Citation
Linear and cortical bone dimensions as indicators of health status in subadults from the Milwaukee County Poor Farm Cemetery

Material Information

Title:
Linear and cortical bone dimensions as indicators of health status in subadults from the Milwaukee County Poor Farm Cemetery
Creator:
Florence, Jessica Lynn
Publication Date:
Language:
English
Physical Description:
xviii, 167 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Anthropology, CU Denver
Degree Disciplines:
Anthropology

Subjects

Subjects / Keywords:
Health status indicators -- Wisconsin -- Milwaukee County ( lcsh )
Anthropometry -- Wisconsin -- Milwaukee County ( lcsh )
Bones -- Analysis -- Wisconsin -- Milwaukee County ( lcsh )
Anthropometry ( fast )
Bones -- Analysis ( fast )
Health status indicators ( fast )
Wisconsin -- Milwaukee County ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 160-167).
General Note:
Department of Anthropology
Statement of Responsibility:
Jessica Lynn Florence.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
166344909 ( OCLC )
ocn166344909
Classification:
LD1193.L43 2007m F56 ( lcc )

Full Text
/ '
LINEAR AND CORTICAL BONE DIMENSIONS AS INDICATORS OF
HEALTH STATUS IN SUBADULTS FROM THE MILWAUKEE
COUNTY POOR FARM CEMETERY
by
Jessica Lynn Florence
B.A., University of Wisconsin Milwaukee, 2002
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center
in partial fulfillment
of the requirements for the degree of
Master of Arts
Anthropology
2007


2007 by Jessica Lynn Florence
All rights reserved.


This thesis for the Master of Arts
degree by
Jessica Lynn Florence
has been approved
by
~7


Florence, Jessica, L. (Master of Arts, Anthropology)
Linear and Cortical Bone Dimensions as Indicators of Health Status in Subadults
from the Milwaukee County Poor Farm Cemetery
Thesis directed by Assistant Professor Charles Musiba
ABSTRACT
This study examines the health status of children (fetal to 12 years of age) buried
in the Milwaukee County Poor Farm Cemetery located in Wauwatosa, Wisconsin.
The Poor Farm Cemetery served as a place of interment during the late 19th and early
20th centuries not only for individuals who resided at the Milwaukee County
Institutional Grounds, but also for the community poor, the unidentified, and those
who died in the County without kin.
The main hypothesis tested in this study is that if indeed the subadult population
buried at the Poor Farm Cemetery suffered from poor health and stress then patterns
of linear and cortical bone growth should reveal this. Skeletal growth profiles,
plotting measurements of length against dental age, were created for the humerus,
radius, ulna, femur, and tibia. Comparison of the profiles with modem and
archaeological samples provided evidence that some age groups suffered from poor
health based on having reduced diaphyseal lengths. Evidence of a reduced level of
general health was also demonstrated by comparing growth rates between the modem
and Poor Farm Cemetery samples. However, skeletal growth profiles created for the
humerus, femur, and tibia involving measurements of cortical thickness plotted
against age revealed normal patterns of cortical bone growth.
The following conclusions were drawn from this study: (1) only the linear growth
data suggest that the subadult population buried in the Poor Farm Cemetery suffered
from poor health and stress; (2) linear growth is a more sensitive indicator of stress
than cortical growth in subadults from the Poor Farm Cemetery; (3) non-specific
markers of stress are not always associated with reductions in linear and cortical bone
growth. In fact, subadults lacking pathologies can demonstrate greater reductions in
long bone length and cortical mass than those with pathologies. It appears that this is
the first study to not only use a digitizing program to obtain measures of cortical
thickness from subadult X-rays, but also provide long bone cortical data on a sample


of children living in the U.S. around the turn of the 20th century that may be used by
other investigators in future studies.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed


DEDICATION PAGE
To my loving husband, Ryan, for his unwavering support throughout the
completion of this thesis as well as over the years. Thank you for constantly believing
in me and encouraging me to do my best. The journey would not have been possible
without you by my side.


ACKNOWLEDGMENT
I would like to, first and foremost, thank Dr. Norman Sullivan in the Department
of Social and Cultural Sciences at Marquette University for allowing me access to the
subadult bones used in this study. Many hours were spent measuring these bones in
his office and I am indebted to his generosity for arranging space for me to work and
letting me retrieve information stored in the MCIG database.
Thanks are also extended to the faculty at Marquette University School of
Dentistry, in particular to the Associate Dean, Dr. Anthony M. Iacopino, for giving
permission to use the schools radiology equipment, and Rodney D. Daering for his
time and assistance in operating the equipment. I appreciate Dr. Sullivan contacting
these individuals on my behalf.
I am grateful to Lauren Devitt for lending a hand with the sorting and carrying of
bones across campus, and Sean Dougherty, who found time while working on his
doctoral dissertation to share with me details about Milwaukee Countys institutions
that he collected from the archives. He also provided information that was missing for
some of the subadult remains.
Finally, I would like to thank the members of my thesis committee, Dr. David
Tracer, Dr. Chris Beekman, and my thesis advisor, Dr. Charles Musiba, for their time
and guidance. I am thankful for the support and opportunities that they, as well as
other faculty and staff members in the Anthropology Department, have given me
throughout my graduate career.


TABLE OF CONTENTS
Figures.......................................................xi
Tables........................................................xvii
CHAPTER
1. INTRODUCTION................................................1
2. PREVIOUS RESEARCH ON LONG BONES.............................7
Research on Linear Growth in Living Populations..........7
Previous Studies of Linear Bone Growth in
Archaeological Populations..............................10
Research on Cortical Bone Growth in Living Populations..19
Previous Studies of Cortical Bone Growth in
Archaeological Populations..............................23
Summary.................................................27
3. HISTORICAL BACKGROUND OF MILWAUKEE COUNTY..........29
A History of the Development and Health of Milwaukee
County in the 19th and Early 20th Centuries.............29
History of Milwaukee Countys Institutions .............36
History and Archaeology of the Milwaukee County
Poor Farm Cemetery......................................54
The Population Buried at the Milwaukee County
Poor Farm Cemetery..................................60
Summary.................................................63
vm


4. MATERIALS AND METHODS........................................65
The MCIG Subadult Sample.................................65
Types of Pathologies Present.........................67
Age Determination....................................71
Measurements of Subadult Long Bones..................72
Statistical Analyses.................................76
5. RESULTS......................................................84
Linear Bone Growth Analysis..............................84
Cortical Bone Growth Analysis............................98
6. DISCUSSION..................................................113
Linear Bone Growth......................................113
Cortical Bone Growth....................................119
Hypotheses Revisited....................................122
7. CONCLUSION..................................................125
APPENDIX
A. Scatter Plots of Individual Diaphyseal Lengths for the
MCIG Subadult Sample........................................129
B. The Z-score Distributions of Individual Diaphyseal
Lengths for the MCIG Subadult Sample........................134
C. Mean Curves for Diaphyseal Length and 95 Percent
Confidence Intervals for the MCIG Subadult Sample...........139
D. Relative Percent Increase in Long Bone Diaphyseal Length
Growth: MCIG Subadults vs. Modem Denver Children............144
IX


E. Mean Curves of Cortical Bone Growth for the MCIG
Subadult Sample..............................................147
F. The Z-score Distributions of Individual Percent Cortical
Areas for the MCIG Subadult Sample...........................157
REFERENCES............................................................160
x


LIST OF FIGURES
Figure
3.1 1848 Map of Milwaukee County Including the City of Milwaukee,
Town of Wauwatosa, and the County Institutional Grounds.............31
3.2 Map Showing the Three Sections of Milwaukee, 1846....................33
3.3 The County Poor Farm Depicted in a 19th Century Oil Painting.........38
3.4 The County Insane Asylum Built in 1880...............................42
3.5 Youth Participating in Outdoor Activities at the Home for
Dependent Children in the 1920s.....................................45
3.6 Reasons for Placement into the Home for Dependent
Children, 1900......................................................46
3.7 Reasons for Placement into the Home for Dependent
Children, 1910......................................................47
3.8 Reasons for Placement into the Home for Dependent
Children, 1920......................................................47
3.9 Number of Subadult Deaths at the Home for Dependent
Children, 1904-1920.................................................50
3.10 The Remodeled Almshouse in 1924.....................................52
3.11 Map Showing the Three Cemeteries Located on the
Milwaukee County Institutional Grounds..............................56
3.12 Location of Excavated Graves at the Milwaukee County
Poor Farm Cemetery..................................................59
3.13 Places of Death Listed for Subadults in the Register of Burials
at the Milwaukee County Poor Farm, 1899-1907........................62


4.1 Radiograph of a Left Femur (lot no. 6180) in the A-P Plane........74
5.1 Mean Curves for Diaphyseal Length of Humerus for the MCIG
Sample and Modem Denver Sample...................................89
5.2 Mean Curves for Diaphyseal Length of Radius for the MCIG
Sample and Modern Denver Sample..................................89
5.3 Mean Curves for Diaphyseal Length of Ulna for the MCIG
Sample and Modem Denver Sample...................................90
5.4 Mean Curves for Diaphyseal Length of Femur for the MCIG
Sample and Modem Denver Sample...................................90
5.5 Mean Curves for Diaphyseal Length of Tibia for the MCIG
Sample and Modern Denver Sample..................................91
5.6 Mean Curves for Diaphyseal Length of Humerus for the MCIG
Sample and St. Thomas Sample....................................92
5.7 Mean Curves for Diaphyseal Length of Radius for the MCIG
Sample and St. Thomas Sample....................................93
5.8 Mean Curves for Diaphyseal Length of Ulna for the MCIG
Sample and St. Thomas Sample....................................93
5.9 Mean Curves for Diaphyseal Length of Femur for the MCIG
Sample and St. Thomas Sample....................................94
5.10 Mean Curves for Diaphyseal Length of Tibia for the MCIG
Sample and St. Thomas Sample....................................94
5.11 Mean Curve for Femur Cortical Thickness for the MCIG Sample
and Modem, Healthy Children from Finland.........................107
5.12 Mean Curve for Tibia Cortical Thickness for the MCIG Sample
and Modern, Healthy Children from Finland........................107
5.13 Mean Curve for Femur Cortical Index for the MCIG Sample
and Modern, Healthy Children from Finland........................108
xii


5.14 Mean Curve for Tibia Cortical Index for the MCIG Sample
and Modem, Healthy Children from Finland.............................108
5.15 Mean Curve for Femur Cortical Thickness for the MCIG Sample
and Wharram Percy Sample.............................................110
5.16 Mean Curve for Femur Cortical Index for the MCIG Sample
and Wharram Percy Sample.............................................110
5.17 Mean Curve for Tibia Percent Cortical Area for the MCIG
Sample and Nubia Sample..............................................112
A. 1 Scatter Plot of the Humerus Plotting Diaphyseal Lengths for
Each Defined Age Group...............................................129
A.2 Scatter Plot of the Radius Plotting Diaphyseal Lengths for
Each Defined Age Group...............................................130
A.3 Scatter Plot of the Ulna Plotting Diaphyseal Lengths for
Each Defined Age Group...............................................131
A.4 Scatter Plot of the Femur Plotting Diaphyseal Lengths for
Each Defined Age Group...............................................132
A. 5 Scatter Plot of the Tibia Plotting Diaphyseal Lengths for
Each Defined Age Group...............................................133
B. 1 The Z-score Distribution of Diaphyseal Lengths for the Humerus
Calculated from the Mean Age Group Values............................134
B.2 The Z-score Distribution of Diaphyseal Lengths for the Radius
Calculated from the Mean Age Group Values............................135
B.3 The Z-score Distribution of Diaphyseal Lengths for the Ulna
Calculated from the Mean Age Group Values............................136
B.4 The Z-score Distribution of Diaphyseal Lengths for the Femur
Calculated from the Mean Age Group Values............................137
B.5 The Z-score Distribution of Diaphyseal Lengths for the Tibia
Calculated from the Mean Age Group Values............................138
xiii


C. 1 Mean Curve for Diaphyseal Length and 95 Percent Confidence
Intervals Plotted Against Mean Values for Each Age Group
for the Humerus.....................................................139
C.2 Mean Curve for Diaphyseal Length and 95 Percent Confidence
Intervals Plotted Against Mean Values for Each Age Group
for the Radius......................................................140
C.3 Mean Curve for Diaphyseal Length and 95 Percent Confidence
Intervals Plotted Against Mean Values for Each Age Group
for the Ulna........................................................141
C.4 Mean Curve for Diaphyseal Length and 95 Percent Confidence
Intervals Plotted Against Mean Values for Each Age Group
for the Femur.......................................................142
C. 5 Mean Curve for Diaphyseal Length and 95 Percent Confidence
Intervals Plotted Against Mean Values for Each Age Group
for the Tibia.......................................................143
D. 1 Relative Percent Increase in Diaphyseal Length Growth
of the Humerus......................................................144
D.2 Relative Percent Increase in Diaphyseal Length Growth
of the Radius.......................................................145
D.3 Relative Percent Increase in Diaphyseal Length Growth
of the Ulna ........................................................145
D.4 Relative Percent Increase in Diaphyseal Length Growth
of the Femur........................................................146
D. 5 Relative Percent Increase in Diaphyseal Length Growth
of the Tibia........................................................146
E. 1 Mean Curves for Total Subperiosteal Width and Medullary
Cavity Width Plotted for the Humerus................................147
E.2 Mean Curves for Total Subperiosteal Area and Medullary
Cavity Area Plotted for the Humerus.................................148


