gerven DP, Sandford MK, Hummert JR. 1981. Mortality
and culture change in Nubias Batn el Hajar. J Hum Evol 10:
395408.
Wald ER, Milmoe GJ, Bowen AD, Ledesma-Medina J, Salamon
N, Bluestone CD. 1981. Acute maxillary sinusitis in children.
N Engl J Med 304:749755.
Wall CE. 1991. Evidence of weaning stress and catch-up growth
in the long bones of a central California Amerindian sample.
Ann Hum Biol 18:922.
Wapler U, Crubezy E, Schultz M. 2004. Is cribra orbitalia synonymous with anemia? Analysis and interpretation of cranial
pathology in Sudan. Am J Phys Anthropol 123:333339.

Weinberg ED. 1992. Iron withholding in prevention of disease.

In: Stuart-Macadam P, Kent S, editors. Diet, demography and
disease: changing perspectives of anaemia. New York: Aldine
de Gruyter. p 105150.
WHO Working Group. 1986. Use and interpretation of anthropometric indicators of nutritional status. Bulletin WHO 64:
929941.
Wood JW, Milner GR, Harpending HC, Weiss KM. 1992.
The osteological paradox. Problems of inferring prehistoric health from skeletal samples. Curr Anthropol 33:
343370.
Wright D. 1979. Acute sinusitis. In: Ballantyne J, Groves J,
editors. Scott-Brownes diseases of the ear, nose and throat.
London: Butterworths. p 243271.