1996 Skeletal growth of children from the iron age.pdf

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 100:389-396 (1996)

Skeletal Growth of Children From the Iron Age Site at K2

(South Africa)

MARYNA STEYN AND MACIEJ HENNEBERG

Department of Anatomy, University of Pretoria, Pretoria, 0001, (M.S.)

and Biological Anthropology Research Programme, University of the

Witwatersrand, Medical School, Parktown, 2193 (M.H.) South Africa

KEY WORDS

bone

Community health, Allometry, Mapungubwe, Long

ABSTRACT Cross-sectional growth data were obtained from the skeletal

remains of children from the Iron Age site of K2 near the Limpopo River.

Standard measurements of the diaphyseal lengths of the long bones from

both limbs were recorded and compared to published skeletal data. For this

purpose, data on Eskimo and Aleut skeletons, Libben skeletons, and skeletons

from Indian Knoll and Altenerding were used. An attempt to study growth

allometrically was made. K2 children were growing as well as children from

these other groups. Comparison of data for K2 children with those on living

South African “Cape Coloured” rural children, studied during the late 198Os,

shows the similarity of growth of both groups. o 1996 Wiley-Liss, Inc.

It is only recently that juvenile skeletons

have been included in population studies of

prehistoric people, placing population differ-

ences and similarities in growth and devel-

opment under study. These kinds of investi-

gations are mainly used to examine the

adequacy of growth, in order to gain insight

into overall community health and adapta-

tion to the environment (Johnston, 1968;

Johnston and Zimmer, 1989). Realizing the

value of such studies in reconstructing

health changesand nutritional statusof past

populations, a number of studies have been

published (e.g. Johnston, 1962; y’Edynak,

1976; Merchant and Ubelaker, 1977; Sun-

dick, 1978; Jantz and Owsley, 1984; Lovejoy

et al., 1990; Hoppa, 1992; Saunders et al.,

1993; Miles and Bulman, 1994).

The assessment of growth in prehistoric

populations, however, is not without prob-

lems. Skeletal samples are generally numer-

ically restricted due to poor preservation of

young children’s skeletons and low mortality

in the adolescent age groups. Difficulties

may also arise from lack of proper standards

for aging. Sexing of juvenile skeletons has

always been difficult. The possible presence

of secular trends makes comparisons be-

tween various groups difficult (Johnston,

1968; Sundick, 1978).

One of the most serious problems arises

from the fact that the juvenile skeletal sam-

ple represents only those individuals who

died in childhood, and not necessarily the

normal, healthy children in the community.

The mere presence of an individual in the

skeletal sample indicates that he/she had

a disease or died of other causes, and the

possibility exists that this disease might

have retarded the growth of the individual.

According to Buikstra and Cook (1980),

these juveniles represent the “minima”

rather than the “modes” of those who sur-

vived. Wood etal. (1992) also argued that the

bony lesions observed and long-bone growth

might not reflect the true health status of a

Received November 7, 1994; accepted December 17, 1995.

Address reprint requests to M. Steyn, Department ofAnatomy,

P.O. Box 2034, University of Pretoria, 0001, South Africa.

Maciej Henneberg’scurrent address is Department ofhatomy

and Histology, University of Adelaide Medical School, Adelaide

5005, South Australia.

0 1996 WILEY-LISS, INC.

390

M. STEYN AND M. HENNEBERG

group of people, They stated that “If mortal-

ity slackens, only the most frail (i.e., short-

statured) individuals die. Periods of low

mortality are therefore characterized by

comparatively low mean stature among the

dead” (p. 351).

Lovejoy et al. (1990),however, believe that

children aremore likely to die of acute rather

than chronic diseases, which would not have

altered their maturation in any way. Sun-

dick (1978) also found that children with

more frequent illnesses compare well, as far

asheight is concerned, with those who were

healthier (also Henneberg et al., 1984).In a

review paper on the issue of mortality bias

and its influence on growth data, Saunders

and Hoppa (1992) found that although the

potential for such a bias exists, its effects

are likely to be small.