E.3 Mean Curve for Cortical Thickness Plotted for the Humerus................148
E.4 Mean Curves for Cortical Area Plotted for the Humerus....................149
E.5 Mean Curve for Cortical Index Plotted for the Humerus....................149
E.6 Scatter Plot of the Humerus for Percent Cortical Area Plotted for
Each Defined Age Group.................................................150
E.7 Mean Curves for Total Subperiosteal Width and Medullary
Cavity Width Plotted for the Femur.....................................151
E.8 Mean Curves for Total Subperiosteal Area and Medullary
Cavity Area Plotted for the Femur......................................151
E.9 Mean Curve for Cortical Thickness Plotted for the Femur..................152
E.10 Mean Curves for Cortical Area Plotted for the Femur.....................152
E. 11 Mean Curve for Cortical Index Plotted for the Femur....................153
E. 12 Scatter Plot of the Femur for Percent Cortical Area Plotted for
Each Defined Age Group.................................................153
E. 13 Mean Curves for Total Subperiosteal Width and Medullary
Cavity Width Plotted for the Tibia.....................................154
E.14 Mean Curves for Total Subperiosteal Area and Medullary
Cavity Area Plotted for the Tibia......................................154
E. 15 Mean Curve for Cortical Thickness Plotted for the Tibia................155
E. 16 Mean Curves for Cortical Area Plotted for the Tibia....................155
E. 17 Mean Curve for Cortical Index Plotted for the Tibia....................156
E. 18 Scatter Plot of the Tibia for Percent Cortical Area Plotted for
Each Defined Age Group.................................................156
F. 1 The Z-score Distribution of Percent Cortical Areas for the
Humerus Calculated from the Mean Age Group Values......................157
xv


F.2 The Z-score Distribution of Percent Cortical Areas for the
'Femur Calculated from the Mean Age Group Values.......................158
F.3 The Z-score Distribution of Percent Cortical Areas for the
Tibia Calculated from the Mean Age Group Values........................159


LIST OF TABLES
Table
3.1 Causes of death for subadults living at the Home for
Dependent Children.......................................................49
3.2 Nationality of Almshouse inmates, 1906....................................53
3.3 Movement of Almshouse population, 1906....................................53
4.1 Details about the MCIG subadult sample....................................66
4.2 Types of non-specific pathological conditions observed in the
MCIG subadult sample.....................................................67
4.3 Cortical measurements obtained from long bones in the
MCIG sample..............................................................75
4.4 Statistical summary of diaphyseal length data for modern
Denver children (Maresh 1970)............................................78
4.5 Statistical summary of diaphyseal length data for the
St. Thomas skeletal sample (Saunders et al. 1993b)......................78
4.6 Statistical summary of cortical thickness and cortical index for
femora from modern, healthy children (Virtama and Helela 1969)...........82
4.7 Statistical summary of cortical thickness and cortical index for
tibias from modern, healthy children (Virtama and Helela 1969)...........83
4.8 Statistical summary of cortical thickness and cortical index for
subadult femora from Wharram Percy (Mays 1999a)..........................83
4.9 Statistical summary of percent cortical area for subadult tibias
from Nubia (Hummert 1983)................................................83
xvii


5.1 Sample size, mean maximum diaphyseal length (mm), standard
deviation, and 95% confidence interval of the mean for each of
the five long bones in the MCIG subadult sample..........................85
5.2 Summary of significant results of likelihood-ratio tests for outliers
for the variable maximum diaphyseal length................................86
5.3 Relative increase in percent for MCIG (M) and modern
Denver children (D).......................................................97
5.4 Paired t tests comparisons of maximum diameter (D) and total
subperiosteal width (T) measurements for the age groups fetal
and 1.0 2.9 months......................................................99
5.5 Wilcoxon signed rank tests comparisons of maximum diameter (D)
and subperiosteal bone width (T) measurements for various
age groups...............................................................100
5.6 Means and standard deviations of attained cortical bone growth
for the MCIG humerus.....................................................102
5.7 Means and standard deviations of attained cortical bone growth
for the MCIG femur.......................................................102
5.8 Means and standard deviations of attained cortical bone growth
for the MCIG tibia.......................................................103
5.9 Summary of significant results of likelihood-ratio tests for outliers
for percent cortical area (%CA)..........................................105
xvi n


CHAPTER 1
INTRODUCTION
Human remains, particularly long bones from archaeological sites, provide a
valuable retrospective picture of health and behavior in the past that can be used to
query historical records of health status and behaviors of once-living populations.
Health status is highly influenced by many factors including nutrition, disease,
behavior, socioeconomic status, family size, and the environment. Using diachronic
evidence and other comparative studies of human remains recovered from
archaeological sites, anthropologists are able to determine lifetime events such as the
nutritional and health status of past populations. Many studies over the recent years
using human skeletal remains have looked at the role that almshouses in the United
States played in shaping the biology of individuals considered to be the unfortunate
and poorest members of society (Grauer and McNamara 1995; Higgins 2001; Higgins
and Sirianni 1995; Phillips 2001; Sirianni and Higgins 1995; Wesolowsky 1991).
Katz (1986, 1995) argues that the establishment of almshouses by state governments
throughout the 19th century in the United States served not only to provide relief (in
the form of food, shelter, and clothing) for the growing number of destitute people
living in different cities and counties, but also to isolate individuals from the outside
world as a deterrence from laziness, intemperance, and pauperism.
1


Although almshouses were intended to provide temporary assistance for those in
need, it was often the case that they became home for the sick, elderly, pregnant, and
mentally ill for many years with some individuals never leaving (Higgins 2001). Even
more unfortunate is the fact that historical documents and analyses of skeletal remains
from the nations various almshouse cemeteries reveal that many of these institutions
failed to properly provide for their inmates and protect them from poor health and
disease (Higgins 2003; Higgins and Sirianni 1995). Interment records kept by
almshouse cemeteries along with skeletal remains from these cemeteries strongly
indicate that many residents (including children and infants) suffered from nutritional
stress and various types of gastrointestinal and respiratory diseases, aside from being
particularly vulnerable to epidemic diseases such as typhus and cholera whenever
they swept through major cities (Higgins 2001; Higgins and Sirianni 1995).
Studies of poorhouses in New York and Chicago indicate that child and infant
mortality was high in these institutions, with in some cases almost one-half of
subadults under the age of 15 dying within the first year of life (Grauer and
McNamara 1995; Higgins 2003; Sirianni and Higgins 1995). Gastrointestinal diseases
such as inflamed bowel, diarrhea, and dysentery at the Monroe County Almshouse
in New York, for example, have been reported to have accounted for a large
percentage of deaths among young inmates, along with respiratory diseases,
particularly tuberculosis, which was the leading cause of death among infants
(Higgins and Sirianni 1995). Many children over the age of one also died from acute
2


infectious diseases such as typhus, cholera, and measles as a result of squalid living
conditions (Higgins 2003). Higgins (2001) further argues that many almshouses
seemingly offered little in terms of prenatal care to women who came to the
institutions to give birth. Evidence for this is based on the high rates of neonatal
mortality observed at the Erie County Poorhouse in New York. Furthermore, the
presence of enamel hypoplasias in the deciduous teeth of subadult skeletons from the
Monroe County Almshouse has been attributed to poor prenatal conditions and early
childhood stress (Higgins and Sirianni 1995).
Skeletal remains from almshouse cemeteries provide an opportunity to test various
hypotheses about the living conditions and health status of low income populations
residing in Wisconsin during the 19th and early 20th centuries. The Poorhouse
Cemetery, the Almshouse Cemetery, Potters Field, and the Poor Farm Cemetery are
all different names that have been used to refer to the cemetery located on the
Milwaukee County Institutional Grounds (MCIG) in Wauwatosa, Wisconsin. The
Milwaukee County Poor Farm Cemetery is one of three locations on the MCIG that
was used as a place of interment for the institutional paupers of Milwaukee County,
along with the community poor, unidentified individuals found in the County, and
individuals who died in the County without kin. Excavations of the cemetery
(conducted by Great Lakes Archaeological Research Center, Inc) that were carried
out in 1991 and 1992 recovered a total of 1,649 burials of which 588 were subadults.
All of the skeletal remains, which span the period 1882 to 1925, are currently stored
3


at Marquette University and provide a retrospective picture of health and behavior of
the MCIG past population.
This study focuses on the health of the children who were interred at the
Milwaukee County Poor Farm Cemetery. More specifically, this study asks the
question, are long bone measurements indicative of poor health status in the subadults
recovered from the Milwaukee Poor Farm Cemetery? Analyses of the subadult
remains demonstrate little evidence of there being obvious pathologies present;
however, this might be attributed to the fact that many of the subadults died at a very
young age, thus precluding the development of skeletal markers indicating disease. In
addition, Ortner (1998) points out that, in general, only a small percentage of
individuals who have an infectious disease will exhibit skeletal evidence of the
disease. This is because most infectious diseases that leave their mark on the skeleton
are the result of chronic conditions in which the individual lived with the disease for
many years. In light of these different possible explanations of why the subadults
interred at the Poor Farm Cemetery demonstrate little evidence of skeletal
pathologies, one may still attempt to determine the health status of this subadult
population by examining the patterns of linear and cortical bone growth present.
Although some researchers might disagree that these different variables can provide
accurate information about the health status of a population, my hypothesis is that if
indeed the subadult population buried at the Milwaukee County Poor Farm Cemetery
4


suffered from poor health and stress, then the patterns of linear and cortical bone
growth present should reveal this.
In addition to describing the general health status of the subadults interred at the
Milwaukee County Poor Farm Cemetery, this study hopes to contribute to the field of
biological anthropology in the following ways:
1. highlighting growth parameters (linear or cortical bone growth) that seem
most useful for assessing the overall health of the MCIG subadult population,
which can be applied to future studies of other similar archaeological
populations.
2. addressing some of the methodological issues posed in doing cortical bone
research such as the compatibility of measurements taken from dry specimens
and those taken from X-rays, and the benefits of using a digitizing program to
obtain cortical bone measurements.
3. providing cortical bone data on a sample of children living in the United
States during the late 19th and early 20th centuries that appears to be currently
lacking in the literature.
This study is organized into seven chapters where Chapter One provides an overall
introduction and a brief review of the health status of almshouse populations. Chapter
Two discusses some of the previous research that has been done on long bones to
infer the health of populations. Chapter Three provides a brief historical background
of the development and health of Milwaukee County during the 19th and early 20th
5


centuries and introduces the archaeology of the Milwaukee County Poor Farm
Cemetery. Chapter Four describes the subadult sample and methods used. Chapter
Five provides the results of the statistical analyses of the linear and cortical bone data
collected. Chapter Six interprets the results of the statistical analyses, and Chapter
Seven provides a summary of the research conclusions generated by this study.
6


CHAPTER 2
PREVIOUS RESEARCH ON LONG BONES
Research on Linear Growth in Living Populations
Numerous studies of normal, healthy living children reveal that changes in either
length or height with age follow a predictable pattern. Usually during the first year of
life linear growth is very rapid, and growth velocities are at their highest (Bogin
1988; Johnston 1978). After 12 months of postnatal life, growth velocities
dramatically decline and begin to level off around the age of three (Bogin 1988).
Growth rates at ages three and above usually continue at a slow and steady rate, with
some individuals showing a slight decrease around the age of 11 (Bogin 1988). A
reversal in the rate of growth occurs after the 11th year, marking the transition from
late childhood to adolescence, which is commonly referred to as the adolescent
growth spurt (Bogin 1988). This pattern describes the velocity curve of growth that is
commonly observed among healthy children. Distance curves of linear growth, on the
other hand, normally show continuous increases in both height and long bone length
from birth until adulthood or when growth is complete (Bogin 1988; Johnston 1978;
Maresh 1943).
Although infant and childhood growth tends to follow a predicable pattern, studies
of contemporary populations indicate that, in addition to heredity, nutrition and
7


various environmental factors can strongly influence the amount and rate of linear
growth (Bogin 1988). For example, modern studies within developed and developing
nations have consistently demonstrated that height differentials occur between
children of low and high socioeconomic status due to differences in nutritional status,
disease rate, and medical care between the two groups (Stinson 2000). Growth
differences have also been observed between urban and rural populations, especially
in developing countries. Urban children in developing countries are typically taller
than their counterparts in rural areas as a result of having access to a better food
supply and basic municipal services (Bogin 1988; Stinson 2000). However, the
reverse seems to have been true for children growing up in industrialized nations
during the 19th and early 20th centuries. According to Stinson (2000), urban children
in industrialized nations during that period were generally shorter than rural children
due to undesirable living conditions such as overcrowding, poor sanitation, and
pollution which contributed to poor health and disease.
Many studies of contemporary populations reveal that arrested growth almost
always begins early in life, particularly within the first two or three years (Beaton
1992; Martorell 1995). Studies of infants and children living in low income countries
have found that linear skeletal growth failure is usually detectable by six months of
age and complete by the second year, after which catch-up growth may occur if
conditions are not too severe (Beaton 1992; Martorell 1995). However, some
8


researchers have shown that achievement of complete catch-up growth is rare due to
the following reasons:
i. children who undergo growth faltering often fail to leave the environment that
contributed to their shortness,
ii. one usually cannot experience prolonged linear growth rates later in life that
are as high as those experienced during infancy (Stinson 2000).
The fact that growth rates are the highest during infancy explains why children are
most affected by growth faltering during the first two or three years of life. The rapid
rate of growth means that nutritional needs are highest during the early years and the
developing infant is more susceptible to malnutrition caused by insufficient
supplemental foods or foods that are low in nutritional value (Martorell 1995; Stinson
2000). Infants and young children are also more susceptible to infectious diseases
because of their weaker immune systems, which can disrupt growh (Martorell 1995).
Furthermore, it is often the case that malnutrition and infection interact synergistically
with one another thus interfering with proper long bone growth in the young. For
example, poor nutritional status can lead to impaired immunocompetence and
reduced resistance to infection, and exposure to infectious diseases can lead to
malabsorption, anorexia, and protein catabolism in order to generate the acute phase
proteins needed in the immune response (King and Ulijaszek 1999).
9


Previous Studies of Linear Bone Growth in
Archaeological Populations
In growth-related studies of archaeological populations, data are primarily derived
from long bone measurements, which represent a simple measure of stature (Saunders
and Hoppa 1993). These studies, just as in studies of living children, often assume
that differences found among populations in terms of stature and rates of growth
reflect the health status and level of well-being of those populations (Buikstra and
Cook 1980; Saunders and Barrans 1999). What is to follow is a survey of some of the
different articles published over the years that have looked at linear bone growth in
past human populations. These articles provide insight as to the ways in which linear
bone growth can be analyzed, and how long bone growth patterns can be interpreted
in order to infer past health.
One of the earliest investigations of linear bone growth from an archaeological
skeletal sample includes Johnstons 1962 study of juveniles from the Indian Knoll site
in Kentucky. This particular site dates approximately 5000 yearsago and contain s the
skeletal remains of an American Indian population that subsisted entirely by hunting
and gathering. Long bone measurements were obtained from 165 individuals ranging
in age from fetal to 5.5 years of age that could be aged accurately using dental and
osseous criteria. The majority of individuals included in the study died before the age
of one indicating a high level of infant mortality. In order to interpret the growth
curves developed for the Indian Knoll sample, Johnston (1962) compared the sample
10


means, standard deviations, and percentages of relative growth to Mareshs (1955)
data on healthy white children from the United States. Johnston (1962) noted a
similar steady increase in long bone length for both populations, but the Indian Knoll
sample displayed slower growth rates with mean differences in length statistically
significant after two years of age. Johnston (1962) concluded that the depressed
growth rates found in the Indian Knoll sample were likely the result of genetic
tendencies toward population shortness reinforced by living in a rigorous
environment.
Armelagos et al. (1972) looked at linear bone growth in a sample of skeletons
recovered from various Sudanese Nubian cemeteries, dating from 350 B.C. to 1400
A.D. In this study, Armelagos et al. (1972) were specifically concerned with
quantifying the rate and course of growth of six long bone shafts including the femur,
tibia, humerus, radius, ulna, and clavicle. Shaft length was measured for 61
individuals ranging in age from six months to 16 years. Ages were determined
according to dental eruption patterns, and the means, standard deviations, and
percentages of relative increase were calculated for each bone in the different
developmental age groups. Armelagos et al. (1972) discovered that all of the bones
went through a continuous incremental increase in length, with growth being the
greatest for individuals between the ages of six months and one year, and 13 and 16
years. Comparisons between the Nubian means and Mareshs (1955) data on
American boys in the fiftieth percentile range revealed that, although both
11


populations undergo a period of decelerated growth in early childhood, American
boys are taller than prehistoric Nubians. Armelagos et al. (1972) suggested that
moderate nutritional deficiencies along with genetic differences may explain why
Nubian subadults are shorter than American boys.
Other researchers have examined linear bone growth patterns in early American
Indian populations in order to assess the amount of variation present between
populations of different time periods. For example, Cooks 1984 study of prehistoric
American Indians from west central Illinois compared femur lengths of juveniles
(aged from birth to six years) between Middle and early Late Woodland hunter and
gatherers, late Late Woodland transitional maize agriculturalists, and Mississippian
intensive maize agriculturalists. Regressions of diaphyseal length to dental age for
these groups revealed that the late Late Woodland population has the shortest femora
for age compared to the earlier and later populations. It was also observed that
individuals who were short for their age had higher frequencies of stress indicators
such as cribra orbitalia and circular caries than those who had long femora for their
age. Cook (1984) concluded that growth retardation among the transitional maize
agriculturalists was the result of nutritional deficiencies brought on by the
introduction of a low protein quality maize diet during the late Late Woodland period,
and population pressure on scarce resources.
Jantz and Owsley (1984b) analyzed linear bone growth variation among ten
skeletal samples of Arikara recovered from archaeological sites located in the Middle
12


Missouri subarea of South Dakota. Each sample came from one of three variants
comprising the Coalescent Tradition: the Extended Coalescent, dating from 1550 to
1675; the Postcontact Coalescent from 1675 to 1780; and the Disorganized
Coalescent from 1780 to 1862. Developmental ages were determined using the dental
standards of Moorees et al. (1963a, 1963b). Only individuals between the ages of
approximately six months and 12 years were included in the study, and measurements
were taken on the femur, humerus, tibia, and radius. Regressions of age on bone
length for each variant showed that the Postcontact Coalescent had slightly longer
bones in early childhood than the Extended, but increasingly longer bones lengths
(with the exception of the radius) during later childhood. The Disorganized
Coalescent by late childhood, on the other hand, had bone lengths either below or in
between the other two variants. Overall, Jantz and Owsley (1984b) claimed that their
study demonstrated the ecological sensitivity of long bone growth variation. They
attributed the improved growth status of the Postcontact groups in relation to the
Extended to climatic changes and acquisition of the horse which led to surpluses of
food and better nutrition. Growth failure of the Disorganized Coalescent groups
during late childhood was considered by the authors to be the result of high levels of
morbidity and undemutrition caused by food shortagesand the introduction of
epidemic diseases.
Lovejoy et al. (1990) measured the length of all six major long bone diaphyses in
children ranging in age from birth to 12 years from a Libben Late Woodland
13


population. Lovejoy et al. (1990) were interested in growth velocity in order to
determine whether growth rates of long bones in the Libben sample differed from
those reported by Maresh (1955) for healthy American children. In this study, a total
of 152 individuals were included in the Libben sample, and all were aged according
to published dental standards, which the authors modified slightly to account for
dental differences between American Indians and Caucasians. Lovejoy et al. (1990)
determined the mean length of every long bone in each age group. Each mean was
then divided by the mean adult length of the corresponding bone in order to obtain
percentages of achieved adult length as a means to control for genetic differences in
growth between the populations. The results of the study revealed a strong depression
of growth velocity in the Libben sample relative to modem childrerduring the first
three years of life. After the age of three, growth rates appeared to be approximately
equal between the two groups. Although Lovejoy et al. (1990) recognized that
malnutrition could lead to growth faltering during the early years, they alternatively
suggested that high levels of systemic infectious disease was a more probable agent
causing growth retardation at Libben. Their conclusion was based on previous studies
that showed the Libben diet to be quantitatively and qualitatively adequate, and also
on the observation that over 50 percent of individuals aged zero to three years showed
signs of periostitis.
Hoppa (1992) measured the diaphyseal lengths of all six major long bones in 69
subadults excavated from a 10th century Anglo Saxon cemetery from the site of
14


Raunds in England. The purpose of Hoppas (1992) study was to not only evaluate
the pattern of skeletal growth in an Anglo-Saxon population, but also present long
bone data that could be used to address methodological problems inherent in
estimating age from diaphyseal length. Individual ages, spanning the range from one
month to 18.5 years, were estimated using the dental standards of Moorrees et al.
(1963a, 1963b) and Anderson et al. (1976). Skeletal growth profiles for the humerus,
radius, ulna, femur, tibia, and fibula were createdby plotting the mean diaphyseal
lengths for every one-year age cohort, and these profiles were compared to similar
growth curves established for modem Caucasians (Maresh 1955), 9th century Slavic
(Stloukal and Hanakova 1978), and 6th to 7th century individuals from Altenerding in
West Germany (Sundick 1978). Hoppa (1992) observed that all of the early
populations have smaller means for each age cohort than modem Caucasians starting
around the age of one. In fact, statistical comparisons between the overall sample
means for the femur revealed that that the modem sample is significantly larger for
age than the other samples, and the Raunds sample is significantly larger than the
Slavic sample. Since the Raunds infants fall within the post-neonatal period, Hoppa
suggested that death for these individuals was probably the result of extrinsic factors
such as poor diet and disease, which would have also played a role in the deaths of
older children and adults.
Saunders et al. (1993b) examined diaphyseal length data collected from a large
sample of subadult skeletons from a 19th century church cemetery in Belleville,
15


Ontario. The church cemetery was in use from 1821 to 1874 and mainly served as a
burial place for those who were members, or rfeatives of member^ of St. Thomas
Anglican Church. Many of the people who settled in Belleville during the early 19th
century were immigrants from the British Isles, Ireland, and Western Europe, as well
as descendents of the United Empire Loyalists from the United States.
The St. Thomas subadult sample included 216 skeletons estimated up to 15 years
of age using the tooth standards of Moorrees et al. (1963a, 1963b). The diaphyseal
lengths of all six major long bones were measured, and skeletal growth profiles were
created by plotting the distributions of these lengths against estimates of
chronological age. Saunders et al. (1993b) compared the growth profiles of the church
sample to the modem standards of Maresh (1970) and to two other archaeological
samples including a protohistoric Arikara sample from South Dakota (Merchant and
Ubelaker 1977), and a collection of subadults from the 10th century Raunds Anglo-
Saxon site in England (Hoppa 1992). The means and 95 percent confidence intervals
for each sample were also calculated and compared to one another.
The results of the study revealed little overlap in the confidence limits of the St.
Thomas sample and the other two archaeological samples. Differences between the
samples were especially notable after the age of one in the Raunds sample and after
the age of two in the Arikara sample. Saunders et al. (1993b), however, found much
overlap between the skeletal growth profiles and confidence limits of the modem and
St. Thomas samples indicating that 19th century St. Thomas children followed a
16


long bone growth pattern similar to that of modem North American children. The
authors did observe that growth rates for the St. Thomas sample tended to be slightly
lower than modem standards between birth and two years of age. Saunders et al.
(1993b) suggested that poor maternal health during pregnancy associated with lower
standards of health care during the 19th century may have led to slower growth rates
during the early years for children from Belleville as well as malnutrition and disease,
which apparently had more of an impact on the health of Arikara and Raunds
children.
Miles and Bulman (1994) examined linear bone growth in a collection of
immature skeletal remains from the island of Ensay in the Western Isles of Scotland.
In their study, the sample included 120 individuals ranging in age from six months of
fetal life to 20 years of age that lived between 1500 and 1850 A.D. Individual ages
were estimated based on the state of epiphyseal fusion and tooth development, and
the diaphyses of all six major long bones were measured, along with the lengths of
the first metacarpal, first metatarsal, calcaneus, talus, and cuboid.
Scatterplots prepared from each of the sets of measurements revealed a curve of
rapid increase during the perinatal and infancy period, followed by a period of slower
growth after the first year, and a final period of rapid increase corresponding to the
adolescent growth spurt around 17 years of age. Miles and Bulman (1994) compared
the growth patterns of the diaphyses to curves prepared from Mareshs (1955) data on
North American whites using combined male-female weighted means, as well as
17


curves prepared from data on ancient Slavs (Stloukal and Hanakova 1978), 6th to 7th
century Germans (Sundick 1978), early American Indians (Jantz and Owsley 1984;
Merchant and Ubelaker 1977; Sundick 1978), and pre-19th century Eskimos and
Aleut juveniles (YEdynak 1946). These comparisons revealed a pattern of
diaphyseal growth in Ensay children that was more similar to ancient populations
than to modem North American whites. A noticeable difference in diaphyseal lengths
between Ensay and modem white juveniles was observed as early as four months of
age.
The diaphyseal length means of the full-term and younger fetal Ensay long bones
were additionally compared to those reported by Fazekas and Kosa (1978) for modem
Hungarian fetuses. Although few fetuses were recovered from the Ensay population,
the authors argued that the amount of difference between the data sets was too small
to support the claim that the Ensay children were small at birth. However, the
postnatal growth pattern for the Ensay sample provided enough evidence to indicate
that apart from the early infancy period, Ensay children were small in stature. Miles
and Bulman (1994) suggested that factors other than undemutrition probably led to
slow long bone growth in the Ensay children, including hard work and living in
damp, smoky, airless houses, which would have promoted ill-health and made them
susceptible to disease. Miles and Bulman (1994) concluded that a diet comprised
primarily of fish would have prevented children from suffering from malnutrition.
18