No major differences have ever been de-

tected between the pattern and direction of

growth when modern and prehistoric groups

are concerned, although many prehistoric

peoples appear to be shorterthan their mod-

ern counterparts (Frayer, 1984; Saunders,

1992).This, however, is not anabsolute rule,

as in many populations body size has been

found to decline since the Upper Paleolithic

to Medieval times (Frayer, 1984; Jacobs

1985; Brown, 1992)or to remain stable over

thousands of years (Henneberg and Henne-

berg, 1990, Henneberg et al., 1992). While

the documentation of age and sex of juvenile

skeletons is not reliable enough to allow de-

tailed analysis of the growth process itself

(Johnston and Zimmer, 1989), studies of

growth of children, based on skeletal sam-

ples, can be used to provide general informa-

tion on the growth pattern in a given popula-

tion, and thus allow some conclusions to be

drawn on the relationshipbetween a popula-

tion and its environment. Average differ-

ences in growth between groups are most

probably due to differences in the environ-

ment, although there are indications that

population-characteristic body proportions

are established early in childhood (y’Edy-

nak, 1976).

Ina recent paper, Sciulli (1994)pointed out

that all bones are not equally affected by nu-

tritionalanddisease stress,andthatthemost

rapidly growing bones are more influenced

than others. The bones of the lower limb (in

the sequence of femur; fibula; tibia) will

therefore be relatively smallerinunfavorable

conditions than those of the upper limb. This

canbe tested by means of allometric analysis,

e.g., studyingthe growth of thefemuragainst

that of the humerus.

The aim of this study is to compare the

infracranial growth of children from the Iron

Age site of K2 to that of others, in order to

gain insight into their overall health status.

The Iron Age site of Mapungubwe is situ-

ated in the Northern Transvaal, close to the

border between SouthAfrica, Botswana and

Zimbabwe. It consists of two valleys: K2 and

the Southern Terrace, with the Mapun-

gubwe Hill towering over the latter.The val-

ley called K2 was inhabited from about AD

1000 to 1200. Later, most of the occupation

shifted to an adjacent valley, the Southern

Terrace. During this time, the Mapungubwe

Hill itself was occupied (Eloff, unpublished)

by people of exalted social status. The com-

plex was abandoned around 1300.

The valley of K2 was densely inhabited,

and it has a distinct midden in the middle

of the settlement which is up to 6 m deep.

The Mapungubwe complex of sites was the

center of economic and political power dur-

ing the period of its occupation (Hall, 1987).

Ivory and bone tools were traded with East

Coast merchants. Large herds of cattle were

kept, and it can be expected that the meat

of domesticated animals and milk formed a

substantial part of the diet (Eloff, unpub-

lished; Voigt, 1983).This was supplemented

by sorghum, millet and beans.

The decline of the settlements coincides

with the rise of the Great Zimbabwean Em-

pire. A more complete description of the site

can be found in Henneberg and Steyn(1994).

Excavations were conducted at the sites

by the University of Pretoria from the 1930s.

They yielded the 106 skeletons used in this

study. This is the largest collection of skele-

tons from a single Iron Age site in Southern

Africa. Ninety-four of these skeletons came

from K2 and twelve from the Mapungubwe

Hill. No skeletons from the SouthernTerrace

have yet been discovered. Previous studies

of thismaterial have mostly centered around

the establishment of racial affinity (Gallo-

way, 1959; Rightmire, 1970; De Villiers, un-

published).

SKELETAL GROWTH AT K2

39 1

This assessment of the growth of K2 skele-

tons forms part of a larger study, in which

various aspects pertaining to demographic

dynamics, health, nutrition and adaptation

are addressed (Henneberg and Steyn, 1994;

Steyn, unpublished; Steyn and Henneberg,

1995, 1996). Paleodemographic analysis in-

dicated a high rate of natural increase, mor-

tality typical for preindustrial societies,

newborn life-expectancy around 20 years,

and less than 50% survivorship to sexual

maturity (Henneberg and Steyn, 1994). Al-

though the people were not free from disease,

it seems they were relatively healthy (Steyn

and Henneberg, 1996).

17 fibulae available for observation. All dia-

physeal lengths of complete long bones were

measured using an osteometric board. Re-

sults were then plotted on a graph by dental

age, and compared with available data from

other skeletal samples. For these compari-

sons, data on Libben (Lovejoy et al. 19901,

the Eskimo and Aleut (y’Edynak, 1976), as

well as those of children from Indian Knoll

and Altenerding, Germany (Sundick, 1978)

were used. These particular samples were

chosen because they represent a wide vari-

ety of populations from different geographi-

cal areas. Also, aging techniques for these

studies differed; for example, the Eskimo

and Aleut samples (y’Edynak, 1976) were

aged by a combination of dental and skeletal

indicators, while Libben (Lovejoy et al.