Research on Cortical Bone Growth in Living
Populations
Patterns of cortical bone growth and maintenance have been well established in
living people of both sexes for all ages (Frisancho et al. 1970a; Gam 1970; Johnston
and Malina 1966). Studies involving radiographic measurements of tubular bones
from clinically-normal individuals reveal that bone apposition at the subperiosteal
surface during infancy and childhood closely resembles, in pattern, diaphyseal growth
in that there is an initial period of rapid increase, representing an extension of late
prenatal growth, and a later period of more moderate increase that continues until the
onset of the adolescent growth spurt. According to Gam (1970), rapid early rates of
subperiosteal apposition are well documented in subjects from Central American
nations and the United States. The data from these countries reveal that the initial
rapid rate of appositional growth ends around six months of life. Afterwards, the rate
of subperiosteal gain declines until the age of two, at which point it settles to a lesser
long-term rate. As with the subperiosteal surface, changes at the endosteal surface are
also disproportionately rapid during the first six months of life. During this period the
medullary cavity width rapidly increases, and then, by the age of one, the rate of
endosteal resorption decreases, remaining at a steady rate until adolescence.
All of these changes at both the subperiosteal and endosteal surfaces during
infancy and childhood consequently lead to changes in cortical thickness. Gam
(1970:46-47) points out that during part of the first year of life, between three and
19


nine months of age, cortical bone actually becomes thinner. This is because the
neonatal rate of subperiosteal apposition decreases faster than the neonatal rate of
endosteal loss. After nine months of age cortical bone thickness begins to increase at
a steady rate again.
The adolescent spurt is evident in both sexes when subperiosteal diameter,
medullary width, and cortical thickness are taken into account. The adolescent spurt
in subperiosteal diameter normally occurs between the ages of 10 and 12 in girls and
12 and 14 in boys in the United States (Frisancho et al. 1970a). Among children
living in rural communities in Central America, the onset of the adolescent spurt
appears to take place later between the ages of 12 and 14 in females and 13 and 16 in
males (Frisancho et al. 1970a). At the endosteal surface, childhood increase in
medullary cavity width continues until the age of 12 in girls and 16 in boys in the
United States (Frisancho et al. 1970a). After this point, the endosteal surface enters
the phase of bone apposition, which causes the width of the medullary cavity to
decrease until the fourth decade of life (Gam 1970). Cortical thickness also undergoes
a growth spurt in both sexes as a result of changes at both the subperiosteal and
endosteal surfaces. According to Gam (1970), the spurt in cortical gain is earlier and
more pronounced in females because they add at both bone surfaces earlier and
quicker than males. By the age of 40, cortical bone has reached its maximum
thickness in both sexes, and cortical bone loss occurs as endosteal resorption begins
to exceed the rate of increase in total subperiosteal diameter (Gam 1970).
20


Researchers, however, point out that cortical bone loss can also occur in children,
particularly among those suffering from genetic disorders, hormonal imbalances,
disease, and nutritional stress (Gam 1970; Hummert 1983). Cortical loss in these
cases can result from either reduced subperiosteal apposition, increased endosteal
surface resorption, or both (Gam 1970). Gam and colleagues (1964), for example,
radiographed the second metacarpal in Guatemalan children hospitalized for
treatment of acute protein-calorie malnutrition. Compared to Guatemalan children
who are known to have a low average intake of dietary protein, Gam et al. (1964)
noted a significant reduction in cortical bone thickness among the hospitalized
children. Even during the recovery period the authors observed that some of the
hospitalized children experienced further decreases in cortical thickness while bone
length increased, providing evidence for actual bone loss and not just failure to gain.
Additional research by Gam and co-workers (1969) revealed that cortical thinning
was the result of having increased medullary cavity widths caused by marked
endosteal resorption, rather than reduced subperiosteal diameters. Interestingly,
cortical thinning as a result of increased endosteal resorption has also been observed
in individuals suffering from hyperparathyroidism, renal tubular acidosis, and long-
term inactivity (Gam 1970). On the other hand, in a variety of malabsorption
syndromes and chromosomal abnormalities cortical thickness and area are reduced
due to both increased bone loss at the endosteal surface and reduced subperiosteal
gain during the juvenile phase (Gam 1970).
21


The idea that nutritional stress can lead to reduced cortical thickness in the bones
of children has also been supported by other research studies. Himes et al. (1975), for
example, gathered radiographic data on rural Guatemalan children between the ages
of one and seven suffering from chronic, mild to moderate protein-calorie
malnutrition and compared it to data on well-nourished white children living in the
United States. Himes et al. (1975) discovered that the Guatemalan children
significantly lagged behind U.S. children in periosteal diameter, cortical thickness,
cortical area, and percent cortical area. However, the Guatemalan children eventually
demonstrated catch-up growth in percent cortical area. According to Himes et al.
(1975), retardation in cortical bone growth among the Guatemalan children occurred
even when stature and weight were controlled for indicating that diminished bone
growth was not proportionate to either stature or weight. Himes et al. (1975)
suggested that chronic, mild to moderate protein-calorie malnutrition may be a more
important factor in slowing subperiosteal diameter growth than acute protein-calorie
malnutrition since medullary diameters were very similar between the rural
Guatemalan and U.S. children.
Overall, studies on cortical bone growth among living subjects reveal that cortical
thickness can serve as a useful osteological indicator of nutritional and disease stress
much in the same way that stature or enamel hypoplasias, for example, are often used
to determine these types of stresses. The fact that bones can react to stress in different
ways means that it is important to take into consideration as many skeletal markers as
22


possible in order to properly assess the health of a population or group of people. By
developing and comparing cortical bone growth profiles between different groups or
populations, one may be able to obtain evidence of the timing and extent of growth
faltering of one group or population relative to another, as well as additional
information on the general trends in cortical bone growth over time.
Previous Studies of Cortical Bone Growth in
Archaeological Populations
The study of cortical bone thickness in excavated human skeletal remains has a
long history, primarily involving the study of ancient hominids and early members of
the genus Homo (Day 1971; Kennedy 1985; Pfeiffer and Zehr 1996; Ruff et al. 1993;
Weidenreich 1941). Focus on cortical bone thickness in anatomically modem human
skeletal remains, however, has only recently received significant attention with many
researchers looking at patterns of cortical bone growth in subadult skeletons. Cook
(1979) provides one of the earliest studies of cortical bone growth and thickness in
subadult remains. While examining the skeletal effects that subsistence change had on
individuals living in the Lower Illinois Valley region during Woodland and
Mississippian times, Cook (1979) found a significant loss of cortical thickness at the
midshaft of the femur in Late Woodland children between two and three years of age.
Due to the fact that cortical bone loss was significant for children of weaning age,
Cook suggested that a diet supplemented with carbohydrate vegetable food sources,
23


such as maize, had a negative impact on the health of children interfering with their
ability to achieve proper cortical bone growth.
Hummert (1983) in another study examined the relationship between longitudinal
and cortical bone growth in a sample of 174 children excavated from two medieval
Christian cemeteries at Kulubnarti in Sudanese Nubia (550 1450 A.D.).
Developmental ages, ranging from birth to 16 years, were assigned to children based
on dental formation and eruption sequences using standards developed by Ubelaker
(1978). In addition, cross-sections of bone were cut from the midshaft of the left tibia
from each individual in order to obtain various cortical measurements including total
subperiosteal area, cortical area, medullary area, and percent cortical area.
Hummerts 1983 results showed similar patterns of growth for the variables
diaphyseal length, total subperiosteal area, and cortical area indicating that growth in
length and width was well-maintained in children from both cemeteries. Distance
curves for total subperiosteal area also revealed that Kulubnarti children were normal
in terms of demonstrating rapid postnatal growth, moderate childhood growth, and an
adolescent spurt. Curves for medullary area, however, showed marked increases in
medullary size during adolescence which seemed more extreme than would be
expected if there was a delay in the onset of the appositional phase. Decreases in
percent cortical area provided additional evidence of excessive endosteal resorption
during adolescence, and also during periods of early childhood. Hummert (1983)
concluded that the large amounts of endosteal resorption concomitant with evidence
24


of cribra orbitalia indicated that children in some age groups suffered from
malnutrition. Decreases in percent cortical area throughout the infancy period also
revealed that the youngest individuals at Kulubnarti were malnourished, possibly as a
result of inadequacies in the quantity and quality of food resources and malabsorption
of available nutrients.
A more recent study by Mays (1999a) investigated longitudinal and appositional
bone growth in subadult skeletal remains excavated from a medieval church site in
northern England. Burials from the site were those of ordinary peasants who lived in
the village of Wharram Percy sometime between the 10th and 16th centuries. A total of
327 skeletons under the age of 18 years were recovered, most of them in excellent
condition. Age estimations for the subadult remains were determined using the dental
calcification standards of Schour and Massler (1941).
The goal of Mays 1999a study was to demonstrate the importance of analyzing
longitudinal and appositional bone growth together, instead of in isolation, as a means
for understanding the pattern of growth and health of a population. Longitudinal
growth was studied by plotting femur diaphyseal length against dental age, which
ranged from six months to 17.5 years. The results were compared to growth profiles
produced for four other populations of European origin dating between the 3rd 4th
and 19th centuries A.D., and also to Mareshs (1955) data on 20th century children
from the United States. The skeletal growth profiles revealed that, at all ages, the
Wharram Percy sample has shorter femur lengths than U.S. children and 19th century
25


European children. Growth profiles for Wharram Percy and the other three European
samples, however, appeared very similar. In his study, Mays (1999a) interpreted the
disparity in diaphyseal lengths between the earlier samples, including Wharram
Percy, and the 19th and 20th century data as being consistent with a secular increase in
stature in children over the past two centuries due to improvements in nutrition and a
decrease in the presence of disease.
Appositional growth in the Wharram Percy subadult sample was analyzed at the
midshaft of the femur using radiogrammetry. Total subperiosteal width and medullary
width were measured from anteroposterior radiographs, and cortical thickness,
cortical area, percentage cortical area, and cortical index were calculated. Growth
profiles were created by plotting the various cortical variables against age, which
ranged from perinatal to 17.5 years.
The results of Mays 1999a study were compared to similar data collected by
Virtama and Helela (1969) from a modem Finnish population. Comparisons revealed
that, although both groups show the same general pattern of slow increase in both
cortical thickness and cortical index with age, the Wharram Percy sample
significantly lags behind the modem sample in every age category. In fact, Mays
observed that for cortical thickness, Wharram Percy children lag behind modem
children to a greater degree than they do for the variable diaphyseal length, indicating
that cortical bone growth is more affected by the stresses experienced by children
living in Wharram Percy than diaphyseal growth.
26


Mays (1999a) claimed that by studying both longitudinal and appositional bone
growth in conjunction he was able to provide multiple lines of evidence supporting
the claim that children from Wharram Percy suffered from poorer health than modem
children, as a result of living in a harsh climate and being exposed to periodic food
shortages. Mays (1999a) was also able to conclude that cortical bone apposition is a
more sensitive indicator of stress among the children at Wharram Percy than long
bone length. However, whether appositional bone growth is always more sensitive to
stress than longitudinal growth is currently unclear. Other studies looking at
malnutrition in living children demonstrate that, in some cases, stature is more
negatively affected by dietary stress than cortical bone growth (Briers et al. 1975),
and cortical thickness can even experience faster catch-up growth than height (Barr et
al. 1972). More research is necessary in order to elucidate the reasons for why cortical
bone and long bone length may be differentially affected by same type of stress.
Summary
From the brief survey of archaeological studies presented above, it is clear that
interpretations of reduced linear and cortical bone growth in subadult skeletal
populations vary in complexity. Some interpretations simply involve an
acknowledgement that genetic and environmental factors affect patterns of growth,
while other interpretations include explanations of either malnutrition or infection, or
a combination of these factors. For the most part, these interpretations are drawn from
the types of archaeological and documentary evidence available as well as the
27


presence of non-specific indicators of stress and their understood causes (King and
Ulijaszek 1999). Interpretations of growth profiles also depend on the researcher
having knowledge of the normal changes that occur during linear and cortical bone
growth, which primarily comes from studies of contemporary populations. The
important point is that all of the archaeological studies described above are able to
demonstrate the usefulness of analyzing long bone growth patterns in order to
determine the health status of a population.
The subadult population buried at the Milwaukee County Poor Farm Cemetery
serves as the focus of this study for the following reasons: one, little is known
regarding the health status of those who were laid to rest here; and two, research of
this populations long bones provides one with the opportunity to determine how
various known environmental factors related to human behavior, diet, and disease in
the late 19th and early 20th centuries in Wisconsin may have directly affected the
growth of those perceived as members of the lower class. It is apparent from the
growth studies conducted by Hummert (1983) and Mays (1999a) that long bones can
respond to environmental insults in various ways; therefore, it seems appropriate to
examine the patterns of linear and cortical bone growth in the MCIG subadult
population in order to fully understand the impact that environmental factors may
have had on the populations health.
28


CHAPTER 3
HISTORICAL BACKGROUND OF
MILWAUKEE COUNTY
When it comes to interpreting growth patterns in an archaeological population, it is
important to recognize that the process leading to reduced linear and cortical bone
growth is often complex. Multiple factors such as age, level of immunity, disease
ecology, cultural practices, sanitation, and disease management can all contribute to
impaired growth, and, therefore, should be taken into consideration when trying to
describe the health status of population. What is to follow is an overview of the
development and health of Milwaukee County during the 19th and early 20th centuries,
along with a discussion of the establishment of Milwaukee Countys institutions
including the Poor Farm Cemetery. It is hoped that much of the information obtained
from historical records and published literature can shed light on the factors that may
have had a negative impact on the growth of those who were interred at the Poor
Farm Cemetery.
A History of the Development and Health of Milwaukee
County in the 19l and Early 20th Centuries
Soon after the creation of Milwaukee County in 1834, most people came to the
area from far away places to purchase land for farming (Olson 1987). However, many
people were also drawn to the area because of its location at the confluence of three
rivers (the Milwaukee, Menomonee, and Kinnickinnic) with Lake Michigan, and its
29


potential as a site of a large city (Olson 1987; Figure 3.1). Eastern Americans moved
into the area first as permanent settlers and were soon followed by Europeans looking
for new opportunities and cheap land (Olson 1987). Wheat and other grains were the
primary crops grown in the County by the earliest settlers; however, by the late 1870s
many farmers began turning to fruits, vegetables, and other cash crops that could be
sold in the City of Milwaukee in addition to the raising of farm animals such as pigs,
cows, and chickens (Olson 1987). Evidence of urbanization and industrialization was
also apparent by this time in the County due to the existence of schools, churches,
post offices, saw-mills, cement-mills, flour-mills, and highway and railroad
intersections in towns such as Wauwatosa and Oak Creek (Olson 1987). Within the
City of Milwaukee alone, the population rapidly grew from 1,712 in 1840 to 45,246
in 1860 (Olson 1987). By 1880 the population was 115,587, and reached 285,315 by
the turn of the 20th century (Leavitt 1996).
30


R XXI. E.
lX XXII c.
Figure 3.1 1878 Map of Milwaukee County Including the City of Milwaukee,
Town of Wauwatosa, and the County Institutional Grounds (modified
after Snyder, Van Vechten & Co.).
31


Much of Milwaukees population growth during the second half of the 19th century
can be attributed to the citys attractiveness as a place for immigrants to live and find
work. Many of these immigrants like in the early 19th century came from the eastern
United States, Germany, Eastern Europe, Britain, and Ireland (Olson 1987). In 1861
the population was evenly divided between native and foreign-bom with 70 percent
of those foreign-bom having German origins and 14 percent having Irish origins
(Olson 1987). By 1914, both the City and County of Milwaukee continued to receive
a substantial number of foreign immigrants, but the percentage native-born
outnumbered those foreign-bom due to the rapid growth of the former (Olson 1987).
Many immigrants who traveled across the Atlantic to Wisconsin tended to establish
their own ethnic communities throughout the Milwaukee area beginning in the 1850s.
For example, German neighborhoods were predominant on the west side of
Milwaukee between the Milwaukee and Menomonee rivers, whereas Polish
immigrants established residence on the south side of Milwaukee (Leavitt 1996;
Olson 1987; Figure 3.2). Prior to the turn of the century other ethnic groups including
Italians, Hungarians, Greeks, Bohemians, and eastern European Jews began arriving,
although in smaller numbers, and settling in and near downtown neighborhoods
(Leavitt 1996; Olson 1987). According to Olson (1987), many of these ethnic
communities were intent on maintaining the language, religion, beliefs, and cultural
traditions of their homeland and passing them onto their American-born children.
32


Figure 3.2 Map Showing the Three Sections of Milwaukee, 1846. (Leavitt, Judith
Walzer. The Healthiest City. 1996. Reprinted by permission of the
University of Wisconsin Press).
Milwaukee also experienced significant economic growth during the second half
of the 19th century. Clothing factories, iron works, slaughter houses, and metal trades
were established throughout the city, providing numerous employment opportunities
for people, aside from contributing to ethnic divisions and income disparities between
managers and unskilled workers (Leavitt 1996). Projects involving both private and
public funds included the improvement of Milwaukees inner harbor for better
shipping access from Lake Michigan, building of canals in the Menomonee Valley
for industrial use, and development of the Citys modem water system between 1869
33


and 1874 (Olson 1987). By the 1890s Milwaukee had become a key producer of steel,
beer, and other products that could be shipped by water or rail to cities such as
Chicago and St. Paul (Olson 1987).
Along with increased population and economic growth, however, came the
negative aspects of industrialization and urbanization that plagued Milwaukee
County, particularly its major city. Vessels and trains from outlying cities introduced
new diseases into the Milwaukee area, and conditions of urban life including
crowded, dark, unventilated housing; inadequate or absent water supplies; unflushed
open sewers; overflowing privy vaults; and streets covered with horse manure and
garbage helped to spread these diseases and maintain ones already present (Leavitt
1996). Some of the various types of diseases that people were periodically exposed to
were smallpox and cholera (Leavitt 1996). Other diseases including diphtheria,
typhoid fever, scarlet fever, and tuberculosis were endemic in the area (Leavitt 1996).
In fact, tuberculosis was one of the leading causes of death each year among people
living in Milwaukee, in some cases reaching as high as 12.9 and 14.1 percent of the
death rate between the years 1882 and 1898 (Leavitt 1996:28). Diarrheal diseases,
which included a wide variety of intestinal illnesses besides cholera, also contributed
to the death rate year after year, especially among infants receiving contaminated
milk. Diphtheria, on the other hand, was one of the most feared diseases of childhood
due to the severity of its symptoms and how quick it killed children, usually within a
matter of a few days. In order to illustrate how all of these diseases impacted children,
34


Leavitt (1996:29-30) points out that in 1871, 61 percent of all the deaths in
Milwaukee occurred among children under the age of five, and consistently stayed
close to this rate for many years to follow. By the early 20th century, this figure had
fallen to 35 percent, but was still quite large considering that in 1900 children under
the age of five comprised only 12 percent of Milwaukees population.
Contributing to the high death rate in Milwaukee during the 19th century was the
fact that health officials, in their attempt to control and combat various diseases, often
encountered resistance to new health policies by foreign immigrants and the poor.
Many of the policies introduced, which ranged from staying away from the funerals
of those who died of smallpox, sending sick children to the Citys isolation hospital,
and getting vaccinated, contradicted accustomed behaviors and cultural beliefs
practiced by particular ethnic groups (Leavitt 1996). Health officials also had a
difficult time trying to convince people living in outlying communities within the city
to change and improve their diets and sanitation practices. Much of this difficulty
could be blamed on lack of money and available medical care in these areas, which
made ill health and death from infectious diseases almost impossible to avoid (Leavitt
1996).
General improvements in the health of Milwaukees residents, including decreases
in child mortality and increases in life expectancy, were not apparent until the late
19th and early 20th centuries. These improvements were the result of many factors
such as the expanding of various urban services like city water, sewer pipes, and
35


street paving into urban immigrant wards and rural areas to prevent the rapid spread
of disease (Leavitt 1996). Controls over food production and distribution by public
officials also served to lower the risk of consuming contaminated milk and meat
(Leavitt 1996). Overall, concerns about city economic health by middle-class
business groups and citizen welfare by Socialists played an important role in
promoting the improvements that were made in living conditions for thousands of
people during this time. In fact, by 1910 the city stood among the seven American
cities showing the lowest death rates (Leavitt 1996:41).
History of Milwaukee Countys Institutions
From the beginning, Milwaukee Countys municipal government and citizens have
accepted communal responsibility for the health and welfare of the sick and poor. The
Countys system of poor relief during the early 19th century rested on the idea shared
by many people throughout the nation at this time that the best way to help the poor
was to compel them to work (Avella 1987). Public charity and handouts, it was
believed, should be provided only as a last resort and for those who needed it the
most. Outdoor relief served as the dominant system for aiding the sick and poor for
many years in the County. Through this system, individuals who filled out
applications requesting help from the County were qualified to receive various items
and services such as firewood, food, and, occasionally, lodging (Avella 1987). Many
citizens and government officials by the mid-19th century, however, were dissatisfied
with this form of assistance for the impoverished. Recognizing that population growth
36


was leading to more people demanding public aid, taxpayers and government officials
began complaining that outdoor relief was too costly to operate and, even more
importantly, ineffective at deterring pauperism (Avella 1987).
In order to help lower the costs of poor relief and better serve the growing number
of paupers in the area, the Milwaukee County Board of Supervisors held a meeting in
1852 to discuss the potential benefits of providing indoor relief and the building of
a county poorhouse to place the poor (Avella 1987). Members of the Board believed
that a poorhouse would provide a place of refuge for people in need as a result of age,
sickness, and disability, while at the same time discouraging families and individuals
who could work from depending on public assistance. Convinced that a poorhouse
would bring about positive changes in the community, the County Board decided in
the same year to purchase for the price of $6,000 the 160-acre farm owned by
Supervisor Hendrik Gregg to serve as the site for the new institution (Avella 1987).
The property, located south of Watertown Plank Road in the southwestern portion of
the Town of Wauwatosa, included the requisite large farm house, as well as bams,
livestock, and a field of crops for inmates to work and help defray the costs of their
keep (Avella 1987; Olson 1987; Figure 3.3). Future purchases of adjacent farm land
eventually expanded the amount of land owned by the County to almost 1,200 acres
making Wauwatosa the location of most of Milwaukee Countys health and social
welfare facilities for many years (Avella 1987).
37


Figure 3.3 The County Poor Farm Depicted in a 19th Century Oil Painting. The
institutions shown include the Almshouse (center), the County Hospital
(right), and the County Insane Asylum (left) (photo courtesy of
Milwaukee County Historical Society).
The Milwaukee County Poor Farm and Almshouse opened its doors in November
of 1852 and had a total of 24 people staying inside the newly renovated farm house
(Avella 1987). Soon after, many more people came to the Poor Farm including
foreign immigrants who were down on their luck, the elderly poor, the sick poor, the
insane, orphans, and children of paupers and county prisoners (Avella 1987). From
the moment the Poor Farm opened, however, county administrators had to face many
unforeseen challenges and problems surrounding its use. For example, the cost of
running of the Almshouse was much higher than the advocates of indoor relief had
anticipated (Avella 1987). Conditions inside of the facility also quickly deteriorated,
38


generating many complaints by people who saw the establishment as inhumane and
ineffective at properly caring for people who needed help. Much of this was caused
by housing all of the different types of inmates in one central place, which created
problems of order and cleanliness. Complaints made by physicians, reformers, and
sometimes the inmates themselves regarding different aspects of the Almshouse
unwittingly attracted the attention of the press and politicians who made the
institution a subject of political debate (Avella 1987).
In response to the publics outcries and dissatisfaction with the Almshouse, the
County Board of Supervisors began initiating changes regarding how the institution
should be run. The Board mandated regular inspections of the Almshouse and
established a standing committee of the Board of Supervisors to oversee conditions
(Avella 1987). The Board also elected a Superintendent of the Almshouse to operate
and manage the facility for an extended period of time (Vogel 1987). Investigations
of the Poor Farm occurred regularly throughout the 19th century and contributed to
the many changes that were eventually made in the types of services rendered to the
poor.
One major change that helped improve living conditions at the Almshouse was the
decision to separate the sick and contagiously ill from each other and from the
healthy. Separation of these groups was made possible by sending the contagiously ill
paupers to St. Johns Infirmary located on the north side of Milwaukee and the other
sick poor to a specific wing of the Almshouse designated as an infirmary (Avella
39


1987). Space in the infirmary, however, was limited and the Almshouse soon had
more sick poor arriving than they could properly care for.
Recognizing the need for a larger facility, a separate thirty-one bed hospital was
erected next to the Almshouse in 1860. Like most hospitals during this time, the
newly created County Hospital was largely viewed by the public as a charity ward for
those who could not afford in-home medical care. Especially looked down upon by
citizens of the community was the lying-in or maternity ward, where unwed
mothers came to give birth to their illegitimate children and hide their shame
(Avella 1987:202). Public officials during their periodic visits to the hospital were
quick to point out the uncleanliness of the place and defects in the construction and
maintenance of the building, including flaws in the ventilation and sewage systems
(Avella 1987). Even when additions were made to the hospital in 1868 to help solve
the problem of overcrowding, the facility still continued to lack proper sanitation and
care for patients (Avella 1987).
Improvements in the operation of the hospital and care of patients could finally be
seen after 1876 beginning with the appointment of Dr. Fisk H. Day as Superintendent
of the Hospital. For the first time, management of the hospital was put into the hands
of a trained physician who made it his duty to look after not only hospital patients,
but also the inmates staying in the Almshouse. Day tried to make sure that both the
sick and poor received humane treatment, and in doing so, he led the way for many
other needed reforms in the Countys institutions (Richards 1997). Some of these
40


reforms, however, came sooner than others, and began with changes in the way
medical care was to be provided for inmates.
In 1880 the County Hospital experienced a disastrous fire forcing the Board to
authorize the construction of a new hospital in addition to a separate facility to house
the insane (Figure 3.4). In the wake of new discoveries during the 1870s that germs
and bacteria were the causes of disease, new standards in patient care meant that these
newly-built facilities had to provide their patients with services in clean, orderly, and
sanitized conditions (Avella 1987). Patient care was further upgraded in 1887 when
Dr. M.E. Connell and his wife Dr. Anna Gregory Connell established a class on the
instruction and training of nurses for service (Avella 1987:206). In the following
year, the county grounds had its own school of nursing, and soon after the County
Hospital began assuming full responsibility for the training and certification of
qualified nurses.
41


Figure 3.4 The County Insane Asylum Built in 1880 (photo courtesy of Milwaukee
County Historical Society).
Additional changes in the operation of other county institutions and the care of
paupers throughout the mid- to late 19th century could also be observed in the
Almshouse, especially among the children. Of particular concern to county officials
and the public was the safety of children who were sent to live at the Almshouse.
Reformers and county officials alike tended to agree that conditions in the Almshouse
were not properly suited to improve the lives of children or accommodate their
special needs. As a result, some of the children that arrived at the Almshouse were
apprenticed to local businesses and craftsmen, or fostered out to families who wanted
to take care of them (Avella 1987). For those children that remained, there was at
42


least some attempt to separate them from the adults for part of the day by sending
them to the local school house located on county grounds. Separating children and
adults, however, was not always possible, especially when periodic epidemics swept
through the area causing an inflated number of orphaned minors to arrive at the
Almshouse (Avella 1987).
State inspections of the Almshouse, along with repeated claims that poorhouses in
general prevented children from becoming respectable and honest citizens, eventually
led the state legislature in 1875 to pass a law forbidding the placement of children
over the age of five and under sixteen in county or city poorhouses (Avella 1987). In
response to the new law, the County began sending orphaned individuals to city
orphanages and state and private industrial schools where they could be boarded and
educated. Temporary residents of the Almshouse, however, posed a serious problem
for the County because orphanages and industrial schools only accommodated
children who needed permanent homes. The issue of where to place these children
was finally resolved in 1882 with the building of a temporary childrens home on
county grounds. The childrens home took in not only minors who needed temporary
care, but also children who needed to be placed into adoptive homes or work
environments (Avella 1987).
The ability of the childrens home to properly care for and help indigent youths,
however, soon reached its limits. According to Avella (1987:209), immigration,
acute financial distress, and the effects of another disastrous cholera outbreak in the
43


1890s brought scores of dependent youngsters to the overcrowded and dangerous
situation in the temporary childrens home. In order to deal with the situation the
County Board went ahead with plans to establish a larger and more permanent facility
on the grounds. The Home for Dependent Children opened its doors in 1898 and was
intended to provide the same services as the old childrens home serving as a
temporary shelter for minors until either their parents could take them or they could
be placed into a foster home or work place (Avella 1987). Later additions to the
building eventually allowed the facility to accept infants and provide hospital care.
Sadly, many of the children that came to the Home for Dependent Children never
returned to their parents or found foster homes and were compelled to become
permanent residents. In order to accommodate some of these individuals, the facility
provided school instruction and recreational programs (Avella 1987; Figure 3.5).
44


Figure 3.5 Youth Participating in Outdoor Activities at the Home for Dependent
Children in the 1920s (photo courtesy of Milwaukee County Historical
Society).
Figures 3.6 through 3.8 are adapted from the Annual Report of the Home for
Dependent Children and provide the reasons why children were admitted during the
years of 1900, 1910, and 1920 (Dougherty et al. 2005). All three figures demonstrate
that few of the children committed to the Home for Dependent Children were
classified as true orphans. Instead, many of the children committed to the facility
were brought there as a result of being removed from the custody of their parents.
Many reformers throughout the nation felt during this time that it was the duty of
state and county governments to intervene between parents and their children not only
when parents drank, stole, or seemed otherwise immoral and neglectful, but also
45


when parents were so poor that they had to ask for relief (Katz 1986). Breaking
families apart, it was often argued, provided the opportunity to break the cycle of
pauperism and prevent parents from passing on their lax ways and general distaste for
work to their children (Katz 1986). Based on the information recorded in the Annual
Report of the Home of Dependent Children 1900,1910, and 1920 it appears that
Milwaukee County did not feel differently about using family breakup as a reform
strategy and encouraged the removal of children from seemingly unfit homes and
incompetent parents (Dougherty et al. 2005).
90
Orphan Foundling Abandoned Illegitimate Uncontrollable Parental
deficiencies
Figure 3.6 Reasons for Placement into the Home for Dependent Children, 1900.
46


Orphan Foundling Abandoned Illegitimate Uncontrollable Parental
deficiencies
Figure 3.7 Reasons for Placement into the Home for Dependent Children, 1910.
Orphan Foundling Abandoned Illegitimate Uncontrollable Parental
deficiencies
Figure 3.8 Reasons for Placement into the Home for Dependent Children, 1920.
47


Despite the efforts made by the County to ensure that children who came to the
institutional grounds received proper care and a safe environment, many residents
living at the Home for Dependent Children still suffered from ill health and disease.
Table 3.1 provides the causes of death for 253 former inmates listed in the Annual
Report for Dependent Children 1904-1905, 1908-1913, and 1916-1920 (Dougherty et
al. 2005). For better understanding, causes of death were categorized according to the
WHO International Statistical Classification of Diseases and Related Health
Problems, 10th Revision. Nutritional disorders such as malnutrition and marasmus
caused the highest percentage of deaths for infants less than six months. However,
deaths due to respiratory illnesses, usually reported as pneumonia, occurred more
frequently among children between the ages of one and six. Other diseases such as
those of the digestive system and infectious diseases including tuberculosis,
diphtheria, and congenital syphilis had a negative impact on the health of children of
all different ages being frequently reported as causes of death for children in almost
every age cohort.
48


Table 3.1 Causes of death for subadults living at the Home for Dependent Children.
(Percent and number of deaths per age cohort).
Cause of Death < 1 Month 1-3 Months 3-6 Months 6-9 Months 9-12 Months 1-6 Years 6-12 Years 12-18 Years
Congenital Malformations 0 8.7(6) 0 0 0 0 0 50.0(1)
Perinatal Conditions 0 8.7(6) 8.2(5) 3.7(1) 5.9(1) 0 0 0
Nutritional Disorders 100(2) 39.1(27) 31.1(19) 3.7(1) 17.6(3) 0 0 0
Infectious Diseases 0 13.0(9) 24.6(15) 37.7(10) 17.6(3) 24.6(17) 33.3(2) 50.0(1)
Diseases of the Respiratory System 0 5.8(4) 16.4(10) 29.6(8) 35.3(6) 63.8(44) 16.7(1) 0
Diseases of the Digestive System 0 23.2(16) 18.0(11) 22.2(6) 17.6(3) 5.8(4) 33.3(2) 0
Diseases of the Genitourinary System 0 0 0 0 5.9(1) 2.9(2) 0 0
Miscellaneous 0 1.4(1) 1.6(1) 3.7(1) 0 2.9(2) 16.7(1) 0
Among those living at the Home for Dependent Children, infants suffered the most
from ill health and disease. Figure 3.9 (adapted from Dougherty et al. 2005, Annual
Report of the Home for Dependent Children 1904-1905, 1908-1913, and 1916-1920)
illustrates that for many years more deaths were reported for infants than for any
other age group, reaching as high as 27 deaths in one year. Looking at the same
graph, the second highest number of deaths was reported for individuals between one
49


and six years of age. These trends are not surprising if one takes into consideration
the fact that infants and younger children tend to be more vulnerable to poor health
and disease than older children. It is also worth pointing out that year after year the
annual reports indicate that more minors over the age of one were admitted into the
Home for Dependent Children between the years 1901 and 1920 than infants,
suggesting that high mortality among infants was not the result of there being a
greater of number of infants admitted into the institution.
Year
Figure 3.9 Number of Subadult Deaths at the Home for Dependent Children, 1904-
1920.
50


By the turn of the 20th century all of Milwaukee Countys institutions were located
in Wauwatosa. Among these institutions were the County Hospital, the Milwaukee
County Asylum for the Chronically Insane, the Milwaukee County Hospital for the
Acute Insane (originally named the County Insane Asylum), and the Almshouse,
which was rebuilt in 1890 and had an addition put on in 1907 (Figure 3.10). Tables
3.2 and 3.3 are adapted from the Annual Report of the Milwaukee County Farm,
Almshouse, October 1st, 1905 to September 30th, 1906 and provide the nationalities
and number of inmates living at the Almshouse in 1906. Although the available
annual reports on the Almshouse spanning the period from 1890 to 1910 do not
mention the presence of children, it may be the case that some of the women who
came to live at the Almshouse were allowed to have their children stay with them, but
these children were not accounted for due to the state law passed in 1875 forbidding
the placement of children over the age of five and under sixteen in county or city
poorhouses. Since this law did not exclude children under the age of five from living
in poorhouses, one cannot know for sure how many infants and other young children
were residing at the facility throughout the years the Poor Farm Cemetery was in use.
Examination of the annual reports for various years reveal inconsistencies as far as
what type of information was chosen to be recorded, which makes it hard to believe
that infants and children were not occasionally residing at the Almshouse after 1875.
Furthermore, inmate records from several poorhouses in New York reveal that
children of all ages continued to occupy these institutions after the state legislature
51


passed the Childrens Act in 1875 requiring the removal of children between the ages
of two and sixteen from poorhouses (Higgins 2001).
Figure 3.10 The Remodeled Almshouse in 1924. The Almshouse was referred to
as the Milwaukee County Infirmary after 1917 (photo courtesy of
Milwaukee County Historical Society).
52


Table 3.2 Nationality of Almshouse inmates, 1906.
Nationality N Frequency (%)
Austrian 2 0.6
Bohemian 2 0.6
Canadian 3 0.9
Dane 3 0.9
English 11 3.3
French 1 0.3
German 163 49.1
Hollander 5 1.5
Hungarian 1 0.3
Italian 2 0.6
Irish 28 8.4
Norwegian 3 0.9
Polish 50 15.1
Russian 1 0.3
Scotch 2 0.6
Swedish 3 0.9
Swiss 6 1.8
United States 46 13.9
Total 332 100
Table 3.3 Movement of Almshouse population, 1906.
Male Female Total
Inmates in Almshouse January 1st 312 75 387
Inmates received during January 21 1 22
Inmates discharged during January 14 4 18
53


Throughout the early 1900s additional facilities were built on the institutional
grounds with specialized services to accommodate different types of people and their
special needs. One such facility included a tuberculosis sanatorium containing
separate wings for the tubercular insane. Changes were also made to the names of
certain institutions over time. In 1917, the Almshouse was renamed the Milwaukee
County Infirmary. The Milwaukee County Home for Dependent Children was
changed to the Milwaukee County Home for Children, the County Hospital became
County General Hospital, the Milwaukee County Hospital for the Acute Insane was
renamed the Milwaukee County Hospital for the Mentally Diseased, and the
Milwaukee County Asylum for the Chronically Insane became the Milwaukee
County Asylum for Mental Diseases (Avella 1987). By the last decades of the 20th
century most of Milwaukee Countys institutions were no longer in operation and the
grounds were transformed into a medical complex with modem buildings and new
kinds of health care services.
History and Archaeology of the Milwaukee County
Poor Farm Cemetery
When the County Almshouse first began operation in 1852, there was no mention
in any written report of there being a cemetery associated with the farm (Richards
1997). The earliest recorded evidence of there being a cemetery present on the
grounds is a death certificate dated to 1872 describing an infant buried at the Poor
Farm Cemetery (Richards 1997). A more detailed description of the nature and
54


condition of the cemetery can be found in the Proceedings of the Board of
Supervisors of Milwaukee County for the year 1878. In this report the size of the
original cemetery is stated to be less than two and a half acres, with almost one third
completely filled with graves and located on high ground, and approximately 10 to 15
graves located in low, wet meadow (Richards 1997). Additional comment is made by
the Board of Supervisors about the depth of the graves, seemingly concerned that
only a depth of two feet ten inches was allotted for each coffin (Richards 1997). By
August of 1882 this cemetery was abandoned in favor of a new cemetery located
northwest of the old one. The second cemetery was in use from 1882 through 1925. A
third cemetery was established in 1925 north of the other two cemeteries and was in
use until 1974. Figure 3.11 illustrates the location of the three cemeteries in relation
to one another and the Menomonee River and modem roads.
55


Figure 3.11 Map Showing the Three Cemeteries Located on the Milwaukee County
Institutional Grounds (reprinted with permission of Richards 1997).
56


The Poor Farm Cemetery, discussed in this study, refers to the second cemetery
located on the Milwaukee County Institutional Grounds that was utilized from 1882
through 1925. Excavation of the cemetery took place in 1991 and 1992 as the result
of a construction project taking place on the grounds of the Milwaukee County
Medical Complex to expand hospital facilities, which in its early stages disturbed an
unknown number of human burials. Permission was granted to the Associate Hospital
Administrator by the Director of the State Historical Society of Wisconsin to remove
the burials in the proposed area of construction, and fieldwork was conducted by the
Great Lakes Archaeological Research Center, Inc. (Richards 1997).
Interestingly, no one knows for sure who reported the disturbed burials that finally
led to the excavation of the Poor Farm Cemetery. It has been suggested that many
people including Milwaukee County officials, local residents, nursing students, and
construction crew who took part in various projects in the area such as the building of
the School of Nursing Residence in 1933 (which has since been destroyed and
replaced by a parking structure) and the digging of trenches for city utility lines and
water mains were all aware that the area was used as a pauper cemetery.
Unfortunately, the precise number of burials that were disturbed by these construction
activities prior to their authorized removal is unknown but has been estimated at
2,985 (Richards 1997).
Excavation of the remnant cemetery involved dividing the entire construction site
into Areas A-N. Every area except Area I and Area M was excavated and assigned a
57


separate series of lot numbers with each burial receiving its own number (Richards
1997). A total of 1,649 human burials were recovered by the end of the
archaeological project of which 588 were subadults. Figure 3.12 shows the location of
all the graves that were removed. All of the graves (except for those that were
disturbed) were oriented east-west and arranged in rows and columns. Newspaper
articles dating to the late 19th century and county records indicate that the head of
each grave had once been marked by a wooden post and rectangular brass tag that
contained an embossed number. The total area excavated during the archaeological
project comprised of 122,640 square feet or 2.82 acres (Richards 1997). Still
unexcavated and intact is the western portion of the cemetery which measures 34,851
square feet (Richards 1997). All of the human remains removed from the cemetery
are currently stored at Marquette University and undergoing analysis.
58


Figure 3.12 Location of Excavated Graves at the Milwaukee County Poor Farm
Cemetery (reprinted with permission of Richards 1997).
59


The Population Buried at the Milwaukee County
Poor Farm Cemetery
Buried at the Poor Farm Cemetery were residents of the Countys various
institutions, the community poor, individuals without kin, the unidentified, and
victims of murder and accidents. Much of the written information that is available
about the individuals buried in this cemetery comes from the Register of Burials at
the Milwaukee County Poor Farm that was in use from February of 1882 and until
June of 1974. Unfortunately, in most cases it is impossible to the link the individuals
listed in the ledger to specific burials, primarily because most of the brass burial tags
and the wooden posts that they were mounted on have deteriorated and disappeared
from the premises, and also because the nature of the information recorded in the
Register is inconsistent. For example, in some instances the same grave number was
assigned twice, or the letters placed before numbers were not used in consecutive
order, skipping from A-45 through A-100 to D-l through D-100 and E-l through E-
100 back to A-l through A-100 (Richards 1997). Consequently, few individuals
excavated from the cemetery have been identified. Only four burial tags were
recovered during excavations of the cemetery that could confidently be linked with
specific burials, and recovered artifacts such as initialed rings have helped in the
identification of several others (Richards 1997).
Inconsistencies found in other types of information recorded in the Register of
Burials also limit ones ability to develop a complete demographic profile of the
60


population buried in the cemetery. In 1894, the Milwaukee County Rules and
Regulations for the County Farm and Almshouse was published that included Rule 17
regarding the preparation and recording of all future pauper burials. Rule 17
stipulated, among other things, that the Superintendent should keep a record of all
burials at the County Farm, place a painted and numbered head board for each grave,
make sure that no grave is less than six feet deep, and ensure that each burial record
contained the name of the deceased, the date and cause of death, number of burial
permit, and number of grave (Richards 1997). The extent to which these rules were
followed in subsequent years varies as one examines the information recorded in the
Register of Burials. According to Richards (1997), it is not until 1897 that cause of
death is listed more consistently. Some of the more common causes of the death that
were written down include anemia, blue baby, bronchitis, colonitis, heat exhaustion,
imperfect circulation, inanition, infection, mitral stenosis, pneumonia, tuberculosis,
and stillborn. For particular years, other information such as age at death and place of
death were also recorded in the Register. Figure 3.13 demonstrates that the majority
of individuals 21 years of age or younger buried at the Poor Farm Cemetery between
1899 and 1907 were recorded in the Register of Burials as non-institutional or non-
resident deaths (Dougherty et al. 2005). These individuals living in the County were
most likely buried at the cemetery because their parents or other family members
could not afford the burial fees of a traditional cemetery. Richards (1997) also points
that there are many subadults listed in the Register of Burials as abandoned
61


stillborn, drowned stillborn, or unknown with cause of death recorded as
homicide, which means that these individuals were found dead in the community and
interred at the Poor Farm Cemetery at the expense of the County. The second most
frequently listed place of death for subadults in the Register are county institutions
such as the County Hospital or Home for Dependent Children. The least common
place of death listed includes institutions not located on the Milwaukee County
Institutional Grounds such as any of the childrens homes located in the City of
Milwaukee (Dougherty et al. 2005).
259 (59%)
Milwaukee County Other Institutions City/County (Non-
Institutions institutional)
Figure 3.13 Places of Death Listed for Subadults in the Register of Burials at the
Milwaukee County Poor Farm, 1899-1907.
62


Summary
Historical records and published literature reveal that, for the most part,
Milwaukee County made a valid effort during the late 19th and early 20th centuries to
ensure the well-being of those who came to depend on its care. This is demonstrated
through the continuous improvements that were made to county facilities and in the
types of services rendered to individuals who came to live on county grounds.
Unfortunately, many individuals, particularly infants and children, still apparently
suffered from poor health and disease based on reports kept by the Home for
Dependent Children. Other documentary sources indicate that children buried at the
Poor Farm Cemetery may have fell victim to diseases such as smallpox and cholera,
in addition to other diseases including diphtheria, typhoid fever, scarlet fever, and
tuberculosis, which were endemic in the area.
Although much is known of the different types of diseases and illnesses that
plagued the young living in Milwaukee and at the Home for Dependent Children,
much still remains unknown regarding the health status of the infants and children
buried in the Poor Farm Cemetery, particularly those who died at county facilities
other than the Home for Dependent Children or in the community. This can partly be
blamed on the way information was recorded in the Register of Burials. Many entries
dating between the years that the cemetery was in use (1882-1925) fail to list
information such as age at death and cause of death. One can also blame the fact that
medical records are absent for individuals buried in the cemetery. Consequently, one
63


cannot be certain of what other types of illnesses individuals may have endured
during their lifetime that would have had a negative effect on their growth. The
information that is available regarding the level of health care paupers received, and
the nature of living conditions for people residing on the Poor Farm and in
impoverished neighborhoods in Milwaukee County, suggests that many individuals
buried in the Poor Farm Cemetery suffered frequently from ill health and stress.
Analyzing the populations linear and cortical bone growth patterns will help
determine whether or not this was the case.
64


CHAPTER 4
MATERIALS AND METHODS
The purpose of the present study is to develop a better understanding of the health
status of the subadult population buried at the Milwaukee County Poor Farm
Cemetery. In order to achieve this purpose this study asks: are long bone
measurements indicative of poor health status in the subadults recovered from the
Poor Farm Cemetery. As previous studies involving human skeletal remains
demonstrate, the study of linear and appositional bone growth can yield valuable
insights into the health and nutrition status of a population. Therefore, diaphyseal
length and cortical bone data were collected from a sample of subadult long bones
from the Poor Farm Cemetery in order to assess the skeletal growth of the population.
This data was also used as a basis for examining and evaluating methodological
problems inherent to studies of cortical bone growth, specifically the compatibility of
measurements taken from dry bones and X-rays.
The MCIG Subadult Sample
The skeletal material used in this study originated from the 1991 to 1992
excavation of the Milwaukee County Poor Farm Cemetery, discussed in detail in
Chapter Three, and included only those subadults (n = 126) with at least one long
bone, such as the humerus, radius, ulna, femur, and tibia. The fibula was excluded
from the study due to its rare presence in the collection and difficulty in being
65


correctly orientated and sided. Out of the 588 subadults disinterred from the Poor
Farm Cemetery, only 269 (46%) individuals ranging in age from fetal to 12 years of
age had preserved teeth that could be used to estimate age. Out of this number, 126
(21%) individuals had at least one long bone that could be measured, enabling them
to be included in the study (Table 4.1). The low frequency of preserved subadult long
bones in the collection is in part due to the disturbances caused by the various
construction projects within the cemetery over the years, the nature of the coffins that
individuals were buried in, which were constructed mostly out of inexpensive woods
such as pine and elm, and the elevation at which coffins were buried (Richards 1997;
Trage 1994). Many of the coffins were badly rotted when discovered, and, in a few
cases, only an outline could be recognized by the discoloration of the soil or the
coffin hardware left behind (Hutchins 1998; Richards 1997). Sex determination of the
individuals included in the study was not attempted since precise techniques for
determining sex in juvenile remains are currently unavailable.
Table 4.1 Details about the MCIG subadult sample.
No. of % of subadult % of subadult
individuals sample population
Subadults in collection Subadults in collection with 588 269 100 46
preserved dentition Included in study with 100
126 21
preserved dentition
Included in study with one 22 17 4
or more pathology
66


Types of Pathologies Present
Few subadults disinterred from the Poor Farm Cemetery show signs of having
visible pathologies. Out of the 126 subadults that provided at least one long bone
measurement and had preserved teeth, only 22 (17%) exhibited at least one type of
observable skeletal pathology (Table 4.1). Table 4.2 illustrates the different types of
pathologies, or non-specific indicators of stress, found and their frequencies, all of
which were compiled in accordance with the guidelines recommended by Buikstra
and Ubelaker (1994).
Table 4.2 Types of non-specific pathological conditions observed in the
MCIG subadult sample.
Pathology____________________________________________n_
Enamel hypoplasias/hypocalcifications 16
Harris lines 11
Dental caries/abscesses 9
Porotic hyperostosis 2
Periostitis 1
Combination of two or more different pathologies 9
Looking at Table 4.2, dental defects such as enamel hypoplasias and
hypocalcifications are the most common pathologies observed in the subadult sample
and are typically thought by researchers to reflect systemic stress caused by either
periods of malnutrition or infectious disease during dental development (Buikstra and
Ubelaker 1994). Enamel hypoplasia, more specifically, is a deficiency in enamel
67


thickness resulting from premature degeneration of ameloblasts, and can be observed
macroscopically on the surface of a tooth crown as horizonatal grooves, or
depressions of isolated or aligned pits (Goodman and Armelagos 1985; Ribot and
Roberts 1996). Hypocalcifications or opacities, on the other hand, result from
imperfect enamel mineralization and appear as transverse lines or oval areas on the
external surfaces of tooth crowns (Buikstra and Ubelaker 1994).
Harris lines are the second most common pathology seen in the subadult sample.
Harris lines appear as horizontal lines or bands of increased bone density at the ends
of bones and can more commonly be found on the ribs, femur, fibula, and tibia, the
latter serving as the site for where the lines documented for the MCIG sample appear
(Mays 1995; Ribot and Roberts 1996). They form when growth is arrested at the
cartilage plate and are sometimes resorbed by successive bone remodeling throughout
life (White 2000). Harris lines, however, only provide general statements about the
variety of possible stressors including episodes of childhood illness and nutritional
deficiencies that an individual may have once endured (Mays 1995; White 2000). In
some cases, they are present when no disease is observed making their precise
etiology problematic (Ribot and Roberts 1996).
Dental caries and abscesses are the third most common types of pathologies
observed in the subadult sample. Dental caries are characterized by the progressive
decalcification of enamel or dentine that is caused by many factors including diet
(White 2000). The macroscopic appearance of caries can vary from opaque spots on
68


the crown to large areas of tooth decay. Abscesses, on the other hand, indicate
inflammation of the pulp chamber following excessive attrition or dental caries
(Buikstra and Ubelaker 1994). Typically when found, an abscess will appear as a
cavity or localized hole within the alveolar bone near the tooth root apices. Abscesses
may also result from enamel cracks after trauma or from spontaneous idiopathic
phenomena (Buikstra and Ubelaker 1994).
Only two cases of porotic hyperostosis were observed in the subadult sample
studied. Porotic hyperostosis is a condition characterized by the presence of small
porous lesions of various size and density on the cranial vault (Stuart-Macadam
1998). The lesions are usually bilaterally symmetrical and present on the parietal
bones and in the orbits of the frontal bone; however, they also can appear on the
frontal and occipital bones (Stuart-Macadam 1998; White 2000). Typically when the
lesions appear in the orbits the condition is referred to as cribra orbitalia, which
characterizes the two cases found in the subadult sample. In general, the lesions seem
to be caused by anemia-associated hypertrophy of the diploe between the inner and
outer tables of bone, and are most often observed in immature individuals (White
2000). According to Stuart-Macadam (1998), this has to do with the fact that the
pressures associated with enlarging bone marrow to produce more red blood cells are
more effective at changing the appearance of bone in children. Clinical and
archaeological investigations reveal that porotic hyperostosis can result from many
factors such as poor nutrition, infectious disease, and parasitism which lead to a
69


deficiency of iron in the body and consequently anemia (White 2000). Porosity of the
cranial vault can also result from deficiency-related diseases including rickets and
scurvy, and inherited hemolytic anemias such as sickle-cell or thalassemia,
illustrating that the etiology of porotic hyperostosis is difficult to specify (Buikstra
and Ubelaker 1994; Ribot and Roberts 1996).
Periostitis is the least common indicator of stress found in the subadult sample
studied, with only one case reported, appearing on the individuals long bones.
Periostitis is characterized by inflammation of the periosteum and can be caused by
trauma or the presence of a non-specific systemic infection (Grauer et al. 1998; White
2000). The inflammation leads to the formation of skeletal striations and raised areas
of new woven or immature bone at the exterior surface (Ribot and Roberts 1996).
Infants, in particular, are often vulnerable to periosteal reactions because they have a
periosteum that is very vascular and loosely bound making it susceptible to bacterial
infections (Ribot and Roberts 1996).
Many researchers claim that there appears to be an association between some of
the different types of non-specific stress indicators discussed above and decreases in
linear and cortical bone growth. For example, Goodman and colleagues (1988) found
in their study of Mexican adolescents a correlation between enamel hypoplasia and
height. Mays (1995) discovered in his analysis of femora from juveniles between the
ages of two and 17 that individuals with Harris lines tend to have thinner cortical
bone based on measures of cortical index. Recognizing that such relationships exist,
70


an additional hypothesis is made that individuals demonstrating at least one visible
non-specific indicator of stress in the MCIG subadult sample should exhibit evidence
of reduced linear and cortical bone growth, and to a greater extent than those
individuals in the sample who do not demonstrate non-specific indicators of stress.
Testing this hypothesis would help in understanding the relationship that exists
between bone growth and visible non-specific stress indicators, and how the latter
possibly interfere with linear and cortical bone growth differently.
Age Determination
Subadult ages had already been estimated for individuals with preserved teeth
prior to this study and consequently did not need to be performed by the author. Ages
were estimated using the developmental standards of Gustafson and Koch (1974) and
Moorrees et al. (1963a, 1963b). All available dentitions were also aged according to
the regression formulas created by Liversidge et al. (1993). Final age for each
individual was determined by averaging the ages estimated by the different dental
standards in order to ensure accuracy. Each individual that was aged and had at least
one long bone measurement was assigned to a specific age group which included the
following: fetal, birth to 0.9 months, one to 2.9 months, three to 5.9 months, six to
11.9 months, one to 2.9 years, 3 to 5.9 years, 6 to 8.9 years, and 9 to 11.9 years.
These age groups were selected not only to obtain groups with sufficient sample sizes
to compensate for the small number of individuals at each one-year interval, but also
to facilitate comparisons with other studies utilizing similar age groups.
71


Measurements of Subadult Long Bones
The protocol for obtaining subadult long bone measurements has been previously
described by Buikstra and Ubelaker (1994). Measurements recorded for dry bones
included maximum diaphyseal length of the humerus, radius, ulna, femur, and tibia,
as well as maximum diameter at the midshaft in the anteroposterior and mediolateral
planes for the humerus, femur, and tibia. Dry measurements were taken by the author
using a digital sliding caliper and, in some instances, an osteometric board if the
length of the bone exceeded the length of the caliper. All measurements were made
on the left side and recorded to the nearest 0.01 millimeter, but when this was not
possible, the right side was utilized in order to not sacrifice a skeleton.
In order to obtain measurements of cortical thickness, the right and left humerus,
femur, and tibia from each individual was radiographed by the author in standardized
position in the anteroposterior and mediolateral planes. The use of two planes of
measurement largely eliminates the errors involved in extrapolating from one-
dimensional bone diameters two-dimensional bone areas (Lazenby 1995; Ruff and
Jones 1981; VanGerven et al. 1969). Radiographs were taken with the Kodak 8000c
Digital Panoramic and Cephalometric System using Kodak Digital Imaging Software
6.04 (73 kV/15 mA exposure power; 0.40 second time exposure) provided by
Marquette University School of Dentistry.
From the radiographs, measurements of total subperiosteal width and medullary
cavity width were taken at the midpoint of the diaphysis of each bone in each plane
72


using the software program, tpsDigl (Rohlf 2004). The procedure for obtaining these
measurements included, one, smoothing each image in order to remove any fine-scale
noise present and enhance the lines representing the subperiosteal and endosteal
borders; two, setting the scale factor for each bone, which was based on the
diaphyseal length measurement taken; three, digitizing the location of two sets of
landmarks, the junction between the medullary cavity and cortical bone marking the
endosteal border, and parallel to this, the subperiosteal border; and four, measuring
the linear distances between these landmarks. Figure 4.1 illustrates the locations of
the digitized landmarks used to obtain total subperiosteal width and medullary cavity
width.
73


1
4 5
222426.':!
1P
'-IB
14 I'
16 17181^
Figure 4.1 Radiograph of a Left Femur (lot no. 6180) in the A-P Plane. Numbers
22 and 23, in this case, represent digitized landmarks used to measure
total subperiosteal width. Numbers 24 and 25 represent digitized
landmarks used to measure medullary cavity width. (Picture and
landmarks, including the distances between landmarks, have been
enlarged here.)
Once total subperiosteal width and medullary cavity width were obtained from
each bone, total subperiosteal area, medullary area, cortical area and percent cortical
area were calculated using the formulas detailed by Ruff and Jones (1981). In
74


addition, cortical thickness and cortical index were calculated using the formulas
detailed by Mays (1999a). Table 4.3 provides a breakdown of all of the cortical
measurements taken and their abbreviations. It should be noted that total
subperiosteal width, taken from the radiographs, and maximum diameter at the
midshaft, obtained from the dry bones, represent the same measurement.
Measurements taken from the radiographs were also recorded to the nearest 0.01
millimeter.
Table 4.3 Cortical measurements obtained from long bones in the MCIG sample.
Measurement___________________________Abbreviation
Total subperiosteal width T
Medullary width M
Cortical thickness CT
Total subperiosteal area TA
Medullary area MA
Cortical area CA
Percent cortical area %CA
Cortical index Cl
In order to control for poor preservation, measurements from radiographs and dry
bones were taken only when exposure of underlying trabecular bone was absent or
minimal. Specimens that had explicit morphological conditions which affected the
normal shape of the ends or midshaft of the bone were not included. Measurements
were also taken three times and averaged to reduce intra-observer error (VanGerven
et al. 1985).
75


Statistical Analyses
Linear Bone Growth
The means, standard deviations, and 95 percent confidence intervals for the
variable maximum diaphyseal length were computed for each long bone in each
developmental age group. Ninety-five percent confidence intervals were plotted
against mean values, and scatter plots for each long bone were prepared by plotting
individual diaphyseal lengths against estimated age. In order to identify whether some
MCIG values represented outliers, individual z-scores were calculated from the mean
age group values and standard deviations. The discordancy of these z-scores was
tested using a series of likelihood-ratio tests (Zhang and Yu 2004).
Comparisons of long bone measurements between archaeological groups and
between archaeological and modem populations can provide interesting information
about the timing and types of stresses experienced by individuals in a population,
along with general trends in linear bone growth over time. Recognizing this, mean
values of diaphyseal length for the MCIG sample were plotted against the mean
values established for modem children published by Maresh (1970), and the mean
values determined by Saunders et al. (1993b) for 19th century Belleville children.
Summary statistics for Mareshs data and the St. Thomas sample are shown in
Tables 4.4 and 4.5. For each sample, some of the age groups from the original data
were combined together so that they could be directly compared with the MCIG
sample. In addition, individuals in the younger age groups in the MCIG sample were
76


lumped together to establish mean values for the age groups 0-6 months and 6-12
months, which are used by Saunders et al. (1993) to represent the infancy period.
Mareshs data published in her 1970 article are based on diaphyseal length
measurements taken from serial radiographs of healthy white children ranging from
two months to 12 years of age. Length measurements incorporating the proximal and
distal epiphyses were also taken by the author from radiographs of children aged 10
to 18 years, but these are not included in Table 4.4. All long bone measurements
taken during Mareshs study were recorded in centimeters. As a result, these were
converted to millimeters so that they could be compared to the MCIG data. Maresh
(1970) also determined the means and standard deviations for males and females
separately, so, for comparative purposes, male and female means in each age cohort
were combined. Overall, Maresh (1970) provides one of the most comprehensive
studies of long bone growth to date. The longitudinal study began in 1935 at the Child
Research Council, University of Colorado School of Medicine, and contains
approximately 5,980 sets of long bone measurements.
77


Table 4.4 Statistical summary of diaphyseal length data for modem Denver
children (Maresh 1970).
Developmental age n Humerus Mean s.d. n Radius Mean s.d. n Ulna Mean s.d.
0-6 mos 63 76.25 0.42 63 61.73 0.31 63 69.33 0.33
6-12 mos 73 87.60 0.48 73 69.20 0.35 73 77.40 0.38
1 3 yrs 76 122.31 0.55 76 92.26 0.44 76 103.75 0.47
3 6 yrs 76 165.38 0.77 76 123.60 0.60 76 136.88 0.62
6 9 yrs 77 206.52 0.95 77 153.56 0.74 77 168.48 0.76
9- 12 yrs 78 243.22 1.22 78 180.87 0.95 78 197.83 1.00
Developmental age n Femur Mean s.d. n Tibia Mean s.d.
0-6 mos 63 93.68 0.45 63 75.95 0.50
6-12 mos 73 111.65 0.48 73 89.95 0.53
1 3 yrs 76 162.01 0.67 76 131.23 0.61
3 6 yrs 76 229.18 1.06 76 186.20 0.98
6 9 yrs 77 295.05 1.40 77 239.40 1.31
6 12 yrs 78 353.07 1.82 78 288.43 1.73
Table 4.5 Statistical summary of diaphyseal length data for the St. Thomas
skeletal sample (Saunders et al. 1993b).
Developmental Humerus Radius Ulna
age n Mean s.d. n Mean s.d. n Mean s.d.
0-6 mos 12 71.36 5.22 11 58.36 3.47 10 64.05 3.78
6-12 mos 25 92.62 10.76 24 70.44 7.34 20 79.40 7.34
1 3 yrs 24 126.57 11.81 21 93.68 7.89 18 102.26 8.10
3 6 yrs 6 170.30 12.87 6 125.58 10.84 4 135.02 13.44
6 9 yrs 6 222.07 18.19 6 162.61 15.72 5 175.94 15.45
9-12 yrs 3 261.11 8.44 2 186.94 5.73 1 210.33 5.00
Developmental Femur Tibia
age n Mean s.d. n Mean s.d.
0-6 mos 15 89.26 9.43 9 76.50 5.87
6-12 mos 32 113.47 13.23 29 96.79 10.16
1 3 yrs 27 162.94 15.66 25 132.58 12.93
3 6 yrs 7 236.41 17.65 6 188.82 13.05
6 9 yrs 8 304.29 22.66 7 246.18 23.15
9- 12 yrs 3 364.61 18.14 3 293.94 18.99
78


In addition to establishing distance curves for all five long bones in the MCIG
sample, growth velocities in terms of relative percent of increase were calculated for
the different long bones using the mean values of the age groups 0-5.9 months, 6 -
11.9 months, 1 2.9 years, 3 5.9 years, 6 8.9 years, and 9 11.9 years. Growth
velocities were also calculated for modem children using the means shown in Table
4.4 from Maresh (1970) and plotted against the MCIG percentages. This was done to
determine whether the MCIG sample displays similar growth rates for each long bone
as modem, healthy children.
Cortical Bone Growth
The fact that some of the confidence intervals for mean diaphyseal length have
discrete mean ranges, or at least display lack of complete overlap with other
confidence intervals, meant that they could be of use in determining the ages of some
of the individuals in the collection without dental remains, but with diaphyseal length
measurements. Utilization of these confidence intervals increased the sample size of
subadults to 168, thus improving the distributions used to calculate the cortical bone
growth data. Only individuals lacking dental remains and visible pathologies were
aged using the diaphyseal length data in order to prevent inaccurate age estimations.
The means and standard deviations for total subperiosteal width, medullary cavity
width, total subperiosteal area, medullary area, cortical area, cortical thickness,
cortical index, and percent cortical area were computed for the humerus, femur, and
tibia in each age group. Growth profiles for the different cortical bone dimensions
79


were created by plotting the means against estimated age. Individual values of percent
cortical area were also plotted in order to identify potential outliers. Since percent
cortical area controls for bone size, as well as differences in cortical thickness in the
anteroposterior and mediolateral planes, it provides the most accurate measurement of
the amount of cortical bone present at a specific site. Individual z-scores were
calculated using the mean age group values and standard deviations. The discordancy
of these z-scores was tested using a series of likelihood-ratio tests (Zhang and Yu
2004).
Differences between maximum diameter measurements taken on dry bones and
total subperiosteal width measurements taken on X-rays were evaluated for the age
groups fetal and 1.0 2.9 months using paired t tests, and for the other age groups
containing fewer than 30 cases using Wilcoxon signed rank tests. Comparisons were
made between measurements taken in the anteroposterior plane, and between
measurements taken in the mediolateral plane. This served to determine the accuracy
of diameter measurements and diaphyseal length measurements taken on dry bones,
and determine if the digitizing program, tpsDigl, is useful for obtaining cortical bone
measurements from X-rays.
Comparisons of cortical bone measurements between archaeological groups and
between archaeological and modem populations potentially can provide interesting
information about the timing of stresses experienced by individuals in a population,
as well as general trends in cortical growth over time, just as linear bone
80


measurements can. However, as previous studies interested in cortical bone growth
show, little research has focused on subadult long bones. Most of the research looking
at cortical bone growth in subadults involves measurements of the second metacarpal
(Frisancho et al. 1970a; Gam 1970; Gam et al. 1964; and Himes et al. 1975). Taking
into account that actual dimensions of total subperiosteal area, medullary area,
cortical area, and percent cortical area vary between bones (Hummert 1983), it was
decided that direct comparisons between the humerus, femur, and tibia with data on
the second metacarpal would probably be misleading. Therefore, mean values of
cortical thickness and cortical index for the MCIG femur and tibia based on
anteroposterior radiographs were plotted against the mean values established for the
femur and tibia in modem children published by Virtama and Helela (1969), and the
mean values for subadult femora from a medieval peasant population published by
Mays (1999a). In addition, mean values of percent cortical area for the MCIG tibia
were plotted against the mean values determined by Hummert (1983) for children
from two medieval Christian cemeteries at Kulubnarti in Sudanese Nubia (550 1450
A.D.). Summary statistics for Virtama and Helelas data, Mays data, and the Nubian
sample are presented in Tables 4.6 4.9. Some age groups, if they could be, were
combined from the original data in order to make them more comparable with the
MCIG sample. In addition, individuals from the younger age groups in the MCIG
sample were lumped together so they could be compared to similar age groups found
in the other samples.
81


Virtama and Helelas 1969 data are based on cortical bone measurements taken
from serial anteroposterior radiographs of healthy individuals living in the
southwestern part of Finland. The purpose of their study, which began in 1963, was to
obtain standard values that could be used in clinical settings for radiographic
estimation of the amount of mineral content present in bones. A total of 38,013
radiographed bones were examined. These bones belonged to individuals ranging in
age from one to 90 years old; however, only data from individuals between the ages
of one and 12 are shown in Table 4.6 and 4.7 and were used for comparison with the
MCIG results. All measurements in Virtama and Helelas 1969 study were recorded
in millimeters and reduced by 0.5 mm to control for radiographic enlargement. Since
mean age group values were calculated separately for males and females in their
study, the means for males and females in each cohort were combined in Tables 4.6
and 4.7. Overall, Virtama and Helela (1969) provide one of the most comprehensive
studies of cortical bone growth to date, including in their study measures of cortical
thickness from various bones in the human body.
Table 4.6 Statistical summary of cortical thickness and cortical index for femora
from modem, healthy children (Virtama and Helela 1969).
Developmental aee Cortical Thickness (mm) Cortical Index
n Mean Mean
1 3 yrs 12 5.93 51.00
3 6 yrs 5 8.03 54.33
6 9 yrs 4 9.78 54.50
9-12 yrs 4 12.07 53.50
82