19901, Alterneding, and Indian Knoll (Sun-

dick, 1978)were mostly aged by teeth alone.

Allometric growth was observed by plot-

ting diaphyseal length of one long bone

against that of another one, irrespective of

the age of the individual. In this way, the

effect of possible erroneous ageing is elimi-

nated, while differential growth of bones

may be indicative of environmental influ-

ences (Sciulli, 1994). In general terms this

method completely eliminates the chrono-

logical age from the growth study and con-

centrates only on the proportional growth of

one body part against the other. When the

same individual had, for example, lengths

available for both a humerus and a femur,

the length of the femur was plotted on a

graph against the length of the humerus.

These allometric plots were compared to

the same groups as before. The datafor these

groups are published as averages from each

year of chronological age. Use of such aver-

ages for comparison with allometric plots of

individuals may be questioned. Fortunately

small samples in published data resulted in

some age categories being represented by

only one individual. These individual data

were plotted against K2 (Fig. 5).

Although it is not possible to compare skel-

etal samples to living groups in a straightfor-

ward way (Johnston and Zimmer, 19891,

long-bone lengths of K2 children were also

plotted on the same graph as limb segment

lengths of rural Cape “Coloured” children

from South Africa (M. Henneberg, unpub-

MATERIALS AND METHODS

The state of preservation of the remains

of the 106 individuals ranged from only a

few teeth to almost complete skeletons. The

Mapungubwe skeletons particularly are in

a very bad condition, while the K2 ones,

found in the ash heaps, are in a much better

condition. Therefore, the Mapungubwe skel-

etons, with only five subadult individuals,

were not included in the study of growth. It

seems that the custom of burying the dead

in ash heaps, which create an alkaline envi-

ronment, contributed to the good preserva-

tion of a large number of infant and child

skeletons.

In a study of growth, reliable age-at-death

estimates are important. Skeletal age can-

not be used as an indicator of chronological

age here, as it would lead to circular reason-

ing (Johnston and Zimmer, 1989; Saunders,

1992).Thus, the dental age was used. Dental

age also varies in different environments

and populations, but to a lesser degree than

does skeletal age (Sundick, 1978). Only

bones of individuals with teeth preserved

well enough to allow determination of dental

age were used in this study. Each individual

was thus aged according to the eruption and

formation of teeth (Ubelaker, 1989; El-

Nofely and Iscan, 1989).In the case of new-

born individuals, only those with tooth

germs available were used.

Incompletely preserved skeletons of 45

subadult individuals of dental age from 0 to

18 years were used. There were 30 humeri,

26 radii, 24 tibiae, 22 ulnae, 19 femora, and

392

M. STEYN AND M. HENNEBERG

TABLE 1. Diaphyseal lengths in millimetres

Age group

(vears)

Humerus

Radius

Ulna

Femur

Tibia

Fibula

n Mean

s n Mean

s n Mean s n Mean

n Mean s

Newbordfetal

Newborn-0.5

0.5-1.5

1.5-2.5

2.5-3.5

3.5-4.5

4.5-5.5

5.5-6.5

6.5-7.5

7.5-8.5

8.5-9.5

9.5-10.5

10.5-11.5

67.4 7.37 3

85.2 0.85 1

97.9 6.90 5

107.0 2

118.5 7.50 2

144.8 10.59 3

150.0 5.00 2

158.0 0

56.3 7.04 3 62.9 6.75 3 75.9 7.37 4 66.4 6.36 2 59.9 2.90

66.3 1 76.3 1 100.5 1 89.0

82.6 6.81 3 93.3 5.34 5 125.0 6.90 4 108.3 7.41 4 97.7 7.85

85.0 1.00 2 95.8 1.25 1 130.0 1 107.5 1 114.0

113.8 14.75 2 124.4 12.60 2 147.5 10.00 1 131.5

112.6 2.41 2 120.0 7.50 3 197.4 7.85 4 158.7 9.94 2 161.2 4.75

120.2 4.75 1 130.0

1 193.0

130.5

1 142.0

2 229.0 6.00 2 188.8 4.25 2 188.2 3.65

191.5

157.0

1 172.0

1 219.3

182.0

141.8

1 157.0

1 257.0

1 210.0

1 204.0

198.0

161.0

2 173.7 3.65

1 230.0

2 227.5 1.50

210.5

181.5

1 198.8

1 245.0

11.5-12.5

12.5-13.5

1 249.3

1 202.0

1 207.5

1 381.3

1 325.5

1 284.0

13.5-14.5

1 260.5

1 175.0

1 352.0

1 294.0

14.5-15.5

15.5-16.5

16.5-17.5

2 286.8 6.75 1 215.5

1 251.0

1 317.5

lished data). This cannot be used to make

comparisons on absolute lengths, but may

provide information on the similarity, or lack

thereof, of growth patterns between these

prehistoric and modern groups.

DISCUSSION

Studies of skeletal growth are often ham-

pered by small sample sizes. This is also

true of the K2 sample. In the case of the

K2 material, however, such an investigation

was deemed to be important because it is

the largest collection of its kind from this

part of the world, and is also to the authors’

knowledge the first such study from south-

ern Africa. Even though the sample sizes are

small, it does seem as if the children from

K2 were a bit taller than those, for example,

from the Eskimo and Indian sample groups.

Of the four groups compared, the Libben

group seems to be the closest in length to the

children from K2, followed by Altenerding.

If it is true that long-bone length reflects

something of the living conditions, then it

seems as if the children from K2 were as

well off, if not better, as those of the other

samples they were compared with. The sam-

ple sizes presented in this study, however,

are not large enough to fully justify this con-

clusion beyond suggestion.

This conclusion is corroborated by the re-

sults of the paleopathological study, which

indicated that the people were relatively

healthy (Steyn and Henneberg, 1996). Al-

though the people were not free of disease,

incidences of cribra orbitalia, enamel hypo-

plasia, and transverse radio-opaque lines

RESULTS

Data for the lengths of the various long

bones at different ages is shown in Table 1.

On all the plots of long bone lengths against

skeletal age, K2 children’s bones were as

long, or even slightly longer, than those of

the other groups. It seems that data for Lib-

ben (Lovejoy et al., 1990) are the closest to

our observations. Figures 1and 2 illustrate

the growth of the femur and humerus, and

Figures 3 and 4 those of the distal limb seg-

ments, namely ulna and tibia.

As far as the comparisons of allometric

growth are concerned (see Fig. 5 asanexam-

ple of the length of the femur vs. that of the

humerus), it seems that children from K2

grew in a fashion similar to those from other

populations, although the number of individ-

uals that could be used for comparison is

small. The same can be said for the compari-

son with the growth of K2 and Cape “Col-

oured” children, once allowance has been

made for soft tissues and epiphyses (Figs. 6

and 7 present examples of lengths of the

forearm and upper leg).

SKELETAL GROWTH AT K2

393

500

Femur: length In mm.

400 -

300 -

Indlan Knoll

Eaklmo and AIbut

Altenbrdlng

Llbben

200 -

100 -

0 2 4 8 8 10 12 14 18 18

ago (yr3

Fig. 1. Diaphyseal growth of the femur.

360

Humerue: length

In mm.

300

- I a K2

Indlan Knoll

Eeklmo and Abut

Alhnerdlng

Llbbbn

......

0 2 4 6 8 10 12 14 18 18

ago (yrd

Fig. 2. Diaphyseal growth of the humerus.

were not very high. Few signs of chronic dis-

ease, in the form of subperiosteal bone

growth, were observed. This was especially

true in the case of the children, where only

two, showing subperiosteal bone growth, out

of a total of 44 children with long bones, had

any signs of disease at all. Chronic diseases

might have been more common in adults,

where four out of the eight adults with long

bones were affected. It has been suggested

that these bony lesions might have been due

to the presence of treponemal disease (Steyn

and Henneberg, 1995).

It is generally accepted that the health of

people declined with the transition from a

hunting-and-gathering to agriculturalist

mode of subsistence (e.g., Cohen and Arme-

lagos, 1984). In the case of the people from

the Mapungubwe complex of sites, however,

it may be proposed that the more sedentary

lifestyle did not lead to deterioration in gen-

eral health, due to the presence of large

herds of domesticated animals. This ensured

the presence of a constant protein-rich food

supply (Steyn and Henneberg, 1996).

Although the comparison of growth with

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: